[Bug 1129035] New: Very high load caused by GTM on Dell 7290
vainfo
vdpauinfo display: :0 screen: 0
http://bugzilla.suse.com/show_bug.cgi?id=1129035 Bug ID: 1129035 Summary: Very high load caused by GTM on Dell 7290 Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: X.Org Assignee: xorg-maintainer-bugs@forge.provo.novell.com Reporter: vpelcak@suse.com QA Contact: xorg-maintainer-bugs@forge.provo.novell.com Found By: --- Blocker: --- Hello. I used to use Lenovo X230 and replaced ith with Dell 7290. I have 2 monitors connected to it. Whenever I'm doing GoToMeetings session displayed on one of the connected monitors (Dell U2412Mb) in recent chromium, all 8 threads are loaded at ~50% and UI becomes unresponsive and coolers start to freak out. # lspci | grep -i vga 00:02.0 VGA compatible controller: Intel Corporation UHD Graphics 620 (rev 07) # glxinfo | grep render direct rendering: Yes GLX_MESA_multithread_makecurrent, GLX_MESA_query_renderer, GLX_INTEL_swap_event, GLX_MESA_copy_sub_buffer, GLX_MESA_query_renderer, Extended renderer info (GLX_MESA_query_renderer): OpenGL renderer string: Mesa DRI Intel(R) UHD Graphics 620 (Kabylake GT2) GL_ARB_conditional_render_inverted, GL_ARB_conservative_depth, GL_NV_conditional_render, GL_NV_depth_clamp, GL_ARB_conditional_render_inverted, GL_ARB_conservative_depth, GL_MESA_window_pos, GL_NV_blend_square, GL_NV_conditional_render, GL_EXT_render_snorm, GL_EXT_robustness, GL_EXT_separate_shader_objects, GL_OES_element_index_uint, GL_OES_fbo_render_mipmap, When I try to play some HD video in VLC, load is not so high (but still significant (one thread reports 50% in htop). Message by vlc is here: libva info: VA-API version 1.4.0 libva info: va_getDriverName() returns 0 libva info: User requested driver 'iHD' libva info: Trying to open /usr/lib64/dri/iHD_drv_video.so libva info: Found init function __vaDriverInit_1_3 libva info: va_openDriver() returns 0 [00007f2f00d9a530] avcodec decoder: Using Intel iHD driver - 1.0.0 for hardware decoding libva info: VA-API version 1.4.0 libva info: va_getDriverName() returns 0 libva info: User requested driver 'iHD' libva info: Trying to open /usr/lib64/dri/iHD_drv_video.so libva info: Found init function __vaDriverInit_1_3 libva info: va_openDriver() returns 0 vainfo: VA-API version: 1.4 (libva 2.3.0) vainfo: Driver version: Intel iHD driver - 1.0.0 vainfo: Supported profile and entrypoints VAProfileNone : VAEntrypointVideoProc VAProfileNone : VAEntrypointStats VAProfileMPEG2Simple : VAEntrypointVLD VAProfileMPEG2Simple : VAEntrypointEncSlice VAProfileMPEG2Main : VAEntrypointVLD VAProfileMPEG2Main : VAEntrypointEncSlice VAProfileH264Main : VAEntrypointVLD VAProfileH264Main : VAEntrypointEncSlice VAProfileH264Main : VAEntrypointFEI VAProfileH264Main : VAEntrypointEncSliceLP VAProfileH264High : VAEntrypointVLD VAProfileH264High : VAEntrypointEncSlice VAProfileH264High : VAEntrypointFEI VAProfileH264High : VAEntrypointEncSliceLP VAProfileVC1Simple : VAEntrypointVLD VAProfileVC1Main : VAEntrypointVLD VAProfileVC1Advanced : VAEntrypointVLD VAProfileJPEGBaseline : VAEntrypointVLD VAProfileJPEGBaseline : VAEntrypointEncPicture VAProfileH264ConstrainedBaseline: VAEntrypointVLD VAProfileH264ConstrainedBaseline: VAEntrypointEncSlice VAProfileH264ConstrainedBaseline: VAEntrypointFEI VAProfileH264ConstrainedBaseline: VAEntrypointEncSliceLP VAProfileVP8Version0_3 : VAEntrypointVLD VAProfileHEVCMain : VAEntrypointVLD VAProfileHEVCMain : VAEntrypointEncSlice VAProfileHEVCMain : VAEntrypointFEI VAProfileHEVCMain10 : VAEntrypointVLD VAProfileHEVCMain10 : VAEntrypointEncSlice VAProfileVP9Profile0 : VAEntrypointVLD VAProfileVP9Profile2 : VAEntrypointVLD libva info: VA-API version 1.4.0 libva info: va_getDriverName() returns 0 libva info: User requested driver 'iHD' libva info: Trying to open /usr/lib64/dri/iHD_drv_video.so libva info: Found init function __vaDriverInit_1_3 libva info: va_openDriver() returns 0 API version: 1 Information string: OpenGL/VAAPI backend for VDPAU Video surface: name width height types ------------------------------------------- 420 4096 4096 NV12 YV12 UYVY YUYV Y8U8V8A8 V8U8Y8A8 422 4096 4096 NV12 YV12 UYVY YUYV Y8U8V8A8 V8U8Y8A8 444 4096 4096 NV12 YV12 UYVY YUYV Y8U8V8A8 V8U8Y8A8 Decoder capabilities: name level macbs width height ---------------------------------------------------- MPEG1 --- not supported --- MPEG2_SIMPLE --- not supported --- MPEG2_MAIN --- not supported --- H264_BASELINE 51 16384 2048 2048 H264_MAIN 51 16384 2048 2048 H264_HIGH 51 16384 2048 2048 VC1_SIMPLE --- not supported --- VC1_MAIN --- not supported --- VC1_ADVANCED --- not supported --- MPEG4_PART2_SP --- not supported --- MPEG4_PART2_ASP --- not supported --- DIVX4_QMOBILE --- not supported --- DIVX4_MOBILE --- not supported --- DIVX4_HOME_THEATER --- not supported --- DIVX4_HD_1080P --- not supported --- DIVX5_QMOBILE --- not supported --- DIVX5_MOBILE --- not supported --- DIVX5_HOME_THEATER --- not supported --- DIVX5_HD_1080P --- not supported --- H264_CONSTRAINED_BASELINE 51 16384 2048 2048 H264_EXTENDED --- not supported --- H264_PROGRESSIVE_HIGH --- not supported --- H264_CONSTRAINED_HIGH --- not supported --- H264_HIGH_444_PREDICTIVE --- not supported --- HEVC_MAIN --- not supported --- HEVC_MAIN_10 --- not supported --- HEVC_MAIN_STILL --- not supported --- HEVC_MAIN_12 --- not supported --- HEVC_MAIN_444 --- not supported --- Output surface: name width height nat types ---------------------------------------------------- B8G8R8A8 16384 16384 y R8G8B8A8 16384 16384 y R10G10B10A2 16384 16384 y B10G10R10A2 16384 16384 y A8 16384 16384 y Bitmap surface: name width height ------------------------------ B8G8R8A8 16384 16384 R8G8B8A8 16384 16384 R10G10B10A2 16384 16384 B10G10R10A2 16384 16384 A8 16384 16384 Video mixer: feature name sup ------------------------------------ DEINTERLACE_TEMPORAL - DEINTERLACE_TEMPORAL_SPATIAL - INVERSE_TELECINE - NOISE_REDUCTION - SHARPNESS - LUMA_KEY - HIGH QUALITY SCALING - L1 - HIGH QUALITY SCALING - L2 - HIGH QUALITY SCALING - L3 - HIGH QUALITY SCALING - L4 - HIGH QUALITY SCALING - L5 - HIGH QUALITY SCALING - L6 - HIGH QUALITY SCALING - L7 - HIGH QUALITY SCALING - L8 - HIGH QUALITY SCALING - L9 - parameter name sup min max ----------------------------------------------------- VIDEO_SURFACE_WIDTH - VIDEO_SURFACE_HEIGHT - CHROMA_TYPE - LAYERS - attribute name sup min max ----------------------------------------------------- BACKGROUND_COLOR - CSC_MATRIX - NOISE_REDUCTION_LEVEL - SHARPNESS_LEVEL - LUMA_KEY_MIN_LUMA - LUMA_KEY_MAX_LUMA - -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c1
--- Comment #1 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c2
--- Comment #2 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c3
Stefan Dirsch
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c4
--- Comment #4 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c5
--- Comment #5 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c6
Vit Pelcak
Hmm. Any reasons why are you setting all these i915 kernel module options and use xf86-video-intel instead of modesetting X driver, which is our default on Kabylake? I suggest to use our defaults first ....
Other things I noticed when looking at your description.
I see that you're already using intel-media-driver as backend for libva (which is our default on newer platforms now instead of intel-vaapi-driver). Indeed performance should be better with intel-media-driver than with intel-vaapi-driver, but maybe it's not. It's rather new ...
I was wondering why vdpauinfo works on your Intel GPU system. Seems you have installed libvdpau_va_gl1 wrapper, which maps things from VDPAU to libVA. Maybe that's the issue here, if GTM tries to use VDPAU, then everything goes through libVA. Maybe if VDPAU would fail, it would fallback to libVA and then could use the direct path. But this is only a shot in the dark! I don't know anything about GTM.
