When I try to make the digitallogic/adl855pc target I receive an error with the following message.
/usr/bin/ld: section .reset [fffdfff0 -> fffdffff] overlaps section .rom [fffd632e -> fffe1d7f] /usr/bin/ld: section .id [fffdffce -> fffdffef] overlaps section .rom [fffd632e -> fffe1d7f] /usr/bin/ld: linuxbios: section .id lma 0xfffdffce overlaps previous sections /usr/bin/ld: linuxbios: section .text lma 0xfffdfff0 overlaps previous sections /usr/bin/ld: linuxbios: section .data lma 0xfffdfff0 overlaps previous sections /usr/bin/ld: linuxbios: section .reset lma 0xfffdfff0 overlaps previous sectionscollect2: ld returned 1 exit status make[1]: *** [linuxbios] Error 1 make[1]: Leaving directory `/root/temp/LinuxBIOSv2/targets/digitallogic/adl855pc/adl855pc/normal' make: *** [normal/linuxbios.rom] Error 1
When I use abuild it says that it builds fine. Any idea why I can't get this to link using make?
Also, is the linuxbios.rom file in the abuild directory good? Or do you suspect that it has the same linking error I get when I use make.
Thanks, Jon
You have run out of room. What is your flash size and payload?
ron
I did not modify the digitallogic/adl855pc target, I just wanted to see if I could biuld it before modifying it. This is straight from svn.
Jon
On 11/20/06, ron minnich rminnich@gmail.com wrote:
You have run out of room. What is your flash size and payload?
ron
-- linuxbios mailing list linuxbios@linuxbios.org http://www.openbios.org/mailman/listinfo/linuxbios
On Mon, Nov 20, 2006 at 03:20:20PM -0500, Jon Dufresne wrote:
I did not modify the digitallogic/adl855pc target, I just wanted to see if I could biuld it before modifying it. This is straight from svn.
Confirmed. I get the same results.
Uwe.
Any suggestion on how I could attempt to fix this would be appreciated.
On 11/20/06, Uwe Hermann uwe@hermann-uwe.de wrote:
On Mon, Nov 20, 2006 at 03:20:20PM -0500, Jon Dufresne wrote:
I did not modify the digitallogic/adl855pc target, I just wanted to see if I could biuld it before modifying it. This is straight from svn.
Confirmed. I get the same results.
Uwe.
http://www.hermann-uwe.de | http://www.holsham-traders.de http://www.crazy-hacks.org | http://www.unmaintained-free-software.org
-----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.5 (GNU/Linux)
iD8DBQFFYhaeXdVoV3jWIbQRAimLAKCjgI7TIehjHxehS37JlGPb5l5ezwCcDouG Oa4j9QankD0dIqox39P6xqA= =tAyS -----END PGP SIGNATURE-----
-- linuxbios mailing list linuxbios@linuxbios.org http://www.openbios.org/mailman/listinfo/linuxbios
On 11/20/06, Jon Dufresne jon.dufresne@gmail.com wrote:
Any suggestion on how I could attempt to fix this would be appreciated.
[rminnich@q adl855pc]$ pwd /home/rminnich/src/LinuxBIOSv2/targets/digitallogic/adl855pc [rminnich@q adl855pc]$ more Config.lb # Sample config file for adl855pc # This will make a target directory of ./adl855pc
target adl855pc mainboard digitallogic/adl855pc
option DEFAULT_CONSOLE_LOGLEVEL=10 option MAXIMUM_CONSOLE_LOGLEVEL=10
set to 3
ron
Is there a way to get a "map" of sorts , or any kind of description, of how buildrom allocates the memory for different parts of the rom. I have a feeling that parts starting to late or too early and wanted to know if there was a convenient tool to analyze this.
Thanks, Jon
On 11/20/06, Jon Dufresne jon.dufresne@gmail.com wrote:
I did not modify the digitallogic/adl855pc target, I just wanted to see if I could biuld it before modifying it. This is straight from svn.
blast. Something has changed. This target never worked, due to lack of chipset info from the vendor. I apologize for this, but you're going to have to dig a little.
thanks
ron
So are you saying this is an bug in the chipset software or the mainboard software?
Thanks for the help!
On 11/20/06, ron minnich rminnich@gmail.com wrote:
On 11/20/06, Jon Dufresne jon.dufresne@gmail.com wrote:
I did not modify the digitallogic/adl855pc target, I just wanted to see if I could biuld it before modifying it. This is straight from svn.
blast. Something has changed. This target never worked, due to lack of chipset info from the vendor. I apologize for this, but you're going to have to dig a little.
thanks
ron
* Jon Dufresne jon.dufresne@gmail.com [061120 20:39]:
/usr/bin/ld: section .reset [fffdfff0 -> fffdffff] overlaps section .rom [fffd632e -> fffe1d7f]
This looks like a linker bug in the Debian linker. Why is it putting .reset at 0xfffdfff0 instead of 0xfffffff0 ?
When I use abuild it says that it builds fine. Any idea why I can't get this to link using make?
Please compare the Config.lb generated by abuild with the one you are using.. It's different configurations.
Also, is the linuxbios.rom file in the abuild directory good? Or do you suspect that it has the same linking error I get when I use make.
Not unless you build it with a payload
svn co svn://linuxbios.org/repos/trunk/LinuxBIOSv2 cd LinuxBIOSv2/util/abuild svn co svn://linuxbios.org/testsystem/LinuxBIOS-payloads ./abuild -t digitallogic/adl855pc -p `pwd`/LinuxBIOS-payloads
Stefan Reinauer wrote:
- Jon Dufresne jon.dufresne@gmail.com [061120 20:39]:
/usr/bin/ld: section .reset [fffdfff0 -> fffdffff] overlaps section .rom [fffd632e -> fffe1d7f]
This looks like a linker bug in the Debian linker. Why is it putting .reset at 0xfffdfff0 instead of 0xfffffff0 ?
When I use abuild it says that it builds fine. Any idea why I can't get this to link using make?
Jon, Sorry it took me a while to get back to you on this. I've been Traveling.
This is the classic out-of-room in the linuxbios IMAGE error. Not to be confused with out-of-room in your ROM.
You can fix this by increasing your ROM_IMAGE_SIZE.
Note that this option is sort of misnamed. Very non-obvious as to what it does. This controls the size of the area we allocate out for the Linuxbios+payload.
Try:
option ROM_IMAGE_SIZE=0x16000
Rather than 0x10000 as in Config.lb
0x16000 is what abuild is using. It can probably be smaller. In fact a good exercise for the reader is to find out how small it can be and still contain a FILO payload.
Stefan:
Why is abuild preferring to use its own custom Config.lb rather than the Config.lb that is in the target directory?
Success! Thanks so much for everyone that gave me help, I really appreciate it.
So if ROM_IMAGE_SIZE = LB + payload what does ROM_SIZE mean? I assume it means the entire size of the ROM, which in my case is 512KB. How is this different than ROM_IMAGE_SIZE. What does the ROM hold in addition to LB + payload.
Richard: No need to apologize, you're under no obligation to help. I consider everything I get a bonus to the already written software.
-Jon
On 11/21/06, Richard Smith smithbone@gmail.com wrote:
Stefan Reinauer wrote:
- Jon Dufresne jon.dufresne@gmail.com [061120 20:39]:
/usr/bin/ld: section .reset [fffdfff0 -> fffdffff] overlaps section .rom [fffd632e -> fffe1d7f]
This looks like a linker bug in the Debian linker. Why is it putting .reset at 0xfffdfff0 instead of 0xfffffff0 ?
When I use abuild it says that it builds fine. Any idea why I can't get this to link using make?
Jon, Sorry it took me a while to get back to you on this. I've been Traveling.
This is the classic out-of-room in the linuxbios IMAGE error. Not to be confused with out-of-room in your ROM.
You can fix this by increasing your ROM_IMAGE_SIZE.
Note that this option is sort of misnamed. Very non-obvious as to what it does. This controls the size of the area we allocate out for the Linuxbios+payload.
Try:
option ROM_IMAGE_SIZE=0x16000
Rather than 0x10000 as in Config.lb
0x16000 is what abuild is using. It can probably be smaller. In fact a good exercise for the reader is to find out how small it can be and still contain a FILO payload.
Stefan:
Why is abuild preferring to use its own custom Config.lb rather than the Config.lb that is in the target directory?
-- Richard A. Smith
On Tue, Nov 21, 2006 at 11:47:29AM -0500, Jon Dufresne wrote:
So if ROM_IMAGE_SIZE = LB + payload what does ROM_SIZE mean? I assume it means the entire size of the ROM, which in my case is 512KB.
Yes, correct.
How is this different than ROM_IMAGE_SIZE. What does the ROM hold in addition to LB + payload.
It can hold a VGA BIOS for example. Or a VSA binary image for the Geodes. Maybe in the future an option ROM image for your SCSI card.
I'm guessing that ROM_IMAGE_SIZE came from the linuxbios_rom label/section/filename.
It's a bit complicated.. While people (rightly) only care about the size of their flash chip, the various binary images that can be included have to be arranged to fit together somehow..
//Peter