coreboot August 2007

coreboot@coreboot.org

78 participants
202 discussions

[LinuxBIOS] [PATCH] Move common utilities to util/
by Uwe Hermann 03 Sep '07

03 Sep '07

Hi, here's my proposal for moving (via 'svn mv') common utilities which are not specific to a certain version of LinuxBIOS into the global util/ (they are now in the v2 util/). Todo ---- abuild -- needs discussion; will it work for v3 eventually? ADLO -- is/should be version-independant analysis buildrom -- not related to Jordan's buildrom? Needed in v2? dump_mmcr getpir lbtdump -- Do we need this at all? Can't lxbios do the same? mptable probe_superio resetcf Projects which should _not_ be moved ------------------------------------ newconfig -- v2-specific optionlist -- v2-specific options -- needed for v2 build, but we have a fork in v3 (FIXME?) romcc -- not used in v3 vgabios -- we use a different/modified version in v3, so leave this Already moved ------------- flashrom lxbios Comments welcome. If nobody complains, I'll move the projects tomorrow or so. Another discussion is needed as to which of the utilities we should draw back into v2/v3 via svn:externals (IMHO _not_ all of them, only often-used ones). After a quick look I'd suggest the following: - adlo (while we're at it, we should rename it from ADLO to adlo, IMO) This will be used often for booting Windows and OpenBSD soon... - flashrom (already done), often used for flashing. - lxbios (already done), often used for changing options. Commens welcome. Uwe. -- http://www.hermann-uwe.de | http://www.holsham-traders.de http://www.crazy-hacks.org | http://www.unmaintained-free-software.org

4 8

[LinuxBIOS] addr_offset help
by Joseph Smith 01 Sep '07

01 Sep '07

Can someone explain the significance of the addr_offset variable in raminit.c on the i440bx and i82810? Also why is this set to 0x1d0 for the Mode register set (MRS)? Thanks - Joe

2 5

[LinuxBIOS] hotplugging flash piece on A8N-E successful
by extinct man 01 Sep '07

01 Sep '07

I don't know whether this can be called "best practice" on mainbords, but with a handle super glued to those SST 49LF040 I repeatedly hot-unplugged a badly flashed (non working) flash and then used flashrom to reflash it. it worked all the time. Even an 8 Mbit piece 49LF080A works with the Asus a8ne, so maybe a kernel may be squeezed in there as well. thus, no need for an external eprommer :-) --Q Wissenswertes für Bastler und Hobby Handwerker. BE A BETTER HEIMWERKER! www.yahoo.de/clever

6 8

[LinuxBIOS] r2755 build service
by LinuxBIOS information 31 Aug '07

31 Aug '07

Dear LinuxBIOS readers! This is the automated build check service of LinuxBIOS. The developer "uwe" checked in revision 2755 to the LinuxBIOS source repository and caused the following changes: Change Log: Fix typo for the ITE IT8712F (trivial). The default for LDN 5 (keyboard), index 0xF0 is not 0x00 but rather 0x08 as per datasheet. Signed-off-by: Uwe Hermann <uwe(a)hermann-uwe.de> Acked-by: Uwe Hermann <uwe(a)hermann-uwe.de> Build Log: Compilation of arima:hdama is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2755&device=hdama&vendor=… Compilation of ibm:e325 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2755&device=e325&vendor=i… Compilation of ibm:e326 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2755&device=e326&vendor=i… Compilation of iei:juki-511p is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2755&device=juki-511p&ven… Compilation of iwill:dk8s2 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2755&device=dk8s2&vendor=… Compilation of iwill:dk8x is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2755&device=dk8x&vendor=i… If something broke during this checkin please be a pain in uwe'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] r2755 - trunk/LinuxBIOSv2/util/probe_superio
by svn＠openbios.org 31 Aug '07

31 Aug '07

Author: uwe Date: 2007-08-31 22:51:00 +0200 (Fri, 31 Aug 2007) New Revision: 2755 Modified: trunk/LinuxBIOSv2/util/probe_superio/probe_superio.c Log: Fix typo for the ITE IT8712F (trivial). The default for LDN 5 (keyboard), index 0xF0 is not 0x00 but rather 0x08 as per datasheet. Signed-off-by: Uwe Hermann <uwe(a)hermann-uwe.de> Acked-by: Uwe Hermann <uwe(a)hermann-uwe.de> Modified: trunk/LinuxBIOSv2/util/probe_superio/probe_superio.c =================================================================== --- trunk/LinuxBIOSv2/util/probe_superio/probe_superio.c 2007-08-30 10:29:15 UTC (rev 2754) +++ trunk/LinuxBIOSv2/util/probe_superio/probe_superio.c 2007-08-31 20:51:00 UTC (rev 2755) @@ -184,7 +184,7 @@ 0x00,NANA,NANA,EOT}}, {0x5, {0x30,0x60,0x61,0x62,0x63,0x70,0x71,0xf0,EOT}, - {0x01,0x00,0x60,0x00,0x64,0x01,0x02,0x00,EOT}}, + {0x01,0x00,0x60,0x00,0x64,0x01,0x02,0x08,EOT}}, {0x6, {0x30,0x70,0x71,0xf0,EOT}, {0x00,0x0c,0x02,0x00,EOT}},

