Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Marc MERLIN]

Hi,

My performance for normal work (2D, although many things like chrome use
the GPU nowadays), and video (mplayer) is between just usable to
horrible.
Performance is bad enough that I'm about to replace the mainboard to get
one with the Nvidia GeForce GT 730M 1GB. I have no idea how good or bad
nouveau is on that, but I'm a bit desperate.

Before I do that, can you let me know if I have my driver or Xorg tuned
wrong somehow?

Playing video with mplayer or youtube makes the whole UI very
unresponsive.
Playing a video with mplayer is mostly ok, but if I stretch the window
just a bit, the rescalling is too slow to play real time and yet I have
a quad core i7-4800MQ, so I doubt it's a CPU problem. Going to full
screeen usually makes things better.

I do use mplayer -vo gl since it's the only way to change
colors/contrast/brightness.

Kernel is 3.19, but problem has been there since I've had the laptop.
X.Org X Server 1.16.4

I do boot my kernel with these options I found recommended to save
power, maybe I should remove some of them because they kill performance?
pcie_aspm=force i915.i915_enable_rc6=7 i915.i915_enable_fbc=1 i915.semaphores=1 i915.lvds_downclock=1 acpi_backlight=vendor

Relevant bits form Xorg.log:
[    48.720] (II) xfree86: Adding drm device (/dev/dri/card0)
[    49.096] (--) PCI:*(0:0:2:0) 8086:0416:17aa:2210 rev 6, Mem @ 0xf0000000/4194304, 0xe0000000/268435456, I/O @ 0x00004000/64
[    49.096] (II) "glx" will be loaded by default.
[    49.096] (II) LoadModule: "glx"
[    49.101] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
[    49.113] (II) Module glx: vendor="X.Org Foundation"
[    49.113] 	compiled for 1.16.4, module version = 1.0.0
[    49.113] 	ABI class: X.Org Server Extension, version 8.0
[    49.113] (**) AIGLX enabled
[    49.113] (II) LoadModule: "intel"
[    49.114] (II) Loading /usr/lib/xorg/modules/drivers/intel_drv.so
[    49.122] (II) Module intel: vendor="X.Org Foundation"
[    49.122] 	compiled for 1.15.99.904, module version = 2.21.15
[    49.122] 	Module class: X.Org Video Driver
[    49.122] 	ABI class: X.Org Video Driver, version 18.0
[    49.122] (II) intel: Driver for Intel(R) Integrated Graphics Chipsets:
	i810, i810-dc100, i810e, i815, i830M, 845G, 854, 852GM/855GM, 865G,
	915G, E7221 (i915), 915GM, 945G, 945GM, 945GME, Pineview GM,
	Pineview G, 965G, G35, 965Q, 946GZ, 965GM, 965GME/GLE, G33, Q35, Q33,
	GM45, 4 Series, G45/G43, Q45/Q43, G41, B43, HD Graphics,
	HD Graphics 2000, HD Graphics 3000, HD Graphics 2500,
	HD Graphics 4000, HD Graphics P4000, HD Graphics 4600,
	HD Graphics 5000, HD Graphics P4600/P4700, Iris(TM) Graphics 5100,
	HD Graphics 4400, HD Graphics 4200, Iris(TM) Pro Graphics 5200
[    49.122] (++) using VT number 7

