Technically those hw operations are only available on gen7, as gen8+
support the conversion on the MOV. But, when using the builder to
implement nir operations (example: nir_op_fquantize2f16), it is not
needed to do the gen check. This check is done later, on the final
emission at brw_F32TO16 (brw_eu_emit), choosing between the MOV or the
specific operation accordingly.
So in the middle, during optimization phases those hw operations can
be around for gen8+ too.
Without this patch, several (at least 95) vulkan-cts quantize tests
crashes when using INTEL_DEBUG=optimizer. For example:
dEQP-VK.spirv_assembly.instruction.graphics.opquantize.too_small_vert
---
src/intel/compiler/brw_eu.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/src/intel/compiler/brw_eu.c b/src/intel/compiler/brw_eu.c
index 77400c1..bff37d7 100644
--- a/src/intel/compiler/brw_eu.c
+++ b/src/intel/compiler/brw_eu.c
@@ -499,10 +499,10 @@ static const struct opcode_desc opcode_descs[128] = {
.name = "csel", .nsrc = 3, .ndst = 1, .gens = GEN_GE(GEN8),
},
[BRW_OPCODE_F32TO16] = {
- .name = "f32to16", .nsrc = 1, .ndst = 1, .gens = GEN7 | GEN75,
+ .name = "f32to16", .nsrc = 1, .ndst = 1, .gens = GEN7 | GEN75 | GEN8 |
GEN9,
},
[BRW_OPCODE_F16TO32] = {
- .name = "f16to32", .nsrc = 1, .ndst = 1, .gens = GEN7 | GEN75,
+ .name = "f16to32", .nsrc = 1, .ndst = 1, .gens = GEN7 | GEN75 | GEN8 |
GEN9,
},
/* Reserved - 21-22 */
[BRW_OPCODE_BFREV] = {
--
2.9.3
_______________________________________________
mesa-dev mailing list
mesa-dev@lists.freedesktop.org
https://lists.freedesktop.org/mailman/listinfo/mesa-dev
