Last week I filed a bug report for the same hardware and code path, unloading i915. It had been broken since v3.13-rc1 and it was solved with two commits from master (v3.15-rc8 at the time). I will spare repeating the same hardware details, you can check them @ https://bugzilla.kernel.org/show_bug.cgi?id=77321 Yesterday I did a go at v3.15 and it fails with a NULL dereference (last time was a WARN_ON) on i915_gem_obj_to_ggtt, matching the same timing of previous bug, although on Gem management instead of display outputs. Below is the trace and the entire dmesg is @ http://bpaste.net/raw/361794/ Let me know if I can be of further assistance ;) [ 1907.660322] vga_switcheroo: disabled [ 1907.660384] BUG: unable to handle kernel NULL pointer dereference at 00000000000000c0 [ 1907.660398] IP: [<ffffffffc01d6f73>] i915_gem_obj_to_ggtt+0x0/0x4a [i915] [ 1907.660447] PGD 0 [ 1907.660453] Oops: 0000 [#1] SMP [ 1907.660463] Dumping ftrace buffer: [ 1907.660472] (ftrace buffer empty) [ 1907.660477] Modules linked in: btusb ccm brcmsmac brcmutil mac80211 nouveau i915(-) bcma ttm drm_kms_helper [last unloaded: btusb] [ 1907.660503] CPU: 3 PID: 8540 Comm: rmmod Not tainted 3.15.0-mbp #19 [ 1907.660510] Hardware name: Apple Inc. MacBookPro6,2/Mac-F22586C8, BIOS MBP61.88Z.0057.B0F.1112091028 12/09/11 [ 1907.660520] task: ffff880243962010 ti: ffff88023d30c000 task.ti: ffff88023d30c000 [ 1907.660528] RIP: 0010:[<ffffffffc01d6f73>] [<ffffffffc01d6f73>] i915_gem_obj_to_ggtt+0x0/0x4a [i915] [ 1907.660570] RSP: 0018:ffff88023d30fd20 EFLAGS: 00010246 [ 1907.660576] RAX: 0000000000000000 RBX: 0000000000000000 RCX: ffff88023d30fd28 [ 1907.660583] RDX: 0000000080000000 RSI: ffffffffc01cb788 RDI: 0000000000000000 [ 1907.660590] RBP: ffff88023d30fd48 R08: 0000000000000000 R09: ffff88023d30fd28 [ 1907.660597] R10: ffff88023d30fd00 R11: 0000000000000001 R12: 00000000000006e0 [ 1907.660604] R13: ffff880243802ec0 R14: 00007fffc270af68 R15: 0000000000000000 [ 1907.660612] FS: 00007fa35f4fd700(0000) GS:ffff88024fcc0000(0000) knlGS:0000000000000000 [ 1907.660621] CS: 0010 DS: 0000 ES: 0000 CR0: 000000008005003b [ 1907.660627] CR2: 00000000000000c0 CR3: 00000002439b5000 CR4: 00000000000007e0 [ 1907.660634] Stack: [ 1907.660638] ffffffffc01d72b5 ffff880242170000 00000000000006e0 ffff880243802ec0 [ 1907.660649] 00007fffc270af68 ffff88023d30fd70 ffffffffc01cbe4c ffff880242170000 [ 1907.660660] ffff880243a06800 ffff880243a06868 ffff88023d30fd98 ffffffffc021fb7f [ 1907.660670] Call Trace: [ 1907.660709] [<ffffffffc01d72b5>] ? i915_gem_object_ggtt_unpin+0x13/0x76 [i915] [ 1907.660744] [<ffffffffc01cbe4c>] i915_gem_context_fini+0xc8/0xd7 [i915] [ 1907.660802] [<ffffffffc021fb7f>] i915_driver_unload+0x185/0x278 [i915] [ 1907.660813] [<ffffffffab445120>] drm_dev_unregister+0x23/0x93 [ 1907.660821] [<ffffffffab445254>] drm_put_dev+0x4a/0x55 [ 1907.660846] [<ffffffffc01b94e1>] i915_pci_remove+0x18/0x1a [i915] [ 1907.660857] [<ffffffffab3b9f45>] pci_device_remove+0x40/0x8a [ 1907.660866] [<ffffffffab45ce8b>] __device_release_driver+0x84/0xda [ 1907.660875] [<ffffffffab45d4b3>] driver_detach+0x85/0xab [ 1907.660883] [<ffffffffab45cbcb>] bus_remove_driver+0x64/0x81 [ 1907.660892] [<ffffffffab45d9e5>] driver_unregister+0x41/0x44 [ 1907.660900] [<ffffffffab3b9413>] pci_unregister_driver+0x17/0x64 [ 1907.660909] [<ffffffffab446dc3>] drm_pci_exit+0x40/0x87 [ 1907.660952] [<ffffffffc0221a62>] i915_exit+0x20/0x150 [i915] [ 1907.660963] [<ffffffffab0fece5>] SyS_delete_module+0x13c/0x1c2 [ 1907.660973] [<ffffffffab7cb98f>] tracesys+0xe1/0xe6 [ 1907.660978] Code: c7 83 c8 29 00 00 12 5b 1d c0 48 c7 83 c0 29 00 00 6c 1a 1d c0 c7 83 d0 29 00 00 02 00 00 00 e8 5b 4b f7 ea 5b 41 5c 41 5d 5d c3 <48> 8b 97 c0 00 00 00 48 8d 87 c0 00 00 00 48 39 c2 75 19 55 be [ 1907.661044] RIP [<ffffffffc01d6f73>] i915_gem_obj_to_ggtt+0x0/0x4a [i915] [ 1907.661084] RSP <ffff88023d30fd20> [ 1907.661088] CR2: 00000000000000c0 [ 1907.661117] ---[ end trace 79c4f4e26cd5edbe ]---
Sorry for the slow response. Please try v3.17-rc, drm-intel-nightly from http://cgit.freedesktop.org/drm-intel, and finally https://github.com/ideak/linux/commits/suspend-fix and report back.
Timeout, closing. Please reopen if the problem persists with recent kernels.