I have a 512 kByte FlashROM but use 256 kByte images so far. I
concatenate two of them, so even if address pin A18 is dangling, the CPU
reads the image correctly.
Now, if I understand correctly, the util/flashrom utility writes (-w)
using the Flash programming algorithms, which do not touch A18. And if I
understand correctly, it used
mmap() to verify the image, which should use A18.
Am I correct that by following this procedure, I can find out whether
A18 is connected on my motherboard to the FlashROM socket? :
dd if=/dev/random of=/tmp/random.bin bs=1024 count=512
flashrom -w /tmp/random.bin
flashrom -v tmp/random.bin
VERIFIED => A18 is connected
*do not do this at home, it destroys your BIOS*
I found this script to be helpful to me in getting a VGA BIOS binary for
leon@nehemiah:/tmp$ cat dissect.sh
for OFFSET in 76607 107793 112921 113603 123297 136064 148043 186053; do
dd if=i0100113.bin of=/tmp/rest.lha bs=1 skip=$OFFSET
lha tv rest.lha
Also, to get a rough first idea of the offsets:
leon@nehemiah:/tmp$ grep -b -a -o lh5 i0100113.bin
On Wed, 2006-02-22 at 15:12 -0300, Jardel Silveira wrote:
> As a matter of fact, this target doesn´t seem to turn on my board.
> The power led is off.
> I´ve an FS2 JTAG interface and may be I can do some necessary tests.
I just committed the raminit.c for GX2. Please take a look at it. I
presume you have access to the developer web site and docs there?
I think I have implemented the DRAM init sequence in their GeodeROM
Porting Guide and LX/GX processor data book correctly. But I may be
wrong. I also thing probably I have to implement the PLL reset in
order for DRAM to work.
BTW, hook a serial console. I think the power led is controlled by
software. It is off all the way in LinuxBIOS (yes, there is something
on the serial console.)
Li-Ta Lo <ollie(a)lanl.gov>
Los Alamos National Lab
I noticed in the v1 tree that you support the LEX cv860a motherboard,
yet there have been no updates to that tree in a long time (2 years?).
I was dissapointed to see that it has not beed added to the v2 tree.
Does this mean that there are no plans to include it?
As I am new to this, I am uncertain of the ramifications in version
difference from v1 to v2. Any insight would be greatly appreciated.
* Jardel Silveira <jardel(a)lesc.ufc.br> [060223 22:54]:
> Dear Mr. Li-Ta Lo,
> I´ve access only to the svn respository. I think don´t have access to the
> developer web site and docs there.
All information is freely available at www.linuxbios.org. There is no
such thing as a developer web site. You only need to log in to the
website to actually _change_ the content.
coresystems GmbH · Brahmsstr. 16 · D-79104 Freiburg i. Br.
Tel.: +49 761 7668825 · Fax: +49 761 7664613
Email: info(a)coresystems.de · http://www.coresystems.de/
My board is close to the MII, but not exact. I have a different south
So thinking that I can get the north bridge up, I am hoping that I can
get my board to at least give me something out of the console, from
there I can start debugging.
I dont think the dd if=/dev/mem.... to grab the video bios will work for
me, different address range's.
Currently I have linuxbios 2177, same as you, but I dont have a
computable flash chip (SST39SF040) currently not supported, so I am
using an external burner.
The biggest bug I am currently stuck on is the linuxbios.rom keeps
coming out at 192k and not 256k... No clue what I have wrong. If i can
get 256k then just `cat linuxbios.rom >> linuxbios.rom.512` do that
twice and I should have a 512 rom. Can I snag a copy of your
Config.lb? My current payload is /etc/hosts, I will switch to filo when
I can get console running. For now I would just like to see ANY signs
of life out of the console. Console is tested and fully running, I
currently dont even use a screen, just serial console.
Leon Woestenberg wrote:
> Hello Adam,
> found your question on the LinuxBIOS list (see below).
> I just retried the EPIA-M LinuxBIOS code again and it worked for my
> EPIA-M 10000 board. Actually, this is the first time I got it working.
> SVN HEAD tree (rev 2177) with filo 4.2.0 with Linux 2.6.11 (or so).
> Note that I did *not* get VGA to work under the early LinuxBIOS boot
> and FILO boot, it came up when Linux booted. I controlled FILO through
> a null-modem serial connection.
> I have attached my notes below.
> Please if you proceed with your EPIA board, keep me in the loop with
> your experiences. It's hard to find experiences from stuff that worked.
> note: Seaching for the VGA issue, it seems I should take the VGA BIOS
> from some specific version of VIA's BIOS...
> Leon Woestenberg.
> [LinuxBIOS] Status of the epia-m
> *Adam Talbot* talbotx at comcast.net
> /Tue Feb 21 23:13:44 CET 2006/
> * Previous message: [LinuxBIOS] This should be an easy error...
> * Next message: [LinuxBIOS] EPIA M COM2 problem + possible
> * *Messages sorted by:* [ date ]
> [ thread ]
> [ subject ]
> [ author ]
>Is the current epia-m code running?
>rev 2177 of linuxbiosv2
>Debugging output from the EPIA:
>115200,8n1,first serial port
>Extract VGA BIOS:
>dd if=/dev/mem of=/video.bios.bin \
>bs=1 count=65536 skip=790528
>Linux sits on:
>/dev/hdc2 on / type ext3 (rw,errors=remount-ro)
>In the 'util/flash_and_burn' directory is the source for the 'flash_rom'
> utility, which is great for re-programming the flash chips on the
> EPIA-M / MII.
>Correction: should be in 'util/flashrom' directory since 2101 or so.
>apt-get install pciutils
>apt-get install pciutils-dev
>leon@nehemiah:~/sandbox/linuxbios/util/flashrom$ sudo -s
>Calibrating delay loop... ok
>No LinuxBIOS table found.
>Enabling flash write on VT8235...tried to set 0x45 to 0x55 on VT8235 failed (WARNING ONLY)
>SST39SF020A found at physical address: 0xfffc0000
>Flash part is SST39SF020A
>OK, only ENABLING flash write, but NOT FLASHING.
>SST39SF020A is a 256k words of 8 bit = 2 Mbit
>ST29SF040A is a 512k words of 8 bit = 4 Mbit
>Note that the EPIA will address the second half of the 4 Mbit Flash! To be sure, copy the bin to
>both halves of the flash.
> The main configuration file for the epia-m is in
> If you need to make any changes to the configuration, for example you wish to
> locate filo.elf in a place other than '/filo.elf', or during the more advanced
> steps of this HOWTO, then these changes are made to this file.
> You need to re-run the './buildtarget via/epia-m' after any such change.
> Change the configuration file to match the full path of the filo.elf file.
> payload /home/leon/sandbox/linuxbios/filo.elf
> Regenerate the build:
> ./buildtarget via/epia-m
>cp filo.elf ../linuxbios
>serial.c:54.34: serial.c:66.29: console.c:6.21: console.c:16.26: console.c:51.36: console.c:72.71: bist.h:11.34: auto.c:133.28: 0x81a44e0 noop Internal compiler warning: awakening noop?
Has anyone had any luck with LinuxBios on the i845GV chipset? I found
a few references here and there on the mailing list archive, but nothing
Mike Anderson 703.585.9384
CTO/Chief Scientist mike(a)ThePTRGroup.com
www.ThePTRGroup.com Embedded, Real-Time Solutions
"Software development is like making a baby.
You can't make a baby in one month by impregnating nine women.
Some things just take time."
Jardel Silveira wrote:
> I have an geode gx2 amd rdk thinclient (GEODE GX + CS5535). Is it
> possible run rumba target in this system ?
> I´ve maked rumba target and tested it, but CRT didn´t work.
> Thanks for yours help.
amazing. Did you get any serial output or anything? This target is very
new, Ollie Lo has been working like crazy to get it going just this week.
let me know how far it got ...
Im trying to make linuxbios work on a kt600 mb by trying to understand how
the bios initializes the board during bootup. Any tips?
www.smsglobal.net SMS Global Ltd Short Message Service For Seafarers