[clang][CodeGen] Remove "unsafe-fp-math" attribute support #162779
Merged
+15
−43
Conversation
This file contains hidden or 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
llvmbot
added
clang
paperchalice
requested review from
andykaylor,
arsenm,
efriedma-quic,
fmayer and
ojhunt
|
@llvm/pr-subscribers-clang @llvm/pr-subscribers-clang-codegen Author: None (paperchalice) ChangesThese global flags block furthur improvements for clang, users should always use fast-math flags Full diff: https://github.com/llvm/llvm-project/pull/162779.diff 8 Files Affected:
diff --git a/clang/lib/CodeGen/CGCall.cpp b/clang/lib/CodeGen/CGCall.cpp
index c5371e4c841cb..df28641904c04 100644
--- a/clang/lib/CodeGen/CGCall.cpp
+++ b/clang/lib/CodeGen/CGCall.cpp
@@ -2012,13 +2012,6 @@ static void getTrivialDefaultFunctionAttributes(
FuncAttrs.addAttribute("no-infs-fp-math", "true");
if (LangOpts.NoHonorNaNs)
FuncAttrs.addAttribute("no-nans-fp-math", "true");
- if (LangOpts.AllowFPReassoc && LangOpts.AllowRecip &&
- LangOpts.NoSignedZero && LangOpts.ApproxFunc &&
- (LangOpts.getDefaultFPContractMode() ==
- LangOptions::FPModeKind::FPM_Fast ||
- LangOpts.getDefaultFPContractMode() ==
- LangOptions::FPModeKind::FPM_FastHonorPragmas))
- FuncAttrs.addAttribute("unsafe-fp-math", "true");
if (CodeGenOpts.SoftFloat)
FuncAttrs.addAttribute("use-soft-float", "true");
FuncAttrs.addAttribute("stack-protector-buffer-size",
diff --git a/clang/lib/CodeGen/CGCall.h b/clang/lib/CodeGen/CGCall.h
index 935b5086f5983..1ef8a3f114573 100644
--- a/clang/lib/CodeGen/CGCall.h
+++ b/clang/lib/CodeGen/CGCall.h
@@ -410,10 +410,10 @@ class ReturnValueSlot {
/// This is useful for adding attrs to bitcode modules that you want to link
/// with but don't control, such as CUDA's libdevice. When linking with such
/// a bitcode library, you might want to set e.g. its functions'
-/// "unsafe-fp-math" attribute to match the attr of the functions you're
+/// "denormal-fp-math" attribute to match the attr of the functions you're
/// codegen'ing. Otherwise, LLVM will interpret the bitcode module's lack of
-/// unsafe-fp-math attrs as tantamount to unsafe-fp-math=false, and then LLVM
-/// will propagate unsafe-fp-math=false up to every transitive caller of a
+/// denormal-fp-math attrs as tantamount to denormal-fp-math=ieee, and then LLVM
+/// will propagate denormal-fp-math=ieee up to every transitive caller of a
/// function in the bitcode library!
