Bug 57661
| Summary: | System freezes several times a day / "PTE Read access is not set" in the logs | ||
|---|---|---|---|
| Product: | Drivers | Reporter: | Dimitri Chausson (dimitri.chausson) |
| Component: | Video(DRI - Intel) | Assignee: | Ville Syrjala (ville.syrjala) |
| Status: | RESOLVED INSUFFICIENT_DATA | ||
| Severity: | normal | CC: | alan, daniel, intel-gfx-bugs |
| Priority: | P3 | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Kernel Version: | 3.9.0 | Subsystem: | |
| Regression: | No | Bisected commit-id: | |
| Attachments: | Output of the dmesg command | ||
Shortly before the last freeze: there is the following message in /var/log/kern.log: --------------------------------------------- May 6 21:07:32 keats kernel: [ 6510.874948] dmar: DMAR:[DMA Write] Request device [04:00.0] fault addr ffffe000 May 6 21:07:32 keats kernel: [ 6510.874948] DMAR:[fault reason 05] PTE Write access is not set May 6 21:07:33 keats kernel: [ 6511.855350] [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for forcewake to ack request. --------------------------------------------- This last message is new (at least I did not notice it before) The clock displayed in Xfce froze at 21:07:49 so it is pretty close to the time stamp of the last message. I can always reproduce this bug while zooming in and out several times on the web page for google maps (maps.google.fr). The last message before freezing is: [drm:__gen6_gt_force_wake_mt_get] *ERROR* Timed out waiting for forcewake to ack request. I tried the command line parameter: i915.i915_enable_rc6=0 but without success. Is it still useful for ivy bridge CPU ? Correction about the previous comment: rc6 state is in fact disabled via the command line parameter. The system is, as expected, much slower when zooming in and out. But the freeze still occurs with the "dmar" messages shown above. Please test with intel_iommu=igfx_off added to the kernel cmdline. Probably broken dmar tables in your bios, so please also check for a bios upgrade ... Timeout. Please reopen if the problem persists with recent kernels. |
Created attachment 100881 [details] Output of the dmesg command I cannot describe steps to reproduce the bug, however it happens quite often. Desktop X session (XFCE), most of the time when browsing the internet. I am using Debian Wheezy, and I had the problem with the kernel provided by them. So I tried to compile a more recent version of Linux. So far I tested: * 3.4.42 * 3.9.0 With the latest version, I get this kind of error messages in /var/log/kern.log : May 6 20:21:32 keats kernel: [ 3750.024929] dmar: DRHD: handling fault status reg 3 May 6 20:21:32 keats kernel: [ 3750.024934] dmar: DMAR:[DMA Read] Request device [04:00.0] fault addr ffff7000 May 6 20:21:32 keats kernel: [ 3750.024934] DMAR:[fault reason 06] PTE Read access is not set May 6 20:21:32 keats kernel: [ 3750.850018] dmar: DRHD: handling fault status reg 2 May 6 20:21:32 keats kernel: [ 3750.850023] dmar: DMAR:[DMA Read] Request device [04:00.0] fault addr fffe7000 May 6 20:21:32 keats kernel: [ 3750.850023] DMAR:[fault reason 06] PTE Read access is not set I am not sure that it is related to the system "hangs"... When it happens, I cannot switch to a text console, the numlock key doesn't activate/deactivate the led, the mouse pointer cannot be moved. I have no other solution than switch off the system (button). I can certainly provide you with more details if you need some. Regards, Dimitri PS: I include the output of dmesg as an attachment.