[    49.133] (==) intel(0): Depth 24, (--) framebuffer bpp 32
[    49.133] (==) intel(0): RGB weight 888
[    49.133] (==) intel(0): Default visual is TrueColor
[    49.133] (--) intel(0): Integrated Graphics Chipset: Intel(R) HD Graphics 4600
[    49.133] (**) intel(0): Relaxed fencing enabled
[    49.133] (**) intel(0): Wait on SwapBuffers? enabled
[    49.133] (**) intel(0): Triple buffering? enabled
[    49.133] (**) intel(0): Framebuffer tiled
[    49.133] (**) intel(0): Pixmaps tiled
[    49.133] (**) intel(0): 3D buffers tiled
[    49.133] (**) intel(0): SwapBuffers wait enabled
[    49.133] (==) intel(0): video overlay key set to 0x101fe
[    49.133] (II) intel(0): Output eDP1 using monitor section Monitor0
[    49.133] (--) intel(0): found backlight control interface /sys/class/backlight/intel_backlight
[    49.134] (II) intel(0): Output VGA1 has no monitor section
[    49.134] (II) intel(0): Output DP1 has no monitor section
[    49.164] (II) intel(0): Output HDMI1 has no monitor section
[    49.164] (II) intel(0): Output DP2 has no monitor section
[    49.164] (II) intel(0): Output HDMI2 has no monitor section
[    49.164] (II) intel(0): EDID for output eDP1
[    49.164] (II) intel(0): Manufacturer: MEI  Model: 96a2  Serial#: 0
[    49.164] (II) intel(0): Year: 2013  Week: 255
[    49.164] (II) intel(0): EDID Version: 1.4
[    49.164] (II) intel(0): Digital Display Input
[    49.164] (II) intel(0): 8 bits per channel
[    49.164] (II) intel(0): Digital interface is DisplayPort
[    49.164] (II) intel(0): Max Image Size [cm]: horiz.: 34  vert.: 19

[    49.201] (II) Loading sub module "dri2"
[    49.201] (II) LoadModule: "dri2"
[    49.201] (II) Module "dri2" already built-in
[    49.201] (==) Depth 24 pixmap format is 32 bpp
[    49.201] (II) intel(0): [DRI2] Setup complete
[    49.201] (II) intel(0): [DRI2]   DRI driver: i965
[    49.201] (II) intel(0): Allocated new frame buffer 2880x1620 stride 11776, tiled
[    49.201] (II) UXA(0): Driver registered support for the following operations:
[    49.201] (II)         solid
[    49.201] (II)         copy
[    49.201] (II)         composite (RENDER acceleration)
[    49.201] (II)         put_image
[    49.201] (II)         get_image
[    49.201] (==) intel(0): Backing store enabled
[    49.201] (==) intel(0): Silken mouse enabled
[    49.201] (II) intel(0): Initializing HW Cursor
[    49.201] (II) intel(0): RandR 1.2 enabled, ignore the following RandR disabled message.
[    49.201] (==) intel(0): DPMS enabled
[    49.201] (==) intel(0): Intel XvMC decoder enabled
[    49.202] (II) intel(0): Set up textured video
[    49.202] (II) intel(0): [XvMC] xvmc_vld driver initialized.
[    49.202] (II) intel(0): direct rendering: DRI2 Enabled
[    49.202] (==) intel(0): hotplug detection: "enabled"
[    49.220] (--) RandR disabled
[    49.223] (II) SELinux: Disabled on system
[    49.266] (II) AIGLX: enabled GLX_MESA_copy_sub_buffer
[    49.266] (II) AIGLX: enabled GLX_ARB_create_context
[    49.266] (II) AIGLX: enabled GLX_ARB_create_context_profile
[    49.266] (II) AIGLX: enabled GLX_EXT_create_context_es2_profile
[    49.266] (II) AIGLX: enabled GLX_INTEL_swap_event
[    49.266] (II) AIGLX: enabled GLX_SGI_swap_control and GLX_MESA_swap_control
[    49.266] (II) AIGLX: enabled GLX_EXT_framebuffer_sRGB
[    49.266] (II) AIGLX: enabled GLX_ARB_fbconfig_float
[    49.266] (II) AIGLX: GLX_EXT_texture_from_pixmap backed by buffer objects
[    49.266] (II) AIGLX: enabled GLX_ARB_create_context_robustness
[    49.267] (II) AIGLX: Loaded and initialized i965
[    49.267] (II) GLX: Initialized DRI2 GL provider for screen 0
[    49.267] (II) intel(0): Setting screen physical size to 762 x 428

Thanks for any suggestions,
Marc
-- 
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
                                      .... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/                         | PGP 1024R/763BE901
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Chris Wilson]

On Thu, Apr 23, 2015 at 03:22:44PM -0700, Marc MERLIN wrote:
> I do boot my kernel with these options I found recommended to save
> power, maybe I should remove some of them because they kill performance?
> pcie_aspm=force i915.i915_enable_rc6=7 i915.i915_enable_fbc=1 i915.semaphores=1 i915.lvds_downclock=1 acpi_backlight=vendor

Those options are broken.

> Relevant bits form Xorg.log:
> [    49.114] (II) Loading /usr/lib/xorg/modules/drivers/intel_drv.so
> [    49.122] (II) Module intel: vendor="X.Org Foundation"
> [    49.122] 	compiled for 1.15.99.904, module version = 2.21.15

Start by grabbing an updated xf86-video-intel.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Marc MERLIN]

Hi Chris, thanks for the answer,

On Thu, Apr 23, 2015 at 11:49:24PM +0100, Chris Wilson wrote:
> On Thu, Apr 23, 2015 at 03:22:44PM -0700, Marc MERLIN wrote:
> > I do boot my kernel with these options I found recommended to save
> > power, maybe I should remove some of them because they kill performance?
> > pcie_aspm=force i915.i915_enable_rc6=7 i915.i915_enable_fbc=1 i915.semaphores=1 i915.lvds_downclock=1 acpi_backlight=vendor
> 
> Those options are broken.
 
Ok, which ones specifically? :)
Without some of those, battery use from the driver has been bad in the past,
but I don't know which one does what exactly :)

- pcie_aspm=force
- i915.i915_enable_rc6=7
- i915.i915_enable_fbc=1
- i915.semaphores=1
- i915.lvds_downclock=1

> > Relevant bits form Xorg.log:
> > [    49.114] (II) Loading /usr/lib/xorg/modules/drivers/intel_drv.so
> > [    49.122] (II) Module intel: vendor="X.Org Foundation"
> > [    49.122] 	compiled for 1.15.99.904, module version = 2.21.15
> 
> Start by grabbing an updated xf86-video-intel.

On debian testing and unstable, it's the latest:
legolas:~# apt-get install xserver-xorg-video-intel/unstable
Reading package lists... Done
Building dependency tree       
Reading state information... Done
xserver-xorg-video-intel is already the newest version.
Selected version '2:2.21.15-2+b2' (Debian:testing [amd64]) for 'xserver-xorg-video-intel'

But now that you mention it, I just found this, so I'll try this out:
legolas:~# apt-get install xserver-xorg-video-intel/experimental
Reading package lists... Done
Building dependency tree       
Reading state information... Done
Selected version '2:2.99.917-1~exp1' (Debian:experimental [amd64]) for 'xserver-xorg-video-intel'

On top of being bleeding edge enoug not to be in debian unstable, does it have 
performance enhancements that you are hopeful will help my situation?

Thanks,
Marc
-- 
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
                                      .... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/  
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Chris Wilson]

On Thu, Apr 23, 2015 at 05:31:01PM -0700, Marc MERLIN wrote:
> Hi Chris, thanks for the answer,
> 
> On Thu, Apr 23, 2015 at 11:49:24PM +0100, Chris Wilson wrote:
> > On Thu, Apr 23, 2015 at 03:22:44PM -0700, Marc MERLIN wrote:
> > > I do boot my kernel with these options I found recommended to save
> > > power, maybe I should remove some of them because they kill performance?
> > > pcie_aspm=force i915.i915_enable_rc6=7 i915.i915_enable_fbc=1 i915.semaphores=1 i915.lvds_downclock=1 acpi_backlight=vendor
> > 
> > Those options are broken.
>  
> Ok, which ones specifically? :)
> Without some of those, battery use from the driver has been bad in the past,
> but I don't know which one does what exactly :)
> 
> - pcie_aspm=force

Probably saved your battery life, though the kernel should now be fixed.

