Re: [PATCH] tree-optimization/110991 - unroll size estimate after vectorization

2023-08-14 Thread Jan Hubicka via Gcc-patches
> The following testcase shows that we are bad at identifying inductions > that will be optimized away after vectorizing them because SCEV doesn't > handle vectorized defs. The following rolls a simpler identification > of SSA cycles covering a PHI and an assignment with a binary operator > with

Re: [PATCH] tree-optimization/110991 - unroll size estimate after vectorization

2023-08-14 Thread Richard Biener via Gcc-patches
On Mon, 14 Aug 2023, Richard Biener wrote: > The following testcase shows that we are bad at identifying inductions > that will be optimized away after vectorizing them because SCEV doesn't > handle vectorized defs. The following rolls a simpler identification > of SSA cycles covering a PHI and

[PATCH] tree-optimization/110991 - unroll size estimate after vectorization

2023-08-14 Thread Richard Biener via Gcc-patches
The following testcase shows that we are bad at identifying inductions that will be optimized away after vectorizing them because SCEV doesn't handle vectorized defs. The following rolls a simpler identification of SSA cycles covering a PHI and an assignment with a binary operator with a constant