[AMDGPU] Remove default wavefrontsize attribute in asm parser #97617
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
If a non-default wavefrontsize is requested in the command line we end up with both wavefrontsize32 and wavefrontsize64 set in the feature bits, one from the command line and one from the target defaults. Because of this many of our tests has -mattr=-wavefrontsize32,+wavefrontsize64 idiom. Remove the default feature bit in this case.
|
A similar code exists in the llc and needed in the disasm. |
|
@llvm/pr-subscribers-backend-amdgpu Author: Stanislav Mekhanoshin (rampitec) ChangesIf a non-default wavefrontsize is requested in the command line we end up with both wavefrontsize32 and wavefrontsize64 set in the feature bits, one from the command line and one from the target defaults. Because of this many of our tests has -mattr=-wavefrontsize32,+wavefrontsize64 idiom. Remove the default feature bit in this case. Patch is 190.13 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/97617.diff 116 Files Affected:
diff --git a/llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp b/llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
index b08957d22ee74..18f2144211d0f 100644
--- a/llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
+++ b/llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
@@ -1408,7 +1408,20 @@ class AMDGPUAsmParser : public MCTargetAsmParser {
copySTI().ToggleFeature("southern-islands");
}
- setAvailableFeatures(ComputeAvailableFeatures(getFeatureBits()));
+ StringRef FS = getSTI().getFeatureString();
+ FeatureBitset FB = getFeatureBits();
+ int isWave32 = FS.contains("+wavefrontsize32");
+ int isWave64 = FS.contains("+wavefrontsize64");
+ // Reset default wavefrontsize features.
+ if (isWave32 + isWave64 == 1) {
+ if (isWave32 && FB[AMDGPU::FeatureWavefrontSize64])
+ copySTI().ToggleFeature(AMDGPU::FeatureWavefrontSize64);
+ else if (isWave64 && FB[AMDGPU::FeatureWavefrontSize32])
+ copySTI().ToggleFeature(AMDGPU::FeatureWavefrontSize32);
+ FB = getFeatureBits();
+ }
+
+ setAvailableFeatures(ComputeAvailableFeatures(FB));
AMDGPU::IsaVersion ISA = AMDGPU::getIsaVersion(getSTI().getCPU());
if (ISA.Major >= 6 && isHsaAbi(getSTI())) {
diff --git a/llvm/test/MC/AMDGPU/expressions-gfx10.s b/llvm/test/MC/AMDGPU/expressions-gfx10.s
index fc029a3d8e519..efc8dc9479884 100644
--- a/llvm/test/MC/AMDGPU/expressions-gfx10.s
+++ b/llvm/test/MC/AMDGPU/expressions-gfx10.s
@@ -1,5 +1,5 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck %s --check-prefix=GFX10
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck -check-prefix=NOGFX10 --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck %s --check-prefix=GFX10
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck -check-prefix=NOGFX10 --implicit-check-not=error: %s
i1=1
diff --git a/llvm/test/MC/AMDGPU/gfx1013.s b/llvm/test/MC/AMDGPU/gfx1013.s
index a4180c3d93a23..f8bad215fdc5a 100644
--- a/llvm/test/MC/AMDGPU/gfx1013.s
+++ b/llvm/test/MC/AMDGPU/gfx1013.s
@@ -1,4 +1,4 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1013 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1013 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck %s
image_bvh64_intersect_ray v[5:8], v[1:12], s[8:11]
// CHECK: [0x01,0x9f,0x9c,0xf1,0x01,0x05,0x02,0x00]
diff --git a/llvm/test/MC/AMDGPU/gfx1030_unsupported.s b/llvm/test/MC/AMDGPU/gfx1030_unsupported.s
index 9112a30b0b7bd..da1fb512fec98 100644
--- a/llvm/test/MC/AMDGPU/gfx1030_unsupported.s
+++ b/llvm/test/MC/AMDGPU/gfx1030_unsupported.s
@@ -1,5 +1,5 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1030 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1030 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1030 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1030 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --implicit-check-not=error: %s
v_mul_lo_i32 v0, v1, v2
// CHECK: :[[@LINE-1]]:{{[0-9]+}}: error: instruction not supported on this GPU
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_dpp16.s b/llvm/test/MC/AMDGPU/gfx10_asm_dpp16.s
index 77df1f62b2243..d8987be832d4c 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_dpp16.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_dpp16.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
v_mov_b32_dpp v5, v1 quad_perm:[3,2,1,0] row_mask:0x0 bank_mask:0x0
// GFX10: [0xfa,0x02,0x0a,0x7e,0x01,0x1b,0x00,0x00]
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_dpp8.s b/llvm/test/MC/AMDGPU/gfx10_asm_dpp8.s
index 260d06d2e4cb5..944e2b3526464 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_dpp8.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_dpp8.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
v_mov_b32_dpp v5, v1 dpp8:[0,1,2,3,4,5,6,7]
// GFX10: encoding: [0xe9,0x02,0x0a,0x7e,0x01,0x88,0xc6,0xfa]
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_ds.s b/llvm/test/MC/AMDGPU/gfx10_asm_ds.s
