[llvm] [AMDGPU][True16][CodeGen] update test for buildbot failure (PR #131028)
via llvm-commits
llvm-commits at lists.llvm.org
Wed Mar 12 13:57:22 PDT 2025
llvmbot wrote:
<!--LLVM PR SUMMARY COMMENT-->
@llvm/pr-subscribers-backend-amdgpu
Author: Brox Chen (broxigarchen)
<details>
<summary>Changes</summary>
This is a NFC patch
https://github.com/llvm/llvm-project/pull/103366 hit a buildbot failure with i1-to-bf16.ll. Update the test to fix the build.
Also remove duplicated comments added in https://github.com/llvm/llvm-project/pull/103366
---
Full diff: https://github.com/llvm/llvm-project/pull/131028.diff
2 Files Affected:
- (modified) llvm/lib/Target/AMDGPU/GCNPreRAOptimizations.cpp (-3)
- (modified) llvm/test/CodeGen/AMDGPU/i1-to-bf16.ll (+12-4)
``````````diff
diff --git a/llvm/lib/Target/AMDGPU/GCNPreRAOptimizations.cpp b/llvm/lib/Target/AMDGPU/GCNPreRAOptimizations.cpp
index 1a00b5a846b2f..4deb2a9485e4d 100644
--- a/llvm/lib/Target/AMDGPU/GCNPreRAOptimizations.cpp
+++ b/llvm/lib/Target/AMDGPU/GCNPreRAOptimizations.cpp
@@ -25,9 +25,6 @@
/// This pass also adds register allocation hints to COPY.
/// The hints will be post-processed by SIRegisterInfo::getRegAllocationHints.
/// When using True16, we often see COPY moving a 16-bit value between a VGPR_32
-/// This pass also adds register allocation hints to COPY.
-/// The hints will be post-processed by SIRegisterInfo::getRegAllocationHints.
-/// When using True16, we often see COPY moving a 16-bit value between a VGPR_32
/// and a VGPR_16. If we use the VGPR_16 that corresponds to the lo16 bits of
/// the VGPR_32, the COPY can be completely eliminated.
///
diff --git a/llvm/test/CodeGen/AMDGPU/i1-to-bf16.ll b/llvm/test/CodeGen/AMDGPU/i1-to-bf16.ll
index a46d629c02b85..59b0ba2469a20 100644
--- a/llvm/test/CodeGen/AMDGPU/i1-to-bf16.ll
+++ b/llvm/test/CodeGen/AMDGPU/i1-to-bf16.ll
@@ -108,13 +108,14 @@ define amdgpu_ps i32 @s_uitofp_i1_to_bf16(i1 inreg %num) {
; GFX11-NEXT: v_cndmask_b32_e64 v0, 0, 1.0, s0
; GFX11-NEXT: v_readfirstlane_b32 s0, v0
; GFX11-NEXT: v_cmp_u_f32_e32 vcc_lo, v0, v0
+; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(SALU_CYCLE_1)
; GFX11-NEXT: s_bfe_u32 s1, s0, 0x10010
-; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_4) | instid1(SALU_CYCLE_1)
; GFX11-NEXT: s_add_i32 s1, s1, s0
; GFX11-NEXT: s_bitset1_b32 s0, 22
; GFX11-NEXT: s_addk_i32 s1, 0x7fff
; GFX11-NEXT: s_and_b32 s2, vcc_lo, exec_lo
; GFX11-NEXT: s_cselect_b32 s0, s0, s1
+; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
; GFX11-NEXT: s_lshr_b32 s0, s0, 16
; GFX11-NEXT: ; return to shader part epilog
;
@@ -125,6 +126,7 @@ define amdgpu_ps i32 @s_uitofp_i1_to_bf16(i1 inreg %num) {
; GFX12-NEXT: s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1)
; GFX12-NEXT: v_cndmask_b32_e64 v0, 0, 1.0, s0
; GFX12-NEXT: v_readfirstlane_b32 s0, v0
+; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1)
; GFX12-NEXT: s_bfe_u32 s1, s0, 0x10010
; GFX12-NEXT: s_or_b32 s2, s0, 0x400000
; GFX12-NEXT: s_wait_alu 0xfffe
@@ -305,10 +307,11 @@ define amdgpu_ps <2 x i32> @s_uitofp_v2i1_to_v2bf16(<2 x i1> inreg %num) {
; GFX11-NEXT: v_cndmask_b32_e64 v1, 0, 1.0, s0
; GFX11-NEXT: v_readfirstlane_b32 s2, v0
; GFX11-NEXT: v_cmp_u_f32_e64 s1, v0, v0
-; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_2) | instid1(SALU_CYCLE_1)
+; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_2)
; GFX11-NEXT: v_readfirstlane_b32 s0, v1
; GFX11-NEXT: v_cmp_u_f32_e32 vcc_lo, v1, v1
; GFX11-NEXT: s_bfe_u32 s3, s0, 0x10010
+; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
; GFX11-NEXT: s_add_i32 s3, s3, s0
; GFX11-NEXT: s_bitset1_b32 s0, 22
; GFX11-NEXT: s_addk_i32 s3, 0x7fff
@@ -338,6 +341,7 @@ define amdgpu_ps <2 x i32> @s_uitofp_v2i1_to_v2bf16(<2 x i1> inreg %num) {
; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2)
; GFX12-NEXT: v_readfirstlane_b32 s2, v0
; GFX12-NEXT: v_readfirstlane_b32 s0, v1
+; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1)
; GFX12-NEXT: s_bfe_u32 s1, s0, 0x10010
; GFX12-NEXT: s_or_b32 s3, s0, 0x400000
; GFX12-NEXT: s_wait_alu 0xfffe
@@ -1161,13 +1165,14 @@ define amdgpu_ps i32 @s_sitofp_i1_to_bf16(i1 inreg %num) {
; GFX11-NEXT: v_cndmask_b32_e64 v0, 0, -1.0, s0
; GFX11-NEXT: v_readfirstlane_b32 s0, v0
; GFX11-NEXT: v_cmp_u_f32_e32 vcc_lo, v0, v0
+; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(SALU_CYCLE_1)
; GFX11-NEXT: s_bfe_u32 s1, s0, 0x10010
-; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1) | instskip(SKIP_4) | instid1(SALU_CYCLE_1)
; GFX11-NEXT: s_add_i32 s1, s1, s0
; GFX11-NEXT: s_bitset1_b32 s0, 22
; GFX11-NEXT: s_addk_i32 s1, 0x7fff
; GFX11-NEXT: s_and_b32 s2, vcc_lo, exec_lo
; GFX11-NEXT: s_cselect_b32 s0, s0, s1
+; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
; GFX11-NEXT: s_ashr_i32 s0, s0, 16
; GFX11-NEXT: ; return to shader part epilog
;
@@ -1178,6 +1183,7 @@ define amdgpu_ps i32 @s_sitofp_i1_to_bf16(i1 inreg %num) {
; GFX12-NEXT: s_delay_alu instid0(SALU_CYCLE_1) | instskip(NEXT) | instid1(VALU_DEP_1)
; GFX12-NEXT: v_cndmask_b32_e64 v0, 0, -1.0, s0
; GFX12-NEXT: v_readfirstlane_b32 s0, v0
+; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1)
; GFX12-NEXT: s_bfe_u32 s1, s0, 0x10010
; GFX12-NEXT: s_or_b32 s2, s0, 0x400000
; GFX12-NEXT: s_wait_alu 0xfffe
@@ -1358,10 +1364,11 @@ define amdgpu_ps <2 x i32> @s_sitofp_v2i1_to_v2bf16(<2 x i1> inreg %num) {
; GFX11-NEXT: v_cndmask_b32_e64 v1, 0, -1.0, s0
; GFX11-NEXT: v_readfirstlane_b32 s2, v0
; GFX11-NEXT: v_cmp_u_f32_e64 s0, v0, v0
-; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_2) | instid1(SALU_CYCLE_1)
+; GFX11-NEXT: s_delay_alu instid0(VALU_DEP_3) | instskip(SKIP_1) | instid1(VALU_DEP_2)
; GFX11-NEXT: v_readfirstlane_b32 s1, v1
; GFX11-NEXT: v_cmp_u_f32_e32 vcc_lo, v1, v1
; GFX11-NEXT: s_bfe_u32 s3, s1, 0x10010
+; GFX11-NEXT: s_delay_alu instid0(SALU_CYCLE_1)
; GFX11-NEXT: s_add_i32 s3, s3, s1
; GFX11-NEXT: s_bitset1_b32 s1, 22
; GFX11-NEXT: s_addk_i32 s3, 0x7fff
@@ -1391,6 +1398,7 @@ define amdgpu_ps <2 x i32> @s_sitofp_v2i1_to_v2bf16(<2 x i1> inreg %num) {
; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_2) | instskip(NEXT) | instid1(VALU_DEP_2)
; GFX12-NEXT: v_readfirstlane_b32 s2, v0
; GFX12-NEXT: v_readfirstlane_b32 s0, v1
+; GFX12-NEXT: s_delay_alu instid0(VALU_DEP_1)
; GFX12-NEXT: s_bfe_u32 s1, s0, 0x10010
; GFX12-NEXT: s_or_b32 s3, s0, 0x400000
; GFX12-NEXT: s_wait_alu 0xfffe
``````````
</details>
https://github.com/llvm/llvm-project/pull/131028
More information about the llvm-commits
mailing list