July 2010 - openfirmware - mail.coreboot.org

[commit] r1898 - cpu/x86/pc/olpc/via
by repository service July 28, 2010

July 28, 2010

Author: quozl Date: Thu Jul 29 08:41:26 2010 New Revision: 1898 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1898 Log: Q3A47 Modified: cpu/x86/pc/olpc/via/fw-version.fth Modified: cpu/x86/pc/olpc/via/fw-version.fth ============================================================================== --- cpu/x86/pc/olpc/via/fw-version.fth Thu Jul 29 08:41:04 2010 (r1897) +++ cpu/x86/pc/olpc/via/fw-version.fth Thu Jul 29 08:41:26 2010 (r1898) @@ -1,3 +1,3 @@ \ The overall firmware revision macro: FW_MAJOR A -macro: FW_MINOR 46 +macro: FW_MINOR 47

1 0

[commit] r1897 - dev/mmc/sdhci
by repository service July 28, 2010

July 28, 2010

Author: quozl Date: Thu Jul 29 08:41:04 2010 New Revision: 1897 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1897 Log: SDHCI - Increased AData workaround timeout to 1200 uS, thus making the system work with 8G Adata cards while even further destroying write performance for every card. Modified: dev/mmc/sdhci/sdhci.fth Modified: dev/mmc/sdhci/sdhci.fth ============================================================================== --- dev/mmc/sdhci/sdhci.fth Thu Jul 22 09:02:07 2010 (r1896) +++ dev/mmc/sdhci/sdhci.fth Thu Jul 29 08:41:04 2010 (r1897) @@ -904,7 +904,7 @@ else ( block# r: #blocks fresh? ) \ Ugly ugly workaround for a problem with AData Class 6 SD cards \ They hang (data timeout) if you start a write too soon after a read. - " d# 900 us" evaluate + " d# 1200 us" evaluate r> if r> pre-write-erase else r> drop then ( block# ) write-multiple true to writing? ( ) then ( )

1 0

[commit] r1896 - in cpu/arm: . mmp2 mmp2/build
by repository service July 22, 2010

