coreboot January 2008

coreboot@coreboot.org

79 participants
347 discussions

Re: [LinuxBIOS] different versions of the GA-M57SLI-S4 (PLCC vs SPI)
by Carl-Daniel Hailfinger 10 Jan '08

10 Jan '08

On 23.09.2007 23:16, echelon(a)free.fr wrote: > Quoting Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net>: >> OK, I have the 0.2 and 0.3 datasheet of IT8716F and I can't find any >> difference in interfaces between the datasheets. Please be more specific >> what you think has been replaced (including page numbers). > Simply the pinout.. (please compare the pin-out diagram btw the 2 version of > the DS). I did. Pages 7-23 (Pin configuration) are exactly the same for 0.2 and 0.3. Please tell me the page number where you see any differences. > In fact, there could be another explanation : I talked about this issue with a > hardware specialist at my work (he has worked into the IC design..) and his > hypothesis is that the component which is actually used on this motherboard is a > "custom" revision especially crafted by ITE for Gigabyte.. (yes, they had > "volume" so they could afford to make a deal for a custom revision..) Maybe. I still want to find out why your data sheet seems to differ from mine. Regards, Carl-Daniel

8 20

[LinuxBIOS] r3045 build service
by LinuxBIOS information 10 Jan '08

10 Jan '08

Dear LinuxBIOS readers! This is the automated build check service of LinuxBIOS. The developer "duwe" checked in revision 3045 to the LinuxBIOS source repository and caused the following changes: Change Log: This patch removes '\n' from the help output since this looks a bit strange. After the patch [...] The line length is still below 80 characters. Signed-off-by: Bernhard Walle <bernhard.walle(a)gmx.de> Acked-by: Torsten Duwe <duwe(a)lst.de> Build Log: Compilation of amd:serengeti_cheetah_fam10 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=3045&device=serengeti_che… If something broke during this checkin please be a pain in duwe's neck until the issue is fixed. If this issue is not fixed within 24h the revision should be backed out. Best regards, LinuxBIOS automatic build system

1 0

[LinuxBIOS] [PATCH] Flashrom: "Fix" help output
by Bernhard Walle 10 Jan '08

10 Jan '08

This patch removes '\n' from the help output since this looks a bit strange: ----------------------------------------------------------------------------- ... -i | --image <name>: only flash image name from flash layout If no file is specified, then all that happens is that flash info is dumped. ----------------------------------------------------------------------------- After the patch: ----------------------------------------------------------------------------- ... -i | --image <name>: only flash image name from flash layout If no file is specified, then all that happens is that flash info is dumped. ----------------------------------------------------------------------------- The line length is still below 80 characters. Signed-off-by: Bernhard Walle <bernhard.walle(a)gmx.de> --- flashrom.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) --- a/flashrom.c +++ b/flashrom.c @@ -206,7 +206,7 @@ void usage(const char *name) " -f | --force: force write without checking image\n" " -l | --layout <file.layout>: read rom layout from file\n" " -i | --image <name>: only flash image name from flash layout\n" - "\n" " If no file is specified, then all that happens\n" + "\n" " If no file is specified, then all that happens" " is that flash info is dumped.\n\n"); exit(1); }

2 1

[LinuxBIOS] r3045 - trunk/util/flashrom
by svn＠openbios.org 10 Jan '08

10 Jan '08

Author: duwe Date: 2008-01-11 01:32:07 +0100 (Fri, 11 Jan 2008) New Revision: 3045 Modified: trunk/util/flashrom/flashrom.c Log: This patch removes '\n' from the help output since this looks a bit strange. After the patch [...] The line length is still below 80 characters. Signed-off-by: Bernhard Walle <bernhard.walle(a)gmx.de> Acked-by: Torsten Duwe <duwe(a)lst.de> Modified: trunk/util/flashrom/flashrom.c =================================================================== --- trunk/util/flashrom/flashrom.c 2008-01-10 17:59:25 UTC (rev 3044) +++ trunk/util/flashrom/flashrom.c 2008-01-11 00:32:07 UTC (rev 3045) @@ -206,7 +206,7 @@ " -f | --force: force write without checking image\n" " -l | --layout <file.layout>: read rom layout from file\n" " -i | --image <name>: only flash image name from flash layout\n" - "\n" " If no file is specified, then all that happens\n" + "\n" " If no file is specified, then all that happens" " is that flash info is dumped.\n\n"); exit(1); }