> - i915.i915_enable_rc6=7

Broken, unvalidated mode for your hardware.

> - i915.i915_enable_fbc=1

Buggy. Can be very slow, causes GPU hangs and machine lockups.

> - i915.semaphores=1

Default.

> - i915.lvds_downclock=1

Subject to visual bugs.

Setting (almost) any of the i915 options will taint your kernel (i.e.
mark it as being in a known broken state).

> > > Relevant bits form Xorg.log:
> > > [    49.114] (II) Loading /usr/lib/xorg/modules/drivers/intel_drv.so
> > > [    49.122] (II) Module intel: vendor="X.Org Foundation"
> > > [    49.122] 	compiled for 1.15.99.904, module version = 2.21.15
> > 
> > Start by grabbing an updated xf86-video-intel.
> 
> On debian testing and unstable, it's the latest:
> legolas:~# apt-get install xserver-xorg-video-intel/unstable
> Reading package lists... Done
> Building dependency tree       
> Reading state information... Done
> xserver-xorg-video-intel is already the newest version.
> Selected version '2:2.21.15-2+b2' (Debian:testing [amd64]) for 'xserver-xorg-video-intel'
> 
> But now that you mention it, I just found this, so I'll try this out:
> legolas:~# apt-get install xserver-xorg-video-intel/experimental
> Reading package lists... Done
> Building dependency tree       
> Reading state information... Done
> Selected version '2:2.99.917-1~exp1' (Debian:experimental [amd64]) for 'xserver-xorg-video-intel'
> 
> On top of being bleeding edge enoug not to be in debian unstable, does it have 
> performance enhancements that you are hopeful will help my situation?

Lots.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Marc MERLIN]

On Fri, Apr 24, 2015 at 07:53:29AM +0100, Chris Wilson wrote:
> > On top of being bleeding edge enoug not to be in debian unstable, does it have 
> > performance enhancements that you are hopeful will help my situation?
> 
> Lots.

:)

Ok, so I owe you thanks.

Between the module options and the new driver, things are better.
I can now play and resize a video with mplayer without it dropping to 5fps
or less. I see some tearing in the video being played on occasion whether
it's original size, resized window, or full screen, but otherwise it plays
full speed and my system doesn't slow down to a crawl.

So things are much better, thank you :)

Marc
-- 
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
                                      .... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/  
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Chris Wilson]

On Fri, Apr 24, 2015 at 08:35:33AM -0700, Marc MERLIN wrote:
> On Fri, Apr 24, 2015 at 07:53:29AM +0100, Chris Wilson wrote:
> > > On top of being bleeding edge enoug not to be in debian unstable, does it have 
> > > performance enhancements that you are hopeful will help my situation?
> > 
> > Lots.
> 
> :)
> 
> Ok, so I owe you thanks.
> 
> Between the module options and the new driver, things are better.
> I can now play and resize a video with mplayer without it dropping to 5fps
> or less. I see some tearing in the video being played on occasion whether
> it's original size, resized window, or full screen, but otherwise it plays
> full speed and my system doesn't slow down to a crawl.

(In the absence of driver bugs...)

Whether of not it tears depends upon your window manager. On bare X,
using mplayer -vo xv or -vo gl, should not tear. Under a compositing
window manager, it depends upon how it decides to update the screen. To
force everything to update without tearing, use

echo > /etc/X11/xorg.conf.d/90-intel.conf <<EOF
Section "Device"
	Identifier "igfx"
	Option "TearFree"
EndSection
EOF
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Marc MERLIN]

On Fri, Apr 24, 2015 at 04:53:53PM +0100, Chris Wilson wrote:
> Whether of not it tears depends upon your window manager. On bare X,
> using mplayer -vo xv or -vo gl, should not tear. Under a compositing
> window manager, it depends upon how it decides to update the screen. To
> force everything to update without tearing, use
> 
> echo > /etc/X11/xorg.conf.d/90-intel.conf <<EOF
> Section "Device"
> 	Identifier "igfx"
> 	Option "TearFree"
> EndSection
> EOF

Cool, thanks for the tip and your help.