July 22, 2010

Author: wmb Date: Thu Jul 22 09:02:07 2010 New Revision: 1896 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1896 Log: ARM MMP2 - can now boot Linux. Added: cpu/arm/mmp2/build/Linux-debug.txt Modified: cpu/arm/boot.fth cpu/arm/mmp2/addrs.fth cpu/arm/mmp2/boot.fth cpu/arm/mmp2/build/Makefile cpu/arm/mmp2/devices.fth cpu/arm/mmp2/fw.bth cpu/arm/mmp2/mmuon.fth cpu/arm/scc.fth cpu/arm/target.fth Modified: cpu/arm/boot.fth ============================================================================== --- cpu/arm/boot.fth Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/boot.fth Thu Jul 22 09:02:07 2010 (r1896) @@ -54,14 +54,9 @@ str r1,'user dp \ set here str r8,'user memtop - sub sp,r8,#0x40 + sub sp,r8,#0x40 \ Guard band \ Now the stacks are just below the end of our memory - ps-size-t rs-size-t + # - sub r8,r8,* - sub r8,r8,r12 - str r8,'user limit - str r7,'user syscall-vec str r10,'user #args str r11,'user args @@ -77,7 +72,16 @@ dec sp,* dec sp,#0x20 str sp,'user sp0 + + mov r8,sp + + ps-size-t # + dec r8,* + sub r8,r8,r12 \ Heap size + str r8,'user limit \ Initial heap will be from limit to bottom of stack + inc sp,1cell \ account for the top of stack register + adr ip,'body cold c; Modified: cpu/arm/mmp2/addrs.fth ============================================================================== --- cpu/arm/mmp2/addrs.fth Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/mmp2/addrs.fth Thu Jul 22 09:02:07 2010 (r1896) @@ -2,7 +2,7 @@ h# 2000.0000 constant total-ram-size h# 1fc0.0000 constant fb-pa -h# 40.0000 constant fb-size +h# 20.0000 constant fb-size \ The screen use a little more than 1 MiB at 800x480x24 fb-pa constant available-ram-size @@ -11,7 +11,7 @@ ' (memory?) to memory? \ OFW implementation choices -h# 0040.0000 constant fw-pa +h# 1fe0.0000 constant fw-pa [ifdef] virtual-mode h# f700.0000 constant fw-virt-base @@ -21,17 +21,13 @@ 0 value fw-virt-size [then] -h# 0010.0000 constant /fw-ram +h# 0020.0000 constant /fw-ram h# 0110.0000 constant def-load-base \ The heap starts at RAMtop, which on this system is "fw-pa /fw-ram +" -\ We leave some memory in the /memory available list above the heap -\ for DMA allocation by the sound and USB driver. OFW's normal memory -\ usage thus fits in one 4M page-directory mapping region. - -h# 18.0000 constant heap-size +h# 10.0000 constant heap-size heap-size constant initial-heap-size h# 4000 constant page-table-pa Modified: cpu/arm/mmp2/boot.fth ============================================================================== --- cpu/arm/mmp2/boot.fth Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/mmp2/boot.fth Thu Jul 22 09:02:07 2010 (r1896) @@ -8,15 +8,19 @@ \ Put the arguments in safe registers sub r6,lr,#0x8c \ r6 points to header (lr set by code at origin) mov r7,#0 \ r7: functions - add r8,r6,0x200000 \ r8: memtop - 2MiB above load address + add r8,r6,`/fw-ram` \ r8: memtop - 2MiB above load address \ r9 is up mov r10,#0 \ r10: argc mov r11,r2 \ r11: argv (kernel args) mov r12,`initial-heap-size` \ r12: initial-heap-size - mov r3,#0 + + \ Debug - snapshot some register values + mov r3,#0x100 str r0,[r3],#4 str r1,[r3],#4 str r2,[r3],#4 + str r8,[r3],#4 + str r6,[r3],#4 b 'code start-forth \ Branch to the generic startup code end-code Added: cpu/arm/mmp2/build/Linux-debug.txt ============================================================================== --- /dev/null 00:00:00 1970 (empty, because file is newly added) +++ cpu/arm/mmp2/build/Linux-debug.txt Thu Jul 22 09:02:07 2010 (r1896) @@ -0,0 +1,32 @@ +\ This is a recipe for using OFW to debug into the Linux kernel +\ startup. You can cut/paste groups into the serial window + +fatload mmc 1:1 0x1f800000 ofw.rom ; boot 0x1f800000 +fatload mmc 1:1 0x1fe00000 ofw.rom ; boot 0x1fe00000 + +dev /memory 1f800000 400000 0 claim +load ext:\zimage + +1100134 till \ After uncompress +--bp + +208064 till \ About to turn on MMU +--bp + +text-off ofw-sections \ Put OFW in pdir +vector-base f.0000 80 move \ Clone vector table to xxxf.0000 +c0e page-table@ fff la+ l! \ Alias vector table at ffff.0000 +step \ Turn on MMU, high vector loc + +c0009fb0 till \ paging_init +--bp + +e35e041f c000b6e8 instruction! \ Stop clearing PDIR at Forth loc - Changes MODULES_VADDR +c000b6f0 till \ Just before second pmd_clear loop +--bp +2000.0000 to r14 \ Bump addr to skip OFW mapping + +e35104d4 c000b76c instruction! \ Change VMALLOC_END to device physical base to preserve device access + +\ Now can set breakpoints at other places +c0010130 till \ memzero Modified: cpu/arm/mmp2/build/Makefile ============================================================================== --- cpu/arm/mmp2/build/Makefile Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/mmp2/build/Makefile Thu Jul 22 09:02:07 2010 (r1896) @@ -5,14 +5,14 @@ CLIENTDIR=../../../../clients CLIENTPROGS= -all: mmp2.rom tags +all: ofw.rom tags -mmp2.tag: mmp2.rom +ofw.tag: ofw.rom -tags: mmp2.tag +tags: ofw.tag @${BASEDIR}/forth/lib/toctags ${BASEDIR} ${TAGFILES} -mmp2.rom: FORCE build ${CLIENTPROGS} +ofw.rom: FORCE build ${CLIENTPROGS} ./build $@ ../../${OS}/forth: Modified: cpu/arm/mmp2/devices.fth ============================================================================== --- cpu/arm/mmp2/devices.fth Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/mmp2/devices.fth Thu Jul 22 09:02:07 2010 (r1896) @@ -62,7 +62,7 @@ ' com1 is fallback-device 0 0 " d420b000" " /" begin-package - " screen" name + " display" name fload ${BP}/cpu/arm/mmp2/lcdcfg.fth fload ${BP}/cpu/arm/mmp2/dsi.fth @@ -94,7 +94,7 @@ ' display-remove is-remove ' display-selftest is-selftest end-package -devalias screen /screen +devalias screen /display devalias keyboard /keyboard Modified: cpu/arm/mmp2/fw.bth ============================================================================== --- cpu/arm/mmp2/fw.bth Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/mmp2/fw.bth Thu Jul 22 09:02:07 2010 (r1896) @@ -152,7 +152,8 @@ [ifdef] virtual-mode : (initial-heap) ( -- adr len ) sp0 @ ps-size - dict-limit tuck - ; [else] -: (initial-heap) ( -- adr len ) RAMtop heap-size ; + \ : (initial-heap) ( -- adr len ) RAMtop heap-size ; +: (initial-heap) ( -- adr len ) limit heap-size ; [then] ' (initial-heap) is initial-heap headers @@ -188,6 +189,7 @@ devalias nfs net//obp-tftp:last//nfs fload ${BP}/cpu/arm/linux.fth +h# 20.0000 to linux-params \ The Jasper Linux kernel fails unless the params are between 0x20.0000 and 0x20.4000 d# 2382 to arm-linux-machine-type \ Marvell Jasper \ Add a tag describing the linear frame buffer Modified: cpu/arm/mmp2/mmuon.fth ============================================================================== --- cpu/arm/mmp2/mmuon.fth Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/mmp2/mmuon.fth Thu Jul 22 09:02:07 2010 (r1896) @@ -11,7 +11,7 @@ : map-section ( pa+mode va -- ) - d# 18 rshift page-table@ + l! + d# 18 rshift page-table@ + tuck l! clean-d$-entry ; : map-sections ( pa mode va size -- ) 2>r + 2 or 2r> ( pa+mode va size ) @@ -22,13 +22,18 @@ drop ; +: ofw-sections ( -- ) + h# 0000.0000 h# c0e over fb-pa map-sections \ Cache and write bufferable + fw-pa h# c0e over /fw-ram map-sections \ Cache and write bufferable + fb-pa h# c06 over fb-size map-sections \ Write bufferable + h# d400.0000 h# c02 over h# 0040.0000 map-sections \ I/O - no caching or buffering +; + : setup-sections page-table-pa page-table! page-table-pa /page-table erase - h# 0000.0000 h# c0e 0 fb-pa map-sections \ Cache and write bufferable - fb-pa h# c06 fb-pa fb-size map-sections \ Write bufferable - h# d400.0000 h# c02 h# d400.0000 h# 0100.0000 map-sections \ I/O - no caching or buffering + ofw-sections ; \ Do we need to map SRAM and DDRC ? Modified: cpu/arm/scc.fth ============================================================================== --- cpu/arm/scc.fth Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/scc.fth Thu Jul 22 09:02:07 2010 (r1896) @@ -3,27 +3,84 @@ hex code scc-id@ ( -- n ) psh tos,sp mrc p15,0,tos,cr0,cr0,0 c; +code cache-type@ ( -- n ) psh tos,sp mrc p15,0,tos,cr0,cr0,1 c; +code tlb-type@ ( -- n ) psh tos,sp mrc p15,0,tos,cr0,cr0,3 c; code control@ ( -- n ) psh tos,sp mrc p15,0,tos,cr1,cr0,0 c; code ttbase@ ( -- n ) psh tos,sp mrc p15,0,tos,cr2,cr0,0 c; +code ttbase0@ ( -- n ) psh tos,sp mrc p15,0,tos,cr2,cr0,0 c; +code ttbase1@ ( -- n ) psh tos,sp mrc p15,0,tos,cr2,cr1,0 c; +code ttcontrol@ ( -- n ) psh tos,sp mrc p15,0,tos,cr2,cr2,0 c; code domain-access@ ( -- n ) psh tos,sp mrc p15,0,tos,cr3,cr0,0 c; code fault-status@ ( -- n ) psh tos,sp mrc p15,0,tos,cr5,cr0,0 c; +code i-fault-status@ ( -- n ) psh tos,sp mrc p15,0,tos,cr5,cr1,0 c; code fault-address@ ( -- n ) psh tos,sp mrc p15,0,tos,cr6,cr0,0 c; +code wp-fault-address@ ( -- n ) psh tos,sp mrc p15,0,tos,cr6,cr1,0 c; +code i-fault-address@ ( -- n ) psh tos,sp mrc p15,0,tos,cr6,cr2,0 c; code control! ( n -- ) mcr p15,0,tos,cr1,cr0,0 pop tos,sp c; code ttbase! ( n -- ) mcr p15,0,tos,cr2,cr0,0 pop tos,sp c; +code ttbase0! ( n -- ) mcr p15,0,tos,cr2,cr0,0 pop tos,sp c; +code ttbase1! ( n -- ) mcr p15,0,tos,cr2,cr1,0 pop tos,sp c; +code ttcontrol! ( n -- ) mcr p15,0,tos,cr2,cr2,0 pop tos,sp c; code domain-access! ( n -- ) mcr p15,0,tos,cr3,cr0,0 pop tos,sp c; code fault-status! ( n -- ) mcr p15,0,tos,cr5,cr0,0 pop tos,sp c; +code i-fault-status! ( n -- ) mcr p15,0,tos,cr5,cr1,0 pop tos,sp c; code fault-address! ( n -- ) mcr p15,0,tos,cr6,cr0,0 pop tos,sp c; -code flush-i&d$ ( -- ) mcr p15,0,r0,cr7,cr7,0 c; +code wp-fault-address! ( n -- ) mcr p15,0,tos,cr6,cr1,0 pop tos,sp c; +code i-fault-address! ( n -- ) mcr p15,0,tos,cr6,cr2,0 pop tos,sp c; + +code c7-wfi ( -- ) mcr p15,0,r0,cr7,cr0,4 c; + code flush-i$ ( -- ) mcr p15,0,r0,cr7,cr5,0 c; +code flush-i$-entry ( va -- ) mcr p15,0,tos,cr7,cr5,1 pop tos,sp c; +code flush-i$-entry-way ( sw -- ) mcr p15,0,tos,cr7,cr5,2 pop tos,sp c; +code flush-prefetch ( -- ) mcr p15,0,r0,cr7,cr5,4 c; +code flush-bt$ ( -- ) mcr p15,0,r0,cr7,cr5,6 c; +code flush-bt$-entry ( va -- ) mcr p15,0,tos,cr7,cr5,7 pop tos,sp c; + code flush-d$ ( -- ) mcr p15,0,r0,cr7,cr6,0 c; -code flush-d$-entry ( va -- ) mcr p15,0,tos,cr7,cr6,1 pop tos,sp c; +code flush-d$-entry ( va -- ) mcr p15,0,tos,cr7,cr6,1 pop tos,sp c; +code flush-d$-entry-way ( sw -- ) mcr p15,0,tos,cr7,cr6,2 pop tos,sp c; + +code flush-i&d$ ( -- ) mcr p15,0,r0,cr7,cr7,0 c; +code flush-u$ ( -- ) mcr p15,0,r0,cr7,cr7,0 c; +code flush-u$-entry ( va -- ) mcr p15,0,tos,cr7,cr7,1 pop tos,sp c; +code flush-u$-way ( sw -- ) mcr p15,0,tos,cr7,cr7,2 pop tos,sp c; + +code clean-d$ ( -- ) mcr p15,0,r0,cr7,cr10,0 c; code clean-d$-entry ( va -- ) mcr p15,0,tos,cr7,cr10,1 pop tos,sp c; +code clean-d$-way ( sw -- ) mcr p15,0,tos,cr7,cr10,2 pop tos,sp c; +code test&clean-d$ ( -- ) mcr p15,0,r0,cr7,cr10,3 c; + +code clean-u$ ( -- ) mcr p15,0,r0,cr7,cr11,0 c; +code clean-u$-entry ( va -- ) mcr p15,0,tos,cr7,cr11,1 pop tos,sp c; +code clean-u$-way ( sw -- ) mcr p15,0,tos,cr7,cr11,2 pop tos,sp c; + +code clean&flush-d$ ( -- ) mcr p15,0,r0,cr7,cr14,0 c; +code clean&flush-d$-entry ( va -- ) mcr p15,0,tos,cr7,cr14,1 pop tos,sp c; +code clean&flush-d$-entry-way ( sw -- ) mcr p15,0,tos,cr7,cr14,2 pop tos,sp c; +code test,clean&flush-d$ ( -- ) mcr p15,0,tos,cr7,cr14,3 c; + +code clean&flush-u$ ( -- ) mcr p15,0,r0,cr7,cr15,0 c; +code clean&flush-u$-entry ( va -- ) mcr p15,0,tos,cr7,cr15,1 pop tos,sp c; +code clean&flush-u$-entry-way ( sw -- ) mcr p15,0,tos,cr7,cr15,2 pop tos,sp c; + + + code drain-write-buffer ( -- ) mcr p15,0,r0,cr7,cr10,4 c; +alias data-sync-barrier drain-write-buffer +code drain-write-buffer ( -- ) mcr p15,0,r0,cr7,cr10,4 c; +code data-memory-barrier ( -- ) mcr p15,0,r0,cr7,cr10,5 c; + code flush-i&d-tlb ( -- ) mcr p15,0,r0,cr8,cr7,0 c; +code flush-i&d-tlb-entry ( va -- ) mcr p15,0,tos,cr8,cr7,1 pop tos,sp c; +code flush-i&d-tlb-asid ( as -- ) mcr p15,0,tos,cr8,cr7,2 pop tos,sp c; code flush-i-tlb ( -- ) mcr p15,0,r0,cr8,cr5,0 c; +code flush-i-tlb-entry ( va -- ) mcr p15,0,tos,cr8,cr5,1 pop tos,sp c; +code flush-i-tlb-asid ( as -- ) mcr p15,0,tos,cr8,cr5,2 pop tos,sp c; code flush-d-tlb ( -- ) mcr p15,0,r0,cr8,cr6,0 c; code flush-d-tlb-entry ( va -- ) mcr p15,0,tos,cr8,cr6,1 pop tos,sp c; +code flush-d-tlb-asid ( as -- ) mcr p15,0,tos,cr8,cr6,1 pop tos,sp c; code enable-odd-lfsr ( -- ) mcr p15,0,r0,cr15,cr1,1 c; code enable-even-lfsr ( -- ) mcr p15,0,r0,cr15,cr2,1 c; @@ -53,6 +110,23 @@ /cache-line +loop ; +code turn-off-dcache ( adr len -- ) + pop r0,sp \ tos:len r0:adr + cmp tos,#0 + <> if + begin + mcr p15,0,r0,cr7,cr10,1 \ Clean D$ entry + mcr p15,0,r0,cr7,cr6,1 \ Flush D$ entry + add r0,r0,#32 \ Advance to next line + decs tos,#32 \ Assume cache line size of 32 bytes + 0= until + then + mcr p15,0,r0,cr7,cr10,4 \ Drain write buffer + mrc p15,0,r0,cr1,cr0,0 \ Read control register + bic r0,r0,#4 \ Clear DCache enable bit + mcr p15,0,r0,cr1,cr0,0 \ Write control register +c; + \ System-dependent function to flush the entire cache \ (In normal ARM nomenclature, as used by most of the words in this file, \ "flush" means "invalidate without ensuring that cached data has been Modified: cpu/arm/target.fth ============================================================================== --- cpu/arm/target.fth Thu Jul 22 08:54:57 2010 (r1895) +++ cpu/arm/target.fth Thu Jul 22 09:02:07 2010 (r1896) @@ -36,9 +36,9 @@ /l-t constant /token-t /l-t constant /link-t /token-t constant /defer-t -/n-t 800 * constant user-size-t -/n-t 100 * constant ps-size-t -/n-t 100 * constant rs-size-t +/n-t th c00 * constant user-size-t +/n-t th 200 * constant ps-size-t +/n-t th 200 * constant rs-size-t /l-t constant /user#-t \ 32 bit host Forth compiling 32-bit target Forth