1 0

[LinuxBIOS] [PATCH] fix build race in LBv3
by Robert Millan 10 Jan '08

10 Jan '08

This fixes a build race (make -j2): make: *** No rule to make target `/home/rmh/hacking/linuxbios/LinuxBIOSv3/build/stage0-prefixed.o', \ needed by `/home/rmh/hacking/linuxbios/LinuxBIOSv3/build/linuxbios.initram'. Stop. -- Robert Millan <GPLv2> I know my rights; I want my phone call! <DRM> What use is a phone call, if you are unable to speak? (as seen on /.)

2 1

[LinuxBIOS] r552 - LinuxBIOSv3/arch/x86
by svn＠openbios.org 10 Jan '08

10 Jan '08

Author: hailfinger Date: 2008-01-10 22:13:19 +0100 (Thu, 10 Jan 2008) New Revision: 552 Modified: LinuxBIOSv3/arch/x86/Makefile Log: This fixes a build race (make -j2): make: *** No rule to make target `LinuxBIOSv3/build/stage0-prefixed.o', needed by `LinuxBIOSv3/build/linuxbios.initram'. Stop. Signed-off-by: Robert Millan <rmh(a)aybabtu.com> Acked-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Modified: LinuxBIOSv3/arch/x86/Makefile =================================================================== --- LinuxBIOSv3/arch/x86/Makefile 2008-01-09 23:09:23 UTC (rev 551) +++ LinuxBIOSv3/arch/x86/Makefile 2008-01-10 21:13:19 UTC (rev 552) @@ -135,7 +135,7 @@ $(patsubst %,$(obj)/arch/x86/%,$(STAGE0_CAR_OBJ)) \ $(STAGE0_MAINBOARD_OBJ) $(STAGE0_CHIPSET_OBJ) -$(obj)/stage0.o $(obj)/stage0.init: $(STAGE0_OBJ) +$(obj)/stage0.o $(obj)/stage0.init $(obj)/stage0-prefixed.o: $(STAGE0_OBJ) $(Q)# We need to be careful. If stage0.o gets bigger than $(Q)# 0x4000 - 0x100, we will end up with a 4 gig file. $(Q)# I wonder if that behavior is on purpose.

1 0

[LinuxBIOS] r3044 build service
by LinuxBIOS information 10 Jan '08

10 Jan '08

Dear LinuxBIOS readers! This is the automated build check service of LinuxBIOS. The developer "hailfinger" checked in revision 3044 to the LinuxBIOS source repository and caused the following changes: Change Log: Add a workaround for a bug in some binutils version which strictly interpret whitespace as macro argument delimiter. Since the code is preprocessed by gcc and the tokenizer may insert whitespace, that can fail. http://sourceware.org/bugzilla/show_bug.cgi?id=669 Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Acked-by: Jordan Crouse <jordan.crouse(a)amd.com> Build Log: Compilation of amd:serengeti_cheetah_fam10 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=3044&device=serengeti_che… If something broke during this checkin please be a pain in hailfinger's neck until the issue is fixed. If this issue is not fixed within 24h the revision should be backed out. Best regards, LinuxBIOS automatic build system

1 0

[LinuxBIOS] r3043 build service
by LinuxBIOS information 10 Jan '08

10 Jan '08

Dear LinuxBIOS readers! This is the automated build check service of LinuxBIOS. The developer "hailfinger" checked in revision 3043 to the LinuxBIOS source repository and caused the following changes: Change Log: This patch introduces 4k CAR size granularity for the AMD x86 CAR code. For the old supported CAR sizes, the newly generated code is equivalent, so it should be a no-brainer. Benefits: * a nice code size reduction * less #ifdef clutter for Family 10h * paranoid checks for CAR size * clear abstractions This has been tested by Marc Jones and Jordan Crouse. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Acked-by: Marc Jones <marc.jones(a)amd.com> Build Log: Compilation of amd:serengeti_cheetah_fam10 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=3043&device=serengeti_che… If something broke during this checkin please be a pain in hailfinger's neck until the issue is fixed. If this issue is not fixed within 24h the revision should be backed out. Best regards, LinuxBIOS automatic build system

1 0

[LinuxBIOS] [PATCH] v2: rewrite AMD K* CAR code
by Carl-Daniel Hailfinger 10 Jan '08

10 Jan '08

This patch is an attempt at introducing 4k CAR size granularity for the AMD x86 CAR code. For the old supported CAR sizes, the newly generated code is equivalent, so it should be a no-brainer. Benefits: * a nice code size reduction * less #ifdef clutter for Family 10h * paranoid checks for CAR size * clear abstractions Not tested, as I lack hardware working with v2. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Index: LinuxBIOSv2-CARx86/src/cpu/amd/car/cache_as_ram.inc =================================================================== --- LinuxBIOSv2-CARx86/src/cpu/amd/car/cache_as_ram.inc (Revision 3039) +++ LinuxBIOSv2-CARx86/src/cpu/amd/car/cache_as_ram.inc (Arbeitskopie) @@ -2,6 +2,7 @@ * This file is part of the LinuxBIOS project. * * Copyright (C) 2005-2007 Advanced Micro Devices, Inc. + * Copyright (C) 2008 Carl-Daniel Hailfinger * * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by @@ -120,82 +121,70 @@ jmp clear_fixed_var_mtrr clear_fixed_var_mtrr_out: -#if CacheSize == 0x10000 - /* enable caching for 64K using fixed mtrr */ - movl $0x268, %ecx /* fix4k_c0000*/ - #if CAR_FAM10 == 1 - movl $0x1e1e1e1e, %edx /* WB MEM type */ - #else - movl $0x06060606, %edx /* WB IO type */ - #endif - - movl %edx, %eax - wrmsr - movl $0x269, %ecx - wrmsr +/* 0x06 is the WB IO type for a given 4k segment. + * 0x1e is the MEM IO type for a given 4k segment (K10 and above). + * segs is the number of 4k segments in the area of the particular + * register we want to use for CAR. + * reg is the register where the IO type should be stored. + */ +.macro extractmask segs, reg +.if \segs <= 0 + /* The xorl here is superfluous because at the point of first execution + * of this macro, %eax and %edx are cleared. Later invocations of this + * macro will have a monotonically increasing segs parameter. + */ + xorl \reg, \reg +#if CAR_FAM10 == 1 +.elseif \segs == 1 + movl $0x1e000000, \reg /* WB MEM type */ +.elseif \segs == 2 + movl $0x1e1e0000, \reg /* WB MEM type */ +.elseif \segs == 3 + movl $0x1e1e1e00, \reg /* WB MEM type */ +.elseif \segs >= 4 + movl $0x1e1e1e1e, \reg /* WB MEM type */ +#else +.elseif \segs == 1 + movl $0x06000000, \reg /* WB IO type */ +.elseif \segs == 2 + movl $0x06060000, \reg /* WB IO type */ +.elseif \segs == 3 + movl $0x06060600, \reg /* WB IO type */ +.elseif \segs >= 4 + movl $0x06060606, \reg /* WB IO type */ #endif +.endif +.endm -#if CacheSize == 0xc000 - /* enable caching for 16K using fixed mtrr */ - movl $0x268, %ecx /* fix4k_c4000*/ - #if CAR_FAM10 == 1 - movl $0x1e1e1e1e, %edx /* WB MEM type */ - #else - movl $0x06060606, %edx /* WB IO type */ - #endif - xorl %eax, %eax - wrmsr - /* enable caching for 32K using fixed mtrr */ - movl $0x269, %ecx /* fix4k_c8000*/ - #if CAR_FAM10 == 1 - movl $0x1e1e1e1e, %edx /* WB MEM type */ - #else - movl $0x06060606, %edx /* WB IO type */ - #endif - movl %edx, %eax - wrmsr +/* size is the cache size in bytes we want to use for CAR. + * windowoffset is the 32k-aligned window into CAR size + */ +.macro simplemask carsize, windowoffset + simplemask_helper (((\carsize - \windowoffset) / 0x1000) - 4), %eax + simplemask_helper (((\carsize - \windowoffset) / 0x1000)), %edx +.endm + +#if CacheSize > 0x10000 +#error Invalid CAR size, must be at most 64k. #endif +#if CacheSize < 0x1000 +#error Invalid CAR size, must be at least 4k. This is a processor limitation. +#endif +#if (CacheSize & (0x1000 - 1)) +#error Invalid CAR size, is not a multiple of 4k. This is a processor limitation. +#endif - -#if CacheSize == 0x8000 - /* enable caching for 32K using fixed mtrr */ - movl $0x269, %ecx /* fix4k_c8000*/ - #if CAR_FAM10 == 1 - movl $0x1e1e1e1e, %edx /* WB MEM type */ - #else - movl $0x06060606, %edx /* WB IO type */ - #endif - movl %edx, %eax - wrmsr +#if CacheSize > 0x8000 + /* enable caching for 32K-64K using fixed mtrr */ + movl $0x268, %ecx /* fix4k_c0000*/ + simplemask CacheSize, 0x8000 + wrmsr #endif -#if CacheSize < 0x8000 - /* enable caching for 16K/8K/4K using fixed mtrr */ - movl $0x269, %ecx /* fix4k_cc000*/ - #if CacheSize == 0x4000 - #if CAR_FAM10 == 1 - movl $0x1e1e1e1e, %edx /* WB MEM type */ - #else - movl $0x06060606, %edx /* WB IO type */ - #endif - #endif - #if CacheSize == 0x2000 - #if CAR_FAM10 == 1 - movl $0x1e1e0000, %edx /* WB MEM type */ - #else - movl $0x06060000, %edx /* WB IO type */ - #endif - #endif - #if CacheSize == 0x1000 - #if CAR_FAM10 == 1 - movl $0x1e000000, %edx /* WB MEM type */ - #else - movl $0x06000000, %edx /* WB IO type */ - #endif - #endif - xorl %eax, %eax + /* enable caching for 0-32K using fixed mtrr */ + movl $0x269, %ecx /* fix4k_c8000*/ + simplemask CacheSize, 0 wrmsr -#endif /* enable memory access for first MBs using top_mem */ movl $TOP_MEM, %ecx

