Skip to content

GitLab

Switch branch/tag
  1. 02 Jun, 2013 1 commit
    • Harm Hanemaaijer's avatar
      G2D: Implement an area threshold for using G2D blits. · b3c2fd2c
      Harm Hanemaaijer authored 
      

Due to the overhead of G2D for small screen-to-screen blits, CPU blits
are faster for small areas. This patch introduces are threshold below
which CPU blits are triggered. It is currently set to 1000 for 32bpp
and 2500 for 16bpp based on test results.

Some benchmarks:

1920x1080x16bppx60Hz, ShadowFB disabled:

x11perf -scroll10

Before:
1500000 trep @   0.0239 msec ( 41800.0/sec): Scroll 10x10 pixels
After:
2500000 trep @   0.0110 msec ( 90900.0/sec): Scroll 10x10 pixels

x11perf -copywinwin10

Before:
1200000 trep @   0.0247 msec ( 40500.0/sec): Copy 10x10 from window to window
After:
1800000 trep @   0.0146 msec ( 68600.0/sec): Copy 10x10 from window to window
Signed-off-by: default avatarHarm Hanemaaijer <fgenfb@yahoo.com>
      b3c2fd2c
  3. 22 Apr, 2013 1 commit
    • Siarhei Siamashka's avatar
      test: race condition on DRI2 buffers allocation when going fullscreen · d147e25f
      Siarhei Siamashka authored 
      

This test program exposes a problem related to window resizing
(or going fullscreen), which is may happen exactly between "back"
and "front" DRI2 buffers allocation.

The xtrace log with some annotations:

000:<:004c:  8: DRI2-Request(151,3): CreateDrawable drawable=0x02200001
000:<:004d: 16: DRI2-Request(151,5): GetBuffers drawable=0x02200001 attachments={attachment=BackLeft(0x00000001)};
000:>:004d:52: Reply to GetBuffers: width=480 height=480 buffers={attachment=BackLeft(0x00000001)
                                    name=0x00000157 pitch=1920 cpp=4 flags=0x00000000};

Get the BackLeft buffer.

000:<:004e:  4: Request(43): GetInputFocus
000:>:004e:32: Reply to GetInputFocus: revert-to=PointerRoot(0x01) focus=0x02200001
000:<:004f: 24: Request(16): InternAtom only-if-exists=false(0x00) name='_NET_WM_STATE'
000:>:004f:32: Reply to InternAtom: atom=0xff("_NET_WM_STATE")
000:<:0050: 32: Request(16): InternAtom only-if-exists=false(0x00) name='_NET_WM_STATE_FULLSCREEN'
000:>:0050:32: Reply to InternAtom: atom=0x102("_NET_WM_STATE_FULLSCREEN")
000:<:0051: 44: Request(25): SendEvent propagate=false(0x00) destination=0x00000170 event-mask=SubstructureNotify,SubstructureRedirect
                             ClientMessage(33) format=0x20 window=0x02200001 type=0xff("_NET_WM_STATE")
                             data=0x01,0x00,0x00,0x00,0x02,0x01,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00;
000:<:0052:  4: Request(43): GetInputFocus
000:>:0052: Event DRI2-InvalidateBuffers(102) drawable=0x02200001

Here the X server attempts to notify the client side DRI2 code in the Mali blob
that the DRI2 buffer must be requested again. But this event gets happily ignored.

000:>:0052: Event Expose(12) window=0x02200001 x=0 y=0 width=1920 height=1080 count=0x0000
000:>:0052:32: Reply to GetInputFocus: revert-to=PointerRoot(0x01) focus=0x02200001
000:<:0053:  8: Request(3): GetWindowAttributes window=0x02200001
000:<:0054:  8: Request(14): GetGeometry drawable=0x02200001
000:>:0053:44: Reply to GetWindowAttributes: backing-store=NotUseful(0x00) visual=0x00000021 class=InputOutput(0x0001)
                                             bit-gravity=Forget(0x00) win-gravity=NorthWest(0x01) backing-planes=0xffffffff
                                             backing-pixel=0x00000000 save-under=false(0x00) map-is-installed=true(0x01)
                                             map-state=Viewable(0x02) override-redirect=false(0x00) colormap=0x00000020
                                             all-event-masks=PointerMotion,Exposure,StructureNotify,FocusChange,PropertyChange
                                             your-event-mask=PointerMotion,Exposure do-not-propagate-mask=0 unused=0x0000
000:>:0054:32: Reply to GetGeometry: depth=0x18 root=0x00000170 x=0 y=0 width=1920 height=1080 border-width=0
001:<:000c: 12: Request(98): QueryExtension name='DRI2'
001:>:000c:32: Reply to QueryExtension: present=true(0x01) major-opcode=151 first-event=101 first-error=0
001:<:000d: 32: DRI2-Request(151,8): SwapBuffers drawable=0x02200001 target_msc_hi=0 target_msc_lo=0
                                     divisor_hi=0 divisor_lo=0 remainder_hi=0 remainder_lo=0
001:>:000d: Event DRI2-BufferSwapComplete(101) drawable=0x00000002 ust_hi=35651585 ust_lo=0 msc_hi=0 msc_lo=0 sbc_hi=0 sbc_lo=1

