On Sun, Aug 18, 2019 at 11:57:05PM +0200, Holger Hoffstätte wrote: > On 8/18/19 10:38 PM, Holger Hoffstätte wrote: > > On 8/18/19 8:49 PM, Greg KH wrote: > > > On Sun, Aug 18, 2019 at 02:31:22PM -0400, CKI Project wrote: > > > > > > > > Hello, > > > > > > > > We ran automated tests on a recent commit from this kernel tree: > > > > > > > > Kernel repo: git://git.kernel.org/pub/scm/linux/kernel/git/stable/linux-stable-rc.git > > > > Commit: 61d06c60569f - Linux 5.2.10-rc1 > > > > > > > > The results of these automated tests are provided below. > > > > > > > > Overall result: FAILED (see details below) > > > > Merge: OK > > > > Compile: OK > > > > Tests: FAILED > > > > > > > > All kernel binaries, config files, and logs are available for download here: > > > > > > > > https://artifacts.cki-project.org/pipelines/108998 > > > > > > > > > > > > > > > > One or more kernel tests failed: > > > > > > > > aarch64: > > > > ❌ Boot test > > > > ❌ Boot test > > > > > > > > ppc64le: > > > > ❌ Boot test > > > > ❌ Boot test > > > > > > > > x86_64: > > > > ❌ Boot test > > > > ❌ Boot test > > > > > > > > > > Are these all real? > > > > > > > Hi Greg, > > > > the current 5.2-queue also fails to boot for me when applied to 5.2.9, > > so this is not a false positive. I had a handful of the queued patches in my > > own tree and know which ones work in 5.2.9, but the new ones for -mm look > > like they could cause problems. I'll try a few things.. > > The culprit is "exit-make-setting-exit_state-consistent.patch", which was > successfully added everywhere. This explains why KCI is consistently sad > everywhere, too. :) > > Removing that patch from the queue results in a working kernel (with the > version updated by me): > > $cat /proc/version > Linux version 5.2.10 (root@ragnarok) (gcc version 9.2.0 (Gentoo 9.2.0 p1)) #1 SMP Sun Aug 18 23:50:50 CEST 2019 Thanks for this, looks like kernelci and Linaro's testing are all failing as well, so this isn't isolated. Sasha, the above commit relies on commit b191d6491be6 ("pidfd: fix a poll race when setting exit_state") which is in 5.3-rc, so I'll drop it from everywhere now... thanks, greg k-h
