Open tlaurion opened 1 year ago
Also just to clarify, there are two sources of "logs" with different timestamps, coming either from the kernel boot time 0 or skiboot boot time 0, which here tells us source of logs are mixed and confusing:
[ 251.491634682,5] XSCOM: Busy even after 1000 retries, resetting XSCOM now. Total retries = 3000
[ 251.491641400,3] XSCOM: write-busy error gcid=0x8 pcb_addr=0xa3006 stat=0x1
[ 251.491644569,6] IPMI: dropping non severe PEL event
[ 251.491645319,3] XSCOM: Write failed, ret = -2
[ 251.492036026,3] I2C: Failed to write the MODE_REG
[ 251.492073228,3] I2C: Failed to program the MODE_REG
[ OK ] Finished Remove Stale Onli…ext4 Metadata Check Snapshots.
[ OK ] Started Virtual Machine an…ontainer Registration Service.
[ OK ] Started User Login Management.
[ OK ] Started Avahi mDNS/DNS-SD Stack.
[ OK ] Started Authorization Manager.
Starting Modem Manager...
[ OK ] Started Network Manager.
Starting Network Manager Wait Online...
[ OK ] Started WPA supplicant.
[ OK ] Reached target Network.
Starting CUPS Scheduler...
Starting Virtualization daemon...
Starting OpenBSD Secure Shell server...
Starting Permit User Sessions...
[ OK ] Finished Permit User Sessions.
Starting Light Display Manager...
Starting Hold until boot process finishes up...
[ 10.880587] tg3 0004:01:00.0 enP4p1s0f0: Link is up at 100 Mbps, full duplex
[ 10.880620] tg3 0004:01:00.0 enP4p1s0f0: Flow control is off for TX and off for RX
[ 10.880643] tg3 0004:01:00.0 enP4p1s0f0: EEE is disabled
[ 10.896478] IPv6: ADDRCONF(NETDEV_CHANGE): enP4p1s0f0: link becomes ready
the 251
timestamps AFAIK comes from skiboot, where the next timestamp is 0-10
timestamps matches the timeline and come from booted kernel. Ideally, the source of the logs would be clearer (skiboot) or non-existing at all because no error from skiboot.
Maybe we could put somethign like [ 251.492073228,3] (skiboot) I2C: Failed to program the MODE_REG
to eahc lien printed by skiboot
Ideally there should be no lines printed by skiboot at all at this point. They can be easily differentiated by looking at log level, e.g. ,3
after timestamp.
Ideally there should be no lines printed by skiboot at all at this point. They can be easily differentiated by looking at log level, e.g.
,3
after timestamp.
@krystian-hebel
What are 5 and 6 then? I didn't realize up until now there was 3, 5 and 6 following timestamp
These are different log levels defined in skiboot.h. The lower the number, the more important that line is. Minimal saved level can be configured separately for screen and in-memory console through device tree that coreboot passes to skiboot.
Dasharo version
0.7.0
Dasharo variant
Common (talos-2 target under Heads)
Affected component(s) or functionality
skiboot?
Brief summary
Reproducibility notes:
How reproducible
Always
How to reproduce Download and flash according to above notes. FullConsoleLogFromHeadsToDebianLogin.txt Heads-provisioning-after-flash.txt pre-heads-console-logs.txt
Steps to reproduce the behavior:
And
Expected behavior
Absence of worrisome errors on daily driver. Actual behavior
Errors under booted OS
Screenshots
Additional context
Solutions you've tried