1 0

[commit] r1895 - cpu/x86/pc/olpc/via
by repository service July 21, 2010

July 21, 2010

Author: quozl Date: Thu Jul 22 08:54:57 2010 New Revision: 1895 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1895 Log: Q3A46 Modified: cpu/x86/pc/olpc/via/fw-version.fth Modified: cpu/x86/pc/olpc/via/fw-version.fth ============================================================================== --- cpu/x86/pc/olpc/via/fw-version.fth Thu Jul 22 08:42:19 2010 (r1894) +++ cpu/x86/pc/olpc/via/fw-version.fth Thu Jul 22 08:54:57 2010 (r1895) @@ -1,3 +1,3 @@ \ The overall firmware revision macro: FW_MAJOR A -macro: FW_MINOR 45 +macro: FW_MINOR 46

1 0

[commit] r1894 - dev/mmc/sdhci
by repository service July 21, 2010

July 21, 2010

Author: wmb Date: Thu Jul 22 08:42:19 2010 New Revision: 1894 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1894 Log: SDHCI - Increased AData workaround timeout to 900 uS, thus making the system work with 8G Adata cards while killing the write performance for every card. Modified: dev/mmc/sdhci/sdhci.fth Modified: dev/mmc/sdhci/sdhci.fth ============================================================================== --- dev/mmc/sdhci/sdhci.fth Tue Jul 20 05:49:44 2010 (r1893) +++ dev/mmc/sdhci/sdhci.fth Thu Jul 22 08:42:19 2010 (r1894) @@ -904,7 +904,7 @@ else ( block# r: #blocks fresh? ) \ Ugly ugly workaround for a problem with AData Class 6 SD cards \ They hang (data timeout) if you start a write too soon after a read. - " d# 20 us" evaluate + " d# 900 us" evaluate r> if r> pre-write-erase else r> drop then ( block# ) write-multiple true to writing? ( ) then ( )

1 0

[commit] r1893 - cpu/x86/pc/olpc/via
by repository service July 19, 2010

July 19, 2010

Author: quozl Date: Tue Jul 20 05:49:44 2010 New Revision: 1893 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1893 Log: Q3A45 Modified: cpu/x86/pc/olpc/via/fw-version.fth Modified: cpu/x86/pc/olpc/via/fw-version.fth ============================================================================== --- cpu/x86/pc/olpc/via/fw-version.fth Tue Jul 20 04:45:30 2010 (r1892) +++ cpu/x86/pc/olpc/via/fw-version.fth Tue Jul 20 05:49:44 2010 (r1893) @@ -1,3 +1,3 @@ \ The overall firmware revision macro: FW_MAJOR A -macro: FW_MINOR 44 +macro: FW_MINOR 45

1 0

[commit] r1892 - cpu/x86/pc/olpc/via
by repository service July 19, 2010

July 19, 2010

Author: wmb Date: Tue Jul 20 04:45:30 2010 New Revision: 1892 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1892 Log: OLPC trac 10017 - fixed regression in fs-update caused by NANDblaster changes. Modified: cpu/x86/pc/olpc/via/fsupdate.fth Modified: cpu/x86/pc/olpc/via/fsupdate.fth ============================================================================== --- cpu/x86/pc/olpc/via/fsupdate.fth Tue Jul 20 03:31:21 2010 (r1891) +++ cpu/x86/pc/olpc/via/fsupdate.fth Tue Jul 20 04:45:30 2010 (r1892) @@ -93,7 +93,7 @@ : data: ( "filename" -- ) safe-parse-word ( filename$ ) - nb-zd-#sectors if ( filename$ ) + nb-zd-#sectors -1 <> if ( filename$ ) 2drop " /nb-updater" ( filename$' ) else ( filename$ ) fn-buf place ( )

1 0

[commit] r1891 - cpu/x86/pc/olpc/via
by repository service July 19, 2010

July 19, 2010

Author: quozl Date: Tue Jul 20 03:31:21 2010 New Revision: 1891 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1891 Log: Q3A44 Modified: cpu/x86/pc/olpc/via/fw-version.fth Modified: cpu/x86/pc/olpc/via/fw-version.fth ============================================================================== --- cpu/x86/pc/olpc/via/fw-version.fth Tue Jul 20 01:36:41 2010 (r1890) +++ cpu/x86/pc/olpc/via/fw-version.fth Tue Jul 20 03:31:21 2010 (r1891) @@ -1,3 +1,3 @@ \ The overall firmware revision macro: FW_MAJOR A -macro: FW_MINOR 43 +macro: FW_MINOR 44