I was curious to see how my power use would be, and got about the same power
use at idle with LCD to minimum:
The battery reports a discharge rate of 12.4 W
The estimated remaining time is 7 hours, 27 minutes

Summary: 171.5 wakeups/second,  0.1 GPU ops/seconds, 0.0 VFS ops/sec and 2.0% CPU use

Power est.              Usage       Events/s    Category       Description
  3.73 W      1.9%                      Device         Display backlight
  1.16 W      4.7 ms/s      65.3        Process        [btrfs-transacti]
  741 mW      0.9 ms/s      41.9        Interrupt      [4] block(softirq)
  175 mW    156.0 µs/s       9.9        Process        [rcu_sched]


Then, if I turn my screen off (backlight 0), it actually seems to save another 7W:
The battery reports a discharge rate of 5.06 W
The estimated remaining time is 18 hours, 5 minutes

Summary: 79.4 wakeups/second,  0.2 GPU ops/seconds, 0.0 VFS ops/sec and 1.8% CPU use

Power est.              Usage       Events/s    Category       Description
  229 mW      1.4 ms/s      13.4        Process        [btrfs-transacti]
  181 mW      4.1 ms/s      10.4        Process        /usr/bin/enlightenment
  174 mW    206.2 µs/s      10.2        Interrupt      [4] block(softirq)


So the good news is that the driver sems to do fine on batteries with the
default options now.

Thanks,
Marc
-- 
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
                                      .... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/  
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Marc MERLIN]

[-- Attachment #1: Type: text/plain, Size: 4071 bytes --]

On Fri, Apr 24, 2015 at 04:53:53PM +0100, Chris Wilson wrote:
> Whether of not it tears depends upon your window manager. On bare X,
> using mplayer -vo xv or -vo gl, should not tear. Under a compositing
> window manager, it depends upon how it decides to update the screen. To
> force everything to update without tearing, use
> 
> echo > /etc/X11/xorg.conf.d/90-intel.conf <<EOF
> Section "Device"
> 	Identifier "igfx"
> 	Option "TearFree"
> EndSection
> EOF

So, I added this at the end of my xorg.conf, but Xorg.0.log says:
'intel(0): TearFree disabled'

Any idea why?

[   272.438] (II) xfree86: Adding drm device (/dev/dri/card0)
[   272.852] (--) PCI:*(0:0:2:0) 8086:0416:17aa:2210 rev 6, Mem @ 0xf0000000/4194304, 0xe0000000/268435456, I/O @ 0x00004000/64
[   272.852] (II) "glx" will be loaded by default.
[   272.852] (II) LoadModule: "glx"
[   272.852] (II) Loading /usr/lib/xorg/modules/extensions/libglx.so
[   272.853] (II) Module glx: vendor="X.Org Foundation"
[   272.853] 	compiled for 1.16.4, module version = 1.0.0
[   272.853] 	ABI class: X.Org Server Extension, version 8.0
[   272.853] (**) AIGLX enabled
[   272.853] (II) LoadModule: "intel"
[   272.853] (II) Loading /usr/lib/xorg/modules/drivers/intel_drv.so
[   272.853] (II) Module intel: vendor="X.Org Foundation"
[   272.853] 	compiled for 1.16.2.901, module version = 2.99.917
[   272.853] 	Module class: X.Org Video Driver
[   272.853] 	ABI class: X.Org Video Driver, version 18.0
[   272.853] (II) intel: Driver for Intel(R) Integrated Graphics Chipsets:
	i810, i810-dc100, i810e, i815, i830M, 845G, 854, 852GM/855GM, 865G,
	915G, E7221 (i915), 915GM, 945G, 945GM, 945GME, Pineview GM,
	Pineview G, 965G, G35, 965Q, 946GZ, 965GM, 965GME/GLE, G33, Q35, Q33,
	GM45, 4 Series, G45/G43, Q45/Q43, G41, B43
[   272.854] (II) intel: Driver for Intel(R) HD Graphics: 2000-6000
[   272.854] (II) intel: Driver for Intel(R) Iris(TM) Graphics: 5100, 6100
[   272.854] (II) intel: Driver for Intel(R) Iris(TM) Pro Graphics: 5200, 6200, P6300
[   272.854] (++) using VT number 8

[   272.866] (II) intel(0): Using Kernel Mode Setting driver: i915, version 1.6.0 20141121
[   272.866] (II) intel(0): SNA compiled: xserver-xorg-video-intel 2:2.99.917-1~exp1 (Vincent Cheng <vcheng@debian.org>)
[   272.867] (--) intel(0): Integrated Graphics Chipset: Intel(R) HD Graphics 4600
[   272.867] (--) intel(0): CPU: x86-64, sse2, sse3, ssse3, sse4.1, sse4.2, avx, avx2
[   272.867] (==) intel(0): Depth 24, (--) framebuffer bpp 32
[   272.867] (==) intel(0): RGB weight 888
[   272.867] (==) intel(0): Default visual is TrueColor
[   272.867] (II) intel(0): Output eDP1 using monitor section Monitor0
[   272.867] (--) intel(0): Found backlight control interface intel_backlight (type 'raw') for output eDP1
[   272.867] (II) intel(0): Enabled output eDP1
[   272.867] (II) intel(0): Output VGA1 has no monitor section
[   272.867] (II) intel(0): Enabled output VGA1
[   272.867] (II) intel(0): Output DP1 has no monitor section
[   272.867] (II) intel(0): Enabled output DP1
[   272.867] (II) intel(0): Output HDMI1 has no monitor section
[   272.867] (II) intel(0): Enabled output HDMI1
[   272.867] (II) intel(0): Output DP2 has no monitor section
[   272.867] (II) intel(0): Enabled output DP2
[   272.867] (II) intel(0): Output HDMI2 has no monitor section
[   272.867] (II) intel(0): Enabled output HDMI2
[   272.867] (--) intel(0): Using a maximum size of 256x256 for hardware cursors
[   272.867] (II) intel(0): Output VIRTUAL1 has no monitor section
[   272.867] (II) intel(0): Enabled output VIRTUAL1
[   272.867] (--) intel(0): Output eDP1 using initial mode 2880x1620 on pipe 0
[   272.867] (==) intel(0): TearFree disabled
[   272.867] (==) intel(0): DPI set to (96, 96)
[   272.867] (II) Loading sub module "dri2"
-- 
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
                                      .... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/                         | PGP 1024R/763BE901

[-- Attachment #2: xorg.conf --]
[-- Type: text/plain, Size: 3410 bytes --]

Section "ServerLayout"
	Identifier     "X.org Configured"
	Screen      0  "Screen0" 0 0
	InputDevice    "Mouse0" "CorePointer"
	InputDevice    "Keyboard0" "CoreKeyboard"
EndSection

#Section "Files"
#	ModulePath   "/usr/lib/xorg/modules"
#	FontPath     "/usr/share/fonts/X11/misc"
#	FontPath     "/usr/share/fonts/X11/100dpi/:unscaled"
#	FontPath     "/usr/share/fonts/X11/75dpi/:unscaled"
#	FontPath     "/usr/share/fonts/X11/Type1"
#	FontPath     "/usr/share/fonts/X11/100dpi"
#	FontPath     "/usr/share/fonts/X11/75dpi"
#	FontPath     "/var/lib/defoma/x-ttcidfont-conf.d/dirs/TrueType"
#	FontPath     "built-ins"
#EndSection

