Observations from an Intel Video user

ewaller · 2010-10-20 17:28:56

Last night, I turned on testing and played with the new suite of Intel Drivers / Xorg / Mesa.

I use Compiz with Xfce4 (I am a sucker for eye candy)

Mostly, I was looking to see the effects of telling Compiz to use direct rendering. It appears that the problems I had seen in the past with screens missing update events with direct rendering have been vanquished. Good job.

But...

Certain applications with certain behaviors cause Compiz to quietly crash, requiring that it be restarted (often)

The applications that I noticed were smplayer, Vlc, the Gimp, and Okular. It seems that the crash always occurs when these programs went full screen or were maximized, but there were times Compiz crashed when a window hit an edge of the screen. I am wondering if it has to do with graphic overlays fighting with Compiz when something related to a screen edge happens.

I don't have a clue as to which package I would file a bug report against. Instead, I thought I'd post here to suggest that great progress has been made, but that this suite is not ready for prime time just yet.

karol · 2010-10-20 18:26:41

Would you mind sharing what card exactly are you using?

ewaller · 2010-10-20 19:58:41

karol wrote:

Would you mind sharing what card exactly are you using?

Not in the Least!

00:02.0 VGA compatible controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07) (prog-if 00 [VGA controller])
        Subsystem: Hewlett-Packard Company Device 30f7
        Flags: bus master, fast devsel, latency 0, IRQ 45
        Memory at d0000000 (64-bit, non-prefetchable) [size=4M]
        Memory at c0000000 (64-bit, prefetchable) [size=256M]
        I/O ports at 60f0 [size=8]
        Expansion ROM at <unassigned> [disabled]
        Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
        Capabilities: [d0] Power Management version 3
        Kernel driver in use: i915

00:02.1 Display controller: Intel Corporation Mobile 4 Series Chipset Integrated Graphics Controller (rev 07)
        Subsystem: Hewlett-Packard Company Device 30f7
        Flags: bus master, fast devsel, latency 0
        Memory at d5400000 (64-bit, non-prefetchable) [size=1M]
        Capabilities: [d0] Power Management version 3

ewaller@odin:~ 1003 %sudo lspci -n
...
00:02.0 0300: 8086:2a42 (rev 07)
00:02.1 0380: 8086:2a43 (rev 07)
...

From everything.log:

Oct 17 17:55:21 odin kernel: Linux agpgart interface v0.103
Oct 17 17:55:21 odin kernel: agpgart-intel 0000:00:00.0: Intel GM45 Chipset
Oct 17 17:55:21 odin kernel: agpgart-intel 0000:00:00.0: detected 65532K stolen
memory, trimming to 32768K
Oct 17 17:55:21 odin kernel: agpgart-intel 0000:00:00.0: AGP aperture is 256M @
0xc0000000
Oct 17 17:55:21 odin kernel: [drm] Initialized drm 1.1.0 20060810
Oct 17 17:55:21 odin kernel: i915 0000:00:02.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16
Oct 17 17:55:21 odin kernel: i915 0000:00:02.0: setting latency timer to 64
Oct 17 17:55:21 odin kernel: [drm] detected 63M stolen memory, trimming to 32M
Oct 17 17:55:21 odin kernel: i915 0000:00:02.0: irq 45 for MSI/MSI-X
Oct 17 17:55:21 odin kernel: [drm] set up 32M of stolen space
Oct 17 17:55:21 odin kernel: [drm:init_ring_common] *ERROR* render ring head not reset to zero ctl 00000000 head 02001000 tail 00000000 start 02001000
Oct 17 17:55:21 odin kernel: [drm:init_ring_common] *ERROR* render ring head forced to zero ctl 00000000 head 00000000 tail 00000000 start 02001000
Oct 17 17:55:21 odin kernel: Console: switching to colour frame buffer device 160x50
Oct 17 17:55:21 odin kernel: fb0: inteldrmfb frame buffer device
Oct 17 17:55:21 odin kernel: drm: registered panic notifier