Hello. Yes, I removed intel driver. Problems stay. Actually, I initially already had problems, therefore I tried intel driver. When I test with some 4k 60 fps video on youtube (regardless of the browser), then all cores are heavily loaded even while using modeset driver. Switching screens in GNOME causes quite heavy load on system as well. That suggests that video acceleration is not working well. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1129035
Stefan Dirsch
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c7
--- Comment #7 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c8
--- Comment #8 from Stefan Dirsch
Created attachment 799898 [details] Xorg.log
X.org log after all content of /etc/X11/xorg.conf.d/50-device.conf was removed and modeset driver gets loaded.
Looks good. Nothing obvious, but I would remove "intel_pstate=skylake_hwp" form kernel options as well ... -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c9
--- Comment #9 from Vit Pelcak
(In reply to Vit Pelcak from comment #4)
Created attachment 799898 [details] Xorg.log
X.org log after all content of /etc/X11/xorg.conf.d/50-device.conf was removed and modeset driver gets loaded.
Looks good. Nothing obvious, but I would remove "intel_pstate=skylake_hwp" form kernel options as well ...
Done. Nothing has changed. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c10
--- Comment #10 from Stefan Dirsch
# cat /etc/modprobe.d/50-i915.conf options i915 enable_dc=1 enable_guc=2 enable_fbc=1 enable_psr=1 disable_power_well=0 modeset=1
And removing this? Run dracut after doing this! -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c11
--- Comment #11 from Vit Pelcak
# cat /etc/modprobe.d/50-i915.conf options i915 enable_dc=1 enable_guc=2 enable_fbc=1 enable_psr=1 disable_power_well=0 modeset=1
And removing this? Run dracut after doing this!
Done. No change at all. I just had a GTM session and all 8 threads were ~70% loaded. Just firmware is no longer being loaded. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c12
--- Comment #12 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c13
Stefan Dirsch
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c14
--- Comment #14 from Vit Pelcak
Ok. Last thing you could try is uninstalling intel-media-driver package and install intel-vaapi-driver. In case things work better with old libva backend.
I tried it. juno:~ # vainfo error: XDG_RUNTIME_DIR not set in the environment. libva info: VA-API version 1.4.0 libva info: va_getDriverName() returns 0 libva info: Trying to open /usr/lib64/dri/i965_drv_video.so libva info: Found init function __vaDriverInit_1_4 libva info: va_openDriver() returns 0 vainfo: VA-API version: 1.4 (libva 2.3.0) vainfo: Driver version: Intel i965 driver for Intel(R) Kaby Lake - 2.3.0 I tried it with https://www.youtube.com/watch?v=LXb3EKWsInQ in chromium and FF. I experienced tearing, occasional short freezes and all 8 threads still quite heavily loaded as well as with GTM.
Otherwise I don't know either. We could try to reproduce the issue here with a Dell 7280, which unfortunately still uses Skylake instead of Kabylake. :-(
Possibly things are not related to gfx drivers at all. GTM may work better with Firefox or Google chrome (the proprietary one).
I'll give it a try. Thank you for suggestion. -- You are receiving this mail because: You are on the CC list for the bug.
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c15
--- Comment #15 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c16
Stefan Dirsch
http://bugzilla.suse.com/show_bug.cgi?id=1129035
Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
Stefan Dirsch
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c19
--- Comment #19 from Stefan Dirsch
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c20
--- Comment #20 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c21
--- Comment #21 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c22
--- Comment #22 from Fabian Baumanis
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c23
--- Comment #23 from Vit Pelcak
http://bugzilla.suse.com/show_bug.cgi?id=1129035
http://bugzilla.suse.com/show_bug.cgi?id=1129035#c24
Stefan Dirsch
Hmm.
I have been trying https://www.youtube.com/watch?v=HSsqzzuGTPo
And I see the major difference between chrome and Firefox. With latter being way easier on the system.
Maybe this could be the case.
Thing is, that GTM doesn't work in Firefox. So I cannot compare them. It doesn't explain how come that after moving from X230 to Dell 7290 GTM started to be such a problem.
In theory things should have been improved with better hardware, but I'm pretty sure you're comparing different systems now, i.e. different Firefox/Chrome. Also display now has a bigger resolution, but this should hopefully been compensated by a more powerful GPU. I'm afraid we can't help here. There is nothing obviously wrongly configured or broken. -- You are receiving this mail because: You are on the CC list for the bug.
participants (1)
-
bugzilla_noreply@novell.com