No idea why this issue hasn't been spotted before. Took several hours to
figure out where exactly wrong memory access happens..
P.S. I really want to switch ImBuf->rect from int* to unsigned char*...
The root of the issue goes to the fact that we only can
read RGB EXR files, but they could be YCbCr or just Luma.
Added support for this two cases.
Note: internally EXR would still be 3 channels, so no
big memory save would happen here, at least yet.
Issue was caused by _wstat returning EOVERFLOW error because
of file size didn't fit into stat structure which was using
long datatype.
The idea of this patch is to use _wstat64 and _stat64 structure
which is capable storing 64bit file sizes.
Made it a typedef for stat structure used by BLI_stat function
in order to make code easier to follow and avoid ifdefs all
over the place.
Additionally solved issue with BLI_exists which was wrongly
returning False in cases destination file is larger then 4GB.
The issue was that we can't assume we support the colorspace from the file. The reported file had an invalid colorspace in fact, which was leading to the segfault in Blender.
Thanks for Sergey Sharybin for the help here.
The mask make sure the conversion only happens in a few areas of the
buffer.
New Functions:
* IMB_buffer_byte_from_float_mask
* IMB_buffer_float_from_float_mask
The functions are an adaptation of their maskless counterparts without accepting different profiles for the input and output buffers.
Review: Sergey Sharybin
There is not much sense to have a whole BLI file just to check SSE2 on CPUs...
So idea is to rename it to more generic "BLI_system", and add to it more system-related
utils, like e.g. an include helper for getpid(), which allows to hide unix/windows
internals from rest of the code...
Reviewers: campbellbarton
Reviewed By: campbellbarton
Differential Revision: https://developer.blender.org/D439
Issue is likely caused by thread-unsafe nature of IMB_freeImBuf
which might lead to race condition in some circumstances.
Now made it thread-safe and from Sebastian's tests seems crash is
gone now, so hopefully the root of the issue is finally nailed down.
Issue was caused by the cache limitor which was removing 4k textures from the
memory when accessing other images.
This is pretty much awful situation and solved by making it so only image sequences
and movies ace cache-guarded.
Could be optimized further so images used by viewport are not being freed, but
that's much more tricky to do..
This is a nice candidature for 'a'.
Basically proxy colorspace didn't work well enough.
It is still a bit weird and mainly:
- Proxies for image sequences are built in the image color space.
- Proxies for movies are built in the movie color space.
This could be unified but would need some work in proxy build
to make it not just pipe frames from one FFmpeg context to
another but also apply OCIO on it.
Was a regression since avg_frame_rate changes.
Didn't find reliable way to get stream duration which will
work with both FFmpeg and Libav so added some freaking black
magic to distinguish one from another.
don't use extension checking preference for low level loading functions,
otherwise a user preference may impact a script loading a file for example.
looked into making this an argument and passing it in, but theres no
real advantage and its simpler just to remove.
r_frame_rate is a guessed number defined as "the least common multiple of
all framerates in the stream". It has been deprecated and removed in new
Libav releases.
Use avg_frame_rate instead, which is the average framerate of the
stream.
CODEC_ID_* have been replaced with AV_CODEC_ID_* in new libavcodec
versions.
Update the code to use those new identifiers.
Added a compatibility code to ffmpeg_compat.h
* EXR layers with names like 'Z' without any pass name were not loaded at all
and would break the Combined pass as well.
* EXR pass names longer than 16 characters where writing past the end of the
array and getting invalid names.
- Made them receive number of channels rather than number of planes.
This matches to how ImBuf structure stored planes and channels.
- IMB_premultiply_rect_float() was called with channels passed instead
of planes already :S.