6 15

[LinuxBIOS] r3044 - trunk/LinuxBIOSv2/src/cpu/amd/car
by svn＠openbios.org 10 Jan '08

10 Jan '08

Author: hailfinger Date: 2008-01-10 18:59:25 +0100 (Thu, 10 Jan 2008) New Revision: 3044 Modified: trunk/LinuxBIOSv2/src/cpu/amd/car/cache_as_ram.inc Log: Add a workaround for a bug in some binutils version which strictly interpret whitespace as macro argument delimiter. Since the code is preprocessed by gcc and the tokenizer may insert whitespace, that can fail. http://sourceware.org/bugzilla/show_bug.cgi?id=669 Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Acked-by: Jordan Crouse <jordan.crouse(a)amd.com> Modified: trunk/LinuxBIOSv2/src/cpu/amd/car/cache_as_ram.inc =================================================================== --- trunk/LinuxBIOSv2/src/cpu/amd/car/cache_as_ram.inc 2008-01-10 17:48:25 UTC (rev 3043) +++ trunk/LinuxBIOSv2/src/cpu/amd/car/cache_as_ram.inc 2008-01-10 17:59:25 UTC (rev 3044) @@ -160,8 +160,15 @@ * windowoffset is the 32k-aligned window into CAR size */ .macro simplemask carsize, windowoffset + .set gas_bug_workaround,(((\carsize - \windowoffset) / 0x1000) - 4) + extractmask gas_bug_workaround, %eax + .set gas_bug_workaround,(((\carsize - \windowoffset) / 0x1000)) + extractmask gas_bug_workaround, %edx +/* Without the gas bug workaround, the entire macro would consist only of the + * two lines below. extractmask (((\carsize - \windowoffset) / 0x1000) - 4), %eax extractmask (((\carsize - \windowoffset) / 0x1000)), %edx + */ .endm #if CacheSize > 0x10000

1 0

← Newer
1
...
24
25
26
27
28
29
30
...
35
Older →

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

coreboot January 2008