search

AMDESE / AMDSEV

AMD Secure Encrypted Virtualization
294 stars 85 forks source link

SEV-SNP: RMPUPDATE failed (SEV: too many firmware commands are in progress) #205

Closed csegarragonz closed 7 months ago

csegarragonz commented 7 months ago

Hi!

Our SEV-SNP enabled host started doing unexpected warm resets a couple of days ago. At the beginning, the host would successfully boot after reset, but now it hangs after a RMPUPDATE failure. (Note that both the reset, and of course the reset and hang, are unexpected)

The RMPUPDATE failure reads something like:

SEV-SNP: RMPUPDATE failed ...

and then it hangs in:

ccp .. SEV: too many firmware commands are in progress

any idea of what could be happening?

FWIW, the only thing I can see in the journalctl are the aforementioned:

Feb 21 17:59:09 milan2 kernel: ccp 0000:01:00.5: SEV: too many firmware commands are in-progress, no command buffers available.

I also include a screenshot of the virtual console from iDRAC of the host that fails to boot:

image

csegarragonz commented 7 months ago

Hi,

This morning the machine is unreachable after a power cycle last night with a different error:

image

Any idea of what could be going on?

mdroth commented 7 months ago

This issue should be fixed in the latest linux/QEMU trees that were pushed yesterday, so please give those a try.

csegarragonz commented 7 months ago

It does, many thanks!