The condition which triggers this error message in dmesg is normal. The error message should be tuned down, so that the automated tests don't pick it up and flag an error. One way to fix this: http://lists.freedesktop.org/archives/intel-gfx/2015-November/080399.html
Highest priority to get rid of dmesg noise in BAT testing, assign to Imre who has the patch.
Temporary fixed with a private branch patch on topic/core-for-CI: commit 64a93eaebb44c35c50e10144923189c8558d043d Author: Imre Deak <imre.deak@intel.com> Date: Wed Nov 18 12:56:28 2015 +0200 PCI / PM: tune down RPM suspend error message with EBUSY and EAGAIN retval We "only" need to make sure here that the patch goes to the right upstream maintainer.
Tested this with commit c9026da2526abed2b3cd1309d9bad7c009826aab Author: Maarten Lankhorst <maarten.lankhorst@ubuntu.com> Date: Thu Dec 3 16:14:29 2015 +0100 drm-intel-nightly: 2015y-12m-03d-15h-13m-50s UTC integration manifest The messages on the kernel ring buffer are printed if debug log level is enabled -- which is by default on dmesg -- assuming that piglit uses warn this should not appear any more. Can be easily triggered each time by running tests/pm_rpm using gem-execbuf-stress subtest. Think this can be closed for now.
So apparently, this reached v5 on m-l. http://lists.freedesktop.org/archives/intel-gfx/2015-November/081630.html. It is fixed in drm-intel-nightly (w/ v2), but will eventually be replaced by v5.
We have the fix in a topical branch to shut up the warnings in nightly, but we need to follow-up the commit making it to upstream linux via the right subsystem.
The fix is merged now to Linus' tree and included in -nightly, so closing this.
Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.