///
/// With the exception of fast-math attrs, this will only make the attributes
diff --git a/clang/lib/CodeGen/CodeGenFunction.cpp b/clang/lib/CodeGen/CodeGenFunction.cpp
index acf8de4dee147..88628530cf66b 100644
--- a/clang/lib/CodeGen/CodeGenFunction.cpp
+++ b/clang/lib/CodeGen/CodeGenFunction.cpp
@@ -183,11 +183,6 @@ void CodeGenFunction::CGFPOptionsRAII::ConstructorHelper(FPOptions FPFeatures) {
mergeFnAttrValue("no-infs-fp-math", FPFeatures.getNoHonorInfs());
mergeFnAttrValue("no-nans-fp-math", FPFeatures.getNoHonorNaNs());
mergeFnAttrValue("no-signed-zeros-fp-math", FPFeatures.getNoSignedZero());
- mergeFnAttrValue(
- "unsafe-fp-math",
- FPFeatures.getAllowFPReassociate() && FPFeatures.getAllowReciprocal() &&
- FPFeatures.getAllowApproxFunc() && FPFeatures.getNoSignedZero() &&
- FPFeatures.allowFPContractAcrossStatement());
}
CodeGenFunction::CGFPOptionsRAII::~CGFPOptionsRAII() {
diff --git a/clang/test/CodeGen/backend-unsupported-error.ll b/clang/test/CodeGen/backend-unsupported-error.ll
index 2de2c87aa720d..47b397b6bb40b 100644
--- a/clang/test/CodeGen/backend-unsupported-error.ll
+++ b/clang/test/CodeGen/backend-unsupported-error.ll
@@ -21,7 +21,7 @@ entry:
ret i32 %call, !dbg !15
}
-attributes #0 = { nounwind noinline "disable-tail-calls"="false" "less-precise-fpmad"="false" "frame-pointer"="all" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "unsafe-fp-math"="false" "use-soft-float"="false" }
+attributes #0 = { nounwind noinline "disable-tail-calls"="false" "less-precise-fpmad"="false" "frame-pointer"="all" "no-infs-fp-math"="false" "no-nans-fp-math"="false" "stack-protector-buffer-size"="8" "use-soft-float"="false" }
!llvm.dbg.cu = !{!0}
!llvm.module.flags = !{!9, !10}
diff --git a/clang/test/CodeGen/fp-function-attrs.cpp b/clang/test/CodeGen/fp-function-attrs.cpp
index e92c26cdb75f7..3775bd5452d78 100644
--- a/clang/test/CodeGen/fp-function-attrs.cpp
+++ b/clang/test/CodeGen/fp-function-attrs.cpp
@@ -36,7 +36,7 @@ float test_reassociate_off_pragma(float a, float b, float c) {
return tmp;
}
-// CHECK: define{{.*}} float @_Z27test_reassociate_off_pragmafff(float noundef nofpclass(nan inf) %a, float noundef nofpclass(nan inf) %b, float noundef nofpclass(nan inf) %c) [[NO_UNSAFE_ATTRS:#[0-9]+]]
+// CHECK: define{{.*}} float @_Z27test_reassociate_off_pragmafff(float noundef nofpclass(nan inf) %a, float noundef nofpclass(nan inf) %b, float noundef nofpclass(nan inf) %c)
// CHECK: fadd nnan ninf nsz arcp contract afn float {{%.+}}, {{%.+}}
// CHECK: fadd fast float {{%.+}}, {{%.+}}
@@ -49,10 +49,9 @@ float test_contract_on_pragma(float a, float b, float c) {
return tmp;
}
-// CHECK: define{{.*}} float @_Z23test_contract_on_pragmafff(float noundef nofpclass(nan inf) %a, float noundef nofpclass(nan inf) %b, float noundef nofpclass(nan inf) %c) [[NO_UNSAFE_ATTRS:#[0-9]+]]
+// CHECK: define{{.*}} float @_Z23test_contract_on_pragmafff(float noundef nofpclass(nan inf) %a, float noundef nofpclass(nan inf) %b, float noundef nofpclass(nan inf) %c)
// CHECK: fmul fast float {{%.+}}, {{%.+}}
// CHECK: fadd reassoc nnan ninf nsz arcp afn float {{%.+}}, {{%.+}}
