[PATCH] D157712: [AMDGPU] Autogenerate & pre-commit tests for D156301 and D157388

[Matt Arsenault via Phabricator via llvm-commits]

arsenm added inline comments.


================
Comment at: llvm/test/CodeGen/AMDGPU/global_atomic_optimizer_fp_rtn.ll:123
+
+define amdgpu_ps float @global_atomic_fmax_div_value_scope_agent(ptr addrspace(1) inreg %ptr, float %val) #1{
+; IR-ITERATIVE-LABEL: @global_atomic_fmax_div_value_scope_agent(
----------------
If you're testing a divergent value, might as well have the pointer be divergent too


================
Comment at: llvm/test/CodeGen/AMDGPU/global_atomic_optimizer_fp_rtn.ll:136
+
+define amdgpu_ps float @global_atomic_fadd_uni_value_scope_defalut(ptr addrspace(1) inreg %ptr, float inreg %val) #2 {
+; IR-ITERATIVE-LABEL: @global_atomic_fadd_uni_value_scope_defalut(
----------------
should be a separate strictfp test, the strictfp is hidden


================
Comment at: llvm/test/CodeGen/AMDGPU/global_atomic_optimizer_fp_rtn.ll:149
+
+define amdgpu_ps float @global_atomic_fadd_div_value_scope_defalut(ptr addrspace(1) inreg %ptr, float %val) #2 {
+; IR-ITERATIVE-LABEL: @global_atomic_fadd_div_value_scope_defalut(
----------------
Typo defalut. Also call it system


================
Comment at: llvm/test/CodeGen/AMDGPU/global_atomics_optimizer_fp_no_rtn.ll:149
+
+define amdgpu_ps void @global_atomic_fadd_div_value_scope_defalut(ptr addrspace(1) inreg %ptr, float %val) #2 {
+; IR-ITERATIVE-LABEL: @global_atomic_fadd_div_value_scope_defalut(
----------------
Ditto for everything in the return version


================
Comment at: llvm/test/CodeGen/AMDGPU/global_atomics_optimizer_fp_no_rtn.ll:165
+attributes #1 = { strictfp "denormal-fp-math-f32"="preserve-sign,preserve-sign" "amdgpu-unsafe-fp-atomics"="true" }
+attributes #2 = { strictfp}
----------------
Missing space


Repository:
  rG LLVM Github Monorepo

CHANGES SINCE LAST ACTION
  https://reviews.llvm.org/D157712/new/

https://reviews.llvm.org/D157712