ewaller@odin:~ 1006 %sudo grep intel /var/log/Xorg.0.log
[    24.653] (==) Matched intel as autoconfigured driver 0
[    24.653] (II) LoadModule: "intel"
[    24.653] (II) Loading /usr/lib/xorg/modules/drivers/intel_drv.so
[    24.688] (II) Module intel: vendor="X.Org Foundation"
[    24.713] (II) intel: Driver for Intel Integrated Graphics Chipsets: i810,
[    24.805] (II) intel(0): Creating default Display subsection in Screen section
[    24.805] (==) intel(0): Depth 24, (--) framebuffer bpp 32
[    24.805] (==) intel(0): RGB weight 888
[    24.805] (==) intel(0): Default visual is TrueColor
[    24.805] (II) intel(0): Integrated Graphics Chipset: Intel(R) GM45
[    24.805] (--) intel(0): Chipset: "GM45"
[    24.805] (==) intel(0): video overlay key set to 0x101fe
[    24.830] (II) intel(0): Output VGA1 has no monitor section
[    24.936] (II) intel(0): Output LVDS1 has no monitor section
[    24.937] (II) intel(0): found backlight control interface /sys/class/backlight/acpi_video0
[    24.946] (II) intel(0): Output HDMI1 has no monitor section
[    24.946] (II) intel(0): Output DP1 has no monitor section
[    24.947] (II) intel(0): Output DP2 has no monitor section
[    24.966] (II) intel(0): EDID for output VGA1
[    25.073] (II) intel(0): EDID for output LVDS1
[    25.073] (II) intel(0): Manufacturer: CMO  Model: 1425  Serial#: 0
[    25.073] (II) intel(0): Year: 2007  Week: 14
[    25.073] (II) intel(0): EDID Version: 1.3
[    25.073] (II) intel(0): Digital Display Input
[    25.073] (II) intel(0): Max Image Size [cm]: horiz.: 30  vert.: 19
[    25.073] (II) intel(0): Gamma: 2.20
[    25.073] (II) intel(0): No DPMS capabilities specified
[    25.073] (II) intel(0): Supported color encodings: RGB 4:4:4 YCrCb 4:4:4
[    25.073] (II) intel(0): First detailed timing is preferred mode
[    25.073] (II) intel(0): redX: 0.588 redY: 0.337   greenX: 0.315 greenY: 0.534
[    25.073] (II) intel(0): blueX: 0.152 blueY: 0.125   whiteX: 0.313 whiteY: 0.329
[    25.073] (II) intel(0): Manufacturer's mask: 0
[    25.073] (II) intel(0): Supported detailed timing:
[    25.073] (II) intel(0): clock: 69.3 MHz   Image Size:  303 x 190 mm
[    25.073] (II) intel(0): h_active: 1280  h_sync: 1319  h_sync_end 1345 h_blank_end 1410 h_border: 0
[    25.073] (II) intel(0): v_active: 800  v_sync: 802  v_sync_end 806 v_blanking: 819 v_border: 0
[    25.073] (II) intel(0):  N141I3-L02
[    25.073] (II) intel(0):  CMO
[    25.073] (II) intel(0):  N141I3-L02
[    25.073] (II) intel(0): EDID (in hex):
[    25.073] (II) intel(0):     00ffffffffffff000daf251400000000
[    25.073] (II) intel(0):     0e110103801e13780a9f059656508827
[    25.073] (II) intel(0):     20505400000001010101010101010101
[    25.073] (II) intel(0):     010101010101111b008250201330271a
[    25.073] (II) intel(0):     24002fbe10000018000000fe004e3134
[    25.073] (II) intel(0):     3149332d4c30320a2020000000fe0043
[    25.073] (II) intel(0):     4d4f0a202020202020202020000000fe
[    25.073] (II) intel(0):     004e31343149332d4c30320a20200070
