diff options
author | Francisco Jerez <[email protected]> | 2016-05-15 22:59:04 -0700 |
---|---|---|
committer | Francisco Jerez <[email protected]> | 2016-05-23 14:05:20 -0700 |
commit | 07e67cc266bf016883ff91b3002925d0833836ed (patch) | |
tree | bbbd6cc300b758f5969f004e2251575dcaa3566d /src | |
parent | 284c8fbcef8d631a0a727f0b7a8b34f9e4ad215d (diff) |
i965/fs: Move scratch block size calculation into the caller of emit_(un)spill.
This makes emit_(un)spill even more stupid by removing the logic that
decides what execution size each scratch read or write send message
should have and instead relying on the caller to specify an
appropriate execution size via the builder argument. This makes sense
because the caller will need to act differently based on the scratch
message width (e.g. emit an additional unspill before the instruction
if the execution width and channel layout of the spill doesn't match
the instruction's).
Reviewed-by: Jason Ekstrand <[email protected]>
Diffstat (limited to 'src')
-rw-r--r-- | src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp | 44 |
1 files changed, 23 insertions, 21 deletions
diff --git a/src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp b/src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp index 0e0c176357d..099df8b815a 100644 --- a/src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp +++ b/src/mesa/drivers/dri/i965/brw_fs_reg_allocate.cpp @@ -769,12 +769,9 @@ emit_unspill(const fs_builder &bld, fs_reg dst, uint32_t spill_offset, unsigned count) { const brw_device_info *devinfo = bld.shader->devinfo; - const fs_visitor *v = static_cast<const fs_visitor *>(bld.shader); - unsigned reg_size = 1; - if (v->dispatch_width == 16 && count % 2 == 0) - reg_size = 2; - - const fs_builder ibld = bld.group(reg_size * 8, 0); + const unsigned reg_size = dst.component_size(bld.dispatch_width()) / + REG_SIZE; + assert(count % reg_size == 0); for (unsigned i = 0; i < count / reg_size; i++) { /* The Gen7 descriptor-based offset is 12 bits of HWORD units. Because @@ -786,12 +783,11 @@ emit_unspill(const fs_builder &bld, fs_reg dst, */ bool gen7_read = (devinfo->gen >= 7 && devinfo->gen < 9 && spill_offset < (1 << 12) * REG_SIZE); - fs_inst *unspill_inst = ibld.emit(gen7_read ? - SHADER_OPCODE_GEN7_SCRATCH_READ : - SHADER_OPCODE_GEN4_SCRATCH_READ, - dst); + fs_inst *unspill_inst = bld.emit(gen7_read ? + SHADER_OPCODE_GEN7_SCRATCH_READ : + SHADER_OPCODE_GEN4_SCRATCH_READ, + dst); unspill_inst->offset = spill_offset; - unspill_inst->regs_written = reg_size; if (!gen7_read) { unspill_inst->base_mrf = spill_base_mrf(bld.shader); @@ -807,16 +803,13 @@ static void emit_spill(const fs_builder &bld, fs_reg src, uint32_t spill_offset, unsigned count) { - const fs_visitor *v = static_cast<const fs_visitor *>(bld.shader); - unsigned reg_size = 1; - if (v->dispatch_width == 16 && count % 2 == 0) - reg_size = 2; - - const fs_builder ibld = bld.group(reg_size * 8, 0); + const unsigned reg_size = src.component_size(bld.dispatch_width()) / + REG_SIZE; + assert(count % reg_size == 0); for (unsigned i = 0; i < count / reg_size; i++) { fs_inst *spill_inst = - ibld.emit(SHADER_OPCODE_GEN4_SCRATCH_WRITE, ibld.null_reg_f(), src); + bld.emit(SHADER_OPCODE_GEN4_SCRATCH_WRITE, bld.null_reg_f(), src); src.reg_offset += reg_size; spill_inst->offset = spill_offset + i * reg_size * REG_SIZE; spill_inst->mlen = 1 + reg_size; /* header, value */ @@ -949,7 +942,10 @@ fs_visitor::spill_reg(int spill_reg) inst->src[i].nr = unspill_dst.nr; inst->src[i].reg_offset = 0; - emit_unspill(ibld, unspill_dst, subset_spill_offset, regs_read); + const unsigned width = + dispatch_width == 16 && regs_read % 2 == 0 ? 16 : 8; + emit_unspill(ibld.group(width, 0), unspill_dst, + subset_spill_offset, regs_read); } } @@ -970,15 +966,21 @@ fs_visitor::spill_reg(int spill_reg) inst->no_dd_clear = false; inst->no_dd_check = false; + const unsigned width = + dispatch_width == 16 && inst->regs_written % 2 == 0 ? 16 : 8; + + /* Builder used to emit the scratch messages. */ + const fs_builder ubld = ibld.group(width, 0); + /* If our write is going to affect just part of the * inst->regs_written(), then we need to unspill the destination * since we write back out all of the regs_written(). */ if (inst->is_partial_write()) - emit_unspill(ibld, spill_src, subset_spill_offset, + emit_unspill(ubld, spill_src, subset_spill_offset, inst->regs_written); - emit_spill(ibld.at(block, inst->next), spill_src, + emit_spill(ubld.at(block, inst->next), spill_src, subset_spill_offset, inst->regs_written); } } |