Am 30.01.2018 um 09:21 schrieb Gert Wollny:
> Although gl_SampleMaskIn is declared as an array in GLSL, it is
> effectively a 32 bit mask on all hardware supported by mesa, so the
> array indexing is ignored (Thanks Glenn Kennard for the explanation).
>
> Add a comment that the assert is not made superfluos by the else branch.
>
> Corrects: piglit spec@arb_gpu_shader5@execution@samplemaskin-indirect for
> debug builds (it already passed in release).
>
> Signed-off-by: Gert Wollny <gw.foss...@gmail.com>
> ---
>
> src/gallium/drivers/r600/sb/sb_bc_parser.cpp | 8 +++++++-
> 1 file changed, 7 insertions(+), 1 deletion(-)
>
> diff --git a/src/gallium/drivers/r600/sb/sb_bc_parser.cpp
> b/src/gallium/drivers/r600/sb/sb_bc_parser.cpp
> index 970e4141d5..b92e058daf 100644
> --- a/src/gallium/drivers/r600/sb/sb_bc_parser.cpp
> +++ b/src/gallium/drivers/r600/sb/sb_bc_parser.cpp
> @@ -125,7 +125,9 @@ int bc_parser::parse_decls() {
> return 0;
> }
>
> - if (pshader->indirect_files & ~((1 << TGSI_FILE_CONSTANT) | (1 <<
> TGSI_FILE_SAMPLER))) {
> + if (pshader->indirect_files &
> + ~((1 << TGSI_FILE_CONSTANT) | (1 << TGSI_FILE_SAMPLER) |
> + (1 << TGSI_FILE_SYSTEM_VALUE))) {
So I suppose this is ok because none of the other system values
could be arrays.
That said, isn't that actually a bug in glsl to tgsi translation? the
sample mask is never an array in gallium/tgsi (and not declared as
such), so I think it should not turn up as indirect when properly
translated, reading the value should not end up as indexed.
Roland
> assert(pshader->num_arrays);
>
> @@ -135,6 +137,10 @@ int bc_parser::parse_decls() {
> sh->add_gpr_array(a.gpr_start, a.gpr_count,
> a.comp_mask);
> }
> } else {
> + /* When the above assert is disabled and proper array
> info
> + * is missing for some reason then, as a fallback, make
> sure
> + * that all GPRs can be accessed indirectly.
> + */
> sh->add_gpr_array(0, pshader->bc.ngpr, 0x0F);
> }
> }
>
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
