77 Commits

Author SHA1 Message Date
Mathias Agopian
022a43b15b to help debugging [2461567] Home screen redraw messed up
log SF's idea of the front buffer in dumpsys.
2010-02-26 18:59:23 -08:00
Mathias Agopian
8c20ca39c1 remove a dependency of surfaceflinger on libskia
libskia was only used for a small part of SkTransform. We now implement
Transform is surfaceflinger directly.
2010-02-22 17:30:40 -08:00
Mathias Agopian
66c77a5b42 Add support for physically rotated displays
This feature is currently controled by a system property.
"ro.sf.hwrotation" can be set to either 90 or 270. It'll cause
SF to rotate the screen by 90 and 270 degres respectively.

That is, if the driver reports 800x480 for instance, and
ro.sf.hwrotation is set to 90, applications will "see" a
480x800 display and will run in portrait.

This is implemented by introducing an extra "display"
transformation in the GraphicPlane.
2010-02-08 15:49:35 -08:00
Mathias Agopian
0e44976aa4 fix [2297155] SurfaceFlinger's freeze-timeout doesn't work
There was bug in the logic that calculated the relative timeout, the start time was
reset each time an event was received, which caused the timeout to never occur if
an application was constantly redrawing.

Now we always check for a timeout when we come back from the waitEvent() and
process the "anti-freeze" if needed, regardless of whether an event was received.
2009-12-01 17:23:28 -08:00
Mathias Agopian
6950e428fe fix [2167050] glTexImage2D code path buggy in SurfaceFlinger
When EGLImage extension is not available, SurfaceFlinger will fallback to using
glTexImage2D and glTexSubImage2D instead, which requires 50% more memory and an
extra copy. However this code path has never been exercised and had some bugs
which this patch fix.

Mainly the scale factor wasn't computed right when falling back on glDrawElements.
We also fallback to this mode of operation if a buffer doesn't have the adequate
usage bits for EGLImage usage.

This changes only code that is currently not executed. Some refactoring was needed to
keep the change clean. This doesn't change anything functionaly.
2009-10-06 17:00:25 -07:00
Mathias Agopian
0c4cec7e4d Attempt to fix [2152536] ANR in browser
The ANR is caused by SurfaceFlinger waiting for buffers of a removed surface to become availlable.
When it is removed from the current list, a Surface is marked as NO_INIT, which causes SF to return
immediately in the above case. For some reason, the surface here wasn't marked as NO_INIT.

This change makes the code more robust by always (irregadless or errors) setting the NO_INIT status
in all code paths where a surface is removed from the list.

Additionaly added more information in the logs, should this happen again.
2009-10-02 18:12:30 -07:00
Mathias Agopian
98a9c56acf attempt to work around [2155085, 2150246] stuck in closeTransaction() 2009-09-30 15:15:07 -07:00
Mathias Agopian
8851617aea remove unused code 2009-09-29 22:32:36 -07:00
Mathias Agopian
81384bf927 don't emit GL commands when sf is in freeze mode
We were emitting GL commands, calling composition complete and releasing clients
without ever calling eglSwapBuffers(), which is completely wrong on non-direct
renders. This could cause transient drawing artifacts when unfreezing the
screen (upon orientaion change for instance) and could also block the clients
for ever as they are waiting for their previous buffer to be rendered.
2009-09-27 22:47:32 -07:00
Mathias Agopian
f2d28b7485 minor SurfaceFlinger code cleanup and remove unnecessary tests 2009-09-24 14:57:26 -07:00
Mathias Agopian
cc934763c3 turn dithering off if it's not needed 2009-09-23 19:16:27 -07:00
Mathias Agopian
2e4b68d57b fix [2133214] STOPSHIP: revert I4a06bb4f: workaround for [2113743] Sholes: frozen then runtime restart going to list view
Revert "workaround for [2113743] Sholes: frozen then runtime restart going to list view"

This reverts commit 4a06bb4f3355b0ef2b76aa883704da9d154c44ae.
2009-09-23 16:59:57 -07:00
Mathias Agopian
4a06bb4f33 workaround for [2113743] Sholes: frozen then runtime restart going to list view 2009-09-20 17:08:45 -07:00
Mathias Agopian
b1a1874625 Android side of the fix for [2121211] Sholes has stutter during animations
a new method, compostionComplete() is added to the framebuffer hal, it is used by surfaceflinger to signal the driver that the composition is complete, BEFORE it releases its client. This gives a chance to the driver to
2009-09-17 16:18:16 -07:00
Mathias Agopian
bcef9ac35d add basic time stats for surfaces lock time 2009-09-17 01:35:28 -07:00
Mathias Agopian
53973fda84 fix [2117464] SF can crash when calling dumpsys 2009-09-14 15:59:16 -07:00
Mathias Agopian
248b5bd51e fix [2112575] stuck on DequeueCondition for a surface that doesn't exist anymore
this also fixes part of [2111536] Device is soft rebooted after ending the call through voice dialer
2009-09-10 19:41:18 -07:00
Mathias Agopian
9779b221e9 fix [2068105] implement queueBuffer/lockBuffer/dequeueBuffer properly
Rewrote SurfaceFlinger's buffer management from the ground-up.
The design now support an arbitrary number of buffers per surface, however the current implementation is limited to four. Currently only 2 buffers are used in practice.

