Gerald Turner <gturner@unzane.com> writes: > Gerald Turner <gturner@unzane.com> writes: > >> Gerald Turner <gturner@unzane.com> writes: >> >>> I downgraded to linux-image-2.6.32-5-686_2.6.32-29_i386.deb¹ and >>> resume From hibernate works flawlessly again. >>> >>> Could the bug be caused by one of these changes in ..-30? >>> >>> * drm/i915: Free hardware status page on unload when physically >>> mapped. >>> * drm/i915/overlay: Ensure that the reg_bo is in the GTT prior to >>> writing. >>> * drm/i915/sdvo: Always add a 30ms delay to make SDVO TV detection >>> reliable. >>> >> >> I rebuilt -30 with these three patches disabled: >> >> debian/patches/series/30: >> #+ bugfix/all/drm-i915-free-hardware-status-page-on-unload-when-physically-mapped.patch >> #+ bugfix/all/drm-i915-overlay-ensure-that-the-reg_bo-is-in-the-gtt-prior-to-writing.patch >> #+ bugfix/all/drm-i915-sdvo-always-add-a-30ms-delay-to-make-sdvo-tv-detection-reliable.patch >> >> Hibernate/resume is working. >> >> Will try building with fewer patches disabled to isolate which patch >> is breaking EeePC resume. > > I tried another build with only one patch disabled: > drm-i915-free-hardware-status-page-on-unload-when-physically-mapped.patch > …resume fails again. > > I suspect the bug is triggered by: > drm-i915-overlay-ensure-that-the-reg_bo-is-in-the-gtt-prior-to-writing.patch > > I'll do another build with only that patch disabled in a few hours. Tested -30 kernel built with only having patch drm-i915-overlay-ensure-that-the-reg_bo-is-in-the-gtt-prior-to-writing.patch dropped and resume from hibernate works. The patch is described as: Just makes sure that writes are not being aliased by the CPU cache and do make it out to main memory. Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk> Signed-off-by: Andi Kleen <ak@linux.intel.com> Reviewed-by: Daniel Vetter <daniel.vetter@ffwll.ch> Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=24977 Signed-off-by: Greg Kroah-Hartman <gregkh@suse.de> I followed the bugzilla link, the bug is about a green window shown instead of video - I have no such problem on this hardware (945GM) without the patch (tested command: mplayer 1259_X_in_squeeze.ogv -vo xv:port=88). BTW, looks like this my bug report may also be a duplicate of #610589 and that Sebastian Andrzej Siewior has reported a similar regression with this patch (however EeePC immediately resets during resume). That makes 5 or 6 accounts of i915 hardware with broken hibernate/resume on -30 kernel. -- Gerald Turner Email: gturner@unzane.com JID: gturner@unzane.com GPG: 0xFA8CD6D5 21D9 B2E8 7FE7 F19E 5F7D 4D0C 3FA0 810F FA8C D6D5
Attachment:
pgpdkS1ItYKlo.pgp
Description: PGP signature