Messages in this thread | | | Date | Sat, 27 Mar 2010 10:54:36 +0100 | Subject | Re: i915 lockup / extreme delay | From | Karl Vogel <> |
| |
On Mon, Mar 22, 2010 at 4:34 PM, Eric Anholt <eric@anholt.net> wrote: > On Mon, 22 Mar 2010 09:11:06 +0100, Karl Vogel <karl.vogel@gmail.com> wrote: >> On Mon, Mar 22, 2010 at 5:20 AM, Eric Anholt <eric@anholt.net> wrote: >> > On Sat, 20 Mar 2010 14:41:41 +0100, Karl Vogel <karl.vogel@gmail.com> wrote: >> >> The 'effect' is that only the mouse pointer works in the X server. The >> >> cpu usage on the laptop during the sluggishness is minimal. When I >> >> suspend the game with winedbg, the X server slowly becomes responsive again. >> >> >> >> The output from latencytop seems to point to i915 being the culprit: >> > >> > If there's some code doing glFlush()es, it's probably that code at >> > fault. You don't need to do that unless you're doing frontbuffer >> > rendering, and if you're doing frontbuffer rendering you should really >> > be doing backbuffer rendering. I don't see a kernel issue here. >> >> That doesnt explain why the box completely locks up on 2.6.34-rc2 >> though, where only a cold reboot works. > > Missed that part of the message. If there's a regression, bisect > please.
Apparently the crash was caused by a hardware bug in the intel chipset which is 8086:2a40 rev 07. While doing the bisect I got an error:
DRHD: handling fault status reg 2 DMAR:[DMA Write] Request device [00:02.0] fault addr dd69a000 DMAR:[fault reason 05] PTE Write access is not set
After some googling around, I found this bugzilla entry which explains it:
https://bugzilla.redhat.com/show_bug.cgi?id=538163#c58
The issue appears that the graphics chip is corrupting memory:
"Unfortunately, this particular chipset sometimes reads from the GTT, does the translation, then writes the translated address back to the _original_ GTT instead of to the shadow GTT. That's why you're seeing real physical addresses where you should have 'virtual DMA addresses', and you get the faults. "
Adding "intel_iommu=igfx_off" to the kernel command line resolved the issue. The fedora kernel automatically disables this when it detects this particular chipset revision.
As for the freeze/slowdown right after booting, sysprof shows that more than 77% of the time is spent inside: drm_mode_getconnector
[/usr/bin/Xorg] 0.00% 80.29% ioctl 0.00% 78.47% - - kernel - - 0.01% 78.47% system_call_fastpath 0.00% 77.15% sys_ioctl 0.00% 77.15% do_vfs_ioctl 0.01% 77.15% vfs_ioctl 0.00% 77.14% drm_ioctl 0.01% 77.14% drm_mode_getconnector 0.00% 77.02% drm_helper_probe_single_connector_modes 0.00% 77.02% intel_lvds_get_modes 0.00% 62.46% intel_ddc_get_modes 0.00% 62.46% drm_get_edid 0.00% 62.45% drm_ddc_read_edid 0.00% 62.45% drm_do_probe_ddc_edid 0.00% 62.45% i2c_transfer 0.00% 62.45% bit_xfer 0.01% 62.44% sclhi 0.00% 25.76% set_clock 0.08% 12.86% __udelay 0.00% 12.78% get_clock 0.11% 0.11% __const_udelay 0.01% 0.01% set_clock 0.12% 12.54% __const_udelay 0.01% 12.42% __delay 0.00% 0.00% __udelay 0.00% 12.08% acknak 0.00% 7.68% sdahi 0.01% 2.30% try_address 0.00% 1.20% i2c_outb 0.00% 0.59% get_data 0.11% 0.11% i2c_stop 0.00% 0.09% i2c_repstart 0.00% 0.07% i2c_start 0.00% 0.01% __const_udelay 0.01% 0.01% __udelay 0.01% 0.01% drm_add_edid_modes 0.00% 0.00% drm_mode_connector_update_edid_property 0.00% 0.00% intel_hdmi_detect 0.00% 8.01% intel_dp_detect 0.00% 4.91% intel_tv_detect 0.00% 1.23% intel_lvds_detect 0.00% 0.38% intel_crt_detect 0.00% 0.03% drm_get_connector_name 0.00% 0.01% i915_gem_execbuffer 0.00% 0.06% i915_gem_mmap_ioctl 0.00% 0.01% i915_gem_set_domain_ioctl 0.00% 0.01% drm_gem_close_ioctl 0.00% 0.01% drm_mode_object_find 0.00% 0.00% i915_gem_busy_ioctl 0.00% 0.00% lock_kernel 0.00% 0.00% i915_gem_create_ioctl 0.00% 0.00% copy_to_user 0.00% 0.00% copy_user_generic_string 0.00% 0.00%
| |