Can someone tell me what the starting sequence is with LinuxBIOS?
Reset jumps to crt0.s. crt0.s calls auto.E. auto.E is built using romcc, so source-level debugging is not possible. The statement locations can be found using the Lxxxx labels in linuxbios.map. Then hardwaremain is called. Hardwaremain is the first C function called.
Crt0 and auto run out of FLASH. Hardwaremain is the first function called after LinuxBIOS is copied to RAM.
Is this correct? Maybe this will help get the American Arium to work.
Steve
On Tue, 26 Oct 2004 Stephen.Kimball@bench.com wrote:
Can someone tell me what the starting sequence is with LinuxBIOS?
Reset jumps to crt0.s. crt0.s calls auto.E. auto.E is built using romcc, so source-level debugging is not possible.
ah well :-)
The statement locations can be found using the Lxxxx labels in linuxbios.map. Then hardwaremain is called. Hardwaremain is the first C function called.
yep.
Crt0 and auto run out of FLASH. Hardwaremain is the first function called after LinuxBIOS is copied to RAM.
sounds good so far.
ron
Hi, I have a board with intel E7501 and Xeon CPU. We want to boot the linux kernel from the IDE drive(we don't need the compressed linux kernel part). What should I change in the config file to make it do that?
Thanks, Gin
* Gin ginlin@nexcom.com.tw [041027 04:40]:
Hi, I have a board with intel E7501 and Xeon CPU. We want to boot the linux kernel from the IDE drive(we don't need the compressed linux kernel part). What should I change in the config file to make it do that?
use filo as a payload.
Stefan
I found this last night.
http://www.linuxjournal.com/article.php?sid=7857
Something is wrong with his setup as 9 seconds to get to the kernel is way too slow. I'm getting that on my setup in less than a second. But I don't know any thing about the EPIA so I can't help him much.
On Wed, 27 Oct 2004, Richard Smith wrote:
I found this last night.
http://www.linuxjournal.com/article.php?sid=7857
Something is wrong with his setup as 9 seconds to get to the kernel is way too slow. I'm getting that on my setup in less than a second. But I don't know any thing about the EPIA so I can't help him much.
I send them a note offering help, no word back.
I don't know why it's so slow.
ron
From: "Ronald G. Minnich" rminnich@lanl.gov To: Richard Smith rsmith@bitworks.com Date: Wed, 27 Oct 2004 13:42:40 -0600 (MDT) Subject: Re: Can some EPIA-M person fix this?
On Wed, 27 Oct 2004, Richard Smith wrote:
I found this last night.
http://www.linuxjournal.com/article.php?sid=7857
Something is wrong with his setup as 9 seconds to get to the kernel is way too slow. I'm getting that on my setup in less than a second. But I don't know any thing about the EPIA so I can't help him much.
I send them a note offering help, no word back.
I don't know why it's so slow.
Don't worry about them. They have enough money to hire somebody who can make it work, but they barely contribute it back :( I also was not able to have it <7sec, but I probably fix that before the NY as I need it badly. It will not be compatible with linuxbios though( epibios may be ? ) Dmitry/
Dmitry Borisov wrote:
Don't worry about them. They have enough money to hire somebody who can make it work, but they barely contribute it back :( I also was not able to have it <7sec, but I probably fix that before the NY as I need it badly. It will not be compatible with linuxbios though( epibios may be ? )
Well the problem is that its bad publicity. Of course any publicity is acutally good for LB but it would be nice to have those numbers be really small and the article talking about how LB was the best solution and as a bonus its open source rather than just an OK solution thats open source.
From: Richard Smith To: Dmitry Borisov Date: Wed, 27 Oct 2004 16:10:32 -0500 Subject: Re: Can some EPIA-M person fix this?
Well the problem is that its bad publicity. Of course any publicity is acutally good for LB but it would be nice to have those numbers be really small and the article talking about how LB was the best solution and as a bonus its open source rather than just an OK solution thats open source.
Well, The problem is in companies that don't want their solution to get better. VIA does not give any details about CLE chipset at all. This way you cannot expect LB work faster/better than legacy software. So when even such innovative company as carbot claims that they are unable to use VIA solutions along the way, it may force somebody to think. Ask Ron about some other vendors who's more responsible than VIA and how fast LB works there... I'm also thinking to switch to some other motherboard with better docs support. Dmitry/
On Wed, Oct 27, 2004 at 01:42:40PM -0600, Ronald G. Minnich wrote:
I send them a note offering help, no word back.