Section "Module"
	#Load  "vnc"
EndSection

Section "InputDevice"
	Identifier  "Keyboard0"
	Driver      "kbd"
        Option          "XkbRules"      "xorg"
	Option          "XkbModel"      "pc105"
	Option          "XkbLayout"     "us"
	Option  "XkbOptions" "terminate:ctrl_alt_bksp"
EndSection

# This is needed by the new dbus keyboard driver.
Section "InputClass"
    Identifier          "Keyboard Defaults"
    MatchIsKeyboard	"yes"
    Option              "XkbOptions" "terminate:ctrl_alt_bksp"
EndSection


Section "InputDevice"
	Identifier  "Mouse0"
	Driver      "mouse"
	Option	    "Protocol" "auto"
	Option	    "Device" "/dev/input/mice"
	Option	    "ZAxisMapping" "4 5 6 7"
EndSection

Section "Monitor"
	Identifier   "Monitor0"
	VendorName   "Monitor Vendor"
	ModelName    "Monitor Model"
EndSection


Section "Device"
        ### Available Driver options are:-
        ### Values: <i>: integer, <f>: float, <bool>: "True"/"False",
        ### <string>: "String", <freq>: "<f> Hz/kHz/MHz"
        ### [arg]: arg optional
        #Option     "NoAccel"            	# [<bool>]
        #Option     "SWcursor"           	# [<bool>]
        #Option     "ColorKey"           	# <i>
        #Option     "CacheLines"         	# <i>
        #Option     "Dac6Bit"            	# [<bool>]
        #Option     "DRI"                	# [<bool>]
        #Option     "NoDDC"              	# [<bool>]
        #Option     "ShowCache"          	# [<bool>]
        #Option     "XvMCSurfaces"       	# <i>
        #Option     "PageFlip"           	# [<bool>]
	Identifier  "Intel GM45"
	Driver      "intel"
	VendorName  "Intel Corporation"
	BoardName   "Mobile 4 Series Chipset Integrated Graphics Controller"
EndSection



Section "Screen"
	Identifier "Screen0"
	Device     "Card0"
	Monitor    "Monitor0"
	SubSection "Display"
		Viewport   0 0
		Depth     24
	EndSubSection
EndSection

# Custom

Section "DRI"
	Mode	0666
EndSection

# http://www.x.org/wiki/IntelGraphicsDriver
# http://intellinuxgraphics.org/
Section "ServerFlags"
	Option  "AIGLX" "on"
	# this would disable hotplugging that creates this:
	# (WW) Hotplugging is on, devices using drivers 'kbd', 'mouse' or 'vmmouse' will be disabled.
	# (WW) Disabling Mouse0
	# (WW) Disabling Keyboard0
	# More info here: https://wiki.archlinux.org/index.php/Xorg#Disabling_Input_Hot-plugging
	#Option "AutoAddDevices" "False"
	# https://wiki.ubuntu.com/X/Backtracing
	Option "NoTrapSignals" "true"
	#For instance, to start Xorg from within gdb (over a ssh connection), start gdb:
	#sudo gdb /usr/bin/Xorg 2>&1 | tee gdb-Xorg.txt
	#inside gdm, start up Xorg:
	#(gdb) run -keeptty -dumbSched
EndSection

Section "Extensions"
	Option		"Composite"	"Enable"
EndSection

Section "Device"
      Identifier "igfx"
      Option "TearFree"
EndSection

[-- Attachment #3: Type: text/plain, Size: 159 bytes --]

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Chris Wilson]

On Sat, Apr 25, 2015 at 11:31:57AM -0700, Marc MERLIN wrote:
> On Fri, Apr 24, 2015 at 04:53:53PM +0100, Chris Wilson wrote:
> > Whether of not it tears depends upon your window manager. On bare X,
> > using mplayer -vo xv or -vo gl, should not tear. Under a compositing
> > window manager, it depends upon how it decides to update the screen. To
> > force everything to update without tearing, use
> > 
> > echo > /etc/X11/xorg.conf.d/90-intel.conf <<EOF
> > Section "Device"
> > 	Identifier "igfx"
> > 	Option "TearFree"
> > EndSection
> > EOF
> 
> So, I added this at the end of my xorg.conf, but Xorg.0.log says:
> 'intel(0): TearFree disabled'

You created a second Device section for the option that didn't match the
first (which is used by default). So X is simply not using Option
"TearFree".

