mesa merge requestshttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests2024-01-12T18:38:49Zhttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/26964nir: Fix cast2024-01-12T18:38:49ZMatt Turnernir: Fix cast```
nir: Fix cast
We were wrongly telling `nir_const_value_as_uint()` that `iter` had
`bit_size` bits, but in one case it is explicitly i64. This works on
little endian platforms, but caused the nir_loop_unroll_test.fadd{,_rev}
tests to...```
nir: Fix cast
We were wrongly telling `nir_const_value_as_uint()` that `iter` had
`bit_size` bits, but in one case it is explicitly i64. This works on
little endian platforms, but caused the nir_loop_unroll_test.fadd{,_rev}
tests to fail on big endian platforms.
Bug: https://bugs.gentoo.org/921297
Fixes: 268ad47c111 ("nir/loop_analyze: Handle bit sizes correctly in calculate_iterations")
```
```
nir/tests: Reenable tests that failed on big-endian
These tests were disabled due to the bug fixed in the previous commit.
```Marge BotMarge Bothttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/19051mesa: Fix format transform on big endian platforms2023-01-25T00:33:03ZJakub KulĂkmesa: Fix format transform on big endian platformsTwo code paths in `_mesa_format_convert()` are correct only on little endian platforms where packed and array formats match. However, on big endian (e.g. on SPARC) those produce incorrect result.
Closes: https://gitlab.freedesktop.org/m...Two code paths in `_mesa_format_convert()` are correct only on little endian platforms where packed and array formats match. However, on big endian (e.g. on SPARC) those produce incorrect result.
Closes: https://gitlab.freedesktop.org/mesa/mesa/-/issues/6001Marge BotMarge Bothttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/15556llvmpipe: fix color rendering on big endian.2022-04-11T20:16:41ZJocelyn Falempellvmpipe: fix color rendering on big endian.Some colors were missing/inverted on big endian machine.
because blend_type.length > src_fmt->nr_channels.
In my case, blend_type has 4 channels (rgba) but src_fmt has only 3.
So the from_lsb was wrong by 1, and channels got messe...Some colors were missing/inverted on big endian machine.
because blend_type.length > src_fmt->nr_channels.
In my case, blend_type has 4 channels (rgba) but src_fmt has only 3.
So the from_lsb was wrong by 1, and channels got messed up.
(blue was always 255, green -> red, and blue -> green).
Fixes: #6204
Signed-off-by: Jocelyn Falempe <jfalempe@redhat.com>Marge BotMarge Bothttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/9921ci: Bump the llvmpipe test timeout to 240 seconds2021-03-31T15:02:47ZAdam Jacksonajax@nwnk.netci: Bump the llvmpipe test timeout to 240 secondslp_test_blend occasionally exceeds this on s390x, but we don't really
want to lose what little (hah) big-endian coverage we have.
Related: mesa/mesa#3437lp_test_blend occasionally exceeds this on s390x, but we don't really
want to lose what little (hah) big-endian coverage we have.
Related: mesa/mesa#3437Marge BotMarge Bothttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/8394CI: Build more drivers in the cross-arch jobs2021-02-11T17:30:26ZAdam Jacksonajax@nwnk.netCI: Build more drivers in the cross-arch jobsThis is mostly to get additional -Werror coverage to avoid introducing
unforced ILP32 or big-endian errors. i386 adds lavapipe, r600, nouveau,
zink, and all the classic drivers. ppc64le adds lavapipe and zink, and
also adds -Werror for s...This is mostly to get additional -Werror coverage to avoid introducing
unforced ILP32 or big-endian errors. i386 adds lavapipe, r600, nouveau,
zink, and all the classic drivers. ppc64le adds lavapipe and zink, and
also adds -Werror for symmetry with the other cross builds. s390x just
adds lavapipe for now, zink currently fails in the format-roundtrip
test which isn't totally surprising.Marge BotMarge Bothttps://gitlab.freedesktop.org/mesa/mesa/-/merge_requests/7776util/format: fix value declarations for big endian2021-07-16T20:17:44ZKarol Herbstkherbst@redhat.comutil/format: fix value declarations for big endianBefore we can fix big endian, we kind of have to agree on how pixel data
should be laid out in memory.
For this we have to use packed data for !is_array declared formats and per
element values for is_array ones.
Also we need proper PAC...Before we can fix big endian, we kind of have to agree on how pixel data
should be laid out in memory.
For this we have to use packed data for !is_array declared formats and per
element values for is_array ones.
Also we need proper PACKED_1xn definitions for BE.Marge BotMarge Bot