Y4M loader: use custom y4m parser #14
Loading…
Reference in New Issue
No description provided.
Delete Branch "NRK:y4m"
Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?
avoid dependency on liby4m, which has some quality issues and isn't
available on any distros according to repology.
additionally, the new loader now supports:
Fixes: #13
Sample multi-frame y4m: https://samples.mplayerhq.hu/yuv4mpeg2/example.y4m.bz2
Sample mono y4m:
ffmpeg -i test/images/icon-64.png -pix_fmt gray8 mono.y4m@ -128,0 +347,5 @@pf = __imlib_GetFrame(im);if (!pf)return LOAD_OOM;pf->frame_count = fcount;pf->loop_count = 0; /* TODO wtf is this thing? */Oops, forgot to fix this up. I'll look into what this is tomorrow :)
31bba57db8to85d3bdc22bLooks good to me.
Is it ready for merging?
Can I squash the two commits?
Yeah, everything works fine as far as I tested. Wanted to remove dependency on libyuv as well, but that'll have to wait for some other day :)
Okay by me.
I just got this:
make[4]: Entering directory '/local/stuff/src/E/git/legacy/imlib2/build/cast-align/src/modules/loaders'
/bin/sh ../../../libtool --tag=CC --tag=disable-static --mode=compile gcc -DHAVE_CONFIG_H -I. -I../../../../../src/modules/loaders -I../../.. -I../../.. -I../../../../../src/lib -D PACKAGE_DATA_DIR="/usr/share/imlib2" -W -Wall -Wcast-align -Wpointer-arith -Wshadow -Wwrite-strings -Wno-unused-parameter -Wmissing-prototypes -Wmissing-declarations -Wstrict-prototypes -Werror -fvisibility=hidden -g -O2 -Wcast-align=strict -fsanitize=undefined -MT y4m_la-loader_y4m.lo -MD -MP -MF .deps/y4m_la-loader_y4m.Tpo -c -o y4m_la-loader_y4m.lo
test -f 'loader_y4m.c' || echo '../../../../../src/modules/loaders/'loader_y4m.clibtool: compile: gcc -DHAVE_CONFIG_H -I. -I../../../../../src/modules/loaders -I../../.. -I../../.. -I../../../../../src/lib -D PACKAGE_DATA_DIR="/usr/share/imlib2" -W -Wall -Wcast-align -Wpointer-arith -Wshadow -Wwrite-strings -Wno-unused-parameter -Wmissing-prototypes -Wmissing-declarations -Wstrict-prototypes -Werror -fvisibility=hidden -g -O2 -Wcast-align=strict -fsanitize=undefined -MT y4m_la-loader_y4m.lo -MD -MP -MF .deps/y4m_la-loader_y4m.Tpo -c ../../../../../src/modules/loaders/loader_y4m.c -fPIC -DPIC -o .libs/y4m_la-loader_y4m.o
../../../../../src/modules/loaders/loader_y4m.c: In function 'y4m_parse_frame':
../../../../../src/modules/loaders/loader_y4m.c:296:48: error: 'sdiv' may be used uninitialized [-Werror=maybe-uninitialized]
296 | res->u_stride = res->v_stride = res->w / sdiv;
| ^
../../../../../src/modules/loaders/loader_y4m.c:233:33: note: 'sdiv' was declared here
233 | ptrdiff_t npixels, sdiv, voff;
| ^~~~
../../../../../src/modules/loaders/loader_y4m.c:295:20: error: 'voff' may be used uninitialized [-Werror=maybe-uninitialized]
295 | res->v = p + voff;
|
^~~~../../../../../src/modules/loaders/loader_y4m.c:233:39: note: 'voff' was declared here
233 | ptrdiff_t npixels, sdiv, voff;
| ^~~~
cc1: all warnings being treated as errors
Looks like bogus warnings and seem to be triggered by some combination of options.
I'd like not to get them though :)
I can fix or you do. Got to go now though.
What version of GCC is that? I don't get that warning (gcc 12.1.1) and that's clearly false positive since the only case those variable aren't initialized is if the color space was mono. (But if the color space was mono, those variables aren't used at all).
I guess you can initialize them to 0 at the top, it should avoid the warnings.
2a314057c4to2cea59b3daSquashed the commits and hopefully the compiler warning should be fixed now too.
Warnings gone, pushed, thanks :)
I'm on gcc 13.1.1. Looks like -fsanitize=undefined triggers it.
@kw Also just noticed, the tests should now allow y4m memory as well. Attached a patch.
Right, forgot, pushed, thanks :)
Pull request closed