On Debian Jessie,freeze noticed with such logs in dmesg: [54320.265767] CPU: 0 PID: 829 Comm: Xorg Tainted: P W O 3.14-1-amd64 #1 Debian 3.14.12-1 [54320.265768] Hardware name: Apple Inc. MacBookAir6,1/Mac-35C1E88140C3E6CF, BIOS MBA61.88Z.0099.B09.1402071141 02/07/2014 [54320.265769] 0000000000000009 ffffffff814ba733 ffff88026519dc98 ffffffff8105f032 [54320.265770] ffff880263518028 ffff88026519dce8 0000000000070084 0000000000000292 [54320.265772] 0000000000000000 ffffffff8105f097 ffffffffa0948a74 ffff880200000018 [54320.265773] Call Trace: [54320.265775] [<ffffffff814ba733>] ? dump_stack+0x41/0x51 [54320.265777] [<ffffffff8105f032>] ? warn_slowpath_common+0x72/0x90 [54320.265778] [<ffffffff8105f097>] ? warn_slowpath_fmt+0x47/0x50 [54320.265784] [<ffffffffa091b0ab>] ? hsw_write32+0x4b/0x150 [i915] [54320.265789] [<ffffffffa091c428>] ? gen6_read32+0x48/0x120 [i915] [54320.265794] [<ffffffffa08d8a4d>] ? intel_crtc_update_cursor+0x29d/0x3f0 [i915] [54320.265799] [<ffffffffa08dbb10>] ? intel_crtc_cursor_move+0x50/0x60 [i915] [54320.265804] [<ffffffffa031605f>] ? drm_mode_cursor_common+0x8f/0x160 [drm] [54320.265809] [<ffffffffa0319c9f>] ? drm_mode_cursor_ioctl+0x3f/0x50 [drm] [54320.265812] [<ffffffffa030b89d>] ? drm_ioctl+0x42d/0x530 [drm] [54320.265816] [<ffffffffa02fec30>] ? evdev_read+0xd0/0x2f0 [evdev] [54320.265818] [<ffffffff81197e9f>] ? do_vfs_ioctl+0x2cf/0x4b0 [54320.265820] [<ffffffff81185e21>] ? vfs_read+0x91/0x160 [54320.265822] [<ffffffff811980f9>] ? SyS_ioctl+0x79/0x90 [54320.265823] [<ffffffff814c80b9>] ? system_call_fastpath+0x16/0x1b [54320.265824] ---[ end trace 77f62fde486db570 ]--- [54320.265827] [drm:hsw_unclaimed_reg_clear] *ERROR* Unknown unclaimed register before writing to 70084 [54320.265828] [drm:hsw_unclaimed_reg_check] *ERROR* Unclaimed write to 70084
Please either attach the whole dmesg or include the full WARN. Now tell us more about the freeze since that is not directly related to that warning.
Created attachment 103316 [details] All dmesg output More precisions: the cursor crash appears even more frequently (as you can see in the dmesg) such that the system becomes unusuable and need to be restarted: the bug appears only after a while maybe after wake up from suspend to ram. I can send you more info if you want to.
Created attachment 103335 [details] Another dmesg output when system is freezing (or cursor always blocked, screen not refreshed)
I can reproduce this bug and am working on a test case and a Kernel fix.
Created attachment 103369 [details] [review] Possible fix Hi Can you confirm this patch fixes your bug? It fixes the bug I can reproduce... I also implemented a test case to reproduce it: http://lists.freedesktop.org/archives/intel-gfx/2014-July/049481.html Thanks, Paulo
Hello all, Many thanks for your patch. I have to test it now. Since it is the first time I report a kernel bug, I do not really know the good practices to test the patch you send me. Which documentation / tutorial / procedure do you advice I follow ? Many thanks in advance for your comprehension. Best regards for you and your works. Alexandre
commit d6dd6843ff4a57c662dbc378b9f99a9c034b0956 Author: Paulo Zanoni <paulo.r.zanoni@intel.com> Date: Fri Aug 15 15:59:32 2014 -0300 drm/i915: fix plane/cursor handling when runtime suspended pushed to drm-intel-fixes branch of http://cgit.freedesktop.org/drm-intel
Closing resolved+fixed after two years.
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.