[    25.073] (II) intel(0): Not using default mode "320x240" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "400x300" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "512x384" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "640x480" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "640x512" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "800x600" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "896x672" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "928x696" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "960x720" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "700x525" (doublescan mode not supported)
[    25.073] (II) intel(0): Not using default mode "1024x768" (doublescan mode not supported)
[    25.073] (II) intel(0): Printing probed modes for output LVDS1
[    25.073] (II) intel(0): Modeline "1280x800"x60.0   69.29  1280 1319 1345 1410  800 802 806 819 -hsync -vsync (49.1 kHz)
[    25.073] (II) intel(0): Modeline "1024x768"x60.0   65.00  1024 1048 1184 1344  768 771 777 806 -hsync -vsync (48.4 kHz)
[    25.073] (II) intel(0): Modeline "800x600"x60.3   40.00  800 840 968 1056  600 601 605 628 +hsync +vsync (37.9 kHz)
[    25.073] (II) intel(0): Modeline "800x600"x56.2   36.00  800 824 896 1024  600 601 603 625 +hsync +vsync (35.2 kHz)
[    25.073] (II) intel(0): Modeline "640x480"x59.9   25.18  640 656 752 800  480 490 492 525 -hsync -vsync (31.5 kHz)
[    25.082] (II) intel(0): EDID for output HDMI1
[    25.082] (II) intel(0): EDID for output DP1
[    25.083] (II) intel(0): EDID for output DP2
[    25.083] (II) intel(0): Output VGA1 disconnected
[    25.083] (II) intel(0): Output LVDS1 connected
[    25.083] (II) intel(0): Output HDMI1 disconnected
[    25.083] (II) intel(0): Output DP1 disconnected
[    25.083] (II) intel(0): Output DP2 disconnected
[    25.083] (II) intel(0): Using exact sizes for initial modes
[    25.084] (II) intel(0): Output LVDS1 using initial mode 1280x800
[    25.084] (II) intel(0): Using default gamma of (1.0, 1.0, 1.0) unless otherwise stated.
[    25.084] (II) intel(0): Pageflipping enabled in kernel, but disabled in X
[    25.084] (II) intel(0): Don't panic: https://bugzilla.redhat.com/588421
[    25.084] (==) intel(0): DPI set to (96, 96)
[    25.098] (II) intel(0): [DRI2] Setup complete
[    25.098] (II) intel(0): [DRI2]   DRI driver: i965
[    25.098] (**) intel(0): Tiling enabled
[    25.098] (**) intel(0): SwapBuffers wait enabled
[    25.098] (==) intel(0): VideoRam: 262144 KB
[    25.098] (II) intel(0): Allocated new frame buffer 1280x800 stride 5120, tiled
[    25.130] (==) intel(0): Backing store disabled
[    25.130] (==) intel(0): Silken mouse enabled
[    25.130] (II) intel(0): Initializing HW Cursor
[    25.160] (II) intel(0): RandR 1.2 enabled, ignore the following RandR disabled message.
[    25.194] (==) intel(0): DPMS enabled
[    25.194] (==) intel(0): Intel XvMC decoder enabled
[    25.194] (II) intel(0): Set up textured video
[    25.194] (II) intel(0): [XvMC] xvmc_vld driver initialized.
[    25.194] (II) intel(0): direct rendering: DRI2 Enabled
[    25.241] (II) intel(0): Setting screen physical size to 338 x 211
[   129.999] (II) intel(0): EDID vendor "CMO", prod id 5157
[   129.999] (II) intel(0): Printing DDC gathered Modelines:
[   129.999] (II) intel(0): Modeline "1280x800"x0.0   69.29  1280 1319 1345 1410  800 802 806 819 -hsync -vsync (49.1 kHz)
[   133.512] (II) intel(0): EDID vendor "CMO", prod id 5157
[   133.513] (II) intel(0): Printing DDC gathered Modelines:
[   133.513] (II) intel(0): Modeline "1280x800"x0.0   69.29  1280 1319 1345 1410  800 802 806 819 -hsync -vsync (49.1 kHz)
ewaller@odin:~ 1007 %

