intel issues
https://gitlab.freedesktop.org/drm/intel/-/issues
2023-12-11T09:28:16Z
https://gitlab.freedesktop.org/drm/intel/-/issues/9753
Up to 11% perf drop in (all) onscreen GPU bound 3D benchmarks
2023-12-11T09:28:16Z
Eero Tamminen
Up to 11% perf drop in (all) onscreen GPU bound 3D benchmarks
Between following drm-tip commits:
* 2023-11-01 10:53:03 UTC: ac3eb5045c: drm-tip: 2023y-11m-01d-10h-52m-15s UTC integration manifest
* 2023-11-02 15:16:16 UTC: 76cfd6fff7: drm-tip: 2023y-11m-02d-15h-14m-32s UTC integration manifest
Per...
Between following drm-tip commits:
* 2023-11-01 10:53:03 UTC: ac3eb5045c: drm-tip: 2023y-11m-01d-10h-52m-15s UTC integration manifest
* 2023-11-02 15:16:16 UTC: 76cfd6fff7: drm-tip: 2023y-11m-02d-15h-14m-32s UTC integration manifest
Performance in GPU bound _onscreen_ 3D benchmarks dropped noticeably, and no benchmark improved.
BXT J4205 (HD Graphics 505) perf drops:
* 11% _SynMark2_ CSCloth
* 10% _SynMark2_ TexMem128 & Batch3
* 9% _SynMark2_ GSCloth, Batch2, Batch1, Batch0, GeomTriStrip, VSDiffuse8, GeomTriList, VSDiffuse1 and VSTangent
* 8% _GfxBench5_ Tess & ALU2, _SynMark2_ PSBum8, PSBump2, PSPhong, FillTexMulti and Deferred
* 7% _SynMark2_ PSPom, TexFilterTri and TexFilterAniso, _GpuTest0.7_ FurMark & PixMark Piano
* 6% _GpuTest0.7_ PixMark Volplosion, GiMark, JuliaFP64 and TessMark64, _GpuMemBW_ TexelRate, Write, Blend, FLOPS and PixelRate
* 5% _SyMark2_ TerrrainPanInst & ShMapVsm
* 4% _GfxBench5_ Manhattan 3.1, CarChase and T-Rex, _SynMark2_ ZBuffer, TerrainPanTess, Batch4. GeomPoint, CSDof, GpuTest0.7 JuliaFP32
* 3-4% _Unigine_ Heaven & Valley
* etc
Average power usage during the run of all benchmarks rose also several percents.
GLK results regress in similar way.
Setup / other notes:
* All tests run in fullscreen on FullHD display
* Regression is same with Mesa Git versions from few days ago, and few months ago
* Those GEN9 BXT (2MB L2) and GLK (4MB L2) run Ubuntu 22.04.3 LTS with distro Unity/Compiz/X as desktop
* GEN12 TGL (8MB LLC) using same Ubuntu 22.04.3 LTS setup with Unity/Compiz/X did _not_ show regressions
* Another BXT, running ClearLinux with Weston, did _not_ show regression either
* (no data from any other GPUs / machines)
Reason for differences between those BXT setups could be that:
* Ubuntu uses "powersave" scaling governor, ClearLinux "performance" one
* => Because perf drop happens only in onscreen cases, not in any of the offscreen ones, I don't think this to be a factor
* Weston/Wayland has proper modifier support, but AFAIK that is disabled in upstream & Ubuntu X server
* => Render buffer compression won't be used for onscreen X window buffers
=> regression could be specific either to older GENs, or iGPUs that do not have LLC, **and** handling of non-compressed screen buffers?
https://gitlab.freedesktop.org/drm/intel/-/issues/7036
igt@Few tests - fail - Failed assertion: !mismatch || igt_skip_crc_compare
2023-10-25T06:13:18Z
Prudhvi Kambhampati
igt@Few tests - fail - Failed assertion: !mismatch || igt_skip_crc_compare
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_6682/shard-apl1/igt@i915_pipe_stress@stress-xrgb8888-untiled.html
````
Starting subtest: stress-xrgb8888-untiled
(i915_pipe_stress:17153) igt_debugfs-CRITICAL: Test assertion failure function...
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_6682/shard-apl1/igt@i915_pipe_stress@stress-xrgb8888-untiled.html
````
Starting subtest: stress-xrgb8888-untiled
(i915_pipe_stress:17153) igt_debugfs-CRITICAL: Test assertion failure function igt_assert_crc_equal, file ../../../usr/src/igt-gpu-tools/lib/igt_debugfs.c:491:
(i915_pipe_stress:17153) igt_debugfs-CRITICAL: Failed assertion: !mismatch || igt_skip_crc_compare
Subtest stress-xrgb8888-untiled failed.
