Dual Opteron 248s w/ 8GB RAM on Tyan K8W (S2885)
I am having a problem with getting 8GB of RAM to work with the 2.6.5 kernel that Red Hat uses in their new Fedora Core 2 test3 release. Yes, not vanilla, but hopefully close enough (otherwise I can try it with a custom kernel). The systems (I'm testing on two mostly identical systems) that are experiencing the problem have these specs:
Motherboard: Tyan K8W (S2885)
Processors: 2 x Opteron 248
Memory: 8 x 1GB
BIOS version: 1.02 (2-3-2004 - latest available)
SCSI card: Adaptec 29320
About 80% of the time I get this error when booting:
...
RAMDISK: Compressed image found at block 0
crc error
VFS: Cannot open root device "<NULL>" or unknown-block(3,2) Please append a correct "root=" boot option
Kernel panic: VFS: Unable to mount root fs on unknown-block(3,2)
The other 20% of the time it boots up fine, but that's not enough consistency for these computers final task (beowulf compute nodes). Also I haven't really stressed the kernel on these boots.
The systems seem to work fine if I pull out half the memory so that they only has 4GB of RAM. I've also tested a similar setup using a Rioworks Arima HDAMA board with 8GB of RAM and it worked. So I'm led to believe that this is some kind of driver issue or possibly a bios problem...
Any help/ideas are of course appreciated.
Thanks,
Bryan
Signature · 3 lines
Aspen Systems, Inc. | http://www.aspsys.com/ Production Engineer | Phone: (303)431-4606 [email protected] | Fax: (303)431-7196
A message that is not in the archive (sent off-list or not yet received): [email protected]
Quoted 2 lines · Denis Vlasenko
Latest memtest86 rumored to support testing more than 4Gb or RAM. You may try it and see whether it detects all the RAM.
During the 20% chance of it actually booting up, I've been able to capture /proc/meminfo. It reports MemTotal as 7642992 kB. I've been told that Tyan boards allocate almost 0.5GB for some reason for their own use... You can check out the logs I made from the bugzilla report I sent to RedHat:
https://bugzilla.redhat.com/bugzilla/show_bug.cgi?id=121868
You'll notice that when it has 4GB in the system, the MemTotal is 3530868 kB... The Rioworks board recognized close to the full amount with both 4GB and 8GB installed.
I've also tested Mandrake 10.0rc1 and SuSE 9.0, and they have the same issues that RedHat has...
Bryan
Signature · 3 lines
Aspen Systems, Inc. | http://www.aspsys.com/ Production Engineer | Phone: (303)431-4606 [email protected] | Fax: (303)431-7196
Quoted 3 lines · Bryan Stillwell
During the 20% chance of it actually booting up, I've been able to capture
/proc/meminfo. It reports MemTotal as 7642992 kB. I've been told that Tyan boards allocate almost 0.5GB for some reason for their
interesting. where'd you hear that? mine has 2gigs and is using just under 24megs. i wonder what it's doing with all that.
Signature · 2 lines
Tom Vier <[email protected]> DSA Key ID 0x15741ECE
Quoted 7 lines · Tom Vier
Quoted 3 lines · Bryan Stillwell
During the 20% chance of it actually booting up, I've been able to capture
/proc/meminfo. It reports MemTotal as 7642992 kB. I've been told that Tyan boards allocate almost 0.5GB for some reason for theirinteresting. where'd you hear that? mine has 2gigs and is using just under 24megs. i wonder what it's doing with all that.
I don't have any theories on what they use it for, but one of the other production engineers at my new job says it has been that way with multiple previous boards from Tyan we've used.
Bryan
Signature · 3 lines
Aspen Systems, Inc. | http://www.aspsys.com/ Production Engineer | Phone: (303)431-4606 [email protected] | Fax: (303)431-7196
Quoted 8 lines · Bryan Stillwell
Quoted 7 lines · Tom Vier
Quoted 3 lines · Bryan Stillwell
During the 20% chance of it actually booting up, I've been able to capture
/proc/meminfo. It reports MemTotal as 7642992 kB. I've been told that Tyan boards allocate almost 0.5GB for some reason for theirinteresting. where'd you hear that? mine has 2gigs and is using just under 24megs. i wonder what it's doing with all that.
It can't be. The only CPU(s) that would know anything about it and thus be able to use that RAM are running the operating system.
It is quite obviously a BAD BIOS that doesn't follow the rules about reporting RAM so some is wasted --- or you have 4GB on the main-board so some gets wasted to make room for the address-space necessary for your PCI and AGP cards.
Quoted 7 lines
I don't have any theories on what they use it for, but one of the other production engineers at my new job says it has been that way with multiple previous boards from Tyan we've used.
Bryan
Cheers,
Dick Johnson
Penguin : Linux version 2.4.26 on an i686 machine (5557.45 BogoMips).
Note 96.31% of all statistics are fiction.
Quoted 2 lines · Richard B. Johnson
It can't be. The only CPU(s) that would know anything about it and thus be able to use that RAM are running the operating system.
unless it's reserving it for cross-boot data. i remember there being some error (ecc, machine checks?) logging option in the bios setup. *shrug* i don't know much about x86(-64) bios. openboot and srm do save state, since you can drop back to them without rebooting.
Signature · 2 lines
Tom Vier <[email protected]> DSA Key ID 0x15741ECE