-
Notifications
You must be signed in to change notification settings - Fork 0
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Talos II : One CPU one ram module 0.4.1 release - SCOM stopped working #81
Comments
Will redo this one, might have not
Correctly. |
This should be fixed in v0.5.0. Here we forgot to remove hardcoded CPU version, sorry about that... And by the way, thanks for testing! |
Just for clarity:
Redid test:
Same result |
We will clarify the regions in the documentations as a part of the #79
So this qualifies to be closed. @krystian-hebel let's backport the fix and confirm, then close. |
Interestingly enough, the fix is already present in 0.4.1: https://github.com/Dasharo/coreboot/blame/raptor-cs_talos-2/rel_v0.4.1/src/soc/ibm/power9/romstage.c#L399 However, the commit hash reported by binaries ("coreboot-4.14-387-g7258fa59c0") doesn't match anything in the tree, so whoever produced those did something strange... @pietrushnic @macpijan @IgorBagnucki CC @tlaurion please try with these: https://cloud.3mdeb.com/index.php/s/MSLKxazwKsCoi68 |
@krystian-hebel : There seems to be a stop of boot loops after the 5th manual With current https://cloud.3mdeb.com/index.php/s/MSLKxazwKsCoi68 tested ROM: root@talos:~# cat /var/log/obmc-console.log (exerpt):
Was able to Which may or not be helpful here.
|
Will take a closer look later, but now it definitely is a different issue than before, and different than #80. The one from previous comment reports recoverable error in cache chiplet, while #80 reported checkstop for core chiplet, although in a core that is connected to the same cache chiplet. For now I'll wait for info from my supervisors as to what to do with bad 0.4.1 binaries released, then we will decide if we want to continue debugging here or open new issue. |
We believe we observe the same issue here: #80 as in this one. The problem is rather not the dual CPU itself - in such a case, it would work just fine on the v0.4.1 version provided by @krystian-hebel The reported problem is most likely related to the memory (and/or CPU? - you've got slightly different - older - revision). Any chance you've got more memory modules do try out, or can use different slots, as suggested here: #80 (comment) Of course, hostboot deals with this setup, so this should be fixable on the firmware level. It is a matter of fiding out the root cause. |
Basically, I think this issue can be closed while HCL is published on dasharo universe, specifying the platform that was tested (CPUs memory and board revision). Otherwise, #80, in my test case, would be a ~duplicate and that release will continue to not work. We cannot change the past (0.4.1) where 0.5 tests will lead to a newer release. |
@macpijan this issue should be closed |
Dasharo version
0.4.1
Dasharo variant
Workstation, 1CPU one ram module. bootblock + coreboot 0.4.1 release
No-flashing instructions per #79
Affected component(s) or functionality
SCOM stopped working
Brief summary
SCOM stops working after step 14.5
How reproducible
At all times booting from non-flashed testing #79
How to reproduce
Laptop:
First SSH session to BMC:
Second SSH session to BMC:
Expected behavior
SCOM not stopping, next steps continuing
Actual behavior
The text was updated successfully, but these errors were encountered: