flashrom

flashrom@flashrom.org

5 participants
9104 discussions

[Patch] Board enable for Asus P4P800-E Deluxe.
by Luc Verhaegen 20 Aug '09

20 Aug '09

First time i successfully dug one out of an AMI Bios. int16h ax=0xE0 is indeed the ticket. This one has its int16h vector at the standard location it seems: 0xF000:0xE82E. Not sure how general this is for AMI. It does seem that i will require dumps of the F segment for this to work though. Luc Verhaegen.

4 3

"lspci -vvnnxx" and "superiotool -d" dumps from an Asus m2n68 motherboard needing a board enable
by Putlinuxonit 20 Aug '09

20 Aug '09

Hi All, This board has it's chipset detected but the eeprom flash chip isn't even though it's an MXIC25L8005 chip which is listed in "flashrom -L" Just thought i'll put up the dumps. Do let me know if you need further testing. Udu E. Ogah

1 0

Not sure where to put this?
by Marko Kraljevic 20 Aug '09

20 Aug '09

Hi, I'm new to the list. I've been playing with flashrom some, and it's great. Thanks for the work guys. I've put a ZIF socket on a 3com NIC, for external programming. It seems to be working great. I've taken a picture, maybe it would be useful for the wiki or something. I'm not a terribly good programmer, but I'm pretty good with hardware generally... I'm going to be doing some more hardware work in the future. I think I'll make an ISA card with 32 and 28pin ZIF, and PLCC socket, on the 440BX board I have, addr is off the ISA bus, so I'll just buffer that. Data seems to go to the superIO chip, so I'll have to have a cheater socket under the stock BIOS ROM to steal /CS D0:7, etc. I think I'll buffer all the lines on the card, so that the bus doesn't get fried in case of a fried chip or whatnot. It will be a sort of general purpose flashing board. I think I'll try and implement support for the older 27xxx EPROMs at some point, which shouldn't be too difficult I don't think? anyways, here's the picture if anyone wants it: http://electrontube.org/img/3com_prog.jpg Mark

2 3

Asus A7v600-X works with stable v0.9.0 but not with bleeding edge
by Putlinuxonit 20 Aug '09

20 Aug '09

Hi All, I'd like to report than an Asus A7V600-X with bios 1009 works with flashrom 0.9.0 stable. The stock version found on the latest stable realease of systemrescuecd 1.2.3. There are some caveats however... find below a few dumps for your review. root@sysresccd% ./flashrom *flashrom v0.9.0* No coreboot table found. *Found chipset "VIA VT8237"*, enabling flash write... OK. This chipset supports the following protocols: Non-SPI. Calibrating delay loop... OK. *Found chip "PMC Pm49FL002" (256 KB, LPC,FWH)* at physical address 0xfffc0000. No operations were specified. The strange thing is that when i tried using the bleeding edge r694 i get the following root@sysresccd /root % ./flashrom -w fimrwarea7v600-x.old *flashrom v0.9.0-r694* No coreboot table found. Found chipset "VIA VT8237", enabling flash write... OK. This chipset supports the following protocols: Non-SPI. Calibrating delay loop... OK. Found chip "PMC Pm49FL002" (256 KB, LPC,FWH) at physical address 0xfffc0000. Flash image seems to be a legacy BIOS. Disabling checks. Writing flash chip... Programming page: ERASE FAILED at 0x00000000! Expected=0xff, Read=0x03, failed byte count from 0x00000000-0x00003fff: 0x33a6 ERASE FAILED! ERASE FAILED! FAILED! Your flash chip is in an unknown state. Get help on IRC at irc.freenode.net channel #flashrom or mail flashrom(a)flashrom.org ------------------------------------------------------------ DO NOT REBOOT OR POWEROFF! i then use the stock 0.9.0 version to reprogramme and i get this root@sysresccd /root % flashrom -w fimrwarea7v600-x.old Calibrating delay loop... OK. No coreboot table found. Found chipset "VIA VT8237", enabling flash write... OK. Found chip "PMC Pm49FL002" (256 KB) at physical address 0xfffc0000. Flash image seems to be a legacy BIOS. Disabling checks. Programming page: 0015 at address: 0x0003c000 i* verify if it's written ok* root@sysresccd /root % flashrom -v fimrwarea7v600-x.old Calibrating delay loop... OK. No coreboot table found. Found chipset "VIA VT8237", enabling flash write... OK. Found chip "PMC Pm49FL002" (256 KB) at physical address 0xfffc0000. Flash image seems to be a legacy BIOS. Disabling checks. Verifying flash... VERIFIED. root@sysresccd /root % flashrom -R *flashrom v0.9.0* so basically, When i use the stock v0.9.0 to dump and write back the firmware on my A7v600-x, everything is fine. when i try to use the current svn. r694, it fails. i had to reflash with stable 0.9.0 after the r694 failed. I've successfully rebooted and verified no damage done to motherboard. interestingly, I downloaded the latest version of the bios (1009) off asus website, same version as i'm running and when i unzip it yeilds an *.awd ( i guess award) file. This file is supposed to be the same version as the one i'm running but when i try both stock v0.9.0 and r694 on it. they both fail. Beats me. maybe the awardflasher file format is different from what I get when i do* "flashrom -r firmware-old.bin"* * * I've also compared md5 checksums on both the firmware-old.bin and the one downloaded off the asus website but they both give different results even though they are meant to be the same version (1009). Anyway, hats-off to the devs for all the good work. idlogin -- Udu E. Ogah .~~~~~~~~~~. attachments I've attached flashrom -V dumps for both 0.9.0 and r694 latest Asus bios for A7v600-X

2 4

[RFC] flashrom 0.9.1
by Carl-Daniel Hailfinger 19 Aug '09

19 Aug '09

Hi, there are no more pending flashrom patches for 0.9.1 and I would like to release it today or tomorrow. Draft 0.9.1 announcement is at http://www.coreboot.org/Flashrom/0.9.1 . Please test! Regards, Carl-Daniel -- http://www.hailfinger.org/

1 0

[PATCH] Fix romentry layout for programmers
by Carl-Daniel Hailfinger 19 Aug '09

19 Aug '09

Flashrom has the ability to use layout files with romentries, but this feature was not adapted to the programmer infrastructure and had undefined behaviour for flasher!=internal. The romentry handling had an off-by-one error which caused all copies to end up one byte short. Fix these issues. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Index: flashrom-romentries_programmer/flash.h =================================================================== --- flashrom-romentries_programmer/flash.h (Revision 692) +++ flashrom-romentries_programmer/flash.h (Arbeitskopie) @@ -422,7 +422,7 @@ int show_id(uint8_t *bios, int size, int force); int read_romlayout(char *name); int find_romentry(char *name); -int handle_romentries(uint8_t *buffer, uint8_t *content); +int handle_romentries(uint8_t *buffer, struct flashchip *flash); /* cbtable.c */ int coreboot_init(void); Index: flashrom-romentries_programmer/flashrom.c =================================================================== --- flashrom-romentries_programmer/flashrom.c (Revision 692) +++ flashrom-romentries_programmer/flashrom.c (Arbeitskopie) @@ -930,8 +930,7 @@ // This should be moved into each flash part's code to do it // cleanly. This does the job. - /* FIXME: Adapt to the external flasher infrastructure. */ - handle_romentries(buf, (uint8_t *) flash->virtual_memory); + handle_romentries(buf, flash); // //////////////////////////////////////////////////////////// Index: flashrom-romentries_programmer/layout.c =================================================================== --- flashrom-romentries_programmer/layout.c (Revision 692) +++ flashrom-romentries_programmer/layout.c (Arbeitskopie) @@ -196,11 +196,11 @@ return -1; } -int handle_romentries(uint8_t *buffer, uint8_t *content) +int handle_romentries(uint8_t *buffer, struct flashchip *flash) { int i; - // This function does not safe flash write cycles. + // This function does not save flash write cycles. // // Also it does not cope with overlapping rom layout // sections. @@ -220,10 +220,9 @@ if (rom_entries[i].included) continue; - /* FIXME: Adapt to the external flasher infrastructure. */ - memcpy(buffer + rom_entries[i].start, - content + rom_entries[i].start, - rom_entries[i].end - rom_entries[i].start); + flash->read(flash, buffer + rom_entries[i].start, + rom_entries[i].start, + rom_entries[i].end - rom_entries[i].start + 1); } return 0; -- http://www.hailfinger.org/

2 2

[commit] r694 - trunk
by svn＠flashrom.org 19 Aug '09

19 Aug '09

Author: hailfinger Date: 2009-08-19 17:19:18 +0200 (Wed, 19 Aug 2009) New Revision: 694 Modified: trunk/flash.h trunk/flashrom.c trunk/layout.c Log: Flashrom has the ability to use layout files with romentries, but this feature was not adapted to the programmer infrastructure and had undefined behaviour for flasher!=internal. The romentry handling had an off-by-one error which caused all copies to end up one byte short. Fix these issues. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Acked-by: Stefan Reinauer <stepan(a)coresystems.de> Modified: trunk/flash.h =================================================================== --- trunk/flash.h 2009-08-19 15:03:28 UTC (rev 693) +++ trunk/flash.h 2009-08-19 15:19:18 UTC (rev 694) @@ -422,7 +422,7 @@ int show_id(uint8_t *bios, int size, int force); int read_romlayout(char *name); int find_romentry(char *name); -int handle_romentries(uint8_t *buffer, uint8_t *content); +int handle_romentries(uint8_t *buffer, struct flashchip *flash); /* cbtable.c */ int coreboot_init(void); Modified: trunk/flashrom.c =================================================================== --- trunk/flashrom.c 2009-08-19 15:03:28 UTC (rev 693) +++ trunk/flashrom.c 2009-08-19 15:19:18 UTC (rev 694) @@ -930,8 +930,7 @@ // This should be moved into each flash part's code to do it // cleanly. This does the job. - /* FIXME: Adapt to the external flasher infrastructure. */ - handle_romentries(buf, (uint8_t *) flash->virtual_memory); + handle_romentries(buf, flash); // //////////////////////////////////////////////////////////// Modified: trunk/layout.c =================================================================== --- trunk/layout.c 2009-08-19 15:03:28 UTC (rev 693) +++ trunk/layout.c 2009-08-19 15:19:18 UTC (rev 694) @@ -196,11 +196,11 @@ return -1; } -int handle_romentries(uint8_t *buffer, uint8_t *content) +int handle_romentries(uint8_t *buffer, struct flashchip *flash) { int i; - // This function does not safe flash write cycles. + // This function does not save flash write cycles. // // Also it does not cope with overlapping rom layout // sections. @@ -220,10 +220,9 @@ if (rom_entries[i].included) continue; - /* FIXME: Adapt to the external flasher infrastructure. */ - memcpy(buffer + rom_entries[i].start, - content + rom_entries[i].start, - rom_entries[i].end - rom_entries[i].start); + flash->read(flash, buffer + rom_entries[i].start, + rom_entries[i].start, + rom_entries[i].end - rom_entries[i].start + 1); } return 0;

1 0

[PATCH] Use : instead of = for programmer parameters
by Carl-Daniel Hailfinger 19 Aug '09

19 Aug '09

Current programmer parameter syntax is -p programmer=parameter Unfortunately, many parameters are of the form variable=val, so we get commandlines like this: flashrom -p it87spi=port=0x820 and this looks horrible. Using : instead of = would make such parameters look better: flashrom -p it87spi:port=0x820 As a side benefit, this patch mentions the programmer name in the error message if it is unknown. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Index: flashrom-programmer_param_colon/flashrom.8 =================================================================== --- flashrom-programmer_param_colon/flashrom.8 (Revision 692) +++ flashrom-programmer_param_colon/flashrom.8 (Arbeitskopie) @@ -150,7 +150,7 @@ .sp The dummy programmer has an optional parameter specifying the bus types it should support. For that you have to use the -.B "flashrom -p dummy=type" +.B "flashrom -p dummy:type" syntax where .B type can be any comma-separated combination of @@ -158,11 +158,11 @@ in any order. .sp Example: -.B "flashrom -p dummy=lpc,fwh" +.B "flashrom -p dummy:lpc,fwh" .sp If you have multiple supported PCI cards which can program flash chips (NICs, SATA/IDE controllers, etc.) in your system, you must use the -.B "flashrom -p xxxx=bb:dd.f" +.B "flashrom -p xxxx:bb:dd.f" syntax to explicitly select one of them, where .B xxxx is the name of the programmer @@ -174,7 +174,7 @@ is the PCI function number of the desired NIC. .sp Example: -.B "flashrom -p nic3com=05:04.0" +.B "flashrom -p nic3com:05:04.0" .sp Currently the following programmers support this mechanism: .BR nic3com , @@ -183,14 +183,14 @@ The it87spi programmer has an optional parameter which will set the I/O base port of the IT87* SPI controller interface to the port specified in the parameter. For that you have to use the -.B "flashrom -p it87spi=port=portnum" +.B "flashrom -p it87spi:port=portnum" syntax where .B portnum is an I/O port number which must be a multiple of 8. .sp The ft2232spi programmer has an optional parameter specifying the controller type and interface/port it should support. For that you have to use the -.B "flashrom -p ft2232spi=model,port=interface" +.B "flashrom -p ft2232spi:model,port=interface" syntax where .B model can be any of @@ -208,9 +208,9 @@ The serprog programmer has an optional parameter specifying either a serial device/baud combination or an IP/port combination for communication with the programmer. For serial, you have to use the -.B "flashrom -p serprog=/dev/device:baud" +.B "flashrom -p serprog:/dev/device:baud" syntax and for IP, you have to use -.B "flashrom -p serprog=ip:port" +.B "flashrom -p serprog:ip:port" instead. More information about serprog is available in serprog-protocol.txt in the source distribution. .sp Index: flashrom-programmer_param_colon/flashrom.c =================================================================== --- flashrom-programmer_param_colon/flashrom.c (Revision 692) +++ flashrom-programmer_param_colon/flashrom.c (Arbeitskopie) @@ -708,7 +708,7 @@ namelen = strlen(name); if (strncmp(optarg, name, namelen) == 0) { switch (optarg[namelen]) { - case '=': + case ':': programmer_param = strdup(optarg + namelen + 1); break; case '\0': @@ -725,7 +725,7 @@ } } if (programmer == PROGRAMMER_INVALID) { - printf("Error: Unknown programmer.\n"); + printf("Error: Unknown programmer %s.\n", optarg); exit(1); } break; -- http://www.hailfinger.org/

2 2

[commit] r693 - trunk
by svn＠flashrom.org 19 Aug '09

19 Aug '09

Author: hailfinger Date: 2009-08-19 17:03:28 +0200 (Wed, 19 Aug 2009) New Revision: 693 Modified: trunk/flashrom.8 trunk/flashrom.c Log: Current programmer parameter syntax is -p programmer=parameter Unfortunately, many parameters are of the form variable=val, so we get commandlines like this: flashrom -p it87spi=port=0x820 and this looks horrible. Using : instead of = would make such parameters look better: flashrom -p it87spi:port=0x820 As a side benefit, this patch mentions the programmer name in the error message if it is unknown. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Acked-by: Stefan Reinauer <stepan(a)coresystems.de> Modified: trunk/flashrom.8 =================================================================== --- trunk/flashrom.8 2009-08-19 13:55:34 UTC (rev 692) +++ trunk/flashrom.8 2009-08-19 15:03:28 UTC (rev 693) @@ -150,7 +150,7 @@ .sp The dummy programmer has an optional parameter specifying the bus types it should support. For that you have to use the -.B "flashrom -p dummy=type" +.B "flashrom -p dummy:type" syntax where .B type can be any comma-separated combination of @@ -158,11 +158,11 @@ in any order. .sp Example: -.B "flashrom -p dummy=lpc,fwh" +.B "flashrom -p dummy:lpc,fwh" .sp If you have multiple supported PCI cards which can program flash chips (NICs, SATA/IDE controllers, etc.) in your system, you must use the -.B "flashrom -p xxxx=bb:dd.f" +.B "flashrom -p xxxx:bb:dd.f" syntax to explicitly select one of them, where .B xxxx is the name of the programmer @@ -174,7 +174,7 @@ is the PCI function number of the desired NIC. .sp Example: -.B "flashrom -p nic3com=05:04.0" +.B "flashrom -p nic3com:05:04.0" .sp Currently the following programmers support this mechanism: .BR nic3com , @@ -183,14 +183,14 @@ The it87spi programmer has an optional parameter which will set the I/O base port of the IT87* SPI controller interface to the port specified in the parameter. For that you have to use the -.B "flashrom -p it87spi=port=portnum" +.B "flashrom -p it87spi:port=portnum" syntax where .B portnum is an I/O port number which must be a multiple of 8. .sp The ft2232spi programmer has an optional parameter specifying the controller type and interface/port it should support. For that you have to use the -.B "flashrom -p ft2232spi=model,port=interface" +.B "flashrom -p ft2232spi:model,port=interface" syntax where .B model can be any of @@ -208,9 +208,9 @@ The serprog programmer has an optional parameter specifying either a serial device/baud combination or an IP/port combination for communication with the programmer. For serial, you have to use the -.B "flashrom -p serprog=/dev/device:baud" +.B "flashrom -p serprog:/dev/device:baud" syntax and for IP, you have to use -.B "flashrom -p serprog=ip:port" +.B "flashrom -p serprog:ip:port" instead. More information about serprog is available in serprog-protocol.txt in the source distribution. .sp Modified: trunk/flashrom.c =================================================================== --- trunk/flashrom.c 2009-08-19 13:55:34 UTC (rev 692) +++ trunk/flashrom.c 2009-08-19 15:03:28 UTC (rev 693) @@ -708,7 +708,7 @@ namelen = strlen(name); if (strncmp(optarg, name, namelen) == 0) { switch (optarg[namelen]) { - case '=': + case ':': programmer_param = strdup(optarg + namelen + 1); break; case '\0': @@ -725,7 +725,7 @@ } } if (programmer == PROGRAMMER_INVALID) { - printf("Error: Unknown programmer.\n"); + printf("Error: Unknown programmer %s.\n", optarg); exit(1); } break;

1 0

[PATCH] Improve error messages, protect the user
by Carl-Daniel Hailfinger 19 Aug '09

19 Aug '09

Disallow erase/write for known bad chips so people won't try without a clear understanding. If we want them to test, we can tell them to patch/edit flashchips.c. If write/erase failed, warn the user to get help and not shutdown/reboot the computer. Warn that the result of a forced read is often garbage. Too many users believed that a forced read meant that everything was fine. Signed-off-by: Carl-Daniel Hailfinger <c-d.hailfinger.devel.2006(a)gmx.net> Index: flashrom-errormessages/flashrom.c =================================================================== --- flashrom-errormessages/flashrom.c (Revision 676) +++ flashrom-errormessages/flashrom.c (Arbeitskopie) @@ -221,6 +221,7 @@ { size_t size = flash->total_size * 1024; /* Flash registers live 4 MByte below the flash. */ + /* FIXME: This is incorrect for nonstandard flashbase. */ flash->virtual_registers = (chipaddr)programmer_map_flash_region("flash chip registers", (0xFFFFFFFF - 0x400000 - size + 1), size); } @@ -475,6 +476,15 @@ return 0; } +void emergency_help_message() +{ + fprintf(stderr, "Your flash chip is in an unknown state.\n" + "Get help on IRC at irc.freenode.net channel #flashrom or\n" + "mail flashrom(a)flashrom.org\n" + "------------------------------------------------------------\n" + "DO NOT REBOOT OR POWEROFF!\n"); +} + void usage(const char *name) { printf("usage: %s [-VfLzhR] [-E|-r file|-w file|-v file] [-c chipname]\n" @@ -743,6 +753,7 @@ printf("Run flashrom -L to view the hardware supported in this flashrom version.\n"); exit(1); } + printf("Please note that forced reads most likely contain garbage.\n"); return read_flash(flashes[0], filename); } // FIXME: flash writes stay enabled! @@ -808,14 +819,31 @@ buf = (uint8_t *) calloc(size, sizeof(char)); if (erase_it) { - if (erase_flash(flash)) + if (flash->tested & TEST_BAD_ERASE) { + fprintf(stderr, "Erase is not working on this chip. " + "Aborting.\n"); return 1; + } + if (erase_flash(flash)) { + emergency_help_message(); + return 1; + } } else if (read_it) { if (read_flash(flash, filename)) return 1; } else { struct stat image_stat; + if (flash->tested & TEST_BAD_ERASE) { + fprintf(stderr, "Erase is not working on this chip " + "and erase is needed for write. Aborting.\n"); + return 1; + } + if (flash->tested & TEST_BAD_WRITE) { + fprintf(stderr, "Write is not working on this chip. " + "Aborting.\n"); + return 1; + } if ((image = fopen(filename, "r")) == NULL) { perror(filename); exit(1); @@ -854,14 +882,18 @@ ret = flash->write(flash, buf); if (ret) { fprintf(stderr, "FAILED!\n"); + emergency_help_message(); return 1; } else { printf("COMPLETE.\n"); } } - if (verify_it) + if (verify_it) { ret = verify_flash(flash, buf); + if (ret) + emergency_help_message(); + } programmer_shutdown(); -- http://www.hailfinger.org/

2 6

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

flashrom