The main new feature is to be able to dequeue all buffers at once (very important when there are only two). 

A client can dequeue all buffers until there are none available, it can lock all buffers except the last one that is used for composition. The client will block then, until a new buffer is enqueued.

The current implementation requires that buffers are locked in the same order they are dequeued and enqueued in the same order they are locked. Only one buffer can be locked at a time.

eg. Allowed sequence:   DQ, DQ, LOCK, Q, LOCK, Q
eg. Forbidden sequence: DQ, DQ, LOCK, LOCK, Q, Q
2009-09-07 16:32:45 -07:00
Mathias Agopian
ffae4fcc78 attempt to fix [2099362] Possible SurfaceFlinger crash 2009-09-04 19:50:23 -07:00
Mathias Agopian
a8d49178f9 dumpsys SurfaceFlinger will now always dump SF's state, even if SF is deadlocked
(in this case the state is dumped without the proper locks held which could result to a crash)

in addition, the last transaction and swap times are printed to the dump as well as the time spent
*currently* in these function. For instance, if SF is unresponsive because eglSwapBuffers() is stuck,
this will show up here.
2009-08-26 16:55:50 -07:00
Mathias Agopian
18b6b49ea5 fix a bug that caused the PixelFormat viewed by Surface to be wrong.
what happened is that the efective pixel format is calculated by SF but Surface nevew had access to it directly.
in particular this caused query(FORMAT) to return the requested format instead of the effective format.
2009-08-19 17:46:26 -07:00
Mathias Agopian
33863dd9e1 resolved conflicts for merge of ac38dfc5 to master 2009-07-28 14:20:21 -07:00
Mathias Agopian
12cedff50c fix [2017532] Partial Update leaves residual image. 2009-07-28 10:57:27 -07:00
Mathias Agopian
07e456d871 am 4d2dbebf: fix for [1885684] E/SurfaceFlinger( 60): not enough memory for layer bitmap size=4294938624
Merge commit '4d2dbebf3d08209f751585d8cc367369e2f6e32f'

* commit '4d2dbebf3d08209f751585d8cc367369e2f6e32f':
  fix for [1885684] E/SurfaceFlinger( 60): not enough memory for layer bitmap size=4294938624
2009-07-09 20:05:11 -07:00
Mathias Agopian
4d2dbebf3d fix for [1885684] E/SurfaceFlinger( 60): not enough memory for layer bitmap size=4294938624 2009-07-09 18:16:43 -07:00
Mathias Agopian
d763b5d919 get rid of references to MemoryDealer in SurfaceFlinger 2009-07-02 18:50:51 -07:00
Mathias Agopian
359140c171 free gralloc buffers as soon as possible (when a surface is not visible any longer), client who have the buffers still mapped won't crash, btu may see garbage data 2009-07-02 18:45:29 -07:00
Mathias Agopian
ecfa7ccd4f fix a bug that caused artifacts when SWAP_RECTANGLE was enabled 2009-06-29 18:49:56 -07:00
Mathias Agopian
0aed7e97a3 fix [1947273] the DimLayer causes the whole screen to update during transactions 2009-06-29 14:44:11 -07:00
Mathias Agopian
2f157b1919 fix [1947273] the DimLayer causes the whole screen to update during transactions 2009-06-28 02:54:16 -07:00
Mathias Agopian
8c9687ae29 fix an update bug with SHOW_UPDATE debug feature. Fix a problem with the debug binder codes too 2009-06-26 19:06:36 -07:00
Mathias Agopian
6edf5af578 fix a memory corruption where a SF Client could be used after it's been destroyed 2009-06-19 17:00:27 -07:00
Mathias Agopian
c8fb5b1979 release the last reference to surfaces explicitely instead of letting it go implicitely when the message is destroyed 2009-06-19 16:24:02 -07:00
Mathias Agopian
151e859e0f new Permission class used to improve permission checks speed (by caching results) 2009-06-15 21:56:51 -07:00
Mathias Agopian
2d5ee25556 fix a bunch of problems with destroying surfaces.
now, all destruction path, go through the purgatory which is emptied when ~ISurface is called, but we also make sure to remove the surface from the current list from there (in case a client forgot to request the destruction explicitely).
2009-06-04 18:46:21 -07:00
Mathias Agopian
947f4f4d38 merge master to master_gl 2009-05-22 02:16:08 -07:00
Android (Google) Code Review
e755d64977 am 69a6c3eb: Merge change 2292 into donut
Merge commit '69a6c3ebd46a8de0e896747757f364322c12608d'

* commit '69a6c3ebd46a8de0e896747757f364322c12608d':
  split boot animation out of SurfaceFlinger
2009-05-21 19:36:53 -07:00
Mathias Agopian
627e7b50be split boot animation out of SurfaceFlinger
Conflicts:

	data/etc/platform.xml
2009-05-21 19:21:59 -07:00
Mathias Agopian
0795272aa2 move libbinder's header files under includes/binder 2009-05-20 12:55:03 -07:00
Mathias Agopian
6158b1bf03 Region now has its own implementation instead of relying on SkRegion, which allows us to break libui's dependency on libcorecg. 2009-05-17 23:34:16 -07:00
Mathias Agopian
2e20bffbab created an new EGL extension called ANDROID_swap_rectangle
ANDROID_swap_rectangle allows to specify the rectangle affected by eglSwapBuffers(), anything outside of this rectangle is unchanged. in particular EGL_BUFFER_DESTROYED only applies to that rectangle. This extension as well as EGL_BUFFER_PRESERVED allow major optimizations on surfaceflinger, which can redraw only the dirty area during compositing.

However, ANDROID_swap_rectangle allows further optimizations in EGL by reducing the amount of copy-back needed. ANDROID_swap_rectangle is particularily important for software implementations.
2009-05-07 15:07:33 -07:00
Mathias Agopian
dff8e58d47 update surfaceflinger, libui and libagl to the new gralloc api
- Currently the lock/unlock path is naive and is done for each drawing operation (glDrawElements and glDrawArrays). this should be improved eventually.
- factor all the lock/unlock code in SurfaceBuffer.
- fixed "showupdate" so it works even when we don't have preserving eglSwapBuffers().
- improved the situation with the dirty-region and fixed a problem that caused GL apps to not update.
- make use of LightRefBase() where needed, instead of duplicating its implementation
- add LightRefBase::getStrongCount()
- renamed EGLNativeWindowSurface.cpp to FramebufferNativeWindow.cpp

- disabled copybits test, since it clashes with the new gralloc api

- Camera/Video will be fixed later when we rework the overlay apis
2009-05-04 14:17:04 -07:00
Android (Google) Code Review
a73b41ae60 am 0124a15: Merge change 626 into donut
Merge commit '0124a150157d69a9f0ea109be2390afb5399b8a7'

* commit '0124a150157d69a9f0ea109be2390afb5399b8a7':
  fix 1803886 android15 Translucent GLSurfaceView demo does not display properly
2009-04-29 13:35:37 -07:00
Mathias Agopian
bdbe6024a5 make use of the perfectly fine List.h instead of our own reimplementation of a linked list. 2009-04-28 03:17:50 -07:00
Mathias Agopian
0c6b5f6ec3 fix 1803886 android15 Translucent GLSurfaceView demo does not display properly 2009-04-27 18:50:06 -07:00
Mathias Agopian
81b0aa696a get rid of an old hack to work around a bug around glDeleteTextures() in the adreno drivers 2009-04-24 16:31:11 -07:00
Mathias Agopian
ef07dda442 get rid of the "show fps" debug option 2009-04-24 16:31:11 -07:00
Mathias Agopian
a3aa6c9aa9 Surfaces are now destroyed properly in SurfaceFlinger.
First, the window manager tells us when a surface is no longer needed. At this point, several things happen:
- the surface is removed from the active/visible list
- it is added to a purgatory list, where it waits for all clients to release their reference
- it destroys all data/state that can be spared

Later, when all clients are done, the remains of the Surface are disposed off: it is removed from the purgatory and destroyed.
In particular its gralloc buffers are destroyed at that point (when we're sure nobody is using them anymore).
2009-04-24 16:30:38 -07:00
Mathias Agopian
6ead5d9f14 a brand new MessageQueue for SurfaceFlinger. 2009-04-24 16:22:36 -07:00
Mathias Agopian
6cf0db228c more Surface lifetime management
Surfaces are now destroyed once all references from the clients are gone, but they go through a partial destruction as soon as the window manager requests it.
This last part is still buggy. see comments in SurfaceFlinger::destroySurface()
2009-04-24 15:00:41 -07:00