mirror of
https://github.com/brl/mutter.git
synced 2024-12-24 20:12:06 +00:00
f53fb5e2e0
This allows apps to catch out-of-memory errors when allocating textures. Textures can be pretty huge at times and so it's quite possible for an application to try and allocate more memory than is available. It's also very possible that the application can take some action in response to reduce memory pressure (such as freeing up texture caches perhaps) so we shouldn't just automatically abort like we do for trivial heap allocations. These public functions now take a CoglError argument so applications can catch out of memory errors: cogl_buffer_map cogl_buffer_map_range cogl_buffer_set_data cogl_framebuffer_read_pixels_into_bitmap cogl_pixel_buffer_new cogl_texture_new_from_data cogl_texture_new_from_bitmap Note: we've been quite conservative with how many apis we let throw OOM CoglErrors since we don't really want to put a burdon on developers to be checking for errors with every cogl api call. So long as there is some lower level api for apps to use that let them catch OOM errors for everything necessary that's enough and we don't have to make more convenient apis more awkward to use. The main focus is on bitmaps and texture allocations since they can be particularly large and prone to failing. A new cogl_attribute_buffer_new_with_size() function has been added in case developers need to catch OOM errors when allocating attribute buffers whereby they can first use _buffer_new_with_size() (which doesn't take a CoglError) followed by cogl_buffer_set_data() which will lazily allocate the buffer storage and report OOM errors. Reviewed-by: Neil Roberts <neil@linux.intel.com> (cherry picked from commit f7735e141ad537a253b02afa2a8238f96340b978) Note: since we can't break the API for Cogl 1.x then actually the main purpose of cherry picking this patch is to keep in-line with changes on the master branch so that we can easily cherry-pick patches. All the api changes relating stable apis released on the 1.12 branch have been reverted as part of cherry-picking this patch so this most just applies all the internal plumbing changes that enable us to correctly propagate OOM errors. |
||
|---|---|---|
| .. | ||
| conform | ||
| data | ||
| micro-perf | ||
| Makefile.am | ||
| README | ||
Outline of test categories:
The conform/ tests:
-------------------
These tests should be non-interactive unit-tests that verify a single
feature is behaving as documented. See conform/ADDING_NEW_TESTS for more
details.
Although it may seem a bit awkward; all the tests are built into a
single binary because it makes building the tests *much* faster by avoiding
lots of linking.
Each test has a wrapper script generated though so running the individual tests
should be convenient enough. Running the wrapper script will also print out for
convenience how you could run the test under gdb or valgrind like this for
example:
NOTE: For debugging purposes, you can run this single test as follows:
$ libtool --mode=execute \
gdb --eval-command="b test_cogl_depth_test" \
--args ./test-conformance -p /conform/cogl/test_cogl_depth_test
or:
$ env G_SLICE=always-malloc \
libtool --mode=execute \
valgrind ./test-conformance -p /conform/cogl/test_cogl_depth_test
By default the conformance tests are run offscreen. This makes the tests run
much faster and they also don't interfere with other work you may want to do by
constantly stealing focus. CoglOnscreen framebuffers obviously don't get tested
this way so it's important that the tests also get run onscreen every once in a
while, especially if changes are being made to CoglFramebuffer related code.
Onscreen testing can be enabled by setting COGL_TEST_ONSCREEN=1 in your
environment.
The micro-bench/ tests:
-----------------------
These should be focused performance tests, ideally testing a
single metric. Please never forget that these tests are synthetic and if you
are using them then you understand what metric is being tested. They probably
don't reflect any real world application loads and the intention is that you
use these tests once you have already determined the crux of your problem and
need focused feedback that your changes are indeed improving matters. There is
no exit status requirements for these tests, but they should give clear
feedback as to their performance. If the framerate is the feedback metric, then
the test should forcibly enable FPS debugging.
The data/ directory:
--------------------
This contains optional data (like images) that can be referenced by a test.
Misc notes:
-----------
• All tests should ideally include a detailed description in the source
explaining exactly what the test is for, how the test was designed to work,
and possibly a rationale for the approach taken for testing.
• When running tests under Valgrind, you should follow the instructions
available here:
http://live.gnome.org/Valgrind
and also use the suppression file available inside the data/ directory.