View COLDFIRE2UM_3299573.PDF datasheet online --- IC-ON-LINE

Datasheet File OCR Text:

motorola, 1998 all rights reserved. coldfire is a registered trademark of motorola, inc. all other trademarks are the property of their respective owners. motorola reserves the right to make changes without further notice to any products herein. motorola makes no warranty, represe ntation or guarantee regarding the suitability of its products for any particular purpose, nor does motorola assume any liability arising out of the applicati on or use of any product or circuit, and specifically disclaims any and all liability, including without limitation consequential or incidental damages. "typical" para meters can and do vary in different applications. all operating parameters, including "typicals" must be validated for each customer application by customer's tec hnical experts. motorola does not convey any license under its patent rights nor the rights of others. motorola products are not designed, intended, or authoriz ed for use as components in systems intended for surgical implant into the body, or other applications intended to support or sustain life, or for any other applic ation in which the failure of the motorola product could create a situation where personal injury or death may occur. should buyer purchase or use motorola products for any such unintended or unauthorized application, buyer shall indemnify and hold motorola and its officers, employees, subsidiaries, affiliates, and di stributors harmless against all claims, costs, damages, and expenses, and reasonable attorney fees arising out of, directly or indirectly, any claim of personal injury or death associated with such unintended or unauthorized use, even if such claim alleges that motorola was negligent regarding the design or manufacture of t he part. motorola and are registered trademarks of motorola, inc. motorola, inc. is an equal opportunity/affirmative action employer. coldfire 2/2m integrated microprocessor users manual ? f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual iii documentation feedback fax 512-895-8593documentation comments only (no technical questions please) http: / / www.mot.com/hpesd/docs_survey.htmldocumentation feedback only the technical communications department welcomes your suggestions for improving our documentation and encourages you to complete the documentation feedback form at the world wide web address listed above. your help helps us measure how well we are serving your information requirements. the technical communications department also provides a fax number for you to submit any questions or comments about this document or how to order other documents. please provide the part number and revision number (located in upper right-hand corner of the cover) and the title of the document. when referring to items in the manual, please reference by the page number, paragraph number, figure number, table number, and line number if needed. please do not fax technical questions to this number. when sending a fax, please provide your name, company, fax number, and phone number including area code. for internet access: web only: http: // www.mot.com/aesop for hotline questions: fax (us or canada): 1-800-248-8567 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
iv coldfire2/2m users manual motorola applications and technical information for questions or comments pertaining to technical information, questions, and applications, please contact one of the following sales offices nearest you. sales offices field applications engineering available through all sales offices united states alabama , huntsville (205) 464-6800 arizona , tempe (602) 897-5056 california , agoura hills (818) 706-1929 california , los angeles (310) 417-8848 california , irvine (714) 753-7360 california , rosevllle (916) 922-7152 california , san diego (619) 541-2163 california , sunnyvale (408) 749-0510 colorado , colorado springs (719) 599-7497 colorado , denver (303) 337-3434 connecticut , wallingford (203) 949-4100 florida , maitland (407) 628-2636 florida , pompano beach/ fort lauderdale (305) 486-9776 florida , clearwater (813) 538-7750 georgla , atlanta (404) 729-7100 idaho , boise (208) 323-9413 illinois , chicago/hoffman estates (708) 490-9500 indlana , fort wayne (219) 436-5818 indiana , indianapolis (317) 571-0400 indiana , kokomo (317) 457-6634 iowa , cedar rapids (319) 373-1328 kansas , kansas city/mission (913) 451-8555 maryland , columbia (410) 381-1570 massachusetts , marborough (508) 481-8100 massachusetts , woburn (617) 932-9700 michigan , detroit (313) 347-6800 minnesota , minnetonka (612) 932-1500 missouri , st. louis (314) 275-7380 new jersey , fairfield (201) 808-2400 new york , fairport (716) 425-4000 new york , hauppauge (516) 361-7000 new york , poughkeepsie/fishkill (914) 473-8102 north carolina , raleigh (919) 870-4355 ohio , cleveland (216) 349-3100 ohio , columbus/worthington (614) 431-8492 ohio , dayton (513) 495-6800 oklahoma , tulsa (800) 544-9496 oregon , portland (503) 641-3681 pennsylvania , colmar (215) 997-1020 philadelphia/horsham (215) 957-4100 tennessee , knoxville (615) 584-4841 texas , austin (512) 873-2000 texas , houston (800) 343-2692 texas , plano (214) 516-5100 virginia , richmond (804) 285-2100 washington , bellevue (206) 454-4160 seattle access (206) 622-9960 wisconsin , milwaukee/brookfield (414) 792-0122 canada british columbia , vancouver (604) 293-7605 ontario , toronto (416) 497-8181 ontario , ottawa (613) 226-3491 quebec , montreal (514) 731-6881 international australia , melbourne (61-3)887-0711 australia , sydney (61(2)906-3855 brazil , sao paulo 55(11)815-4200 china , beijing 86 505-2180 finland , helsinki 358-0-35161191 car phone 358(49)211501 france , paris/vanves 33(1)40 955 900 germany , langenhagen/ hanover 49(511)789911 germany , munich 49 89 92103-0 germany , nuremberg 49 911 64-3044 germany , sindelfingen 49 7031 69 910 germany , wiesbaden 49 611 761921 hong kong , kwai fong 852-4808333 tai po 852-6668333 india , bangalore (91-812)627094 israel , tel aviv 972(3)753-8222 italy , milan 39(2)82201 japan , aizu 81(241)272231 japan , atsugi 81(0462)23-0761 japan , kumagaya 81(0485)26-2600 japan , kyushu 81(092)771-4212 japan , mito 81(0292)26-2340 japan , nagoya 81(052)232-1621 japan , osaka 81(06)305-1801 japan, sendai 81(22)268-4333 japan, tachikawa 81(0425)23-6700 japan, tokyo 81(03)3440-3311 japan , yokohama 81(045)472-2751 korea , pusan 82(51)4635-035 korea , seoul 82(2)554-5188 malaysia , penang 60(4)374514 mexico , mexico city 52(5)282-2864 mexico , guadalajara 52(36)21-8977 marketing 52(36)21-9023 customer service 52(36)669-9160 netherlands , best (31)49988 612 11 puerto rico , san juan (809)793-2170 singapore (65)2945438 spain , madrid 34(1)457-8204 or 34(1)457-8254 sweden , solna 46(8)734-8800 switzerland , geneva 41(22)7991111 switzerland , zurich 41(1)730 4074 talwan , taipei 886(2)717-7089 thailand , bangkok (66-2)254-4910 united kingdom , aylesbury 44(296)395-252 full line representatives colorado , grand junction cheryl lee whltely (303) 243-9658 kansas , wichita melinda shores/kelly greiving (316) 838 0190 nevada , reno galena technology group (702) 746 0642 new mexico , albuquerque s&s technologies, lnc. (505) 298-7177 utah , salt lake city utah component sales, inc. (801) 561-5099 washington , spokane doug kenley (509) 924-2322 argentina , buenos aires argonics, s.a. (541) 343-1787 hybrid components resellers elmo semiconductor (818) 768-7400 minco technology labs inc. (512) 834-2022 semi dice inc. (310) 594-4631 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual v preface the coldfire2/2m integrated microprocessor users manual describes the programming, capabilities, and operation of the coldfire2/2m device. refer to the mcf5200 coldfire family programmers reference manual rev. 1.0 for information on the coldfire family of microprocessors. throughout this document, the coldfire2/2m integrated microprocessor is referred to as the coldfire2/2m. contents this user manual is organized as follows: section 1: overview section 2: signal summary section 3: master bus operations section 4: exception processing section 5: integrated memories section 6: multiply-accumulate unit section 7: debug support section 8: test operation section 9: instruction execution timing section 10: electrical characteristics appendix a: register summary appendix b: new mac instructions index f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual vii table of contents paragraph page number title number section 1 overview 1.1 flexcore integrated processors ..................................................................1-2 1.1.1 flexcore advantages .............................................................................1-4 1.1.2 flexcore module types .........................................................................1-4 1.2 development cycle......................................................................................1-5 1.3 system architecture.....................................................................................1-8 1.3.1 internal bus structure.............................................................................1-8 1.3.1.1 master bus.........................................................................................1-8 1.3.1.2 slave bus...........................................................................................1-9 1.3.1.3 external bus ......................................................................................1-9 1.3.1.4 test bus.............................................................................................1-9 1.3.2 system functional blocks ......................................................................1-9 1.3.2.1 alternate master ................................................................................1-9 1.3.2.2 coldfire2/2m .....................................................................................1-9 1.3.2.3 i-cache data array ..........................................................................1-10 1.3.2.4 i-cache tag array ...........................................................................1-10 1.3.2.5 master bus arbiter (marb) .............................................................1-10 1.3.2.6 rom array.......................................................................................1-11 1.3.2.7 slave modules .................................................................................1-11 1.3.2.8 sram array.....................................................................................1-11 1.3.2.9 system bus controller (sbc) ..........................................................1-11 1.4 programming model ..................................................................................1-11 1.4.1 integer unit user programming model .................................................1-11 1.4.1.1 data registers (d0 C d7) ................................................................1-12 1.4.1.2 address registers (a0 C a6) ...........................................................1-12 1.4.1.3 stack pointer (a7,sp)......................................................................1-12 1.4.1.4 program counter (pc).....................................................................1-12 1.4.1.5 condition code register (ccr) ......................................................1-12 1.4.2 mac unit user programming model ....................................................1-13 1.4.2.1 accumulator (acc)..........................................................................1-14 1.4.2.2 mask register (mask) ....................................................................1-14 1.4.2.3 mac status register (macsr).......................................................1-14 1.4.3 supervisor programming model ...........................................................1-14 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number viii coldfire2/2m users manual motorola 1.4.3.1 status register (sr)........................................................................ 1-14 1.4.3.2 cache control register (cacr)...................................................... 1-15 1.4.3.3 access control registers (acr0, acr1)........................................ 1-15 1.4.3.4 vector base register (vbr)............................................................ 1-15 1.4.3.5 rom base address register (rombar0) ..................................... 1-15 1.4.3.6 sram base address register (rambar0).................................... 1-15 1.5 integer data formats................................................................................. 1-16 1.6 organization of data in registers.............................................................. 1-16 1.6.1 organization of integer data formats in registers .............................. 1-16 1.6.2 organization of integer data formats in memory ................................ 1-17 1.7 addressing mode summary ...................................................................... 1-18 1.8 instruction set summary ........................................................................... 1-19 section 2 signal summary 2.1 introduction.................................................................................................. 2-1 2.2 master bus signals...................................................................................... 2-3 2.2.1 68k interrupt acknowledge mode enable (iack_68k).......................... 2-3 2.2.2 master address bus (maddr[31:0]) ..................................................... 2-3 2.2.3 master arbiter control (marbc[1:0])..................................................... 2-3 2.2.4 master freeze (mfrzb) ........................................................................ 2-4 2.2.5 master kill (mkillb) .............................................................................. 2-4 2.2.6 master read data bus (mrdata[31:0]) ............................................... 2-4 2.2.7 master read data input enable (mie) ................................................... 2-4 2.2.8 master read/write (mrwb)................................................................... 2-4 2.2.9 master reset (mrstb) .......................................................................... 2-4 2.2.10 master size (msiz[1:0]) ......................................................................... 2-4 2.2.11 master transfer acknowledge (mtab) .................................................. 2-5 2.2.12 master transfer error acknowledge (mteab)....................................... 2-5 2.2.13 master transfer modifier (mtm[2:0])...................................................... 2-5 2.2.14 master transfer start (mtsb) ................................................................ 2-6 2.2.15 master transfer type (mtt[1:0]) ........................................................... 2-6 2.2.16 master write data bus (mwdata[31:0])............................................... 2-6 2.2.17 master write data output enable (mwdataoe).................................. 2-6 2.3 general control signals .............................................................................. 2-6 2.3.1 clock (clk) ............................................................................................ 2-6 2.3.2 interrupt priority level (iplb[2:0]) .......................................................... 2-6 2.4 integrated memory signals.......................................................................... 2-7 2.4.1 instruction cache signals....................................................................... 2-7 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number motorola coldfire2/2m users manual ix 2.4.1.1 instruction cache address bus (ich_addr[14:2])...........................2-7 2.4.1.2 instruction cache data chip-select (ichd_csb) .............................2-7 2.4.1.3 instruction cache data input bus (ichd_di[31:0]) ...........................2-7 2.4.1.4 instruction cache data output bus (ichd_do[31:0]).......................2-7 2.4.1.5 instruction cache data strobe (ichd_st)........................................2-7 2.4.1.6 instruction cache data read/write (ichd_rwb).............................2-7 2.4.1.7 instruction cache size (ich_sz[2:0])................................................2-8 2.4.1.8 instruction cache tag chip-select (icht_csb)...............................2-8 2.4.1.9 instruction cache tag input bus (icht_di[31:8]) .............................2-8 2.4.1.10 instruction cache tag output bus (icht_do[31:8]) ........................2-8 2.4.1.11 instruction cache tag strobe (icht_st)..........................................2-9 2.4.1.12 instruction cache tag read/write (icht_rwb) ..............................2-9 2.4.2 integrated rom signals .........................................................................2-9 2.4.2.1 rom address bus (rom_addr[14:2]) ............................................2-9 2.4.2.2 rom data output bus (rom_do[31:0])...........................................2-9 2.4.2.3 rom enable (rom_enb[1:0]) ..........................................................2-9 2.4.2.4 rom size (rom_sz[2:0]) .................................................................2-9 2.4.2.5 rom valid (rom_vld)...................................................................2-10 2.4.3 integrated sram signals .....................................................................2-10 2.4.3.1 sram address bus (sram_addr[14:2]) ......................................2-10 2.4.3.2 sram chip-select (sram_csb)....................................................2-10 2.4.3.3 sram data input bus (sram_di[31:0]) .........................................2-11 2.4.3.4 sram data output bus (sram_do[31:0]).....................................2-11 2.4.3.5 sram size (sram_sz[2:0]) ...........................................................2-11 2.4.3.6 sram strobe (sram_st[3:0]) .......................................................2-11 2.4.3.7 sram read/write (sram_rwb[3:0]) ............................................2-11 2.5 debug signals ...........................................................................................2-11 2.5.1 break point (bkptb)............................................................................2-11 2.5.2 debug data (ddata[3:0])....................................................................2-12 2.5.3 development serial clock (dsclk).....................................................2-12 2.5.4 development serial input (dsi)............................................................2-12 2.5.5 development serial output (dso) .......................................................2-12 2.5.6 processor status (pst[3:0]).................................................................2-12 2.6 test signals ...............................................................................................2-12 2.6.1 integrated memory test signals...........................................................2-12 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number x coldfire2/2m users manual motorola 2.6.1.1 test address bus (test_addr[14:2]) .......................................... 2-13 2.6.1.2 test control (test_ctrl) ............................................................. 2-13 2.6.1.3 test idata read (test_idata_rd) ............................................ 2-13 2.6.1.4 test idata write (test_idata_wrt) ......................................... 2-13 2.6.1.5 test instruction cache read hit (test_rhit)............................... 2-13 2.6.1.6 test invalidate inhibit (test_ivld_inh)........................................ 2-13 2.6.1.7 test itag write (test_itag_wrt).............................................. 2-13 2.6.1.8 test kta mode enable (test_kta).............................................. 2-13 2.6.1.9 test mode enable (test_mode) .................................................. 2-13 2.6.1.10 test sram read (test_sram_rd) ............................................ 2-13 2.6.1.11 test sram write (test_sram_wrt).......................................... 2-13 2.6.1.12 test read (test_rd) .................................................................... 2-13 2.6.1.13 test rom read (test_rom_rd) ................................................ 2-13 2.6.1.14 test write inhibit (test_wr_inh)................................................. 2-13 2.6.2 scan signal description ....................................................................... 2-13 2.6.2.1 scan enable (scan_enable)....................................................... 2-14 2.6.2.2 scan exercise array (scan_xarray).......................................... 2-14 2.6.2.3 scan input (scan_in[15:0]) ........................................................... 2-14 2.6.2.4 scan mode (scan_mode) ............................................................ 2-14 2.6.2.5 scan output (scan_out[15:0])..................................................... 2-14 2.6.2.6 scan test ring clock (tr_clk)..................................................... 2-14 2.6.2.7 scan test ring core mode enable (tr_core_en) ..................... 2-14 2.6.2.8 scan test ring data input 0 (tr_di0)............................................ 2-14 2.6.2.9 scan test ring data input 1 (tr_di1)............................................ 2-14 2.6.2.10 scan test ring data output 0 (tr_do0) ....................................... 2-14 2.6.2.11 scan test ring data output 1 (tr_do1) ....................................... 2-14 2.6.2.12 scan test ring enable (tr_en)..................................................... 2-14 2.6.2.13 scan test ring mode (tr_mode) ................................................. 2-14 section 3 master bus operation 3.1 signal description........................................................................................ 3-1 3.1.1 68k interrupt acknowledge mode enable (iack_68k).......................... 3-1 3.1.2 master address bus (maddr[31:0]) ..................................................... 3-1 3.1.3 master arbiter control (marbc[1:0])..................................................... 3-1 3.1.4 master freeze (mfrzb) ........................................................................ 3-2 3.1.5 master kill (mkillb) .............................................................................. 3-2 3.1.6 master read data bus (mrdata[31:0]) ............................................... 3-2 3.1.7 master read data input enable (mie) ................................................... 3-2 3.1.8 master read/write (mrwb)................................................................... 3-2 3.1.9 master reset (mrstb) .......................................................................... 3-2 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number motorola coldfire2/2m users manual xi 3.1.10 master size (msiz[1:0])..........................................................................3-2 3.1.11 master transfer acknowledge (mtab) ..................................................3-2 3.1.12 master transfer error acknowledge (mteab) .......................................3-3 3.1.13 master transfer modifier (mtm[2:0])......................................................3-3 3.1.14 master transfer start (mtsb) ................................................................3-3 3.1.15 master transfer type (mtt[1:0]) ...........................................................3-3 3.1.16 master write data bus (mwdata[31:0]) ...............................................3-4 3.1.17 master write data output enable (mwdataoe) ..................................3-4 3.2 data transfer mechanism ...........................................................................3-4 3.2.1 transfer type control signals................................................................3-4 3.2.1.1 coldfire2/2m access.........................................................................3-4 3.2.1.2 alternate master access....................................................................3-5 3.2.1.3 emulator mode access......................................................................3-5 3.2.1.4 interrupt acknowledge access ..........................................................3-5 3.2.1.5 cpu space access ...........................................................................3-5 3.2.2 data bus requirements .........................................................................3-5 3.3 data transfers .............................................................................................3-6 3.3.1 byte, word, and longword read transfers ...........................................3-6 3.3.2 byte, word, and longword write transfers ...........................................3-9 3.3.3 line read transfer...............................................................................3-11 3.3.4 line write transfers .............................................................................3-14 3.4 misaligned operands.................................................................................3-18 3.5 invalid master bus cycles .........................................................................3-20 3.6 pipeline stalls ............................................................................................3-20 3.7 interrupt acknowledge bus cycles ............................................................3-21 3.7.1 interrupt acknowledge bus cycle (terminated normally) ....................3-22 3.7.2 spurious interrupt acknowledge bus cycle .........................................3-27 3.8 master bus exception control cycles .......................................................3-27 3.8.1 bus errors.............................................................................................3-28 3.8.2 fault-on-fault halt................................................................................3-29 3.9 reset operation.........................................................................................3-29 3.10 master bus arbitration ...............................................................................3-30 3.10.1 master bus arbitration algorithm..........................................................3-30 3.10.1.1 park on coldfire2/2m......................................................................3-30 3.10.1.2 park on alternate master .................................................................3-30 3.10.1.3 park on current master ...................................................................3-31 3.10.2 bus arbitration programming model.....................................................3-31 section 4 exception processing 4.1 exception processing overview ..................................................................4-1 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number xii coldfire2/2m users manual motorola 4.1.1 exception stack frame definition .......................................................... 4-3 4.1.1.1 self-aligning stack ............................................................................ 4-4 4.1.2 exception vectors .................................................................................. 4-5 4.1.3 multiple exceptions ................................................................................ 4-6 4.1.4 fault-on-fault halt.................................................................................. 4-6 4.2 exceptions ................................................................................................... 4-7 4.2.1 reset exception ..................................................................................... 4-7 4.2.2 access error exception .......................................................................... 4-7 4.2.3 address error exception ........................................................................ 4-8 4.2.4 illegal instruction exception.................................................................... 4-9 4.2.5 privilege violation exception .................................................................. 4-9 4.2.6 trace exception ..................................................................................... 4-9 4.2.7 unimplemented opcode exception........................................................ 4-9 4.2.8 debug interrupt .................................................................................... 4-10 4.2.9 format error exceptions ...................................................................... 4-10 4.2.10 trap instruction exceptions................................................................ 4-10 4.2.11 interrupt exception ............................................................................... 4-10 4.2.11.1 level seven interrupts..................................................................... 4-11 4.2.11.2 spurious, autovectored, and uninitialized interrupts....................... 4-12 section 5 integrated memories 5.1 instruction cache......................................................................................... 5-1 5.1.1 instruction cache signal description ..................................................... 5-1 5.1.1.1 instruction cache address bus (ich_addr[14:2]) .......................... 5-2 5.1.1.2 instruction cache data chip-select (ichd_csb)............................. 5-2 5.1.1.3 instruction cache data input bus (ichd_di[31:0]) ........................... 5-2 5.1.1.4 instruction cache data output bus (ichd_do[31:0]) ...................... 5-2 5.1.1.5 instruction cache data strobe (ichd_st)........................................ 5-2 5.1.1.6 instruction cache data read/write (ichd_rwb) ............................ 5-3 5.1.1.7 instruction cache size (ich_sz[2:0]) ............................................... 5-3 5.1.1.8 instruction cache tag chip-select (icht_csb)............................... 5-3 5.1.1.9 instruction cache tag input bus (icht_di[31:8])............................. 5-3 5.1.1.10 instruction cache tag output bus (icht_do[31:8]) ........................ 5-3 5.1.1.11 instruction cache tag strobe (icht_st) ......................................... 5-4 5.1.1.12 instruction cache tag read/write (icht_rwb) .............................. 5-4 5.1.2 instruction cache physical organization................................................ 5-4 5.1.3 interaction with other modules.............................................................. 5-4 5.1.4 cache miss fetch algorithm/line fills ................................................... 5-4 5.1.5 cacheability............................................................................................ 5-5 5.1.6 invalidating cache entries...................................................................... 5-5 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number motorola coldfire2/2m users manual xiii 5.1.7 cache coherency...................................................................................5-6 5.1.8 reset ......................................................................................................5-6 5.1.9 instruction cache programming model ..................................................5-6 5.2 access control registers ............................................................................5-8 5.2.1 acr programming model.......................................................................5-8 5.3 rom module..............................................................................................5-10 5.3.1 rom signal description .......................................................................5-10 5.3.1.1 rom address bus (rom_addr[14:2]) ..........................................5-11 5.3.1.2 rom data output bus (rom_do[31:0]).........................................5-11 5.3.1.3 rom enable (rom_enb[1:0]) ........................................................5-11 5.3.1.4 rom size (rom_sz[2:0]) ...............................................................5-12 5.3.1.5 rom valid (rom_vld)...................................................................5-12 5.3.2 rom programming model....................................................................5-12 5.4 sram module............................................................................................5-14 5.4.1 sram signal description .....................................................................5-14 5.4.1.1 sram address bus (sram_addr[14:2]) ......................................5-15 5.4.1.2 sram chip-select (sram_csb)....................................................5-16 5.4.1.3 sram data input bus (sram_di[31:0]) .........................................5-16 5.4.1.4 sram data output bus (sram_do[31:0]).....................................5-16 5.4.1.5 sram size (sram_sz[2:0]) ...........................................................5-16 5.4.1.6 sram strobe (sram_st[3:0]) .......................................................5-16 5.4.1.7 sram read/write (sram_rwb[3:0]) ............................................5-16 5.4.2 sram programming model..................................................................5-16 section 6 multiply-accumulate unit 6.1 introduction ..................................................................................................6-1 6.2 mac programming model ...........................................................................6-2 6.2.1 accumulator (acc).................................................................................6-2 6.2.2 mac status register (macsr)..............................................................6-2 6.2.3 mask register (mask) ...........................................................................6-3 6.3 shifting operations ......................................................................................6-4 6.4 overflow mode.............................................................................................6-4 6.5 mac instruction set summary ....................................................................6-5 section 7 debug support 7.1 signal description........................................................................................7-1 7.1.1 break point (bkptb)..............................................................................7-1 7.1.2 debug data (ddata[3:0])......................................................................7-2 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number xiv coldfire2/2m users manual motorola 7.1.3 development serial clock (dsclk)....................................................... 7-2 7.1.4 development serial input (dsi).............................................................. 7-2 7.1.5 development serial output (dso) ......................................................... 7-2 7.1.6 processor status (pst[3:0])................................................................... 7-2 7.2 real-time trace.......................................................................................... 7-2 7.2.1 processor status signal encoding ......................................................... 7-3 7.2.1.1 continue execution (pst = $0) ......................................................... 7-3 7.2.1.2 begin execution of an instruction (pst = $1).................................... 7-3 7.2.1.3 entry into user mode (pst = $3) ...................................................... 7-3 7.2.1.4 begin execution of pulse or wddata instructions (pst = $4) ..... 7-3 7.2.1.5 begin execution of taken branch (pst = $5)................................... 7-4 7.2.1.6 begin execution of rte instruction (pst = $7) ................................ 7-5 7.2.1.7 begin data transfer (pst = $8 - $a) ................................................ 7-5 7.2.1.8 exception processing (pst = $c) ..................................................... 7-5 7.2.1.9 emulator-mode exception processing (pst = $d) ........................... 7-5 7.2.1.10 processor stopped (pst = $e) ......................................................... 7-5 7.2.1.11 processor halted (pst = $f) ............................................................ 7-5 7.3 background debug mode (bdm) ................................................................ 7-5 7.3.1 cpu halt ................................................................................................ 7-6 7.3.2 bdm serial interface .............................................................................. 7-7 7.3.2.1 receive packet format ..................................................................... 7-8 7.3.2.2 transmit packet format .................................................................... 7-9 7.3.3 bdm command set ............................................................................... 7-9 7.3.3.1 bdm command set summary .......................................................... 7-9 7.3.3.2 coldfire bdm commands............................................................... 7-10 7.3.3.3 command sequence diagram ........................................................ 7-11 7.3.3.4 command set descriptions............................................................. 7-12 7.3.3.4.1 read a/d register (rareg/rdreg) ....................................... 7-13 7.3.3.4.2 write a/d register (wareg/wdreg)...................................... 7-13 7.3.3.4.3 read memory location (read)................................................. 7-14 7.3.3.4.4 write memory location (write) ............................................... 7-16 7.3.3.4.5 dump memory block (dump).................................................... 7-17 7.3.3.4.6 fill memory block (fill) ............................................................ 7-19 7.3.3.4.7 resume execution (go) ............................................................ 7-21 7.3.3.4.8 no operation (nop)................................................................... 7-21 7.3.3.4.9 read control register (rcreg) ............................................... 7-22 7.3.3.4.10 write control register (wcreg)............................................... 7-23 7.3.3.4.11 read debug module register (rdmreg) ................................ 7-24 7.3.3.4.12 write debug module register (wdmreg)................................ 7-25 7.3.3.4.13 unassigned opcodes................................................................. 7-25 7.4 real-time debug support......................................................................... 7-26 7.4.1 theory of operation ............................................................................. 7-26 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number motorola coldfire2/2m users manual xv 7.4.1.1 emulator mode ................................................................................7-27 7.4.1.2 reuse of debug module hardware .................................................7-28 7.4.2 programming model .............................................................................7-28 7.4.2.1 address breakpoint registers (ablr, abhr) ................................7-29 7.4.2.2 address attribute register (aatr)..................................................7-30 7.4.2.3 program counter breakpoint register (pbr, pbmr) .....................7-32 7.4.2.4 data breakpoint register (dbr, dbmr) .........................................7-33 7.4.2.5 trigger definition register (tdr) ....................................................7-34 7.4.2.6 configuration/status register (csr)...............................................7-36 7.4.3 concurrent bdm and processor operation..........................................7-39 7.4.4 motorola recommended bdm pinout ..................................................7-39 7.4.5 differences between the coldfire2/2m bdm and cpu32 bdm ..........7-40 section 8 test operation 8.1 integrated memory testing..........................................................................8-1 8.1.1 test bus signal description ...................................................................8-1 8.1.1.1 test address bus (test_addr[14:2]).............................................8-1 8.1.1.2 test control (test_ctrl) ...............................................................8-1 8.1.1.3 test idata read (test_idata_rd) ..............................................8-1 8.1.1.4 test idata write (test_idata_wrt) ...........................................8-2 8.1.1.5 test instruction cache read hit (test_rhit).................................8-2 8.1.1.6 test invalidate inhibit (test_ivld_inh) ..........................................8-2 8.1.1.7 test itag write (test_itag_wrt)................................................8-2 8.1.1.8 test kta mode enable (test_kta) ................................................8-2 8.1.1.9 test mode enable (test_mode) ....................................................8-2 8.1.1.10 test sram read (test_sram_rd)...............................................8-2 8.1.1.11 test sram write (test_sram_wrt)............................................8-2 8.1.1.12 test read (test_rd) ......................................................................8-2 8.1.1.13 test rom read (test_rom_rd)...................................................8-2 8.1.1.14 test write inhibit (test_wr_inh) ...................................................8-2 8.1.2 theory of operation................................................................................8-2 8.1.3 test mode...............................................................................................8-3 8.1.4 instruction cache tag ram testing.......................................................8-3 8.1.4.1 instruction cache tag ram write function ......................................8-3 8.1.4.2 instruction cache tag ram read function ......................................8-5 8.1.5 instruction cache data ram testing .....................................................8-6 8.1.5.1 instruction cache data ram write function .....................................8-6 8.1.5.2 instruction cache data ram read function.....................................8-8 8.1.6 instruction cache kta mode testing.....................................................8-9 8.1.7 rom testing ........................................................................................8-11 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number xvi coldfire2/2m users manual motorola 8.1.7.1 rom read function........................................................................ 8-11 8.1.8 sram testing ...................................................................................... 8-13 8.1.8.1 sram write function ...................................................................... 8-13 8.1.8.2 sram read function...................................................................... 8-15 8.2 scan testing.............................................................................................. 8-16 8.2.1 scan signal description ....................................................................... 8-16 8.2.1.1 scan enable (scan_enable)....................................................... 8-16 8.2.1.2 scan exercise array (scan_xarray).......................................... 8-16 8.2.1.3 scan input (scan_in[15:0]) ........................................................... 8-16 8.2.1.4 scan mode (scan_mode) ............................................................ 8-17 8.2.1.5 scan output (scan_out[15:0])..................................................... 8-17 8.2.1.6 scan test ring clock (tr_clk)..................................................... 8-17 8.2.1.7 scan test ring core mode enable (tr_core_en) ..................... 8-17 8.2.1.8 scan test ring data input 0 (tr_di0)............................................ 8-17 8.2.1.9 scan test ring data input 1 (tr_di1)............................................ 8-17 8.2.1.10 scan test ring data output 0 (tr_do0) ....................................... 8-17 8.2.1.11 scan test ring data output 1 (tr_do1) ....................................... 8-17 8.2.1.12 scan test ring enable (tr_en)..................................................... 8-17 8.2.1.13 scan test ring mode (tr_mode) ................................................. 8-17 8.2.2 test ring .............................................................................................. 8-17 8.3 burn-in testing .......................................................................................... 8-18 8.4 data retention testing.............................................................................. 8-18 section 9 instruction execution timing 9.1 timing assumptions .................................................................................... 9-1 9.2 move instruction execution times............................................................. 9-2 9.3 standard one-operand instruction execution times.................................. 9-4 9.4 standard two-operand instruction execution times.................................. 9-5 9.5 miscellaneous instruction execution times................................................. 9-7 9.6 mac instruction execution timing .............................................................. 9-8 9.7 branch instruction execution times ............................................................ 9-8 section 10 electrical chacteristics 10.1 definitions of specifications....................................................................... 10-1 10.1.1 current ................................................................................................. 10-1 10.1.2 voltage ................................................................................................. 10-1 10.1.3 capacitance ......................................................................................... 10-2 10.1.4 ac switching parameters and waveforms .......................................... 10-2 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
table of contents (continued) paragraph page number title number motorola coldfire2/2m users manual xvii 10.2 coldfire2 data sheet ................................................................................10-6 appendix a register summary a.1 register access methods ........................................................................... a-1 a.2 register formats ........................................................................................ a-2 appendix b new mac instructions b.1 enhanced integer multiply instructions....................................................... b-1 b.2 new mac instructions ................................................................................ b-1 b.3 new register instructions......................................................................... b-12 b.4 operation code map ................................................................................ b-22 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual xix list of illustrations figure page number title number 1-1 flexcore integrated processor typical die layout...........................................1-3 1-2 design system overview ..................................................................................1-7 1-3 coldfire2/2m system diagram .........................................................................1-8 1-4 coldfire2/2m block diagram ..........................................................................1-10 1-5 integer unit user programming model............................................................1-12 1-6 condition code register (ccr)......................................................................1-13 1-7 mac unit user programming model...............................................................1-13 1-8 supervisor programming model......................................................................1-14 1-9 status register (sr) .......................................................................................1-15 1-10 organization of integer data formats in data registers ................................1-16 1-11 organization of integer data formats in address registers...........................1-17 1-12 memory operand addressing .........................................................................1-18 2-1 coldfire2/2m detailed block diagram ..............................................................2-1 3-1 byte, word, and longword read transfer flowchart .......................................3-7 3-2 normal transfer (without wait states) ..............................................................3-8 3-3 byte, word, and longword write transfer flowchart .......................................3-9 3-4 normal write transfer (with wait states) .........................................................3-10 3-5 line read transfer flowchart.........................................................................3-12 3-6 line read transfer (without wait states).........................................................3-13 3-7 line write transfer flowchart .........................................................................3-15 3-8 line write transfer (without wait states).........................................................3-16 3-9 line write transfer (with wait states)..............................................................3-17 3-10 example of a misaligned longword transfer ................................................3-18 3-11 example of a misaligned word transfer .........................................................3-18 3-12 misaligned word read transfer .....................................................................3-19 3-13 example master bus wait state .....................................................................3-21 3-14 interrupt acknowledge bus cycle flowchart...................................................3-23 3-15 coldfire mode interrupt acknowledge bus cycle...........................................3-24 3-16 68k mode interrupt acknowledge bus cycle..................................................3-26 3-17 bus exception cycle .......................................................................................3-29 3-18 initial power-on reset ....................................................................................3-30 4-1 exception processing flowchart .......................................................................4-2 4-2 exception stack frame form............................................................................4-3 5-1 example 8 kbyte instruction cache interface diagram ....................................5-2 5-2 cache control register (cacr) .......................................................................5-6 5-3 access control register (acr0, acr1) ...........................................................5-9 5-4 example 8 kbyte rom interface diagram ......................................................5-11 5-5 rom base address register (rombar0).....................................................5-12 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
list of illustrations (continued) figure page number title number xx coldfire2/2m users manual motorola 5-6 example 8 kbyte sram interface diagram.................................................... 5-15 5-7 sram base address register (rambar0) ................................................... 5-17 6-1 mac flow diagram........................................................................................... 6-2 6-2 mac status register (macsr) ........................................................................ 6-3 6-3 mac mask register (mask) ............................................................................ 6-4 7-1 processor/debug module interface .................................................................. 7-1 7-2 example pst diagram ..................................................................................... 7-4 7-3 bdm serial transfer ......................................................................................... 7-8 7-4 bdm signal sampling ....................................................................................... 7-8 7-5 receive bdm packet ........................................................................................ 7-8 7-6 transmit bdm packet ....................................................................................... 7-9 7-7 command sequence diagram........................................................................ 7-12 7-8 debug programming model ............................................................................ 7-29 7-9 address breakpoint low register (ablr)...................................................... 7-29 7-10 address breakpoint high register (abhr) .................................................... 7-30 7-11 address attribute register (aatr) ................................................................. 7-30 7-12 program counter breakpoint register (pbr) ................................................. 7-32 7-13 program counter breakpoint mask register (pbmr) .................................... 7-33 7-14 data breakpoint register (dbr)..................................................................... 7-33 7-15 data breakpoint mask register (dbmr) ........................................................ 7-33 7-16 trigger definition register (tdr) ................................................................... 7-34 7-17 configuration/status register (csr) .............................................................. 7-37 7-18 recommended bdm connector ..................................................................... 7-40 8-1 test instruction cache tag write cycles.......................................................... 8-4 8-2 test instruction cache tag read cycles ......................................................... 8-5 8-3 test instruction cache data write cycles ........................................................ 8-7 8-4 test instruction cache data read cycles ........................................................ 8-8 8-5 kta mode cycles ........................................................................................... 8-10 8-6 test rom read cycles .................................................................................. 8-12 8-7 test sram write cycles ................................................................................ 8-14 8-8 test sram read cycles ................................................................................ 8-15 10-1 t plh and t phl measurements .......................................................................... 10-2 10-2 t r and t f measurements ................................................................................... 10-2 10-3 internal cell three-state measurements and example circuits ..................... 10-3 10-4 t rec recovery time.......................................................................................... 10-4 10-5 t su and t h measurements between data and a control signal ....................... 10-4 10-6 t su and t h measurements between data and clock signals ........................... 10-4 10-7 switching waveforms showing t w(l) and t w(h) measurements ....................... 10-5 a-1 address attribute register (aatr) ...................................................................a-2 a-2 address breakpoint high register (abhr) ......................................................a-2 a-3 address breakpoint low register (ablr)........................................................a-2 a-4 access control register (acr0, acr1)...........................................................a-3 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
list of illustrations (continued) figure page number title number motorola coldfire2/2m users manual xxi a-5 cache control register (cacr) ...................................................................... a-3 a-6 condition code register (ccr)....................................................................... a-3 a-7 configuration/status register (csr) ............................................................... a-4 1-8 data breakpoint mask register (dbmr) ......................................................... a-4 1-9 data breakpoint register (dbr) ...................................................................... a-4 1-10 mac status register (macsr) ....................................................................... a-4 1-11 mac mask register (mask)............................................................................ a-5 1-12 program counter breakpoint mask register (pbmr)...................................... a-5 1-13 program counter breakpoint register (pbr) .................................................. a-5 a-14 sram base address register (rambar0) .................................................... a-5 a-15 rom base address register (rombar0)...................................................... a-6 a-16 status register (sr) ........................................................................................ a-6 a-17 trigger definition register (tdr)..................................................................... a-6 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual xxiii list of tables table page number title number 1-1 movec register map.....................................................................................1-14 1-2 integer data formats ......................................................................................1-16 1-3 effective addressing modes and categories ..................................................1-19 1-4 notational conventions ...................................................................................1-19 1-5 instruction set summary.................................................................................1-21 2-1 signal summary................................................................................................2-2 2-2 master arbiter control encoding .......................................................................2-4 2-3 master bus transfer size encoding..................................................................2-5 2-4 master bus transfer modifier encoding............................................................2-5 2-5 master bus transfer type encoding.................................................................2-6 2-6 interrupt levels and mask values.....................................................................2-7 2-7 cache configuration encoding .........................................................................2-8 2-8 valid tag ram data signals.............................................................................2-8 2-9 valid rom address bits....................................................................................2-9 2-10 rom configuration encoding .........................................................................2-10 2-11 valid sram address bits................................................................................2-10 2-12 sram configuration encoding .......................................................................2-11 2-13 processor status encoding.............................................................................2-12 3-1 master arbiter control encoding .......................................................................3-1 3-2 master bus transfer size encoding..................................................................3-2 3-3 master bus transfer modifier encoding............................................................3-3 3-4 master bus transfer type encoding.................................................................3-4 3-5 mrdata requirements for read transfers.....................................................3-5 3-6 mwdata bus requirements for write transfers.............................................3-6 3-7 allowable line access patterns ......................................................................3-11 3-8 memory alignment cycles ..............................................................................3-19 3-9 mtab and mteab assertion results.............................................................3-27 4-1 stack pointer alignment....................................................................................4-5 4-2 exception vector assignments .........................................................................4-5 4-3 exception priority groups .................................................................................4-6 4-4 interrupt levels and mask values...................................................................4-11 5-1 cache configuration encoding .........................................................................5-3 5-2 valid tag ram data signals.............................................................................5-3 5-3 initial fetch offset and clnf bits.....................................................................5-5 5-4 cache line fill encoding ..................................................................................5-8 5-5 valid rom address bits..................................................................................5-11 5-6 rom configuration encoding .........................................................................5-12 5-7 valid rom base address bits ........................................................................5-13 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