1 0

[LinuxBIOS] Please update http://www.linuxbios.org/Development_Guidelines
by Markus Boas 30 Aug '07

30 Aug '07

I don't know how to make a valid patch. I have added support for the Winbond W29EE011 Chip. This is an old device. But I found it on same mainboards. Greeding Markus

6 15

[LinuxBIOS] [PATCH] probe_superio: add ITE support
by Carl-Daniel Hailfinger 30 Aug '07

30 Aug '07

This adds probing for almost all recent (since 1999) ITE Super I/O chips to probe_superio. Not much of the configuration is dumped, however I did verify against all ITE datasheets (including those not available any more) that the probing was non-destructive. More information can be extracted easily, however this needs loads of datasheet surfing. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> --- util/probe_superio/probe_superio.c (Revision 2744) +++ util/probe_superio/probe_superio.c (Arbeitskopie) @@ -3,6 +3,7 @@ * * Copyright (C) 2006 Ronald Minnich <rminnich(a)gmail.com> * Copyright (C) 2006 coresystems GmbH <info(a)coresystems.de> + * Copyright (C) 2007 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 @@ -32,11 +33,16 @@ * if we need this. */ -unsigned char regval(unsigned short port, unsigned short reg) { +unsigned char regval(unsigned short port, unsigned char reg) { outb(reg, port); return inb(port+1); } +void regwrite(unsigned short port, unsigned char reg, unsigned char val) { + outb(reg, port); + outb(val, port+1); +} + void dump_ns8374(unsigned short port) { printf("Enables: 21=%02x, 22=%02x, 23=%02x, 24=%02x, 26=%02x\n", @@ -46,15 +52,13 @@ /* check COM1. This is all we care about at present. */ printf("COM 1 is Globally %s\n", regval(port,0x26)&8 ? "disabled" : "enabled"); /* select com1 */ - outb(0x7, port); - outb(3, port+1); + regwrite(port, 0x07, 0x03); printf("COM 1 is locally %s\n", regval(port, 0x30) & 1 ? "enabled" : "disabled"); printf("COM1 60=%02x, 61=%02x, 70=%02x, 71=%02x, 74=%02x, 75=%02x, f0=%02x\n", regval(port, 0x60), regval(port, 0x61), regval(port, 0x70), regval(port, 0x71), regval(port, 0x74), regval(port, 0x75), regval(port, 0xf0)); /* select gpio */ - outb(0x7, port); - outb(7, port+1); + regwrite(port, 0x07, 0x07); printf("GPIO is %s\n", regval(port, 0x30) & 1 ? "enabled" : "disabled"); printf("GPIO 60=%02x, 61=%02x, 70=%02x, 71=%02x, 74=%02x, 75=%02x, f0=%02x\n", regval(port, 0x60), regval(port, 0x61), regval(port, 0x70), regval(port, 0x71), @@ -85,38 +89,33 @@ printf("2b=%02x\n", regval(port, 0x2b)); /* select UART 1 */ - outb(0x07, port); - outb(0x01, port+1); + regwrite(port, 0x07, 0x01); printf("UART1 is %s\n", regval(port, 0x30) & 1 ? "enabled" : "disabled"); printf("UART1 base=%02x%02x, irq=%02x, mode=%s\n", regval(port, 0x60), regval(port, 0x61), regval(port, 0x70)&0x0f, regval(port, 0xf0)&0x10 ? "RS485":"RS232"); /* select UART 2 */ - outb(0x07, port); - outb(0x02, port+1); + regwrite(port, 0x07, 0x02); printf("UART2 is %s\n", regval(port, 0x30) & 1 ? "enabled" : "disabled"); printf("UART2 base=%02x%02x, irq=%02x, mode=%s\n", regval(port, 0x60), regval(port, 0x61), regval(port, 0x70)&0x0f, regval(port, 0xf0)&0x10 ? "RS485":"RS232"); /* select Parport */ - outb(0x07, port); - outb(0x03, port+1); + regwrite(port, 0x07, 0x03); printf("PARPORT is %s\n", regval(port, 0x30) & 1 ? "enabled" : "disabled"); printf("PARPORT base=%02x%02x, irq=%02x\n", regval(port, 0x60), regval(port, 0x61), regval(port, 0x70)&0x0f); /* select hw monitor */ - outb(0x07, port); - outb(0x04, port+1); + regwrite(port, 0x07, 0x04); printf("HW monitor is %s\n", regval(port, 0x30) & 1 ? "enabled" : "disabled"); printf("HW monitor base=%02x%02x, irq=%02x\n", regval(port, 0x60), regval(port, 0x61), regval(port, 0x70)&0x0f); /* select gpio */ - outb(0x07, port); - outb(0x06, port+1); + regwrite(port, 0x07, 0x05); printf("GPIO is %s\n", regval(port, 0x30) & 1 ? "enabled" : "disabled"); printf("GPIO 70=%02x, e0=%02x, e1=%02x, e2=%02x, e3=%02x, e4=%02x, e5=%02x\n", regval(port, 0x70), regval(port, 0xe0), regval(port, 0xe1), regval(port, 0xe2), @@ -130,7 +129,20 @@ } +void +dump_ite(unsigned short port) +{ + int i; + printf ("ITE chip\n"); + for (i=0x20; i<=0x2c; i++) + printf("index %02x=%02x\n", i, regval(port, i)); + + /* select UART 1 */ + regwrite(port, 0x07, 0x01); + printf("UART1 is %s\n", regval(port, 0x30) & 1 ? "enabled" : "disabled"); +} + void probe_idregs_simple(unsigned short port){ unsigned char id; @@ -167,9 +179,10 @@ void probe_idregs_fintek(unsigned short port){ - unsigned int vid, did; + unsigned int vid, did, success; // Enable configuration sequence (Fintek uses this for example) + // Older ITE chips have the same enable sequence outb(0x87, port); outb(0x87, port); @@ -185,13 +198,10 @@ } did = inb(port+1); - outb(0x21, port); - did = did|(inb(port+1)<<8); + did = did|(regval(port, 0x21)<<8); - outb(0x23, port); - vid = inb(port+1); - outb(0x24, port); - vid = vid|(inb(port+1)<<8); + vid = regval(port, 0x23); + vid = vid|(regval(port, 0x24)<<8); printf("SuperIO found at 0x%02x: vid=0x%04x/did=0x%04x\n", port, vid, did); @@ -199,23 +209,104 @@ return; // printf("%s\n", familyid[id]); + switch(did) { + case 0x1087: // reversed for ITE8710 + success = 1; + dump_ite(port); + // disable configuration + regwrite(port, 0x02, 0x02); + break; + default: + break; + } switch(vid) { case 0x3419: + success = 1; dump_fintek(port, did); break; default: - printf("no dump for 0x%04x/0x%04x\n", vid, did); break; } + if (!success) + printf("no dump for vid 0x%04x, did 0x%04x\n", vid, did); - // disable configuration + // disable configuration (for Fintek, doesn't hurt ITE) outb(0xaa, port); } void +probe_idregs_ite(unsigned short port){ + unsigned int id, chipver; + + // Enable configuration sequence (ITE uses this for newer IT87[012]x) + // IT871[01] uses 0x87, 0x87 -> fintek detection should handle it + // IT8761 uses 0x87, 0x61, 0x55, 0x55/0xaa + // IT86xx series uses different ports + // IT8661 uses 0x86, 0x61, 0x55/0xaa, 0x55/0xaa and 32 more writes + // IT8673 uses 0x86, 0x80, 0x55/0xaa, 0x55/0xaa and 32 more writes + outb(0x87, port); + outb(0x01, port); + outb(0x55, port); + if (port == 0x2e) + outb(0x55, port); + else + outb(0xAA, port); + + outb(0x20, port); + if (inb(port) != 0x20) { + if (inb(port) == 0xff ) + printf ("No SuperIO chip found at 0x%04x\n", port); + else + printf("probing 0x%04x, failed (0x%02x), data returns 0x%02x\n", + port, inb(port), inb(port+1)); + return; + } + + // Read Chip ID Byte 1, we already issued the write to port + id = inb(port+1); + id <<= 8; + + // Read Chip ID Byte 2 + id |= regval(port, 0x21); + + // Read Chip Version, only bit 3..0 for all IT87xx + chipver = regval(port, 0x22) & 0x0f; + + /* ID Mapping Table + unknown -> IT8711 (no datasheet) + unknown -> IT8722 (no datasheet) + 0x8702 -> IT8702 + 0x8705 -> IT8700 or IT8705 + 0x8710 -> IT8710 + 0x8712 -> IT8712 + 0x8716 -> IT8716 or IT8726 (identical except CPU voltage control) + 0x8718 -> IT8718 + */ + printf("ITE? SuperIO found at 0x%02x: id=0x%04x, chipver=0x%01x\n", + port, id, chipver); + + switch(id) { + case 0x8702: + case 0x8705: + case 0x8710: //pointless, IT8710 has different enable sequence + case 0x8712: + case 0x8716: + case 0x8718: + dump_ite(port); + break; + default: + printf("no dump for id 0x%04x\n", id); + break; + } + // disable configuration + regwrite(port, 0x02, 0x02); +} + +void probe_superio(unsigned short port) { probe_idregs_simple(port); probe_idregs_fintek(port); + probe_idregs_ite(port); } int

3 2

[LinuxBIOS] r2754 build service
by LinuxBIOS information 30 Aug '07

30 Aug '07

Dear LinuxBIOS readers! This is the automated build check service of LinuxBIOS. The developer "uwe" checked in revision 2754 to the LinuxBIOS source repository and caused the following changes: Change Log: Add support for the Athlon64 x2 5000+ CPU. A trivial one-liner for the CPU I happen to have. The sales docs said it's a "G1 revision", but the Rev F code works just fine. Signed-off-by: Torsten Duwe <duwe(a)lst.de> Acked-by: Uwe Hermann <uwe(a)hermann-uwe.de> Build Log: Compilation of arima:hdama is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2754&device=hdama&vendor=… Compilation of ibm:e325 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2754&device=e325&vendor=i… Compilation of ibm:e326 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2754&device=e326&vendor=i… Compilation of iei:juki-511p is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2754&device=juki-511p&ven… Compilation of iwill:dk8s2 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2754&device=dk8s2&vendor=… Compilation of iwill:dk8x is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2754&device=dk8x&vendor=i… If something broke during this checkin please be a pain in uwe'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] r2753 build service
by LinuxBIOS information 30 Aug '07

30 Aug '07

Dear LinuxBIOS readers! This is the automated build check service of LinuxBIOS. The developer "uwe" checked in revision 2753 to the LinuxBIOS source repository and caused the following changes: Change Log: Add support for the Winbond W29EE011. Signed-off-by: Markus Boas <ryven(a)ryven.de> Acked-by: Uwe Hermann <uwe(a)hermann-uwe.de> Build Log: Compilation of arima:hdama is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2753&device=hdama&vendor=… Compilation of ibm:e325 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2753&device=e325&vendor=i… Compilation of ibm:e326 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2753&device=e326&vendor=i… Compilation of iei:juki-511p is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2753&device=juki-511p&ven… Compilation of iwill:dk8s2 is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2753&device=dk8s2&vendor=… Compilation of iwill:dk8x is still broken See the error log at http://qa.linuxbios.org/log_buildbrd.php?revision=2753&device=dk8x&vendor=i… If something broke during this checkin please be a pain in uwe'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] Problem with Buildroom and GeodeLX (db800)
by Steffen D. 30 Aug '07

30 Aug '07

Hello together! Im trying to build a linuxbios for my AMD Geode LX db800 board. First time i tryed it directly with linuxbiosv2 and v3. After i find out that v3 wont work i tryed it with v2 and dont get anything working - i tryed it by hand with the lx_vsa and so on. I read a lot of papers i found on the net but didnt get forward so i asked on the irc channel, thanks uwe for you help! Im now trying it with buildrom - my main problem is that i dont get any vga and rs232 output. I have a postcard which shows these values: 75,11, 40, 24, 66, F8, EE It takes about 5 then EE is showed. I have a very big ROM 4Mbit. The resulting image is 512kb - i tryed this and get the post-output. When i try "cat foo.rom foo.rom ... > final.rom" (8 times foo.rom) to fill up 4mbit it gives the same result - no output only on the post-card. You find my .config for buildrom on nopaste - http://nopaste.php-quake.net/3269 If have done everything exactly like in the read me. I have a hardware flash device its called galep-5. In very sure its working (also with my roms) because i have flashed the old bios from this chip and cloned it to another one. Im also testing some times if its still working by flashing the old bios on it and booting it. Here are all details: SuperIO - Winbond83627hG AMD Geode LX 700 / CS5536AD ROM - SST49LF004B (4mbit) Devel-env: debian gcc-4.1 (3.4 wont work at all with lxbios for me!) Im very sure that all the hardware works - im have done everything a lot of times again to verify that i dont made a mistake - also i have tested the hw with a other bios to see if everything included the rom is stoll working - every help is very welcome! Regards, Steffen

3 2

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 August 2007