diff options
| author | Francisco Jerez <[email protected]> | 2016-05-19 00:10:03 -0700 |
|---|---|---|
| committer | Francisco Jerez <[email protected]> | 2016-05-27 23:29:04 -0700 |
| commit | 81bc6de8c0f7faafd0f3b0aee944a14ba3ef0b64 (patch) | |
| tree | 15be42f1cedc34eb105d7d8fb7d0838c5c9a005c /src/mesa/drivers/dri/i965/brw_eu_emit.c | |
| parent | 41562eb8f33558f02ff8f53b3094a0e6d54e4c49 (diff) | |
i965/ir: Make BROADCAST emit an unmasked single-channel move.
Alternatively we could have extended the current semantics to 32-wide
mode by changing brw_broadcast() to emit multiple indexed MOV
instructions in the generator copying the selected value to all
destination registers, but it seemed rather silly to waste EU cycles
unnecessarily copying the exact same value 32 times in the GRF.
The vstride change in the Align16 path is required to avoid assertions
in validate_reg() since the change causes the execution size of the
MOV and SEL instructions to be equal to the source region width.
Reviewed-by: Jason Ekstrand <[email protected]>
Diffstat (limited to 'src/mesa/drivers/dri/i965/brw_eu_emit.c')
| -rw-r--r-- | src/mesa/drivers/dri/i965/brw_eu_emit.c | 12 |
1 files changed, 9 insertions, 3 deletions
diff --git a/src/mesa/drivers/dri/i965/brw_eu_emit.c b/src/mesa/drivers/dri/i965/brw_eu_emit.c index 2a3b0b03d80..1c2ccb4e163 100644 --- a/src/mesa/drivers/dri/i965/brw_eu_emit.c +++ b/src/mesa/drivers/dri/i965/brw_eu_emit.c @@ -3426,6 +3426,10 @@ brw_broadcast(struct brw_codegen *p, const bool align1 = brw_inst_access_mode(devinfo, p->current) == BRW_ALIGN_1; brw_inst *inst; + brw_push_insn_state(p); + brw_set_default_mask_control(p, BRW_MASK_DISABLE); + brw_set_default_exec_size(p, align1 ? BRW_EXECUTE_1 : BRW_EXECUTE_4); + assert(src.file == BRW_GENERAL_REGISTER_FILE && src.address_mode == BRW_ADDRESS_DIRECT); @@ -3476,19 +3480,21 @@ brw_broadcast(struct brw_codegen *p, */ inst = brw_MOV(p, brw_null_reg(), - stride(brw_swizzle(idx, BRW_SWIZZLE_XXXX), 0, 4, 1)); + stride(brw_swizzle(idx, BRW_SWIZZLE_XXXX), 4, 4, 1)); brw_inst_set_pred_control(devinfo, inst, BRW_PREDICATE_NONE); brw_inst_set_cond_modifier(devinfo, inst, BRW_CONDITIONAL_NZ); brw_inst_set_flag_reg_nr(devinfo, inst, 1); /* and use predicated SEL to pick the right channel. */ inst = brw_SEL(p, dst, - stride(suboffset(src, 4), 0, 4, 1), - stride(src, 0, 4, 1)); + stride(suboffset(src, 4), 4, 4, 1), + stride(src, 4, 4, 1)); brw_inst_set_pred_control(devinfo, inst, BRW_PREDICATE_NORMAL); brw_inst_set_flag_reg_nr(devinfo, inst, 1); } } + + brw_pop_insn_state(p); } /** |