Edit: Note that I backed down off of Testing to get some other things done. If you need logs from the system while running the testing versions, I'll need to get those tonight (GMT-7). These logs are brought to you courtesy of ssh.

Last edited by ewaller (2010-10-20 20:01:25)

JGC · 2010-10-21 21:39:14

It's possible that you get freezes with the current driver in testing. It's a bug in the kernel related to pageflipping, I opened a bugreport about that, but our kernel maintainers haven't looked at it yet. Kernel 2.6.36 which was released today contains both the patches referenced in that bugreport, so we'll have to wait for 2.6.36 to appear in the repositories before the new intel driver is working the way it should.

DarksideEE7 · 2010-10-22 23:57:09

I recently reformatted and installed Arch on my laptop with an Intel 4500MHD, where before it was running quite well. This install was about a year old, and I've played around and tinkered so much that instead of trying to fix certain problems, I thought it would be faster to simply wipe and reinstall.

Anyhow, upon setting up KDE and installing my drivers, I noticed that there were some problems where thumbnails in the 'peek' function of the Smooth Tasks plasmoid would render partial images only in black and white. This also happened with the ctrl+10 function. When this happens FPS grinds to a halt.

I then upgraded to the testing driver and it fixed this. However I frequently have X lockup, and I have to press ctrl+alt+f2, then kill X and restart. Sometimes the keyboard stops responding, then I have to ssh in from my desktop and restart the entire laptop, very annoying.

Now I just enabled testing and I'm updating all of the packages, in hopes that xorg-server in testing may help (or other packages).

grassmonk · 2010-10-23 00:22:09

I am using an Intel core i3, with KDE 4.5, and Mesa 7.9 and Intel 2.13.0 from testing. I was experiencing no problems except with multi-screen, where enabling compositing would appear to offset one of my displays, but only visually (e.g. clicking the mouse would click where the stuff on the screen was actually supposed to be, not where it was displayed).

I tried my hand at building the new kernel 2.6.36 and everything is working great now, except for one minor annoyance, which doesn't affect actual use: during boot or when enabling compositing, my secondary screen turns on and off a few times, but once the system is up and running stays on. I expect the official Arch kernel will be better, but until then I am happy with what I have running.

DarksideEE7 · 2010-10-23 00:51:06

grassmonk wrote:

I am using an Intel core i3, with KDE 4.5, and Mesa 7.9 and Intel 2.13.0 from testing. I was experiencing no problems except with multi-screen, where enabling compositing would appear to offset one of my displays, but only visually (e.g. clicking the mouse would click where the stuff on the screen was actually supposed to be, not where it was displayed).
I tried my hand at building the new kernel 2.6.36 and everything is working great now, except for one minor annoyance, which doesn't affect actual use: during boot or when enabling compositing, my secondary screen turns on and off a few times, but once the system is up and running stays on. I expect the official Arch kernel will be better, but until then I am happy with what I have running.

What graphics chipset are you running? Kernel module in use?

i915 here. I also get this really strange interference, where the bottom half or third of the screen is completely scrambled, then it stays that way until I move the mouse or zoom in, then back out. Really annoying.

Intel driver performance is just embarrassing.

axel668 · 2010-10-23 06:22:15

That's it, putting Intel drivers on "ignore" until Kernel 2.6.36 comes in ...

grassmonk · 2010-10-23 07:08:46

DarksideEE7 wrote:

What graphics chipset are you running? Kernel module in use?
i915 here. I also get this really strange interference, where the bottom half or third of the screen is completely scrambled, then it stays that way until I move the mouse or zoom in, then back out. Really annoying.
Intel driver performance is just embarrassing.