index abe3586e3d0ea..ccdcd2262f9a4 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_ds.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_ds.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_DS.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_err.s b/llvm/test/MC/AMDGPU/gfx10_asm_err.s
index ef12ba2a66b19..eee26196623d6 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_err.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_err.s
@@ -2,8 +2,8 @@
// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx701 %s 2>&1 | FileCheck --check-prefixes=GFX6-7,GFX6-8,GFX6-9 --implicit-check-not=error: %s
// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx801 %s 2>&1 | FileCheck --check-prefixes=GFX6-8,GFX6-9,GFX8-9 --implicit-check-not=error: %s
// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx900 %s 2>&1 | FileCheck --check-prefixes=GFX6-9,GFX8-9 --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10 --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10 --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefixes=GFX10 --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10 --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_DS.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_flat.s b/llvm/test/MC/AMDGPU/gfx10_asm_flat.s
index 7340e48db3e84..4a375c224b122 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_flat.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_flat.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_FLAT.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_mubuf.s b/llvm/test/MC/AMDGPU/gfx10_asm_mubuf.s
index b77f8e0a31927..95c22ef8ce283 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_mubuf.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_mubuf.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_MUBUF.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_smem.s b/llvm/test/MC/AMDGPU/gfx10_asm_smem.s
index 25c9e8f35093e..b582de83a2f29 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_smem.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_smem.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_SMEM.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_sop.s b/llvm/test/MC/AMDGPU/gfx10_asm_sop.s
index 8f1cde76c6aa3..c35b04c20c8c3 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_sop.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_sop.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_SOP1.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vop1.s b/llvm/test/MC/AMDGPU/gfx10_asm_vop1.s
index 5a1673ef1ccef..3cc25501ff7c0 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vop1.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vop1.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=GFX10-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=GFX10-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=GFX10-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=GFX10-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOP1.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vop2.s b/llvm/test/MC/AMDGPU/gfx10_asm_vop2.s
index bf8e18ec14512..3dcf288bbbaa5 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vop2.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vop2.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOP2.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vop3.s b/llvm/test/MC/AMDGPU/gfx10_asm_vop3.s
index b4256f1a628d5..c151bf99b76c5 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vop3.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vop3.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10-ERR,W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10-ERR,W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefixes=GFX10-ERR,W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10-ERR,W64-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOP3.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vopc.s b/llvm/test/MC/AMDGPU/gfx10_asm_vopc.s
index 40618d1ea5235..8f42a47834961 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vopc.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vopc.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOPC - v_cmp_* opcodes.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vopc_e64.s b/llvm/test/MC/AMDGPU/gfx10_asm_vopc_e64.s
index 669e79cdab10a..f182de8fea984 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vopc_e64.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vopc_e64.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOPC, VOP3 variant.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vopc_sdwa.s b/llvm/test/MC/AMDGPU/gfx10_asm_vopc_sdwa.s
index 8808ab9ff2f9c..50bd8c70ea5c0 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vopc_sdwa.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vopc_sdwa.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
-// RUN: not llvm-mc -triple=a...
[truncated]
|
|
@llvm/pr-subscribers-mc Author: Stanislav Mekhanoshin (rampitec) ChangesIf a non-default wavefrontsize is requested in the command line we end up with both wavefrontsize32 and wavefrontsize64 set in the feature bits, one from the command line and one from the target defaults. Because of this many of our tests has -mattr=-wavefrontsize32,+wavefrontsize64 idiom. Remove the default feature bit in this case. Patch is 190.13 KiB, truncated to 20.00 KiB below, full version: https://github.com/llvm/llvm-project/pull/97617.diff 116 Files Affected:
diff --git a/llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp b/llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
index b08957d22ee74..18f2144211d0f 100644
--- a/llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
+++ b/llvm/lib/Target/AMDGPU/AsmParser/AMDGPUAsmParser.cpp
@@ -1408,7 +1408,20 @@ class AMDGPUAsmParser : public MCTargetAsmParser {
copySTI().ToggleFeature("southern-islands");
}
- setAvailableFeatures(ComputeAvailableFeatures(getFeatureBits()));
+ StringRef FS = getSTI().getFeatureString();
+ FeatureBitset FB = getFeatureBits();
+ int isWave32 = FS.contains("+wavefrontsize32");
+ int isWave64 = FS.contains("+wavefrontsize64");
+ // Reset default wavefrontsize features.
+ if (isWave32 + isWave64 == 1) {
+ if (isWave32 && FB[AMDGPU::FeatureWavefrontSize64])
+ copySTI().ToggleFeature(AMDGPU::FeatureWavefrontSize64);
+ else if (isWave64 && FB[AMDGPU::FeatureWavefrontSize32])
+ copySTI().ToggleFeature(AMDGPU::FeatureWavefrontSize32);
+ FB = getFeatureBits();
+ }
+
+ setAvailableFeatures(ComputeAvailableFeatures(FB));
AMDGPU::IsaVersion ISA = AMDGPU::getIsaVersion(getSTI().getCPU());
if (ISA.Major >= 6 && isHsaAbi(getSTI())) {
diff --git a/llvm/test/MC/AMDGPU/expressions-gfx10.s b/llvm/test/MC/AMDGPU/expressions-gfx10.s
index fc029a3d8e519..efc8dc9479884 100644
--- a/llvm/test/MC/AMDGPU/expressions-gfx10.s
+++ b/llvm/test/MC/AMDGPU/expressions-gfx10.s
@@ -1,5 +1,5 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck %s --check-prefix=GFX10
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck -check-prefix=NOGFX10 --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck %s --check-prefix=GFX10
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck -check-prefix=NOGFX10 --implicit-check-not=error: %s
i1=1
diff --git a/llvm/test/MC/AMDGPU/gfx1013.s b/llvm/test/MC/AMDGPU/gfx1013.s
index a4180c3d93a23..f8bad215fdc5a 100644
--- a/llvm/test/MC/AMDGPU/gfx1013.s
+++ b/llvm/test/MC/AMDGPU/gfx1013.s
@@ -1,4 +1,4 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1013 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1013 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck %s
image_bvh64_intersect_ray v[5:8], v[1:12], s[8:11]
// CHECK: [0x01,0x9f,0x9c,0xf1,0x01,0x05,0x02,0x00]
diff --git a/llvm/test/MC/AMDGPU/gfx1030_unsupported.s b/llvm/test/MC/AMDGPU/gfx1030_unsupported.s
index 9112a30b0b7bd..da1fb512fec98 100644
--- a/llvm/test/MC/AMDGPU/gfx1030_unsupported.s
+++ b/llvm/test/MC/AMDGPU/gfx1030_unsupported.s
@@ -1,5 +1,5 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1030 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1030 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1030 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1030 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --implicit-check-not=error: %s
v_mul_lo_i32 v0, v1, v2
// CHECK: :[[@LINE-1]]:{{[0-9]+}}: error: instruction not supported on this GPU
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_dpp16.s b/llvm/test/MC/AMDGPU/gfx10_asm_dpp16.s
index 77df1f62b2243..d8987be832d4c 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_dpp16.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_dpp16.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
v_mov_b32_dpp v5, v1 quad_perm:[3,2,1,0] row_mask:0x0 bank_mask:0x0
// GFX10: [0xfa,0x02,0x0a,0x7e,0x01,0x1b,0x00,0x00]
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_dpp8.s b/llvm/test/MC/AMDGPU/gfx10_asm_dpp8.s
index 260d06d2e4cb5..944e2b3526464 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_dpp8.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_dpp8.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
v_mov_b32_dpp v5, v1 dpp8:[0,1,2,3,4,5,6,7]
// GFX10: encoding: [0xe9,0x02,0x0a,0x7e,0x01,0x88,0xc6,0xfa]
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_ds.s b/llvm/test/MC/AMDGPU/gfx10_asm_ds.s
index abe3586e3d0ea..ccdcd2262f9a4 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_ds.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_ds.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_DS.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_err.s b/llvm/test/MC/AMDGPU/gfx10_asm_err.s
index ef12ba2a66b19..eee26196623d6 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_err.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_err.s