**** DEBUG ****
(i915_pipe_stress:17153) igt_kms-DEBUG: plane C.2: Setting property "SRC_H" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: plane C.2: Setting property "CRTC_X" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: plane C.2: Setting property "CRTC_Y" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: plane C.2: Setting property "CRTC_W" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: plane C.2: Setting property "CRTC_H" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: plane C.2: Setting property "FB_ID" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: plane C.2: Setting property "CRTC_ID" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: plane C.2: Setting property "rotation" to 0x1/1
(i915_pipe_stress:17153) igt_kms-DEBUG: display: DP-1: preparing atomic, pipe: A
(i915_pipe_stress:17153) igt_kms-DEBUG: DP-1: Setting property "CRTC_ID" to 0x3b/59
(i915_pipe_stress:17153) igt_kms-DEBUG: DP-1: Setting property "Broadcast RGB" to 0x1/1
(i915_pipe_stress:17153) igt_kms-DEBUG: DP-1: Setting property "Content Protection" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: DP-1: Setting property "HDR_OUTPUT_METADATA" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: display: DP-2: preparing atomic, pipe: None
(i915_pipe_stress:17153) igt_kms-DEBUG: DP-2: Setting property "CRTC_ID" to 0x0/0
(i915_pipe_stress:17153) igt_kms-DEBUG: DP-2: Setting property "Broadcast RGB" to 0x1/1
````
https://gitlab.freedesktop.org/drm/intel/-/issues/5465
igt@kms_setmode@basic - fail - Failed assertion: fabs(mean - expected) < max(...
2024-03-28T07:17:18Z
LAKSHMINARAYANA VUDUM
igt@kms_setmode@basic - fail - Failed assertion: fabs(mean - expected) < max(line_time(kmode), 1.718 * stddev)
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_6400/shard-snb7/igt@kms_setmode@basic@pipe-a-vga-1.html
```
Starting subtest: basic
Starting dynamic subtest: pipe-A-VGA-1
(kms_setmode:1498) CRITICAL: Test assertion failure function check_ti...
https://intel-gfx-ci.01.org/tree/drm-tip/IGT_6400/shard-snb7/igt@kms_setmode@basic@pipe-a-vga-1.html
```
Starting subtest: basic
Starting dynamic subtest: pipe-A-VGA-1
(kms_setmode:1498) CRITICAL: Test assertion failure function check_timings, file ../tests/kms_setmode.c:535:
(kms_setmode:1498) CRITICAL: Failed assertion: fabs(mean - expected) < max(line_time(kmode), 1.718 * stddev)
(kms_setmode:1498) CRITICAL: vblank interval differs from modeline! expected 16665.6us, measured 16717us +- 3.531us, difference 51.4us (14.6 sigma, 2.5 scanlines)
Dynamic subtest pipe-A-VGA-1 failed.
```
https://gitlab.freedesktop.org/drm/intel/-/issues/4817
igt@i915_suspend@basic-s2idle-without-i915|igt@i915_suspend@basic-s3-without-...
2024-01-09T11:25:00Z
LAKSHMINARAYANA VUDUM
igt@i915_suspend@basic-s2idle-without-i915|igt@i915_suspend@basic-s3-without-i915 - incomplete - PM: suspend entry (s2idle|deep)
https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_21849/shard-skl10/igt@i915_suspend@sysfs-reader.html
```
<6> [186.481914] Console: switching to colour dummy device 80x25
<6> [186.482708] [IGT] i915_suspend: executing
<6> [186.562211] ...
https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_21849/shard-skl10/igt@i915_suspend@sysfs-reader.html
```
<6> [186.481914] Console: switching to colour dummy device 80x25
<6> [186.482708] [IGT] i915_suspend: executing
<6> [186.562211] [IGT] i915_suspend: starting subtest sysfs-reader
<7> [187.048168] i915 0000:00:02.0: [drm:intel_backlight_device_get_brightness [i915]] get backlight PWM = 187
<7> [187.148242] i915 0000:00:02.0: [drm:intel_backlight_device_get_brightness [i915]] get backlight PWM = 187
<7> [187.595700] i915 0000:00:02.0: [drm:intel_backlight_device_get_brightness [i915]] get backlight PWM = 187
<7> [187.683227] i915 0000:00:02.0: [drm:intel_backlight_device_get_brightness [i915]] get backlight PWM = 187
<6> [188.171866] PM: suspend entry (deep)
<7> [188.191744] i915 0000:00:02.0: [drm:intel_backlight_device_get_brightness [i915]] get backlight PWM = 187
<6> [188.200114] Filesystems sync: 0.028 seconds
```
https://gitlab.freedesktop.org/drm/intel/-/issues/4767
igt@kms_fbcon_fbt@(fbc|psr)(-suspend)? - fail - Failed assertion: feature->wa...
2024-02-01T11:17:53Z
SAI NANDAN
igt@kms_fbcon_fbt@(fbc|psr)(-suspend)? - fail - Failed assertion: feature->wait_until_update(drm)
https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_964/fi-kbl-soraka/igt@kms_fbcon_fbt@fbc-suspend.html
````
Starting subtest: fbc-suspend
(kms_fbcon_fbt:1157) CRITICAL: Test assertion failure function subtest, file ../tests/i915/kms_fbcon_...
https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_964/fi-kbl-soraka/igt@kms_fbcon_fbt@fbc-suspend.html
````
Starting subtest: fbc-suspend
(kms_fbcon_fbt:1157) CRITICAL: Test assertion failure function subtest, file ../tests/i915/kms_fbcon_fbt.c:367:
(kms_fbcon_fbt:1157) CRITICAL: Failed assertion: feature->wait_until_update(drm)
Subtest fbc-suspend failed.
https://gitlab.freedesktop.org/drm/intel/-/issues/4082
System hangs during parallel media transcode operations after enabling VT-d
2022-11-29T16:45:48Z
Eero Tamminen
System hangs during parallel media transcode operations after enabling VT-d
After enabling "VT-d" from BIOS (which i915 enables IOMMU for GFX) for BXT J4205, that machine started to suffer from system hangs.
They happen when running multiple GPU transcode operations in **parallel** (I'm testing these with 5 par...
After enabling "VT-d" from BIOS (which i915 enables IOMMU for GFX) for BXT J4205, that machine started to suffer from system hangs.
They happen when running multiple GPU transcode operations in **parallel** (I'm testing these with 5 parallel ones).
Transcode could be either heavy HEVC one:
`ffmpeg -hwaccel vaapi -hwaccel_output_format vaapi -i Netflix_FoodMarket_4096x2160_10bit_420_100mbs_600.h265 -c:v hevc_vaapi -b:v 20M -an -vframes 300 -y 0099_4K20.h265`
Or somewhat more lightweight AVC one:
`ffmpeg -hwaccel vaapi -hwaccel_output_format vaapi -i 1920x1080i_29.97_20mb_mpeg2_high.mpv -c:v h264_vaapi -b:v 6000K -compression_level 7 -an -vframes 1200 -y 0024_HD17i7_1.0.h264`
And repeating that parallel test-case for few times.
There's nothing in "dmesg -w" output over remote ssh connection when this happens, the connection just dies.
I've enabled VT-d also on SkullCanyon (SKL GT4e) and TGL-H (GT1), but those do not suffer from the same problem. I think GLK would though, as it's also on Atom.
These system hangs happen both with latest drm-tip and few months old drm-tip version so this kernel bug may have been there for longer time, or be HW/FW issue needing WA.
(I first thought this was due to media-driver update, but rolling back to media stack version before VT-d change did not help.)
https://gitlab.freedesktop.org/drm/intel/-/issues/4046
[KASAN] DRM-Tip 5.14 stack-out-of-bounds in unwind_next_frame
2021-09-10T11:37:01Z
Tomi Sarvela
[KASAN] DRM-Tip 5.14 stack-out-of-bounds in unwind_next_frame
Recent DRM-Tip 5.14-rc7 KASAN run on CI had hit on SKL RKL and BWR platforms.
Short log on SKL:
```
<3> [649.943340] ==================================================================
<3> [649.944993] BUG: KASAN: stack-out-of-bounds in ...
Recent DRM-Tip 5.14-rc7 KASAN run on CI had hit on SKL RKL and BWR platforms.
Short log on SKL:
```
<3> [649.943340] ==================================================================
<3> [649.944993] BUG: KASAN: stack-out-of-bounds in unwind_next_frame+0x1372/0x17a0
<3> [649.945003] Read of size 8 at addr ffffffff83c07ea0 by task swapper/0/0
<3> [649.945011]
<3> [649.945014] CPU: 0 PID: 0 Comm: swapper/0 Not tainted 5.14.0-rc7-g329f62b04cd6-kasan_277+ #1
<3> [649.945023] Hardware name: System manufacturer System Product Name/Z170 PRO GAMING, BIOS 0802 09/02/2015
<3> [649.945032] Call Trace:
<3> [649.945036] <IRQ>
<3> [649.945039] dump_stack_lvl+0x56/0x7b
<3> [649.945045] print_address_description.constprop.10.cold.14+0xe/0x2e2
<3> [649.945054] ? unwind_next_frame+0x1372/0x17a0
