On Mon, Jan 25, 2021 at 09:13:59PM +0000, Chris Wilson wrote: > Quoting Mike Rapoport (2021-01-25 21:04:56) > > On Mon, Jan 25, 2021 at 08:34:34PM +0000, Chris Wilson wrote: > > > Quoting Linus Torvalds (2021-01-25 01:06:40) > > > > Mike Rapoport (3): > > > ... > > > > mm: fix initialization of struct page for holes in memory layout > > > > > > We have half a dozen or so different machines in CI that are silently > > > failing to boot, that we believe is bisected to this patch. > > > > > > 17:56 tsa : ickle: dolphin: I hit the following patch in my > > > bisection, and the hang is also dependent on kconfig > > > 17:56 tsa : first bad commit: > > > [d3921cb8be29ce5668c64e23ffdaeec5f8c69399] mm: fix initialization of > > > struct page for holes in > > > memory layout > > > 17:57 tsa : couldn't reproduce on older CI kconfig, current > > > one does it > > > > > > https://gitlab.freedesktop.org/gfx-ci/i915-infra/-/blob/master/kconfig/debug > > > > > > Here's a boot dmesg from some affected machines from just before the merge > > > with rc5: > > > https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9676/shard-skl1/boot18.txt > > > https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9676/fi-skl-6600u/boot.html > > > https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9676/fi-bsw-cyan/boot.html > > > https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_9676/fi-bdw-samus/boot.html > > > > Is there any way to get early console from these machines? > > 12:16 tsa : none of those have good hook for serial > > Nothing on the console and no serial console option, and panics before > netconsole. > > Maybe some early_printk and boot_delay if you think there's something to > see with those, but I'll have to ask Tomi nicely tomorrow.
I think there could be an early panic at some point of mm initialization. So if it was possible to see early printks somehow that would have been helpful. > -Chris -- Sincerely yours, Mike.