@@ -2,8 +2,8 @@
// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx701 %s 2>&1 | FileCheck --check-prefixes=GFX6-7,GFX6-8,GFX6-9 --implicit-check-not=error: %s
// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx801 %s 2>&1 | FileCheck --check-prefixes=GFX6-8,GFX6-9,GFX8-9 --implicit-check-not=error: %s
// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx900 %s 2>&1 | FileCheck --check-prefixes=GFX6-9,GFX8-9 --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10 --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10 --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefixes=GFX10 --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10 --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_DS.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_flat.s b/llvm/test/MC/AMDGPU/gfx10_asm_flat.s
index 7340e48db3e84..4a375c224b122 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_flat.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_flat.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_FLAT.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_mubuf.s b/llvm/test/MC/AMDGPU/gfx10_asm_mubuf.s
index b77f8e0a31927..95c22ef8ce283 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_mubuf.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_mubuf.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_MUBUF.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_smem.s b/llvm/test/MC/AMDGPU/gfx10_asm_smem.s
index 25c9e8f35093e..b582de83a2f29 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_smem.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_smem.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_SMEM.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_sop.s b/llvm/test/MC/AMDGPU/gfx10_asm_sop.s
index 8f1cde76c6aa3..c35b04c20c8c3 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_sop.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_sop.s
@@ -1,5 +1,5 @@
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
//===----------------------------------------------------------------------===//
// ENC_SOP1.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vop1.s b/llvm/test/MC/AMDGPU/gfx10_asm_vop1.s
index 5a1673ef1ccef..3cc25501ff7c0 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vop1.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vop1.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=GFX10-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=GFX10-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=GFX10 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=GFX10-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=GFX10-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOP1.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vop2.s b/llvm/test/MC/AMDGPU/gfx10_asm_vop2.s
index bf8e18ec14512..3dcf288bbbaa5 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vop2.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vop2.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOP2.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vop3.s b/llvm/test/MC/AMDGPU/gfx10_asm_vop3.s
index b4256f1a628d5..c151bf99b76c5 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vop3.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vop3.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10-ERR,W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10-ERR,W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefixes=GFX10,W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefixes=GFX10,W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefixes=GFX10-ERR,W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefixes=GFX10-ERR,W64-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOP3.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vopc.s b/llvm/test/MC/AMDGPU/gfx10_asm_vopc.s
index 40618d1ea5235..8f42a47834961 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vopc.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vopc.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOPC - v_cmp_* opcodes.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vopc_e64.s b/llvm/test/MC/AMDGPU/gfx10_asm_vopc_e64.s
index 669e79cdab10a..f182de8fea984 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vopc_e64.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vopc_e64.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 -show-encoding %s | FileCheck --check-prefix=W32 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32 %s 2>&1 | FileCheck --check-prefix=W32-ERR --implicit-check-not=error: %s
+// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize64 %s 2>&1 | FileCheck --check-prefix=W64-ERR --implicit-check-not=error: %s
//===----------------------------------------------------------------------===//
// ENC_VOPC, VOP3 variant.