list of tables (continued) figure page number title number xxiv coldfire2/2m users manual motorola 5-8 valid sram address bits ............................................................................... 5-15 5-9 sram configuration encoding ....................................................................... 5-16 5-10 valid sram base address bits ...................................................................... 5-17 6-1 mask addressing mode .................................................................................... 6-4 6-2 accumulator result in saturation mode............................................................ 6-5 6-3 mac instruction set summary.......................................................................... 6-6 7-1 processor status encoding .............................................................................. 7-2 7-2 cpu-generated message encoding ................................................................ 7-9 7-3 bdm command summary .............................................................................. 7-10 7-4 bdm size field encoding ............................................................................... 7-11 7-5 control register map ...................................................................................... 7-22 7-6 definition of drc encoding - read ................................................................. 7-24 7-7 definition of drc encoding - write ................................................................. 7-25 7-8 ddata, csr[31:28] breakpoint response.................................................... 7-26 7-9 shared bdm/breakpoint hardware................................................................. 7-28 7-10 misaligned data operand references............................................................ 7-34 7-11 bdm connector correlation............................................................................ 7-40 9-1 misaligned operand references ...................................................................... 9-2 9-2 move byte and word execution times............................................................. 9-3 9-3 move long execution times ............................................................................ 9-3 9-4 one operand instruction execution times ....................................................... 9-4 9-5 two operand instruction execution times ....................................................... 9-5 9-6 miscellaneous instruction execution times ...................................................... 9-7 9-7 mac instruction execution times..................................................................... 9-8 9-8 general branch instruction execution times.................................................... 9-8 9-9 bra, bcc instruction execution times............................................................ 9-8 a-1 register summary ............................................................................................a-1 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
preliminary motorola coldfire2/2m users manual xxv list of acronyms acronym definition bcd ........................................................ binary coded decimal bdm........................................................ background debug mode cad ........................................................ computer-aided design cpu ........................................................ central processing unit dsp ........................................................ digital signal processing et ........................................................... execution time iack ....................................................... interrupt acknowledge ifp .......................................................... instruction fetch pipeline lru......................................................... least recently used lsb ......................................................... least significant bit lsw ........................................................ least significant word mac........................................................ multiply-accumulate marb ..................................................... master bus arbiter msb ........................................................ most significant bit msw ....................................................... most significant word oep ........................................................ operand execution pipeline risc ....................................................... reduced instruction set computer rom ....................................................... read-only memory sbc ........................................................ system bus controller sram ..................................................... static random access memory ta ........................................................... transfer acknowledge ts ........................................................... transfer start f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 1-1 section 1 overview this is a summary of the use and operation of the flexcore coldfire? microprocessor core (referred to as the coldfire2) and flexcore coldfire microprocessor core with the multiply- accumulate unit (mac), referred to as the coldfire2m. it also contains a detailed set of timing and electrical specifications. all references to coldfire2/2m will apply to both the coldfire2 and the coldfire2m devices. refer to the coldfire programmers reference manual rev 1.0 (mcf5200prm/ad) for detailed information on the operation of the instruction set and addressing modes. the coldfire2/2m is part of the flexcore program, a semicustom, standard-cell based design program. based on the concept of variable-length reduced instruction set computer (risc) technology, coldfire combines the architectural simplicity of conventional 32-bit risc with a memory-saving, variable-length instruction set. in the flexcore program, high- volume manufacturers can create their own integrated microprocessor containing a core processor (such as the coldfire2/2m) and their own proprietary technology. a flexcore integrated processor allows significant reductions in component count, power consumption, board space, and costresulting in higher system reliability and performance. the main features of the coldfire2/2m processor include: ? 32-bit address bus which can directly address up to 4 gbytes ? 32-bit data bus ? variable-length risc ? optimized instruction set for high-level language constructs ? sixteen general-purpose 32-bit data- and address- registers ? multiply accumulate (mac) unit for dsp applications (coldfire2m only) ? supervisor/user modes for system protection ? vector-base register to relocate exception-vector table ? special core interfacing signals for integrated memories ? full debug support the coldfire2/2m has 32-bit address and data busses. the 32-bit address bus allows direct addressing of up to 4 gbytes. a misalignment unit provides support for misaligned data accesses, and an optional bus arbitration unit provides support for additional bus masters. the coldfire2/2m also supports an integrated instruction cache, sram, and rom (maximum of 32 kbyte each.) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-2 coldfire2/2m users manual motorola the coldfire2/2m supports a subset of the 68000 instruction set, and the coldfire2m has the added functionality of a multiply accumulate (mac) unit for dsp applications. the removed instructions include binary coded decimal, bit field, logical rotate, decrement and branch, integer division, and integer multiply with a 64-bit result. in addition, only the twelve addressing modes supported by the 68000 are supported. user-mode code developed for the coldfire2 processor will run unchanged on 68020, 68030, 68040, and 68060 processors. the new mac instructions on the coldfire2m processor will not run on other processors. a complete debug module is integrated into the coldfire2/2m. this unit provides real-time trace, background debug mode, and real-time debug support. this includes a parallel processor status port, a subset of the background debug mode (bdm) functionality found on motorolas 683xx family of parts, and real-time breakpoint capability. this built-in debug support results in a standard debug interface to established tools for all coldfire - based processors. 1.1 flexcore integrated processors the flexcore design methodology allows designers of high-volume digital systems and third-party technology providers to place their proprietary circuitry on-chip with a motorola microprocessor core. by using flexcore, a designer can reduce the total system cost, component count, and power consumption while providing higher performance and greater reliability. custom logic, memory, and peripheral modules can be added to a core processor to produce the most cost-effective solution for a designer's system. core processors provide special power-management features such as 5 v, 3.3 v, and static operation. the 68000 family of processors have a proven architecture with a broad base of application and system software support, including real-time kernels, operating systems, and compilers, in addition to a wide range of tools to support software development. in the future, additional processing architectures will be included in the flexcore program, including powerpc and digital signal processing (dsp). figure 1-1 shows a typical die layout of a flexcore integrated processor. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-3 complete product lines can be created using flexcore by implementing one base design using a variety of core processors. designers already familiar with 680x0 family design can easily migrate to flexcore processors as the core processors use similar interfaces. additionally, many peripheral modules and memory elements are available for integration. motorola has developed a complete design system for the customer that includes both a broad cell-based library and effective computer-aided design (cad) tools. by building on motorola's proven 680x0 microprocessor architecture and superior manufacturing capabilities, flexcore offers designers the best path to higher system integration. flexcore custom processors are ideal for: ? high-volume users of 8-, 16-, and 32-bit integrated solutions requiring higher system performance whose needs are not met by standard 68300 family devices. ? designers of high-volume applications who need to reduce cost, space, and/or power consumption. ? third-party technology providers who want to deliver their proprietary application- specific technology to a worldwide marketplace. to develop a solution that best suits system requirements in the shortest time frame, integrated processor design is performed by the designer using a methodology created, tested, and documented by motorola. the resulting netlist is then laid out by motorola, verified, and fabricated in silicon. this enables flexcore integrated processors to be produced quickly and cost-effectively, with the resulting device integrating many of the discrete functions needed in the final system. to implement the application-specific logic, the designer uses motorola's standard-cell library. this library offers an extensive range of design elements, memory configurations, and an expanding array of peripheral modules. each cell in the library has been designed figure 1-1. flexcore integrated processor typical die layout customer-designed logic special-function or memory block special function or memory block flexcore family processor core f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-4 coldfire2/2m users manual motorola for optimum size, power, and performance. the added flexibility of high-speed, high-density cells allows the designer to achieve the most cost-effective solution while satisfying critical timing requirements. the standard cell library has been thoroughly characterized and maintained to ensure a smooth transition from a simulated design to working silicon. if both motorola and the customer have the desire, a custom part may also become a standard product. standard products are sold on the open market, allowing costs to be spread over additional units, resulting in lower component prices for high-volume users. third-party technology providers can use the same methodology to combine their application-specific systems expertise with a core processor. the resulting device is manufactured by motorola and can be delivered to the marketplace through either the technologists or motorolas marketing and sales channels. refer to the design system users guide for semicustom & flexcore for more information on the flexcore design methodology. 1.1.1 flexcore advantages developers face challenges in reducing product cost. by incorporating user-designed logic and motorola-supplied functions into a single flexcore processor, a system designer can realize significant savings in cost, power consumption, board space, and pin count. the equivalent functionality can easily require 20 separate components. each component might have 16C64 pins, totaling over 350 connections. each connection is a candidate for a bad solder joint or misrouted trace. each component is another part to qualify, purchase, inventory, and maintain. each component requires a share of the printed circuit board. each component draws poweroften to drive large buffers and circuit board traces to get signals to another chip. each component must be individually placed and attached to a printed circuit board. the signals between the core processor unit and a peripheral might not be compatible nor run from the same clock, requiring time delays or other special design considerations. in a flexcore integrated processor, the major functions and glue logic are all properly connected internally, timed with the same clock, and fully tested. only essential signals are brought out to pins. the processor is then assembled and tested in a surface-mount package for the smallest possible footprint. 1.1.2 flexcore module types the three types of flexcore modules are: ? hard module not alterable laid out has a technology file containing timing data has a defined test scheme ? soft module netlist not alterable other than by clock tree insertion f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-5 not laid out has a defined test scheme simulation test fixture is provided ? parameterizable module alterable via insertion of predefined parameters behavioral model definition of parameters defines test scheme customer selects parameter values and motorola synthesizes the design the coldfire2/2m is available as a hard module only. 1.2 development cycle there are several steps that must be followed in order to create a flexcore integrated mi- croprocessor with a coldfire2/2m. figure 1-2 illustrates the standard cell design flow. these steps include: ? schematic capture on workstationuse the engineering workstation to implement the required system functions with a coldfire2/2m, memory blocks, peripherals, and cells from the motorola standard cell library. ? verilog modulesoptionally use verilog to implement complex user modules or system interconnect of standard cells. ? generate test patternsgenerate the stimulus and test patterns for the design to be used during functional simulation. ? module compilationuse motorola software to generate compiled modules (sram, rom, etc.) ? functional simulationensure that the logic of the system is functionally sound by simulating the design. (no timing information is yet associated with the simulations, and all propagation delays are preset to a unit delay.) ? logic synthesisthe behavioral and structural level description of the design is mapped to a more efficient structural description using synopsys. this final description is a netlist of standard cell components. ? electrical rule checkmotorola software verifies the electrical integrity of the design. this includes checking connectivity, fan-out, edge rates, and other electrical rules. ? calculate node delays motorola software calculates the estimated propagation delays of each node in the circuit. the design software estimates delays based on the fanout, input rise/fall times, drive characteristics, and estimated interconnect capacitances of the netlist. ? path delay analysiswith path delay information, the delays between the clocked elements of the circuit can be determined, and the critical paths that limit the clock rate can be identified. checking for setup, hold, and pulse-width violations can also be accomplished. ? perform real-time simulationthe real-time simulation is run to verify full functionality using the estimated propagation delays calculated by the design tools. ? package and pinout definitiondevelop the package and pin definition file. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-6 coldfire2/2m users manual motorola ? extract test vectorsthe simulator records the input/output patterns generated during the real-time simulation. the test vectors that motorola will use to test the prototypes are derived from these patterns. ? post testkit verificationsimulate the design using the extracted test vectors to ensure proper operation. ? perform fault gradingdetermine the fault coverage of the extracted test vectors. ? timing constraintschip-level timing constraints are created to be used during floorplanning, clock tree synthesis, placement, and routing. ? floor planning/clock tree synthesisfloor planning and clock tree synthesis are performed by motorola using the design timing constraints provided by the designer. ? automatic place & routethe circuits physical layout is created by motorola from the netlist using automatic place and route software. ? post-layout delay calculationafter the cells are placed and routed, the interconnect resistances and capacitances are extracted by motorola. extracted elements replace those estimated earlier during the pre-layout calculation of the node delays. ? re-simulatethe circuit is re-simulated with verilog to ensure no problems have arisen due to a change in load conditions. ? in-place optimizationif the new delay information causes the simulation to fail, synopsys is used to further optimize the layout. placement and routing is then repeated. this cycle continues until the final layout, with post-layout delay, satisfies the design goals. ? post testkit simulationsimulate the design using the extracted test vectors to ensure proper operation. ? re-extract test vectorsextract the test vectors again in order to account for timing changes due to more accurate delay analysis after layout and routing. ? power simulationa power simulation is performed to determine if the design meets the necessary design goals. power simulation should also be run early in the design cycle to ensure design goals are met. ? netlist comparisonthe netlist after place and route is compared by motorola to the original netlist to check for connectivity errors. ? pattern, mask, and wafer generationmotorola generates the patterns, masks, and wafers. ? assembly/testparts are assembled by motorola and tested using the test vectors. ? ship tested prototypestested prototypes are shipped from motorola to the customer. ? final test programthe final test is performed by motorola. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-7 figure 1-2. design system overview automatic place & route post-layout delay calculation netlist comparison pattern, mask and wafer generation assembly / test ship tested prototypes motorola customer logic synthesis generate test patterns calculate node delays electrical rule check perform fault grading perform real-time simulation path delay analysis extract test vectors re-simulate final test program schematic capture on workstation functional simulation post testkit verification package and pinout definition extract test vectors post testkit simulation verilog modules power simulation floor planning / clock tree synthesis timing constraints steps to be done in parallel module compilation in-place optimization optional steps to be done in parallel f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-8 coldfire2/2m users manual motorola 1.3 system architecture most flexcore custom processors will be designed according to a standard system architecture. a block diagram of this architecture is shown in figure 1-3 . this architecture contains the standardized busses and modules discussed below. for more information on coldfire system configurations refer to the mcf5204 users manual (mcf5204um/ad) and the mcf5206 users manual (mcf5206um/ad.) 1.3.1 internal bus structure 1.3.1.1 master bus. the master bus is the primary data interface for the coldfire2/2m. it is a basic two-cycle unidirectional bus. devices on the master bus are capable of initiating bus transactions. this bus includes a 32-bit address bus, 32-bit read data bus, 32-bit write data bus, and various control signals. none of the signals can be tri-stated. as a result, an optional master bus arbiter is required if multiple masters are present in the system. refer to section 3 master bus operation for more information. figure 1-3. coldfire2/2m system diagram i-cache data array i-cache tag array sram array rom array system bus controller coldfire2/2m slave module slave module slave module alternate master master bus arbiter (optional) (optional) master bus master bus master bus slave bus external bus test bus f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-9 1.3.1.2 slave bus. the slave bus is a simplified bus that provides the interface between the internal master bus and the on-chip peripheral modules. the system bus controller (sbc) transfers information between the master bus and the slave bus and is the slave bus master. only the sbc can initiate slave bus transactions. the slave bus includes device select lines, interrupt lines, output enables, write enables, interrupt vector enables, and other control signals to interface to slave modules. all signals are unidirectional and can not be tri-stated. 1.3.1.3 external bus. the external bus provides the interface between the internal master bus and external resources. this bus has no predefined requirements. it can be asynchronous or synchronous. the address and data bus widths may be different than the internal master bus. the sbc provides the necessary translation between the master and external busses. 1.3.1.4 test bus. the test bus provides an interface for performing extensive tests of the integrated memories. signals are provided to control reading and writing of the integrated srams, sroms, instruction cache tag ram, and the instruction cache data ram. 1.3.2 system functional blocks 1.3.2.1 alternate master. any device that is required to initiate bus transactions is required to be on the master bus. all master bus devices except the coldfire2/2m that initiate bus transactions are considered to be alternate masters. use of alternate masters requires a master bus arbiter module. examples of alternate masters include dma controllers and coprocessors. 1.3.2.2 coldfire2/2m. the coldfire2/2m is the primary cpu for any coldfire custom processor. it has dedicated busses for instruction cache and integrated memories. the master bus is the data interface bus used for all data movement to and from the cpu and is usually connected to the system bus controller. clock and debug signals are connected directly to i/o pins. see section 2 signal summary for more information on all of the coldfire2/2m interface signals. a block diagram of the cpu is shown in figure 1-4 . f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-10 coldfire2/2m users manual motorola 1.3.2.3 i-cache data array. the optional instruction cache data array is a compiled ram used to hold cache data. it is connected directly to the coldfire2/2m via a dedicated bus. refer to section 5.1 instruction cache for more information on the cache configuration and interface signals. 1.3.2.4 i-cache tag array. the optional instruction cache tag array is a compiled ram used to hold tag information for the cache. it is connected directly to the coldfire2/2m via a dedicated bus. refer to section 5.1 instruction cache for more information on the cache configuration and the interface signals. 1.3.2.5 master bus arbiter (marb). the optional master bus arbiter is required to support multiple masters on the master bus. it performs bus control and multiplexing for the unidirectional signals on the master bus. refer to section 3.10 master bus arbitration for more information. figure 1-4. coldfire2/2m block diagram coldfire core debug module multiply accumulate unit misalignment unit master bus controller instruction cache controller rom controller sram controller sram rom instruction cache interface interface interface master bus debug interface internal bus test bus interface (coldfire2m only) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-11 1.3.2.6 rom array. the optional rom array is a compiled rom used to hold boot code, critical code, and monitor code. it is connected directly to the coldfire2/2m via a dedicated bus. refer to section 5.3 rom module for more information on the rom configuration and the interface signals. 1.3.2.7 slave modules. the slave modules are on-chip peripherals. they communicate with the coldfire2/2m via the slave bus and sbc. slave modules are always bus slaves and cannot initiate bus transactions except via interrupts. examples of slave modules include serial ports, parallel ports, and timers. 1.3.2.8 sram array. the optional sram array is a compiled ram used to hold critical variables and the stack. it is connected directly to the coldfire2/2m processor via a dedicated bus. refer to section 5.4 sram module for more information on the sram configuration and the interface signals. 1.3.2.9 system bus controller (sbc). the system bus controller is responsible for providing overall control of the slave and external busses. the system bus controller is the single slave-bus master and interrupt controller, a possible external bus master, bus arbiter and interrupt controller, and a master-bus slave and interrupt controller. the sbc provides programmable registers to configure the memory map and interrupt control. the sbc provides master bus cycle termination for accesses to slave modules on the slave bus. it also generates interrupts on the master bus when requested by slaves on the slave bus, and it responds to interrupt acknowledge cycles on the master bus. 1.4 programming model the coldfire2/2m programming model consists of three register groups: integer unit user, mac unit user, and supervisor. programs executing in the user mode use only the registers in the integer and mac groups. system software executing in the supervisor mode can access all registers and use the control registers in the supervisor group to perform supervisor functions. the following paragraphs provide a brief description of the registers in the user and supervisor models. refer to appendix a register summary . 1.4.1 integer unit user programming model figure 1-5 illustrates the integer portion of the user programming model. it consists of the following registers: ? 16 general-purpose 32-bit registers (d0 C d7, a0 C a7) ? 32-bit program counter (pc) ? 8-bit condition code register (ccr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-12 coldfire2/2m users manual motorola 1.4.1.1 data registers (d0 C d7) . registers d0Cd7 are used as data registers for bit (1 bit), byte (8 bits), word (16 bits), and long word (32 bits) operations and may also be used as index registers. 1.4.1.2 address registers (a0 C a6) . these registers can be used as software stack pointers, index registers, or base address registers and may be used for word and long word operations. 1.4.1.3 stack pointer (a7,sp) . coldfire2/2m supports a single hardware stack pointer (a7) used during stacking for subroutine calls, returns, and exception handling. the initial value of a7 is loaded from the reset exception vector, address $0. the same register is used for user mode and supervisor mode, and may be used for word and long word operations. a subroutine call saves the pc on the stack, and the return restores the pc from the stack. both the pc and the status register (sr) are saved on the stack during the processing of exceptions and interrupts. the return from exception instruction restores the sr and pc values from the stack. 1.4.1.4 program counter (pc). the pc contains the address of the currently executing instruction. during instruction execution and exception processing, the processor automatically increments the contents of the pc or places a new value in the pc, as appropriate. for some addressing modes, the pc can be used as a pointer for pc-relative operand addressing. 1.4.1.5 condition code register (ccr). the ccr is the least significant byte of the processor status register (sr), as shown in figure 1-6 . bits 4C0 represent indicator flags 31 16 15 8 7 0 data register 0 (d0) data register 1 (d1) data register 2 (d2) data register 3 (d3) data register 4 (d4) data register 5 (d5) data register 6 (d6) data register 7 (d7) address register 0 (a0) address register 1 (a1) address register 2 (a2) address register 3 (a3) address register 4 (a4) address register 5 (a5) address register 6 (a6) stack pointer (sp,a7) program counter (pc) condition code register (ccr) figure 1-5. integer unit user programming model f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-13 based on results generated by processor operations. bit 4, the extend bit (x bit), is also used as an input operand during multiprecision arithmetic computations. field definitions: x[4]extend condition code assigned the value of the carry bit for arithmetic operations; otherwise not affected or set to a specified result. n[3]negative condition code set if the most significant bit of the result is set; otherwise cleared. z[2]zero condition code set if the result equals zero; otherwise cleared. v[1]overflow condition code set if an arithmetic overflow occurs implying that the result cannot be represented in the operand size; otherwise cleared. c[0]carry condition code set if a carryout of the most significant bit of the operand occurs for an addition, or if a borrow occurs in a subtraction; otherwise cleared. 1.4.2 mac unit user programming model figure 1-7 illustrates the mac portion of the user programming model available on the coldfire2m only. it consists of the following registers: ? 32-bit accumulator (acc) ? 16-bit mask register (mask) ? 8-bit mac status register (macsr) bits 7 6 5 4 3 2 1 0 field ---xnzvc reset -------- r/w r r r r/w r/w r/w r/w r/w figure 1-6. condition code register (ccr) 31 16 15 8 7 0 accumulator (acc) mask register (mask) mac status register (macsr) figure 1-7. mac unit user programming model f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-14 coldfire2/2m users manual motorola see section 6.2 mac programming model for more details. 1.4.2.1 accumulator (acc). this is a 32-bit special purpose register used to accumulate the results of mac operations. 1.4.2.2 mask register (mask). this is a 16-bit special purpose register used to hold the address mask for mac load operations. 1.4.2.3 mac status register (macsr). this is an 8-bit special purpose register used to hold the status of mac operations. 1.4.3 supervisor programming model system programers use the supervisor programming model to implement sensitive operating system functions. the following paragraphs briefly describe the registers in the supervisor programming model. all accesses that affect the control features of the coldfire2/2m are in the supervisor programming model, which consist of the registers available to users as well as the registers listed in figure 1-8 . most of the control registers are accessed via the movec instruction using the control register definitions shown in table 1-1 . these are a subset of the registers defined in the coldfire programmers reference manual rev. 1.0 (mcf5200prm/ad). refer to appendix a register summary for a description of the access methods for all of the register in the coldfire2/2m. 1.4.3.1 status register (sr). figure 1-9 illustrates the sr, which stores the processor status and contains the condition codes that reflect the results of a previous operation. the low-order byte of the sr is the condition code register (ccr.) 31 16 15 0 cache control register (cacr) access control register 0 (acr0) access control register 1 (acr1) vector base register (vbr) rom base address register (rombar0) sram base address register (rambar0) status register (sr) figure 1-8. supervisor programming model table 1-1. movec register map rc[11:0] register definition $002 cache control register (cacr) $004 access control register 0 (acr0) $005 access control register 1 (acr1) $801 vector base register (vbr) $c00 rom base address register (rombar0) $c04 ram base address register 0 (rambar0) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-15 field definitions: t[15]trace enable when set, the processor will perform a trace exception after every instruction; otherwise no trace exception is performed. s[13]supervisor / user state denotes the processor privilege mode: supervisor mode (s set) or user mode (s cleared). m[12]master / interrupt state this bit is cleared by an interrupt exception, and can be set by software during execution of the rte or move to sr instructions. i[10:8]interrupt priority mask defines the current interrupt priority. interrupt requests are inhibited for all priority levels less than or equal to the current priority, except the level seven request, which cannot be masked. 1.4.3.2 cache control register (cacr). the cacr controls the cache operation. this includes cache enable, cache freeze, cache invalidate, cache mode, and default write protect. see section 5.1.11 cache control register (cacr) for more information. 1.4.3.3 access control registers (acr0, acr1). the acrs allow definition of access attributes for two definable memory regions. these attributes include burst control, instruction caching, and write protection. these attributes override the defaults in the cacr. see section 5.2.1 acr programming model for more information. 1.4.3.4 vector base register (vbr). the vbr contains the base address of the exception vector table in memory. the displacement of an exception vector is added to the value in this register to access the vector table. only the upper 12 bits of the vbr are used and the lower 20 bits are filled with zeros. this forces the exception vector table to be aligned on a 1 mbyte boundary. this register is reset to zero. 1.4.3.5 rom base address register (rombar0). the rombar0 register configures the internal rom module. this includes the base address, code space masks, and enable. see section 5.3.2 rom programming model for more information. 1.4.3.6 sram base address register (rambar0). the rambar0 register configures the internal sram module. this includes the base address, write protect, code bits 15 14 13 12 11 10 8 7 5 4 3 2 1 0 field t-sm- i - xnzvc reset 00110 7 0 00000 r/w r/w r r/w r/w r r/w r r/w r/w r/w r/w r/w figure 1-9. status register (sr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-16 coldfire2/2m users manual motorola space masks, and enable. see section 5.4.3 ram programming model for more information. 1.5 integer data formats table 1-2 lists the operand data formats that are supported by the integer unit. integer unit operands can reside in registers, memory, or instructions. the operand size for each instruction is either explicitly encoded in the instruction or implicitly defined by the instruction operation. 1.6 organization of data in registers the following paragraphs describe data organization within the data, address, and control registers. 1.6.1 organization of integer data formats in registers figure 1-10 shows the integer format for data registers. each integer data register is 32 bits wide. byte and word operands occupy the lower 8- and 16-bit portions of integer data registers, respectively. long-word operands occupy the entire 32 bits of integer data registers. a data register that is either a source or destination operand only uses or changes the appropriate lower 8 or 16 bits in byte or word operations, respectively. the remaining high-order portion does not change. the least significant bit (lsb) of all integer sizes is zero, the most significant bit (msb) of a longword integer is 31, the msb of a word integer is 15, and the msb of a byte integer is 7. because address registers and stack pointers are 32-bits wide, address registers cannot be used for byte-size operands. when an address register is a source operand, either the low- table 1-2. integer data formats operand data format size bit 1 bit byte integer 8 bits word integer 16 bits long-word integer 32 bits figure 1-10. organization of integer data formats in data registers lsb msb 10 31 30 bit (0 modulo (offset) < 31,offset of 0 = msb) < _ 0 7 31 byte 0 31 16-bit word 0 31 long word 15 low-order word long word lsb lsb msb msb lsb msb not used not used f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-17 order word or the entire longword operand is used, depending on the operation size. when an address register is the destination operand, the entire register becomes affected, despite the operation size. if the source operand is a word size, it is sign-extended to 32 bits and then used in the operation to an address register destination. address registers are primarily for addresses and address computation support. the instruction set (see section 1.8 instruction set summary ) explains how to add, compare, and move the contents of address registers. figure 1-11 illustrates the integer format for address registers. control registers vary in size according to function. some control registers have undefined bits reserved for future definition by motorola. those particular bits read as zeros and must be written as zeros for future compatibility. all operations to the sr and ccr are word-size operations. for all ccr operations, the upper byte is read as all zeros and is ignored when written, despite privilege mode. 1.6.2 organization of integer data formats in memory the coldfire2/2m uses a big-endian addressing scheme. the byte-addressable organization of memory allows lower addresses to correspond to higher order bytes. the address n of a long-word data item corresponds to the address of the high order word. the lower order word is located at address n + 2. the address n of a word data item corresponds to the address of the high order byte. the lower order byte is located at address n + 1. this organization is shown in figure 1-12 . 31 16 15 0 sign-extended 16-bit address operand 31 0 full 32-bit address operand figure 1-11. organization of integer data formats in address registers f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-18 coldfire2/2m users manual motorola 1.7 addressing mode summary the addressing modes are grouped into categories according to the mode of use. data addressing modes refer to data operands. memory addressing modes refer to memory operands. alterable addressing modes refer to alterable (writable) operands. control addressing modes refer to memory operands without an associated size. these categories sometimes combine to form new categories that are more restrictive. two combined classifications are alterable memory (both alterable and memory) and data alterable (both alterable and data). table 1-3 lists a summary of effective addressing modes and their categories. only the twelve addressing modes supported by the 68000 are available on the coldfire2/2m. figure 1-12. memory operand addressing 31 23 15 7 0 byte $00000000 word $00000000 long word $00000000 byte $00000001 byte $00000002 byte $00000003 word $00000002 byte $00000004 word $00000004 long word $00000004 byte $00000005 byte $00000006 byte $00000007 word $00000006 byte $fffffffc word $fffffffc long word $fffffffc byte $fffffffd byte $fffffffe byte $ffffffff word $fffffffe f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-19 1.8 instruction set summary table 1-4 lists the notational conventions used throughout this manual unless otherwise specified. table 1-5 lists the coldfire2/2m instruction set by opcode. this instruction set is a reduced version of the 68000 instruction set. the removed instructions include bcd, bit field, logical rotate, decrement and branch, integer division, and integer multiply with a 64- bit result. nine new mac instructions have been added. table 1-3. effective addressing modes and categories addressing modes syntax mode field reg. field category data memory control alterable register direct data address dn an 000 001 reg. no. reg. no. x x x register indirect address address with postincrement address with predecrement address with displacement (an) (an)+ C(an) (d 16 , an) 010 011 100 101 reg. no. reg. no. reg. no. reg. no. x x x x x x x x x x x x x x address register indirect with index 8-bit displacement (d 8 , an, xn) 110 reg. no. x x x x program counter indirect with displacement (d 16 , pc) 111 010 x x x program counter indirect with index 8-bit displacement (d 8 , pc, xn) 111 011 x x x absolute data addressing short long (xxx).w (xxx).l 111 111 000 001 x x x x x x immediate # 111 100 x x table 1-4. notational conventions opcode wildcards cc logical condition (example: ne for not equal) register operands an any address register n (example: a3 is address register 3) ay,ax source and destination address registers, respectively dn any data register n (example: d5 is data register 5) dy,dx source and destination data registers, respectively rn any address or data register ry,rx any source and destination registers, respectively rw any second source register rc any control register (example vbr is the vector base register) register/port names acc mac accumulator ddata debug data port ccr condition code register (lower byte of status register) macsr mac status register mask mask register pc program counter pst processor status port sr status register f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-20 coldfire2/2m users manual motorola miscellaneous operands # immediate data following the instruction word(s). effective address assemble program label list of registers (example: d3Cd0) shift operation: shift left (<<), shift right (>>) operand data size: byte (b), word (w), longword (l) operations + arithmetic addition or postincrement indicator C arithmetic subtraction or predecrement indicator arithmetic multiplication ~ invert; operand is logically complemented l logical and v logical or logical exclusive or << shift left (example: d0 << 3 is shift d0 left 3 bits) >> shift right (example: d0 >> 3 is shift d0 right 3 bits) ? source operand is moved to destination operand ? two operands are exchanged sign-extended all bits of the upper portion are made equal to the high-order bit of the lower portion if then else test the condition. if true, the operations after then are performed. if the condition is false and the optional else claus e is present, the operations after else are performed. if the condition is false and else is omitted, the instruction performs no operation. refer to the bcc instruction description as an example. & and |or subfields and qualifiers {} optional operation () identifies an indirect address d n displacement value, n-bits wide (example: d 16 is a 16-bit displacement) address calculated effective address (pointer) bit bit selection (example: bit 3 of d0) lsb least significant bit (example: msb of d0) lsw least significant word msb most significant bit msw most significant word condition code register bit names c carry bit in ccr n negative bit in ccr v overflow bit in ccr x extend bit in ccr z zero bit in ccr table 1-4. notational conventions (continued) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-21 table 1-5. instruction set summary instruction operand syntax operand size operation add dy,,x y,dx 32 32 source + destination ? destination adda y,ax 32 source + destination ? destination addi #,dx 32 immediate data + destination ? destination addq #,x 32 immediate data + destination ? destination addx dy,dx 32 source + destination + x ? destination and dy,x y,dx 32 32 source l destination ? destination andi #,dx 32 immediate data l destination ? destination asl dx,dy #,dn 32 32 x/c (dy << dx) 0 x/c (dy << #) 0 asr dx,dy ,dx 32 32 msb ? (dy >> dx) ? x/c msb ? (dy >> #) ? x/c bcc 8,16 if condition true, then pc + d n ? pc bchg dy,x #,x 8,32 8,32 ~( of destination) ? z, bit of destination bclr dy,x #,x 8,32 8,32 ~( of destination) ? z, 0 ? bit of destination bra 8,16 pc + d n ? pc bset dy,x #,x 8,32 8,32 ~( of destination) ? z, 1 ? bit of destination bsr 8,16 sp C 4 ? sp; pc ? (sp); pc + d n ? pc btst dy,x #,x 8,32 8,32 ~( of destination) ? z clr x 8,16,32 0 ? destination cmpi #,dx 8,16,32 destination C immediate data cmp y,dx 32 destination - source cmpa y,ax 32 destination - source cpush (an) 32 push and invalidate cache line eor dy,x 32 source destination ? destination eori #,dx 32 immediate data destination ? destination ext dx dx 8 ? 16 16 ? 32 sign-extended destination ? destination extb dx 8 ? 32 sign-extended destination ? destination halt none none enter halted state jmp y none y ? pc jsr y none sp C 4 ? sp; next pc ? (sp); y ? pc lea y,ax 32 address of ? an link ax,# 32 sp C 4 ? sp; ax ? (sp); sp ? ax; sp + d16 ? sp lsl dx,dy #,dx 32 32 x/c (dy << dx) 0 x/c (dx << #) 0 lsr dx,dy #,dx 32 32 0 ? (dy >> dx) ? x/c 0 ? (dx >> #) ? x/c mac ? rw,rx , 16 16 + 32 ? 32 32 32 + 32 ? 32 acc + (rw rx){ << 1 | >> 1} ? acc macl ? rw,rx,,,ry 16 16 + 32 ? 32 32 32 + 32 ? 32 acc + (rw rx){ << 1 | >> 1} ? acc; ({& mask}) ? ry move y,x 8,16,32 source ? destination move from acc ? acc,rx 32 acc ? rn move from ccr dx 16 ccr ? destination note: ? available on the coldfire2m only. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview 1-22 coldfire2/2m users manual motorola move from macsr ? macsr,rx macsr,ccr 32 8 macsr ? rx macsr ? ccr move from mask ? mask,rx 32 mask ? rx move from sr dx 16 sr ? destination move to acc ? rx,acc #,acc 32 32 rx ? acc # ? acc move to ccr dy,ccr, #,ccr 16 source ? ccr # ? ccr move to macsr ? rn,macsr #,macsr 32 rn ? macsr # ? macsr move to mask ? rn,mask #,mask 32 32 rn ? mask # ? mask move to sr dy,sr #,sr 16 source ? sr # ? sr movea y,ax 16,32 ? 32 source ? destination movec rn,rc 32 rn ? rc movem list,x y,list 32 32 listed registers ? destination source ? listed registers moveq #,dx 8 ? 32 sign-extended immediate data ? destination msac ? rw,rx, 32 - 16 16 ? 32 32 - 32 32 ? 32 acc - (rw rx){ << 1 | >> 1} sf ? acc msacl ? rw,rx,,,ry 32 - 16 16 ? 32 32 - 32 32 ? 32 acc - (rw rx){ << 1 | >> 1} sf ? acc; ({& mask}) ? ry muls y,dx y, di 16 x 16 ? 32 32 x 32 ? 32 source destination ? destination (signed or unsigned) mulu y,dx y, di 16 x 16 ? 32 32 x 32 ? 32 source x destination ? destination (signed or unsigned) neg x 32 0 C destination ? destination negx x 32 0 C destination C x ? destination nop none none pc + 2 ? pc; pipeline synchronized not x 32 ~ destination ? destination or dy,x y,dx 32 source v destination ? destination ori #,dx 32 immediate data v destination ? destination pea y 32 sp C 4 ? sp; y ? (sp) pulse none none generate unique pst value rte none none (sp+2) ? sr; (sp+4) ? pc; sp+ 8 ? pc rts none none (sp) ? pc; sp + 4 ? sp scc dx 8 if condition true, then 1's ? destination; else 0's ? destination stop # 16 immediate data ? sr; enter stopped state sub dy,x y,dx 32 32 destination - source ? destination suba ,ax 32 destination - source ? destination subi #,dx 32 destination C immediate data ? destination subq #,x 32 destination C immediate data ? destination subx dy,dx 32 destination C source C x ? destination swap dx 16 msw of dx ? lsw of dx trap none none sp C 4 ? sp; pc ? (sp); sp C 2 ? sp; sr ? (sp); sp C 2 ? sp; format ? (sp); vector address ? pc table 1-5. instruction set summary (continued) instruction operand syntax operand size operation note: ? available on the coldfire2m only. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
overview motorola coldfire2/2m users manual 1-23 trapf none # none 16 32 pc + 2 ? pc; pc + 4 ? pc; pc + 6 ? pc tst y 8,16,32 set integer condition codes unlk ax 32 ax ? sp; (sp) ? ax; sp + 4 ? sp wddata y 8,16,32 (y) ? ddata port wdebug y 64 y ? debug; y + 4 ? debug table 1-5. instruction set summary (continued) instruction operand syntax operand size operation note: ? available on the coldfire2m only. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 2-1 section 2 signal summary 2.1 introduction this section describes the coldfire2/2m input and output signals. figure 2-1 shows the coldfire2/2m along with the signal interface. the signals are listed alphabetically in table 2-1 . all coldfire2/2m signals are unidirectional and synchronous. figure 2-1. coldfire2/2m detailed block diagram maddr[31:0] mrdata[31:0] mwdata[31:0] msiz[1:0] mtt[1:0] mtm[2:0] mie mrwb mtsb mtab mteab mrstb ich_sz rom_sz sram_sz ich_addr[14: 2 ichd_csb ichd_di[31:0] ichd_do[31:0 ] ichd_st ichd_rwb icht_csb icht_di[31:8] icht_do[31:8] icht_st icht_rwb bkptb ddata[3:0] dsclk dsi dso pst[3:0] rom_addr rom_do rom_enb rom_vld sram_addr sram_csb sram_di sram_do sram_st sram_rwb clk iplb[2:0] test_addr[14:2] test_idata_rd test_ivld_inh test_kta test_mode test_sram_rd test_rd test_ctrl bus controller misc i-cache unit sram unit rom unit debug cpu core mac unit unit scan_xarray test_itag_wrt test_idata_wr t test_sram_wr t test_rom_rd test_wr_inh mfrzb mkillb marbc[1:0] mwdataoe test_rhit s can_enable scan_mode mem config iack_68k (coldfire2m) si[15:0] so[15:0] tr_mode tr_se tr_sdi[1:0] core_test trclk test port tr_sd0[1:0] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary 2-2 coldfire2/2m users manual motorola table 2-1. signal summary signal mnemonic input/output active state 68k interrupt acknowledge mode iack_68k input high break point bkptb input low clock clk input high debug data ddata [3:0] output high development serial clock dsclk input high development serial data input dsi input high development serial data output dso output high instruction cache address bus ich_addr[ 14:2] output high instruction cache data chip-select ichd_csb output low instruction cache data input bus ichd_d i [31:0] output high instruction cache data output bus ichd_do [31:0] input high instruction cache data strobe ichd_st output high instruction cache data write enable ichd_rwb output low instruction cache size ich_sz [2:0] input high instruction cache tag chip-select icht_csb output low instruction cache tag input bus icht_di[ 31:8] output high instruction cache tag output bus icht_do [31:8] input high instruction cache tag strobe icht_st output high instruction cache tag write enable icht_rwb output low interrupt priority level iplb [2:0] input low master address bus maddr [31:0] output high master arbiter control marbc [1:0] output high master freeze mfrzb output low master kill mkillb output low master read data bus mrdata [31:0] input high master read data input enable mie input high master read/write mrwb output low master reset mrstb input low master size msiz [1:0] output high master transfer acknowledge mtab input low master transfer error acknowledge mteab input low master transfer modifier mtm [2:0] output high master transfer start mtsb output low master transfer type mtt [1:0] output high master write data bus mwdata [31:0] output high master write data output enable mwdataoe output high processor status pst [3:0] output high rom address bus rom_addr [14:2] output high rom data output bus rom_do [31:0] input high rom enable rom_enb [1:0] output low rom size rom_sz [2:0] input high rom valid rom_vld input high scan enable scan_enable input high scan exercise array scan_xarray input high scan input si [15:0] input high scan mode scan_mode input high scan output so [15:0] input high scan test ring clock tr_clk input high f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary motorola coldfire2/2m users manual 2-3 2.2 master bus signals 2.2.1 68k interrupt acknowledge mode enable (iack_68k) this active-high input signal enables the 68k interrupt acknowledge mode. in this mode, the master address bus, maddr[31:0], mtt[1:0], and mtm[2:0] signals mimic the 68k address bus and function codes during interrupt acknowledge and cpu space bus cycles. this is a static input. refer to section 3.7.1 interrupt acknowledge bus cycle for more information. 2.2.2 master address bus (maddr[31:0]) during a normal bus cycle, this 32-bit output bus provides the address of the first item of a bus transfer. it is capable of addressing 4 gbytes of address space. 2.2.3 master arbiter control (marbc[1:0]) these output signals can be used to specify the mode of operation for an optional arbitration module. they reflect the bit positions [17:16] in the cacr. if an optional arbitration module is not used, these signals can be used as general purpose output signals. the cacr is accessible in supervisor mode as control register $002 using the movec instruction. scan test ring core mode enable core_test input high scan test ring data input tr_sdi[1:0] input high scan test ring data output tr_sdo[1:0] output high scan test ring enable tr_se input high scan test ring mode tr_mode input high sram address bus sram_addr [14:2] output high sram chip-select sram_csb output low sram data input bus sram_di [31:0] output high sram data output bus sram_do [31:0] input high sram size sram_sz [2::0] input high sram strobe sram_st [3::0] output high sram read/write sram_rwb [3:0] output low test address bus test_addr [14:2] input high test control test_ctrl input high test invalidate inhibit test_ivld_inh input high test itag write test_itag_wrt input high test instruction cache read hit test_rhit output high test idata read test_idata_rd input high test idata write test_idata_wrt input high test kta mode enable test_kta input high test mode enable test_mode input high test sram read test_sram_rd input high test sram write test_sram_wrt input high test read test_rd input high test rom read test_rom_rd input high test write inhibit test_wr_inh input high table 2-1. signal summary (continued) signal mnemonic input/output active state f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary 2-4 coldfire2/2m users manual motorola 2.2.4 master freeze (mfrzb) this active-low output signal indicates that the core has been halted. mfrzb is not part of the m-bus protocol. it is simply a signal that can be used to alert timers or other peripheral modules that the core has been halted. 2.2.5 master kill (mkillb) this active-low output signal qualifies mtsb (i.e. it can assert in other cycles but is only significant in a cycle where mstb is asserted). when mkillb is asserted simultaneously with mtsb assertion, this indicates a hit in a k-bus memory and that the external cycle must be inhibited. this means the current master bus transaction is no longer required and should be ignored (mtab should not be asserted). mkillb is asserted late in the mtsb cycle. note that if there is no k-bus resident memory (icache, sram, or rom), mkillb never asserts. see table 2-2 for mtsb/mkillb interaction in table format. 2.2.6 master read data bus (mrdata[31:0]) these input signals provide the read data path between the system and the coldfire2/2m. the read data bus is 32-bits wide and can transfer 8, 16 or 32 bits of data per bus transfer. during a line transfer, the data bus is time-multiplexed across multiple clock cycles to transfer 128 bits. 2.2.7 master read data input enable (mie) this active-high input signal enables the capturing of mrdata[31:0]. power consumption can be reduced by minimizing signal switching in the coldfire2/2m by negating mie when mrdata[31:0] is invalid. mie must be asserted during all functional operation with one master and during k-bus memory testing. 2.2.8 master read/write (mrwb) this output signal indicates the direction of the data transfer for the current bus cycle. a high level indicates a read cycle and a low level indicates a write cycle. 2.2.9 master reset (mrstb) this active-low input signal instructs all master bus modules, including the coldfire2/2m, to enter reset mode. the coldfire2/2m will then initiate a reset exception. 2.2.10 master size (msiz[1:0]) these output signals indicate the data size for the bus transfer. refer to table 2-3 for the bus size encoding. table 2-2. m-bus protocol with respect to mtsb and mkillb mtsb mkillb m-bus protocol 0 1 initiate m-bus transfer 0 0 nop 1 x nop f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary motorola coldfire2/2m users manual 2-5 2.2.11 master transfer acknowledge (mtab) this active-low input signal is asserted by master bus slaves to indicate the successful completion of a requested bus transfer. 2.2.12 master transfer error acknowledge (mteab) this active-low input signal is asserted by master bus slaves to indicate an error condition for the current bus transfer. if mtab and mteab are both asserted, the cycle is terminated with an error because mteab always has precedence over mtab. 2.2.13 master transfer modifier (mtm[2:0]) these output signals provide supplemental information for each transfer type. the exact meaning depends on the mtt[1:0], and iack_68k signals as shown in table 2-4 . table 2-3. master bus transfer size encoding msiz[1:0] transfer size 00 longword (4 bytes) 01 byte (1 byte) 10 word (2 bytes) 11 line (16 bytes) table 2-4. master bus transfer modifier encoding mtm[2:0] mtt[1:0] coldfire iack mode 1 68k iack mode 2 000 0x reserved reserved 001 0x user data access user data access 010 0x user code access user code access 011 - 100 0x reserved reserved 101 0x supervisor data access supervisor data access 110 0x supervisor code access supervisor code access 111 0x reserved interrupt acknowledge/cpu space access 000 - 100 10 reserved reserved 101 10 emulator mode data access emulator mode data access 110 10 emulator mode code access emulator mode code access 111 10 reserved reserved 000 11 cpu space access reserved 001 11 interrupt acknowledge level 1 reserved 010 11 interrupt acknowledge level 2 reserved 011 11 interrupt acknowledge level 3 reserved 100 11 interrupt acknowledge level 4 reserved 101 11 interrupt acknowledge level 5 reserved 110 11 interrupt acknowledge level 6 reserved 111 11 interrupt acknowledge level 7 reserved notes: 1. 68k interrupt acknowledge mode signal negated (iack_68k = 0) 2. 68k interrupt acknowledge mode signal asserted (iack_68k = 1) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary 2-6 coldfire2/2m users manual motorola 2.2.14 master transfer start (mtsb) this active-low output signal indicates the start of each bus transfer. 2.2.15 master transfer type (mtt[1:0]) these output signals indicate the type of access of the current bus cycle. the exact meaning depends on the iack_68k signal as shown in table 2-5 . 