[Mesa-dev] [PATCH v2] nvc0/ir: be more careful about preserving modifiers in SHLADD creation
Samuel Pitoiset
samuel.pitoiset at gmail.com
Thu Oct 13 13:53:24 UTC 2016
On 10/12/2016 08:42 PM, Ilia Mirkin wrote:
> src2 was being given the wrong modifier, and we were not properly
> managing the modifier on the SHL source either.
>
> Signed-off-by: Ilia Mirkin <imirkin at alum.mit.edu>
> ---
> src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp | 12 +++++-------
> 1 file changed, 5 insertions(+), 7 deletions(-)
>
> diff --git a/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp b/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp
> index d88bb34..737bda3 100644
> --- a/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp
> +++ b/src/gallium/drivers/nouveau/codegen/nv50_ir_peephole.cpp
> @@ -2163,7 +2163,6 @@ LateAlgebraicOpt::tryADDToSHLADD(Instruction *add)
> Value *src1 = add->getSrc(1);
> ImmediateValue imm;
> Instruction *shl;
> - Modifier mod[2];
> Value *src;
> int s;
>
> @@ -2182,20 +2181,19 @@ LateAlgebraicOpt::tryADDToSHLADD(Instruction *add)
> src = add->getSrc(s);
> shl = src->getUniqueInsn();
>
> - if (shl->bb != add->bb || shl->usesFlags() || shl->subOp)
> + if (shl->bb != add->bb || shl->usesFlags() || shl->subOp || shl->src(0).mod)
SHL can't have any modifiers, why do you check it here? I would say it's
just for safety, but if the compiler adds modifiers to SHL something is
really wrong...
Looks good now.
Reviewed-by: Samuel Pitoiset <samuel.pitoiset at gmail.com>
> return false;
>
> if (!shl->src(1).getImmediate(imm))
> return false;
>
> - mod[0] = add->src(0).mod;
> - mod[1] = add->src(1).mod;
> -
> add->op = OP_SHLADD;
> add->setSrc(2, add->src(!s));
> - add->src(2).mod = mod[s];
> -
> + // SHL can't have any modifiers, but the ADD source may have had
> + // one. Preserve it.
> add->setSrc(0, shl->getSrc(0));
> + if (s == 1)
> + add->src(0).mod = add->src(1).mod;
> add->setSrc(1, new_ImmediateValue(shl->bb->getProgram(), imm.reg.data.u32));
> add->src(1).mod = Modifier(0);
>
>
--
-Samuel
More information about the mesa-dev
mailing list