1 0

[commit] r1890 - dev/mmc/sdhci
by repository service July 19, 2010

July 19, 2010

Author: wmb Date: Tue Jul 20 01:36:41 2010 New Revision: 1890 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1890 Log: SDHCI driver - set the data timeout early in the card attach process because that seems to prevent spurious data timeouts from the Via VX855 chip. Modified: dev/mmc/sdhci/sdhci.fth Modified: dev/mmc/sdhci/sdhci.fth ============================================================================== --- dev/mmc/sdhci/sdhci.fth Tue Jul 20 01:33:51 2010 (r1889) +++ dev/mmc/sdhci/sdhci.fth Tue Jul 20 01:36:41 2010 (r1890) @@ -123,7 +123,6 @@ ; : internal-clock-off ( -- ) h# 2c cw@ 1 invert and h# 2c cw! ; - : card-clock-on ( -- ) h# 2c cw@ 4 or h# 2c cw! ; : card-clock-off ( -- ) h# 2c cw@ 4 invert and h# 2c cw! ; @@ -227,6 +226,11 @@ 0 instance value timeout? : .sderror ( isr -- ) + debug? if + ." Chip registers: " cr + chip h# 80 " ldump" evaluate cr + then + esr@ dup esr! ( isr esr ) dup 1 and if ( isr esr ) @@ -254,6 +258,8 @@ : isr-event? ( mask -- happened? ) h# 8000 or ( mask' ) isr@ 2dup and if ( mask' isr ) + dup h# 8000 and if dup .sderror then ( mask isr ) + \ Only clear the bits we will handle this time. \ If additional ISR bits are set, leave them set because \ later code will be waiting for them. In practice, the @@ -261,9 +267,8 @@ \ bit - mask 2 - which "2 wait" will handle. \ But we do go ahead and clear card removal/insertion \ events, because we don't handle them elsewhere. - swap h# c0 or over and isr! ( isr ) + swap h# c0 or and isr! ( ) - dup h# 8000 and if .sderror else drop then ( ) true else ( mask' isr ) \ DMA interrupt - the transfer crossed an address boundary @@ -282,6 +287,8 @@ 8 and if 0 cl@ 0 cl! 8 isr! then ( mask ) repeat ( mask isr ) + dup h# 8000 and if dup .sderror then ( mask isr ) + \ Only clear the bits we will handle this time. \ If additional ISR bits are set, leave them set because \ later code will be waiting for them. In practice, the @@ -289,9 +296,7 @@ \ bit - mask 2 - which "2 wait" will handle. \ But we do go ahead and clear card removal/insertion \ events, because we don't handle them elsewhere. - swap h# c0 or over and isr! ( isr ) - - dup h# 8000 and if .sderror else drop then ( ) + swap h# c0 or and isr! ( ) ; : wait-ready ( -- ) @@ -729,6 +734,7 @@ card-power-on d# 40 ms \ This delay is just a guess (20 was barely too slow for a Via board) card-inserted? 0= if card-power-off intstat-off false true exit then card-clock-slow d# 10 ms \ This delay is just a guess + reset-card \ Cmd 0 set-operating-conditions if intstat-off true true exit then false @@ -781,6 +787,8 @@ : attach-card ( -- okay? ) setup-host + set-timeout + ['] power-up-card catch if true true then if ( retry? ) \ The first try at powering up failed. if ( ) @@ -816,8 +824,6 @@ select-card \ Cmd 7 - Select - set-timeout - configure-transfer intstat-off

1 0

[commit] r1889 - dev
by repository service July 19, 2010

July 19, 2010

Author: wmb Date: Tue Jul 20 01:33:51 2010 New Revision: 1889 URL: http://tracker.coreboot.org/trac/openfirmware/changeset/1889 Log: OLPC keyboard driver - Don't send the F4 (enable scan) command as it sometimes confuses the EC. Modified: dev/pckbd.fth Modified: dev/pckbd.fth ============================================================================== --- dev/pckbd.fth Tue Jul 20 01:32:54 2010 (r1888) +++ dev/pckbd.fth Tue Jul 20 01:33:51 2010 (r1889) @@ -450,12 +450,12 @@ \ 7 set-leds d# 100 ms 0 set-leds then -[then] \ Leave the keyboard in scan set 2 (its default state), but also leave \ the 8042 in the mode where it translates to scan set 1. keyboard-present? if enable-scan then +[then] true to keyboard-probed? ;

1 0