00:02.0 VGA compatible controller: Intel Corporation Core Processor Integrated Graphics Controller (rev 02) (prog-if 00 [VGA controller])
        Subsystem: Hewlett-Packard Company Device 144b
        Flags: bus master, fast devsel, latency 0, IRQ 40
        Memory at c0000000 (64-bit, non-prefetchable) [size=4M]
        Memory at b0000000 (64-bit, prefetchable) [size=256M]
        I/O ports at 5050 [size=8]
        Expansion ROM at <unassigned> [disabled]
        Capabilities: [90] MSI: Enable+ Count=1/1 Maskable- 64bit-
        Capabilities: [d0] Power Management version 2
        Capabilities: [a4] PCI Advanced Features
        Kernel driver in use: i915
        Kernel modules: i915

glxinfo says

Mesa DRI Intel(R) Ironlake Mobile GEM 20100330 DEVELOPMENT

It seems I am also getting random video freezes when watching video fullscreen (in smplayer). The system is still completely responsive, and going out of fullscreen starts the video again, so... I dunno. This is with compositing on; with compositing suspended the video hasn't stopped yet.

RomikT · 2010-11-09 22:46:00

grassmonk wrote:

It seems I am also getting random video freezes when watching video fullscreen (in smplayer). The system is still completely responsive, and going out of fullscreen starts the video again, so... I dunno. This is with compositing on; with compositing suspended the video hasn't stopped yet.

I'm getting such freezes too.

kernel module - i915
arch - x86_64

kernel26 - 2.6.36-3
mesa - 7.9-1
xf86-video-intel - 2.13.0-4

ewaller · 2010-11-10 04:25:21

Here is an update. I have the following installed:

glproto-git 20101108-1
intel-dri-git 20101108-1
kernel26-vanilla 2.6.36-2
libdrm-git 20101108-1
libgl-git 20101108-1
xf86-video-intel-git 20101108-1

Compiz with direct rendering works (so far) perfectly -- including with the programs with which I had problems in the OP.
Kwin is a lot better. It enables effects and uses direct rendering without any fuss. Unfortunately, some effects cause hard freezes.

Edit: Oh, and I just watched 15 minutes of "The Matrix" on SMplayer with linear interpolation de-interlacing. No freezes, nice video.

Last edited by ewaller (2010-11-10 04:28:51)

Square · 2010-11-10 04:46:08

Try disabling the tiling support in driconf. This solved crashes with 3D applications for me.

RomikT · 2010-11-10 07:42:09

Disabling tiling helped with fullscreen video freezes. However I had to put this option into global /etc/drirc as ~/.drirc was completely ignored.

UPDATE: Unfortunatelly it didn't help. Freezes just become much rarer, but didn't disappear.

Last edited by RomikT (2010-11-10 20:09:34)

Arch Linux

#1 2010-10-20 17:28:56

Observations from an Intel Video user

#2 2010-10-20 18:26:41

Re: Observations from an Intel Video user

#3 2010-10-20 19:58:41

Re: Observations from an Intel Video user

#4 2010-10-21 21:39:14

Re: Observations from an Intel Video user

#5 2010-10-22 23:57:09

Re: Observations from an Intel Video user

#6 2010-10-23 00:22:09

Re: Observations from an Intel Video user

#7 2010-10-23 00:51:06

Re: Observations from an Intel Video user

#8 2010-10-23 06:22:15

Re: Observations from an Intel Video user

#9 2010-10-23 07:08:46

Re: Observations from an Intel Video user

#10 2010-11-09 22:46:00

Re: Observations from an Intel Video user

#11 2010-11-10 04:25:21

Re: Observations from an Intel Video user

#12 2010-11-10 04:46:08

Re: Observations from an Intel Video user

#13 2010-11-10 07:42:09

Re: Observations from an Intel Video user

Board footer