From 4f4e9ba1661528bed8e956a4931ae154e6612824 Mon Sep 17 00:00:00 2001 From: Tobias Klausmann Date: Wed, 4 Jun 2014 00:35:50 +0200 Subject: nvc0/ir: Handle OP_POPCNT when folding constant expressions Signed-off-by: Tobias Klausmann [imirkin: make sure to only fold 1-arg popcnt in opnd] Reviewed-by: Ilia Mirkin --- src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp | 13 +++++++++++++ 1 file changed, 13 insertions(+) (limited to 'src/gallium/drivers/nouveau/codegen') diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp b/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp index a91d698c5b2..b89da43136e 100644 --- a/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp +++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp @@ -540,6 +540,9 @@ ConstantFolding::expr(Instruction *i, } break; } + case OP_POPCNT: + res.data.u32 = util_bitcount(a->data.u32 & b->data.u32); + break; default: return; } @@ -957,6 +960,16 @@ ConstantFolding::opnd(Instruction *i, ImmediateValue &imm0, int s) i->subOp = 0; break; } + case OP_POPCNT: { + // Only deal with 1-arg POPCNT here + if (i->srcExists(1)) + break; + uint32_t res = util_bitcount(imm0.reg.data.u32); + i->setSrc(0, new_ImmediateValue(i->bb->getProgram(), res)); + i->setSrc(1, NULL); + i->op = OP_MOV; + break; + } default: return; } -- cgit v1.2.3