<3> [649.945061] ? unwind_next_frame+0x1372/0x17a0
<3> [649.945068] kasan_report.cold.15+0x83/0xdf
<3> [649.945075] ? unwind_next_frame+0x1372/0x17a0
<3> [649.945082] unwind_next_frame+0x1372/0x17a0
<3> [649.945089] ? arch_cpu_idle_exit+0x30/0x30
<3> [649.945096] ? deref_stack_reg+0x70/0x70
<3> [649.945102] ? __module_address+0x3a/0x370
<3> [649.945108] ? execlists_submission_tasklet+0x334/0x6360 [i915]
<3> [649.945219] ? execlists_submission_tasklet+0x334/0x6360 [i915]
<3> [649.945328] ? execlists_submission_tasklet+0x271/0x6360 [i915]
<3> [649.945437] ? is_module_text_address+0x26/0x40
<3> [649.945444] ? execlists_submission_tasklet+0x334/0x6360 [i915]
<3> [649.945559] ? kernel_text_address+0x69/0x120
<3> [649.945566] ? __thaw_task+0x70/0x70
<3> [649.945573] arch_stack_walk+0x86/0xf0
<3> [649.945581] ? arch_cpu_idle_exit+0x30/0x30
<3> [649.945588] stack_trace_save+0x85/0xb0
<3> [649.945595] ? stack_trace_consume_entry+0x160/0x160
<3> [649.945603] kasan_save_stack+0x19/0x40
<3> [649.945609] ? kasan_save_stack+0x19/0x40
<3> [649.945615] ? kasan_record_aux_stack+0xb0/0xc0
<3> [649.945621] ? insert_work+0x43/0x330
<3> [649.945627] ? __queue_work+0x39c/0xd40
<3> [649.945633] ? queue_work_on+0x78/0x90
<3> [649.945639] ? intel_engine_add_retire+0x1fd/0x280 [i915]
<3> [649.945750] ? __execlists_schedule_out+0xa55/0xe30 [i915]
<3> [649.945854] ? execlists_submission_tasklet+0x334/0x6360 [i915]
<3> [649.945962] ? tasklet_action_common.isra.18+0x202/0x2e0
<3> [649.945970] ? __do_softirq+0x1cb/0x84a
<3> [649.945976] ? irq_exit_rcu+0x13b/0x150
<3> [649.945981] ? common_interrupt+0x9b/0xc0
<3> [649.945988] ? asm_common_interrupt+0x1e/0x40
<3> [649.945994] ? arch_cpu_idle_exit+0x30/0x30
<3> [649.946001] ? debug_object_activate+0x327/0x4a0
<3> [649.946008] ? lock_downgrade+0x6e0/0x6e0
<3> [649.946015] ? do_raw_spin_lock+0x121/0x290
<3> [649.946022] ? rwlock_bug.part.2+0x90/0x90
<3> [649.946029] ? rcu_read_lock_bh_held+0xb0/0xb0
<3> [649.946036] ? do_raw_spin_unlock+0x4f/0x250
<3> [649.946043] ? _raw_spin_unlock_irqrestore+0x3d/0x60
<3> [649.946050] ? debug_object_activate+0x327/0x4a0
<3> [649.946057] ? debug_object_assert_init+0x380/0x380
<3> [649.946064] kasan_record_aux_stack+0xb0/0xc0
<3> [649.946071] insert_work+0x43/0x330
<3> [649.946077] __queue_work+0x39c/0xd40
<3> [649.946085] queue_work_on+0x78/0x90
<3> [649.946091] intel_engine_add_retire+0x1fd/0x280 [i915]
<3> [649.946203] ? engine_retire+0xc0/0xc0 [i915]
<3> [649.946313] ? lrc_check_regs+0x11a/0x570 [i915]
<3> [649.946423] __execlists_schedule_out+0xa55/0xe30 [i915]
<3> [649.946529] execlists_submission_tasklet+0x334/0x6360 [i915]
<3> [649.946647] ? execlists_reset_cancel+0xc40/0xc40 [i915]
<3> [649.946754] ? rcu_read_lock_sched_held+0x9c/0xd0
<3> [649.946762] ? rcu_read_lock_bh_held+0xb0/0xb0
<3> [649.946769] ? find_held_lock+0x33/0x1c0
<3> [649.946775] ? wake_bit_function+0x180/0x180
<3> [649.946784] tasklet_action_common.isra.18+0x202/0x2e0
<3> [649.946792] __do_softirq+0x1cb/0x84a
<3> [649.946799] irq_exit_rcu+0x13b/0x150
<3> [649.946805] common_interrupt+0x9b/0xc0
<3> [649.946812] </IRQ>
<3> [649.946816] asm_common_interrupt+0x1e/0x40
<3> [649.946822] RIP: 0010:do_idle+0x0/0x520
<3> [649.946828] Code: 66 2e 0f 1f 84 00 00 00 00 00 c3 0f 1f 44 00 00 66 2e 0f 1f 84 00 00 00 00 00 c7 05 9e 5d 10 03 01 00 00 00 fb c3 0f 1f 40 00 <41> 57 41 56 41 55 41 54 55 53 48 bb 00 00 00 00 00 fc ff df 65 4c
<3> [649.946845] RSP: 0018:ffffffff83c07ec0 EFLAGS: 00000296
<3> [649.946852] RAX: 0000000000000000 RBX: 0000000000000093 RCX: ffffffff82fdde5a
<3> [649.946860] RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffffffff83c21500
<3> [649.946867] RBP: 1ffffffff0780fdb R08: fffffbfff07842a1 R09: fffffbfff07842a1
<3> [649.946875] R10: ffffffff83c21507 R11: fffffbfff07842a0 R12: 00000000003d08f0
<3> [649.946882] R13: ffffffff83c21500 R14: 1ffffffff0780fd1 R15: 0000000092f3bc93
<3> [649.946891] ? schedule_idle+0x5a/0x90
<3> [649.946898] cpu_startup_entry+0x14/0x20
<3> [649.946904] start_secondary+0x22e/0x2c0
<3> [649.946911] ? set_cpu_sibling_map+0x1340/0x1340
<3> [649.946917] ? cpu_startup_entry+0x14/0x20
<3> [649.946924] ? start_kernel+0x376/0x394
<3> [649.946931] secondary_startup_64_no_verify+0xb0/0xbb
<3> [649.946941]
<3> [649.946944] KASAN internal error: frame info validation failed; invalid marker: 16
<3> [649.946951]
<3> [649.946954] Memory state around the buggy address:
<3> [649.946960] ffffffff83c07d80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
<3> [649.946968] ffffffff83c07e00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
<3> [649.946975] >ffffffff83c07e80: 00 00 00 00 f1 01 f2 f2 f2 00 00 f1 f1 f1 f1 04
<3> [649.946983] ^
<3> [649.946988] ffffffff83c07f00: f2 f2 f2 00 00 00 00 00 00 00 00 00 00 00 00 00
<3> [649.946996] ffffffff83c07f80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
<3> [649.947003] ==================================================================
```
Full log on SKL RKL and BWR:
https://intel-gfx-ci.01.org/tree/drm-tip/kasan_277/fi-skl-6700k2/igt@gem_exec_whisper@basic-queues-all.html
https://intel-gfx-ci.01.org/tree/drm-tip/kasan_277/fi-rkl-11600/igt@gem_exec_whisper@basic-queues-all.html
https://intel-gfx-ci.01.org/tree/drm-tip/kasan_277/fi-bwr-2160/igt@kms_flip@2x-flip-vs-wf_vblank@ab-vga1-dvi-d1.html
https://gitlab.freedesktop.org/drm/intel/-/issues/4045
[KASAN] DRM-Tip 5.14 stack-out-of-bounds in tick_sched_handle.isra.19
2021-09-10T11:40:50Z
Tomi Sarvela
[KASAN] DRM-Tip 5.14 stack-out-of-bounds in tick_sched_handle.isra.19
Recent DRM-Tip 5.14-rc7 KASAN run on CI had hit on BSW and RKL platforms.
Short log on RKL:
```
<3> [319.731364] ==================================================================
<3> [319.731434] BUG: KASAN: stack-out-of-bounds in tick...
Recent DRM-Tip 5.14-rc7 KASAN run on CI had hit on BSW and RKL platforms.
Short log on RKL:
```
<3> [319.731364] ==================================================================
<3> [319.731434] BUG: KASAN: stack-out-of-bounds in tick_sched_handle.isra.19+0x10b/0x120
<3> [319.731466] Read of size 8 at addr ffffffff83c07ea0 by task swapper/0/0
<3> [319.731491]
<3> [319.731500] CPU: 0 PID: 0 Comm: swapper/0 Tainted: G U 5.14.0-rc7-g329f62b04cd6-kasan_277+ #1
<3> [319.731536] Hardware name: ASUS System Product Name/TUF GAMING Z590-PLUS WIFI, BIOS 0811 04/06/2021
<3> [319.731568] Call Trace:
<3> [319.731579] <IRQ>
<3> [319.731590] dump_stack_lvl+0x56/0x7b
<3> [319.731609] print_address_description.constprop.10.cold.14+0xe/0x2e2
<3> [319.731637] ? tick_sched_handle.isra.19+0x10b/0x120
<3> [319.731659] ? tick_sched_handle.isra.19+0x10b/0x120
<3> [319.731681] ? tick_sched_do_timer+0xf0/0xf0
<3> [319.731700] kasan_report.cold.15+0x83/0xdf
<3> [319.731721] ? tick_sched_handle.isra.19+0x10b/0x120
<3> [319.731745] tick_sched_handle.isra.19+0x10b/0x120
<3> [319.731767] tick_sched_timer+0xac/0xd0
<3> [319.731785] __hrtimer_run_queues+0x48d/0xa20
<3> [319.731811] ? enqueue_hrtimer+0x2a0/0x2a0
<3> [319.731830] ? ktime_get_update_offsets_now+0xdb/0x2c0
<3> [319.731857] hrtimer_interrupt+0x2ec/0x790
<3> [319.731884] __sysvec_apic_timer_interrupt+0x141/0x4e0
<3> [319.731908] sysvec_apic_timer_interrupt+0x85/0xb0
<3> [319.731930] </IRQ>
<3> [319.731941] asm_sysvec_apic_timer_interrupt+0x12/0x20
<3> [319.731964] RIP: 0010:cpu_startup_entry+0x14/0x20
<3> [319.731985] Code: 0a 48 81 ff d0 13 ff 82 0f 92 c0 c3 66 2e 0f 1f 84 00 00 00 00 00 53 89 fb e8 08 f6 ff ff 89 df e8 b1 2f f5 ff e8 4c f6 ff ff <eb> f9 cc cc cc cc cc cc cc cc cc cc 48 b8 00 00 00 00 00 fc ff df
<3> [319.732056] RSP: 0018:ffffffff83c07ec8 EFLAGS: 00000296
<3> [319.732080] RAX: 0000000000000000 RBX: 0000000000000093 RCX: ffffffff82fdde5a
<3> [319.732108] RDX: 0000000000000000 RSI: 0000000000000008 RDI: ffffffff83c21500
<3> [319.732136] RBP: 1ffffffff0780fdb R08: fffffbfff07842a1 R09: fffffbfff07842a1
<3> [319.732165] R10: ffffffff83c21507 R11: fffffbfff07842a0 R12: 00000000003bb140
<3> [319.732193] R13: ffffffff83c21500 R14: 1ffffffff0780fd1 R15: 000000007505c460
<3> [319.732226] ? schedule_idle+0x5a/0x90
<3> [319.732247] ? cpu_startup_entry+0x14/0x20
<3> [319.732265] start_secondary+0x22e/0x2c0
<3> [319.732284] ? set_cpu_sibling_map+0x1340/0x1340
<3> [319.732312] ? cpu_startup_entry+0x14/0x20
<3> [319.732339] ? start_kernel+0x376/0x394
<3> [319.732361] secondary_startup_64_no_verify+0xb0/0xbb
<3> [319.732393]
<3> [319.732401] KASAN internal error: frame info validation failed; invalid marker: 16
<3> [319.732432]
<3> [319.732440] Memory state around the buggy address:
<3> [319.732461] ffffffff83c07d80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
<3> [319.732491] ffffffff83c07e00: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
<3> [319.732521] >ffffffff83c07e80: 00 00 00 00 f1 01 f2 f2 f2 00 00 f1 f1 f1 f1 04
<3> [319.732551] ^
<3> [319.732569] ffffffff83c07f00: f2 f2 f2 00 00 00 00 00 00 00 00 00 00 00 00 00
<3> [319.732599] ffffffff83c07f80: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00
<3> [319.732629] ==================================================================
```
Full logs on two platforms:
https://intel-gfx-ci.01.org/tree/drm-tip/kasan_277/fi-bsw-cyan/igt@kms_cursor_crc@pipe-a-cursor-256x256-offscreen.html
https://intel-gfx-ci.01.org/tree/drm-tip/kasan_277/fi-rkl-guc/igt@kms_addfb_basic@unused-modifier.html
https://gitlab.freedesktop.org/drm/intel/-/issues/2180
Up to 14% perf drop in Media transcode tests (on devices using powersave gove...
2022-11-25T13:46:15Z
Eero Tamminen
Up to 14% perf drop in Media transcode tests (on devices using powersave governor)
Setup:
* HW: BXT/APL J4205 (18EU)
* OS: Ubuntu 20.04
* XDM / Desktop: lightdm / Unity
* X server: git head (= dmabuf / E2E RBC enabled)
* kernel: drm-tip git
Between following kernel commits:
* 999bc17a2471df17a3af3001d094cf6d5d4849b0 d...
Setup:
* HW: BXT/APL J4205 (18EU)
* OS: Ubuntu 20.04
* XDM / Desktop: lightdm / Unity
* X server: git head (= dmabuf / E2E RBC enabled)
* kernel: drm-tip git
Between following kernel commits:
* 999bc17a2471df17a3af3001d094cf6d5d4849b0 drm-tip: 2020y-06m-13d-09h-30m-45s UTC integration manifest
* 593c112156feb0f6159814f2276a32c90f243823 drm-tip: 2020y-06m-15d-12h-41m-08s UTC integration manifest
(I can reproduce the drop just by booting between these kernel builds.)
Kernel performance dropped on BXT J4205 both in Media and 3D GPU tests:
- 8-14% in most of the transcode tests (both single and multi-stream tests)
(Only GPU test that improves is QSV HEVC downscale + discard test, but even that test improves only with FFmpeg/QSV, not with MEdiaSDK/QSV. I think they do threading differently.)
Because of which tests are impacted, this issue seems to concern only tests that aren't fully GPU or CPU bound, but ones that are somewhere in between.
=> I assume the reason for the perf drop is kernel changing "powersave" scaling governor to "ondemand" one, because with ClearLinux (which uses "performance" governor), there was no change in any of the tests on BXT.
As I have only GEN9 HW for this, I don't know whether perf dropped also on Atoms with newer GEN. And I'm missing usable perf data for several other benchmarks and for GLK, because of this kernel bug: https://gitlab.freedesktop.org/drm/intel/-/issues/1205
EDIT: This is also visible on GLK, but it has huge variance, so impact is visible only on longer term perf trend. Of the Core machines, this is most visible on SKL GT4e, where the impact was up to 2-3%.
https://gitlab.freedesktop.org/drm/intel/-/issues/2017
KMS tests - dmesg-warn/dmesg-fail - *ERROR* Fault errors on pipe .*: 0x00000080
2024-01-22T11:54:53Z
LAKSHMINARAYANA VUDUM
KMS tests - dmesg-warn/dmesg-fail - *ERROR* Fault errors on pipe .*: 0x00000080
https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_511/fi-skl-guc/igt@kms_plane_scaling@pipe-c-scaler-with-rotation.html#dmesg-warnings341
````
7> [27.507630] [drm:drm_mode_addfb2] [FB:129]
<7> [27.509052] i915 0000:00:02.0: [drm:skl_update...
https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_511/fi-skl-guc/igt@kms_plane_scaling@pipe-c-scaler-with-rotation.html#dmesg-warnings341
````
7> [27.507630] [drm:drm_mode_addfb2] [FB:129]
<7> [27.509052] i915 0000:00:02.0: [drm:skl_update_scaler [i915]] scaler_user index 2.6: staged scaling request for 8x8->1920x1200 scaler_users = 0x40
<7> [27.509118] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] ddb ( 0 - 0) -> ( 0 - 860), size 0 -> 860
<7> [27.509148] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] level wm0, wm1, wm2, wm3, wm4, wm5, wm6, wm7, twm, swm -> *wm0,*wm1,*wm2,*wm3,*wm4,*wm5,*wm6,*wm7, twm, swm
<7> [27.509185] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] lines 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 -> 0, 16, 16, 16, 16, 16, 16, 16, 0, 0
<7> [27.509221] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] blocks 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 -> 2, 3, 3, 3, 3, 3, 3, 3, 0, 0
<7> [27.509246] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] min_ddb 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 -> 3, 4, 4, 4, 4, 4, 4, 4, 0, 0
<7> [27.509296] i915 0000:00:02.0: [drm:intel_atomic_setup_scalers [i915]] Attached scaler id 2.0 to PLANE:73
<3> [27.524049] i915 0000:00:02.0: [drm] *ERROR* Fault errors on pipe C: 0x00000080
<3> [27.524077] i915 0000:00:02.0: [drm] *ERROR* Fault errors on pipe C: 0x00000080
<3> [27.524081] DMAR: DRHD: handling fault status reg 3
<3> [27.524130] DMAR: [DMA Read] Request device [00:02.0] PASID ffffffff fault addr 0 [fault reason 06] PTE Read access is not set
<7> [27.524354] i915 0000:00:02.0: [drm:i915_fifo_underrun_reset_write [i915]] Re-arming FIFO underruns on pipe C
<7> [27.524544] i915 0000:00:02.0: [drm:skl_update_scaler [i915]] scaler_user index 2.6: Staged freeing scaler id 0 scaler_users = 0x0
<7> [27.524585] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] ddb ( 0 - 860) -> ( 0 - 0), size 860 -> 0
<7> [27.524609] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] level *wm0,*wm1,*wm2,*wm3,*wm4,*wm5,*wm6,*wm7, twm, swm -> wm0, wm1, wm2, wm3, wm4, wm5, wm6, wm7, twm, swm
<7> [27.524632] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] lines 0, 16, 16, 16, 16, 16, 16, 16, 0, 0 -> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
<7> [27.524654] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] blocks 2, 3, 3, 3, 3, 3, 3, 3, 0, 0 -> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
<7> [27.524676] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] min_ddb 3, 4, 4, 4, 4, 4, 4, 4, 0, 0 -> 0, 0, 0, 0, 0, 0, 0, 0, 0, 0
<7> [27.541029] [drm:drm_mode_addfb2] [FB:130]
<7> [27.542506] i915 0000:00:02.0: [drm:skl_update_scaler [i915]] scaler_user index 2.6: staged scaling request for 8x8->1920x1200 scaler_users = 0x40
<7> [27.542549] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] ddb ( 0 - 0) -> ( 0 - 860), size 0 -> 860
<7> [27.542572] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] level wm0, wm1, wm2, wm3, wm4, wm5, wm6, wm7, twm, swm -> *wm0,*wm1,*wm2,*wm3,*wm4,*wm5,*wm6,*wm7, twm, swm
<7> [27.542595] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] lines 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 -> 0, 16, 16, 16, 16, 16, 16, 16, 0, 0
<7> [27.542617] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] blocks 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 -> 2, 3, 3, 3, 3, 3, 3, 3, 0, 0
<7> [27.542639] i915 0000:00:02.0: [drm:skl_compute_wm [i915]] [PLANE:73:plane 1C] min_ddb 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 -> 3, 4, 4, 4, 4, 4, 4, 4, 0, 0
<7> [27.542686] i915 0000:00:02.0: [drm:intel_atomic_setup_scalers [i915]] Attached scaler id 2.0 to PLANE:73
<3> [27.557393] i915 0000:00:02.0: [drm] *ERROR* Fault errors on pipe C: 0x00000080
<3> [27.557413] DMAR: DRHD: handling fault status reg 3
<3> [27.557424] i915 0000:00:02.0: [drm] *ERROR* Fault errors on pipe C: 0x00000080
````
https://gitlab.freedesktop.org/drm/intel/-/issues/562
[regression] (recoverable) GPU resets & fails in slower 3D benchmarks
2022-07-28T15:57:48Z
Bugzilla Migration User
[regression] (recoverable) GPU resets & fails in slower 3D benchmarks
## Submitted by Eero Tamminen `@eero-t`
Assigned to **Chris Wilson `@ickle`**
**[Link to original bug (#112169)](https://bugs.freedesktop.org/show_bug.cgi?id=112169)**
## Description
Setup:<br>
* HW: e.g. J4205 (BXT) and 7567U-i7 ...
## Submitted by Eero Tamminen `@eero-t`
Assigned to **Chris Wilson `@ickle`**
**[Link to original bug (#112169)](https://bugs.freedesktop.org/show_bug.cgi?id=112169)**
## Description
Setup:<br>
* HW: e.g. J4205 (BXT) and 7567U-i7 (KBL GT3e)<br>
* OS: Ubuntu 18.04<br>
* SW: latest Git versions of drm-tip kernel, X server and Mesa<br>
<br>
Between following drm-tip commits:<br>
* 2019-10-23 16:27:36 863a8a1bef: drm-tip: 2019y-10m-23d-16h-23m-44s UTC integration manifest<br>
* 2019-10-27 20:46:21 54520983c6: drm-tip: 2019y-10m-27d-20h-45m-34s UTC integration manifest<br>
<br>
Some of the slower 3D benchmarks started to GPU hang:<br>
------------------------------------------<br>
[ 4937.365687] Iteration 1/3: GpuTest /test=pixmark_piano /width=1366 /height=768 /msaa=1 /no_scorebox /benchmark /benchmark_duration_ms=35000<br>
[ 4937.904763] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4938.016748] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4938.128707] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4938.213360] Iteration 2/3: GpuTest /test=pixmark_piano /width=1366 /height=768 /msaa=1 /no_scorebox /benchmark /benchmark_duration_ms=35000<br>
[ 4938.736916] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4938.848767] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4938.960794] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4939.013764] Iteration 3/3: GpuTest /test=pixmark_piano /width=1366 /height=768 /msaa=1 /no_scorebox /benchmark /benchmark_duration_ms=35000<br>
[ 4939.536891] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4939.648780] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4940.560756] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4940.623625] Iteration 1/3: GpuTest /test=pixmark_volplosion /width=1366 /height=768 /msaa=1 /no_scorebox /benchmark /benchmark_duration_ms=35000<br>
[ 4941.112755] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4941.224772] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4941.504756] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4941.569851] Iteration 2/3: GpuTest /test=pixmark_volplosion /width=1366 /height=768 /msaa=1 /no_scorebox /benchmark /benchmark_duration_ms=35000<br>
[ 4942.032771] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4942.600708] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4942.864828] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4942.920061] Iteration 3/3: GpuTest /test=pixmark_volplosion /width=1366 /height=768 /msaa=1 /no_scorebox /benchmark /benchmark_duration_ms=35000<br>
[ 4943.408815] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4943.520753] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
[ 4943.632747] i915 0000:00:02.0: Resetting rcs0 for preemption time out<br>
------------------------------------------<br>
<br>
These are mainly Piano & Volplosion in public GpuTest v0.7 test-suite, and most of the tests in the internal GPU MemBW test-suite.<br>
<br>
Notes:<br>
* There's no i915 error state<br>
* Screen updates in the failing benchmarks happen on average at single digit (2-8) FPS on BXT<br>
* These aren't slowest benchmarks, e.g. offscreen GfxBench tests run at <1 FPS, and don't have GPU resets
### See also
* https://bugs.freedesktop.org/show_bug.cgi?id=111424
Chris Wilson
Chris Wilson
https://gitlab.freedesktop.org/drm/intel/-/issues/180
All tests - dmesg-warn/dmesg-fail.incomplete - *ERROR* DPCD read failed, addr...
2023-05-22T10:30:37Z
Bugzilla Migration User
All tests - dmesg-warn/dmesg-fail.incomplete - *ERROR* DPCD read failed, address .*
## Submitted by Martin Peres `@mupuf`
Assigned to **Swati2 Sharma `@swati2.sharma`**
**[Link to original bug (#108566)](https://bugs.freedesktop.org/show_bug.cgi?id=108566)**
## Description
https://intel-gfx-ci.01.org/tree/drm-tip...
## Submitted by Martin Peres `@mupuf`
Assigned to **Swati2 Sharma `@swati2.sharma`**
**[Link to original bug (#108566)](https://bugs.freedesktop.org/show_bug.cgi?id=108566)**
## Description
https://intel-gfx-ci.01.org/tree/drm-tip/drmtip_133/fi-apl-guc/igt@gem_workarounds@suspend-resume-context.html<br>
<br>
`<3>` [226.775927] [drm:lspcon_write_infoframe [i915]] *ERROR* DPCD read failed, address 0x5df<br>
`<3>` [226.775994] [drm:lspcon_write_infoframe [i915]] *ERROR* Failed to write AVI infoframes
Swati2 Sharma
Swati2 Sharma