[PATCH] D106053: [CodeGen] Remove pending AssertZext AssertSext in promoting FP_TO_INT
Xiang Zhang via Phabricator via llvm-commits
llvm-commits at lists.llvm.org
Thu Jul 15 18:48:06 PDT 2021
xiangzhangllvm added inline comments.
================
Comment at: llvm/test/CodeGen/X86/tmp/rep.ll:14
+VPlannedBB181vector_func.i7zxzx: ; preds = %VPlannedBB181vector_func.i6zxzx
+ %i376 = fptoui <8 x float> %i375 to <8 x i8>
+ %i377 = fmul <8 x float> %2, <float 2.000000e+00, float 2.000000e+00, float 2.000000e+00, float 2.000000e+00, float 2.000000e+00, float 2.000000e+00, float 2.000000e+00, float 2.000000e+00>
----------------
craig.topper wrote:
> Or is it this fptoui that overflowed?
Yes, It is. We can't control the load value of %1 %2
================
Comment at: llvm/test/CodeGen/X86/tmp/rep.ll:27
+ %i386 = fmul <8 x float> %i385, <float 2.550000e+02, float 2.550000e+02, float 2.550000e+02, float 2.550000e+02, float 2.550000e+02, float 2.550000e+02, float 2.550000e+02, float 2.550000e+02>
+ %i387 = fptoui <8 x float> %i386 to <8 x i8>
+
----------------
craig.topper wrote:
> Can you provide the values for %call.i24.i, %i384, %i385, %i386 in your failing case?
This is just a small reproduce, in our project, if we clear the high 8 bits of the related element, the project will run pass.
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D106053/new/
https://reviews.llvm.org/D106053
More information about the llvm-commits
mailing list