Note that your Xorg is recent enough for the existing xorg.conf to be
entirely superfluous.
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Marc MERLIN]

On Sat, Apr 25, 2015 at 09:02:12PM +0100, Chris Wilson wrote:
> > So, I added this at the end of my xorg.conf, but Xorg.0.log says:
> > 'intel(0): TearFree disabled'
> 
> You created a second Device section for the option that didn't match the
> first (which is used by default). So X is simply not using Option
> "TearFree".
> 
> Note that your Xorg is recent enough for the existing xorg.conf to be
> entirely superfluous.

Indeed, I'm an idiot, sorry about this.
It's been so long since I touched that file, and as you say most of it is
not needed anymore.

Thanks,
Marc
-- 
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
                                      .... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/  
_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: Very bad performance from intel chipset on Thinkpad T540p (2880x1620)

[Dave Gordon]

On 24/04/15 17:18, Marc MERLIN wrote:
> On Fri, Apr 24, 2015 at 04:53:53PM +0100, Chris Wilson wrote:
>> Whether of not it tears depends upon your window manager. On bare X,
>> using mplayer -vo xv or -vo gl, should not tear. Under a compositing
>> window manager, it depends upon how it decides to update the screen. To
>> force everything to update without tearing, use
>>
>> echo > /etc/X11/xorg.conf.d/90-intel.conf <<EOF
>> Section "Device"
>> 	Identifier "igfx"
>> 	Option "TearFree"
>> EndSection
>> EOF
> 
> Cool, thanks for the tip and your help.
> 
> I was curious to see how my power use would be, and got about the same power
> use at idle with LCD to minimum:
> The battery reports a discharge rate of 12.4 W
> The estimated remaining time is 7 hours, 27 minutes
> 
> Summary: 171.5 wakeups/second,  0.1 GPU ops/seconds, 0.0 VFS ops/sec and 2.0% CPU use
> 
> Power est.              Usage       Events/s    Category       Description
>   3.73 W      1.9%                      Device         Display backlight
>   1.16 W      4.7 ms/s      65.3        Process        [btrfs-transacti]
>   741 mW      0.9 ms/s      41.9        Interrupt      [4] block(softirq)
>   175 mW    156.0 µs/s       9.9        Process        [rcu_sched]
> 
> 
> Then, if I turn my screen off (backlight 0), it actually seems to save another 7W:

On many small devices (phones, tablets, notebooks, modern laptops) the
backlight consumes about as much as everything else put together! So #1
energy-saving tip is always "turn the brightness down" followed by "turn
the screen off at every opportunity".

.Dave.

> The battery reports a discharge rate of 5.06 W
> The estimated remaining time is 18 hours, 5 minutes
> 
> Summary: 79.4 wakeups/second,  0.2 GPU ops/seconds, 0.0 VFS ops/sec and 1.8% CPU use
> 
> Power est.              Usage       Events/s    Category       Description
>   229 mW      1.4 ms/s      13.4        Process        [btrfs-transacti]
>   181 mW      4.1 ms/s      10.4        Process        /usr/bin/enlightenment
>   174 mW    206.2 µs/s      10.2        Interrupt      [4] block(softirq)
> 
> 
> So the good news is that the driver sems to do fine on batteries with the
> default options now.
> 
> Thanks,
> Marc

_______________________________________________
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx