Commit Graph

445 Commits

Author SHA1 Message Date
Louis Huemiller
5d86b53dc1 HWC Stres Test Enhancments
+ YV12 width/height divisable by 2
  + Row width determined by getStride()
  + Misc comment and whitespace fixes
  + printf changed to testPrintI

Change-Id: I27551141bafbfb258eb1b6b7f63c5295b1f9501a
2010-12-14 10:31:57 -08:00
Louis Huemiller
b526594d84 Hardware Composer Stress Test Enhancements
Enhancements include:
  + Change default delay after each set operation
    from 0.1 to 0.0 seconds.  This significantly
    increases the default rate at which operations
    are performed.

  + Make sourceCrop to displayFrame scale
    factor of 1.0 a frequent occurance.

  + Enable use of RGB888 and YV12 graphic formats.

  + Add initCheck() call after creation of a
    GraphicBuffer, to confirm it was created
    without any errors.

Change-Id: I1606dff3924a23bdd5cb27ba302bebb4f200d768
2010-12-13 13:19:13 -08:00
Mathias Agopian
695b66f082 fix build.
Change-Id: I49262ac143da297efcaec4eed04052274c73b7f8
2010-12-08 17:44:07 -08:00
Mathias Agopian
2fa3af5b43 am 054ae782: am e1752f41: Merge "fix [3258603] \'Quadrant\' benchmark crashes inside eglQueryContext on GB/Crespo" into gingerbread
* commit '054ae78245fe9b4c14bbc372ae77cd0a977ded73':
  fix [3258603] 'Quadrant' benchmark crashes inside eglQueryContext on GB/Crespo
2010-12-08 17:20:39 -08:00
Mathias Agopian
18c35abd39 am e1752f41: Merge "fix [3258603] \'Quadrant\' benchmark crashes inside eglQueryContext on GB/Crespo" into gingerbread
* commit 'e1752f41d34249c891564889b07963262b1bd8bf':
  fix [3258603] 'Quadrant' benchmark crashes inside eglQueryContext on GB/Crespo
2010-12-08 16:20:23 -08:00
Mathias Agopian
c3ce880972 fix [3258603] 'Quadrant' benchmark crashes inside eglQueryContext on GB/Crespo
eglQueryContext(..., EGL_CONFIG_ID, ...) is dereferencing an uninitialized pointer
due to a typo.

Change-Id: I100addf3150f19cb6dfbce9987fb5239dd240878
2010-12-08 15:34:02 -08:00
Dan Bornstein
47c9507c91 Merge "Fix another set of out-of-date headers." 2010-12-07 16:39:45 -08:00
Dan Bornstein
4bb407d47e Fix another set of out-of-date headers.
This time in XML files.

Change-Id: Iebe36810c3a2d7a6ff354f43808b1b8649b85637
2010-12-07 16:27:55 -08:00
Dan Bornstein
a5651f24b8 Merge "Fix newlines in this file." 2010-12-07 16:27:31 -08:00
Dan Bornstein
22b03897a8 Fix newlines in this file.
They were CRLF instead of just LF.

Change-Id: I2d06d19b4d8ca969527d93bd996103c161e48a79
2010-12-07 16:23:05 -08:00
Mathias Agopian
b0c48b8993 am 0484a673: am daef0a04: Merge "[3229973, 3247470, ...] set EGLNativeWindowSurface\'s format in EGL" into gingerbread
* commit '0484a673c5839d2be9e615e52c6a45febdf1127b':
  [3229973, 3247470, ...] set EGLNativeWindowSurface's format in EGL
2010-12-06 18:25:58 -08:00
Mathias Agopian
d6a9c7d2db am daef0a04: Merge "[3229973, 3247470, ...] set EGLNativeWindowSurface\'s format in EGL" into gingerbread
* commit 'daef0a040de7a9825bdbcba7b2eae66195e3a95f':
  [3229973, 3247470, ...] set EGLNativeWindowSurface's format in EGL
2010-12-06 18:22:57 -08:00
Mathias Agopian
644bb2a29e [3229973, 3247470, ...] set EGLNativeWindowSurface's format in EGL
(there are multiple bugs this should fix)

we now use the EGL_NATIVE_VISUAL_ID of a config to set
the ANativeWindow's format from eglCreateWindowSurface(),
this guarantees that the surface's format will match
whatever EGLConfig the user chose.

this should fix all current and future config/surface format
mismatch and allow users to easily select 32-bits surfaces.

Change-Id: I3835d0eb70c75eeecded3c3509a0a8207554c98b
2010-12-03 14:15:40 -08:00
Louis Huemiller
365b2c66a2 Hardware Composer Stress Test
Change-Id: Ief08b22dacd052483f10754b3ab61aa09ca28b17
2010-11-29 14:08:57 -08:00
Mathias Agopian
a8422af09e am 17fd21fc: Merge "DO NOT MERGE. workaround [3177481] eglGetProcAddress() returns the wrong pointer for some GL extensions" into gingerbread
* commit '17fd21fc0638d8734867b6f189efbabb227f3ae1':
  DO NOT MERGE. workaround [3177481] eglGetProcAddress() returns the wrong pointer for some GL extensions
2010-11-10 14:15:57 -08:00
Mathias Agopian
70745094cc Merge "DO NOT MERGE. workaround [3177481] eglGetProcAddress() returns the wrong pointer for some GL extensions" into gingerbread 2010-11-10 14:09:17 -08:00
David 'Digit' Turner
ab5426e53d am b02ca27f: am b7a109c0: Merge "Refresh EGL header to work with the NDK" into gingerbread
* commit 'b02ca27f05a47fad75c57d6d8c52ccc5b21a8043':
  Refresh EGL header to work with the NDK
2010-11-09 16:42:27 -08:00
David 'Digit' Turner
2346d0223f am b7a109c0: Merge "Refresh EGL header to work with the NDK" into gingerbread
* commit 'b7a109c02735bb0d9e0bcb8f84e4aa3923701460':
  Refresh EGL header to work with the NDK
2010-11-09 16:25:58 -08:00
Mathias Agopian
744026fed6 DO NOT MERGE. workaround [3177481] eglGetProcAddress() returns the wrong pointer for some GL extensions
We just make sure eglGetProcAddress() will return NULL for
 glEGLImageTargetTexture2DOES
 glEGLImageTargetRenderbufferStorageOES

which is better than returning the address of the wrong implementation.

the correct fix is more involved.

Change-Id: I585a1f40e564f862e5dd382224609ccd069cd3b5
2010-11-09 14:41:13 -08:00
David 'Digit' Turner
dfd0759ff6 Refresh EGL header to work with the NDK
This change allows to use this header with the NDK's standalone toolchain.
For the record, the NDK toolchain defines __ANDROID__ as a compiler built-in
macro, this is however not the case currently for the prebuilt binaries that
are being used by the full Android build system, which otherwise defines ANDROID.

This change allows the header to be used by all toolchains properly. Note however
that we should properly should change our toolchain and sources to provide and
rely on __ANDROID__ instead of ANDROID though.

Change-Id: Iaa1aa1146985b5f24dcf3a83d9ddb9b4b59dc328
2010-11-07 23:36:28 +01:00
Jamie Gennis
b27cb7f5bf Merge "OpenGL ES 1 YUV texturing test" 2010-10-28 15:44:11 -07:00
Jamie Gennis
e6bb73a1cd Merge "YUV texturing test." 2010-10-28 15:44:01 -07:00
Jack Palevich
6132b3703d Trace glUniformXXv and glUniformMatrixXfv calls.
Change-Id: I653021d2174fa1a7343d10b0fd2c29b3e5147be6
2010-10-27 15:37:01 -07:00
Nuno Subtil
ce0a5ca423 OpenGL ES 1 YUV texturing test
Change-Id: I295e5bce16a8ab088b99929f788f2fbbe8492b8d
2010-10-27 15:14:56 -07:00
Michael I. Gold
6ecdb49b25 YUV texturing test.
Change-Id: Ib2a6b7eb2a5fc98ecaace032d6c63d50613aaa0d
2010-10-27 14:51:43 -07:00
Kristian Monsen
72c384e0a1 Fix for simulator build
The code looks correct, but causes an internal compiler error on
simulator builds on mine (an other) gLucid desktops.

Change-Id: Iaf0d8120778ec1c6fc739ef13e6eaaac53e3bcc8
2010-10-27 22:49:34 +01:00
Jack Palevich
8c026c039f Merge "OpenGL tracing." 2010-10-26 18:37:10 -07:00
Jack Palevich
a2dd6cf599 OpenGL tracing.
Enable with:

adb shell setprop debug.egl.trace 1

Change-Id: Icfbc795f5260141510975228e72234e9aab56a85
2010-10-26 15:21:24 -07:00
Mathias Agopian
698346ed59 am d9588e9c: am a7b745ca: Merge "fix eglChooseConfig() for special attributes." into gingerbread 2010-10-26 12:23:46 -07:00
Mathias Agopian
3e98e38510 am a7b745ca: Merge "fix eglChooseConfig() for special attributes." into gingerbread 2010-10-26 12:19:43 -07:00
Mathias Agopian
639716736f fix eglChooseConfig() for special attributes.
affects software renderer (emu) only.

per EGL spec:
  If EGL_MAX_PBUFFER_WIDTH, EGL_MAX_PBUFFER_HEIGHT,
  EGL_MAX_PBUFFER_PIXELS, or EGL_NATIVE_VISUAL_ID are specified
  in attrib list, then they are ignored

Change-Id: I3d5214ce896986da2f11d2eb1889afc41731eaa6
2010-10-25 15:51:24 -07:00
Mathias Agopian
2e6153ee11 am 49af98ed: am aa6df366: Merge "fix [3119687] [Maps][4.6.0] Force close in street view" into gingerbread
Merge commit '49af98ed877e8a2db5a6b5be811c798e6f73dff7'

* commit '49af98ed877e8a2db5a6b5be811c798e6f73dff7':
  fix [3119687] [Maps][4.6.0] Force close in street view
2010-10-21 17:23:24 -07:00
Mathias Agopian
613493c424 am aa6df366: Merge "fix [3119687] [Maps][4.6.0] Force close in street view" into gingerbread
Merge commit 'aa6df3669deab2a804c6f8cc1ea3fd5af0a28685' into gingerbread-plus-aosp

* commit 'aa6df3669deab2a804c6f8cc1ea3fd5af0a28685':
  fix [3119687] [Maps][4.6.0] Force close in street view
2010-10-21 17:07:15 -07:00
Mathias Agopian
6af358e0a7 fix [3119687] [Maps][4.6.0] Force close in street view
The EGLConfig attributes MUST be sorted, because they're used in a binary search.
A recent change introduced a bug where 2 of the configs had improperly
sorted attributes.

Change-Id: I1ac53e4463d62f27125ca9f82ed946e6c98ddba0
2010-10-21 15:58:25 -07:00
Mathias Agopian
e3c5974719 am 63ac8f53: am 2b2da526: Merge "Add a test application to dump all the EGLConfig available" into gingerbread
Merge commit '63ac8f5377833303a7f8c4d60f0d74daf903e6ea'

* commit '63ac8f5377833303a7f8c4d60f0d74daf903e6ea':
  Add a test application to dump all the EGLConfig available
2010-10-21 13:48:50 -07:00
Mathias Agopian
f6715ba6de am 6d90e89a: am a0795fc9: Merge "software renderer: fix [3117351] GL driver doesn\'t set the EGL_NATIVE_VISUAL_ID" into gingerbread
Merge commit '6d90e89afa47293cf5fe18f512cce43d2fa6df82'

* commit '6d90e89afa47293cf5fe18f512cce43d2fa6df82':
  software renderer: fix [3117351] GL driver doesn't set the EGL_NATIVE_VISUAL_ID
2010-10-21 13:47:59 -07:00
Mathias Agopian
41e5645b1a am add9db83: am f4cd9be0: Merge "fix [2236865] STOPSHIP: bump soft gl and egl minor version number" into gingerbread
Merge commit 'add9db83fdbf939f0610f8b357788fd3e8ad9ab5'

* commit 'add9db83fdbf939f0610f8b357788fd3e8ad9ab5':
  fix [2236865] STOPSHIP: bump soft gl and egl minor version number
2010-10-21 10:47:32 -07:00
Mathias Agopian
483d824271 am 2b2da526: Merge "Add a test application to dump all the EGLConfig available" into gingerbread
Merge commit '2b2da52608303b149d22418865e67c8030c70e73' into gingerbread-plus-aosp

* commit '2b2da52608303b149d22418865e67c8030c70e73':
  Add a test application to dump all the EGLConfig available
2010-10-20 21:28:34 -07:00
Mathias Agopian
a6da12c071 Merge "Add a test application to dump all the EGLConfig available" into gingerbread 2010-10-20 20:05:28 -07:00
Mathias Agopian
3f9f60fa42 am a0795fc9: Merge "software renderer: fix [3117351] GL driver doesn\'t set the EGL_NATIVE_VISUAL_ID" into gingerbread
Merge commit 'a0795fc9bdf97703780cbf0ec47c06eccb642e72' into gingerbread-plus-aosp

* commit 'a0795fc9bdf97703780cbf0ec47c06eccb642e72':
  software renderer: fix [3117351] GL driver doesn't set the EGL_NATIVE_VISUAL_ID
2010-10-20 18:22:17 -07:00
Mathias Agopian
d8e5ceb963 software renderer: fix [3117351] GL driver doesn't set the EGL_NATIVE_VISUAL_ID
Change-Id: I371bc0abfee6065247bc6f04b0e0451645eb2dbf
2010-10-20 17:21:43 -07:00
Mathias Agopian
16a3e39cc0 Add a test application to dump all the EGLConfig available
Change-Id: Iddf27f2a9dd51733f6b90081a523ca37bfe310fa
2010-10-20 17:10:47 -07:00
Mathias Agopian
309be7131f am f4cd9be0: Merge "fix [2236865] STOPSHIP: bump soft gl and egl minor version number" into gingerbread
Merge commit 'f4cd9be0390135269e22e42fd2e1fe61fdcf9863' into gingerbread-plus-aosp

* commit 'f4cd9be0390135269e22e42fd2e1fe61fdcf9863':
  fix [2236865] STOPSHIP: bump soft gl and egl minor version number
2010-10-19 18:26:28 -07:00
Mathias Agopian
141550bc18 fix [2236865] STOPSHIP: bump soft gl and egl minor version number
Change-Id: Ibf46b18d7493348e189f2c93e969a6cac61f68b5
2010-10-19 14:47:08 -07:00
Ari Hirvonen
551dc26ebd libagl: eglSwapInterval fix
SW egl returned always EGL_FALSE even when the display was valid.
This made calls to eglSwapInterval to fail on HW since the wrapper
fails if either SW or HW egl fails.

Letting SW eglSwapInterval to return true when display is valid is
ok since spec says that interval parameter is silently clamped.

Change-Id: Ib4aa1dbfccf510a0f36621796e79a4d46852ee6a
2010-10-06 16:44:37 -07:00
Jim Shuma
257d2bbcbf Initial checkin of a framerate test app
Measures the time to do nothing but clear the screen in an
OpenGL Java app. Spews frame time statistics to the log.

Control long frame time by setting debug.longframe_ms;
e.g., adb shell setprop debug.longframe_ms 33

Change-Id: I435a3cb170048349e00417c44db42121c68744b2
2010-09-28 14:59:20 -07:00
Mathias Agopian
30faafb7d3 am 4eb1ad5e: am 524a6d8e: Merge "better fix for [3028370] GL get error should return a valid error if no context is bound." into gingerbread
Merge commit '4eb1ad5e98c7b36f7ac4ec8c3270f9763afd107e'

* commit '4eb1ad5e98c7b36f7ac4ec8c3270f9763afd107e':
  better fix for [3028370] GL get error should return a valid error if no context is bound.
2010-09-23 21:41:30 -07:00
Mathias Agopian
303f3eb352 am 524a6d8e: Merge "better fix for [3028370] GL get error should return a valid error if no context is bound." into gingerbread
Merge commit '524a6d8e9bb4df9cec0520ac1d9733642b13e76f' into gingerbread-plus-aosp

* commit '524a6d8e9bb4df9cec0520ac1d9733642b13e76f':
  better fix for [3028370] GL get error should return a valid error if no context is bound.
2010-09-23 18:17:08 -07:00
Mathias Agopian
6f0871222f better fix for [3028370] GL get error should return a valid error if no context is bound.
it turns out that we cannot return INVALID_OPERATION from glGetError() because the
GL spec says that it must be called in a loop until it returns GL_NO_ERROR.

now, we always return 0 from GL functions called from a thread with no
context bound. This means that glGetError() will return NO_ERROR in this case,
which is better than returning a random value (which could trap the app in a loop).

if this happens in the main thread of a process, we LOG an error message once.

Change-Id: Id59620e675a890286ef62a257c02b06e0fdcaf69
2010-09-23 16:44:48 -07:00
Mathias Agopian
aa22d20554 am 9c4d7d67: am 5c6c5c7a: fix [3028370] GL get error should return a valid error if no context is bound.
Merge commit '9c4d7d677097981a349c718902e29050dad3d59f'

* commit '9c4d7d677097981a349c718902e29050dad3d59f':
  fix [3028370] GL get error should return a valid error if no context is bound.
2010-09-23 13:28:47 -07:00