2.2.16 master write data bus (mwdata[31:0]) these output signals provide the write data path between the coldfire2/2m and the system. the write data bus is 32-bits wide and can transfer 8, 16 or 32 bits of data per bus transfer. during a line transfer, the data bus is time-multiplexed across multiple clock cycles to carry 128 bits. 2.2.17 master write data output enable (mwdataoe) this active-high output signal indicates that the coldfire2/2m is driving the master write data bus. this is used to control optional bidirectional data bus three-state drivers. 2.3 general control signals 2.3.1 clock (clk) this input signal is the synchronous clock for the coldfire2/2m. clk is used to clock or sequence the coldfire2/2m internal logic. 2.3.2 interrupt priority level (iplb[2:0]) these active-low input signals indicate the encoded priority level of the requested interrupt. level seven, which cannot be masked, has the highest priority. level zero indicates no interrupt has been requested. these signals must maintain the interrupt request until the coldfire2/2m acknowledges the interrupt to guarantee that the interrupt is recognized. table 2-6 lists the interrupt levels, the iplb[2:0] states, and the mask value that allows an interrupt at each level. table 2-5. master bus transfer type encoding mtt[1:0] coldfire iack mode 1 68k iack mode 2 00 coldfire2/2m access acknowledge/cpu space/coldfire2/2m access 01 alternate master access alternate master access 10 emulator mode access emulator mode access 11 acknowledge/cpu space access reserved notes: 1. 68k interrupt acknowledge mode signal negated (iack_68k = 0) 2. 68k interrupt acknowledge mode signal asserted (iack_68k = 1) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary motorola coldfire2/2m users manual 2-7 2.4 integrated memory signals these signals interface the coldfire2/2m to an integrated instruction cache, roms and srams. 2.4.1 instruction cache signals the signals interface the coldfire2/2m to an optional integrated instruction cache. 2.4.1.1 instruction cache address bus (ich_addr[14:2]). these registered output signals provide the address of the current bus cycle (i.e. fetch cycle) to the integrated cache rams. ich_addr is only updated on fetch cycles (i.e. ich_addr does not get updated on sram or rom hits). this bus should be connected to the address bus (a) of the two compiled cache rams. 2.4.1.2 instruction cache data chip-select (ichd_csb). this active-low, output signal indicates the cache data ram is currently selected to perform a data transfer with the coldfire2/2m. this bus should be connected to the chip-select (csb) signal of the compiled cache data ram. 2.4.1.3 instruction cache data input bus (ichd_di[31:0]). these output signals provide the write data path between the coldfire2/2m and the cache data ram. the data bus is 32-bits wide and should be connected to the data inputs (dbi) of the compiled cache data ram. 2.4.1.4 instruction cache data output bus (ichd_do[31:0]). these input signals provide the read data path between the cache data ram and the coldfire2/2m. the data bus is 32-bits wide and should be connected to the data outputs (dbo) of the compiled cache data ram. 2.4.1.5 instruction cache data strobe (ichd_st). this output signal initiates a read or write cycle to the cache data ram on a low-to-high transition. this signal should be connected to the strobe input (st) signal of the compiled cache data ram. 2.4.1.6 instruction cache data read/write (ichd_rwb). this output signal indicates the direction of the data transfer to the cache data ram. a high level indicates a table 2-6. interrupt levels and mask values requested interrupt level control line status interrupt mask level required for recognition iplb[2] iplb[1] iplb[0] 0 high high high no request 1 high high low 0 2 high low high 0-1 3 high low low 0-2 4 low high high 0-3 5 low high low 0-4 6 low low high 0-5 7 low low low 0-7 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary 2-8 coldfire2/2m users manual motorola read cycle and a low level indicates a write cycle. it should be connected to the read/write (rwb) signal of the compiled cache data ram. 2.4.1.7 instruction cache size (ich_sz[2:0]). these static inputs specify the size of the compiled cache rams connected to the coldfire2/2m processor. table 2-7 lists the possible cache configurations. ich_sz[2:0] does not affect the cacr in any way; thus a movec instruction will write the cacr regardless of the ich_sz specification (which is contrary to the sram_sz and rom_sz effect during rambar and rombar loading). 2.4.1.8 instruction cache tag chip-select (icht_csb). this active-low output signal indicates the cache tag ram is currently selected to perform a data transfer with the coldfire2/2m. this signal should be connected to the chip-select (csb) signal of the compiled cache tag ram. 2.4.1.9 instruction cache tag input bus (icht_di[31:8]). these output signals provide the write data path between the coldfire2/2m and the cache tag ram. the data bus is 24-bits wide. bit eight is always the valid bit and is always used as seen in the cache configuration shown in table 2-8 . this bus should be connected to the data inputs (dbi) of the compiled cache tag ram. functionally, ich_addr[31:9] is what gets written onto icht_di[31:9] and icht_di[8] is written with the valid state of the entry. table 2-7. cache configuration encoding cache size (bytes) ich_sz[2:0] tag ram (bits) data ram (bits) address data address data none 000 - - - - 512 001 5 24 7 32 1 k 010 6 23 8 32 2 k 011 7 22 9 32 4 k 100 8 21 10 32 8 k 101 9 20 11 32 16k? 110 10 19 12 32 32k? 111 11 18 13 32 note: ?16k and 32k rams may require a reduced operating frequency in hpf65 coldfire2 hard macro. table 2-8. valid tag ram data signals cache size (bytes) valid data bits 512 icht_dx[31:8] 1 k icht_dx[31:10,8] 2 k icht_dx[31:11,8] 4 k icht_dx[31:12,8] 8 k icht_dx[31:13,8] 16k icht_dx[31:14,8] 32k icht_dx[31:15,8] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary motorola coldfire2/2m users manual 2-9 2.4.1.10 instruction cache tag output bus (icht_do[31:8]). these input signals provide the read data path between the cache tag ram and the coldfire2/2m. the data bus is 24-bits wide. bit eight is always the valid bit and is always used regardless of the cache configuration as shown in table 2-8 . this bus should be connected to the data outputs (dbo) of the compiled cache tag ram. unused signals must be tied low. 2.4.1.11 instruction cache tag strobe (icht_st). this output signal initiates a read or write cycle to the cache tag ram on a low-to-high transition. this signal should be connected to the strobe input (st) signal of the compiled cache tag ram. 2.4.1.12 instruction cache tag read/write (icht_rwb). this output signal indicates the direction of the data transfer to the cache tag ram. a high level indicates a read cycle and a low level indicates a write cycle. it should be connected to the read/write (rwb) signal of the compiled cache tag ram. 2.4.2 integrated rom signals these signals interface the coldfire2/2m to the optional integrated roms. 2.4.2.1 rom address bus (rom_addr[14:2]). these output signals provide the address of the current bus cycle to the integrated roms. this bus should be connected to the address bus (a) of the compiled roms. the number of valid address signals depends on the total rom size as shown in table 2-9 . 2.4.2.2 rom data output bus (rom_do[31:0]). these input signals provide the read data path from the integrated roms to the coldfire2/2m. the data bus is 32-bits wide and can transfer 8, 16, or 32 bits of data per bus transfer. during a line transfer, the data lines are time-multiplexed across multiple cycles to carry 128 bits. this bus should be connected to the data outputs (do) of the compiled roms. 2.4.2.3 rom enable (rom_enb[1:0]). these active-low, output signals indicate the roms are currently selected to drive the rom_do[31:0] bus. these signals should be connected individually to the enable signal (romenb) signal of the compiled roms. both are asserted for 32-bit accesses. rom_enb[0] connects to the msw while rom_enb[1] connects to the lsw. table 2-9. valid rom address bits total rom size valid rom_addr bits 0 none 512 rom_addr[8:2] 1 k rom_addr[9:2] 2 k rom_addr[10:2] 4 k rom_addr[11:2] 8 k rom_addr[12:2] 16k rom_addr[13:2] 32k rom_addr[14:2] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary 2-10 coldfire2/2m users manual motorola 2.4.2.4 rom size (rom_sz[2:0]). these static inputs specify the size of the compiled roms connected to the coldfire2/2m. these pins need to stay valid during all operation. table 2-10 lists the possible rom configurations. if the rom_sz pins are zero, the rom cannot be enabled via a cpu space write to rombar. therefore if the rom is enabled while the rom_sz pins are at zero, the processor behaves as if no rom module existed. 2.4.2.5 rom valid (rom_vld). this active-high input signal determines if the rom module should be active immediately after a hard reset. thus, if asserted, the first fetches ($0, $4) go to rom instead of external memory. rom_vld controls the reset value of the rom base address register (i.e. the rom must be based at $0000 if rom_vld is asserted). 2.4.3 integrated sram signals these signals interface the coldfire2/2m to the optional integrated srams. 2.4.3.1 sram address bus (sram_addr[14:2]). these registered output signals provide the address of the current bus cycle to the integrated srams. this bus should be connected to the address bus (a) of the four compiled srams. the number of valid address signals depends on the total sram size as shown in table 2-11 . table 2-10. rom configuration encoding total rom size (bytes) rom_sz[2:0] address (bits) data 1 (bits) none 000 - - 512 001 7 2 @ 16 1 k 010 8 2 @ 16 2 k 011 9 2 @ 16 4 k 100 10 2 @ 16 8 k 101 11 2 @ 16 16k 2 110 12 2 @ 16 32k 2 111 13 2 @ 16 notes: 1. 2 roms, each 16-bits wide 2. 16k and 32k roms may require a reduced operating frequency. table 2-11. valid sram address bits total sram size valid sram_addr bits 0 none 512 sram_addr[8:2] 1 k sram_addr[9:2] 2 k sram_addr[10:2] 4 k sram_addr[11:2] 8 k sram_addr[12:2] 16k sram_addr[13:2] 32k sram_addr[14:2] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary motorola coldfire2/2m users manual 2-11 2.4.3.2 sram chip-select (sram_csb). this active-low output signal indicates the srams are currently selected to perform a data transfer with the coldfire2/2m. this signal should be connected to the chip-select (csb) signal of the four compiled srams. 2.4.3.3 sram data input bus (sram_di[31:0]). these output signals provide the write data path between the coldfire2/2m and the integrated sram. the data bus is 32-bits wide and can transfer 8, 16, or 32 bits of data per bus transfer. during a line transfer, the data lines are time-multiplexed across multiple cycles to carry 128 bits. this bus should be connected to the data inputs (dbi) of the four compiled srams. if only one byte is being written, the byte will be replicated on all 4 lines likewise a word will be replicated in both word positions. 2.4.3.4 sram data output bus (sram_do[31:0]). these input signals provide the read data path between the integrated ram and the coldfire2/2m. the data bus is 32-bits wide and can transfer 8, 16 or 32 bits of data per bus transfer. during a line transfer, the data lines are time-multiplexed across multiple cycles to carry 128 bits. this bus should be connected to the data outputs (dbo) of the four compiled srams. 2.4.3.5 sram size (sram_sz[2:0]). these static inputs specify the size of the compiled srams connected to the coldfire2/2m. these pins need to stay valid during all operation. if the sram_sz pins are zero, the sram cannot be enabled via a cpu space write to rambar. therefore if the ram is enabled while the sram_sz pins are at zero, the processor behaves as if no sram module existed. table 2-12 lists the possible sram configurations. 2.4.3.6 sram strobe (sram_st[3:0]). these output signals initiate a read or write cycle to the integrated srams on a low-to-high transition. these signals should be connected individually to the strobe input (st) signals of the four compiled srams. the st[0] signal connects to the high-order byte and st[3] connects to the low-order byte. table 2-12. sram configuration encoding total sram size (bytes) ram_sz[2:0] address (bits) data 1 (bits) none 000 - - 512 001 7 4@8 1 k 010 8 4@8 2 k 011 9 4@8 4 k 100 10 4@8 8 k 101 11 4@8 16k 2 110 12 4@8 32k 2 111 13 4@8 notes: 1. 4 rams, each 8-bits wide 2. 16k and 32k rams may require a reduced operating frequency. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary 2-12 coldfire2/2m users manual motorola 2.4.3.7 sram read/write (sram_rwb[3:0]). these output signals indicate the direction of the data transfer to the to the integrated srams. a high level indicates a read cycle and a low level indicates a write cycle. they should be connected individually to the read/write (rwb) signal of the four compiled srams. like sram_st[3:0], the sram_rwb[3] signal connects to the high-order byte and sram_st[0] connects to the low-order byte. 2.5 debug signals 2.5.1 break point (bkptb) this active-low, unidirectional input signal is used to request a manual break point. it will cause the processor to enter a halted state after the completion of the current instruction. this status will be reflected on the processor status (pst) outputs. 2.5.2 debug data (ddata[3:0]) these output signals display the captured processor status and break point status. 2.5.3 development serial clock (dsclk) this input signal is used as the development serial clock for the serial interface to the debug module.the maximum frequency is 1/2 the clock (clk) frequency. 2.5.4 development serial input (dsi) this input signal provides the single-bit communication for the debug module commands. 2.5.5 development serial output (dso) this output signal provides single-bit communication for the debug module responses. 2.5.6 processor status (pst[3:0]) these output signals report the processor status. table 2-13 shows the encoding of these signals. these signals indicate the current status of the processor pipeline and, as a result, are not related to the current bus transfer. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary motorola coldfire2/2m users manual 2-13 . 2.6 test signals 2.6.1 signals required to perform scan test this section describes the coldfire2/2m signals dedicated to the scan testing of the coldfire2/2m. all coldfire2/2m signals are unidirectional and synchronous. 2.6.1.1 scan enable (scan_enable). this active-high input signal enables scan testing of the coldfire2/2m. it forces all internal flip-flops to be linked together into sixteen parallel scan chains. this signal must be negated for functional operation. 2.6.1.2 scan exercise array (scan_xarray). this active-high input signal is used to exercise the integrated memory arrays during scan testing. this signal causes random writes to the internal rams by strobing the write strobes while scanning. 2.6.1.3 scan input (si[15:0]). these input signals are connected to the16 internal coldfire2/2m scan chain inputs. 2.6.1.4 scan mode (scan_mode). this active-high, unidirection input signal gates off all memory array outputs during scan testing.scan_mode should be held asserted for the duration of scan testing. 2.6.1.5 scan output (so[15:0]). these output signals are connected to the 16 internal coldfire2/2m scan chain outputs. table 2-13. processor status encoding pst[3:0] definition (hex) (binary) $0 0000 continue execution $1 0001 begin execution of an instruction $2 0010 reserved $3 0011 entry into user-mode $4 0100 begin execution of pulse and wddata instructions $5 0101 begin execution of taken branch $6 0110 reserved $7 0111 begin execution of rte instruction $8 1000 begin 1-byte transfer on ddata $9 1001 begin 2-byte transfer on ddata $a 1010 begin 3-byte transfer on ddata $b 1011 begin 4-byte transfer on ddata $c 1100 exception processing? $d 1101 emulator-mode entry exception processing? $e 1110 processor is stopped, waiting for interrupt? $f 1111 processor is halted ? note: ?these encodings are asserted for multiple cycles. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary 2-14 coldfire2/2m users manual motorola 2.6.1.6 io test ring clock (trclk). this input signal is the synchronous clock used to transition the test ring during scan testing. tr_clk is connected to the clock input of all io test ring registers. 2.6.1.7 io test ring core mode enable (core_test). this active-high input signal enables the core mode of the test ring during scan testing. the test ring is in scan core mode if core_test is asserted and in scan asic mode if core_test is negated. 2.6.1.8 io test ring data input (tr_sdi[1:0]). these input signals are the serial data inputs for test ring chain one and zero. 2.6.1.9 io test ring data output (tr_sdo[1:0]). these output signals are the serial output data from test ring chain one and zero. 2.6.1.10 io test ring enable (tr_se). this active-high input signal enables the test ring. tr_se is connected to the scan enable input of all io test ring scannable registers. 2.6.1.11 io test ring mode (tr_mode). this active-high input signal enables the scan test mode of the test ring. the test ring is in scan test mode if tr_mode is asserted and in normal functional mode if negated. tr_mode should be asserted for the duration of scan testing, and be held negated for the duration of memory testing and during functional operation of the device. 2.6.2 integrated memory test signals this section describes the coldfire2/2m signals dedicated to testing the integrated memories. other signals are required to be either controlled or brought out (muxed) to package pins as well (see section 8 test operation ). 2.6.2.1 test address bus (test_addr[14:2]). these input signals specify an address when testing the integrated memories. 2.6.2.2 test control (test_ctrl). this active-high input signal indicates the test address bus (test_addr[14:2]) will be latched on the next positive clock edge. 2.6.2.3 test idata read (test_idata_rd). this active-high input signal tests the instruction cache data memory read operation. 2.6.2.4 test idata write (test_idata_wrt). this active-high input signal tests the instruction cache data memory write operation. 2.6.2.5 test instruction cache read hit (test_rhit). this active-high output signal indicates a hit has occurred when accessing the instruction cache during memory array testing. 2.6.2.6 test invalidate inhibit (test_ivld_inh). this active-high input signal inhibits the invalidate operation when testing the instruction cache. 2.6.2.7 test itag write (test_itag_wrt). this active-high input signal tests the instruction cache tag memory write operation. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
signal summary 2-15 coldfire2/2m users manual motorola 2.6.2.8 test kta mode enable (test_kta). this active-high input signal allows the instruction cache tag and data arrays to be read in parallel, mimicking the functional operation. this allows testing of the speed path from the tag and data arrays to the core. 2.6.2.9 test mode enable (test_mode). this active-high input signal enables all of the integrated memory test signals. test_mode should be asserted for the duration of memory testing. 2.6.2.10 test sram read (test_sram_rd). this active-high input signal tests the integrated sram memory read operation. 2.6.2.11 test sram write (test_sram_wrt). this active-high input signal tests the integrated sram memory write operation. 2.6.2.12 test read (test_rd). this active-high input signal tests read operations on all of the integrated memories. 2.6.2.13 test rom read (test_rom_rd). this active-high input signal tests the integrated rom memory read operation. 2.6.2.14 test write inhibit (test_wr_inh). this active-high input signal disables the write strobes to the sram and instruction cache compiled rams. test_wr_inh should be negated for the duration of memory test. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 3-1 section 3 master bus operation the master bus provides a basic two cycle bus protocol, similar to that used by previous generations of m68000 microprocessors. basic cycles are defined as the transfer start (ts) cycle and the transfer acknowledge (ta) cycle. the address and control information is driven onto the bus during the ts cycle, and the data is valid during the subsequent ta cycle. by delaying the assertion of the transfer acknowledge signal, the bus automatically inserts wait states to easily accommodate any slave response speed. the following sections detail the signal descriptions, data transfer mechanism, and bus transfer protocols. 3.1 signal description this section describes the coldfire2/2m signals associated with the master bus. all coldfire2/2m signals are unidirectional and synchronous. 3.1.1 68k interrupt acknowledge mode enable (iack_68k) this active-high input signal enables the 68k interrupt acknowledge mode. in this mode, the maddr [31:0], mtt [1:0], and mtm [2:0] signals mimic the 68k address bus and function codes during interrupt acknowledge and cpu space bus cycles. this is a static input. refer to section 3.7.1 interrupt acknowledge bus cycle for more information. 3.1.2 master address bus (maddr[31:0]) during a normal bus cycle, this 32-bit output bus provides the address of the first item of a bus transfer. it is capable of addressing four gbytes of address space. 3.1.3 master arbiter control (marbc[1:0]) these output signals can be used to specify the mode of operation for an optional arbitration module. they reflect the bit positions [17:16] in the cacr. if an optional arbitration module is not used, these signals can be used as general purpose output signals. 3.1.4 master freeze (mfrzb) this active-low output signal indicates that the core has been halted. mfrzb is not part of the m-bus protocol. it is simply a signal that can be used to alert timers or other peripheral modules that the core has been halted. 3.1.5 master kill (mkillb) this active-low output signal qualifies mtsb (i.e. it can assert in other cycles but is only significant in a cycle where mstb is asserted). when mkillb is asserted simultaneously with mtsb assertion, this indicates a hit in a k-bus memory and that the external cycle must f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-2 coldfire2/2m users manual motorola be inhibited. this means the current master bus transaction is no longer required and should be ignored (mtab should not be asserted). mkillb is asserted late in the mtsb cycle. note that if there is no k-bus resident memory (icache, sram, or rom), mkillb never asserts. see table 3-1 for mtsb/mkillb interaction in table format. 3.1.6 master read data bus (mrdata[31:0]) these input signals provide the read data path between the system and the coldfire2/2m device. the read data bus is 32-bits wide and can transfer 8, 16 or 32 bits of data per bus transfer. during a line transfer, the data bus is time-multiplexed across multiple clock cycles to transfer 128 bits. 3.1.7 master read data input enable (mie) this active-high input signal enables the capturing of mrdata [31:0]. power consumption can be reduced by minimizing signal switching in the coldfire2/2m by negating mie when mrdata[31:0] is invalid. mie must be asserted during all one-master functional operation and during k-bus memory testing. 3.1.8 master read/write (mrwb) this output signal indicates the direction of the data transfer for the current bus cycle. a high level indicates a read cycle and a low level indicates a write cycle. 3.1.9 master reset (mrstb) this active-low input signal instructs all master bus modules, including the coldfire2/2m device, to enter reset mode. the coldfire2/2m processor will then initiate a reset exception. 3.1.10 master size (msiz[1:0]) these output signals indicate the data size for the bus transfer. refer to table 3-2 for the bus size encoding. table 3-1. m-bus protocol with respect to mtsb and mkillb mtsb mkillb m-bus protocol 0 1 initiate m-bus transfer 0 0 nop 1 x nop table 3-2. master bus transfer size encoding msiz[1:0] transfer size 00 longword (4 bytes) 01 byte (1 byte) 10 word (2 bytes) 11 line (16 bytes) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-3 3.1.11 master transfer acknowledge (mtab) this active low input signal is asserted by master bus slaves to indicate the successful completion of a requested bus transfer. 3.1.12 master transfer error acknowledge (mteab) this active low input signal is asserted by master bus slaves to indicate an error condition for the current bus transfer. if mtab and mteab are both asserted, the cycle is terminated with an error because mteab always has precedence over mtab. 3.1.13 master transfer modifier (mtm[2:0]) these output signals provide supplemental information for each transfer type. the exact meaning depends on the mtt [1:0], and iack_68k signals as shown in table 3-3 . 3.1.14 master transfer start (mtsb) this active low output signal indicates the start of each bus transfer. 3.1.15 master transfer type (mtt[1:0]) these output signals indicate the type of access of the current bus cycle. the exact meaning depends on the iack_68k signal as shown in table 3-4 . table 3-3. master bus transfer modifier encoding mtm[2:0] mtt[1:0] coldfire iack mode 1 68k iack mode 2 000 0x reserved reserved 001 0x user data access user data access 010 0x user code access user code access 011 - 100 0x reserved reserved 101 0x supervisor data access supervisor data access 110 0x supervisor code access supervisor code access 111 0x reserved interrupt acknowledge/cpu space access 000 - 100 10 reserved reserved 101 10 emulator mode data access emulator mode data access 110 10 emulator mode code access emulator mode code access 111 10 reserved reserved 000 11 cpu space access reserved 001 11 interrupt acknowledge level 1 reserved 010 11 interrupt acknowledge level 2 reserved 011 11 interrupt acknowledge level 3 reserved 100 11 interrupt acknowledge level 4 reserved 101 11 interrupt acknowledge level 5 reserved 110 11 interrupt acknowledge level 6 reserved 111 11 interrupt acknowledge level 7 reserved notes: 1. 68k interrupt acknowledge mode signal negated (iack_68k = 0) 2. 68k interrupt acknowledge mode signal asserted (iack_68k = 1) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-4 coldfire2/2m users manual motorola 3.1.16 master write data bus (mwdata[31:0]) these output signals provide the write data path between the coldfire2/2m and the system. the write data bus is 32-bits wide and can transfer 8, 16 or 32 bits of data per bus transfer. during a line transfer, the data bus is time-multiplexed across multiple clock cycles to carry 128 bits. 3.1.17 master write data output enable (mwdataoe) this active high output signal indicates that the coldfire2/2m is driving the master write data bus. this is used to control optional bidirectional data bus three-state drivers. 3.2 data transfer mechanism 3.2.1 transfer type control signals the transfer type control signals indicate the type of bus transaction occurring on the master bus. this includes the master transfer type ( mtt [1:0]) and master transfer modifier ( mtm [2:0]) signals. the mtt[1:0] signals indicates the type of transfer and the mtm[2:0] signals provide supplemental information. the encodings for mtt[1:0] and mtm[2:0] are shown in table 3-3 and table 3-4 . the transfer type attributes for accesses made through the debug module are determined by the programming of the debug module (see section 7.4.2.2 address attribute register (aatr) ). 3.2.1.1 coldfire2/2m access. if the coldfire2/2m requests a master bus transfer, it drives the mtt [1:0] signals with a coldfire2/2m access encoding. the mtm [2:0] encoding will depend on the privilege mode and address space of the transfer: privilege modesupervisor/user mode access when the supervisor (s) bit in the status register ( sr ) is set, the coldfire2/2m will drive mtm [2:0] with one of the supervisor mode encodings during a master bus transfer. when the s bit in the sr is clear, the coldfire2/2m will drive mtm[2:0] with one of the user mode access encodings. address spacecode/data access if the coldfire2/2m accesses the code space, it will drive mtm [2:0] with one of the code access encodings during a master bus transfer. code space accesses are opcode fetches or operand fetches in one of the pc relative addressing modes. if the coldfire2/2m accesses the data space, it will drive mtm[2:0] with one of the data access encodings. data table 3-4. master bus transfer type encoding mtt[1:0] coldfire iack mode 1 68k iack mode 2 00 coldfire2/2m access acknowledge/cpu space/coldfire2/2m access 01 alternate master access alternate master access 10 emulator mode access emulator mode access 11 acknowledge/cpu space access reserved notes: 1. 68k interrupt acknowledge mode signal negated (iack_68k = 0) 2. 68k interrupt acknowledge mode signal asserted (iack_68k = 1) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-5 space accesses are operand fetches that are not in one of the pc relative addressing modes. 3.2.1.2 alternate master access. when an alternate master requests a master bus transfer, the mtt [1:0] signals should be driven with the alternate master access encoding by the alternate master. the mtm [2:0] encoding is the same as that for the coldfire2/2m access. 3.2.1.3 emulator mode access. accesses made while in emulator mode generate emulator mode accesses on the master bus. this is controlled by the configuration/status register ( csr ) in the debug module. refer to section 7.4.1.1 emulator mode . emulator mode accesses result in he mtt [1:0] signals being driven with the emulator mode access encoding. the encoding of the mtm [2:0] signals will be one of the emulator mode encodings depending on the address space as defined for the coldfire2/2m access. 3.2.1.4 interrupt acknowledge access. interrupt acknowledge bus cycles are indicated as an acknowledge/cpu space access on the mtt [1:0] signals (the encoding depends on the interrupt acknowledge mode). if the coldfire2/2m is in the coldfire interrupt acknowledge mode, the mtm [2:0] signals are driven with the pending interrupt level number. in the 68k interrupt acknowledge mode, the mtm[2:0] signals will be driven high. refer to section 3.7 interrupt acknowledge bus cycles . 3.2.1.5 cpu space access. cpu space accesses are indicated as a acknowledge/ cpu space access on the mtt [1:0] signals. the mtm [2:0] signals are driven with a cpu space encoding. the specific encoding depends on the interrupt acknowledge mode. many debug commands and movec instructions result in cpu space accesses. 3.2.2 data bus requirements the coldfire2/2m designates all operands for transfers on a byte-boundary system. a line- sized operand (16 bytes) is four longwords. for all data transfers, maddr [31:2] indicates the longword base address of the first byte of the reference item. maddr[1:0] indicates the byte offset from this base address. the msiz [1:0] signals along with the low-order two address signals are used to determine how the data bus is used. table 3-5 indicates the mrdata [31:0] requirements for slave devices when responding to read transfers. a - indicates a dont care, i.e. the value is ignored. table 3-5. mrdata requirements for read transfers transfer size msiz[1:0] maddr[1:0] mrdata[31:24] mrdata[23:16] mrdata[15:8] mrdata[7:0] byte 01 00 byte data - - - 01 01 - byte data - - 01 10 - - byte data - 01 11 - - - byte data word 10 00 word data - 10 10 - word data long 00 00 longword data line 11 00 first longword data f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-6 coldfire2/2m users manual motorola some of the system bus controllers (sbc) within the coldfire architecture support dynamically-sized external data transfers, i.e., the slave indicates the width of the data port at the time of the transfer. to support this bus sizing feature, there are certain data replication functions which must be performed by all master devices, including the coldfire2/2m, during write cycles. table 3-6 indicates the mwdata [31:0] requirements for master devices when initiating write transfers. note that the coldfire2/2m device, as well as all masters, places the byte operand on all byte lanes for a byte write cycle, and the word operand on both word lanes for a word write cycle. 3.3 data transfers 3.3.1 byte, word, and longword read transfers for byte, word, and longword read accesses, the coldfire2/2m requests data from a slave device. the bus operations are similar for the different sized accesses, with the msiz [1:0] signals defining the access size. based on the transfer size, the data is placed on the appropriate byte lanes of the read data bus ( mrdata [31:0]). this read transfer is defined in figure 3-1 . table 3-6. mwdata bus requirements for write transfers transfer size msiz[1:0] maddr[1:0] mwdata[31:24] mwdata[23:16] mwdata[15:8] mwdata[7:0] byte 01 -- byte data byte data byte data byte data word 10 -0 word data word data long 00 00 longword data line 11 00 first longword data f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-7 see figure 3-2 for an example of normal read and write master bus transfers without wait states. figure 3-1. byte, word, and longword read transfer flowchart 1) drive address on maddr [31:0] 2) set mrwb to read (mrwb = 1) 3) set msiz [1:0] to byte, word, or long 4) set mtt [1:0], mtm [2:0] to signal the appropriate access type 5) assert mtsb for one clk cycle 1) register the mrdata [31:0] bus 2) recognize the transfer is done 1) decode the address and select the appropriate slave device. insert necessary wait-states. 1) drive the data on the appropriate byte lanes of the mrdata [31:0] bus, based on msiz [1:0], maddr 1:0]. 2) assert mie for one clk cycle. 2) assert mtab for one clk cycle. master slave f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-8 coldfire2/2m users manual motorola clock 1 (c1) the read cycle starts in c1. during the first half of c1, the coldfire2/2m places valid values on the master address bus ( maddr [31:0]) and transfer control signals. the mtt [1:0] and mtm [2:0] signals identify the specific access type. the master read/write ( mrwb ) signal is driven high for a read cycle, and the master size signals ( msiz [1:0]) indicate transfer size. the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c1 to indicate the beginning of a bus cycle. clock 2 (c2) during the first half of c2, the coldfire2/2m negates mtsb . the selected device uses the mrwb and msiz signals to place the data on the master read data bus ( mrdata [31:0]) and assert the master input enable ( mie ) signal. concurrently, the selected device asserts the master transfer acknowledge ( mtab ) signal. at the end of c2, the coldfire2/2m samples the level of mtab and latches the current value on mrdata[31:0]. if mtab is figure 3-2. normal transfer (without wait states) c1 c2 c1 c2 clk a1 a2 maddr mrwb msiz mtt mtm mtsb d2 mwdata mwdataoe d1 mrdata mie mtab mteab read write f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-9 asserted, the transfer terminates. if mtab is not asserted, the coldfire2/2m processor ignores the data and inserts wait states instead of terminating the transfer. the coldfire2/ 2m continues to sample mtab on successive rising edges of clk until it is asserted. the selected device negates the mie and mtab signals in the first half of the next clk cycle. 3.3.2 normal tranfers with mkillb figure3-3 illustrates normal tranfers (no wait states) with mkillb asserted and negated: figure 3-3. normal transfer with mkillb timing diagram (without wait states) clock 1 (c1) the read cycle starts in c1. during the first half of c1, the coldfire2/2m places valid values on the master address bus ( maddr [31:0]) and transfer control signals. the mtt [1:0] and mtm [2:0] signals identify the specific access type. the master read/write ( mrwb ) signal is c7 c6 c5 c4 c3 c2 c1 clk maddr mrwb msiz mtt mtm mtsb mkillb mwdata wdata mrdata mie mtab external read internal integ. memory read external write internal integ. memory write next access mteab f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-10 coldfire2/2m users manual motorola driven high for a read cycle, and the master size signals ( msiz [1:0]) indicate transfer size. the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c1 to indicate the beginning of a bus cycle. mkillb does not assert during this cycle so the access will go external via m-bus. clock 2 (c2) during the first half of c2, the coldfire2/2m processor negates mtsb . the selected device uses the mrwb and msiz signals to place the data on the master read data bus ( mrdata [31:0]) and assert the master input enable ( mie ) signal. concurrently, the selected device asserts the master transfer acknowledge ( mtab ) signal. at the end of c2, the coldfire2/2m processor samples the level of mtab and latches the current value on mrdata[31:0]. if mtab is asserted, the transfer terminates. if mtab is not asserted, the coldfire2/2m ignores the data and inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is asserted. the selected device negates the mie and mtab signals in the first half of the next clk cycle. clock 3 (c3) another read cycle begins in c3; however this read cycle will be to internal integrated memory. during the first half of c3, the coldfire2/2m places valid values on the master address bus ( maddr [31:0]) and transfer control signals. the mtt [1:0] and mtm [2:0] signals identify the specific access type. the master read/write ( mrwb ) signal is driven high for a read cycle, and the master size signals ( msiz [1:0]) indicate transfer size. the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c3 to indicate the beginning of a bus cycle. mkillb asserts late in the cycle. the combination of the assertion of both ( mtsb and mkillb signifies that the access will occur internally in integrated memory and the m-bus transaction is not needed. the internal access will complete in one cycle; therefore another access can begin immediately on the following cycle. clock 4 (c4) a write cycle starts in c4. during the first half of c4, the coldfire2/2m places valid values on the master address bus ( maddr [31:0]) and transfer control signals. the mtt [1:0] and mtm [2:0] signals identify the specific access type. the master read/write ( mrwb ) signal is driven low for a write cycle, and the master size signals ( msiz [1:0]) indicate transfer size. the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c4 to indicate the beginning of a bus cycle. mkillb does not assert during this cycle so the access will go external via m-bus. cycle 5 (c5) during the first half of c5, the coldfire2/2m negates mtsb . the selected device uses the mrwb and msiz signals to take the data off the master write data bus ( mwdata [31:0]. concurrently, the selected device asserts the master transfer acknowledge ( mtab signal. at the end of c5, if mtab is asserted, the transfer terminates. if mtab is not asserted, the coldfire2/2m inserts wait states instead of terminating the transfer. the coldfire2/2m f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-11 continues to sample mtab on successive rising edges of clk until it is asserted. the selected device negates the ( mtab signal in the first half of the next clk cycle. cycle 6 (c6) another write cycle begins in c6; however this write cycle will be to internal integrated memory. during the first half of c6, the coldfire2/2m places valid values on the master address bus ( maddr [31:0]) and transfer control signals. the mtt [1:0] and mtm [2:0] signals identify the specific access type. the master read/write ( mrwb ) signal is driven low for a write cycle, and the master size signals ( msiz [1:0]) indicate transfer size. the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c6 to indicate the beginning of a bus cycle. mkillb asserts late in the cycle. the combination of the assertion of both ( mtsb and mkillb signifies that the access will occur internally in integrated memory and the m-bus transaction is not needed. the internal access will complete in one cycle; therefore another access can begin immediately on the following cycle. cycle 7 (c7) another access can begin in this cycle (( mtsb can assert). 3.3.3 byte, word, and longword write transfers for byte, word, and longword write accesses, the coldfire2/2m transfers data to a slave device. the bus operations are similar for the different sized accesses, with the msiz [1:0] signals defining the access size. based on the transfer size, the data is placed on the appropriate byte lanes of the write data bus ( mwdata [31:0]). this write transfer is defined in figure 3-4 . f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-12 coldfire2/2m users manual motorola see figure 3-2 for an example of a normal write transfer without wait states. figure 3-5 shows an example of a normal write master bus transfer with wait states. note the cxw nomenclature is used to define a wait state during the bus cyclee.g., c2w. figure 3-4. byte, word, and longword write transfer flowchart master 1) drive address on maddr [31:0] 2) set mrwb to write (mrwb = 0) 3) set msiz [1:0] to byte, word, or long 4) set mtt [1:0], mtm 2:0] to signal the appropriate access type 5) assert mtsb for one clk cycle 1) drive data onto mwdata [31:0] 2) assert mwdataoe 1) recognize the transfer is done 2) negate mwdataoe 1) decode the address and select the appropriate slave device. insert necessary wait-states. 1) capture the data from the appropriate byte lanes of the mwdata [31:0] bus, based on msiz [1:0], maddr [1:0] 2) assert mtab for one clk cycle slave f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-13 clock 1 (c1) the write cycle starts in c1. during the first half of c1, the coldfire2/2m places valid values on the master address bus ( maddr [31:0]) and transfer control signals. the master transfer type ( mtt [1:0]) and master transfer modifier ( mtm [2:0]) signals identify the specific access type. the master read/write ( mrwb ) signal is driven low for a write cycle, and the master size signals ( msiz [1:0]) indicate transfer size. the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c1 to indicate the beginning of a bus cycle. clock 2 (c2) during the first half of c2, the coldfire2/2m negates mtsb , places the data on the master write data bus ( mwdata [31:0]), and asserts the master write data output enable ( mwdataoe ) signal. concurrently, the selected device asserts the master transfer acknowledge ( mtab ) signal if it is ready to latch the data. at the end of c2, the selected device latches the current value on mwdata[31:0], and the coldfire2/2m samples the level of mtab. if mtab is asserted, the bus cycle is terminated and the coldfire2/2m negates mwdataoe in the first half on the next clk cycle. if mtab is not asserted, the coldfire2/2m inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is asserted. figure 3-5. normal write transfer (with wait states) c1 c2w c2w c2 clk a1 maddr mrwb msiz mtt mtm mtsb d1 mwdata mwdataoe mtab mteab write f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-14 coldfire2/2m users manual motorola 3.3.4 line read transfer the coldfire2/2m uses a line read access to fetch a four-longword operand using a burst transfer. a line read accesses a block of four longwords, aligned to a 16-byte memory boundary, by supplying a starting address that points to the critical longword in the four- longword block. the address and attributes driven by the coldfire2/2m remain stable throughout the entire transfer. as a result, the slave device must increment maddr [3:2] internally to sequence for each transfer with the address wrapping around at the end of the block. the allowable longword fetch patterns during a line access are shown in table 3-7 . the responding slave device terminates each longword transfer on the mrdata [31:0] bus by asserting the transfer acknowledge control signal, mtab . all devices on the master bus must support burst accesses. the assertion of mteab aborts the line read access. see section 3.8.1 access errors for more information on mteab. a line read burst is defined in figure 3-6 . table 3-7. allowable line access patterns maddr[3:2] longword access order 00 0 - 4 - 8 - c 01 4 - 8 - c - 0 10 8 - c - 0 - 4 11 c - 0 - 4 - 8 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-15 see figure 3-7 for an example of a line read burst master bus transfer without wait states. figure 3-6. line read transfer flowchart master 1) drive address on maddr [31:0] 2) set mrwb to read (mrwb = 1) 3) set msiz [1:0] to line 4) set mtt [1:0] and mtm [2:0] to signal the appropriate access type 5) assert mtsb for one clk cycle 1) register the mrdata [31:0] bus 2) recognize the 1st transfer is done 1) register the mrdata [31:0] bus 2) recognize the 2nd transfer is done 1) register the mrdata [31:0] bus 2) recognize the 3rd transfer is done 1) register the mrdata [31:0] bus 2) recognize the 4th transfer is done 1) decode the address and select the appropriate slave device. insert necessary wait-states. 1) drive the data on mrdata [31:0]. 2) assert mie for one clk cycle. 3) assert mtab for one clk cycle. 1) drive the data on mrdata [31:0]. 2) assert mie for one clk cycle. 3) assert mtab for one clk cycle. 1) drive the data on mrdata [31:0]. 2) assert mie for one clk cycle. 3) assert mtab for one clk cycle. 1) drive the data on mrdata [31:0]. 2) assert mie for one clk cycle. 3) assert mtab for one clk cycle. slave f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-16 coldfire2/2m users manual motorola clock 1 (c1) the line read cycle starts in c1. during the first half of c1, the coldfire2/2m places valid values on the master address bus ( maddr [31:0]) and transfer control signals. the master transfer type ( mtt [1:0]) and master transfer modifier ( mtm [2:0]) signals identify the specific access type. the master read/write ( mrwb ) signal is driven high for a read cycle, and the master size signals ( msiz [1:0]) indicate transfer size (line = $3). the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c1 to indicate the beginning of a bus cycle. clock 2 (c2) during the first half of c2, the coldfire2/2m negates mtsb . the selected device uses the mrwb and msiz signals to place the data on the master read data bus ( mrdata [31:0]) and assert the master input enable ( mie ) signal. the first transfer must supply the longword at the corresponding long-word boundary. concurrently, the selected device asserts master transfer acknowledge ( mtab ) signal. at the end of c2, the coldfire2/2m samples the level of mtab and latches the current value on mrdata[31:0]. if mtab is asserted, the transfer of the first longword terminates. if mtab is not asserted, the coldfire2/2m ignores the data and inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is asserted. figure 3-7. line read transfer (without wait states) c1 c2 c3 c4 c5 clk a1 maddr mrwb $3 msiz mtt mtm mtsb d1 d2 d3 d4 mrdata mie mtab mteab f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-17 clock 3 (c3) the coldfire2/2m holds the address and transfer control signals constant during c3. the selected device must increment the maddr [3:2] signals to reference the next longword to transfer, place the data on mrdata [31:0], assert mie , and assert mtab . at the end of c3, the coldfire2/2m samples the level of mtab and latches the current value on the mrdata[31:0] signals. if mtab is asserted, the transfer terminates. if mtab is not asserted at the end of c3, the coldfire2/2m ignores the latched data and inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is asserted. clock 4 (c4) this clock is identical to c3 except that once mtab is recognized as being asserted, the latched value corresponds to the third longword of data for the burst. clock 5 (c5) this clock is identical to c3 except that once mtab is recognized, the latched value corresponds to the fourth longword of data for the burst. this is the last clock cycle of the line read transaction. the selected device negates the mie and mtab signals in the first half of the next clk cycle. 3.3.5 line write transfers the line write transfer is similar to the line read transfer. this 16-byte burst write transfer is defined in figure 3-8 . f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-18 coldfire2/2m users manual motorola the assertion of mteab will abort a line write transfer. see section 3.8.1 access errors for more information on mteab. it is the slaves responsibility to increment and wrap the maddr [3:2] signals internally. figure 3-8. line write transfer flowchart master 1) drive address on maddr [31:0] 2) set mrwb to write (mrwb = 0) 3) set msiz [1:0] to line 4) set mtt [1:0] and mtm [2:0] to signal the appropriate access type 5) assert mtsb for one clk cycle 1) drive data onto mwdata [31:0] 2) assert mwdataoe 1) recognize the 1st transfer is done 1) drive data onto mwdata [31:0] 1) recognize the 2nd transfer is done 1) drive data onto mwdata [31:0] 2) recognize the 3rd transfer is done 1) drive data onto mwdata [31:0] 1) recognize the 4th transfer is done 2) negate mwdataoe 1) decode the address and select the appropriate slave device 2) insert necessary wait-states 1) capture the data from mwdata [31:0] 2) assert mtab for one clk cycle 1) capture the data from mwdata [31:0] 2) assert mtab for one clk cycle 1) capture the data from mwdata [31:0] 2) assert mtab for one clk cycle 1) capture the data from mwdata [31:0] 2) assert mtab for one clk cycle slave f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-19 see figure 3-9 for an example of a line write master bus transfer without wait states. clock 1 (c1) the line write cycle starts in c1. during the first half of c1, the coldfire2/2m places valid values on the master address bus ( maddr [31:0]) and transfer control signals. the master transfer type ( mtt [1:0]) and master transfer modifier ( mtm [2:0]) signals identify the specific access type. the master read/write ( mrwb ) signal is driven low for a write cycle, and the master size signals ( msiz [1:0]) indicate line size ($3). the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c1 to indicate the beginning of a bus cycle. clock 2 (c2) during the first half of c2, the coldfire2/2m negates mtsb , places the first longword of data on the master write data bus ( mwdata [31:0]), and asserts the master write data output enable ( mwdataoe ) signal. concurrently, the selected device asserts the master transfer acknowledge ( mtab ) signal if it is ready to latch the data. at the end of c2, the coldfire2/ 2m samples the level of mtab, and the selected device latches mwdata[31:0]. if mtab is asserted, the transfer of the first longword terminates. if mtab is not asserted, the coldfire2/2m inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is recognized asserted. figure 3-9. line write transfer (without wait states) c1 c2 c3 c4 c5 clk a1 maddr mrwb $3 msiz mtt mtm mtsb d1 d2 d3 d4 mwdata mwdataoe mtab mteab f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-20 coldfire2/2m users manual motorola clock 3 (c3) the coldfire2/2m holds the address and transfer control signals constant during c3, but drives mwdata [31:0] with the second longword of data. the selected device must increment maddr [3:2] to reference the second longword address, and assert mtab . at the end of c3, the coldfire2/2m samples the level of mtab, and the selected device latches mwdata[31:0]. if mtab is asserted, the transfer terminates. if mtab is not recognized asserted at the end of c3, the coldfire2/2m inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is recognized. clock 4 (c4) this clock is identical to c3 except that once mtab is asserted, the value corresponds to the third longword of data for the burst. clock 5 (c5) this clock is identical to c3 except that once mtab is asserted, the data value corresponds to the fourth longword of data for the burst. this is the last clock cycle of the line write transaction and the coldfire2/2m negates mwdataoe in the first half of the next clk cycle. see figure 3-10 for an example of a line write master bus transfer with wait states. note the cxw nomenclature that is used to define a wait state during the bus cycle, e.g., c2w. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-21 3.4 misaligned operands all coldfire2/2m data formats can be located in memory on any byte boundary. a byte operand is properly aligned at any address, a word operand is misaligned at an odd address, and a longword is misaligned at an address that is not evenly divisible by four. however, since operands can reside at any byte boundary, they can be misaligned. although the coldfire2/2m does not enforce any alignment restrictions for data operands, including pc relative data addressing, some performance degradation occurs when additional bus cycles are required for longword or word operands that are misaligned. for maximum performance, data items should be aligned on their natural boundaries. all instruction words and extension words must reside on word boundaries. attempting to prefetch an instruction word at an odd address causes an address error exception. see section 4.2.3 address error exception for more information. the coldfire2/2m misalignment unit converts misaligned operand accesses that are noncachable to a sequence of aligned accesses. figure 3-11 illustrates the transfer of a longword operand from an odd address requiring more than one bus cycle. in this example, the msiz [1:0] signals specify a byte transfer, and the byte offset is $1. the slave device supplies the byte and acknowledges the data transfer. when the coldfire2/2m starts the second cycle, the msiz[1:0] signals specify a word transfer with a byte offset of $2. the next two bytes are transferred during this cycle. the coldfire2/2m then initiates the third cycle, figure 3-10. line write transfer (with wait states) c1 c2w c2 c3w c3 c4w c4 c5w c5 clk a1 maddr mrwb $3 msiz mtt mtm mtsb d1 d2 d3 d4 mwdata mwdataoe mtab mteab f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-22 coldfire2/2m users manual motorola with the msiz[1:0] signals indicating a byte transfer. the byte offset is now $0; the port supplies the final byte and the operation is complete. this example is similar to the one illustrated in figure 3-12 except that the operand is word sized and the transfer requires only two bus cycles. figure 3-13 illustrates a functional timing diagram for a misaligned word read transfer. 31 24 23 16 15 8 7 0 transfer 1 - byte 3 - - transfer 2 - - byte 2 byte 1 transfer 3 byte 0 - - - figure 3-11. example of a misaligned longword transfer 31 24 23 16 15 8 7 0 transfer 1 - - - byte1 transfer 2 byte 0 - - - figure 3-12. example of a misaligned word transfer figure 3-13. misaligned word read transfer c1 c2 c1 c2 clk a1 a2 maddr mrwb $1 $1 msiz mtt mtm mtsb d1 d2 mrdata mie mtab mteab read 1 read 2 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-23 the combination of operand size and alignment determines the number of bus cycles required to perform a particular memory access. table 3-8 lists the number of bus cycles required for different operand sizes with all possible alignment conditions for read and write cycles. the table confirms that alignment significantly affects bus cycle throughput for noncachable accesses. for example, in figure 3-11 the misaligned longword operand took three bus cycles because the first byte offset = $1. if the byte offset = $0, then it would have taken one bus cycle. the coldfire2/2m system designer and programmer should account for these effects, particularly in time-critical applications. 3.5 invalid master bus cycles the coldfire2/2m starts a master bus transaction before it determines if the address hits in the cache or is mapped to the internal memories. as a result, the coldfire2/2m will assert the mkillb signal (late in that mtsb cycle) to indicate the current bus transaction resulted in a hit in the cache or internal memories. when this signal is asserted, master bus slaves must stop driving the master bus and must not return a mtab or mteab . the current master bus cycle must be inhibited. the general priority scheme is as follows: if (sram hits) sram supplies data to the processor else if (rom hits) rom supplies data to the processor else if (line fill buffer hits) line fill buffer supplies data to the processor else if (icache hits) icache supplies data to the processor else master bus cycle accesses to reference data from non-local memory clock 1 (c1) the read cycle starts in c1. during the first half of c1, the coldfire2/2m places valid values on the master address bus (maddr[31:0]) and transfer control signals. the coldfire2/2m asserts the master transfer start (mtsb) signal during c1 to indicate the beginning of a bus cycle. table 3-8. memory alignment cycles transfer size $0 ? $1 ? $2 ? $3 ? instruction 1 - - - byte operand 1111 word operand 1212 longword operand 1323 note: ?where the byte offset ( maddr [1:0]) equals this encoding. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-24 coldfire2/2m users manual motorola clock 2 (c2) during the first half of c2, the coldfire2/2m negates mtsb. because the mkillb signal was not asserted during the rising clock edge of c2, the selected device uses the mrwb and msiz signals to place the data on the master read data bus (mrdata[31:0]) and assert the master input enable (mie) signal. concurrently, the selected device asserts the master transfer acknowledge (mtab) signal. at the end of c2, the coldfire2/2m samples the level of mtab and captures the current value on mrdata[31:0]. if mtab is asserted, the transfer terminates. if mtab is not asserted, the coldfire2/2m ignores the data and inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is asserted. the selected device negates the mie and mtab signals in the first half of the next clk cycle. clock 3 (c3) the read cycle starts in c3. the coldfire2/2m asserts the mtsb signal, and then the mkillb signal. this signifies that the m-bus cycle must be inhibited; i.e. the access has hit in an internal-bus memory. clock 4 (c4) nop. clock 5 (c5) the read cycle starts in c5. the coldfire2/2m asserts the mtsb signal, and then the mkillb signal. this signifies that the m-bus cycle must be inhibited; i.e. the access has hit in an internal k-bus memory. clock 6 (c6) another read cycle starts in c6. the c1 description applies here. clock 7 (c7) during the first half of c7, the coldfire2/2m negates mtsb. because the mkillb signal was not asserted during the rising clock edge of c7, the selected device uses the mrwb and msiz signals to place the data on the master read data bus (mrdata[31:0]). the rest of the c2 description applies here. 3.6 pipeline stalls in an idealized environment for maximum performance, all coldfire2/2m references are mapped to integrated memory resources and complete in a single clock cycle. any memory reference that generates a master bus access stalls the processor pipeline since the internal transfer cannot be completed in a single clock cycle. this performance degradation factor can be expressed as: pipeline stall = master bus clock cycles - 1 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-25 where the stall is measured in clock cycles, and master bus clock cycles represents the entire master bus transfer time in clock cycles. in the examples shown in figure 3-14 , the read cycle with one wait state stalls the coldfire2/2m for two clock cycles, and the zero wait state write transfer produces a stall of one clock cycle. 3.7 interrupt acknowledge bus cycles when a peripheral device requires the services of the coldfire2/2m or is ready to send information that the coldfire2/2m requires, it can signal the coldfire2/2m to take an interrupt exception. the interrupt exception transfers control to a routine that responds appropriately. the peripheral device uses the active-low interrupt priority level signals ( iplb [2:0]) to signal an interrupt condition to the coldfire2/2m and to specify the priority level for the condition. refer to section 4 exception processing for a discussion on the iplb[2:0] levels. figure 3-14. example master bus wait state c1 c2w c2 c1 c2 clk a1 a2 maddr mrwb msiz mtt mtm mtsb d2 mwdata mwdataoe d1 mrdata mie mtab mteab read write f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-26 coldfire2/2m users manual motorola the status register of the coldfire2/2m, described in section 1.4.3.1 status register (sr) , contains an interrupt priority mask ( iplb [2:o]). the value in the interrupt mask is the highest priority level that the coldfire2/2m ignores. when an interrupt request has a priority higher than the value in the mask, the coldfire2/2m makes the request a pending interrupt. iplb [2:0] must maintain the interrupt request level until the coldfire2/2m acknowledges the interrupt to guarantee that the interrupt is recognized. the coldfire2/2m continuously samples iplb[2:0] on consecutive rising edges of clk . as a result, the iplb[2:0] signals are synchronous and must meet setup and hold times to clk. if the external iplb[2:0] signals are asynchronous, flip-flops should be used to synchronize them before they drive the iplb[2:0] signals on the coldfire2/2m. the coldfire2/2m takes an interrupt exception for a pending interrupt within one instruction boundary after processing any other pending exception with a higher priority. thus, the coldfire2/2m executes at least one instruction in an interrupt exception handler before recognizing another interrupt request. the following paragraphs describe the two kinds of interrupt acknowledge bus cycles that can be executed as part of interrupt exception processing. 3.7.1 interrupt acknowledge bus cycle (terminated normally) when the coldfire2/2m processes an interrupt exception, it performs an interrupt acknowledge bus cycle to obtain the vector number that contains the starting location of the interrupt exception handler. most interrupting devices have programmable vector registers that contain the interrupt vectors for the exception handlers they use. others may have fixed vector numbers. the values driven on maddr [31:0], mtt [1:0], and mtm [2:0] are dependent on the interrupt acknowledge mode. the interrupt acknowledge mode is statically determined by the connection of the 68k interrupt acknowledge mode enable ( iack_68k ) signal. the interrupt acknowledge bus cycle is a read transfer. if the coldfire2/2m is in the coldfire interrupt acknowledge mode (iack_68k negated), it differs from a normal read cycle in the following respects: 1. mtt [1:0]= $3 to indicate a acknowledge/cpu space bus cycle. 2. address signals maddr [31:5] are set to all ones ($3ffffff). the maddr[4:2] signals are set to the pending interrupt number, and the maddr[1:0] signals are driven low. 3. mtm [2:0] are set to the interrupt request level, the inverted values of iplb [2:0]. this will be nonzero for all interrupt acknowledge cycles. if the coldfire2/2m is in the 68k interrupt acknowledge mode (iack_68k asserted), it differs in the following respects: 1. mtt [1:0] = $0 to indicate a acknowledge/cpu space bus cycle. 2. address signals maddr [31:4] are set to all ones ($7ffffff). the maddr[3:1] signals are set to the pending interrupt number, and the maddr[0] signal is driven high. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-27 3. mtm [2:0] = $7 to indicate an interrupt acknowledge cycle. the responding device places the vector number on mrdata [31:24] during the interrupt acknowledge bus cycle, and the cycle is terminated normally with mtab . figure 3-15 illustrates a flowchart diagram for an interrupt acknowledge cycle terminated with mtab. see figure 3-16 for an example of a coldfire mode (iack_68k negated) interrupt acknowledge cycle. the interrupt shown occurs during a normal master bus read cycle with no lower-order interrupts pending. figure 3-15. interrupt acknowledge bus cycle flowchart 1) recognize pending interrupt 2) wait for instruction boundary 1) drive appropriate value on maddr [31:0] 2) set mrwb to read ($1) 3) set msiz [1:0] to byte ($1) 4) set mtt [1:0] to acknowledge 5) drive value on mtm [2:0] 6) assert mtsb for one clk cycle 1) latch vector number 2) recognize the transfer is done request interrupt on iplb [2:0] 1) decode the address and select the interrupting slave device. 2) insert necessary wait-states. 3) remove interrupt on iplb [2:0] 1) drive the vector number on mrdata [31:24] 2) assert mtab for one clk cycle. 3) assert mie for one clk cycle. master slave f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-28 coldfire2/2m users manual motorola clock 1 (c1) the interrupt acknowledge cycle starts in c1. during the first half of c1, the coldfire2/2m drives the master address bus ( maddr [31:5]) high, maddr[1:0] low, the mtt [1:0] signals to $3, and the maddr[4:2] and mtm [2:0] signals to the interrupt level. the master read/ write ( mrwb ) signal is driven high for a read cycle, and the master size signals ( msiz [1:0]) indicate transfer size (byte = $1). the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c1 to indicate the beginning of a bus cycle. figure 3-16. coldfire mode interrupt acknowledge bus cycle c1 c2 clk iplb stack maddr[31:5] int level stack maddr[4:2] stack maddr[1:0] mrwb $1 msiz $0 mtt int level $5 mtm mtsb mwdata mwdataoe vec mrdata[31:24] mrdata[23:0] mie mtab mteab int ack stack f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation motorola coldfire2/2m users manual 3-29 clock 2 (c2) during the first half of c2, the coldfire2/2m negates mtsb . the interrupting device uses the mrwb and msiz signals to place the vector number on the high-order byte of the master read data bus ( mrdata [31:24]) and assert the master input enable ( mie ) signal. concurrently, the interrupting device asserts the master transfer acknowledge ( mtab ) signal. at the end of c2, the coldfire2/2m samples the level of mtab and latches the current value on mrdata[31:24]. if mtab is asserted, the transfer terminates. if mtab is not asserted, the coldfire2/2m ignores the data and inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is asserted. the interrupting device negates the mie and mtab signals in the first half of the next clk cycle. see figure 3-17 for an example of a 68k mode (iack_68k asserted) interrupt acknowledge cycle. the interrupt is shown occurring during a normal master bus read cycle with no lower- order interrupts pending. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-30 coldfire2/2m users manual motorola clock 1 (c1) the interrupt acknowledge cycle starts in c1. during the first half of c1, the coldfire2/2m drives the master address bus ( maddr [31:4]), maddr[0], and the mtm [2:0] signals high, the mtt [1:0] signals low, and maddr[3:1] to the interrupt level. the master read/write ( mrwb ) signal is driven high for a read cycle, and the master size signals ( msiz [1:0]) indicate transfer size (byte = $1). the coldfire2/2m asserts the master transfer start ( mtsb ) signal during c1 to indicate the beginning of a bus cycle. figure 3-17. 68k mode interrupt acknowledge bus cycle c1 c2 clk iplb stack maddr[31:4] int level stack maddr[3:1] stack maddr[0] mrwb $1 msiz $0 mtt int level $5 mtm mtsb mwdata mwdataoe vec mrdata[31:24] mrdata[23:0] mie mtab mteab int ack stack f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-31 coldfire2/2m users manual motorola clock 2 (c2) during the first half of c2, the coldfire2/2m negates mtsb . the interrupting device uses the mrwb and msiz signals to place the vector number on the high-order byte of the master read data bus ( mrdata [31:24]) and assert the master input enable ( mie ) signal. concurrently, the interrupting device asserts the master transfer acknowledge ( mtab ) signal. at the end of c2, the coldfire2/2m samples the level of mtab and latches the current value on mrdata[31:24]. if mtab is asserted, the transfer terminates. if mtab is not asserted, the coldfire2/2m ignores the data and inserts wait states instead of terminating the transfer. the coldfire2/2m continues to sample mtab on successive rising edges of clk until it is asserted. the interrupting device negates the mie and mtab signals in the first half of the next clk cycle. 3.7.2 spurious interrupt acknowledge bus cycle when a device does not respond to an interrupt acknowledge bus cycle with mtab , the external logic typically returns the transfer error acknowledge signal ( mteab ). in this case, it is the responsibility of the external logic to return the spurious interrupt vector number 24 ($18) on mrdata [31:24] and assert mie . the vector number will be latched on the first rising edge of clk after mteab is asserted. because the spurious interrupt vector number is returned on mrdata[31:24] in the same manor as a normal interrupt acknowledge, mtab may be asserted instead of mteab. 3.8 master bus exception control cycles the coldfire2/2m bus architecture requires assertion of mtab from an external device to signal that a bus cycle is complete. mtab is not asserted in the following cases: ? the external device does not respond to a normal bus cycle. ? no interrupt vector is provided during an interrupt acknowledge cycle. ? various other application-dependent errors occur. external circuitry should assert mteab when no device asserts mtab within an appropriate period of time after the coldfire2/2m begins the bus cycle. this terminates the cycle and allows the coldfire2/2m to enter exception processing for the error condition. to properly control termination of a bus cycle for a access error, mtab and/or mteab must be asserted and negated for the same rising edge of clk . table 3-9 lists the control signal combinations and the resulting bus cycle terminations. note that the access error exception taken upon an mteab assertion cannot be masked and will occur for both reads and writes that result in mteab being asserted. access error terminations during burst cycles operate as described in section 3.3.4 line read transfer and section 3.3.5 line write transfers . f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-32 coldfire2/2m users manual motorola 3.8.1 access errors the system hardware can use the mteab signal to abort the current bus cycle when a fault is detected as shown in figure 3-18 . an access error is recognized during a bus cycle in which mteab is asserted. when the coldfire2/2m recognizes an access error condition, the access is terminated immediately. a line access that has mteab asserted for one of the four longword transfers aborts without completing the remaining transfers. when mteab is asserted to terminate a bus cycle, the coldfire2/2m can enter access error exception processing immediately following the bus cycle, or it can defer processing the exception in the following manner. the instruction prefetch mechanism requests instruction words from the instruction memory unit before it is ready to execute them. if an access error occurs on an instruction fetch, the coldfire2/2m does not take the exception until it attempts to use the instruction. should an intervening instruction cause a branch or should a task switch occur, the access error exception for the unused access does not occur. similarly, if an access error is detected on the second, third, or fourth longword transfer for a line read access, an access error exception is taken only if the execution unit is specifically requesting that longword. otherwise, the line is not placed in the cache, and the coldfire2/2m repeats the line access when another access references the line. if a misaligned operand spans two longwords in a line, an access error on either the first or second transfer for the line causes exception processing to begin immediately. an access error termination for any write accesses or for read accesses that reference data specifically requested by the execution unit causes the coldfire2/2m to begin exception processing immediately. refer to section 4exception processing for details of access error exception processing. when an access error terminates an access, the contents of the corresponding cache can be affected. for a cache line read to replace a valid instruction line, the cache line being filled is invalidated before the bus cycle begins and remains invalid if the replacement line access is terminated with an access error. note that if an access is made to a space that is masked, it simply becomes mapped to the next valid space; no access error is generated. see section 5.5, interactions between k- bus memories, for further information. when a write error occurs on a buffered write, an access error will be generated but will not be reported on the instruction that generated the write. access errors only occur because of the following: 1) mteab asserts 2) mmu error; i.e. trying to write in a write protected space table 3-9. mtab and mteab assertion results mtab mteab result dont care asserted access errorterminate and take access error exception this exception cannot be masked asserted negated normal cycle terminate and continue negated negated insert wait states f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-33 coldfire2/2m users manual motorola (rambar, rombar, acr1, acr0, and cacr have write protect bits) 3) accessing cpu space from the debug module while the processor in not halted or stopped 3.8.2 fault-on-fault halt a fault-on-fault error occurs when an access or address error occurs during the exception processing sequence - e.g., the coldfire2/2m attempts to stack several words containing information about the state of the machine while processing an access error exception. if an access error occurs during the stacking operation, the second error is considered a fault-on- fault error. a second access or address error that occurs during execution of an exception handler or later, does not cause a fault-on-fault condition. the coldfire2/2m indicates a fault-on-fault condition by continuously driving the processor status ( pst [3:0]) signals with an encoded value of $f until it is reset. only an external reset operation can restart a halted coldfire2/2m. see section 7.3.1 cpu halt for more information. figure 3-18. bus exception cycle c1 c2 c1 c2 clk a1 a2 maddr mrwb msiz mtt mtm mtsb d2 mwdata mwdataoe d1 mrdata mie mtab mteab error stack f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
master bus operation 3-34 coldfire2/2m users manual motorola 3.9 reset operation an external device asserts the mrstb signal to reset the coldfire2/2m. when power is applied to the system, external circuitry should assert mrstb for a minimum of six clock cycles after v dd and clk are within tolerance. the mrstb signal is not internally synchronized and must meet the specified setup and hold times to clk. if the external reset is asynchronous, two flip-flops should be used to synchronize the signal before it drives mrstb. figure 3-19 shows the general relationship between v dd , mrstb , and the bus signals during the power-on reset operation. resets during normal operation must follow the same requirements as those for power-on reset. 3.10 master bus arbitration the coldfire2/2m implementation is fully compatible with other, optional alternate masters. the m-bus supports arbitration; i.e. where the arbitration is performed in an external module to the coldfire2/2m. this arbitration module could function as a multiplexer-based master switch. the coldfire2/2m and the alternate masters generate master bus requests via their individual bus connections, with the arbitration logic selecting between the masters on a transfer-by-transfer basis. figure 3-19. initial power-on reset c1 vdd clk mrstb bus mtsb t >= 6 cycles t = 22cycles f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 4-1 section 4 exception processing exception processing is the activity performed by the coldfire2/2m in preparing to execute a special routine, called an exception handler, for any condition that causes an exception. exception processing does not include execution of the routine itself. this section describes the processing for each type of integer unit exception, exception priorities, the return from an exception, and bus fault recovery. also described are the formats of the exception stack frames. 4.1 exception processing overview exception processing is the transition from the normal processing of a program to the processing required for any special internal or external condition that preempts normal processing. external conditions that cause exceptions are interrupts from external devices, access errors, resets, and the assertion of the breakpoint signal. internal conditions that cause exceptions are address errors, instruction traps, illegal instructions, privilege violations, tracing, format errors, and interrupts from the debug module. exception processing uses an exception vector table and an exception stack frame. exception processing for the coldfire2/2m is streamlined for performance. differences from prior 68000 family processors are: ? a simplified exception vector table ? reduced relocation capabilities using the vector base register (vbr) ? a fixed-length exception stack frame format ? elimination of simultaneous multiple exception support ? use of a single self-aligning system stack exception processing is comprised of four major steps and can be defined as the time from the detection of the fault condition until the fetch of the first instruction of the exception handler has been initiated. figure 4-1 illustrates a general flowchart for the steps taken by the coldfire2/2m during exception processing. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing 4-2 coldfire2/2m users manual motorola the processing of an exception on the coldfire2/2m is composed of the following steps: 1. the coldfire2/2m makes an internal copy of the status register (sr) and then enters supervisor mode by setting the s-bit and disabling trace mode by clearing the t-bit in the sr. the occurrence of an interrupt exception also forces the master/interrupt bit, figure 4-1. exception processing flowchart entry exit exit enable supervisor determine vector number save contents to exception frame (see note) save internal copy of sr access error? calculate address of exception handler fetch first instruction of exception handler access/address error? begin execution of exception handler halted state y n y n (fault-on-fault) (fault-on-fault) note: this block varies for reset and interrupt excepti o mode & disable trace f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing motorola coldfire2/2m users manual 4-3 m-bit, in the sr to be cleared and the interrupt priority mask, i[2:0], in the sr to be set to the level of the current interrupt request 2. the coldfire2/2m determines the exception vector number. for all faults except interrupts, the coldfire2/2m performs this calculation based on the exception type. for interrupts, the coldfire2/2m performs an interrupt acknowledge (iack) bus cycle to obtain the vector number from a peripheral device. the iack cycle is mapped to a special acknowledge address space with the interrupt level encoded in the address. refer to section 3.7 interrupt acknowledge bus cycles . 3. the coldfire2/2m saves the current context by creating an exception stack frame on the system stack. the coldfire2/2m supports a single stack pointer (sp) in the a7 address register, i.e., there is no notion of separate supervisor- or user- stack pointers. as a result, the exception stack frame is created at a 0-modulo-4 address on the top of the current system stack. additionally, the coldfire2/2m uses a simplified fixed- length exception stack frame for all exceptions. the exception type determines whether the program counter placed in the exception stack frame defines the location of the faulting instruction or the address of the next instruction to be executed. 4. the coldfire2/2m calculates the address of the first instruction of the exception handler. by definition, the exception vector table is aligned on a 1 mbyte boundary. this instruction address is generated by fetching an exception vector from the table located at the address defined in the vector base register. the index into the exception table is calculated as (4 x vector_number). once the exception vector has been fetched, the contents of the vector determine the address of the first instruction of the desired exception handler. after the instruction fetch for the first opcode of the exception handler has been initiated, exception processing terminates and normal instruction processing continues in the exception handler. 4.1.1 exception stack frame definition a diagram of the exception stack frame is shown in figure 4-2 . the first long word of the exception stack frame, pointed to by sp, contains the 16-bit format/vector word (f/v) and the 16-bit status register, and the second long word contains the 32-bit program counter address. 31 28 27 26 25 18 17 16 15 0 format fs[3:2] vector[3:2] fs[1:0] status register program counter[31:0] pc figure 4-2. exception stack frame form f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing 4-4 coldfire2/2m users manual motorola field definitions: formatexception frame format this field is always written with a value of $4, $5, $6, or $7 by the coldfire2/2m indicating a two long-word frame format. the specific value depends on the value of the stack pointer (sp) before the exception occurred. 0100 = original sp[1:0] set to 00 0101 = original sp[1:0] set to 01 0110 = original sp[1:0] set to 10 0111 = original sp[1:0] set to 11 fsfault status this field is defined for access and address errors only and written as zeros for all other types of exceptions. 0000 = not address or access error 0001 = reserved 001x = reserved 0100 = error on instruction fetch 0101 = reserved 011x = reserved 1000 = error on operand write 1001 = attempted write to write-protected space 101x = reserved 1100 = error on operand read 1101 = reserved 111x = reserved vectorexception vector number this 8-bit vector number defines the exception vector number used to index into the exception vector table. for internal exceptions, this number is calculated by the coldfire2/ 2m, and for external interrupts, this number is read during the iack cycle as described in section 3.7 interrupt acknowledge bus cycles . refer to table 4-2 for exception vector assignments. pcprogram counter this field contains the 32-bit program counter at the time of the exception. the instruction at this address is the faulting instruction or the next instruction depending on the type of exception. 4.1.1.1 self-aligning stack. the coldfire2/2m aligns the stack to a longword boundary before the exception stack frame is placed on the stack. as a result, the difference between the sp before the exception and the sp at the start of the exception handler may be more than two longwords, but the location of the exception stack frame is always at the top of the stack. as seen in table 4-1 , the format field reflects the sp alignment and is used when an rte instruction executes to correctly restore the sp to its original value. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing motorola coldfire2/2m users manual 4-5 4.1.2 exception vectors the coldfire2/2m supports a 1024-byte vector table aligned on any 1 mbyte address boundary (see table 4-2 ). the table contains 256 exception vectors where the first 64 are defined by motorola and the remaining 192 are user-defined.the exception table order does not have any priority. it is an area in memory reserved for the storage of the first address used within the different exception handler routines. the reserved areas in the exception table are for future expansion and can be used as long as the application does not care if these areas become valid regions in future coldfire processors. table 4-1. stack pointer alignment original sp [1:0] @ time of exception, format field sp @ 1st instruction of exception handler 00 0100 (original sp) - 8 01 0101 (original sp) - 9 10 0110 (original sp) - 10 11 0111 (original sp) - 11 table 4-2. exception vector assignments vector number vector offset stacked program counter assignment hex decimal hex $0 0 $000 - reset initial stack pointer $1 1 $004 - reset initial program counter $2 2 $008 fault access error $3 3 $00c fault address error $4 4 $010 fault illegal instruction $5-$7 5-7 $014-$01c - reserved $8 8 $020 fault privilege violation $9 9 $024 next trace $a 10 $028 fault unimplemented line-a opcode $b 11 $02c fault unimplemented line-f opcode $c 12 $030 next debug interrupt $d 13 $034 - reserved $e 14 $038 fault format error $f 15 $03c next optional uninitialized interrupt $10-$17 16-23 $040-$05c - reserved $18 24 $060 next optional spurious interrupt $19-$1f 25-31 $064-$07c next optional level 1-7 autovectored interrupts $20-$2f 32-47 $080-$0bc next trap # 0-15 instructions $30-$3f 48-63 $0c0-$0fc - reserved $40-$ff 64-255 $100-$3fc next user-defined interrupts notes: 1. fault refers to the pc of the instruction that caused the exception 2. next refers to the pc of the next instruction that follows the instruction that caused the fault. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing 4-6 coldfire2/2m users manual motorola the coldfire2/2m inhibits sampling for interrupts during exception processing, including the first instruction of the exception handler. this allows the first instruction of any exception handler to effectively disable interrupts, if desired, by raising the interrupt mask level contained in the status register. normally, the end of an exception handler contains an rte instruction. when the coldfire2/ 2m executes the rte instruction, it examines the exception stack frame on top of the stack to determine if it is a valid frame. if the coldfire2/2m determines that it is a valid frame, the sr and pc fields are loaded from the exception stack frame and control is passed to the specified instruction address. if the frame is invalid, a format exception is taken. all exception vectors are located in the supervisor data space. since the vector base register (vbr) provides the base address of the exception vector table, the exception vector table can be located anywhere in memory; it can even be dynamically relocated for each task that an operating system executes. the vbr is reset to zero. refer to section 1.4.3.4 vector base register (vbr) . 4.1.3 multiple exceptions within a coldfire2/2m system, more than one exception can occur at the same time. when this occurs, only the exception with the highest priority will be processed. exceptions can be divided into the four basic groups identified in table 4-3 . these groups are defined by specific characteristics and the priority with which they are handled. as far as the coldfire2/2m is concerned, the interrupt exception will never appear to occur at the same time as another exception. interrupt exceptions will remain pending until the other exception is processed. refer to section 4.2.6 trace exception for an example of simultaneous exceptions. table 4-3. exception priority groups group priorit y exception and relative priority characteristics 0.0 reset the coldfire2/2m aborts all processing (instruction or exception) and does not save old context. 1.0 1.1 1.2 address error instruction access error data access error the coldfire2/2m suspends processing and saves the context. 2.0 2.1 2.2 2.3 a-line unimplemented f-line unimplemented illegal instruction privilege violation exception processing begins before the instruction is executed. 3.0 trap instruction format error exception processing is part of the instruction execution and begins after instruction execution. 4.0 4.1 4.2 trace debug interrupt interrupt exception processing begins when the current instruction is completed. notes: 1. 0.0 is the highest priority. 2. 4.2 is the lowest priority. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing motorola coldfire2/2m users manual 4-7 4.1.4 fault-on-fault halt if the coldfire2/2m encounters any type of fault during the exception processing of another fault, it immediately halts execution with the catastrophic fault-on-fault condition. the coldfire2/2m indicates a fault-on-fault condition by continuously driving the processor status (pst[3:0]) signals with an encoded value of $f until it is reset. only an external reset operation can restart a halted coldfire2/2m. see section 7.3.1 cpu halt for more information. 4.2 exceptions the following paragraphs describe the external interrupt exceptions and the different types of exceptions generated internally by the integer unit. the following exceptions are dis- cussed: ? reset ? access error ? address error ? illegal instruction ? privilege violation ? trace ? unimplemented opcode ? debug interrupt ? format error ? trap instruction ? external interrupt 4.2.1 reset exception asserting the reset input signal to the coldfire2/2m causes a reset exception, vector number $0. the reset exception has the highest priority of any exception; it provides for system initialization. reset also aborts any processing in progress when the reset input is recognized, and the aborted processing cannot be recovered. the reset exception places the coldfire2/2m in the supervisor mode by setting the s-bit and disables tracing by clearing the t-bit in the sr. this exception also clears the m-bit and sets the coldfire2/2ms interrupt priority mask in the sr to the highest level (level 7). next, the vbr, cacr, acrs, rambar0, and rombar0 are initialized to their reset value. this will disable the cache, sram, and optionally the rom (see section 5.3.2 rom programming model .) once the coldfire2/2m is granted the bus, and it does not detect any other alternate masters taking the bus, the core then performs two longword read bus cycles. because the vbr is reset to zero, the first longword is always loaded into the stack pointer from address zero, and the second longword is always loaded into the program counter from address four. after the initial instruction is fetched from memory, program execution begins at the address f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing 4-8 coldfire2/2m users manual motorola in the pc. if an access error or address error occurs before the first instruction begins execution, the coldfire2/2m enters the fault-on-fault halted state. refer to section 3.9 reset operation for more information on initiating a reset exception. 4.2.2 access error exception an access error exception, vector number $2, occurs when a bus cycle terminates with an error condition such as the assertion of the master transfer error acknowledge (mteab) signal. refer to section 3.8 master bus exception control cycles . the exact response to an access error is dependent on the type of memory reference being performed. for an instruction fetch, the coldfire2/2m postpones the reporting of an error until the faulted reference is needed by an instruction to be executed. thus, faults which occur during instruction prefetches which are then followed by a change of instruction flow will not generate an exception. when the coldfire2/2m attempts to execute an instruction with a faulted opword and/or extension words, the access error will be signaled and the instruction aborted. for this type of exception, the programming model has not been altered by the instruction generating the access error. if the access error occurs on an operand read, the coldfire2/2m immediately aborts the current instructions execution and initiates exception processing. in this situation, any address register updates due to the auto-addressing modes, {e.g., (an)+,-(an)}, will already have been performed. thus, the programming model contains the updated an value. in addition, if an access error occurs during the execution of a movem instruction loading registers from memory, any registers already updated before the fault occurs will contain the operands from memory. the coldfire2/2m uses an imprecise reporting mechanism for access errors on write operations. since the actual write cycle may be decoupled from the coldfire2/2ms execution of the instruction requesting the write, the signaling of an access error appears to be decoupled from the instruction which generated the write. accordingly, the pc contained in the exception stack frame merely represents the location in the program when the access error was signaled, not when the offending instruction was executed. all programming model updates associated with the write instruction are completed. the nop instruction can be used for purposes of collecting access errors for writes. this instruction delays its execution until all previous operations, including all pending write operations to internal memory resources, are complete. noncachable writes to the master bus may be buffered but will be completed before the execution of a nop instruction. the nop instruction waits until the pipeline is cleared out, including buffered writes before executing. generally, buffering can provide higher performance, although issues concerning recovery from physical write-errors may become more difficult to resolve. 4.2.3 address error exception any attempted execution transferring control to an odd instruction address (i.e., if bit 0 of the target address is set) results in an address error exception, vector number $3. for conditional branch instructions, the exception is generated regardless of the taken/not-taken resolution of the branch condition. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing motorola coldfire2/2m users manual 4-9 the coldfire2/2m has a misalignment unit which will generate a series of aligned bus cycles to access data requested from a misaligned address. as a result, misaligned operand fetches will not cause an address error exception. any attempted use of a word-sized index register (xn.w) or an invalid scale factor on an indexed effective addressing mode generates an address error. the setting of the extension word valid (ev) bit on an indexed addressing mode instruction will also generate an address error. refer to the coldfire programmers reference manual rev. 1.0 (mcf5200prm/ad). 4.2.4 illegal instruction exception the attempted execution of the $0000 and the $4afc opwords generates an illegal instruction exception, vector number $4. the coldfire2/2m does not provide illegal instruction detection on the extension words on any instruction, including movec. if any other non-supported opcode is executed, the resulting operation is undefined and the coldfire2/2ms behavior is unpredictable. 4.2.5 privilege violation exception the attempted execution of a supervisor mode instruction while in user mode generates a privilege violation exception, vector number $8. see the coldfire programmers reference manual rev. 1.0 for lists of supervisor- and user- mode instructions. 4.2.6 trace exception to aid in program development, the coldfire2/2m provides an instruction-by-instruction tracing capability. while in trace mode, indicated by the setting of the t-bit in the status register (sr[15] = 1), the completion of an instruction execution triggers a trace exception, vector number $9. this functionality allows a debugger to monitor program execution. the single exception to this definition is the stop instruction. a stop instruction that begins execution in trace mode (t-bit in sr set) or enters trace mode because of the stop instruction execution (bit 15 of the stop operand is set) forces a trace exception after it loads the sr. upon return from the trace exception handler, execution continues with the instruction following the stop instruction, and the coldfire2/2m never enters the stopped condition. a stop instruction will enter the stopped state only if the coldfire2/2m is not in trace mode before the stop instruction is executed and the stop instruction does not place the coldfire2/2m in the trace mode. since thecoldfire2/2m does not support hardware stacking of multiple exceptions, it is the responsibility of the operating system to check for trace mode after processing other exception types. as an example, consider the execution of a trap instruction while in trace mode. the coldfire2/2m will initiate the trap exception and then pass control to the corresponding exception handler, clearing the trace bit (t-bit) in the sr. if the system requires that a trace exception be processed, it is the responsibility of the trap exception handler to check for this condition, sr[15] in the exception stack frame, and pass control to the trace exception handler before returning from the original exception. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing 4-10 coldfire2/2m users manual motorola a trace exception does not occur immediately after the execution (move to sr, rte) of the instruction that puts the coldfire2/2m in trace mode. the first trace exception occurs after the subsequent instruction. 4.2.7 unimplemented opcode exception the attempted execution of line a opcodes not used for mac instructions and line f opcodes not used for the debug instructions, including $ffff, generates their unique exception types, vector numbers $a and $b respectively. refer to appendix b new mac instructions for a list of the mac opcodes and the coldfire programmers reference manual rev. 1.0 for a list of the debug instruction opcodes. 4.2.8 debug interrupt the debug module is the only internal interrupt source for the coldfire2/2m. this type of program interrupt, vector number $c, is discussed in detail in section 7.4 real-time debug support . this exception is generated in response to a hardware breakpoint register trigger. the coldfire2/2m does not generate an iack cycle, but rather calculates the vector number internally. 4.2.9 rte & format error exceptions when an rte instruction is executed, the coldfire2/2m first examines the 4-bit format field in the exception stack frame on the stack to validate the frame type. any attempted execution of an rte when the format is not equal to $4, $5, $6, or $7 generates a format error, vector number $e. the exception stack frame for the format error is created without disturbing the original exception stack frame. the pc field in the new exception stack frame will point to the rte instruction. the selection of the format value provides some limited debug support for porting code from 68000 applications. on 680x0 family processors, the status register (sr) was located at the top of the stack. on those processors, bit[30] of the longword addressed by the system stack pointer (sp) is typically zero. thus, if an rte is attempted with a 680x0-type exception stack frame, the coldfire2/2m will generate a format exception. if the format field defines a valid type, the coldfire2/2m: (1) reloads the sr operand, (2) fetches the second long word operand (pc), (3) adjusts the stack pointer by adding the format value to the auto-incremented address after the fetch of the first long word, and then (4) transfers control to the instruction address defined by the pc (fetched in step 2.) 4.2.10 trap instruction exceptions the trap instruction always forces an exception as part of its execution and is useful for implementing system calls. the instruction adds the immediate operand (vector) of the instruction to 32 to obtain the vector number. the range of vector values is 0 - 15, which provides 16 vectors numbered $20 - $2f. refer to the coldfire programmers reference manual rev. 1.0 (mcf5200prm/ad) for more information on the trap instruction. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing motorola coldfire2/2m users manual 4-11 4.2.11 interrupt exception when a peripheral device requires the services of the coldfire2/2m or is ready to send information that the coldfire2/2m requires, it can signal the coldfire2/2m to take an interrupt exception. seven levels of interrupt priorities are provided, numbered 1C7. devices can be chained externally within interrupt priority levels, allowing an unlimited number of peripheral devices to interrupt the coldfire2/2m. the status register contains a 3-bit mask indicating the current interrupt priority, and interrupts are inhibited for all priority levels less than or equal to the current priority (see section 1.4.3.1 status register (sr) .) an interrupt request is made to the coldfire2/2m by encoding the interrupt request levels 1C7 on the three interrupt request level (iplb[2:0]) signals; all signals high indicate no interrupt request. table 4-4 shows the relationship between the actual requested interrupt and the state of the iplb[2:0] signals as well as the interrupt mask levels required for recognition of the requested level. interrupt requests arriving at the coldfire2/2m do not force immediate exception processing, but the requests are made pending. pending interrupts are detected between instruction executions. if the priority of the pending interrupt is lower than or equal to the current coldfire2/2m priority, execution continues with the next instruction, and the requesting interrupt is postponed until the priority of the pending interrupt becomes greater than the current coldfire2/2m priority. if the priority of the pending interrupt is greater than the current coldfire2/2m priority, the exception processing sequence for the requesting interrupt is started. a copy of the status register is saved internally; the privilege mode is set to supervisor mode; tracing is suppressed; and the coldfire2/2m priority level is set to the level of the interrupt being acknowledged. the coldfire2/2m fetches the vector number from the interrupting device by executing an interrupt acknowledge cycle, which displays the level number of the interrupt being acknowledged on the address bus (see section 3.7 interrupt acknowledge bus cycles ). the coldfire2/2m then proceeds with the usual exception processing, saving the exception stack frame on the supervisor stack. the saved value of the program counter is the address of the instruction that would have been executed had the interrupt not been taken. the appropriate interrupt vector is fetched and loaded into the program counter, and normal instruction execution commences in the interrupt handling routine. table 4-4. interrupt levels and mask values requested interrupt level control line status interrupt mask level required for recognition iplb[2] iplb[1] iplb[0] 0 high high high no request 1 high high low 0 2 high low high 0-1 3 high low low 0-2 4 low high high 0-3 5 low high low 0-4 6 low low high 0-5 7 low low low 0-7 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing 4-12 coldfire2/2m users manual motorola interrupt requests should be maintained on the iplb[2:0] signals until the conclusion of the interrupt acknowledge (iack) cycle from the processor to guarantee that the interrupt will be recognized. the interrupt request can be maintained without being recognized again, until the interrupt priority level in the sr is lowered below the currently requested level. this is usually the result of the rte instruction at the end of the interrupt exception handler. once the conclusion of an interrupt acknowledge (iack) cycle occurs, another interrupt may be asserted on the iplb lines.the processor will start fetching the interrupt handler code for the acknowledged interrupt. in the handler, the processor looks at the interrupt mask level and determines if the second interrupt is masked or allowed.if the second interrupt is not masked, the processor will start another iack cycle for the second interrupt. it will go to the second interrupt handler and sample iplb inputs, compare mask, and execute this handler if a higher level interrupt is not pending. once the second interrupt has completed the return from exception (rte) in the handler, the processor will return to the first interrupt handler, where the iplb lines are sampled and compared to the interrupt mask. if no interrupts have higher priority, it will execute this handler and return (rte) to the place in the code where the first interrupt was allowed to be taken. thus, interrupts can be nested, and higher interrupts given priority by the interrupt mask register once they are executing code within the handlers. 4.2.11.1 level 7 interrupts. level 7 interrupts are handled differently than interrupt levels one through six. a level 7 interrupt is a nonmaskable interrupt; therefore, a 7 in the interrupt mask does not disable a level 7 interrupt. level 7 interrupts are edge-triggered by a transition from a lower priority request to the level 7 request, as opposed to interrupt levels one through six, which are level sensitive. therefore, if the interrupt priority level (iplb[2:0]) signals remain at level 7, the coldfire2/ 2m will only recognize one level 7 interrupt since only one transition from a lower level request to a level 7 request occurred. for the coldfire2/2m to recognize a level 7 interrupt followed by another level 7 interrupt, one of the two following sequences must occur: 1. the interrupt request level on the iplb[2:0] signals changes from a lower request level to level 7 and remains at level 7 until the interrupt acknowledge bus cycle begins. later, the interrupt request level returns to a lower interrupt request level and then back to level 7, causing a second transition on the iplb[2:0] signals. 2. the interrupt request level on the iplb[2:0] signals changes from a lower request level to level 7 and remains at level 7. if the interrupt handling routine for the level 7 interrupt lowers the interrupt mask level, a second level 7 interrupt will be recognized even though no transition has occurred on the interrupt control pins. after the level 7 interrupt handling routine completes, the coldfire2/2m will compare the interrupt mask level to the interrupt request level on the iplb[2:0] signals. since the interrupt mask level will be lower than the requested level, the interrupt mask will be set back to level 7. the level 7 request on the iplb[2:0] signals must be held until the second interrupt acknowledge bus cycle has begun to ensure that the interrupt is recognized. 4.2.11.2 spurious, autovectored, and uninitialized interrupts. if when the external logic indicates an access error during the interrupt acknowledge cycle, the f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
exception processing 4-13 coldfire2/2m users manual motorola interrupt is considered spurious (refer to section 3.7.2 spurious interrupt acknowledge bus cycle .) it is the responsibility of the external logic to return the spurious interrupt vector number, $18. external hardware may also be designed to generate autovector interrupts, $64 - $7c, for each interrupt level. many m68000 family peripherals use programmable interrupt vector numbers as part of the interrupt-acknowledge operation for the system. if this vector number is not initialized after reset and the peripheral must acknowledge an interrupt request, the peripheral usually returns the vector number for the uninitialized interrupt vector, $f. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 5-1 section 5 integrated memories the coldfire2/2m has dedicated buses to support three integrated memories: instruction cache, ram, and rom. 5.1 instruction cache the coldfire2/2m has a dedicated bus to support an integrated instruction cache with the following features: ? 0 - 32 kbyte direct-mapped instruction cache ? instruction cache byte size programmed with ich_sz[2:0] static signals ? single-cycle access on cache hits ? physically located on processors high-speed local bus ? non-blocking design to maximize performance ? configurable cache-miss fetch algorithm the cache module services instruction-fetch requests from the coldfire2/2m by either returning matching 32-bit cache entries in a single clock, or by initiating memory requests to service accesses that miss in the cache. the instruction cache size is specified via the ich_sz[2:0] pins which are static signals that need to stay valid for all operation. only the coldfire2/2m can access the cache. a fetch is defined as a read from user or supervisor code space only. 5.1.1 instruction cache hardware organization the instruction cache is an optional direct-mapped single-cycle memory, organized as 32 (512 byte) to 2k (32 kbyte) lines, each containing 4 longwords or 16 bytes. the data array is organized in longwords, 4 bytes per entry. the tag array is organized in lines, one entry per four longwords or line. the cache size is determined by the encoding of the ich_sz[2:0] inputs as shown in table 5-1 . thus the memory storage consists of a n-entry tag array (where n corresponds to the number of lines in the data array) containing addresses and a valid bit, and the data array containing m bytes of instruction data (where m= 512,1k, 2k, 4k, 8k, 16k, or 32k), organized as m/4 x 32 bits. the two memory arrays are accessed in parallel: bits [x:4] of the instruction fetch address providing the index into the tag array, and bits [x:2] addressing the data array (where x ranges from 8 to 14 for 512-byte to 8k-byte i-cache size, see table 5-1 ).the tag array outputs the address mapped to the given cache location along with the valid bit for the line. this address field is compared to bits [31:y] of the instruction fetch address (where y = x + f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-2 coldfire2/2m users manual motorola 1 for a given cache size) from the local bus to determine if a cache hit in the memory array has occurred. if the desired address is mapped into the cache memory, the output of the data array is driven onto the processors local data bus completing the access in a single cycle. the tag array maintains a single valid bit per line entry. accordingly, only entire 16-byte lines are loaded into the instruction cache. the instruction cache also contains a 16-byte fill buffer which provides temporary storage for the last line fetched in response to a cache miss. with each instruction fetch, the contents of the line fill buffer are examined. thus, each instruction fetch address examines both the tag memory array and the line fill buffer to see if the desired address is mapped into either hardware resource, with the line fill buffer having priority over the instruction cache. a cache hit in either the memory array or the line fill buffer is serviced in a single cycle. since the line fill buffer maintains valid bits on a longword basis, hits in the buffer can be serviced immediately without waiting for the entire line to be fetched. if the referenced address is not contained in the memory array nor the line fill buffer, the instruction cache initiates the required external fetch operation. in most situations, this is a 16-byte line-sized burst reference. an external bus cycle is always started simultaneously with the fetch cycle to the instruction cache. if a hit occurs in the instruction cache, the mkillb signal is asserted late in the cycle to kill the external bus cycle. thus an asserted mtsb and mkillb direct the external bus controller to ignore the mtsb. if no hit occurred in the instruction cache (or other k-bus memory), the m-bus cycle uses the normal number of clock cycles beginning with the mtsb issued for the instruction fetch. see section 5.5, interactions between k-bus memories for more details. the hardware implementation is a non-blocking design, meaning the processors local bus is released after the initial access of a miss. thus, the cache, ram, or rom module can service subsequent requests while the remainder of the line is being fetched and loaded into the fill buffer. 5.1.2 instruction cache operation the instruction cache is physically connected to the processors local bus allowing it to service all instruction fetches from the coldfire cpu and certain memory fetches initiated by the debug module. typically, the debug moduless memory references appear as supervisor data accesses, but the unit may be programmed to generate user mode accesses and/or instruction fetches. any instruction fetch access is processed by the instruction cache in the normal manner. 5.1.3 instruction cache signal description the following signals interface the coldfire2/2m to an integrated instruction cache. the cache is comprised of two compiled rams: tag and data. figure 5-1 illustrates an 8 kbyte configuration. all coldfire2/2m signals are unidirectional and synchronous. instruction cache outputs are registered and the instruction cache data is latched into the coldfire2/2m on the falling edge of the clock. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-3 5.1.3.1 instruction cache address bus (ich_addr[14:2]). these registered output signals provide the address of the current bus cycle (i.e. fetch cycle) to the integrated cache rams. ich_addr is only updated on fetch cycles (i.e. ich_addr does not get updated on ram or rom hits). this bus should be connected to the address bus (a) of the two compiled cache rams. ich_addr[n:m] is being provided via maddr[n:m]. 5.1.3.2 instruction cache data chip-select (ichd_csb). this active-low, output signal indicates the cache data ram is currently selected to perform a data transfer with the coldfire2/2m. this bus should be connected to the chip-select (csb) signal of the compiled cache data ram. 5.1.3.3 instruction cache data input bus (ichd_di[31:0]). these output signals provide the write data path between the coldfire2/2m and the cache data ram. the data bus is 32-bits wide and should be connected to the data inputs (dbi) of the compiled cache data ram. 5.1.3.4 instruction cache data output bus (ichd_do[31:0]). these input signals provide the read data path between the cache data ram and the coldfire2/2m. the data bus is 32-bits wide and should be connected to the data outputs (dbo) of the compiled cache data ram. figure 5-1. example of 8 kbyte instruction cache interface diagram ichd_csb ichd_di[31:0] ichd_do[31:0] ichd_st ichd_rwb ich_addr[14:2] icht_csb icht_di[31:8] icht_do[31:8] icht_st icht_rwb csb dbi[19:0] dbo[19:0] st rwb a[8:0] csb dbi[31:0] dbo[31:0] st rwb a[10:0] tag ram cpu core data ram size ich_sz[2:0] [31:13,8] [31:13,8] [12:2] [31:0] [31:0] [2:0] [12:4] [12:2] gnd [12:9] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-4 coldfire2/2m users manual motorola 5.1.3.5 instruction cache data strobe (ichd_st). this output signal initiates a read or write cycle to the cache data ram on a low-to-high transition. this signal should be connected to the strobe input (st) signal of the compiled cache data ram. 5.1.3.6 instruction cache data read/write (ichd_rwb). this output signal indicates the direction of the data transfer to the cache data ram. a high level indicates a read cycle and a low level indicates a write cycle. it should be connected to the read/write (rwb) signal of the compiled cache data ram. 5.1.3.7 instruction cache size (ich_sz[2:0]). these static inputs specify the size of the compiled cache rams connected to the coldfire2/2m. table 5-1 lists the possible cache configurations. ich_sz[2:0] does not affect the cacr in any way; thus a movec instruction will write the cacr regardless of the ich_sz specification (which is contrary to the ram_sz and rom_sz effect during rambar and rombar loading). 5.1.3.8 instruction cache tag chip-select (icht_csb). this active-low output signal indicates the cache tag ram is currently selected to perform a data transfer with the coldfire2/2m. this signal should be connected to the chip-select (csb) signal of the compiled cache tag ram. 5.1.3.9 instruction cache tag input bus (icht_di[31:8]). these output signals provide the write data path between the coldfire2/2m and the cache tag ram. the data bus depends upon the size of the cache; see table 5-1 . bit eight is always the valid bit and is always used as seen in the cache configuration shown in table 5-2 . this bus should be connected to the data inputs (dbi) of the compiled cache tag ram. functionally, maddr[31:9] is written onto icht_di[31:9] and icht_di[8] is written with the valid state of the entry. table 5-1. cache configuration encoding icache sz (bytes) ich_sz[2:0] tag array size tag array address data array size data array address none 000 - - - - 512 001 32x24 ich_addr[8:4] 128x32 ich_addr[8:2] 1 k 010 64x23 ich_addr[9:4] 256x32 ich_addr[9:2] 2 k 011 128x22 ich_addr[10:4] 512x32 ich_addr[10:2] 4 k 100 256x21 ich_addr[11:4] 1kx32 ich_addr[11:2] 8 k 101 512x20 ich_addr[12:4] 2kx32 ich_addr[12:2] 16k? 110 1kx19 ich_addr[13:4] 4kx32 ich_addr[13:2] 32k? 111 2kx18 ich_addr[14:4] 8kx32 ich_addr[14:2] note: ?hpf65 coldfire2 hard macro may require a reduced operating frequency for 16k and 32k sized icache. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-5 5.1.3.10 instruction cache tag output bus (icht_do[31:8]). these input signals provide the read data path between the cache tag ram and the coldfire2/2m. the data bus size depends upon the size of the cache; see table 5-1 . bit eight is always the valid bit and is always used regardless of the cache configuration as shown in table 5-2 . this bus should be connected to the data outputs (dbo) of the compiled cache tag ram. unused signals must be tied low. 5.1.3.11 instruction cache tag strobe (icht_st). this output signal initiates a read or write cycle to the cache tag ram on a low-to-high transition. this signal should be connected to the strobe input (st) signal of the compiled cache tag ram. 5.1.3.12 instruction cache tag read/write (icht_rwb). this output signal indicates the direction of the data transfer to the cache tag ram. a high level indicates a read cycle and a low level indicates a write cycle. it should be connected to the read/write (rwb) signal of the compiled cache tag ram. 5.1.4 interaction with other modules since the instruction cache, high-speed rom and ram modules are connected to the processors local data bus, certain user-defined configurations may result in simultaneous instruction fetch processing. if the referenced address is mapped into the rom or ram module, that module will service the request in a single cycle. in this case, data accessed from the instruction cache is simply discarded, and no external memory references are generated. the ram module has higher priority over the rom module. if the address is not mapped into the ram or rom space, then the request is handled by the instruction cache in the normal fashion. 5.1.5 memory reference attributes for every memory reference generated by the processor or the debug module, a set of effective attributes is determined based on the address and the access control registers (acr0, acr1). this set of attributes includes the cacheable/non-cacheable definition, the precise/imprecise handling of operand writes and the write-protect capability. in particular, each address is compared to the values programmed in the access control registers. if the address matches one of the acr values, the access attributes from that table 5-2. valid tag ram data signals cache size (bytes) valid data bits cpu data placed on valid data bits 512 icht_dx[31:8] {maddr[31:9], valid} 1 k icht_dx[31:10,8] maddr[31:10], valid} 2 k icht_dx[31:11,8] maddr[31:11], valid} 4 k icht_dx[31:12,8] maddr[31:12], valid} 8 k icht_dx[31:13,8] maddr[31:13], valid} 16k icht_dx[31:14,8] maddr[31:14], valid} 32k icht_dx[31:15,8] maddr[31:15], valid} f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-6 coldfire2/2m users manual motorola acr are applied to the reference. if the address does not match either acr, then the default value defined in the cache control register (cacr) is used. the specific algorithm is: if (address = acr0_address including mask) effective attributes = acr0 attributes else if (address = acr1_address including mask) effective attributes = acr1 attributes else effective attributes = cacr default attributes 5.1.6 cache coherency and invalidation the instruction cache does not monitor processor data references for accesses to cached instructions. therefore it is necessary for software to maintain cache coherence by invalidating the appropriate cache entries after modifying code segments. the cache invalidation can be performed in two ways: 1. the assertion of bit 24 in the cache control register, via a cpu space write, forces the entire instruction cache to be marked as invalid. the invalidation operation requires n (n = # of lines) cycles since the cache sequences through the entire tag array clearing a single location (the valid location) each cycle. any subsequent instruction fetch accesses are postponed until the invalidation sequence is complete; i.e. the processor can continue running as long as it does not try and fetch from the instruction cache. if the instruction cache is accessed, processing halts and waits for the invalidation to complete. the cacr can be loaded to not turn on the instruction cache to let processing continue unblocked. 2. the privileged cpushl instruction can be used to invalidate a single cache line. when this instruction is executed, the cache entry defined by bits [x:4] of the source address register in invalidated, provided bit 28 of the cacr is cleared. these invalidation operations may be initiated from the processor or the debug module. 5.1.7 reset a hardware reset clears the cacr disabling the instruction cache. the contents of the tag array are not affected by the reset. accordingly, the system start-up code must explicitly perform a cache invalidation by setting cacr[24] before the cache may be enabled. 5.1.8 line fill buffer and cache miss fetch algorithm as discussed in section 5.1.1, the instruction cache hardware includes a 16-byte line fill buffer for providing temporary storage for the last fetched instruction. with the cache enabled as defined by cacr[31], a cacheable instruction fetch that misses in both the tag memory and the line-fill buffer generates an external fetch. the size of the external fetch is determined by the value contained in the 2-bit clnf field of the cacr, and the miss address. table 5-3 shows the relationship between the clnf bits, the miss address and the size of the external fetch: f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-7 depending on the runtime characteristics of the application and the memory response speed, overall performance may be increased by programming the clnf bits to values {00, 01}. for all cases of a line-sized fetch, the critical longword defined by bits [3:2] of the miss address is accessed first, followed by the remaining three longwords which are accessed by incrementing the longword address in a modulo-16 fashion as shown below: if miss address[3:2] = 00 fetch sequence = {$0, $4, $8, $c} if miss address[3:2] = 01 fetch sequence = {$4, $8, $c, $0} if miss address[3:2] = 10 fetch sequence = {$8, $c, $0, $4} if miss address[3:2] = 11 fetch sequence = {$c, $0, $4, $8} once an external fetch has been initiated and the data loaded into the line-fill buffer, the instruction cache maintains a special most-recently-used indicator which tracks the contents of the fill buffer versus its corresponding cache location. at the time of the miss, the hardware indicator is set, marking the fill buffer as most recently used. if a subsequent access to the cache location defined by bits [x:4] of the fill buffer address occurs, the data in the cache memory array is now most-recently-used, so the hardware indicator is cleared. in all cases, the indicator defines whether the contents of the line-fill buffer or the memory data array are most-recently-used. at the time of the next cache miss, the contents of the line-fill buffer are written into the memory array if the entire line is present, and the fill buffer data is still most-recently-used compared to the memory array. only a complete line can be transferred to the icache array (the array only has one valid bit while the line-fill buffer has a valid bit per longword). this transfer can only occur if the icache is not locked or frozen. this write takes four cycles with ich_addr[3:2] incrementing each cycle. generally, when a fetch misses the cache, the previously fetched instruction line in the line-fill buffer is written to the cache while the master bus is running a fetch cycle. the fill buffer can also be utilized as temporary storage for line-sized bursts of non- cacheable references under control of cacr [10]. with this bit set, a non-cacheable instruction fetch is processed as defined by table 5-3 . for this condition, the fill buffer is loaded and subsequent references can hit in the buffer, but the data is never loaded into the memory array. table 5-3. initial fetch size based on miss address & clnf clnf[1:0] miss address [3:2] (longword address bits) 00 01 10 11 00 line line line longword 01 line line longword longword 1x line line line line f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-8 coldfire2/2m users manual motorola line fills begin with the longword containing the requested instruction and wrap around, as needed, to complete the full 16-byte line request. noncachable accesses can still result in line requests to the master bus because they are buffered in the cache line-fill buffer, but they are not copied into the cache. requests which result in a longword fetch will not be written to the cache. the relationship between cacr bits 31 and 10, and the type of instruction fetch is shown below. 5.1.9 instruction cache programming model the operation of the instruction cache and local bus controller are defined by three supervisor registers: the cache control register (cacr) and two access control registers (acr0, acr1). all three registers may be written from the processor using the privileged movec instruction. additionally, the registers may be accessed from the debug module. from the debug module, the registers may be read or written. in all cases, undefined bits in a register are reserved. these bits should be written as zeroes, and return zeroes when read from the debug module. 5.1.10 cacheability cacheability of instruction accesses is controlled by either the access control registers acr0 and acr1, for accesses matching the address ranges defined by these registers, or by the default cache mode bits in the cacr for all other accesses. only instruction fetches are cached (i.e. code space accesses.) 5.1.11 cache control register ( cacr ) the operation of the instruction cache is controlled by the cache control register (cacr). the cacr also provides a set of default memory access attributes used when a reference address does not map into the space defined by the access control registers. the cacr is accessed as control register $002 using the privileged movec instruction. this instruction provides write-only access to this register from the processor. additionally, table 5-4. instruction cache operation as defined by cacr [31, 10] cacr [31] cacr [10] type of inst fetch description 0 0 - instruction cache is completely disabled; all fetches are word, longword in size. 0 1 - all fetches are treated as non-cacheable and loaded into the line-fill buffer as defined by table 5-3 . 1 - cacheable fetch size is defined by table 5-3 , and contents of the line fill buffer can be written into the memory array. 1 0 non-cacheable all fetches are longword in size, and not loaded into the line-fill buffer. 1 1 non-cacheable fetch size is defined by table 5-3 , and loaded into the line-fill buffer, but are never written into the memory array. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-9 the cacr register may be accessed from the debug module in a similar manner. the entire register is cleared by a hardware reset. field definitions: 5.1.11.1 cenb: cacr[31]cache enable. 0 = cache disabled 1 = cache enabled the memory array of the instruction cache is enabled only if cenb is asserted. if ich_sz[2:0] is set to zero, this bit is set to zero and the icache module is disabled. any attempt to set the cenb bit is disabled when ich_sz[2:0] is set to zero. 5.1.11.2 cpsh: cacr[28] - cpush disable invalidate. 0 = enable invalidation 1 = disable invalidation when the privileged cpushl instruction is executed, the cache entry defined by bits [x:4] of the address is invalidated if cpsh = 0. if cpsh = 1, no operation is performed. 5.1.11.3 cfrz: cacr [ 27]cache freeze. 0 = normal operation 1 = freeze valid cache lines setting this bit effectively freezes the contents of the cache. line fetches may still be initiated and loaded into the line fill buffer, but a valid cache entry is never overwritten. if a given cache location is invalid, the contents of the line-fill buffer may be written into the memory array while cfrz is asserted. bits 31 30 29 28 27 26 25 24 23 18 17 16 field cenb - cpsh cfrz - cinv - park reset 0-00-0 - 0 r/w w-ww-w - w bits 15 11 10 9 8 7 6 5 4 2 1 0 field - cben dcm dbw e - dwp - clnf reset - 000 - 0 - 0 r/w - www - w - w cache control register (cacr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-10 coldfire2/2m users manual motorola 5.1.11.4 cinv: cacr[24]cache invalidate. 0 = no operation 1 = invalidate all cache lines setting this bit forces the cache to invalidate each tag array entry. the invalidation process requires n (n = #lines) machine cycles, with a single cache entry cleared per machine cycle. the state of this bit is always read as a zero. after a hardware reset, the cache must be invalidated before it is enabled. 5.1.11.5 park: cacr[17:16]optional external arbiter control. this field can be used to drive an external master arbiter control module via the marbc[1:0] signals. otherwise, these bits can use used as general purpose output bits. refer to section 3.10 master bus arbitration . if an external master arbiter module is not used, the marbc[1:0] signals may be used as general purpose control signals. 5.1.11.6 cben:cacr [ 10]cache enable non-cacheable instruction bursting. 0 = disable burst fetches on non-cacheable accesses 1 = enable burst fetches on non-cacheable accesses setting this bit allows the line-fill buffer to be loaded with burst transfers under control of clnf[1:0] for non-cacheable accesses. non-cacheable accesses are never written into the memory array. cben in conjunction with cacr[31], cenb, determines the line buffer and icache array status. see table 5-5 for guidance in setting cacr[10]. 5.1.11.7 dcm: cacr[9]default instruction fetch cache mode. 0 = caching enabled 1 = caching disabled this bit defines the default cache mode: 0 is cacheable, 1 is non-cacheable. for more information on the selection of the effective memory attributes, see section 5.1.5. 5.1.11.8 dbwe:cacr[8]default buffered write enable. 0 = disable buffered writes 1 = enable buffered writes table 5-5. cacr[31] and cacr[10] configuration cacr[31] cacr[10] icache/line fill buffer configuration 0 0 icache disabled, line fill buffer disabled 0 1 icache disabled, line fill buffer enabled 1 0 icache enabled, line fill buffer enabled on cachable accesses 1 1 icache enabled, line fill buffer enabled, but get the line-fill buffer even on non-cacheable accesses f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-11 this bit defines the default value for enabling buffered writes. if dbwe = 0, the termination of an operand write cycle on the processors local bus is delayed until the external bus cycle is completed. if dbwe = 1, the write cycle on the local bus in terminated immediately, and the operation buffered in the bus controller. in this mode, operand write cycles are effectively decoupled between the processors local bus and the external bus. generally, the enabling of buffered writes provides higher system performance, but recovery from access errors may be more difficult. for the coldfire cpu, the reporting of access errors on operand writes is always imprecise, and enabling buffered writes simply decouples the write instruction from the signaling of the fault even more. 5.1.11.9 dwp: cacr[5]default write protection. 0 = read and write access permitted 1 = only read access permitted the dwp bit defines the default write-protection attribute. if the effective memory attributes for a given access select the dwp bit, then any attempted write with this bit set is terminated with an access error. 5.1.11.10 clnf:cacr[1:0]cache line fill. this two-bit field determines the size of the instruction fetch memory transfer based on various cacr bits. see section 5.1.8 for additional information. the encoding is shown in table 5-6 . 5.2 access control registers (acr0, acr1) the two access control registers (acr0, acr1) provide a definition of memory reference attributes for two memory regions (one per acr). this set of effective attributes is defined for every memory reference using the acrs or the set of default attributes contained in the cacr. the acrs are examined for every memory reference that is not mapped to the ram or rom module. the acrs are accessed as control registers $004 and $005 using the privileged movec instruction (acr0 = $004, acr1 = $005). this instruction provides write-only access to these registers from the processor. additionally, the acrs may be accessed from the debug module in a similar manner. each acr is disabled by a hardware reset. table 5-6. external fetch size based on miss address & clnf clnf[1:0] miss address[3:2] 00 01 10 11 00 line line line longword 01 line line longword longword 1x line line line line f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-12 coldfire2/2m users manual motorola 5.2.1 acr programming model the acrs are accessible in supervisor mode using the movec instruction to access control register $004 for acr0 and $005 for acr1. all undefined bits are reserved and should always be written as zero. a hardware reset clears all bits in the acrs. field definitions: 5.2.1.1 ab: acr[31:24]address base [31:24]. this 8-bit field is compared to address bits [31:24] from the processors local bus, under control of the acr address mask. if the address matches, the attributes for the memory reference are sourced from the given acr. 5.2.1.2 am: acr [23:16]address mask [23:16]. this 8-bit field masks comparison of the access address with the acr address base bits. setting an am bit ignores the comparison of the corresponding address base bits. 5.2.1.3 en: acr [15]enable . 0 = acr disabled 1 = acr enabled the en bit defines the acr enable. this bit is cleared by hardware reset, disabling the acr. 5.2.1.4 sm: acr [14:13]supervisor mode . 00 = match if user mode 01 = match if supervisor mode 1x = match always; ignore user/supervisor mode this two-bit field allows the given acr to be applied to references based on operating privilege mode of the coldfire processor. the field allows use of the acr for user- references only, supervisor-references only, or all accesses. bits 31 24 23 16 field ab am reset 00 r/w ww bits 15 14 13 12 8 7 6 5 4 3 2 1 0 field en sm - enib cm bufw - wp - reset 00 - 000-0- r/w ww - www-w- access control register (acr0, acr1) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-13 5.2.1.5 cm: acr [6]cache mode. 0 = caching enabled 1 = caching disabled this bit defines the cache mode: 0 is cacheable, 1 is non-cacheable. 5.2.1.6 bwe: acr [5]buffered writes. 0 = disable buffered writes 1 = enable buffered writes this bit defines the value for enabling buffered writes. if bwe = 0, the termination of an operand write cycle on the processors local bus is delayed until the external bus cycle is completed. if bwe = 1, the write cycle on the local bus is terminated immediately, and the operation buffered in the bus controller. in this mode, operand write cycles are effectively decoupled between the processors local bus and the external bus. generally, the enabling of buffered writes provides higher system performance, but recovery from access errors may be more difficult. for the coldfire cpu, the reporting of access errors on operand writes is always imprecise, and enabling buffered writes simply decouples the write instruction from the signaling of the fault even more. 5.2.1.7 wp: acr [2]write protect. 0 = read and write accesses permitted 1 = only read accesses permitted the wp bit defines the write-protection attribute. if the effective memory attributes for a given access select the wp bit, then any attempted write with this bit set is terminated with an access error. 5.3 rom module the coldfire2/2m has a dedicated bus to support an integrated rom with the following features: ? 0 - 32 kbytes rom, organized by rom size/4 x 32 bits ? rom byte-size programmed with rom_sz[2:0] static signals ? single-cycle access ? physically located on processors high-speed bus ? byte, word, longword addressable ? memory-mapping defined by user ? configurable at reset as boot memory the rom module provides a general-purpose memory block that the coldfire2/2m can access in a single clock, and can store either code or data structures. the rom can be programmed to function as a boot memory; this configurable option is available at reset time. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-14 coldfire2/2m users manual motorola the rom size is specified via the rom_sz[2:0] pins which are static signals that need to stay valid for all operation. only the coldfire2/2m can access the rom module. the coldfire2/2m issues the instruction fetches in parallel to all k-bus memories. fetches from the rom module are not cached because they require only one clock cycle to complete. two roms are used to lower power consumption by only driving the required portions of the data bus. 5.3.1 rom signal description these signals interface the coldfire2/2m to two integrated roms. figure 5-2 illustrates an 8 kbyte configuration. all coldfire2/2m signals are unidirectional and synchronous. 5.3.1.1 rom address bus (rom_addr[14:2]). these output signals provide the address of the current bus cycle to the integrated roms. this bus should be connected to the address bus (a) of the compiled roms. the number of valid address signals depends on the total rom size as shown in table 5-7 . figure 5-2. example 8 kbyte rom interface diagram rom_addr[14:2] rom_do[31:0] rom_enb[1:0] rom_sz[2:0] a[10:0] do[15:0] romenb cpu core rom 0 [12:2] [31:0] [2:0] size a[10:0] do[15:0] romenb rom 1 [0] [1] [31:16] [15:0] rom_vld rom_vld f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-15 5.3.1.2 rom data output bus (rom_do[31:0]). these input signals provide the read data path from the integrated roms to the coldfire2/2m. the data bus is 32-bits wide and can transfer 8, 16, or 32 bits of data per bus transfer. during a line transfer, the data lines are time-multiplexed across multiple cycles to carry 128 bits. this bus should be connected to the data outputs (do) of the compiled roms. 5.3.1.3 rom enable (rom_enb[1:0]). these active-low, output signals indicate the roms are currently selected to drive the rom_do[31:0] bus. these signals should be connected individually to the enable signal (romenb) signal of the compiled roms. both are asserted for 32-bit accesses. rom_enb[0] connects to the msw while rom_enb[1] connects to the lsw. 5.3.1.4 rom size (rom_sz[2:0]). these static inputs specify the size of the compiled roms connected to the coldfire2/2m. these pins need to stay valid during all operation. table 5-8 lists the possible rom configurations. if the rom_sz pins are zero, the rom cannot be enabled via a cpu space write to rombar. therefore, if the rom is enabled while the rom_sz pins are at zero, the processor behaves as if no rom module existed. 5.3.1.5 rom valid (rom_vld). this active-high input signal determines if the rom module should be active immediately after a hard reset. thus, if asserted, the first fetches ($0, $4) go to rom instead of external memory. rom_vld controls the reset value of the table 5-7. valid rom address bits total rom size valid rom_addr bits 0 none 512 rom_addr[8:2] 1 k rom_addr[9:2] 2 k rom_addr[10:2] 4 k rom_addr[11:2] 8 k rom_addr[12:2] 16k rom_addr[13:2] 32k rom_addr[14:2] table 5-8. rom configuration encoding total rom size (bytes) rom_sz[2:0] address (bits) data 1 (bits) none 000 - - 512 001 7 2 @ 16 1 k 010 8 2 @ 16 2 k 011 9 2 @ 16 4 k 100 10 2 @ 16 8 k 101 11 2 @ 16 16k 2 110 12 2 @ 16 32k 2 111 13 2 @ 16 notes: 1. 2 roms, each 16-bits wide 2. 16k and 32k roms may require a reduced operating frequency. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-16 coldfire2/2m users manual motorola rom base address register (i.e. the rom must be based at $0000 if rom_vld is asserted). 5.3.2 rom programming model the configuration information in the rom base address register (rombar) controls the rom module operation. the rombar is accessible in supervisor mode as control register $c00 using the movec instruction. all undefined bits are reserved and should always be written as zero. a hardware reset clears only the valid bit if rom_vld is negated. if rom_vld is asserted, rombar is reset to $0121. this activates the rom with a starting address of $0000 and all address spaces, except the cpu space, are allowed. this is shown below where the reset values in parenthesis are valid if rom_vld is asserted during a reset. field definitions: 5.3.2.1 ba: rombar [31:9]base address. defines the base address for the rom module address range. the number of valid base address bits in this field is a function of the rom size as shown in table 5-9 . the base address is reset to $0 if rom_vld is asserted during reset. bits 31 16 field ba reset - (0) r/w w bits 15 9 8 7 6 5 1 0 field ba wp - as v reset - (0) - (1) - - 0 (1) r/w ww-ww note: the reset values in parenthesis are valid if the rom_vld signal is asserted during reset. rom base address register (rombar) table 5-9. valid rom base address bits rom size valid ba bits 512 ba[31:9] 1 k ba[31:10] 2 k ba[31:11] 4 k ba[31:12] 8 k ba[31:13] 16k ba[31:14] 32k ba[31:15] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-17 5.3.2.2 wp: rombar[8]write protect. this bit is reserved for future use. this bit can be used for debug purposes; i.e. it could set a trip point when a write to rom occurred. 0 = no effect 1 = an attempted write access will generate an access error exception in the processor. 5.3.2.3 as rombar[5:1]address space masks. this five-bit field, specified by rombar[5:1], allows certain types of accesses to be masked or inhibited, from accessing the ram module. the mask bits are defined as: as5 - mask cpu space and interrupt acknowledge accesses as4 - mask supervisor code accesses as3 - mask supervisor data accesses as2 - mask user code accesses as1 - mask user data accesses if a given mask bit is set, then references of that type are not allowed to access the rom module. if asn = 0, then accesses of the given type are allowed by the rom. if asn = 1, then accesses of the given type are not allowed by the rom. if an access is made to a space that is masked, it simply becomes mapped to the next valid space. 5.3.2.4 v: rombar [0]valid. the valid bit is specified by rombar[0]. this bit is cleared by a hardware reset. when set, this bit enables the rom module, otherwise the module is disabled. if rom_sz is set to zero, this bit is set to zero and the rom module is disabled. any attempt to set the valid bit is disabled when rom_sz is set to zero. 5.3.3 rom initialization / rom boot after a hardware reset, the contents of the rombar depend upon the rom_vld signal polarity and the rom_sz[2:0] configuration. if rom_sz[2:0] = 0, the valid bit is held cleared, even if a write is attempted during a load to rombar. rom_vld determines if the rom module should be active immediately after a hard reset. thus, if rom_vld is asserted, the first fetches ($0, $4) go to rom instead of external memory. note, if rom_vld is asserted, rom_sz cannot be set to zero. rom_vld controls the reset value of the rom base address register (i.e. the rom must be based at $0000 if rom_vld is asserted). to map the rombar, a load to the rombar mapping the rom module to the desired location within the address space, must be performed. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-18 coldfire2/2m users manual motorola 5.3.4 power management as noted previously, depending upon the configuration defined by the rombar, instruction fetch accesses may be sent to the rom module and the i-cache simultaneously. if the access is mapped to the rom module, it sources the read data and the i-cache access is discarded. if the rom is used only for data operands, power dissipation can be lowered by asserting the asn bits associated with instruction fetches. additionally, if the rom contains only instructions, power dissipation can be reduced by masking operand accesses. consider the following examples of typical rombar settings: data contained in rom rombar[7:0] only code $2b only data $35 both code and data $21 5.4 ram module the coldfire2/2m has a dedicated bus to support an integrated ram with the following features: ? 0 - 32 kbytes ram, organized by ram byte size/4 x 32 bits ? ram byte-size programmed with ram_sz[2:0] static signals ? single-cycle access ? physically located on processors high-speed local bus ? byte, word, longword address capability ? memory mapping defined by user 5.4.1 ram operation the ram module provides a general-purpose memory block that the coldfire2/2m can access in a single cycle. the ram size is specified via the ram_sz[2:0] pins which are static signals that need to stay valid for all operation. the location of the memory block may be specified to any 0-modulo-[icache byte size] address within the four gigabyte address space. the memory is ideal for storing critical code or data structures, or for use as the system stack. since the ram module physically is connected to the processors high-speed local bus, it can service cpu-initiated accesses, or memory referencing commands from the debug module. depending on configuration information, instruction fetches can be sent to both the instruction cache and the ram block simultaneously. if the instruction fetch address is mapped into the region defined by the ram, the ram provides data back to the processor, and the i-cache data is discarded. accesses from the ram module are not cached. generally, the ram is loaded by copying a hex image from another memory region into the ram address space. this copy function can be performed by the processor during f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-19 initialization, or can be performed by an emulator during system debug. the emulator approach uses the bdm serial communication channel to download the hex image from a host machine into the ram directly. 5.4.2 ram signal description these signals interface the coldfire2/2m to the four integrated arrays that comprise the ram. figure 5-3 illustrates an 8 kbyte configuration. all coldfire2/2m signals are unidirectional and synchronous. 5.4.2.1 ram address bus (ram_addr[14:2]). these registered output signals provide the address of the current bus cycle to the integrated rams. this bus should be connected to the address bus (a) of the four compiled figure 5-3. example 8 kbyte ram interface diagram ram 0 ram 3 cpu core ram_addr[14:2] ram_csb ram_di[31:0] ram_do[31:0] ram_st[3:0] ram_rwb[3:0] size ram_sz[2:0] csb dbi[7:0] dbo[7:0] st rwb a[10:0] csb dbi[7:0] dbo[7:0] st rwb [12:2] [31:0] [31:0] [3:0] [3:0] [31:24] [31:24] [7:0] [7:0] a[10:0] [3] [0] [3] [0] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-20 coldfire2/2m users manual motorola rams. the number of valid address signals depends on the total ram size as shown in table 5-10 . 5.4.2.2 ram chip-select (ram_csb). this active-low output signal indicates the rams are currently selected to perform a data transfer with the coldfire2/2m. this signal should be connected to the chip-select (csb) signal of the four compiled rams. 5.4.2.3 ram data input bus (ram_di[31:0]). these output signals provide the write data path between the coldfire2/2m and the integrated ram. the data bus is 32-bits wide and can transfer 8, 16, or 32 bits of data per bus transfer. during a line transfer, the data lines are time-multiplexed across multiple cycles to carry 128 bits. this bus should be connected to the data inputs (dbi) of the four compiled rams. if only one byte is being written, the byte will be replicated on all 4 lines; likewise a word will be replicated in both word positions. 5.4.2.4 ram data output bus (ram_do[31:0]). these input signals provide the read data path between the integrated ram and the coldfire2/2m. the data bus is 32-bits wide and can transfer 8, 16 or 32 bits of data per bus transfer. during a line transfer, the data lines are time-multiplexed across multiple cycles to carry 128 bits. this bus should be connected to the data outputs (dbo) of the four compiled rams. 5.4.2.5 ram size (ram_sz[2:0]). these static inputs specify the size of the compiled rams connected to the coldfire2/2m. these pins need to stay valid during all operation. if the ram_sz pins are zero, the ram cannot be enabled via a cpu space write to rambar. therefore if the ram is enabled while the ram_sz pins are at zero, the processor behaves as if no ram module existed. table 5-11 lists the possible ram configurations. table 5-10. valid ram address bits total ram size valid ram_addr bits 0 none 512 ram_addr[8:2] 1 k ram_addr[9:2] 2 k ram_addr[10:2] 4 k ram_addr[11:2] 8 k ram_addr[12:2] 16k ram_addr[13:2] 32k ram_addr[14:2] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-21 5.4.2.6 ram strobe (ram_st[3:0]). these output signals initiate a read or write cycle to the integrated rams on a low-to-high transition. these signals should be connected individually to the strobe input (st) signals of the four compiled rams. the st[0] signal connects to the high-order byte and st[3] connects to the low-order byte. 5.4.2.7 ram read/write (ram_rwb[3:0]). these output signals indicate the direction of the data transfer to the integrated rams. a high level indicates a read cycle and a low level indicates a write cycle. they should be connected individually to the read/write (rwb) signal of the four compiled rams. like ram_st[3:0], the ram_rwb[3] signal connects to the high-order byte and ram_st[0] connects to the low-order byte. 5.4.3 ram programming model the configuration information in the ram base address register (rambar) controls the operation of the ram module. the rambar is accessed as control register $c04 using the privileged movec instruction. the movec instruction provides write-only access to this register. additionally, the rambar register may be accessed from the debug module in a similar manner. from the debug module, the register may be read or written. all undefined bits are reserved. these bits should be written as zeroes, and return zeroes when read from the debug module. only the valid bit is cleared by a hardware reset. table 5-11. ram configuration encoding total ram size (bytes) ram_sz[2:0] address (bits) data 1 (bits) 0 000 - - 512 001 7 4@8 1 k 010 8 4@8 2 k 011 9 4@8 4 k 100 10 4@8 8 k 101 11 4@8 16k 2 110 12 4@8 32k 2 111 13 4@8 notes: 1. 4 rams, each 8-bits wide 2. 16k and 32k rams may require a reduced operating frequency. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-22 coldfire2/2m users manual motorola 5.4.4 ram base address register w field definitions: 5.4.4.1 ba: rambar [31:9]base address. defines the base address for the ram module address range. the number of valid base address bits in this field are a function of the ram size as shown in table 5-12 . by programming this field, the ram may be located anywhere within the four gigabyte address space of coldfire. 5.4.4.2 wp: rambar [8]write protect. 0 = ram module supports read and write references 1 = ram module supports only read accesses. the write protect field is defined by rambar [8]. if set, this bit allows only read accesses to the ram. any attempted write access will generate an access error exception in the processor. if cleared, the ram supports read and write references. bits 31 16 field ba reset - r/w w bits 15 9 8 7 6 5 1 0 field ba wp - as v reset ----0 r/w ww-ww ram base address register (rambar) table 5-12. valid ram base address bits ram byte size valid ba bits 512 ba[31:9] 1 k ba[31:10] 2 k ba[31:11] 4 k ba[31:12] 8 k ba[31:13] 16k ba[31:14] 32k ba[31:15] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-23 5.4.4.3 as: rambar [5:1]address space masks. this five bit field, specified by rambar [5:1], allows certain types of accesses to be masked, or inhibited, from accessing the ram module. the mask bits are defined as: as5 - mask cpu space and interrupt acknowledge accesses as4 - mask supervisor code accesses as3 - mask supervisor data accesses as2 - mask user code accesses as1 - mask user data accesses if a given mask bit is set, then references of that type are not allowed to access the ram module. these bits are useful for power management as detailed in section 5.4.6. if asn = 0, then accesses of the given type are allowed by the ram. if asn = 1, then accesses of the given type are not allowed by the ram. if an access is made to a space that is masked, it simply becomes mapped to the next valid space. 5.4.4.4 v: rambar [0]valid. this bit indicates when the contents of the rambar are valid. the base address value is not used, and the ram module is not accessible until the v-bit is set. an external bus cycle is generated if the base address field matches the internal core address, and the v-bit is cleared. 0 = contents of rambar are not valid 1 = contents of rambar are valid the valid bit is specified by rambar [0]. this bit is cleared by a hardware reset. when set, this bit enables the ram module, otherwise the module is disabled. the mapping of a given access into the ram used the following algorithm to determine if the access hits in the memory: if (rambar[0] = 1) if (requested address [31:9]* = rambar[31:9]* if (asn of the requested type = 0) access is mapped to the ram module if (access = read) read the ram and return the data if (access = write) if (rambar[8] = 0) write the data into the ram else signal a write-protect access error * see table 5-10 for address bit map pertaining to ram sizes. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories 5-24 coldfire2/2m users manual motorola if ram_sz is set to zero, this bit is set to zero, and the ram module is disabled. any attempt to set the valid bit is disabled when ram_sz is set to zero. 5.4.5 ram initialization after a hardware reset, the contents of the ram module are undefined. the valid bit of the rambar is cleared, disabling the module. if the ram needs to be initialized with instructions or data, the following steps should be performed: 1. load the rambar mapping the ram module to the desired location within the address space. 2. read the source data and write it to the ram. there are various instructions to support this function, including memory-to-memory move instructions, or the movem opcode. the movem instruction is optimized to generate line-sized burst fetches on 0-modulo-16 addresses, so this opcode generally provides maximum performance. 3. after the data has been loaded into the ram, it may be appropriate to load a revised value into the rambar with a new set of attributes. these attributes consist of the write-protect and address space mask fields. these initialization functions can be performed by the coldfire processor, or from an external emulator using the debug module. 5.4.6 power management as noted previously, depending upon the configuration defined by the rambar, instruction fetch accesses may be sent to the ram module and the i-cache simultaneously. if the access is mapped to the ram module, it sources the read data and the i-cache access is discarded. if the ram is used only for data operands, power dissipation can be lowered by asserting the asn bits associated with instruction fetches. additionally, if the ram contains only instructions, power dissipation can be reduced by masking operand accesses. consider the following examples of typical rambar settings: data contained in ram rambar[7:0] only code $2b only data $35 both code and data $21 5.5 interactions between kbus memories depending on configuration information, instruction fetches and operand accesses may be sent to all of the k-bus memories (i.e. ram, rom, and icache) simultaneously. there needs to be consistency between the acrs and the default modes defined by cacr (cacr[9], cacr[8], and cacr[5]). f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
integrated memories motorola coldfire2/2m users manual 5-25 if the access address is mapped into the region defined by the ram (and this region is not masked), the ram provides the data back to the processor, and the i-cache data discarded. accesses from the ram module are not cached. the rom behaves similarly. the ram has priority over the rom. the full priority scheme is as follows: if (ram hits) ram supplies data to the processor else if (rom hits) rom supplies data to the processor else if (line-fill buffer hits) line-fill buffer supplies data to the processor else if (icache hits) icache supplies data to the processor else master bus cycle accesses to reference data from non-k-bus memory f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 6-1 section 6 multiply-accumulate unit this section details the hardware multiply-accumulate (mac) support functions within the coldfire2m. the mac is not present in the coldfire2 hard macro, but is an optional high- speed execution unit that is available within the coldfire core environment. if the mac is present, it executes the integer multiplies at an accelerated speed. the mac unit is designed to provide a common set of dsp operations and to enhance the current multiply instructions in the coldfire architecture. the mac unit provides functionality in three related areas: ? faster signed and unsigned integer multiplies ? new multiply-accumulate operations supporting signed and unsigned operands ? new miscellaneous register operations each is addressed in detail in the following sections. 6.1 introduction the mac unit is designed to optimize the current coldfire multiply instructions, muls and mulu, and provide additional instructions for algorithms that use multiply-accumulate operations. as shown in figure 6-1 , these new instructions multiply two numbers, followed by the addition/subtraction of this product to/from the value contained in the accumulator. the product may be optionally shifted left or right one bit before the addition or subtraction takes place. all arithmetic operations use register-based input operands, and summed values are stored in the internal accumulator. for word operations, the upper or lower word of each input register can be chosen as a source operand. as a result, two word operations will require only one register load operation. some instructions also allow a parallel load to be performed in addition to the mac operation. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
multiply-accumulate unit 6-2 coldfire2/2m users manual motorola the mac module has been designed for 16-bit multiplies to optimize the die size. word operations require two signed or unsigned 16-bit operands and produce a 32-bit result. by iteratively performing multiple 16-bit operations, the mac can perform 32-bit operations. longword operations require two signed or unsigned 32-bit operands and produce a 32-bit result. because the longword operations are iteratively computed, the mac instructions have an effective issue rate of one clock for word operations and three clocks for longword operations. 6.2 mac programming model the mac unit includes three new registers: a 32-bit accumulator (acc), an 8-bit status register (macsr), and a 16-bit mask register (mask). 6.2.1 accumulator (acc) this is a 32-bit special purpose register used to accumulate the results of mac operations. the accumulator is not cleared upon reset. 6.2.2 mac status register (macsr) the status register contains the saturation mode control bit, the negative, zero and overflow flags, as well as the signed/unsigned operation control bit as shown in the next register. figure 6-1. mac flow diagram x +/- operand x operand y shift 0,1,-1 accumulator f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
multiply-accumulate unit motorola coldfire2/2m users manual 6-3 field definitions: omc[7]Coverflow mode control 0 = normal mode 1 = saturation mode this bit is used to enable or disable saturation mode on overflow. this bit is cleared by system reset. refer to section 6.4 overflow mode . s/u[6]Csigned/unsigned operations the s/u bit defines the type of multiply operations that are performed. 0 = signed numbers 1 = unsigned numbers this bit is cleared by system reset. n[3]Cnegative this bit is set if the most significant bit of the result is set, cleared otherwise. this bit is affected only by mac operations. the muls and mulu instructions do not change this value. z[2]Czero this bit is set if the result equals zero, otherwise, it is cleared. this bit is affected only by mac operations. muls and mulu instructions do not change this value. v[1]Coverflow this bit is set if an arithmetic overflow occurs implying that the result cannot be represented in 32 bits. once set, this bit remains set until the acc register is loaded with a new value using the move to acc instruction, or the macsr is explicitly loaded using the move to macsr instruction. the muls and mulu instructions do not change this value. c[0]Ccarry this field is always zero and is reserved for future use. 6.2.3 mask register (mask) this is a 16-bit special purpose register used to mask the lower 16 address bits during mac load operations. the field definition follows. bits 7 6 5 4 3 2 1 0 field omc s/u - - n z v c reset 0000-- -0 r/w r/w r/w r r r/w r/w r/w r mac status register (macsr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
multiply-accumulate unit 6-4 coldfire2/2m users manual motorola field definitions: mask[15:0]Caddress mask this field is logically anded with the lower 16 bits of the effective address during mac instructions with a register load performed using the mask addressing mode modifier. the actions of the mask addressing mode depend on the effective addressing mode as shown in table 6-1 . when loading data into the data registers, the mask register can be used to implement a very efficient circular queue. 6.3 shifting operations the mac unit is capable of shifting a product before the result is added to or subtracted from the accumulator (acc). since there is the possibility of overflowing a 32-bit product, use the following guidelines when performing mac instructions: ? for both word and longword unsigned operations, a zero is shifted into the product on right shifts. ? for signed word operations, the sign bit is shifted into the product on right shifts, unless the product is zero. ? for signed, longword operations, the sign bit is shifted into the product unless an overflow occurs or the product is zero, in which case a zero is shifted in. 6.4 overflow mode when dealing with potential overflow conditions, software overhead can be minimized by enabling hardware support for saturation arithmetic. saturation mode is controlled by the bits 15 0 field mask rese t - r/w r/w mac mask register (mask) table 6-1. mask addressing mode effective addressing mode output address new an (an) an & {0xffff, mask} - (an)+ an (an +4) & {0xffff, mask} -(an) (an - 4) & {0xffff, mask} (an - 4) & {0xffff, mask} (d16, an) (an + d16) & {0xffff, mask} - note: {upper, lower} notation indicates upper and lower order words f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
multiply-accumulate unit motorola coldfire2/2m users manual 6-5 omc bit in the macsr. in saturation mode, if a mac instruction overflows 32 bits during the multiply portion of an operation, the overflow bit (v) will be set in the macsr and the accumulator (acc) will contain the most positive or the most negative value possible. as seen in table 6-2 , the value depends on the signed/unsigned mode (s/u bit in the macsr), multiply result, and the addition/subtraction operation. the overflow value will remain in the acc until the v bit in the macsr is cleared by loading a new value into the acc or macsr. 6.5 mac instruction set summary the mac unit in the coldfire2m enhances the multiply operations that are currently supported by the coldfire2 architecture, adds new multiply-accumulate instructions, and adds register instructions for accessing the new mac registers. table 6-3 below summarizes the mac unit instruction set. refer to appendix b new mac instructions for details. table 6-2. accumulator result in saturation mode signed/unsigned multiply/ result mac instruction mac operation accumulator result signed positive mac , macl addition $7fffffff msac , msacl subtraction $80000000 negative mac , macl addition $80000000 msac , msacl subtraction $7fffffff unsigned - mac , macl addition $ffffffff msac , msacl subtraction $00000000 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
multiply-accumulate unit 6-6 coldfire2/2m users manual motorola table 6-3. mac instruction set summary instruction operand syntax operand size operation muls ,dx 16 x 16 ? 32 32 x 32 ? 32 dx ? dx note: signed operation mulu ,dx 16 x 16 ? 32 32 x 32 ? 32 dx ? dx note: unsigned operation mac ry,rx , 16 16 + 32 ? 32 32 32 + 32 ? 32 acc + (ry rx){ << 1 | >> 1} ? acc macl ry,rx,,,rw 16 16 + 32 ? 32 32 32 + 32 ? 32 acc + (ry rx){ << 1 | >> 1} ? acc; ({& mask }) ? rw msac ry,rx, 32 - 16 16 ? 32 32 - 32 32 ? 32 acc - (ry rx){ << 1 | >> 1} sf ? acc msacl ry,rx,,,rw 32 - 16 16 ? 32 32 - 32 32 ? 32 acc - (ry rx){ << 1 | >> 1} sf ? acc; ({& mask }) ? rw move from acc acc ,rx 32 acc ? rx move from macsr macsr ,rx macsr , ccr 32 8 macsr ? rx macsr ? ccr move from mask mask ,rx 32 mask ? rx move to acc ry, acc #,acc 32 32 ry ? acc # ? acc move to ccr dy, ccr , #,ccr 16 dy ? ccr # ? ccr move to macsr ry, macsr #,macsr 32 ry ? macsr # ? macsr move to mask ry ,mask #,mask 32 32 ry ? mask # ? mask f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 7-1 section 7 debug support this section details the hardware debug support functions within the coldfire2/2m. the general topic of debug support has been divided into three separate areas: ? real-time trace support ? background debug mode (bdm) ? real-time debug support each is addressed in detail in the following sections. the logic required to support these three areas is contained in a debug module, which is shown in the system block diagram in figure 7-1 . 7.1 signal description this section describes the coldfire2/2m signals associated with the debug module. all coldfire2/2m signals are unidirectional and synchronous. 7.1.1 break point (bkptb) this active-low, unidirectional input signal is used to request a manual break point. it will cause the processor to enter a halted state after the completion of the current instruction. this status will be reflected on the processor status ( pst ) outputs. figure 7-1. processor/debug module interface coldfire cpu debug module internal buses core trace port ddata, pst bdm port dsclk, dsi, dso bkptb f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-2 coldfire2/2m users manual motorola 7.1.2 debug data (ddata[3:0]) these output signals display the captured processor status and break point status. 7.1.3 development serial clock (dsclk) this input signal is used as the development serial clock for the serial interface to the debug module.the maximum frequency is 1/2 the clock ( clk ) frequency. 7.1.4 development serial input (dsi) this input signal provides the single-bit communication for the debug module commands. 7.1.5 development serial output (dso) this output signal provides single-bit communication for the debug module responses. 7.1.6 processor status (pst[3:0]) these output signals report the processor status. table 7-1 shows the encoding of these signals. these signals indicate the current status of the processor pipeline and, as a result, are not related to the current bus transfer. . . 7.2 real-time trace in the area of debug functions, one fundamental requirement is support for real-time trace functionality, i.e., definition of the dynamic execution path. the coldfire2/2m solution is to include a parallel output port providing encoded processor status and data to an external development system. this port is partitioned into two nibbles (4 bits): one nibble allows the table 7-1. processor status encoding pst[3:0] definition (hex) (binary) $0 0000 continue execution $1 0001 begin execution of an instruction $2 0010 reserved $3 0011 entry into user-mode $4 0100 begin execution of pu lse orwddata instruction $5 0101 begin execution of taken branch $6 0110 reserved $7 0111 begin execution of rte instruction $8 1000 begin 1-byte transfer on ddata $9 1001 begin 2-byte transfer on ddata $a 1010 begin 3-byte transfer on ddata $b 1011 begin 4-byte transfer on ddata $c 1100 exception processing? $d 1101 emulator-mode entry exception processing? $e 1110 processor is stopped, waiting for interrupt? $f 1111 processor is halted ? note: ?these encodings are asserted for multiple cycles. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-3 processor to transmit information concerning the execution status of the core (processor status , pst[3:0]), while the other nibble allows operand data to be displayed (debug data, ddata[3:0]). the processor status timing is synchronous with the processor clock (clk) and may not be related to the current bus transfer. table 7-1 shows the encoding of these signals. the processor status (pst) outputs can be used with an external image of the program to completely track the dynamic execution path of the machine. the tracking of this dynamic path is complicated by any change-of-flow operation. this is especially evident when the branch target address is calculated based on the contents of a program visible register (variant addressing.) for this reason, the debug data (ddata) outputs will display the target address of a taken branch instruction. because the ddata bus is only 4 bits wide, the address is displayed a nibble at a time across multiple clock cycles. the debug module includes two 32-bit storage elements for capturing the internal coldfire2/ 2m bus information. these two elements effectively form a fifo buffer connecting the internal bus to the external development system through the ddata signals. the fifo buffer captures branch target addresses along with certain operand read/write data for eventual display on the ddata output port one nibble at a time. the execution speed of the coldfire2/2m is affected only when both storage elements contain valid data waiting to be dumped onto the ddata port. in this case, the processor core is stalled until one fifo entry is available. in all other cases, data output on the ddata port does not impact execution speed. 7.2.1 processor status signal encoding the processor status (pst) signals are encoded to indicate a variety of conditions that are not always visible outside of the coldfire2/2m. 7.2.1.1 continue execution (pst = $0). most instructions complete in a single cycle. if an instruction requires more clock cycles, the subsequent clock cycles are indicated by driving the pst outputs with this encoding. 7.2.1.2 begin execution of an instruction (pst = $1). for most instructions, this encoding signals the first cycle of an instructions execution. some instructions generate a different unique encoding. 7.2.1.3 entry into user mode (pst = $3). this encoding indicates the coldfire2/2m has entered user mode. this encoding is signaled after the instruction which causes the user mode to be entered is executed (signaled with the appropriate encoding.) 7.2.1.4 begin execution of pulse or wddata instructions (pst = $4). the coldfire2/2m instruction set architecture includes a pulse opcode. this opcode generates a unique pst encoding, $4, when executed. this instruction can define logic analyzer triggers for debug and/or performance analysis. additionally, a wddata instruction is supported that allows the processor core to write any operand (byte, word, long) directly to the ddata port, independent of any debug module configuration. this opcode also generates the special pst encoding, $4, when executed, but a data transfer on ddata will f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-4 coldfire2/2m users manual motorola also be indicated. the size of the transfer will depend on the format of the wddata instruction. 7.2.1.5 begin execution of taken branch (pst = $5). this encoding is generated whenever a taken branch is executed. the branch target may be optionally displayed on ddata depending on the control parameters contained in the configuration/ status register (csr). the number of bytes of the address to be displayed is also controlled in the csr and indicated during the data transfer on the following clock cycle. the bytes are always displayed in a least-significant to most-significant order. the processor captures only those target addresses associated with taken branches using a variant addressing mode, i.e. all jmp and jsr instructions using address register indirect or indexed addressing modes, all rte and rts instructions as well as all exception vectors. the simplest example of a branch instruction using a variant address is the compiled code for a c language case statement. typically, the evaluation of this statement uses the variable of an expression as an index into a table of offsets, where each offset points to a unique case within the structure. for these types of change-of-flow operations, the coldfire2/2m processor uses the debug pins to output a sequence of information on successive clock cycles 1. identify a taken branch has been executed using the pst pins ($5). 2. using the pst pins, optionally signal the target address is to be displayed on the ddata pins. the encoding ($9, $a, $b) identifies the number of bytes that are displayed. 3. the new target address is optionally available on subsequent cycles using the nibble- wide ddata port. the number of bytes of the target address displayed on this port is a configurable parameter (2, 3, or 4 bytes). another example of a variant branch instruction would be a jmp (a0) instruction. if the csr was programmed to display the lower two bytes of an address, the output of the pst and ddata signals when this instruction executed are shown in figure 7-2 . in the first cycle, pst is driven with a $5 indicating a taken branch with a variant address. in the second cycle, pst is driven with a $9 indicating a two-byte address will be displayed four bits at a time on the ddata signals over the next four clock cycles. the remaining four clock cycles display the lower two-bytes of the address (a0), least significant nibble to most significant nibble. the output of the pst signals after the branch instruction completes will figure 7-2. example pst diagram clk $5 $9 $0 pst $0 $0 a[3:0] a[7:4] a[11:8] a[15:12] ddata f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-5 be dependent on the next instruction in the pipeline. the pst can continue with the next instruction before the address has completely displayed on the ddata because of the ddata fifo. if the fifo is full and the next instruction needs to display something on ddata, the pipeline will stall (pst = $0) until space is available in the fifo. 7.2.1.6 begin execution of rte instruction (pst = $7). the unique encoding is generated whenever the return-from-exception instruction is executed. 7.2.1.7 begin data transfer (pst = $8 - $a). these encodings indicate the number of bytes to be displayed on the ddata port on subsequent clock cycles. this encoding is driven onto the pst port one machine cycle before the actual data is displayed on ddata. 7.2.1.8 exception processing (pst = $c). this encoding is displayed during normal exception processing. exceptions which enter emulation mode (debug interrupt, or optional trace) generate a different encoding. because this encoding defines a multicycle mode, the pst outputs are driven with this value until exception processing is completed. 7.2.1.9 emulator-mode exception processing (pst = $d). exceptions which enter emulation mode (debug interrupt, or optional trace) generate a different this encoding. because this encoding defines a multicycle mode, the pst outputs are driven with this value until exception processing is completed. 7.2.1.10 processor stopped (pst = $e). this encoding is generated as a result of the stop instruction. the coldfire2/2m remains in the stopped state until an interrupt occurs. because this encoding defines a multicycle mode, the pst outputs are driven with this value until the stopped mode is exited. 7.2.1.11 processor halted (pst = $f). this encoding is generated as when the coldfire2/2m is halted (see section 7.3.1 cpu halt .) because this encoding defines a multicycle mode, the pst outputs are driven with this value until the halted mode is exited. 7.3 background debug mode (bdm) the coldfire2/2m supports a modified version of the background debug mode (bdm) functionality found on motorolas cpu32 family of parts. bdm implements a low-level system debugger in the microprocessor hardware. communication with the development system is handled via a dedicated, high-speed serial command interface (bdm port). unless noted otherwise, the bdm functionality provided by the coldfire2/2m is a proper subset of the cpu32 functionality. the main differences include the following: ? coldfire2/2m implements the bdm controller in a dedicated hardware module. although some bdm operations do require the cpu to be halted (e.g. cpu register accesses), other bdm commands such as memory accesses can be executed while the processor is running. ? dsclk, dsi, and dso are treated as synchronous signals, where the inputs, dsclk and dsi, must meet the required input setup and hold timings, and the output, dso, is specified as a delay relative to the rising edge of the processor clock. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-6 coldfire2/2m users manual motorola ? on cpu32 parts, the dso signal can inform hardware that a serial transfer can start. coldfire clocking schemes restrict the use of this bit. because dso changes only when dsclk is high, dso cannot be used to indicate the start of a serial transfer. the development system should use either a free-running dsclk or count the number of clocks in any given transfer. ? the read/write system register commands, rsreg and wsreg, have been replaced by read/write control register commands, rcreg and wcreg. these commands use the register coding scheme from the movec instruction. ? the read/write debug module register commands, rdmreg and wdmreg, have been added to support debug module register accesses. ? call and rst commands are not supported and will generate an illegal command response. ? illegal command responses can be returned using the fill and dump commands. ? for any command performing a byte-sized memory read operation, the upper 8 bits of the response data are undefined. the referenced data is returned in the lower 8 bits of the response. ? the debug module forces alignment for memory-referencing operations: long accesses are forced to a 0-modulo-4 address; word accesses are forced to a 0-modulo-2 address. an address error response can no longer be returned. 7.3.1 cpu halt although some bdm operations can occur in parallel with cpu operation, unrestricted bdm operation requires the cpu to be halted. a number of sources can cause the cpu to halt, including the following as shown in order of priority: 1. the occurrence of the catastrophic fault-on-fault condition automatically halts the processor. the halt status, $f, is posted on the pst port. 2. the occurrence of a hardware breakpoint can be configured to generate a pending halt condition in a manner similar to the assertion of the bkptb signal. in all cases, the occurrence of this type of breakpoint halts the processor in an imprecise manner. once the hardware breakpoint is asserted, the processor halts at the next sample point. see section 7.4.1 theory of operation for more detail. 3. the execution of the halt , also known as bgnd on the 683xx devices, instruction immediately suspends execution and posts the halt status ($f) on the pst outputs. by default, this is a supervisor instruction and attempted execution while in user mode generates a privilege-violation exception. a user halt enable (uhe) control bit is provided in the configuration/status register (csr) to allow execution of halt in user mode. 4. the assertion of the bkptb input pin is treated as a pseudo-interrupt, i.e., the halt condition is made pending until the processor core samples for halts/interrupts. the processor samples for these conditions once during the execution of each instruction. if there is a pending halt condition at the sample time, the processor suspends execution and enters the halted state. the halt status, $f, is reflected in the pst outputs. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-7 there are two special cases involving the assertion of the bkptb pin to be considered. after the master reset signal (mrstb) is negated, the processor waits for 16 clock cycles before beginning reset exception processing. if the bkptb input pin is asserted within the first eight cycles after mrstb is negated, the processor will enter the halt state, signaling that halt status, $f, on the pst outputs. while in this state, all resources accessible via the debug module can be referenced. this is the only opportunity to force the coldfire2/2m into emulation mode via the emu bit in the configuration/status register (csr). once the system initialization is complete, the processor response to a bdm go command is dependent on the set of bdm commands performed while breakpointed. specifically, if the processors pc register was loaded, then the go command simply causes the processor to exit the halted state and pass control to the instruction address contained in the pc. note in this case, the normal reset exception processing is bypassed. conversely, if the pc register was not loaded, then the go command causes the processor to exit the halted state and continue with reset exception processing. the coldfire2/2m also handles a special case with the assertion of bkptb while the processor is stopped by execution of the stop instruction. for this case, the processor exits the stopped mode and enters the halted state. once halted, the standard bdm commands may be exercised. when the processor is restarted, it continues with the execution of the next sequential instruction, i.e., the instruction following the stop opcode. the halt source is indicated in csr[27:24]; for simultaneous halt conditions, the highest priority source is indicated. 7.3.2 bdm serial interface once the cpu is halted and the halt status reflected on the pst outputs (pst[3:0] = $f), the development system may send unrestricted commands to the debug module. the debug module implements a synchronous protocol using a three-pin interface: development serial clock (dsclk), development serial input (dsi), and development serial output (dso). the development system serves as the serial communication channel master and is responsible for generation of the clock (dsclk). the operating range of the serial channel is dc to 1/2 of the processor frequency. the channel uses a full duplex mode, where data is transmitted and received simultaneously by both master and slave devices. the transmission consists of 17-bit packets composed of a status/control bit and a 16-bit data word. as seen in figure 7-3 , data is exchanged on the positive edge of clk when dsclk is high (i.e. dsi is sampled and dso is driven.) the dsclk signal must also be sampled low (on a positive edge of clk) between each bit exchange. the msb is transferred first. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-8 coldfire2/2m users manual motorola both dsclk and dsi are synchronous inputs and must meet input setup and hold times with respect to clk. the dsclk signal essentially acts as a pseudo clock enable and is sampled on the rising edge of clk. if the setup time of dsclk is met, then the internal logic transitions on the rising edge of clk, and dsi is sampled on the same clk rising edge. the dso output is specified as a delay from the dsclk-enabled clk rising edge. all events in the debug modules serial state machine are based on the rising edge of the microprocessor clock (see figure 7-4 below). also refer to the electrical characteristics section of this manual. 7.3.2.1 receive packet format. the basic receive packet of information is 17 bits long,16 data bits plus a status bit, as shown below in figure 7-5 . figure 7-3. bdm serial transfer figure 7-4. bdm signal sampling 16 15 0 s data field [15:0] figure 7-5. receive bdm packet clk dsclk 16 15 14 0 dsi 16 15 14 1 0 dso clk dsclk dsi dso f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-9 status[16] the status bit indicates the status of cpu-generated messages (always single word), with the data field encoded as listed in table 7-2 . data field[15:0] the data field contains the message data to be communicated from the debug module to the development system. the response message is always a single word, with the data field encoded as shown in table 7-2 . 7.3.2.2 transmit packet format. the basic transmit packet of information is 17 bits long,16 data bits plus a control bit, as shown below in figure 7-6 . control[16] the control bit is not used but is reserved by motorola for future use. command and data transfers initiated by the development system should clear bit 16. data field[15:0] the data field contains the message data to be communicated from the development system to the debug module. 7.3.3 bdm command set the coldfire2/2m supports a subset of bdm instructions from the current 683xx parts, as well as extensions to provide access to new hardware features. the bdm commands should not be issued whenever the coldfire2/2m is accessing the debug module registers using the wdebug instruction. 7.3.3.1 bdm command set summary. the bdm command set is summarized in figure 7-6 . subsequent paragraphs contain detailed descriptions of each command. table 7-2. cpu-generated message encoding s bit data message type 0 xxxx valid data transfer 0 $ffff command complete; status ok 1 $0000 not ready with response; come again 1 $0001 tea-terminated bus cycle; data invalid 1 $ffff illegal command 16 15 0 c data field [15:0] figure 7-6. transmit bdm packet f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-10 coldfire2/2m users manual motorola 7.3.3.2 coldfire bdm commands. all coldfire family bdm commands include a 16- bit operation word followed by an optional set of one or more extension words. operation field the operation field specifies the command. table 7-3. bdm command summary command mnemonic description cpu impact 1 page read a/d register rareg/rdreg read the selected address or data register and return the results via the serial interface. halted 7-13 write a/d register wareg/wdreg the data operand is written to the specified address or data register. halted 7-13 read memory location read read the data at the memory location specified by the longword address. cycle steal 7-14 write memory location write write the operand data to the memory location specified by the longword address. cycle steal 7-16 dump memory block dump used in conjunction with the read command to dump large blocks of memory. an initial read is executed to set up the starting address of the block and to retrieve the first result. subsequent operands are retrieved with the dump command. cycle steal 7-17 fill memory block fill used in conjunction with the write command to fill large blocks of memory. an initial write is executed to set up the starting address of the block and to supply the first operand. subsequent operands are written with the fill command. cycle steal 7-19 resume execution go the pipeline is flushed and refilled before resuming instruction execution at the current pc . halted 7-21 no operation nop nop performs no operation and may be used as a null command. parallel 7-21 read control register rcreg read the system control register. halted 7-22 write control register wcreg write the operand data to the system control register. halted 7-23 read debug module register rdmreg read the debug module register. parallel 7-24 write debug module register wdmreg write the operand data to the debug module register. parallel 7-25 note: 1. general command effect and/or requirements on cpu operation: halted - the cpu must be halted to perform this command steal - command generates bus cycles which can be interleaved with cpu accesses parallel - command is executed in parallel with cpu activity refer to command summaries for detailed operation descriptions. 15 1098765432 0 operation 0 r/w op size 0 0 a/d register extension word(s) bdm command format f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-11 r/w field the r/w field specifies the direction of operand transfer. when the bit is set, the transfer is from the cpu to the development system. when the bit is cleared, data is written to the cpu or to memory from the development system. operand size for sized operations, this field specifies the operand data size. all addresses are expressed as 32-bit absolute values. the size field is encoded as listed in table 7-4 . address / data (a/d ) field the a/d field is used in commands that operate on address and data registers in the processor. it determines whether the register field specifies a data or address register. a one indicates an address register; zero, a data register. register field in commands that operate on processor registers, this field specifies which register is selected. the field value contains the register number. extension word(s) (as required): certain commands require extension words for addresses and/or immediate data. addresses require two extension words because only absolute long addressing is permitted. immediate data can be either one or two words in lengthbyte and word data each require a single extension word; longword data requires two words. both operands and addresses are transferred most significant word first. in the following descriptions of the bdm command set, the optional set of extension words is defined as the operand data. 7.3.3.3 command sequence diagram. a command sequence diagram (see figure 7-7 ) illustrates the serial bus traffic for each command. each bubble in the diagram represents a single 17-bit transfer across the bus. the top half in each bubble corresponds to the data transmitted by the development system to the debug module; the bottom half corresponds to the data returned by the debug module in response to the development system commands. command and result transactions are overlapped to minimize latency. the cycle in which the command is issued contains the development system command mnemonic (in this example, read memory location). during the same cycle, the debug module responds with either the lowest order results of the previous command or with a command complete status (if no results were required). table 7-4. bdm size field encoding encoding operand size 00 byte 01 word 10 long 11 reserved f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-12 coldfire2/2m users manual motorola during the second cycle, the development system supplies the high-order 16 bits of the memory address. the debug module returns a not ready response ($10000) unless the received command was decoded as unimplemented, in which case the response data is the illegal command ($1ffff) encoding. if an illegal command response occurs, the development system should retransmit the command. note the response can be ignored unless a memory bus cycle is in progress. otherwise, the debug module can accept a new serial transfer after eight system clock periods. in the third cycle, the development system supplies the low-order 16 bits of a memory address. the debug module always returns the not ready response in this cycle. at the completion of the third cycle, the debug module initiates a memory read operation. any serial transfers that begin while the memory access is in progress return the not ready response. results are returned in the two serial transfer cycles following the completion of the memory access. the data transmitted to the debug module during the final transfer is the opcode for the following command. should a memory access generate a bus error, an error status ($10001) is returned in place of the result data. figure 7-7. command sequence diagram commands transmitted to the debug module command code transmitted during this cycle high-order 16 bits of memory address low-order 16 bits of memory address sequence taken if operation has not completed data unused from this transfer sequence taken if illegal command is received by debug module results from previous command responses from the debug module nonserial-related activity ms addr "not ready" xxx "illegal" ls addr "not ready" next cmd "not ready" read (long) ??? next cmd "not ready" xxx "not ready" xxxxx xxx berr ms result next cmd ls result read memory location next command code sequence taken if bus error occurs on memory access high- and low-order 16 bits of result xxx f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-13 7.3.3.4 command set descriptions. the bdm command set is summarized in table 7-3 . subsequent paragraphs contain detailed descriptions of each command. note all the accompanying bdm commands assume the control bit (c) is zero. the bdm results are defined with the status bit (s) as zero. refer to section 7.3.2 bdm serial interface for information on the serial packet format unassigned command opcodes are reserved by motorola for future expansion. all unused command formats within any revision level will perform a nop and return the illegal command response. 7.3.3.4.1 read a/d register (rareg/rdreg). read the selected address or data register and return the 32-bit result. a bus error response is returned if the cpu core is not halted. formats: \ command sequence: operand data: none result data: the contents of the selected register are returned as a longword value. the data is returned most significant word first. 7.3.3.4.2 write a/d register (wareg/wdreg). the operand data is written to the specified address or data register. all 32 register bits are altered by the write. a bus error response is returned if the cpu core is not halted. 1514131211109876543210 $2 $1 $8 a/d register rareg/rdreg command 1514131211109876543210 data [31:16] data [15:0] rareg/rdreg result xxx ms result next cmd ls result rareg/rdreg ??? xxx berr next cmd "not ready" f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-14 coldfire2/2m users manual motorola command formats: command sequence: operand data: longword data is written into the specified address or data register. the data is supplied most significant word first. result data: command complete status will be indicated by returning the data $ffff (with the status bit cleared) when the register write is complete. 7.3.3.4.3 read memory location (read). read the operand data from the memory location specified by the longword address. the address space is defined by the contents of the low-order 5 bits {tt, tm} of the address attribute register ( aatr ). the hardware forces the low-order bits of the address to zeros for word and longword accesses to ensure that operands are always accessed on natural boundaries: words on 0-modulo-2 addresses, longwords on 0-modulo-4 addresses. formats: 1514131211109876543210 $2 $0 $8 a/d register data [31:16] data [15:0] wareg/wdreg command 1514131211109876543210 $1 $9 $0 $0 address [31:16] address [15:0] byte read command 1514131211109876543210 xxxxxxxx data [7:0] byte read result ms data "not ready" xxx ls data "not ready" next cmd "not ready" wdreg/wareg ??? next cmd "cmd complete" berr f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-15 command sequence: operand data: the single operand is the longword address of the requested memory location. 1514131211109876543210 $1 $9 $4 $0 address [31:16] address [15:0] word read command 1514131211109876543210 data [15:0] word read result 1514131211109876543210 $1 $9 $8 $0 address [31:16] address [15:0] long read command 1514131211109876543210 data [31:16] data [15:0] long read result ms addr "not ready" ls addr "not ready" read (b/w) ??? next cmd "not ready" xxx "not ready" xxx xxx berr result next cmd read memory location ms addr "not ready" ls addr "not ready" read (long) ??? next cmd "not ready" xxx "not ready" xxx xxx berr ms result xxx read memory location next cmd ls result f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-16 coldfire2/2m users manual motorola result data: the requested data is returned as either a word or longword. byte data is returned in the least significant byte of a word result, with the upper byte undefined. word results return 16 bits of significant data; longword results return 32 bits. a value of $0001 (with the status bit set) will be returned if a bus error occurs. 7.3.3.4.4 write memory location (write). write the operand data to the memory location specified by the longword address. the address space is defined by the contents of the low-order 5 bits {tt, tm} of the address attribute register ( aatr ). the hardware forces the low-order bits of the address to zeros for word and longword accesses to ensure that operands are always accessed on natural boundaries: words on 0-modulo-2 addresses, longwords on 0-modulo-4 addresses. formats: 1514131211109876543210 $1 $8 $0 $0 address [31:16] address [15:0] xxxxxxxx data [7:0] byte write command 1514131211109876543210 $1 $8 $4 $0 address [31:16] address [15:0] data [15:0] word write command 1514131211109876543210 $1 $8 $8 $0 address [31:16] address [15:0] data [31:16] data [15:0] long write command f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-17 command sequence: operand data: two operands are required for this instruction. the first operand is a longword absolute address that specifies a location to which the operand data is to be written. the second operand is the data. byte data is transmitted as a 16-bit word, justified in the least significant byte; 16- and 32-bit operands are transmitted as 16 and 32 bits, respectively. result data: successful write operations will be indicated by returning the data $ffff (with the status bit cleared) when the register write is complete. a value of $0001 (with the status bit set) will be returned if a bus error occurs. 7.3.3.4.5 dump memory block (dump). dump is used in conjunction with the read command to dump large blocks of memory. an initial read is executed to set up the starting address of the block and to retrieve the first result. the dump command retrieves subsequent operands. the initial address is incremented by the operand size (1, 2, or 4) and saved in a temporary register (address breakpoint high ( abhr )). subsequent dump commands use this address, perform the memory read, increment it by the current operand size, and store the updated address in abhr. ms addr "not ready" ls addr "not ready" write (b/w) ??? next cmd "not ready" xxx "not ready" xxx xxx berr "cmd complete" next cmd write memory location data "not ready" ms addr "not ready" ls addr "not ready" write (long) ??? next cmd "not ready" xxx "not ready" xxx xxx berr "cmd complete" next cmd write memory location ms data "not ready" ls data "not ready" f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-18 coldfire2/2m users manual motorola note the dump command does not check for a valid address in abhr dump is a valid command only when preceded by another dump, nop or by a read command. otherwise, an illegal command response is returned. the nop command can be used for intercommand padding without corrupting the address pointer. the size field is examined each time a dump command is given, allowing the operand size to be dynamically altered. command formats: 1514131211109876543210 $1 $d $0 $0 byte dump command 1514131211109876543210 xxxxxxxx data [7:0] byte dump result 1514131211109876543210 $1 $d $4 $0 word dump command 1514131211109876543210 data [15:0] word dump result 1514131211109876543210 $1 $d $8 $0 long dump command 1514131211109876543210 data [31:16] data [15:0] long dump result f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-19 command sequence: operand data: none result data: requested data is returned as either a word or longword. byte data is returned in the least significant byte of a word result. word results return 16 bits of significant data; longword results return 32 bits. a value of $0001 (with the status bit set) will be returned if a bus error occurs. 7.3.3.4.6 fill memory block (fill). fill is used in conjunction with the write command to fill large blocks of memory. an initial write is executed to set up the starting address of the block and to supply the first operand. the fill command writes subsequent operands. the initial address is incremented by the operand size (1, 2, or 4) and is saved in address breakpoint high ( abhr ) after the memory write. subsequent fill commands use this address, perform the write, increment it by the current operand size, and store the updated address in abhr. note the fill command does not check for a valid address in abhr fill is a valid command only when preceded by another fill, nop or by a write command. otherwise, an illegal command response is returned. the nop command can be used for intercommand padding without corrupting the address pointer. the size field is examined each time a fill command is processed, allowing the operand size to be altered dynamically. xxx "not ready" next cmd result xxx berr xxx "illegal" next cmd "not ready" next cmd "not ready" read memory location dump (b/w) ??? xxx "not ready" next cmd ms result xxx berr xxx "illegal" next cmd "not ready" next cmd "not ready" read memory location dump (long) ??? next cmd ls result f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-20 coldfire2/2m users manual motorola formats: command sequence: operand data: a single operand is data to be written to the memory location. byte data is transmitted as a 16-bit word, justified in the least significant byte; 16- and 32-bit operands are transmitted as 16 and 32 bits, respectively. 1514131211109876543210 $1 $c $0 $0 xxxxxxxx data [7:0] byte fill command 1514131211109876543210 $1 $c $4 $0 data [15:0] word fill command 1514131211109876543210 $1 $c $8 $0 data [31:16] data [15:0] long fill command next cmd "not ready" "not ready" xxx berr "cmd complete" data "not ready" xxx next cmd xxx "illegal" next cmd "not ready" fill (long) ??? write memory location next cmd "not ready" xxx "not ready" xxx berr "cmd complete" ms data "not ready" next cmd xxx "illegal" next cmd "not ready" ls data "not ready" write memory location fill (b/w) ??? f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-21 result data: command complete status will be indicated by returning the data $ffff (with the status bit cleared) when the register write is complete. a value of $0001 (with the status bit set) will be returned if a bus error occurs. 7.3.3.4.7 resume execution (go). the pipeline is flushed and refilled before resuming normal instruction execution. prefetching begins at the current pc and current privilege level. if either the pc or sr is altered during bdm, the updated value of these registers is used when prefetching begins. formats: command sequence: operand data: none result data: the command complete response ($0ffff) is returned during the next shift operation. 7.3.3.4.8 no operation (nop). nop performs no operation and may be used as a null command where required. formats: command sequence: 1514131211109876543210 $0 $c $0 $0 go command 15 12 11 8 7 4 3 0 $0 $0 $0 $0 nop command go ??? next cmd "cmd complete" nop ??? next cmd "cmd complete" f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-22 coldfire2/2m users manual motorola operand data: none result data: the command complete response, $ffff (with the status bit cleared), is returned during the next shift operation. 7.3.3.4.9 read control register (rcreg). read the selected control register and return the 32-bit result. accesses to the processor/memory control registers are always 32 bits in size, regardless of the implemented register width. the second and third words of the command effectively form a 32-bit address used by the debug module to generate a special bus cycle to access the specified control register. the 12-bit rc field is the same as that used by the movec instruction. formats: rc encoding: 1514131211109876543210 $2 $9 $8 $0 $0 $0 $0 $0 $0 rc rcreg command 1514131211109876543210 data [31:16] data [15:0] rcreg result table 7-5. control register map rc register definition $002 cache control register ( cacr ) $004 access control register 0 ( acr0 ) $005 access control register 1 ( acr1 ) $801 vector base register ( vbr ) $804 mac status register ( macsr )? $805 mac mask register ( mask )? $806 mac accumulator ( acc )? $80e status register ( sr ) $80f program counter ( pc ) $c00 rom base address register ( rombar0 ) $c04 ram base address register ( rambar0 ) note: ?available on the coldfire2m only. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-23 command sequence: operand data: the single operand is the 32-bit rc control register select field. result data: the contents of the selected control register are returned as a longword value. the data is returned most-significant-word first. for those control registers with widths less than 32 bits, only the implemented portion of the register is guaranteed to be correct. the remaining bits of the longowrd are undefined. as an example, a read of the 16-bit sr will return the sr in the lower word and undefined data in the upper word. 7.3.3.4.10 write control register (wcreg). the operand (longword) data is written to the specified control register. the write alters all 32 register bits. formats: 1514131211109876543210 $2 $8 $8 $0 $0 $0 $0 $0 $0 rc data [31:16] data [15:0] wcreg command ext word "not ready" ext word "not ready" rcreg ??? next cmd "not ready" xxx "not ready" xxx xxx berr ms result read memory location next cmd ls result xxx f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-24 coldfire2/2m users manual motorola command sequence: operand data: two operands are required for this instruction. the first long operand selects the register to which the operand data is to be written. the second operand is the data. result data: successful write operations return $ffff. bus errors on the write cycle are indicated by the assertion of bit 16 in the status message and by a data pattern of $0001. 7.3.3.4.11 read debug module register (rdmreg). read the selected debug module register and return the 32-bit result. the only valid register selection for the rdmreg command is the csr (drc = $0). command formats: drc encoding: 1514131211109876543210 $2 $d $8 drc rdmreg bdm command 1514131211109876543210 data [31:16] data [15:0] rdmreg bdm result table 7-6. definition of drc encoding - read drc[3:0] debug register definition mnemonic initial state $0 configuration/status csr $0 $1-$f reserved - C ext word "not ready" ext word "not ready" wcreg ??? next cmd "not ready" xxx "not ready" xxx xxx berr "cmd complete" next cmd write memory location ms data "not ready" ls data "not ready" f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-25 command sequence: operand data: none result data: the contents of the selected debug register are returned as a longword value. the data is returned most-significant-word first. 7.3.3.4.12 write debug module register (wdmreg). the operand (longword) data is written to the specified debug module register. all 32 bits of the register are altered by the write. the dsclk signal must be inactive while cpu accesses are being performed. command format: drc encoding: 1514131211109876543210 $2 $c $8 drc data [31:16] data [15:0] wdmreg bdm command table 7-7. definition of drc encoding - write drc[3:0] debug register definition mnemonic initial state $0 configuration/status csr $0 $1-$5 reserved - C $6 bus attributes and mask aatr $0005 $7 trigger definition tdr $0 $8 pc breakpoint pbr C $9 pc breakpoint mask pbmr C $a-$b reserved C C $c operand address high breakpoint abhr C $d operand address low breakpoint ablr C $e data breakpoint dbr C $f data breakpoint mask dbmr C xxx ms result xxx "illegal" next cmd ls result next cmd "not ready" rdmreg ??? f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-26 coldfire2/2m users manual motorola command sequence: operand data: longword data is written into the specified debug register. the data is supplied most- significant-word first. result data: command complete status ($0ffff) is returned when register write is complete. 7.3.3.4.13 unassigned opcodes. unassigned command opcodes are reserved by motorola. all unused command formats within any revision level will perform a nop and return the illegal command response . 7.4 real-time debug support the coldfire2/2m provides support for the debug of real-time applications. for these types of embedded systems, the processor cannot be halted during debug, but must continue to operate. the foundation of this area of debug support is that while the processor cannot be halted to allow debug, the system can tolerate small intrusions into the real-time operation. as discussed in the previous section, the debug module provides a number of hardware resources to support various hardware breakpoint functions. specifically, three types of breakpoints are supported: pc with mask, operand address range, and data with mask. these three basic breakpoints can be configured into one- or two-level triggers with the exact trigger response also programmable. the debug module programming model is accessible from either the external development system using the serial interface or from the processors supervisor programming model using the wdebug instruction. 7.4.1 theory of operation the breakpoint hardware can be configured to respond to triggers in several ways. the desired response is programmed into the trigger definition register. in all situations where a breakpoint triggers, an indication is provided on the ddata output port, when not displaying captured operands or branch addresses, as shown in table 7-8 . ms data "not ready" xxx "illegal" ls data "not ready" next cmd "not ready" wdmreg ??? next cmd "cmd complete" f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-27 the breakpoint status is also posted in the csr . the new bdm instructions load and configure the desired breakpoints using the appropriate registers. as the system operates, a breakpoint trigger generates a response as defined in the tdr . if the system can tolerate the processor being halted, a bdm entry can be used. with the trc bits of the tdr = $1, the breakpoint trigger causes the core to halt as reflected in the status of pst = $f. for pc breakpoints, the halt occurs before the targeted instruction is executed. for address and data breakpoints, the processor may have executed several additional instructions. as a result, trigger reporting is considered imprecise. if the processor core cannot be halted, the special debug interrupt can be used. with this configuration, trc bits of the tdr = $2 , the breakpoint trigger is converted into a debug interrupt to the processor (see section 4.2.8 debug interrupt .) this interrupt is treated as higher than the nonmaskable level 7 interrupt request. as with all interrupts, it is made pending when the processor samples, once per instruction. again, the hardware forces the pc breakpoint to occur immediately (before the execution of the targeted instruction). this is possible because the pc breakpoint comparison is enabled at the same time the interrupt sampling occurs. for the address and data breakpoints, the reporting is considered imprecise because several additional instructions may be executed after the triggering address or data is seen. once the debug interrupt is recognized, the processor aborts execution and initiates exception processing. at the initiation of the exception processing, the core enters emulator mode. after the standard 8-byte exception stack is created, the processor fetches a unique exception vector, $c, from the vector table. execution continues at the instruction address contained in this exception vector. all interrupts are ignored while in emulator mode. users can program the debug-interrupt handler to perform the necessary context saves using the supervisor instruction set. as an example, this handler may save the state of all the program-visible registers as well as the current context into a reserved memory area. once the required operations are complete, the return-from-exception (rte) instruction is executed and the processor exits emulator mode. once the debug interrupt handler has completed its execution, the external development system can then access the reserved memory locations using the bdm commands to read memory. table 7-8. ddata, csr[31:28] breakpoint response ddata[3:0], csr[31:28] breakpoint status $0 no breakpoints enabled $1 waiting for level 1 breakpoint $2 level 1 breakpoint triggered $3-4 reserved $5 waiting for level 2 breakpoint $6 level 2 breakpoint triggered $7-$f reserved f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-28 coldfire2/2m users manual motorola 7.4.1.1 emulator mode. emulator mode is used to facilitate non-intrusive emulator functionality. this mode can be entered in three different ways: ? the emu bit in the configuration/status register ( csr ) may be programmed to force the coldfire2/2m into emulation mode. this bit is examined only when mrstb is negated and the processor begins reset exception processing. it may be set while the coldfire2/ 2m is halted before the reset exception processing begins. refer to section 7.3.1 cpu halt . ? a debug interrupt always enters emulation mode when the debug interrupt exception processing begins. ? the tcr bit in the csr may be programmed to force the coldfire2/2m into emulation mode when trace exception processing begins. during emulation mode, the coldfire2/2ms exhibits the following properties: ? all interrupts are ignored, including level seven. ? if the map bit in the csr is set, all memory accesses are forced, including the exception stack frame writes and the vector fetch, into a specially mapped address space signalled by tt = $2, tm = $5 or $6. ? if the map bit in the csr is set, all caching of memory accesses is disabled. the return-from-exception (rte) instruction exits emulation mode. the processor status output port provides a unique encoding for emulator mode entry ($d) and exit ($7). 7.4.1.2 reuse of debug module hardware. the debug module implementation provides a common hardware structure for both bdm and breakpoint functionality. several structures are used for both bdm and breakpoint purposes. table 7-9 identifies the shared hardware structures. the shared use of these hardware structures means the loading of the register to perform any specified function is destructive to the shared function. for example, if an operand address breakpoint is loaded into the debug module, a bdm command to access memory overwrites the breakpoint. if a data breakpoint is configured, a bdm write command overwrites the breakpoint contents. 7.4.2 programming model in addition to the existing bdm commands that provide access to the processors registers and the memory subsystem, the debug module contains nine registers to support the table 7-9. shared bdm/breakpoint hardware register bdm function breakpoint function aatr bus attributes for all memory commands attributes for address breakpoint abhr address for all memory commands address for address breakpoint dbr data for all bdm write commands data for data breakpoint f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-29 required functionality. all of these registers are treated as 32-bit quantities, regardless of the actual number of bits in the implementation. the registers, known as the debug control registers, are accessed through the bdm port using two new bdm commands: wdmreg and rdmreg . these commands contain a 4-bit field, drc, which specifies the particular register being accessed. these registers are also accessible from the processors supervisor programming model through the execution of the wdebug instruction. thus, the breakpoint hardware within the debug module may be accessed by the external development system using the serial interface, or by the operating system running on the processor core. it is the responsibility of the software to guarantee that all accesses to these resources are serialized and logically consistent. the hardware provides a locking mechanism in the csr to allow the external development system to disable any attempted writes by the processor to the breakpoint registers (setting ipw = 1). the bdm commands should not be issued whenever the coldfire2/2m is accessing the debug module registers using the wdebug instruction. figure 7-8 illustrates the debug module programming model. 7.4.2.1 address breakpoint registers (ablr, abhr). the address breakpoint registers define a region in the operand address space of the processor that can be used as part of the trigger. the full 32-bits of the ablr and abhr values are compared with the internal address signals of the coldfire2/2m. the trigger definition register ( tdr ) determines if the trigger is the inclusive range bound by ablr and abhr, all addresses outside this range, or the address in ablr only. the abhr is accessible in supervisor mode as debug control register $c using the wdebug instruction and via the bdm port using the rdmreg and wdmreg commands. the ablr is accessible in supervisor mode as debug control register $d using the wdebug instruction and via the bdm port using the wdmreg commands. the abhr is overwritten by the bdm hardware when accessing memory as described in section 7.4.1.2 reuse of debug module hardware . figure 7-8. debug programming model address breakpoint registers pc breakpoint registers data breakpoint registers ablr abhr pbr pbmr dbmr dbr tdr 15 0 31 trigger definition register address attribute register aatr 7 0 15 csr configuration/ status f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-30 coldfire2/2m users manual motorola field definition: address[31:0]Clow address this field contains the 32-bit address which marks the lower bound of the address breakpoint range. field definition: address[31:0]Chigh address this field contains the 32-bit address which marks the upper bound of the address breakpoint range. 7.4.2.2 address attribute register (aatr). the aatr defines the address attributes and a mask to be matched in the trigger. the aatr value is compared with the internal address attribute signals of the coldfire2/2m, as defined by the setting of the tdr . the aatr is accessible in supervisor mode as debug control register $6 using the wdebug instruction and via the bdm port using the wdmreg command. the lower five bits of the aatr are also used for bdm command definition to define the address space for memory references as described in section 7.4.1.2 reuse of debug module hardware . bits 31 0 field address reset - r/w w address breakpoint low register (ablr) bits 31 0 field address reset - r/w w address breakpoint high register (abhr) bits 15 14 13 12 11 10 8 7 6 5 4 3 2 0 field rm szm ttm tmm r sz tt tm reset 00 0 0 00 0 101 r/w ww w w ww w w address attribute register (aatr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-31 field definitions: rm[15]Cread/write mask this field corresponds to the r-field. setting this bit causes r to be ignored in address comparisons. szm[14:13]Csize mask this field corresponds to the sz field. setting a bit in this field causes the corresponding bit in sz to be ignored in address comparisons. ttm[12:11]Ctransfer type mask this field corresponds to the tt field. setting a bit in this field causes the corresponding bit in tt to be ignored in address comparisons. tmm[10:8]Ctransfer modifier mask this field corresponds to the tm field. setting a bit in this field causes the corresponding bit in tm to be ignored in address comparisons. r[7]Cread/write this field is compared with the internal r/w signal of the coldfire2/2m. a high level indicates a read cycle and a low level indicates a write cycle. sz[6:5]size this field is compared to the internal size signals of the coldfire2/2m. these signals indicate the data size for the bus transfer. 00 = longword 01 = byte 10 = word 11 = reserved tt[4:3]transfer type this field is compared with the internal tt signals of the coldfire2/2m. these signals indicate the transfer type for the bus transfer. these signals are always encoded as if the coldfire2/2m is in the coldfire iack mode (see section 3.1.15 master transfer type (mtt[1:0]) .) 00 = coldfire2/2m access 01 = reserved 10 = emulator mode access 11 = acknowledge/cpu space access these bits also define the tt encoding for bdm memory commands. in this case, the 01 encoding generates an alternate master access. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-32 coldfire2/2m users manual motorola tm[2:0]transfer modifier this field is compared with the internal tm signals of the coldfire2/2m. these signals provide supplemental information for each transfer type. these signals are always encoded as if the coldfire2/2m is in the coldfire iack mode (see section 3.1.13 master transfer modifier (mtm[2:0]) .) the encoding for normal coldfire2/2m transfers is: 000 = reserved 001 = user data access 010 = user code access 011 = reserved 100 = reserved 101 = supervisor data access 110 = supervisor code access 111 = reserved the encoding for emulator mode transfers is: 0xx = reserved 100 = reserved 101 = emulator mode data access 110 = emulator mode code access 111 = reserved the encoding for acknowledge/cpu space transfers is: 000 = cpu space access 001 = interrupt acknowledge level 1 010 = interrupt acknowledge level 2 011 = interrupt acknowledge level 3 100 = interrupt acknowledge level 4 101 = interrupt acknowledge level 5 110 = interrupt acknowledge level 6 111 = interrupt acknowledge level 7 these bits also define the tm encoding for bdm memory commands. 7.4.2.3 program counter breakpoint register (pbr, pbmr). the pc breakpoint registers define a region in the code address space of the processor that can be used as part of the trigger. the pbr value is masked by the pbmr value, allowing only those bits in pbr that have a corresponding zero in pbmr to be compared with the processors program counter register, as defined in the tdr . the pbr is accessible in supervisor mode as debug control register $8 using the wdebug instruction and via the bdm port using the rdmreg and wdmreg commands. the pbmr is accessible in supervisor mode as debug control register $9 using the wdebug instruction and via the bdm port using the wdmreg command. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-33 field definition: address[31:0]Cpc breakpoint address this field contains the 32-bit address to be compared with the pc as a breakpoint trigger. field definition: mask[31:0]Cpc breakpoint mask this field contains the 32-bit mask for the pc breakpoint trigger. a zero in a bit position causes the corresponding bit in the pbr to be compared to the appropriate bit of the pc. a one causes that bit to be ignored. 7.4.2.4 data breakpoint register (dbr, dbmr). the data breakpoint registers define a specific data pattern that can be used as part of the trigger into debug mode.the dbr value is masked by the dbmr value, allowing only those bits in dbr that have a corresponding zero in dbmr to be compared with internal data bus of the coldfire2/2m, as defined in the tdr . the dbr is accessible in supervisor mode as debug control register $e using the wdebug instruction and via the bdm port using the rdmreg and wdmreg commands. the dbmr is accessible in supervisor mode as debug control register $f using the wdebug instruction and via the bdm port using the wdmreg command. the dbr is overwritten by the bdm hardware when accessing memory as described in section 7.4.1.2 reuse of debug module hardware . bits 31 0 field address reset - r/w w program counter breakpoint register (pbr) bits 31 0 field mask reset - r/w w program counter breakpoint mask register (pbmr) bits 31 0 field address reset - r/w w data breakpoint register (dbr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-34 coldfire2/2m users manual motorola field definition: address[31:0]Cdata breakpoint value this field contains the 32-bit value to be compared with the internal data bus as a breakpoint trigger. field definition: mask[31:0]Cdata breakpoint mask this field contains the 32-bit mask for the data breakpoint trigger. a zero in a bit position causes the corresponding bit in the dbr to be compared to the appropriate bit of the internal data bus. a one causes that bit to be ignored. the data breakpoint register supports both aligned and misaligned operand references. the relationship between the processor address, the access size, and the corresponding location within the 32-bit data bus is shown in table 7-10 . 7.4.2.5 trigger definition register (tdr). the tdr configures the operation of the hardware breakpoint logic within the debug module and controls the actions taken under the defined conditions. the breakpoint logic may be configured as a one- or two-level trigger, where bits [31:16] of the tdr define the 2nd level trigger and bits [15:0] define the first level trigger. the tdr is accessible in supervisor mode as debug control register $7 using the wdebug instruction and via the bdm port using the wdmreg command. bits 31 0 field mask reset - r/w w data breakpoint mask register (dbmr) table 7-10. misaligned data operand references address[1:0] access size operand location 00 byte data[31:24] 01 byte data[23:16] 10 byte data[15:8] 11 byte data[7:0] 00 word data[31:16] 10 word data[15:0] 00 long data[31:0] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-35 field definitions: trcCtrigger response control the trigger response control determines how the processor is to respond to a completed trigger condition. the trigger response is always displayed on the ddata pins. 00 = display on ddata only 01 = processor halt 10 = debug interrupt 11 = reserved eblCenable breakpoint level if set, this bit serves as the global enable for the breakpoint trigger. if cleared, all breakpoints are disabled. edlwCenable data breakpoint for the data longword if set, this bit enables the data breakpoint based on the entire internal data bus. the assertion of any of the ed bits enables the data breakpoint. if all bits are cleared, the data breakpoint is disabled. edwlCenable data breakpoint for the lower data word if set, this bit enables the data breakpoint based on the low-order word of the internal data bus. edwuCenable data breakpoint for the upper data word if set, this bit enables the data breakpoint trigger based on the high-order word of the internal data bus. bits 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 field trc ebl edl w edw l edw u edl l edl m edu m edu u di eai ear eal epc pci rese t 0 00000000000000 r/w w wwwwwwwwwwwwww bits 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 field - ebl edl w edw l edw u edl l edl m edu m edu u di eai ear eal epc pci rese t - 00000000000000 r/w - wwwwwwwwwwwwww trigger definition register (tdr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-36 coldfire2/2m users manual motorola edllCenable data breakpoint for the lower lower data byte if set, this bit enables the data breakpoint trigger based on the low-order byte of the low- order word of the internal data bus. edlmCenable data breakpoint for the lower middle data byte if set, this bit enables the data breakpoint trigger based on the high-order byte of the low- order word of the internal data bus. edumCenable data breakpoint for the upper middle data byte if set, this bit enables the data breakpoint trigger on the low-order byte of the high-order word of the internal data bus. eduuCenable data breakpoint for the upper upper data byte if set, this bit enables the data breakpoint trigger on the high-order byte of the high-order word of the internal data bus. diCdata breakpoint invert this bit provides a mechanism to invert the logical sense of all the data breakpoint comparators. this can develop a trigger based on the occurrence of a data value not equal to the one programmed into the dbr . eaiCenable address breakpoint inverted if set, this bit enables the address breakpoint based outside the range defined by ablr and abhr . the assertion of any of the ea bits enables the address breakpoint. if all three bits are cleared, this breakpoint is disabled. earCenable address breakpoint range if set, this bit enables the address breakpoint based on the inclusive range defined by ablr and abhr . ealCenable address breakpoint low if set, this bit enables the address breakpoint based on the address contained in the ablr . epcCenable pc breakpoint if set, this bit enables the pc breakpoint. pciCpc breakpoint invert if set, this bit allows execution outside a given region as defined by pbr and pbmr to enable a trigger. if cleared, the pc breakpoint is defined within the region defined by pbr and pbmr. 7.4.2.6 configuration/status register (csr). the csr defines the operating configuration for the processor and memory subsystem. in addition to defining the microprocessor configuration, this register also contains status information from the breakpoint logic. the csr is cleared during system reset. the csr can be read and written to by the external development system and written to by the supervisor programming model. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support motorola coldfire2/2m users manual 7-37 the csr is accessible in supervisor mode as debug control register $0 using the wdebug instruction and via the bdm port using the rdmreg and wdmreg commands. field definitions: status[31:28]Cbreakpoint status this 4-bit field provides read-only status information concerning the hardware breakpoints. this field is defined as follows: 000x = no breakpoints enabled 001x = waiting for level 1 breakpoint 010x = level 1 breakpoint triggered 101x = waiting for level 2 breakpoint 110x = level 2 breakpoint triggered the breakpoint status is also output on the ddata port when not busy displaying other coldfire2/2m data. a write to the tdr resets this field. fof[27]Cfault-on-fault if this read-only status bit is set, a catastrophic halt has occurred and forced entry into bdm. this bit is cleared on a read from the csr. trg[26]Chardware breakpoint trigger if this read-only status bit is set, a hardware breakpoint has halted the processor core and forced entry into bdm. this bit is cleared by reading csr, or when the processor is restarted. bits 31 28 27 26 25 24 23 17 16 field status fof trg hal t bkp t - ipw rese t 0 0000 - 0 r/w 1 r rrrr - r/w bits 15 14 13 12 11 10 9 8 7 6 5 4 3 0 field map trc emu ddc uhe btb - npl ipi ssm - rese t 000 0 0 0 0000 - r/w ? r/w r/w r/w r/w r/w r/w r r/w r/w r/w - note: ?the csr is a write only register from the programming model. it can be read from and written to via the bdm port. configuration/status register (csr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-38 coldfire2/2m users manual motorola halt[25]Cprocessor halt if this read-only status bit is set, the processor has executed the halt instruction and forced entry into bdm. this bit is cleared by reading the csr, or when the processor is restarted. bkpt[24]Cbreakpoint assert if this read-only status bit is set, the bkptb signal was asserted, forcing the processor into bdm. this bit is cleared on a read from the csr, or when the processor is restarted. ipw[16]Cinhibit processor writes to debug registers if set, this bit inhibits any processor-initiated writes to the debug modules programming model registers. this bit can only be modified by commands from the external development system. map[15]Cforce processor references in emulator mode if set, this bit forces the processor to map all references while in emulator mode to a special address space, tt = $2, tm = $5 (data) or $6 (text). if cleared, all emulator-mode references are mapped into supervisor code and data spaces. trc[14]Cforce emulation mode on trace exception if set, this bit forces the processor to enter emulator mode when a trace exception occurs. emu[13]Cforce emulation mode if set, this bit forces the processor to begin execution in emulator mode when a trace exception occurs. refer to section 7.4.1.1 emulator mode . ddc[12:11]Cdebug data control this 2-bit field provides configuration control for capturing operand data for display on the ddata port. the encoding is: 00 = no operand data is displayed 01 = capture all m-bus write data 10 = capture all m-bus read data 11 = capture all m-bus read and write data in all cases, the ddata port displays the number of bytes defined by the operand reference size, i.e., byte displays 8 bits, word displays 16 bits, and long displays 32 bits (one nibble at a time across multiple clock cycles.) refer to section 7.2.1.7 begin data transfer (pst = $8 - $a) . uhe[10]-user halt enable this bit selects the cpu privilege level required to execute the halt instruction. 0 = halt is a privileged, supervisor-only instruction 1 = halt is a non-privileged, supervisor/user instruction f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-39 coldfire2/2m users manual motorola btb[8:9]Cbranch target bytes this 2-bit field defines the number of bytes of branch target address to be displayed on the ddata outputs. the encoding is 00 = 0 bytes 01 = lower two bytes of the target address 10 = lower three bytes of the target address 11 = entire four-byte target address refer to section 7.2.1.5 begin execution of taken branch (pst = $5) . npl[6]Cnon-pipelined mode if set, this bit forces the processor core to operate in a nonpipeline mode of operation. in this mode, the processor effectively executes a single instruction at a time with no overlap. ipi[5]Cignore pending interrupts if set, this bit forces the processor core to ignore any pending interrupt requests signalled while executing in single-instruction-step mode. ssm[4]Csingle-step mode if set, this bit forces the processor core to operate in a single-instruction-step mode. while in this mode, the processor executes a single instruction and then halts. while halted, any of the bdm commands may be executed. on receipt of the go command, the processor executes the next instruction and then halts again. this process continues until the single- instruction-step mode is disabled. reserved - all bits labelled reserved or 0 are currently unused and are reserved for future use. these bits should always be written as 0. 7.4.3 concurrent bdm and processor operation the debug module supports concurrent operation of both the processor and most bdm commands. bdm commands may be executed while the processor is running, except for the operations that access processor/memory registers: ? read/write address and data registers ? read/write control registers for bdm commands that access memory, the debug module requests the coldfire2/2ms internal bus. the processor responds by stalling the instruction fetch pipeline and then waiting until all current bus activity is complete. at that time, the processor relinquishes the internal bus to allow the debug module to perform the required operation. after the conclusion of the debug module bus cycle, the processor reclaims ownership of the bus. by implementing this scheduling mechanism, the processor can minimize the amount of intrusion caused by debug module requests. under certain conditions, the processor may never grant the processor's internal bus to the debug module causing the bdm command to never be performed. specifically, the f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-40 coldfire2/2m users manual motorola processor's internal bus grant may be withheld from the debug module if the processor is executing a tight loop where the entire loop is contained within one aligned longword. examples include: align 4 label1: nop bra.b label1 align 4 label2: bra.w label2 the solution to this scheduling problem is to force the loop to be aligned across a longword boundary. given this alignment, the processor will always correctly grant the processor's internal bus to the debug module. the development system must use caution in configuring the breakpoint registers if the processor is executing. the debug module does not contain any hardware interlocks, so motorola recommends that the tdr be disabled while the breakpoint registers are being loaded. at the conclusion of this process, the tdr can be written to define the exact trigger. this approach guarantees that no spurious breakpoint triggers will occur. because there are no hardware interlocks in the debug unit, no bdm operations are allowed while the cpu is writing the debug registers ( dsclk must be inactive). 7.4.4 motorola recommended bdm pinout the coldfire bdm connector is a 26-pin berg connector arranged 2x13, shown in figure 7-9 . f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-41 coldfire2/2m users manual motorola table 7-11 shows the correlation between the standard coldfire connector and the coldfire2/2ms signals. 7.4.5 differences between the coldfire2/2m bdm and cpu32 bdm 1. dsclk , bkpt , and dsi need to meet the setup and hold times relative to the rising edge of the processor clock to prevent the processor from propagating metastable states. notes: 1. supplied by target 2. pins reserved for bdm developer use. contact developer. 3. * denotes a vectored signal. figure 7-9. recommended bdm connector table 7-11. bdm connector correlation connector coldfire2/2m connector coldfire2/2m +5v +5v dso dso bkpt bkptb gnd gnd clk_cpu clk pst0 pst [0] ddata0 ddata [0] pst1 pst [1] ddata1 ddata [1] pst2 pst [2] ddata2 ddata [2] pst3 pst [3] ddata3 ddata [3] reset mrstb dsclk dsclk tea mteab dsi dsi vcc_cpu v dd 1 3 5 7 9 11 13 15 17 19 21 23 25 2 4 6 8 10 12 14 16 18 20 22 24 26 e veloper reserved2 gnd gnd reset* +5v1 gnd pst2 pst0 ddata2 ddata0* m otorola reserved gnd vcc_cpu bkpt dsclk developer reserv e dsi dso pst3* pst1 ddata3* ddata1 gnd motorola reserve d clk_cpu tea f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
debug support 7-42 coldfire2/2m users manual motorola 2. dso transitions relative to the rising edge of dsclk only. in the cpu32 bdm, dso transitions between serial transfers to indicate to the development system that a command has successfully completed. the coldfire bdm does not support this feature. 3. the development system must note that the dso is not valid during the first rising edge of dsclk . instead, the first rising edge of dsclk causes dso to transmit the msb of dso. a serial transfer is illustrated in figure 7-3 . f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 8-1 section 8 test operation the coldfire2/2m test stategy includes: 1. at-speed parallel scan testing via 16 parallel scan chains to provide high fault coverage 2. the use of an i/o test ring to enable scan testing of the embedded cpu and the surrounding asic logic 3. a methodology to provide testing of the integrated memories connected to the cpu with a minimal required pinout. the following sections describe the signals and methodologies for: ? scan testing ? integrated memory testing iddq testing can also be performed contingent upon the ability to constrain the design to avoid conditions that cause high leakage currents. 8.1 signals required to perform scan test this section describes the coldfire2/2m signals dedicated to the scan testing of the coldfire2/2m. these signals are required to be muxed out to package pins. all coldfire2/ 2m signals are unidirectional and synchronous. 8.1.1 scan enable (scan_enable) this active-high input signal enables scan testing of the coldfire2/2m. it forces all internal flip-flops to be linked together into sixteen parallel scan chains. 8.1.2 scan exercise array (scan_xarray) this active-high input signal is used to exercise the integrated memory arrays during scan testing. this signal causes random writes to the internal rams by strobing the write strobes while scanning. the array output data does not affect the scan via use of the scan_mode pin. 8.1.3 scan input (si[15:0]) these input signals are connected to the16 internal coldfire2/2m scan chain inputs. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-2 coldfire2/2m users manual motorola 8.1.4 scan mode (scan_mode). this active-high, unidirection input signal gates off all memory array outputs during scan testing.scan_mode must be asserted for the duration of scan testing. 8.1.5 scan output (so[15:0]) these output signals are connected to the 16 internal coldfire2/2m scan chain outputs. 8.1.6 i/o test ring clock (trclk) this input signal is the synchronous clock used to transition the test ring during scan testing. tr_clk is connected to the clock input of all i/o test ring registers. 8.1.7 i/o test ring core mode enable (core_test) this active-high input signal enables the core mode of the test ring during scan testing. the test ring is in scan core mode if core_test is asserted and in scan asic mode if core_test is negated. 8.1.8 i/o test ring data input (tr_sdi[1:0]) these input signals are the serial data inputs for the i/o test ring chains. 8.1.9 i/o test ring data output (tr_sdo[1:0]) these output signals are the serial output data from the i/o test ring chains. 8.1.10 i/o test ring enable (tr_se) this active-high input signal enables the test ring. tr_se is connected to the scan enable input of all i/o test ring scannable registers. 8.1.11 i/o test ring mode (tr_mode) this active-high input signal enables the scan test mode of the test ring. the test ring is in scan test mode if tr_mode is asserted and in normal functional mode if negated. tr_mode should be asserted for the duration of scan testing, and be held negated for the duration of memory testing and during functional operation of the device. 8.1.12 test write inhibit (test_wr_inh). optional: asserting this signal will prevent strobing; i.e. writing, of any of the integrated memories. however, as long as scan_mode is asserted, the array outputs are gated off from the coldfire2/2m and will not affect the scan vectors. 8.2 scan operation motorola provides atpg vectors for the coldfire2/2m. the signals listed in the previous section must be brought out to package pins (muxed out) in order for motorola supplied scan vectors to be applied. this section provides an understanding of the coldfire2/2m scan implementation. the following diagram illustrates the coldfire2/2m scan test implementation: f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation motorola coldfire2/2m users manual 8-3 figure 8-1. scan test implementation diagram si/so refer to the 16 parallel scan chains that cover the registers in coldfire2/ 2m.scan_en controls these chains. if these pins, plus all other inputs and outputs of coldfire2/2m were muxed to package pins, atpg coverage of the core would be enabled. the i/o test ring allows stimulus to be applied to embedded coldfire2/2m inputs and allows coldfire2/2m embedded outputs to be observed with a minimum of pins muxed to the chip periphery. si[15:0] is used to load the coldfire2/2m parallel scan chains while tr_sdi[1:0] set up the appropriate values on the coldfire2/2m inputs for each scan vector. so[15:0] and tr_sdo[1:0] are used to observe and verify the scan behavior. the i/o test ring isolates the asic logic from the coldfire2/2m enabling separate scan testing of each. it is comprised of two scan chains, each containing 44 registers (half scannable, half not; see table 8-3 , table 8-4 and figure 8-2). the heads of the chains are tr_sdi[0] and tr_sdi[1]. the tails of the chains are tr_sdo[0] and tr_sdo[1]. the i/o test ring also enables input and output spec testing as follows: the clocks (tr_clk and clk) can be skewed to verify that a transition launched from the scan ring is captured at the input register in a specified period of time. this is then verified by capturing outputs of the coldfire2/2m and shifting the scan data out to compare to expected values. this is also how output specs are verified: the clocks can be skewed a specified amount, and an output transition can be captured by the scan ring. again, the data is shifted out and values compared to expected values. see table 8-1 for scan chain i/o and length information. so[15:0] s1[15:0] output input output output output input input input tr_sdi[1:0] tr_sdo[1:0] test ring cf2 cpu coldfire2/2m f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-4 coldfire2/2m users manual motorola in functional mode, the i/o test ring adds a 2:1 multiplexor delay to the inputs and outputs of the core. each cell in the i/o test ring pairs an input and an output which allows for the sharing of registers. figure 8-2. i/o test ring cell table 8-1. parallel scan chain information scan chain designator scan chain input scan chain output length 0 si[0] so[15] 99 1 si[1] so[14] 99 2 si[2] so[13] 99 3 si[3] so[12] 99 4 si[4] so[11] 99 5 si[5] so[10] 99 6 si[6] so[9] 99 7 si[7] so[8] 99 8 si[8] so[7] 99 9 si[9] so[6] 99 10 si[10] so[5] 99 11 si[11] so[4] 99 12 si[12] so[3] 99 13 si[13] so[2] 99 14 si[14] so[1] 99 15 si[15] so[0] 84 i/o test ring 0 trsdi[0] trsdo[0] 88 i/o test ring 1 trsdi[1] trsdo[1] 88 trsdi trsdo cf2 core asic core_in 0 1 ff2 ff1 d q ck d sdi se ck q 0 1 0 1 asic_out asic_in core_out core_test tr_se trclk tr_mode f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation motorola coldfire2/2m users manual 8-5 the i/o test ring behaves in the same manner for the asic logic as it does for coldfire2/ 2m. table 8-2 lists the i/o test ring control signal configurations for the different modes of testing. table 8-3 and table 8-4 list each signal, which i/o test ring chain it is part of, and how deep (how many cells from trsdi[1:0] it is in the chain. note that an x means that the signal needs to be driven, however high or low is optional. the actual value is a dont care. * needs to driven high or low. the actual value is a dont care. table 8-2. i/o test ring mode configurations load scan coldfire2/2m scan asic shift out functional (non-scan) core_test *x 1 0 *x *x tr_se 1 0 0 1 *x trclk on on on on *x tr_mode *x 1 1 *x 0 tr_sdi[1:0] apply data *x *x *x *x table 8-3. i/o test ring chain 0 (trsdi[0]/trsdo[0]) core input core output cell # (2 regs/ cell) core input core output cell # (2 regs/ cell) mrdata_ie mwdata_oe head of i/o test ring chain 0 mrdata[21] mwdata[21] 23 mrdata[0] mwdata[0] 2 mrdata[22] mwdata[22] 24 mrdata[1] mwdata[1] 3 mrdata[23] mwdata[23] 25 mrdata[2] mwdata[2] 4 mrdata[24] mwdata[24] 26 mrdata[3] mwdata[3] 5 mrdata[25] mwdata[25] 27 mrdata[4] mwdata[4] 6 mrdata[26] mwdata[26] 28 mrdata[5] mwdata[5] 7 mrdata[27] mwdata[27] 29 mrdata[6] mwdata[6] 8 mrdata[28] mwdata[28] 30 mrdata[7] mwdata[7] 9 mrdata[29] mwdata[29] 31 mrdata[8] mwdata[8] 10 mrdata[30] mwdata[30] 32 mrdata[9] mwdata[9] 11 mrdata[31] mwdata[31] 33 mrdata[10] mwdata[10] 12 mteab mkillb 34 mrdata[11] mwdata[11] 13 iplb[0] mtm[0] 35 mrdata[12] mwdata[12] 14 iplb[1] mtm[1] 36 mrdata[13] mwdata[13] 15 iplb2] mtm[2] 37 mrdata[14] mwdata[14] 16 ich_sz[0] pst[0] 38 mrdata[15] mwdata[15] 17 ich_sz[1] pst[1] 39 mrdata[16] mwdata[16] 18 ich_sz[2] pst[2] 40 mrdata[17] mwdata[17] 19 sram_sz[0] ddata[0] 41 mrdata[18] mwdata[18] 20 sram_sz[1] ddata[1] 42 mrdata[19] mwdata[19] 21 sram_sz[2] ddata[2] 43 mrdata[20] mwdata[20] 22 rom_sz[2] test_rhit tail of i/o test ring chain 0 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-6 coldfire2/2m users manual motorola table 8-4. i/o test ring chain 1 (trsdi[1]/trsdo[1]) 8.3 integrated memory testing a test mode is provided to test embedded sram, rom, and/or icache arrays that connect directly to the coldfire2/2m. all integrated memories are tested by placing the coldfire2/ 2m into test mode (test_mode = 1) and performing reads and writes via the test bus and system-bus. these busses allow external access to the integrated memories via functional paths. the path from the test/system bus to the memory arrays is a pipelined path. the test bus input signals serve as multiplexor select signals to enable the correct path from mrdata and test_addr to a particular array, and from the particular array out through mwdata. during test mode, coldfire2/2m is idle and does not execute code. the following section details the signal descriptions, test methodology, and data transfer mechanisms. 8.3.1 test bus signal description this section describes the coldfire2/2m signals dedicated to testing the integrated memories. additional signals are required to be either asserted or negated during memory test and are described here also. if all three types of integrated memories are used; ie. core input core output cell # (2 regs/ cell) core input core output cell # (w regs/ cell) nc mfrzb head of i/o test ring chain 1 test_ctrl maddr[18] 23 mrdata[0] marbc[0] 2 test_rd maddr[19] 24 rom_sz[0] marbc[1] 3 test_data_rd maddr[20] 25 rom_sz[1] pst[3] 4 test_ram_rd maddr[21] 26 mrstb ddata[3] 5 test_rom_rd maddr[22] 27 dsclk dso 6 rom_vld maddr[23] 28 iack_68k mrw 7 test_ivld_inh maddr[24] 29 test_mode mtt[0] 8 nc maddr[25] 30 test_kta mtt[1] 9 nc maddr[26] 31 test_addr[2] maddr[5] 10 test_itag_wrt maddr[27] 32 test_addr[3] maddr[6] 11 test_data_wrt maddr[28] 33 test_addr[4] maddr[7] 12 test_ram_wrt maddr[29] 34 test_addr[5] maddr[8] 13 mtab maddr[30] 35 test_addr[4] maddr[9] 14 bkptb maddr[31] 36 test_addr[7] maddr10] 15 nc maddr[0] 37 test_addr[8] maddr[11] 16 nc maddr[1] 38 test_addr[9] maddr[12] 17 nc maddr[2] 39 test_addr[10] maddr[13] 18 nc maddr[3] 40 test_addr[11] maddr[14] 19 nc maddr[4] 41 test_addr[12] maddr[15] 20 nc msiz[0] 42 test_addr[13] maddr[16] 21 nc msiz[1] 43 test_addr[14] maddr[17] 22 nc mtsb tail of i/o test ring chain 1 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation motorola coldfire2/2m users manual 8-7 sram, rom, and icache, all signals listed here must be brought out to package pins via muxing. if only one or two types of integrated memories are used, a subset of signals listed here must be brought out to package pins. i.e. some of the test input control pins could be negated instead of needing to be muxed out. also, depending upon array size, not all of test_addr may be necessary. all coldfire2/2m signals are unidirectional and synchronous. 8.3.1.1 test address bus (test_addr[14:2]). these input signals are used to specify an address when testing the integrated memories. 8.3.1.2 test control (test_ctrl). this active-high input signal indicates the test address bus (test_addr[14:2]) will be latched on the next positive clock edge. 8.3.1.3 test idata read (test_idata_rd). this active-high input signal is used to test the instruction cache data memory read operation. 8.3.1.4 test idata write (test_idata_wrt). this active-high input signal is used to test the instruction cache data memory write operation. 8.3.1.5 test instruction cache read hit (test_rhit). this active-high output signal indicates a hit has occurred when accessing the instruction cache during memory array testing. 8.3.1.6 test invalidate inhibit (test_ivld_inh). this active-high input signal inhibits the invalidate operation when testing the instruction cache. 8.3.1.7 test itag write (test_itag_wrt). this active-high input signal is used to test the instruction cache tag memory write operation. 8.3.1.8 test kta mode enable (test_kta). this active-high input signal allows the instruction cache tag and data arrays to be read in parallel, mimicking the functional operation. this allows testing of the speed path from the tag and data arrays to the core. 8.3.1.9 test mode enable (test_mode). this active-high input signal is used to enable all of the integrated memory test signals. test_mode should be asserted for the duration of memory testing. 8.3.1.10 test sram read (test_sram_rd). this active-high input signal is used to test the integrated sram memory read operation. 8.3.1.11 test sram write (test_sram_wrt). this active-high input signal is used to test the integrated sram memory write operation. 8.3.1.12 test read (test_rd). this active-high input signal is used to test read operations on all of the integrated memories. 8.3.1.13 test rom read (test_rom_rd). this active-high input signal is used to test the integrated rom memory read operation. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-8 coldfire2/2m users manual motorola 8.3.1.14 test write inhibit (test_wr_inh). this active-high input signal disables the write strobes to the sram and instruction cache compiled rams. test_wr_inh should be negated for the duration of memory test. 8.3.1.15 mie. this active-high input signal enables the capturing of mrdata into the memory arrays. mie should be asserted for the duration of memory testing. 8.3.1.16 tr_mode. this signal needs to be negated to keep the i/o test ring in functional mode. tr_mode should be negated for the duration of memory testing. 8.3.1.17 mwdata[31:0]. the m-bus write data bus is used to observe array data during array reads. 8.3.1.18 mrdata[31:0]. the m-bus read data bus is used to apply write data during array writes. 8.3.1.19 scan_mode. this signal must be negated to allow array output data into the coldfire2/2m. 8.3.1.20 scan_se. this signal must be negated to allow functional operation of the coldfire2/2m. 8.3.1.21 scan_xarray. this signal should be negated to prevent continuous strobing of the coldfire2/2m integrated memories. 8.3.2 memory test theory of operation a write consists of 4 cycles; i.e. it is not until the fourth cycle after applying test_addr and mrdata, with the arrays write control signal active, that data is strobed into the array. data and address can be updated continuously every cycle to fill the array via the pipeline. a read consists of 6 cycles; i.e. it is not until the sixth cycle after applying test_addr, with the arrays read control signal active, that data appears on mwdata. address can be updated continuously every cycle to read the array via the pipeline. both the sram and icache data arrays are accessed as noted above. the rom is read as noted above. two test modes exist for the icache tag array. essentially the tag and data array are written. leaving test_mode asserted, test_kta is then asserted and address applied. test_rhit will be asserted 4 cycles later, icache data array data will subsequently appear on mwdata 2 cycles later. the processor must be placed in reset in accordance with the system reset specification. the sequence (mrstb asserted 6 cycles, stall 8cycles) must be executed both upon entering test mode at power-up, as well as when switching into test mode from any other mode. it is this sequence that directs the cpu to go into an idle mode to allow integrated memory testing. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation motorola coldfire2/2m users manual 8-9 8.3.3 test mode the test mode is entered by asserting test_modeand mie, negating test_ctrl, test_wr_inh, and tr_mode, and resetting the coldfire2/2m as described in section 3.9 reset operation . mrstb must be asserted (low) >= 6 cycles. the first test sequence should not begin until at least eight clock cycles after mrstb is negated. test_mode and mie should remain asserted during all test sequences. test_wr_inh and tr_mode should remain negated during all test sequences. 8.3.4 instruction cache tag ram testing the instruction cache tag ram consists of up to 2k addresses that can be accessed individually through the test bus. testing the compiled tag ram is accomplished by first writing into the tag ram to set the valid bit, and reading the tag ram to check for a cache hit. 8.3.4.1 instruction cache tag ram write function. writing to the instruction cache tag ram is performed though the test bus and mrdata[31:0] after entering test mode. the address and control signals are input on the test bus and the data to be written to the tag ram is input on mrdata[31:0]. writes to the tag ram are performed in a pipelined fashion as shown in figure 8-3 . all input signals are latched on the positive edge of clk and all outputs transition on the positive edge of clk. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-10 coldfire2/2m users manual motorola clock 1 (c1) on the first clock cycle of the tag ram write sequence, the first tag ram address should be driven onto test_addr[14:2] and test_ctrl should be asserted. clock 2 (c2) on the second cycle, the first data value associated with the address asserted in c1 should be driven onto mrdata[31:0] and the next address should be driven onto test_addr[14:2]. clock 3 (c3) on the third cycle, the next address and data values should be driven. clock 4 (c4) on the fourth cycle, the next address and data values should be driven and test_itag_wrt should be asserted. the remaining addresses and data are driven each successive clock cycle and test_itag_wrt should remain asserted until the last data value has been latched. clock 5 (c5) the fifth clock cycle is identical to c4. the first write to the tag ram occurs. figure 8-3. test instruction cache tag write cycles c1 c2 c3 c4 c5 cn clk test_mode a0 a1 a2 a3 a4 a5 test_addr test_ctrl d0 d1 d2 d3 d4 mrdata test_itag_wrt a0 a1 ich_addr icht_csb icht_rwb t0 t1 icht_di icht_st f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation motorola coldfire2/2m users manual 8-11 clock n (cn) the remaining clock cycles in the test are identical to c5. throughout the entire sequence, the data value being driven is associated with the address driven in the previous clock cycle. the format of the data value written to the tag ram depends on the cache configuration. the data value consists of the significant upper bits latched on mrdata with the valid bit (bit eight) set. see section 5.1 instruction cache for more information. 8.3.4.2 instruction cache tag ram read function. reading from the instruction cache tag ram is not performed as a typical input address, read data operation. instead, reading the instruction cache tag ram is performed though the test bus and mrdata[31:0] after entering test mode. the address and control signals are input on the test bus and the expected value of the data from the tag ram is input on mrdata[31:0]. reads from the tag ram are performed in a pipelined fashion as shown in figure 8-4 . all input signals are latched on the positive edge of clk and all outputs transition on the positive edge of clk. figure 8-4. test instruction cache tag read cycles c1 c2 c3 c4 cn clk test_mode a0 a1 a2 a3 a4 test_addr test_ctrl d0 d1 d2 d3 mrdata a0 a1 ich_addr icht_csb icht_rwb icht_st t0 t1 icht_do test_rhit f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-12 coldfire2/2m users manual motorola clock 1 (c1) on the first clock cycle of the tag ram read sequence, the first data ram address should be driven onto test_addr[14:2] and test_ctrl should be asserted. clock 2 (c2) on the second cycle, the first expected data value associated with the address asserted in c1 should be driven onto mrdata[31:0] and the next address should be driven onto test_addr[14:2]. the remaining addresses and expected data are driven each successive clock cycle. clock 3 (c3) the third clock cycle is identical to c2. clock 4 (c4) the fourth clock cycle is identical to c3. the first read from the tag ram occurs in c4, three cycles after the first address was driven. if the value read from the tag ram is equal to the associated value latched on mrdata[31:0] and the valid bit is set, test_rhit will be asserted. clock n (cn) the remaining clock cycles in the test are identical to c4. throughout the entire sequence, test_rhit is associated with the address driven three clock cycles earlier. for this reason, three stall cycles (test_ctrl negated) should occur after the last address is driven before the next sequence to wait for the last test_rhit. 8.3.4.3 instruction cache tag ram write followed by read function. the following timing diagram illustrates the minimum number of cycles necessary to perform a write followed by read of the instruction cache tag ram. a certain sequence is neccessary because of the pipelined datapath to the arrays from mrdata and mwdata. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation motorola coldfire2/2m users manual 8-13 . figure 8-5. i-cache tag write followed by read clock 1 (c1) on the first clock cycle , the data ram address a1 should be driven onto test_addr[14:2] and test_itag_wrt should be asserted. clock 2 (c2) on the second cycle, the d1, data value associated with the address asserted in c1, should be driven onto mrdata[31:0]. the address value driven in c1, a1, can optionally be driven in c2. test_itag_wrt remains asserted. clock 3 (c3) the third clock cycle is a stall cycle necessary because of the pipelined data path. the address driven in c1, a1, must be driven during c3.test_itag_wrt remains asserted. clock 4 (c4) the next write address, a2, should be driven onto test_addr[14:2] during c4. the data driven in c2, d1, must be driven in c4.test_itag_wrt remains asserted. c1 c2 c3 c4 c5 c6 c7 c8 c9 a1 a2 d2 d1 d3 d2 d2 d1 d1 a1 a1 a2 a2 a3 a3 d2 d1 clk test_mode test_ctrl test_itag_wrt test_rd test_addr mrdata test_rhit icht_addr icht_di icht_do write read stall wr rd stall f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-14 coldfire2/2m users manual motorola clock 5(c5) test_itag_wrt must be negated during c5. a1 and d1, associated with c1/c2, are at the icache tag array during c5; i.e. the actual write to the array occurs in this cycle. the data value associated with the address asserted in c4, d2, is driven onto mrdata[31:0] during this cycle as well. you have the option of driving a2 during this cycle. clock 6 (c6) the test_rhit signal will assert during this cycle if the icache tag array data for a1 matches d1, the access in cycles c1 and c2; i.e. it is in this cycle that the actual read occurs. a2 must be driven onto test_addr[14:2] during this cycle. test_itag_wrt asserts and d2 can optionally be driven during this cycle. clock 7 (c7) on c7, d2, must be driven onto mrdata[31:0] and the next data ram address, a3, should be driven onto test_addr[14:2]. this is a stall cycle. test_itag_wrt remains asserted during this cycle. clock 8 (c8) during c8, test_itag_wrt must be negated. a3 can optionally be driven during this cycle; d3 must be driven during this cycle. the actual write of a2/d2 occurs during this cycle. this periodic 3-cycle sequence continues; ie. c4-c6 are repeated during c7-c9, c10-c12, etc. these three cycles are stall, write, then read. test_itag_wrt always negates during the c5 cycle or the write cycle and asserts during the c6-c7 cycles.. test_rhit always asserts if no error during c6 or the read cycle. the address is asserted in the c4 cycle, optionally asserted in the c5 cycle and again asserted during the c6 cycle. the corresponding data to that address is driven in c5, optionally driven in c6 and then driven again in c7. 8.3.5 instruction cache data ram testing the instruction cache data ram consists of up to 8k long words that can be accesses individually through the test bus. testing the compiled data ram is accomplished by first writing test patterns into the data ram, reading the data ram, and verifying the results. 8.3.5.1 instruction cache data ram write function. writing to the instruction cache data ram is performed though the test bus and mrdata[31:0] after entering test mode. the address and control signals are input on the test bus and the data to be written to the data ram is input on mrdata[31:0]. writes to the data ram are performed in a pipelined fashion as shown in figure 8-6 . all input signals are latched on the positive edge of clk and all outputs transition on the positive edge of clk. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation motorola coldfire2/2m users manual 8-15 clock 1 (c1) on the first clock cycle of the data ram write sequence, the first data ram address should be driven onto test_addr[14:2] and test_ctrl should be asserted. clock 2 (c2) on the second cycle, the first data value associated with the address asserted in c1 should be driven onto mrdata[31:0] and the next address should be driven onto test_addr[14:2]. clock 3 (c3) on the third cycle, the next address and data values should be driven and test_idata_wrt should be asserted. clock 4 (c4) the fourth clock cycle is identical to c3. the remaining addresses and data are driven each successive clock cycle and test_idata_wrt should remain asserted until the last data value has been latched. the first write to the data ram occurs. clock n (cn) the remaining clock cycles in the test are identical to c4. figure 8-6. test instruction cache data write cycles c1 c2 c3 c4 cn clk test_mode a0 a1 a2 a3 a4 test_addr test_ctrl d0 d1 d2 d3 mrdata test_idata_wrt a0 a1 ich_addr ichd_csb ichd_rwb d0 d1 ichd_di ichd_st f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-16 coldfire2/2m users manual motorola throughout the entire sequence, the data value being driven is associated with the address driven in the previous clock cycle. 8.3.5.2 instruction cache data ram read function. reading from the instruction cache data ram is performed though the test bus and mwdata[31:0] after entering test mode. the address and control signals are input on the test bus and the data from the data ram is output on mwdata[31:0]. reads from the data ram are performed in a pipelined fashion as shown in figure 8-7 . all input signals are latched on the positive edge of clk and all outputs transition on the positive edge of clk. clock 1 (c1) on the first clock cycle of the data ram read sequence, the first data ram address should be driven onto test_addr[14:2] and test_ctrl should be asserted. clock 2 (c2) on the second cycle, the next addresses should be driven onto test_addr[14:2]. figure 8-7. test instruction cache data read cycles c1 c2 c3 c4 c5 c6 cn clk test_mode a0 a1 a2 a3 a4 a5 a6 test_addr test_ctrl test_idata_rd test_rd a0 a1 a2 a3 ich_addr ichd_csb ichd_rwb ichd_st d0 d1 d2 d3 ichd_do d0 d1 mwdata f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-17 coldfire2/2m users manual motorola clock 3 (c3) the third cycle is identical to c2. clock 4 (c4) on the fourth cycle, the next address should be driven and test_idata_rd should be asserted. the first read from the data ram occurs in c4. clock 5 (c5) on the fifth cycle, the next address should be driven and test_rd should be asserted. the remaining addresses are driven each successive clock cycle and test_idata_rd should remain asserted until the last data value has been read from the data ram. clock 6 (c6) cycle 6 is identical to c5. the mwdata[31:0] signals are driven with the first data, five cycles after the associated address was driven. test_rd should remain asserted until the last data value has been driven onto mwdata[31:0]. clock n (cn) the remaining clock cycles in the test are identical to c6. throughout the entire sequence, the data value being driven is associated with the address driven five clock cycles earlier. for this reason, five stall cycles (test_ctrl negated) should occur after the last address is driven before the next sequence to wait for the last data to be driven. 8.3.5.3 instruction cache data ram write followed by read function. the following timing diagram illustrates the minimum number of cycles necessary to perform a write followed by read of the instruction cache data ram. a certain sequence is neccessary because of the pipelined datapath to the arrays from mrdata and mwdata. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-18 coldfire2/2m users manual motorola figure 8-8. i-cache data write/read clock 1 (c1) on the first clock cycle , a1 (data ram address) should be driven onto test_addr[14:2] and test_idata_wrt should be asserted. clock 2 (c2) on the second cycle, d1 (the data value associated with a1, the address asserted in c1) should be driven onto mrdata[31:0]. a1 must continue to be driven during this cycle. test_idata_wrt continues to be asserted. clock 3 (c3) on the third cycle, the next data ram address, a2, is driven onto test_addr[14:2]. d1 can optionally be driven during this cycle. test_idata_wrt remains asserted during this cycle. clock 4 (c4) d2 should be driven onto mrdata[31:0] during this cycle. test_idata_rd must be asserted here for the length of this test. test_idata_wrt is negated during this cycle. the actual write to the icache data array of the c1/c2 (a1/d1) access occurs during this cycle. a2 must remain driven during this cycle. c1 c2 c3 c4 c5 c6 c7 c8 c9 a1 a2 a3 test_rd test_addr mrdata icht_addr ichd_di ichd_do write read a4 d1 d2 d3 d4 d1 d2 a3 a2 a1 d1 d2 d3 d1 d2 d3 write read read write test_idata_rd test_idata_wrt mwdata f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-19 coldfire2/2m users manual motorola clock 5(c5) the next data ram address, a3, should be driven onto test_addr[14:2] during this cycle. test_rd is asserted here for the length of this test. test_idata_wrt is asserted during this cycle. the acutal read of the icache data array c1/c2 (a1/d1) access occurs during this cycle. d2 can optionally be driven during this cycle. clock 6 (c6) d3, the data associated with the c5 address, a3, is driven onto mrdata[31:0] during this cycle. test_idata_wrt is negated during this cycle .the actual write of the icache data access c3/c4 (a2/d2) occurs during this cycle. a3 must remain driven in this cycle. clock 7 (c7) a4 should be driven onto test_addr[14:2] during this cycle. test_idata_wrt is asserted. the c1/c2 access read data, d1, will be driven onto mwdata[31:0] during this cycle. the actual read of the icache data array c3/c4 access (a2/d2) occurs during this cycle. d3 can optionally be driven in this cycle. clock 8 (c8) d4 must be driven onto mrdata[31:0] during this cycle. a4 must remain driven. test_idata_wrt is negated. the actual write of the icache data array c3/c4 access, a3/d, occurs here. clock 9 (c9) a5 must be driven onto test_addr[14:2] during this cycle. test_idata_wrt is asserted. d4 can optionally be driven in this cycle. the c5/c6 access read data, d2, will be driven onto mwdata[31:0] during this cycle. the actual read of the icache data array c5/c6 access (a3/d3) occurs during this cycle. this periodic 2-cycle sequence continues; ie. c6-c7 are repeated during c8-c9, c10-c11, etc. the c6 cycle is the actual write cyle and the c7 cycle is the actual read cycle where the data from the previous actual write/read sequence is displayed on mwdata. 8.3.6 instruction cache kta mode testing the kta mode tests the read data path for both of the instruction cache (tag and data) rams. this test checks the valid bit and performs a compare between the upper bits of the appropriate tag and the upper bits of the address to determine a cache hit. if the read is a hit, the value from the data ram is returned. this mimicks the reading of the instruction cache during normal operation. both rams are first written using instruction cache data and tag write cycles. both the mrdata[31:15] signals and part of the test_addr[14:2] signals are used to drive data into the array and compare against during the read. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-20 coldfire2/2m users manual motorola a 512 byte icache array uses an icache tag array sized 32x24; thus having 5 address bits. test_addr[8:4] is used to address the tag array, while test_addr[8:2] addresses the icache data array (128x32). {mrdata[31:15], test_addr[14:9]} is used to compare data to the tag array. these 23 bits compare to the 23 bits of address in the tag array, the 24th bit is the valid bit. table 8-5. test kta configurations icache sz tag array sz tag array test_addr mrdata & test_addr bytes locations x # of data bits addr bits used to addr used for data writes/cmprs 512 32x24 5 test_addr[8:4] {mrdata[31:15], test_addr[14:9]} 1k 64x23 6 test_addr[9:4] {mrdata[31:15], test_addr[14:10]} 2k 128x22 7 test_addr[10:4] {mrdata[31:15], test_addr[14:11]} 4k 256x21 8 test_addr[11:4] {mrdata[31:15], test_addr[14:12]} 8k 512x20 9 test_addr[12:4] {mrdata[31:15], test_addr[14:13]} 16k 1kx19 10 test_addr[13:4] {mrdata[31:15], test_addr[14]} 32k 2kx18 11 test_addr[14:4] mrdata[31:15] f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-21 coldfire2/2m users manual motorola clock 1 (c1) on the first clock cycle of the kta mode sequence, the first data ram address should be driven onto test_addr[14:2] and test_ctrl should be asserted. figure 8-9. kta mode cycles c1 c2 c3 c4 c5 c6 cn clk test_mode a0 a1 a2 a3 a4 a5 a6 test_addr test_ctrl t0 t1 t2 t3 t4 t5 mrdata test_kta test_idata_rd test_rd a0 a1 a2 a3 ich_addr icht_csb icht_rwb icht_st t0 t1 t2 t3 icht_do ichd_csb ichd_rwb ichd_st d0 d1 d2 d3 ichd_do test_rhit d0 d1 mwdata f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-22 coldfire2/2m users manual motorola clock 2 (c2) on the second cycle, the next data ram addresses should be driven onto test_addr[14:2]. the mrdata[31:8] signals should be driven with the tag information expected from the tag ram for the line associated with the address driven in c1. clock 3 (c3) the third cycle is identical to c2. clock 4 (c4) on the fourth cycle, the next address should be driven, test_kta should be asserted, and test_idata_rd should be asserted. the first read from the cache rams occur in c4. if the value from the tag ram is equal to the tag value driven in c2, test_rhit will be asserted. clock 5 (c5) on the fifth cycle, the next address should be driven and test_rd should be asserted. the remaining addresses are driven each successive clock cycle and test_kta and test_idata_rd should remain asserted until the last data value has been read from the data ram. clock 6 (c6) cycle 6 is identical to c5. the mwdata[31:0] signals are driven with the first data, five cycles after the associated address was driven. test_rd should remain asserted until the last data value has been driven onto mwdata[31:0]. clock n (cn) the remaining clock cycles in the test are identical to c6. 8.3.7 rom testing the rom consists of up to 8k long words that can be accesses individually through the test bus. testing the compiled rom is accomplished by reading the rom and verifying the results. 8.3.7.1 rom read function. reading from the rom is performed though the test bus and mwdata[31:0] after entering test mode. the address and control signals are input on the test bus and the data from the rom is output on mwdata[31:0]. reads from the rom are performed in a pipelined fashion as shown in figure 8-10 . all input signals are latched on the positive edge of clk and all outputs transition on the positive edge of clk. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-23 coldfire2/2m users manual motorola clock 1 (c1) on the first clock cycle of the rom read sequence, the first rom address should be driven onto test_addr[14:2] and test_ctrl should be asserted. clock 2 (c2) on the second cycle, the next addresses should be driven onto test_addr[14:2]. clock 3 (c3) the third cycle is identical to c2. clock 4 (c4) on the fourth cycle, the next address should be driven and test_rom_rd should be asserted. the first read from the rom does not occur until c4. clock 5 (c5) on the fifth cycle, the next address should be driven and test_rd should be asserted. the remaining addresses are driven each successive clock cycle and test_rom_rd should remain asserted until the last data value has been read from the rom. figure 8-10. test rom read cycles c1 c2 c3 c4 c5 c6 cn clk test_mode a0 a1 a2 a3 a4 a5 a6 test_addr test_ctrl test_rom_rd test_rd a0 a1 a2 a3 rom_addr rom_csb rom_enb d0 d1 d2 d3 rom_do d0 d1 mwdata f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-24 coldfire2/2m users manual motorola clock 6 (c6) cycle six is identical to c5. the mwdata[31:0] signals are driven with the first data, five cycles after the associated address was driven. test_rd should remain asserted until the last data value has been driven onto mwdata[31:0]. clock n (cn) the remaining clock cycles in the test are identical to c6. throughout the entire sequence, the data value being driven is associated with the address driven five clock cycles earlier. for this reason, five stall cycles (test_ctrl negated) should occur after the last address is driven before the next sequence to wait for the last data to be driven. 8.3.8 sram testing the sram consists of up to 8k longwords that can be accesses individually through the test bus. testing the compiled sram is accomplished by first writing test patterns into the sram, reading the sram, and verifying the results. 8.3.8.1 sram write function. writing to the sram is performed though the test bus and mrdata[31:0] after entering test mode. the address and control signals are input on the test bus and the data to be written to the sram is input on mrdata[31:0]. writes to the sram are performed in a pipelined fashion as shown in figure 8-11 . all input signals are latched on the positive edge of clk and all outputs transition on the positive edge of clk. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-25 coldfire2/2m users manual motorola clock 1 (c1) on the first clock cycle of the sram write sequence, the first sram address should be driven onto test_addr[14:2] and test_ctrl should be asserted. clock 2 (c2) on the second cycle, the first data value associated with the address asserted in c1 should be driven onto mrdata[31:0] and the next address should be driven onto test_addr[14:2]. clock 3 (c3) on the third cycle, the next address and data values should be driven and test_sram_wrt should be asserted. the remaining addresses and data are driven each successive clock cycle and test_sram_wrt should remain asserted until the last data value has been latched. clock 4 (c4) on the fourth cycle, the next address and data values should be driven. the first write to the srams occurs in cycle four, three cycles after the first address was driven. figure 8-11. test sram write cycles c1 c2 c3 c4 c5 clk test_mode a0 a1 a2 a3 a4 test_addr test_ctrl d0 d1 d2 d3 mrdata test_sram_wrt a0 a1 sram_addr sram_csb sram_rwb d0 d1 sram_di sram_st f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-26 coldfire2/2m users manual motorola clock n (cn) the remaining clock cycles in the test are identical to c4. throughout the entire sequence, the data value being driven is associated with the address driven in the previous clock cycle. 8.3.8.2 sram read function. reading from the sram is performed though the test bus and the mwdata[31:0] after entering test mode. the address and control signals are input on the test bus and the data from the sram is output on mwdata[31:0]. reads from the sram are performed in a pipelined fashion as shown in figure 8-12 . all input signals are latched on the positive edge of clk and all outputs transition on the positive edge of clk. clock 1 (c1) on the first clock cycle of the sram read sequence, the first sram address should be driven onto test_addr[14:2] and test_ctrl should be asserted. clock 2 (c2) on the second cycle, the next addresses should be driven onto test_addr[14:2]. figure 8-12. test sram read cycles c1 c2 c3 c4 c5 c6 cn clk test_mode a0 a1 a2 a3 a4 a5 a6 test_addr test_ctrl test_sram_rd test_rd a0 a1 a2 a3 sram_addr sram_csb sram_rwb sram_st d0 d1 d2 d3 sram_do d0 d1 mwdata f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-27 coldfire2/2m users manual motorola clock 3 (c3) the third cycle is identical to c2. clock 4 (c4) on the fourth cycle, the next address should be driven and test_sram_rd should be asserted. the first read to the sram does not occur until c4, three cycles after the first address was driven. clock 5 (c5) on the fifth cycle, the next address should be driven and test_rd should be asserted. the remaining addresses are driven each successive clock cycle and test_sram_rd should remain asserted until the last data value has been read from the sram. clock 6 (c6) cycle six is identical to c5. the mwdata[31:0] signals are driven with the first data, five cycles after the associated address was driven. test_rd should remain asserted until the last data value has been driven onto mwdata[31:0]. clock n (cn) the remaining clock cycles in the test are identical to c6. throughout the entire sequence, the data value being driven is associated with the address driven five clock cycles earlier. for this reason, five stall cycles (test_ctrl negated) should occur after the last address is driven before the next sequence to wait for the last data to be driven. 8.3.8.3 sram write followed by read function. the following timing diagram illustrates the minimum number of cycles necessary to perform a write followed by read of the sram. a certain sequence is neccessary because of the pipelined datapath to the arrays from mrdata and mwdata. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-28 coldfire2/2m users manual motorola figure 8-13. sram write followed by read clock 1 (c1) on the first clock cycle , a1, sram address, should be driven onto test_addr[14:2] and test_sram_wrt should be asserted. clock 2 (c2) on the second cycle, d1 (the data value associated with a1, the address asserted in c1) should be driven onto mrdata[31:0]. a1 must continue to be driven during this cycle. test_sram_wrt remains asserted. clock 3 (c3) c3 is a stall cycle. test_sram_wrt must remain asserted clock 4 (c4) c4 is a stall cycle. test_sram_wrt must negate during this cycle which does the actual write of a1/d1 at the array. test_sram_rd must assert during this cycle. a1 test_addr d1 d2 d3 a0 a2 a3 test_ctrl test_kta test_idata_rd test_rd test_rhit mwdata d0 d1 d1 d1 d2 ich_addr[14:2] d0 a3 a2 a1 a0 d0 d1 d2 d3 ichd_do if 8k cache, test_addr[14:13] are used as the lower two bits of data to be compared for the hit signal. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
test operation 8-29 coldfire2/2m users manual motorola clock 5 (c5) the next sram address, a2, should be driven onto test_addr[14:2] during this cycle. test_rd is asserted here; it will remain asserted during the next cycle c5, and then repeat 2 cycles negated, 2 cycles asserted for the duration of this type of test. test_idata_wrt remains negated during this cycle; it will repeat two cycles asserted, two cyles negated. test_sram_rd remains asserted during this cycle. the acutal read of the icache data array c1/c2 (a1/d1) access occurs during this cycle. clock 6 (c6) d2, associated with a2, is driven onto mrdata[31:0] during this cycle. a2 remains driven. test_rd remains asserted, test_sram_rd negates, and test_sram_wrt asserts during this cycle. clock 7 (c7) d1 is driven onto mwdata[31:0]. this is a stall cycle; test_sram_rd remains negated, test_rd negates, and test_sram_wrt remains asserted during this cycle. clock 8 (c8) the repetition of c4-c7 begins on c8. test_sram_rd is asserted, test_rd negated, and test_sram_wrt negated during this cycle. the actual write of a2/d2 at the array occurs during this cycle. clock 9 (c9) c9 mimmicks c5: test_sram_rd remains asserted, test_rd asserts, and test_sram_wrt remains negated. a3 is driven onto test_addr[14:2]. the actual read of a2/d2 at the array occurs during this cycle. clock 10 (c10) c10 mimmicks c6: a3 remains driven on test_addr[14:2]; d3 asserts on mrdata[31:0]. test_sram_rd negates, test_rd remains asserted, and test_sram_wrt asserts during this cycle. clock 11 (c11) c11 mimmicks c7: d2 is driven onto mwdata[31:0]. this is a stall cycle; test_sram_rd remains negated, test_rd negates, and test_sram_wrt remains asserted during this cycle. this periodic 4-cycle sequence continues; ie. c4-c7 are repeated during c8-c11, c12-c15, etc. the c4 cycle is the actual write cyle and the c5 cycle is the actual read cycle. during the c7 cycle the data from the previous actual write/read sequence at the array is displayed on mwdata. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual 9-1 section 9 instruction execution timing this section presents coldfire2/2m instruction execution times in terms of clock cycles. the number of operand references for each instruction is also included, enclosed in parentheses following the number of clock cycles. each timing entry is presented as c (r/w) where: ? c - the number of coldfire2/2m clock cycles, including all applicable operand fetches and writes, as well as all internal cycles required to complete the instruction execution. ? r/w - the number of operand reads (r) and writes (w) required by the instruction. an operation performing a read-modify-write function is denoted as (1/1). this section includes assumptions concerning the timing values and the execution time details and applies to both the coldfire2 and coldfire2m unless otherwise noted. 9.1 timing assumptions the timing data presented in this section have the following assumptions: 1. the operand execution pipeline (oep) is loaded with the opword and all required extension words at the beginning of each instruction execution. this implies that the oep doesnt wait for the instruction fetch pipeline (ifp) to supply opwords and/or extension words. 2. the oep does not experience any sequence-related pipeline stalls. for the coldfire2/ 2m, the most common example of this type of stall involves consecutive store operations, excluding the movem instruction. for all store operations (except movem), certain hardware resources within the coldfire2/2m are marked as busy for two clock cycles after the final dsoc cycle of the store instruction. if a subsequent store instruction is encountered within this 2-cycle window, it will be stalled until the resource again becomes available. thus, the maximum pipeline stall involving consecutive store operations is 2 cycles. the movem instruction uses a different set of resources and this stall does not apply. 3. the oep completes all memory accesses without any stall conditions caused by the memory itself. thus, the timing details provided in this section assume an infinite zero- wait state memory is attached to the coldfire2/2m. 4. all operand data accesses are aligned on the same byte boundary as the operand size: 16-bit operands aligned on 0-modulo-2 addresses, 32-bit operands aligned on 0- modulo-4 addresses. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
instruction execution timing 9-2 coldfire2/2m users manual motorola if the operand alignment fails these guidelines, the misalignment unit is used. the processor core decomposes the misaligned operand reference into a series of aligned accesses as shown in table 9-1 . 9.2 move instruction execution times the execution times for the move.{b,w} instructions are shown in table 9-2 , while table 9-3 provides the timing for move.l. for all tables in this section, the execution time (et) of any instruction using the pc-relative effective addressing modes is exactly equivalent to the time using the comparable an- relative mode. the nomenclature xxx.wl refers to both forms of absolute addressing, xxx.w and xxx.l. table 9-1. misaligned operand references address[1:0] size kbus operations additional c(r/w) x1 word byte, byte 2(1/0) if read 1(0/1) if write x1 long byte, word, byte 3(2/0) if read 2(0/2) if write 10 long word, word 2(1/0) if read 1(0/1) if write f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
instruction execution timing motorola coldfire2/2m users manual 9-3 . table 9-2. move byte and word execution times source destination rx (ax) (ax)+ -(ax) (d16,ax) (d8,ax,xn*sf) xxx.wl dy 1(0/0) 1(0/1) 1(0/1) 1(0/1) 1(0/1) 2(0/1) 1(0/1) ay 1(0/0) 1(0/1) 1(0/1) 1(0/1) 1(0/1) 2(0/1) 1(0/1) (ay) 3(1/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) (ay)+ 3(1/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) -(ay) 3(1/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) (d16,ay) 3(1/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) (d8,ay,xn*sf) 4(1/0) 4(1/1) 4(1/1) 4(1/1) xxx.w 3(1/0) 3(1/1) 3(1/1) 3(1/1) xxx.l 3(1/0) 3(1/1) 3(1/1) 3(1/1) (d16,pc) 3(1/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) (d8,pc,xn*sf) 4(1/0) 4(1/1) 4(1/1) 4(1/1) #xxx 1(0/0) 3(0/1) 3(0/1) 3(0/1) table 9-3. move long execution times source destination rx (ax) (ax)+ -(ax) (d16,ax) (d8,ax,xn*sf) xxx.wl dy 1(0/0) 1(0/1) 1(0/1) 1(0/1) 1(0/1) 2(0/1) 1(0/1) ay 1(0/0) 1(0/1) 1(0/1) 1(0/1) 1(0/1) 2(0/1) 1(0/1) (ay) 2(1/0) 2(1/1) 2(1/1) 2(1/1) 2(1/1) 3(1/1) 2(1/1) (ay)+ 2(1/0) 2(1/1) 2(1/1) 2(1/1) 2(1/1) 3(1/1) 2(1/1) -(ay) 2(1/0) 2(1/1) 2(1/1) 2(1/1) 2(1/1) 3(1/1) 2(1/1) (d16,ay) 2(1/0) 2(1/1) 2(1/1) 2(1/1) 2(1/1) (d8,ay,xn*sf) 3(1/0) 3(1/1) 3(1/1) 3(1/1) xxx.w 2(1/0) 2(1/1) 2(1/1) 2(1/1) xxx.l 2(1/0) 2(1/1) 2(1/1) 2(1/1) (d16,pc) 2(1/0) 2(1/1) 2(1/1) 2(1/1) 2(1/1) (d8,pc,xn*sf) 3(1/0) 3(1/1) 3(1/1) 3(1/1) #xxx 1(0/0) 2(0/1) 2(0/1) 2(0/1) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
instruction execution timing 9-4 coldfire2/2m users manual motorola 9.3 standard one operand instruction execution times table 9-4. one operand instruction execution times opcode effective address rn (an) (an)+ -(an) (d16,an) (d8,an,xn*sf) xxx.wl #xxx clr.b 1(0/0) 1(0/1) 1(0/1) 1(0/1) 1(0/1) 2(0/1) 1(0/1) clr.w 1(0/0) 1(0/1) 1(0/1) 1(0/1) 1(0/1) 2(0/1) 1(0/1) clr.l 1(0/0) 1(0/1) 1(0/1) 1(0/1) 1(0/1) 2(0/1) 1(0/1) ext.w d x 1(0/0) ext.l d x 1(0/0) extb.l d x 1(0/0) neg.l d x 1(0/0) negx.l d x 1(0/0) not.l d x 1(0/0) scc dx 1(0/0) swap d x 1(0/0) tst.b 1(0/0) 3(1/0) 3(1/0) 3(1/0) 3(1/0) 4(1/0) 3(1/0) 1(0/0) tst.w 1(0/0) 3(1/0) 3(1/0) 3(1/0) 3(1/0) 4(1/0) 3(1/0) 1(0/0) tst.l 1(0/0) 2(1/0) 2(1/0) 2(1/0) 2(1/0) 3(1/0) 2(1/0) 1(0/0) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
instruction execution timing motorola coldfire2/2m users manual 9-5 9.4 standard two operand instruction execution times table 9-5. two operand instruction execution times opcode effective address rn (an) (an)+ -(an) (d16,an) (d16,pc) (d8,an,xn*sf) (d8,pc,xn*sf) xxx.wl #xxx add.l ,rx 1(0/0) 3(1/0) 3(1/0) 3(1/0) 3(1/0) 4(1/0) 3(1/0) 1(0/0) add.l dy, 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) addi.l #imm,dx 1(0/0) addq.l #imm, 1(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) addx.l dy,dx 1(0/0) and.l ,dn 1(0/0) 3(1/0) 3(1/0) 3(1/0) 3(1/0) 4(1/0) 3(1/0) 1(0/0) and.l dy, 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) andi.l #imm,dx 1(0/0) asl.l ,dx 1(0/0) 1(0/0) asr.l ,dx 1(0/0) 1(0/0) bchg dy, 2(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) bchg #imm, 2(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) bclr dy, 2(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) bclr #imm, 2(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) bset dy, 2(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) bset #imm, 2(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) btst dy, 2(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) btst #imm, 1(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 1(0/0) cmp.l ,rx 1(0/0) 3(1/0) 3(1/0) 3(1/0) 3(1/0) 4(1/0) 3(1/0) 1(0/0) cmpi.l #imm,dx 1(0/0) eor.l dy, 1(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) eori.l #imm,dx 1(0/0) lea ,ax 1(0/0) 1(0/0) 2(0/0) 1(0/0) lsl.l ,dx 1(0/0) 1(0/0) lsr.l ,dx 1(0/0) 1(0/0) moveq #imm,dx 1(0/0) muls.w ,dx 9(0/0) 1 11(1/0) 1 11(1/0) 1 11(1/0) 1 11(1/0) 1 12(1/0) 1 11(1/0) 1 9(0/0) 1 3(0/0) 2 5(1/0) 2 5(1/0) 2 5(1/0) 2 5(1/0) 2 6(1/0) 2 5(1/0) 2 3(0/0) 2 mulu.w ,dx 9(0/0) 1 11(1/0) 1 11(1/0) 1 11(1/0) 1 11(1/0) 1 12(1/0) 1 11(1/0) 1 9(0/0) 1 3(0/0) 2 5(1/0) 2 5(1/0) 2 5(1/0) 2 5(1/0) 2 6(1/0) 2 5(1/0) 2 3(0/0) 2 muls.l ,dx 18(0/0) 1 20(1/0) 1 20(1/0) 1 20(1/0) 1 20(1/0) 1 --- 5(0/0) 2 7(1/0) 2 7(1/0) 2 7(1/0) 2 7(1/0) 2 --- mulu.l ,dx 18(0/0) 1 20(1/0) 1 20(1/0) 1 20(1/0) 1 20(1/0) 1 --- 5(0/0) 2 7(1/0) 2 7(1/0) 2 7(1/0) 2 7(1/0) 2 --- or.l ,dn 1(0/0) 3(1/0) 3(1/0) 3(1/0) 3(1/0) 4(1/0) 3(1/0) 1(0/0) or.l dy, 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) or.l #imm,dx 1(0/0) sub.l ,rx 1(0/0) 3(1/0) 3(1/0) 3(1/0) 3(1/0) 4(1/0) 3(1/0) 1(0/0) sub.l dy, 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) subi.l #imm,dx 1(0/0) notes: 1. applies to the coldfire2 only. 2. applies to the coldfire2m only. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
instruction execution timing 9-6 coldfire2/2m users manual motorola subq.l #imm, 1(0/0) 3(1/1) 3(1/1) 3(1/1) 3(1/1) 4(1/1) 3(1/1) subx.l dy,dx 1(0/0) table 9-5. two operand instruction execution times (continued) opcode effective address rn (an) (an)+ -(an) (d16,an) (d16,pc) (d8,an,xn*sf) (d8,pc,xn*sf) xxx.wl #xxx notes: 1. applies to the coldfire2 only. 2. applies to the coldfire2m only. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
instruction execution timing motorola coldfire2/2m users manual 9-7 9.5 miscellaneous instruction execution times table 9-6. miscellaneous instruction execution times opcode effective address rn (an) (an)+ -(an) (d16,an) (d8,an,xn*sf) xxx.wl #xxx cpush 9(0/1) link.w ay,#imm 2(0/1) move.w ccr,dx 1(0/0) move.w ,ccr 1(0/0) 1(0/0) move.w sr,dx 1(0/0) move.w ,sr 7(0/0) 7(0/0) 2 move.l 1 ,acc 1(0/0) - - - - - - 1(0/0) move.l 1 ,macsr 1(0/0) - - - - - - 1(0/0) move.l 1 ,mask 1(0/0) - - - - - - 1(0/0) move.l 1 acc, 2(0/0) - - - - - - - move.l 1 macsr, 2(0/0) - - - - - - - move.l 1 mask, 2(0/0) - - - - - - - movec ry,rc 9(0/1) movem.l ,&list 1+n(n/0) 3 1+n(n/0) 3 movem.l &list, 1+n(0/n) 3 1+n(0/n) 3 nop 3(0/0) pea 2(0/1) 2(0/1) 4 3(0/1) 5 2(0/1) pulse 1(0/0) stop #imm 3(0/0) 6 trap #imm 15(1/2) tpf 1(0/0) tpf.w #imm 1(0/0) tpf.l #imm 1(0/0) unlk a x 2(1/0) wddata 3(1/0) 3(1/0) 3(1/0) 3(1/0) 4(1/0) 3(1/0) 3(1/0) wdebug 5(2/0) 5(2/0) notes: 1. applies to the coldfire2m only. 2. if a move.w #imm,sr instruction is executed and imm[13] = 1, the execution time is 1(0/0). 3. n is the number of registers moved by the movem opcode. 4. pea execution times are the same for (d16,pc) 5. pea execution times are the same for (d8,pc,xn*sf) 6. the execution time for stop is the time required until the coldfire2/2m begins sampling continuously for interrupts. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
instruction execution timing 9-8 coldfire2/2m users manual motorola 9.6 mac instruction execution timing these instructions are supported on the coldfire2m only. 9.7 branch instruction execution times table 9-7. mac instruction execution times opcode effective address rn (an) (an)+ -(an) (d16,an) (d16,pc) (d8,an,xn*sf) (d8,pc,xn*sf) xxx.wl #xxx mac.w ry,rx 1(0/0) - - - - - - - mac.l ry,rx 3(0/0) - - - - - - - msac.w ry,rx 1(0/0) - - - - - - - msac.l ry,rx 3(0/0) - - - - - - - macl.w ry,rx,,rw - 2(1/0) 2(1/0) 2(1/0) 2(1/0)? - - - macl.l ry,rx,,rw - 4(1/0) 4(1/0) 4(1/0) 4(1/0)? - - - msacl.w ry,rx,,rw - 2(1/0) 2(1/0) 2(1/0) 2(1/0)? - - - msacl.l ry,rx,,rw - 4(1/0) 4(1/0) 4(1/0) 4(1/0)? - - - note: ?effective address of (d16,pc) not supported table 9-8. general branch instruction execution times opcode effective address rn (an) (an)+ -(an) (d16,an) (d8,an,xi*sf) xxx.wl #xxx bsr 3(0/1) jmp 3(0/0) 3(0/0) 4(0/0) 3(0/0) jsr 3(0/1) 3(0/1) 4(0/1) 3(0/1) rte 8(2/0) rts 5(1/0) table 9-9. bra, bcc instruction execution times opcode forward taken forward not taken backward taken backward not taken bra 2(0/0) 2(0/0) bcc 3(0/0) 1(0/0) 2(0/0) 3(0/0) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual a-1 appendix a register summary a.1 register access methods all of the coldfire2/2m registers are accessed via special instructions or addressing modes. none of them are mapped into the user data address space. table a-1 lists the coldfire2/2m registers and their control register addresses. table a-1. register summary register acronym program access 1 debug access 2 instruction rc command drc address attribute register aatr wdebug $6 wdmreg $6 address breakpoint registers (high) abhr wdebug $c wdmreg $c address breakpoint registers (low) ablr wdebug $d wdmreg $d access control register 0 acr0 movec $004 rcreg,wcreg $004 access control register 1 acr1 movec $005 rcreg,wcreg $005 accumulator 3 acc move to/from acc - rcreg,wcreg $806 address registers a0-a6 move - rareg,wareg - cache control register cacr movec $002 rcreg,wcreg $002 condition code register ccr move to/from ccr - rcreg,wcreg $80e configuration/status register csr wdebug $0 rdmreg,wdmreg $0 data breakpoint mask register dbmr wdebug $f wdmreg $f data breakpoint register dbr wdebug $e wdmreg $e data registers d0-d7 move - rdreg,wdreg - mac status register 3 macsr move to/from macsr - rcreg,wcreg $804 mask register 3 mask move to/from mask - rcreg,wcreg $805 program counter pc - - rcreg,wcreg $80f program counter breakpoint mask register pbmr wdebug $9 wdmreg $9 program counter breakpoint register pbr wdebug $8 wdmreg $8 ram base address register rambar0 movec $c04 rcreg,wcreg $c04 rom base address register rombar0 movec $c00 rcreg,wcreg $c00 notes: 1. refer to the coldfire programmers reference manual (mcf5200prm/ad). 2. refer to section 7.3.3.1 bdm command set summary . 3. coldfire2m only. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
register summary a-2 coldfire2/2m users manual motorola a.1 register formats stack pointer a7,sp move - rareg/wareg - status register sr move to/from sr - rcreg,wcreg $80e trigger definition register tdr wdebug $7 wdmreg $7 vector base register vbr movec $801 rcreg,wcreg $801 bits 15 14 13 12 11 10 8 7 6 5 4 3 2 0 field rm szm ttm tmm r sz tt tm reset 00 0 0 00 0 101 r/w ww w w ww w w figure a-2. address attribute register (aatr) bits 31 0 field address reset - r/w w figure a-3. address breakpoint high register (abhr) bits 31 0 field address reset - r/w w figure a-4. address breakpoint low register (ablr) table a-1. register summary (continued) register acronym program access 1 debug access 2 instruction rc command drc notes: 1. refer to the coldfire programmers reference manual (mcf5200prm/ad). 2. refer to section 7.3.3.1 bdm command set summary . 3. coldfire2m only. f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
register summary motorola coldfire2/2m users manual a-3 bits 31 24 23 16 field ab am rese t 00 r/w ww bits 15 14 13 12 8 7 6 5 4 3 2 1 0 field en sm - enib cm buf w -wp- rese t 00 - 000-0- r/w ww - www-w- figure a-5. access control register (acr0, acr1) bits 31 30 29 28 27 26 25 24 23 18 17 16 field cen b - cps h cfr z - cinv - park rese t 0-00-0 - 0 r/w w-ww-w - w bits 15 11 10 9 8 7 6 5 4 2 1 0 field - cbe n cmo d cbu f - cwr p - clnf rese t - 000 - 0 - 0 r/w - www - w - w figure a-6. cache control register (cacr) bits 7 6 5 4 3 2 1 0 field ---xnzvc reset 00000000 r/w r r r r/w r/w r/w r/w r/w figure a-7. condition code register (ccr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
register summary a-4 coldfire2/2m users manual motorola bits 31 28 27 26 25 24 23 17 16 field status fof trg hal t bkp t - ipw rese t 0 0000 - 0 r/w 1 r rrrr - r/w bits 15 14 13 12 11 10 9 8 7 6 5 4 3 0 field map trc emu ddc uhe btb - npl ipi ssm - rese t 000 0 0 0 0000 - r/w ? r/w r/w r/w r/w r/w r/w r r/w r/w r/w - note: ?the csr is a write only register from the programming model. it can be read from and written to via the bdm port. figure a-8. configuration/status register (csr) bits 31 0 field mask reset - r/w w figure 1-9. data breakpoint mask register (dbmr) bits 31 0 field address reset - r/w w figure 1-10. data breakpoint register (dbr) bits 7 6 5 4 3 2 1 0 field omc s/u - - n z v c reset 0000-- -0 r/w r/w r/w r r r/w r/w r/w r figure 1-11. mac status register (macsr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
register summary motorola coldfire2/2m users manual a-5 bits 15 0 field mask rese t - r/w r/w figure 1-12. mac mask register (mask) bits 31 0 field mask reset - r/w w figure 1-13. program counter breakpoint mask register (pbmr) bits 31 0 field address reset - r/w w figure 1-14. program counter breakpoint register (pbr) bits 31 16 field ba rese t - r/w w bits 15 9 8 7 6 5 4 3 2 1 0 field ba wp - csm scm sdm ucm udm v rese t - - - -----0 r/w w w - wwwwww figure a-15. sram base address register (rambar0) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
register summary a-6 coldfire2/2m users manual motorola bits 31 16 field ba rese t - (0) r/w w bits 15 9 8 7 6 5 4 3 2 1 0 field ba wp - csm scm sdm ucm udm v rese t - (0) - (1) - - (1) - (0) - (0) - (0) - (0) 0 (1) r/w w w - wwwwww note: the reset values in parenthesis are valid if the rom_vld signal is asserted during reset. figure a-16. rom base address register (rombar0) bits 15 14 13 12 11 10 8 7 5 4 3 2 1 0 field t-sm- i - xnzvc reset 00110 7 0 00000 r/w r/w r r/w r/w r r/w r r/w r/w r/w r/w r/w figure a-17. status register (sr) bits 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 field trc ebl edl w edw l edw u edl l edl m edu m edu u di eai ear eal epc pci rese t 0 00000000000000 r/w w wwwwwwwwwwwwww bits 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 0 field - ebl edl w edw l edw u edl l edl m edu m edu u di eai ear eal epc pci rese t - 00000000000000 r/w - wwwwwwwwwwwwww figure a-18. trigger definition register (tdr) f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual b-1 appendix b new mac instructions b.1 enhanced integer-multiply instructions opcodes for the muls and mulu instructions are described in the mcf5200 family programmers reference manual . the only change to these opcodes is the improved execution efficiency. b.2 new mac instructions this section describes the new mac instructions for the coldfire2m. details of each instruction description are arranged in alphabetical order by instruction mnemonic. for notational conventions, refer to table 1-4 in section 1.8 instruction set summary . f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-2 coldfire2/2m users manual motorola mac multiply and accumulate mac operation: acc + ((ry rx){<< 1 | >> 1}) ? acc assembler syntax: mac. ry.

ry.

attributes: size = (word, long) ul = (upper, lower) shift = (<<, >>) description: multiplies two 16- or 32-bit numbers to produce a 32-bit result, then adds this product, optionally shifted left or right one bit, to the accumulator ( acc ). the result is stored back into the accumulator. if 16-bit operands are used, the upper or lower word of each register must be speci?ed. mac status register: omc - not affected. s/u - not affected. n - set if the most signi?cant bit of the result is set, otherwise cleared. z - set if the result is zero, otherwise cleared. v - set if an over?ow is generated, otherwise unchanged. c - always cleared. processor condition codes: not affected instruction format: instruction fields: ry - source y field specifies a source register operand, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. note that bit ordering is 6, 11, 10, 9 (msb to lsb). omc s/u - - n z v c --00***0 1514131211109876543210 1010 rw 00rw00 rx - sz sf 0 u/lw u/lx - f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions motorola coldfire2/2m users manual b-3 rx - source x field specifies a source register operand, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. note that bit ordering is 3, 2, 1, 0 (msb to lsb). sz - size field 0 = word-sized input operands 1 = long-sized input operands sf - scale factor field 00 = none 01 = product << 1 10 = reserved 11 = product >> 1 u/ly -source yword select field this bit determines which 16-bit operand of the source w register is used in the operation for word-sized operations only. 0 = lower word 1 = upper word u/lx - source x word select field this bit determines which 16-bit operand of the source x register is used in the operation for word-sized operations only. 0 = lower word 1 = upper word f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-4 coldfire2/2m users manual motorola macl multiply and accumulate macl with register load operation: acc + ((ry rx){<< 1 | >> 1}) ? acc ({& mask}) ? ry assembler syntax: macl. ry.

,rw macl. ry.

,,rw macl. ry.

ry.

attributes: size = (word, long) ul = (upper, lower) shift = (<<, >>) description: multiplies two 16- or 32-bit numbers to produce a 32-bit result, then subtracts this product, optionally shifted left or right one bit, from the accumulator ( acc ). the result is stored back into the accumulator. if 16-bit operands are used, the upper or lower word of each register must be speci?ed. mac status register: omc - not affected. s/u - not affected. n - set if the most signi?cant bit of the result is set, otherwise cleared. z - set if the result is zero, otherwise cleared. v - set if an over?ow is generated, otherwise unchanged. c - always cleared. processor condition codes: not affected instruction format: instruction fields: ry - operand yfield specifies a source register operand, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. note that bit ordering is 6, 11, 10, 9 (msb to lsb). omc s/u - - n z v c --00***0 1514131211109876543210 1010 ry 00ry00 rx - sz sf 1 u/l y u/lx - f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-8 coldfire2/2m users manual motorola rx - operand x field specifies a source register operand, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. note that bit ordering is 3, 2, 1, 0 (msb to lsb). sz - size field 0 = word-sized input operands 1 = long-sized input operands sf - scale factor field 00 = none 01 = product << 1 10 = reserved 11 = product >> 1 u/ly -source yword select field this bit determines which 16-bit operand of the source yregister is used in the operation for word-sized operations only. 0 = lower word 1 = upper word u/lx - source x word select field this bit determines which 16-bit operand of the source x register is used in the operation for word-sized operations only. 0 = lower word 1 = upper word f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions motorola coldfire2/2m users manual b-9 msacl multiply and subtract msacl with register load operation: acc - ((ry rx){<< 1 | >> 1}) ? acc ({& mask}) ? rw assembler syntax: msacl. ry.

,rw msacl. ry.

,,rw msacl. ry.

,&,rw attributes: size = (word, long) ul = (upper, lower) shift = (<<, >>) ea = effective address description: multiplies two 16- or 32-bit numbers to produce a 32-bit result, then subtracts this product, optionally shifted left or right one bit, from the accumulator ( acc ). the result is stored back into the accumulator. if 16-bit operands are used, the upper or lower word of each register must be speci?ed. in parallel with this operation, a 32-bit operand is fetched from the memory location defined by and loaded into the destination register, rw. if the mask addressing mode is used, the low-order word of is anded with the mask register. refer to section 6.2.3 mask register (mask) . mac status register: omc - not affected. s/u - not affected. n - set if the most signi?cant bit of the result is set, otherwise cleared. z - set if the result is zero, otherwise cleared. v - set if an over?ow is generated, otherwise unchanged. c - always cleared. processor condition codes: not affected omc s/u - - n z v c --00***0 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-10 coldfire2/2m users manual motorola instruction format: instruction fields: ry -source yfield specifies a source register operand, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. note that bit ordering is 15, 14, 13, 12 (msb to lsb). rx - source x field specifies a source register operand, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. note that bit ordering is 3, 2, 1, 0 (msb to lsb). rw - destination field specifies a destination register operand, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. note that bit ordering is 6, 11, 10, 9 (msb to lsb). - effective address of memory operand field sz - size field 0 = word-sized input operands 1 = long-sized input operands sf - scale factor field 00 = none 01 = product << 1 10 = reserved 11 = product >> 1 1514131211109876543210 1010 ry 01ry mode reg rw sz sf 1 u/ly u/lx mam 0 rx addressing mode mode register addressing mode mode register dn - - (xxx).w - - an - - (xxx).l - - (an) 010 reg.num:an # - - (an)+ 011 reg.num:an -(an) 100 reg.num:an (d16,an) 101 reg.num:an (d16,pc) - - (d8,an,xn) - - (d8,pc,xn) - - f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions motorola coldfire2/2m users manual b-11 u/ly -source yword select field this bit determines which 16-bit operand of the source yregister is used in the operation for word-sized operations only. 0 = lower word 1 = upper word u/lx - source x word select field this bit determines which 16-bit operand of the source x register is used in the operation for word-sized operations only. 0 = lower word 1 = upper word mam - mask addressing mode modifier this bit determines if the mask addressing mode should be used. refer to section 6.2.3 mask register (mask) . 0 = normal addressing mode 1 = mask addressing mode b.3 new register instructions this section describes the new register instructions for the coldfire2m. details of each instruction description are arranged in alphabetical order by instruction mnemonic. for notational conventions, refer to table 1-4 in section 1.8 instruction set summary . f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-12 coldfire2/2m users manual motorola move move from acc move from accumulator from acc operation: acc ? rx assembler syntax: move. acc, rx attributes: size = long description: moves a 32-bit value from the accumulator ( acc ) to a register. the size of the operation must be speci?ed as long. mac status register: not affected processor condition codes: not affected instruction format: instruction fields: rx[3:0] specifies the destination register, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. 1514131211109876543 0 101000011000 rx f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions motorola coldfire2/2m users manual b-13 move move from macsr move from mac status reg from macsr operation: macsr ? rx[7:0] 0 ? rx[31:8] assembler syntax: move. macsr, rx attributes: size = long description: moves the contents of the mac status register ( macsr ), zero-extended to long size, into a general-purpose register, rx. the size of the operation must be speci?ed as long. mac status register: not affected processor condition codes: not affected instruction format: instruction fields: rx[3:0] specifies the destination register, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. 1514131211109876543 0 101010011000 rx f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-14 coldfire2/2m users manual motorola move move from mask move from mask from mask operation: mask ? rx[15:0] 0xffff ? rx[31:16] assembler syntax: move. mask, rx attributes: size = long description: moves a 32-bit value from the mask register ( mask ), one-extended to long size, to a register. the size of the operation must be speci?ed as long. mac status register: not affected processor condition codes: not affected instruction format: instruction fields: rx[3:0] specifies the destination register, where $0 is d0,..., $7 is d7, $8 is a0,..., $f is a7. 1514131211109876543 0 101011011000 rx f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions motorola coldfire2/2m users manual b-15 move move to acc move to accumulator to acc operation: source ? acc assembler syntax: move. , acc attributes: size = long description: moves a 32-bit value from a register or an immediate value into the accumulator ( acc ). the size of the operation must be speci?ed as long. mac status register: omc - not affected. s/u - not affected. n - set if the most signi?cant bit of the result is set, otherwise cleared. z - set if the result is zero, otherwise cleared. v - always cleared. c - always cleared. processor condition codes: not affected instruction format: omc s/u - - n z v c - - 00* *00 15141312111098765 0 1010000100 mode reg f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-16 coldfire2/2m users manual motorola instruction fields: - effective address addressing mode mode register addressing mode mode register dn 000 reg.num:dn (xxx).w - - an 001 reg.num:an (xxx).l - - (an) - - # 111 100 (an)+ - - -(an) - - (d16,an) - - (d16,pc) - - (d8,an,xn) - - (d8,pc,xn) - - f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions motorola coldfire2/2m users manual b-17 move move to ccr move to condition code register to ccr operation: macsr[4:0] ? ccr[4:0] assembler syntax: move. macsr,ccr attributes: size = long description: moves the indicator flags of the mac status register ( macsr ) into the processors condition code cegister ( ccr ). the size of the operation must be specified as long. mac status register: not affected processor condition codes: x - not affected. n - set to the value of macsr bit 3, n. z - set to the value of macsr bit 2, z. v - set to the value of macsr bit 1, v. c - set to the value of macsr bit 0, c. instruction format : xnzvc -**** 1514131211109876543210 1010100111000000 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-18 coldfire2/2m users manual motorola move move to macsr move to mac status register to macsr operation: source ? macsr assembler syntax: move. , macsr attributes: size = long description: moves the low-order byte of a 32-bit value from a register or an immediate value into the mac status register ( macsr ). the size of the operation must be speci?ed as long. mac status register: omc - set to the value of bit 7 of the source operand. s/u - set to the value of bit 6 of the source operand. n - set to the value of bit 3 of the source operand. z - set to the value of bit 2 of the source operand. v - set to the value of bit 1 of the source operand. c - always cleared. processor condition codes: not affected instruction format: omc s/u - - n z v c **00***0 15141312111098765 0 1010100100 mode reg f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions motorola coldfire2/2m users manual b-19 instruction fields: - effective address addressing mode mode register addressing mode mode register dn 000 reg.num:dn (xxx).w - - an 001 reg.num:an (xxx).l - - (an) - - # 111 100 (an)+ - - -(an) - - (d16,an) - - (d16,pc) - - (d8,an,xn) - - (d8,pc,xn) - - f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-20 coldfire2/2m users manual motorola move move to mask move to modulus register to mask operation: source ? mask assembler syntax: move. , mask attributes: size = long description: moves the low-order word of a 32-bit value from a register or an immediate value into the mask register ( mask ). the size of the operation must be speci?ed as long. mac status register: not affected processor condition codes: not affected instruction format: instruction fields: - effective address 15141312111098765 0 1010110100 mode reg addressing mode mode register addressing mode mode register dn 000 reg.num:dn (xxx).w - - an 001 reg.num:an (xxx).l - - (an) - - # 111 100 (an)+ - - -(an) - - (d16,an) - - (d16,pc) - - (d8,an,xn) - - (d8,pc,xn) - - f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions motorola coldfire2/2m users manual b-21 b.4 operation code map all mac instructions are mapped into line a, i.e. bits 15-12 of the instruction are 1010 ($a). 1. mac 1514131211109876543210 1010 ry 00ry00 rx - sz sf 0 u/l y u/lx - 2. msac 1514131211109876543210 1010 ry 00ry00 rx - sz sf 1 u/l y u/lx - 3. macl 1514131211109876543210 1010 ry 01ry mode reg ry sz sf 0 u/ly u/lx mam 0 rx 4. msacl 1514131211109876543210 1010 rw 01ry mode reg r y sz sf 1 u/ly u/lx mam 0 rx 5. move to acc 15141312111098765 0 1010000100 mode reg 6. move to macsr 15141312111098765 0 1010100100 mode reg 7. move to mask 15141312111098765 0 1010110100 mode reg 8. move from acc 1514131211109876543 0 101000011000 rx f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
new mac instructions b-22 coldfire2/2m users manual motorola 9. move from macsr 1514131211109876543 0 101010011000 rx 10. move from mask 1514131211109876543 0 101011011000 rx 11. move to ccr 15141312111098765 0 1010000100 mode reg f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual index-1 index a a0 - a7 1-12 aabr 7-29 , 7-30 , 7-32 , 7-33 , a-2 , a-4 , a-5 aatr 7-28 , 7-30 ablr/abhr 7-28 , 7-29 acc 1-14 , 6-2 access alternate master 3-5 emulator mode 3-5 interrupt acknowledge mode 3-5 normal mode 3-4 access control programming model 5-8 registers 5-8 , 5-9 , a-3 access error exception 4-7 accumulator (acc) 1-14 , 6-2 acr 5-8 , 5-9 , a-3 address error exception 4-8 address registers (a0 C a6) 1-12 address space 2-3 , 3-1 , 3-4 addressing mode summary 1-18 alternate master 1-9 alternate master access 3-5 arbitration 3-30 arbitration algorithm 3-30 autovectored interrupts 4-12 b background debug mode see bdm bdm 7-5 command format 7-10 command sequence diagram 7-11 , 7-12 command set 7-9 connector 7-39 cpu32 7-40 dump memory block (dump) 7-17 fill memory block (fill) 7-19 no operation (nop) 7-21 read a/d register (rareg/rdreg) 7-13 read control register (rcreg) 7-22 read debug module register (rdmreg) 7-24 read memory location (read) 7-14 recommended connector 7-40 resume execution (go) 7-21 sampling diagram 7-8 serial interface 7-7 serial transfer diagram 7-8 write a/d register (wareg/wdreg) 7-13 write control register (wcreg) 7-23 write debug module register (wdmreg) 7-25 write memory location (write) 7-16 bkptb 2-11 , 7-1 , 7-6 , 7-7 branch indication on pst 7-3 bus arbitration 3-30 programming model 3-31 bus errors 3-28 c cache see instruction cache cacheability 5-5 cacr 3-31 , 5-6 , a-3 ccr 1-12 , 1-13 , a-3 clk 2-6 clock (clk) 2-6 coherency 5-6 coldfire system diagram 1-8 coldfire2 1-1 coldfire2m 1-1 , 1-13 , 6-1 , b-1 compiled ram 5-1 , 5-4 condition code register (ccr) 1-13 , a-3 configuration encoding 2-8 , 2-10 , 2-11 , 5-3 , 5-12 , 5-16 core block diagram 1-10 cpu halt 7-5 , 7-6 csr 7-36 , 7-37 , a-4 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
index-2 coldfire2/2m users manual motorola d d0 - d7 1-12 data formats 1-16 data registers (d0 C d7) 1-12 data sheet 10-6 data transfer mechanism 3-4 data transfers 3-6 dbr/dbmr 7-28 , 7-33 ddata 2-12 , 7-2 , 7-3 , 7-4 , 7-5 debug module 7-1 bdm 7-5 command set 7-9 dump 7-17 fill 7-19 go 7-21 nop 7-21 rareg/rdreg 7-13 rcreg 7-22 rdmreg 7-24 read 7-14 serial interface 7-7 wareg/wdreg 7-13 wcreg 7-23 wdmreg 7-25 write 7-16 concurrent operation 7-39 cpu halt 7-6 emulator mode 7-27 , 7-38 hardware reuse 7-28 interrupt 4-10 , 7-26 programming model 7-28 real-time support 7-26 real-time trace 7-2 registers address attribute (aatr) 7-30 address attribute breakpoint (aabr) 7-29 , 7-30 , 7-32 , 7-33 , a-2 , a-4 , a-5 address breakpoint (ablr, abhr) 7-29 configuration/status register (csr) 7-36 , 7-37 , a-4 data breakpoint (dbr, dbmr) 7-33 program counter breakpoint (pbr, pbmr) 7-32 trigger definition (tdr) 7-34 , a-6 signals 7-1 break point (bkptb) 2-11 , 7-1 debug data (ddata) 2-12 , 7-2 development serial clock (dsclk) 2-12 , 7-2 development serial input (dsi) 2-12 , 7-2 development serial output (dso) 2-12 , 7-2 processor status (pst) 2-12 , 7-2 theory of operation 7-26 see also registers, signals definitions 10-1 development cycle 1-5 diagrams bdm command sequence 7-12 bdm sampling 7-8 bdm serial transfer 7-8 bus exception 3-29 coldfire system 1-8 core block 1-10 design system overview 1-7 example instruction cache interface 5-2 example rom interface 5-11 example sram interface 5-15 flexcore integrated processor 1-3 instruction cache data read 8-8 instruction cache data write 8-7 instruction cache tag read 8-5 instruction cache tag write 8-4 integer address formats 1-17 integer data formats 1-16 interrupt acknowledge 3-24 , 3-26 kta mode 8-10 line read transfer 3-13 line write transfer 3-16 , 3-17 mac flow 6-2 memory operand addressing 1-18 misaligned transfer 3-19 processor status 7-4 processor/debug module interface 7-1 read transfer 3-8 recommended bdm connector 7-40 reset 3-30 rom read 8-12 signals block 2-1 sram read 8-15 sram write 8-14 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual index-3 wait state 3-21 write transfer 3-10 dsclk 2-12 , 7-2 , 7-8 dsi 2-12 , 7-2 dso 2-12 , 7-2 dump 7-17 e electrical chacteristics 10-1 emulator mode 7-5 , 7-27 , 7-38 emulator mode access 3-5 enhanced integer multiply instructions b-1 exception processing 4-1 , 7-5 flowchart 4-2 overview 4-1 self-aligning stack 4-4 stack frame definition 4-3 exception vector table 4-5 exceptions 4-7 access error 4-7 address error 4-8 bus cycle diagram 3-29 control cycles 3-27 debug interrupt 4-10 , 7-26 format error 4-10 illegal instruction 4-9 interrupt 4-10 priorities 4-6 privilege violation 4-9 reset 4-7 simultaneous 4-6 trace 4-9 trap instruction 4-10 unimplemented opcode 4-9 vector number 4-4 , 4-5 external bus 1-9 f fault-on-fault halt 3-29 , 4-6 , 7-6 fill 7-19 flexcore 1-8 advantages 1-4 development cycle 1-5 integrated processor diagram 1-3 integrated processors 1-2 module types 1-4 flowcharts exception processing 4-2 interrupt acknowledge 3-23 line read transfer 3-12 line write transfer 3-15 read transfer 3-7 write transfer 3-9 format error exceptions 4-10 freezing the instruction cache 5-7 g general control signals 2-6 clock (clk) 2-6 interrupt priority level (iplb) 2-6 go 7-21 h halt 7-6 halt instruction 7-5 hard module 1-4 i iack_68k 2-3 , 2-5 , 2-6 , 3-1 , 3-3 , 3-4 , 3-22 ich_addr 2-7 , 5-2 ich_sz 2-8 , 5-3 ichd_csb 2-7 , 5-2 ichd_di 2-7 , 5-2 ichd_do 2-7 , 5-2 ichd_rwb 2-7 , 5-3 ichd_st 2-7 , 5-2 icht_csb 2-8 , 5-3 icht_di 2-8 , 5-3 icht_do 2-8 , 5-3 icht_rwb 2-9 , 5-4 icht_st 2-9 , 5-4 illegal instruction exception 4-9 instruction cache 1-10 , 5-1 cacheability 5-5 coherency 5-6 configuration encoding 2-8 , 5-3 data ram testing 8-6 example interface diagram 5-2 freezing 5-7 interaction with other modules 5-4 invalidating 5-5 , 5-7 kta mode testing 8-9 line fill encoding 5-8 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
index-4 coldfire2/2m users manual motorola miss fetch algorithm 5-4 , 5-8 physical organization 5-4 programming model 5-6 registers cache control (cacr) 5-6 , a-3 reset 5-6 signals 5-1 address bus (ich_addr) 2-7 , 5-2 cache tag output bus (icht_do) 2-8 , 5-3 data chip select (ichd_csb) 2-7 , 5-2 data input bus (ichd_di) 2-7 , 5-2 data output bus (ichd_do) 2-7 , 5-2 data read/write (ichd_rwb) 2-7 , 5-3 data strobe (ichd_st) 2-7 , 5-2 size(ich_sz) 2-8 , 5-3 tag chip select (icht_csb) 2-8 , 5-3 tag input bus (icht_di) 2-8 , 5-3 tag read/write (icht_rwb) 2-9 , 5-4 tag strobe (icht_st) 2-9 , 5-4 tag ram testing 8-3 write protection 5-8 see also registers, signals instruction execution timing 9-1 assumptions 9-1 branch 9-8 mac 9-8 miscellaneous 9-7 move 9-2 one operand 9-4 two operand 9-5 instruction set summary 1-19 , 6-5 integer data formats 1-16 integer programming model 1-11 integrated memories 5-1 testing 8-1 integrated memory test signals 2-12 integrated processors 1-2 interrupt acknowledge diagram 3-24 , 3-26 flowchart 3-23 interrupt acknowledge access 3-5 interrupt acknowledge bus cycle 3-22 interrupt acknowledge mode 2-3 , 2-5 , 2-6 , 3-1 , 3-3 , 3-4 , 3-22 interrupt priority level (iplb) 2-6 , 4-10 interrupts 4-10 autovectored 4-12 debug 4-10 , 7-26 level seven 4-11 spurious 3-27 , 4-12 uninitialized 4-12 invalidating cache entries 5-5 , 5-7 iplb 2-6 , 4-10 k kta mode 8-9 l level seven interrupts 4-11 line access patterns 3-11 line fills 5-4 line read transfer diagram 3-13 flowchart 3-12 line write transfer diagram 3-16 , 3-17 flowchart 3-15 m mac 9-8 , b-2 mac status register (macsr) 6-2 mac status register (macsr) 1-14 mac unit 6-1 enhanced integer multiply instructions b-1 instruction set summary 6-5 instruction timing 9-8 introduction 6-1 new mac instructions b-1 new register instructions b-12 overflow mode 6-3 , 6-4 programming model 1-13 , 6-2 registers acc 6-2 macsr 6-2 mask 6-3 saturation mode 6-3 , 6-4 shifting 6-4 macl 9-8 , b-4 macsr 1-14 , 6-2 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual index-5 maddr 2-3 , 3-1 , 3-8 , 3-10 , 3-13 , 3-16 , 3-24 , 3-26 marb 1-10 marbc 2-3 , 3-1 mask 6-3 mask (mask) 6-3 mask addressing mode 6-4 master bus 1-8 , 3-1 arbitration 1-10 , 3-30 arbitration algorithm 3-30 bus errors 3-28 data transfer mechanism 3-4 data transfers 3-6 exception control cycles 3-27 line read transfers 3-11 line write transfers 3-14 misaligned operands 3-18 normal read transfers 3-6 normal write transfers 3-9 requirements for read transfer 3-5 requirements for write transfers 3-6 reset 3-29 signals 3-1 68k iack mode enable (iack_68k) 2-3 , 2-5 , 2-6 , 3-1 , 3-3 , 3-4 , 3-22 abiter control (marbc) 2-3 , 3-1 address bus (maddr) 2-3 , 3-1 freeze (mfrzb) 2-4 , 3-2 kill (mkillb) 2-4 , 3-2 read data bus (mrdata) 2-4 , 3-2 read data input enable (mie) 2-4 , 3-2 read/write (mrwb) 2-4 , 3-2 reset (mrstb) 2-4 , 3-2 size (msiz) 2-4 , 3-2 transfer acknowledge (mtab) 2-5 , 3-2 transfer error acknowledge (mteab) 2-5 , 3-3 transfer modifier (mtm) 2-5 , 3-3 transfer start (mtsb) 2-6 , 3-3 transfer type (mtt) 2-6 , 3-3 write data bus (mwdata) 2-6 , 3-4 transfer modifier encoding 2-5 , 3-3 transfer size encoding 2-5 , 3-2 transfer type 3-4 transfer type encoding 2-6 , 3-4 write data output enable (mwdataoe) 2-6 , 3-4 see also registers, signals memory operand addressing diagram 1-18 memory organization 1-17 mfrzb 2-4 , 3-2 mie 2-4 , 3-2 , 3-8 , 3-13 , 3-25 , 3-27 misaligned operands 3-18 misaligned transfer diagram 3-19 misalignment unit 3-18 miss fetch algorithm 5-4 , 5-8 mkillb 2-4 , 3-2 module types 1-4 move from acc 9-7 , b-13 move from macsr 9-7 , b-14 move from mask 9-7 , b-15 move to acc 9-7 , b-16 move to ccr 9-7 , b-18 move to macsr b-19 move to macsr 9-7 move to mask 9-7 , b-21 movec instruction 3-31 , 5-6 , 5-8 , 5-12 , 5-17 , 7-22 , a-1 mrdata 2-4 , 3-2 , 3-8 , 3-13 , 3-25 , 3-27 mrstb 2-4 , 3-2 mrwb 2-4 , 3-2 , 3-8 , 3-10 , 3-13 , 3-16 , 3-24 , 3-26 msac 9-8 , b-7 msacl 9-8 , b-9 msiz 2-4 , 3-2 , 3-8 , 3-10 , 3-13 , 3-16 , 3-24 , 3-26 mtab 2-5 , 3-2 , 3-8 , 3-13 , 3-16 , 3-25 , 3-27 mteab 2-5 , 3-3 mtm 2-5 , 3-3 , 3-13 mtsb 2-6 , 3-3 , 3-8 , 3-10 , 3-13 , 3-16 , 3-24 , 3-26 mtt 2-6 , 3-3 , 3-8 , 3-10 , 3-13 , 3-16 , 3-24 , 3-26 muls and mulu 9-5 , b-1 multiple exceptions 4-6 multiply-accumulate unit see mac unit mwdata 2-6 , 3-4 , 3-10 , 3-16 mwdataoe 2-6 , 3-4 , 3-10 , 3-16 n new mac instructions b-1 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
index-6 coldfire2/2m users manual motorola new register instructions b-12 nop 4-8 , 7-21 normal access 3-4 notational conventions 1-19 o overflow mode 6-3 , 6-4 p parameterizable module 1-5 pbmr 7-32 pbr 7-32 pc 1-12 performance 3-20 pipeline stalls 3-20 power 2-4 , 3-2 privilege mode 1-15 , 3-4 , 4-2 , 7-3 privilege violation exception 4-9 processor status 2-12 , 7-2 , 7-3 processor status diagram 7-4 processor status encoding 7-3 program counter (pc) 1-12 programming model 1-11 access control 5-8 bus arbitration 3-31 debug module 7-28 instruction cache 5-6 integer 1-11 mac unit 1-13 , 6-2 rom module 5-12 sram module 5-16 supervisor 1-14 pst 2-12 , 7-2 , 7-3 pulse instruction 2-12 , 7-2 , 7-3 r rambar0 5-17 , a-5 rareg a-1 rareg/rdreg 7-13 rcreg 7-22 , a-1 rdmreg 7-24 , a-1 read 7-14 read transfer diagram 3-8 flowchart 3-7 read transfers 3-6 , 3-11 real-time debug support 7-26 real-time trace 7-2 registers access control acr 1-15 , 5-8 , 5-9 , a-3 access methods a-1 debug module aabr 7-29 , 7-30 , 7-32 , 7-33 , a-2 , a-4 , a-5 aatr 7-28 , 7-30 ablr/abhr 7-28 , 7-29 csr 7-36 , 7-37 , a-4 dbr/dbmr 7-28 , 7-33 pbr/pbmr 7-32 tdr 7-34 , a-6 instruction cache cacr 1-15 , 5-6 , a-3 integer unit a0 - a6 1-12 ccr 1-13 , a-3 d0 - d7 1-12 pc 1-12 sp 1-12 mac unit acc 1-14 , 6-2 macsr 1-14 , 6-2 mask 6-3 rom module rombar0 1-15 , 5-12 , a-6 sram module rambar0 1-15 , 5-17 , a-5 summary a-1 supervisor sr 1-15 , 4-2 , a-6 vbr 1-15 , 4-6 reset 2-4 , 3-2 , 3-29 , 3-30 , 4-7 , 5-6 , 5-8 , 5-12 , 5-17 rom module 1-11 , 5-10 configuration encoding 2-10 , 5-12 example interface diagrams 5-11 programming model 5-12 registers rom base address (rombar0) 5-12 , a-6 signals 5-10 activate (rom_actb) 2-10 , 5-12 address bus (rom_addr) 2-9 , 5-11 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual index-7 data output bus (rom_do) 2-9 , 5-11 enable (rom_enb) 2-9 , 5-11 size (rom_sz) 2-9 , 5-12 testing 8-11 see also registers, signals rom_actb 2-10 , 5-12 rom_addr 2-9 , 5-11 rom_do 2-9 , 5-11 rom_enb 2-9 , 5-11 rom_sz 2-9 , 5-12 rombar0 5-12 , a-6 rte instruction 2-12 , 4-6 , 4-10 , 7-2 , 7-4 , 7-5 , 7-27 s saturation mode 6-3 , 6-4 sbc 1-11 , 3-6 scan signals enable (scan_enable) 2-14 , 8-16 exercise array (scan_xarray) 2-14 , 8-16 input (scan_in) 2-14 , 8-16 mode (scan_mode) 2-14 , 8-17 output (scan_out) 2-14 , 8-17 scan_enable 2-14 , 8-16 scan_in 2-14 , 8-16 scan_mode 2-14 , 8-17 scan_out 2-14 , 8-17 scan_xarray 2-14 , 8-16 shifting operations 6-4 signals summary 2-1 block diagram 2-1 debug module 7-1 bkptb 2-11 , 7-1 ddata 2-12 , 7-2 dsclk 2-12 , 7-2 dsi 2-12 , 7-2 dso 2-12 , 7-2 pst 2-12 , 3-29 , 4-6 , 7-2 general control 2-6 clk 2-6 iplb 2-6 , 4-10 instruction cache 5-1 ich_addr 2-7 , 5-2 ich_sz 2-8 , 5-3 ichd_csb 2-7 , 5-2 ichd_di 2-7 , 5-2 ichd_do 2-7 , 5-2 ichd_rwb 2-7 , 5-3 ichd_st 2-7 , 5-2 icht_csb 2-8 , 5-3 icht_di 2-8 , 5-3 icht_do 2-8 , 5-3 icht_rwb 2-9 , 5-4 icht_st 2-9 , 5-4 master bus 3-1 iack_68k 2-3 , 2-5 , 2-6 , 3-1 , 3-3 , 3-4 , 3-22 maddr 2-3 , 3-1 marbc 2-3 , 3-1 mfrzb 2-4 , 3-2 mie 2-4 , 3-2 mkillb 2-4 , 3-2 , 3-20 mrdata 2-4 , 3-2 , 3-5 , 3-6 mrstb 2-4 , 3-2 , 3-30 mrwb 2-4 , 3-2 msiz 2-4 , 3-2 , 3-6 , 3-9 mtab 2-5 , 3-2 , 3-27 mteab 2-5 , 3-3 , 3-28 mtm 2-5 , 3-3 , 3-4 mtsb 2-6 , 3-3 mtt 2-6 , 3-3 , 3-4 mwdata 2-6 , 3-4 , 3-6 , 3-9 mwdataoe 2-6 , 3-4 rom module 5-10 rom_actb 2-10 , 5-12 rom_addr 2-9 , 5-11 rom_do 2-9 , 5-11 rom_enb 2-9 , 5-11 rom_sz 2-9 , 5-12 scan scan_enable 2-14 , 8-16 scan_in 2-14 , 8-16 scan_mode 2-14 , 8-17 scan_out 2-14 , 8-17 scan_xarray 2-14 , 8-16 sram module 5-14 sram_addr 2-10 , 5-15 sram_csb 2-10 , 5-16 sram_di 2-11 , 5-16 sram_do 2-11 , 5-16 sram_rwb 2-11 , 5-16 sram_st 2-11 , 5-16 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
index-8 coldfire2/2m users manual motorola sram_sz 2-11 , 5-16 test bus 2-12 test_addr 2-13 , 8-1 , 8-4 test_ctrl 2-13 , 8-1 , 8-4 test_ich_rhit 2-13 , 8-2 , 8-6 , 8-11 test_idata_rd 2-13 , 8-1 , 8-9 test_idata_wrt 2-13 , 8-2 , 8-7 test_itag_wrt 2-13 , 8-2 , 8-4 test_ivld_inh 2-13 , 8-2 test_kta 2-13 , 8-2 , 8-11 test_mode 2-13 , 8-2 , 8-3 test_read 2-13 , 8-2 , 8-9 test_rom_rd 2-13 , 8-2 , 8-12 test_sram_rd 2-13 , 8-2 , 8-16 test_sram_wrt 2-13 , 8-2 , 8-14 test_wr_inh 2-13 , 8-2 signed/unsigned mac operations 6-3 slave bus 1-9 slave modules 1-11 soft module 1-4 sp 1-12 , 4-4 spurious interrupt 3-27 spurious interrupts 4-12 sr 1-15 , 3-4 , 4-2 , 4-10 , a-6 sram module 1-11 , 5-14 configuration encoding 2-11 , 5-16 example interface diagram 5-15 programming model 5-16 registers base address (rambar0) 5-17 , a-5 signals 5-14 address bus (sram_addr) 2-10 , 5-15 chip select (sram_csb) 2-10 , 5-16 data input bus (sram_di) 2-11 , 5-16 data output bus (sram_do) 2-11 , 5-16 read/write (sram_rwb) 2-11 , 5-16 size (sram_sz) 2-11 , 5-16 strobe (sram_st) 2-11 , 5-16 testing 8-13 write protection 5-13 , 5-17 see also registers, signals sram_addr 2-10 , 5-15 sram_csb 2-10 , 5-16 sram_di 2-11 , 5-16 sram_do 2-11 , 5-16 sram_rwb 2-11 , 5-16 sram_st 2-11 , 5-16 sram_sz 2-11 , 5-16 stack frame 4-3 stack pointer (a7,sp) 1-12 status register (sr) 1-15 , 4-2 , a-6 stop instruction 4-9 , 7-5 , 7-7 supervisor programming model 1-14 system bus controller (sbc) 1-11 , 3-6 t tdr 7-34 , a-6 test bus 8-1 signals 2-12 address bus (test_addr) 2-13 , 8-1 , 8-4 control (test_ctrl) 2-13 , 8-1 , 8-4 idata read (test_idata_rd) 2-13 , 8-1 , 8-9 idata write (test_idata_wrt) 2-13 , 8-2 , 8-7 instruction cache read hit (test_ich_rhit) 2-13 , 8-2 , 8-6 , 8-11 invalidate inhibit (test_ivld_inh) 2-13 , 8-2 itag write (test_itag_wrt) 2-13 , 8-2 , 8-4 kta mode enable (test_kta) 2-13 , 8-2 , 8-11 mode enable (test_mode) 2-13 , 8-2 , 8-3 read (test_rd) 2-13 , 8-2 , 8-9 rom read (test_rom_rd) 2-13 , 8-2 , 8-12 sram read (test_sram_rd) 2-13 , 8-2 , 8-16 sram write (test_sram_wrt) 2-13 , 8-2 , 8-14 write inhibit (test_wr_inh) 2-13 , 8-2 test mode 8-3 test_addr 2-13 , 8-1 , 8-4 test_ctrl 2-13 , 8-1 , 8-4 test_ich_rhit 2-13 , 8-2 , 8-6 , 8-11 test_idata_rd 2-13 , 8-1 , 8-9 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .
motorola coldfire2/2m users manual index-9 test_idata_wrt 2-13 , 8-2 , 8-7 test_itag_wrt 2-13 , 8-2 , 8-4 test_ivld_inh 2-13 , 8-2 test_kta 2-13 , 8-2 , 8-11 test_mode 2-13 , 8-2 , 8-3 test_read 2-13 , 8-2 , 8-9 test_rom_rd 2-13 , 8-2 , 8-12 test_sram_rd 2-13 , 8-2 , 8-16 test_sram_wrt 2-13 , 8-2 , 8-14 test_wr_inh 2-13 , 8-2 testing cache data ram 8-6 cache tag ram 8-3 integrated memories 8-1 kta mode 8-9 rom 8-11 sram 8-13 trace exception 4-9 trace mode 4-2 , 4-9 transfer modifier encoding 2-5 , 3-3 transfer size encoding 2-5 , 3-2 transfer type 3-4 transfer type encoding 2-6 , 3-4 transfers line read 3-11 line write 3-14 normal read 3-6 normal write 3-9 trap instruction exceptions 4-9 , 4-10 u unimplemented opcode exception 4-9 uninitialized interrupts 4-12 v variant addressing 7-3 , 7-4 vbr 1-15 , 4-6 vector base register (vbr) 1-15 , 4-6 vector number 4-5 w wait state diagram 3-21 wait states 3-10 , 3-17 , 3-20 wareg a-1 wareg/wdreg 7-13 wcreg 7-23 , a-1 wddata instruction 2-12 , 7-2 , 7-3 wdebug instruction a-1 wdmreg 7-25 , a-1 write 7-16 write protection 5-8 , 5-10 , 5-13 , 5-17 write transfer diagram 3-10 flowchart 3-9 write transfers 3-9 , 3-14 f r e e s c a l e s e m i c o n d u c t o r , i freescale semiconductor, inc. f o r m o r e i n f o r m a t i o n o n t h i s p r o d u c t , g o t o : w w w . f r e e s c a l e . c o m n c . . .

▲Up To Search▲

Price & Availability of COLDFIRE2UM

	To Download COLDFIRE2UM Datasheet File
If you can't view the Datasheet, Please click here to try to view without PDF Reader .