diff --git a/llvm/test/MC/AMDGPU/gfx10_asm_vopc_sdwa.s b/llvm/test/MC/AMDGPU/gfx10_asm_vopc_sdwa.s
index 8808ab9ff2f9c..50bd8c70ea5c0 100644
--- a/llvm/test/MC/AMDGPU/gfx10_asm_vopc_sdwa.s
+++ b/llvm/test/MC/AMDGPU/gfx10_asm_vopc_sdwa.s
@@ -1,7 +1,7 @@
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=+wavefrontsize32,-wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W32 %s
-// RUN: not llvm-mc -triple=amdgcn -mcpu=gfx1010 -mattr=-wavefrontsize32,+wavefrontsize64 -show-encoding %s | FileCheck --check-prefix=W64 %s
-// RUN: not llvm-mc -triple=a...
[truncated]
|
rampitec
commented
Jul 3, 2024
| @@ -12,13 +12,13 @@ v_cmp_class_f16 vcc, vcc_hi, v255 | |||
| v_cmp_class_f16 vcc, vcc_lo, v255 | |||
| // GFX11: v_cmp_class_f16_e64 | |||
|
|
|||
| v_cmp_class_f16 vcc_lo, v127, v255 | |||
| v_cmp_class_f16 vcc, v127, v255 | |||
There was a problem hiding this comment.
This was a bug and vcc_lo should have never be allowed as sdst with wave64, but we had both wavefrontsize32 and wvefrontsize64 set, thus it was accepted.
arsenm
reviewed
Jul 3, 2024
| int isWave64 = FS.contains("+wavefrontsize64"); | ||
| // Reset default wavefrontsize features. | ||
| if (isWave32 + isWave64 == 1) { | ||
| if (isWave32 && FB[AMDGPU::FeatureWavefrontSize64]) |
There was a problem hiding this comment.
We shouldn't really be directly checking the bits. The + and -s are sequentially evaluated in the subtarget construction to set the WavefrontSize field. I would expect the same to happen down in MC. I don't know why the subtarget evaluation logic isn't an an MCSubtarget superclass
There was a problem hiding this comment.
We do exactly the same in the codegen: https://llvm.org/doxygen/AMDGPUSubtarget_8cpp_source.html#l00085
There was a problem hiding this comment.
That's also hacky pre-processing of the input features strings, which should also be avoidable. The actual interpretation logic is in ParseSubtargetFeatures
There was a problem hiding this comment.
We can remove FeatureWavefrontSize32 from gfx10 and above, but then we need to add default wavefront size feature everywhere, and we need to parse feature string for this anyway. We are already doing it in clang and flang, but not in any other tools.
There was a problem hiding this comment.
Like this: #98400. Actually w/o additional string parsing.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
If a non-default wavefrontsize is requested in the command line we end up with both wavefrontsize32 and wavefrontsize64 set in the feature bits, one from the command line and one from the target defaults. Because of this many of our tests has -mattr=-wavefrontsize32,+wavefrontsize64 idiom.
Remove the default feature bit in this case.