QEMU x86_64 Q35 Automated Test Failure

[Raptor Engineering Automated Test Stand]

The QEMU x86_64 Q35 fails verification as of commit 8ccdeaeb207031eea3881511acfaf3e949678f74

The following tests failed:
CBMEM_TIMESTAMP_ACCESS_FAILURE
CBMEM_OBJECT_TABLE_TRUNCATED

Commits since last successful test:
8ccdeae haswell: Link stage_cache_external_region into ramstage, too
634899c resource: Adjust memory resources high earlier
e6a9290 resource: Enhance resource reporting
134b616 resource: Report correct secondary resource window
ade2c5e winent/mb6047: symbolic arguments for acpi_create_madt_lapic_nmis()

<8 commits skipped>

2436bda cpu: get rid of socket source code
99cc1aa Mediawiki editing warning
e9b7e25 util/xcompile/xcompile: Allow to override `HOSTCC` variable
939dc84 crossgcc: Re-download the archive if it is incomplete
6548ae9 cbfstool/Makefile*: Use `LDFLAGS` instead of `LINKFLAGS`

See attached boot log for details

This message was automatically generated from Raptor Engineering's QEMU x86_64 Q35 test stand
Raptor Engineering offers coreboot consulting services!  Please visit https://www.raptorengineeringinc.com for more information

Please contact Timothy Pearson at Raptor Engineering  regarding any issues stemming from this notification

REACTS: Also upload failed logs to board status (was: QEMU x86_64 Q35 Automated Test Failure)

[Paul Menzel]

Dear coreboot folks,

Am Montag, den 04.05.2015, 23:47 -0500 schrieb REACTS:
> The QEMU x86_64 Q35 fails verification as of commit 8ccdeaeb207031eea3881511acfaf3e949678f74
>
> The following tests failed:
> CBMEM_TIMESTAMP_ACCESS_FAILURE
> CBMEM_OBJECT_TABLE_TRUNCATED
>
> Commits since last successful test:

[…]

as this test run still booted to GNU/Linux, files could still be
uploaded to the board status repository. As the name of the board status
repository is not “good board status” repository, I suggest that failing
logs are still uploaded with the tag *[failed]* in the commit summary
and the failing tests (basically the email message) in the commit
message body.

In each board directory, a separate directory `failed` could be added or
`failed` be added to the directory name of the commit.

A separate branch could also be thought of.

The Wiki scripts [1] would probably need to be updated to ignore the
failed entries.

Thanks,

Paul

REACTS: Also upload failed logs to board status

[Timothy Pearson]

On 05/05/2015 12:26 AM, Paul Menzel wrote:
> Dear coreboot folks,
>
> Am Montag, den 04.05.2015, 23:47 -0500 schrieb REACTS:
> > The QEMU x86_64 Q35 fails verification as of commit 8ccdeaeb207031eea3881511acfaf3e949678f74
> >
> > The following tests failed:
> > CBMEM_TIMESTAMP_ACCESS_FAILURE
> > CBMEM_OBJECT_TABLE_TRUNCATED
> >
> > Commits since last successful test:
>
> […]
>
> as this test run still booted to GNU/Linux, files could still be
> uploaded to the board status repository. As the name of the board status
> repository is not “good board status” repository, I suggest that failing
> logs are still uploaded with the tag *[failed]* in the commit summary
> and the failing tests (basically the email message) in the commit
> message body.
>
> In each board directory, a separate directory `failed` could be added or
> `failed` be added to the directory name of the commit.
>
> A separate branch could also be thought of.
>
> The Wiki scripts [1] would probably need to be updated to ignore the
> failed entries.
>
> Thanks,
>
> Paul

Before enabling anything like that a failed flag is an absolute must. 
The addition of such a flag would change the board-status repository 
from its current function (reporting known good boards only) to a much 
more useful repository of actual board status for specific timestamps.

If/when the board-status utility is modified to add a --failed option or 
similar let me know and I will enable reporting of failed tests to the 
board-status repository.

Thanks!