Here the DRI2 code from the Mali blob tries to swap buffers (with the
hope that the allocated BackLeft would go to front)

001:>:000d:32: Reply to SwapBuffers: swap_hi=0 swap_lo=4096
000:<:0055:  8: DRI2-Request(151,3): CreateDrawable drawable=0x02200001
000:<:0056: 16: DRI2-Request(151,5): GetBuffers drawable=0x02200001 attachments={attachment=BackLeft(0x00000001)};
000:>:0056:52: Reply to GetBuffers: width=1920 height=1080 buffers={attachment=BackLeft(0x00000001)
                                    name=0x00000159 pitch=7680 cpp=4 flags=0x00000000};

And requests for the new BackLeft DRI2 buffer.

000:<:0057:  4: Request(43): GetInputFocus
000:>:0057:32: Reply to GetInputFocus: revert-to=PointerRoot(0x01) focus=0x02200001
000:<:0058:  8: Request(3): GetWindowAttributes window=0x02200001
000:<:0059:  8: Request(14): GetGeometry drawable=0x02200001
000:>:0058:44: Reply to GetWindowAttributes: backing-store=NotUseful(0x00) visual=0x00000021 class=InputOutput(0x0001)
                                             bit-gravity=Forget(0x00) win-gravity=NorthWest(0x01) backing-planes=0xffffffff
                                             backing-pixel=0x00000000 save-under=false(0x00) map-is-installed=true(0x01)
                                             map-state=Viewable(0x02) override-redirect=false(0x00) colormap=0x00000020
                                             all-event-masks=PointerMotion,Exposure,StructureNotify,FocusChange,PropertyChange
                                             your-event-mask=PointerMotion,Exposure do-not-propagate-mask=0 unused=0x0000
000:>:0059:32: Reply to GetGeometry: depth=0x18 root=0x00000170 x=0 y=0 width=1920 height=1080 border-width=0
000:<:005a:  8: Request(3): GetWindowAttributes window=0x02200001
000:<:005b:  8: Request(14): GetGeometry drawable=0x02200001
000:>:005a:44: Reply to GetWindowAttributes: backing-store=NotUseful(0x00) visual=0x00000021 class=InputOutput(0x0001)
                                             bit-gravity=Forget(0x00) win-gravity=NorthWest(0x01) backing-planes=0xffffffff
                                             backing-pixel=0x00000000 save-under=false(0x00) map-is-installed=true(0x01)
                                             map-state=Viewable(0x02) override-redirect=false(0x00) colormap=0x00000020
                                             all-event-masks=PointerMotion,Exposure,StructureNotify,FocusChange,PropertyChange
                                             your-event-mask=PointerMotion,Exposure do-not-propagate-mask=0 unused=0x0000
000:>:005b:32: Reply to GetGeometry: depth=0x18 root=0x00000170 x=0 y=0 width=1920 height=1080 border-width=0
001:<:000e: 32: DRI2-Request(151,8): SwapBuffers drawable=0x02200001 target_msc_hi=0 target_msc_lo=0
                                     divisor_hi=0 divisor_lo=0 remainder_hi=0 remainder_lo=0
001:>:000e: Event DRI2-BufferSwapComplete(101) drawable=0x00000002 ust_hi=35651585 ust_lo=0 msc_hi=0 msc_lo=0 sbc_hi=0 sbc_lo=2

And here it is simply swapping the buffers.

001:>:000e:32: Reply to SwapBuffers: swap_hi=0 swap_lo=4096
000:<:005c:  8: Request(3): GetWindowAttributes window=0x02200001
000:<:005d:  8: Request(14): GetGeometry drawable=0x02200001
000:>:005c:44: Reply to GetWindowAttributes: backing-store=NotUseful(0x00) visual=0x00000021 class=InputOutput(0x0001)
                                             bit-gravity=Forget(0x00) win-gravity=NorthWest(0x01) backing-planes=0xffffffff
                                             backing-pixel=0x00000000 save-under=false(0x00) map-is-installed=true(0x01)
                                             map-state=Viewable(0x02) override-redirect=false(0x00) colormap=0x00000020
                                             all-event-masks=PointerMotion,Exposure,StructureNotify,FocusChange,PropertyChange
                                             your-event-mask=PointerMotion,Exposure do-not-propagate-mask=0 unused=0x0000
000:>:005d:32: Reply to GetGeometry: depth=0x18 root=0x00000170 x=0 y=0 width=1920 height=1080 border-width=0

And now it is polling for the change of window geometry. The same
"SwapBuffers -> GetGeometry -> SwapBuffers" pattern keeps repeating.
Signed-off-by: default avatarSiarhei Siamashka <siarhei.siamashka@gmail.com>
      d147e25f
  5. 30 Mar, 2013 1 commit
  7. 28 Mar, 2013 1 commit
  9. 26 Mar, 2013 3 commits
  11. 22 Mar, 2013 2 commits
  13. 21 Mar, 2013 1 commit
    • Siarhei Siamashka's avatar
      G2D: accelerate CopyArea between different pixmaps in framebuffer · cc1b1410
      Siarhei Siamashka authored 
      

Now source and destination pixmaps don't need to be the same for
using G2D acceleration (as long as both of them are allocated in
the framebuffer). This allows using G2D to copy pixels from DRI2
buffers to the framebuffer on the fallback path (when the window
of an OpenGL ES application is partially overlapped by some other
windows). Though it only works when composite extension is
disabled, for example by adding the following to xorg.conf:

    Section "Extensions"
        Option "Composite" "Disable"
    EndSection

If composite extension is enabled, windows have backing pixmaps, and
we have a longer chain of copies:

   DRI2 buffer -> backing pixmap -> framebuffer

Because backing pixmap is not allocated in a physically contiguous
memory, it can't be copied using G2D yet.
Signed-off-by: default avatarSiarhei Siamashka <siarhei.siamashka@gmail.com>
      cc1b1410
  15. 19 Mar, 2013 1 commit
  17. 18 Mar, 2013 1 commit
    • Siarhei Siamashka's avatar
      G2D: Hardware acceleration for XCopyArea (initially 32bpp only) · ecfeb4aa
      Siarhei Siamashka authored 
      

Wrap CreateGC function to add a hook for CopyArea operation, which
can be accelerated using G2D for the buffers inside of the visible
part of the framebuffer. In the future we may try to also ensure
that DRI2 buffers are copied using G2D instead of CPU in the case
if we hit the fallback path and can't avoid this copy.

Benchmark using "x11perf -scroll500 -copywinwin500":

=== ShadowFB (software rendering) ===

   3000 reps @   2.0308 msec (   492.0/sec): Scroll 500x500 pixels
   3000 reps @   1.9741 msec (   507.0/sec): Scroll 500x500 pixels
   3000 reps @   1.9826 msec (   504.0/sec): Scroll 500x500 pixels
   3000 reps @   1.9830 msec (   504.0/sec): Scroll 500x500 pixels
   3000 reps @   1.9965 msec (   501.0/sec): Scroll 500x500 pixels
  15000 trep @   1.9934 msec (   502.0/sec): Scroll 500x500 pixels

   1600 reps @   3.3054 msec (   303.0/sec): Copy 500x500 from window to window
   1600 reps @   3.3179 msec (   301.0/sec): Copy 500x500 from window to window
   1600 reps @   3.2263 msec (   310.0/sec): Copy 500x500 from window to window
   1600 reps @   3.2491 msec (   308.0/sec): Copy 500x500 from window to window
   1600 reps @   3.2357 msec (   309.0/sec): Copy 500x500 from window to window
   8000 trep @   3.2669 msec (   306.0/sec): Copy 500x500 from window to window

=== G2D (hardware acceleration) ===

   3000 reps @   2.1949 msec (   456.0/sec): Scroll 500x500 pixels
   3000 reps @   2.1929 msec (   456.0/sec): Scroll 500x500 pixels
   3000 reps @   2.1923 msec (   456.0/sec): Scroll 500x500 pixels
   3000 reps @   2.1889 msec (   457.0/sec): Scroll 500x500 pixels
   3000 reps @   2.1941 msec (   456.0/sec): Scroll 500x500 pixels
  15000 trep @   2.1926 msec (   456.0/sec): Scroll 500x500 pixels

   2800 reps @   1.8114 msec (   552.0/sec): Copy 500x500 from window to window
   2800 reps @   1.8103 msec (   552.0/sec): Copy 500x500 from window to window
   2800 reps @   1.8160 msec (   551.0/sec): Copy 500x500 from window to window
   2800 reps @   1.8099 msec (   553.0/sec): Copy 500x500 from window to window
   2800 reps @   1.8126 msec (   552.0/sec): Copy 500x500 from window to window
  14000 trep @   1.8120 msec (   552.0/sec): Copy 500x500 from window to window

CPU usage remains low when running this test with G2D acceleration enabled.
Signed-off-by: default avatarSiarhei Siamashka <siarhei.siamashka@gmail.com>
      ecfeb4aa
  19. 17 Mar, 2013 2 commits
  21. 16 Mar, 2013 1 commit
  23. 15 Mar, 2013 2 commits
  25. 14 Mar, 2013 2 commits
  27. 13 Mar, 2013 3 commits
  29. 12 Mar, 2013 1 commit
  31. 11 Mar, 2013 1 commit
  33. 07 Mar, 2013 1 commit
  35. 23 Feb, 2013 1 commit
  37. 18 Feb, 2013 2 commits
    • Siarhei Siamashka's avatar
      Add support for hardware ARGB cursors up to 32x32 size · 06ae9b68
      Siarhei Siamashka authored 
      

Actually they are converted to 32x32 with 256 color palette. In the
case if we have more than 256 unique colors, the color components
of the pixels are reduced from 8-bit to 7-bit, then to 6-bit if
necessary and so on (until we reduce the number of unique colors
so that they can fit the palette). In the worst case we may
theoretically end up with just 2 bits per A, R, G and B channels,
but in practice 7 or 6 bits seem to be enough.
Signed-off-by: default avatarSiarhei Siamashka <siarhei.siamashka@gmail.com>
      06ae9b68
    • Siarhei Siamashka's avatar
      Disable hardware layers when software ARGB cursor is used · 0ab0a9e8
      Siarhei Siamashka authored 
      

The modern desktops may use ARGB cursors. As the current
sunxi display controller support code can't handle this
type of cursor yet, the X server fallbacks to a software
cursor which is not visible under layers and ruining user
experience.

This patch adds empty implementations for "UseHWCursorARGB"
and "LoadCursorARGB" functions which just return error for
now (so that the X server still fallbacks to software cursor).
However we also introduce callback functions responsible for
notifying the DRI2 code about enabling/disabling the use of
hardware cursor. So that now hardware overlays are disabled
when switching to software cursor and re-enabled again when
switching back to hardware cursor.
Signed-off-by: default avatarSiarhei Siamashka <siarhei.siamashka@gmail.com>
      0ab0a9e8
  39. 12 Feb, 2013 1 commit
  41. 01 Feb, 2013 1 commit
  43. 24 Jan, 2013 5 commits
  45. 20 Jan, 2013 4 commits
    • Siarhei Siamashka's avatar
      Support for sunxi hardware cursor · 64d51c0f
      Siarhei Siamashka authored 
      Hardware cursor is necessary because it is also visible on
top of sunxi disp layers, while software cursor is not.

FIXME: there is one minor problem with negative cursor
positions. The hardware does not support them, so such
positions are just set to 0 for now. In the future this
can be solved better by changing the cursor picture and
showing only the parts which are visible on screen.
      64d51c0f
    • Siarhei Siamashka's avatar
      Added supplementary wrapper functions for sunxi display controller ioctls · bf7c7a6c
      Siarhei Siamashka authored 
      Note: the header file "sunxi_disp_ioctl.h" is GPL licensed. So until
it is gets a MIT/X11 replacement, the DDX driver is GPL licensed as
a whole. The individual source files still have their own license.
Also in order to avoid any possible confusion, the MIT/X11 license
header from COPYING has been added to "fbdev.c" and "fbdev_priv.h".
      bf7c7a6c
    • Siarhei Siamashka's avatar
      Move driver private data into a separate header file · b0813ddf
      Siarhei Siamashka authored 
      It is going to be included by multiple different source files.
      b0813ddf
    • Siarhei Siamashka's avatar
      Rename "fbdev" -> "sunxifb" and update man page · 703aea28
      Siarhei Siamashka authored 
      As there is no way for the hardware specific bits to be accepted
in xf86-video-fbdev, we need a new driver with its own name.
      703aea28
  47. 25 Sep, 2012 1 commit