I don't know why it's so slow.
We looked at it earlier and found several very large sleeps and decompression were the primary culprits. With the sleeps disabled/shorted and compression turned off, it was much faster, though still nothing like as fast as the original epia.
linuxbios@mikebell.org wrote:
On Wed, Oct 27, 2004 at 01:42:40PM -0600, Ronald G. Minnich wrote:
I send them a note offering help, no word back.
I don't know why it's so slow.
We looked at it earlier and found several very large sleeps and decompression were the primary culprits. With the sleeps disabled/shorted and compression turned off, it was much faster, though still nothing like as fast as the original epia.
So what's his new serial log look like now? Is it still 300mS to enable the RAM?
Ronald G. Minnich wrote:
I don't know why it's so slow.
I had about the same timing some months ago, but with freebios2 and the patch from Nick Barker I get about 1 second to go to FILO. This is with the loglevel set to 0, it was *much* longer with loglevel set to 9.
I think they did set loglevel too high, judging by some of what they said.
lotsa prints will slow you down a LOT
ron
On Wed, 2004-10-27 at 14:02 -0500, Richard Smith wrote:
I found this last night.
http://www.linuxjournal.com/article.php?sid=7857
Something is wrong with his setup as 9 seconds to get to the kernel is way too slow. I'm getting that on my setup in less than a second. But I don't know any thing about the EPIA so I can't help him much.
I am to make one of those for myself and would be glad to have something that'd work for sure (using an MII 12000 here) currently, booting my gentoo is about 2 minutes...
On Thu, 28 Oct 2004, Amaury Jacquot wrote:
I am to make one of those for myself and would be glad to have something that'd work for sure (using an MII 12000 here) currently, booting my gentoo is about 2 minutes...
I'm about to try the same wsith an MII 10000.... :D
-marc
Marc Nicholas wrote:
On Thu, 28 Oct 2004, Amaury Jacquot wrote:
I am to make one of those for myself and would be glad to have something that'd work for sure (using an MII 12000 here) currently, booting my gentoo is about 2 minutes...
I'm about to try the same wsith an MII 10000.... :D
-marc
ok, I'll wait ;D
On Wed, 27 Oct 2004, Gin wrote:
I have a board with intel E7501 and Xeon CPU. We want to boot the linux kernel from the IDE drive(we don't need the compressed linux kernel part). What should I change in the config file to make it do that?
use FILO or ETherboot, or (best) enable FILO in etherboot, and you can boot kernel from ext2 partitions.
ron
On Wed, 2004-10-27 at 10:48, Ronald G. Minnich wrote:
On Wed, 27 Oct 2004, Gin wrote:
I have a board with intel E7501 and Xeon CPU. We want to boot the linux kernel from the IDE drive(we don't need the compressed linux kernel part). What should I change in the config file to make it do that?
use FILO or ETherboot, or (best) enable FILO in etherboot, and you can boot kernel from ext2 partitions.
ron
To use FILO or Etherboot you change the payload. So you need to change the freebios2/targets/<vendor>/<board>/Config.lb file to point to the FILO or Etherboot payload and make linuxbios.rom.
Steve
On Wed, 27 Oct 2004, Steve Kimball wrote:
To use FILO or Etherboot you change the payload. So you need to change the freebios2/targets/<vendor>/<board>/Config.lb file to point to the FILO or Etherboot payload and make linuxbios.rom.
don't forget you can make your own!
I have
targets/arima/hdama/Config.filo.lb
and I do this:
./buildtarget arima/hdama/Config.filo.lb
File is this:
# Sample config file for the Arima HDAMA
target hdama.filo mainboard arima/hdama
romimage "normal" option USE_FALLBACK_IMAGE=0 option ROM_IMAGE_SIZE=0x10000 option LINUXBIOS_EXTRA_VERSION=".0Normal" payload /opt/filo/normal/filo.elf end
romimage "fallback" option USE_FALLBACK_IMAGE=1 option ROM_IMAGE_SIZE=0x10000 option LINUXBIOS_EXTRA_VERSION=".0Fallback" payload /opt/filo/fallback/filo.elf end
buildrom ./linuxbios.rom ROM_SIZE "normal" "fallback"
-
Adrien Moulin -
Amaury Jacquot -
Dmitry Borisov -
Gin -
linuxbios＠mikebell.org -
Marc Nicholas -
Richard Smith -
Ronald G. Minnich -
Stefan Reinauer -
Stephen.Kimball＠bench.com -
Steve Kimball