-// CHECK: attributes [[FAST_ATTRS]] = { {{.*}}"no-infs-fp-math"="true" {{.*}}"no-nans-fp-math"="true" "no-signed-zeros-fp-math"="true" {{.*}}"unsafe-fp-math"="true"{{.*}} }
-// CHECK: attributes [[PRECISE_ATTRS]] = { {{.*}}"no-infs-fp-math"="false" {{.*}}"no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false" {{.*}}"unsafe-fp-math"="false"{{.*}} }
-// CHECK: attributes [[NO_UNSAFE_ATTRS]] = { {{.*}}"no-infs-fp-math"="true" {{.*}}"no-nans-fp-math"="true" "no-signed-zeros-fp-math"="true" {{.*}}"unsafe-fp-math"="false"{{.*}} }
+// CHECK: attributes [[FAST_ATTRS]] = { {{.*}}"no-infs-fp-math"="true" {{.*}}"no-nans-fp-math"="true" "no-signed-zeros-fp-math"="true"{{.*}} }
+// CHECK: attributes [[PRECISE_ATTRS]] = { {{.*}}"no-infs-fp-math"="false" {{.*}}"no-nans-fp-math"="false" "no-signed-zeros-fp-math"="false"{{.*}} }
diff --git a/clang/test/CodeGen/func-attr.c b/clang/test/CodeGen/func-attr.c
index 1b36c51b8ce9c..42b7f1c785a64 100644
--- a/clang/test/CodeGen/func-attr.c
+++ b/clang/test/CodeGen/func-attr.c
@@ -1,18 +1,18 @@
// RUN: %clang_cc1 -triple x86_64-linux-gnu -ffast-math \
// RUN: -ffp-contract=fast -emit-llvm -o - %s | \
-// RUN: FileCheck %s --check-prefixes=CHECK,CHECK-UNSAFE,FINITEONLY
+// RUN: FileCheck %s --check-prefixes=CHECK,FINITEONLY
// RUN: %clang_cc1 -triple x86_64-linux-gnu -funsafe-math-optimizations \
// RUN: -ffp-contract=fast -emit-llvm -o - %s | \
-// RUN: FileCheck %s --check-prefixes=CHECK,CHECK-UNSAFE,NOFINITEONLY
+// RUN: FileCheck %s --check-prefixes=CHECK,NOFINITEONLY
// RUN: %clang_cc1 -triple x86_64-linux-gnu -funsafe-math-optimizations \
// RUN: -ffp-contract=on -emit-llvm -o - %s | \
-// RUN: FileCheck %s --check-prefixes=CHECK,CHECK-NOUNSAFE,NOFINITEONLY
+// RUN: FileCheck %s --check-prefixes=CHECK,NOFINITEONLY
// RUN: %clang_cc1 -triple x86_64-linux-gnu -funsafe-math-optimizations \
// RUN: -ffp-contract=off -emit-llvm -o - %s | \
-// RUN: FileCheck %s --check-prefixes=CHECK,CHECK-NOUNSAFE,NOFINITEONLY
+// RUN: FileCheck %s --check-prefixes=CHECK,NOFINITEONLY
float foo(float a, float b) {
return a+b;
@@ -24,6 +24,4 @@ float foo(float a, float b) {
// CHECK: attributes [[ATTRS]] = {
// CHECK-SAME: "no-signed-zeros-fp-math"="true"
// CHECK-SAME: "no-trapping-math"="true"
-// CHECK-UNSAFE-SAME: "unsafe-fp-math"="true"
-// CHECK-NOUNSAFE-NOT: "unsafe-fp-math"="true"
// CHECK-SAME: }
diff --git a/clang/test/CodeGenCUDA/propagate-attributes.cu b/clang/test/CodeGenCUDA/propagate-attributes.cu
index 6dfd44487d1dc..a7e6b09ff97db 100644
--- a/clang/test/CodeGenCUDA/propagate-attributes.cu
+++ b/clang/test/CodeGenCUDA/propagate-attributes.cu
@@ -12,18 +12,17 @@
// RUN: %clang_cc1 -x cuda %s -emit-llvm -mlink-builtin-bitcode %t.bc -o - \
// RUN: -fcuda-is-device -triple nvptx-unknown-unknown \
-// RUN: | FileCheck %s --check-prefix=CHECK --check-prefix=NOFTZ --check-prefix=NOFAST
+// RUN: | FileCheck %s --check-prefix=CHECK --check-prefix=NOFTZ
// RUN: %clang_cc1 -x cuda %s -emit-llvm -mlink-builtin-bitcode %t.bc \
// RUN: -fdenormal-fp-math-f32=preserve-sign -o - \
// RUN: -fcuda-is-device -triple nvptx-unknown-unknown \
-// RUN: | FileCheck %s --check-prefix=CHECK --check-prefix=FTZ \
-// RUN: --check-prefix=NOFAST
+// RUN: | FileCheck %s --check-prefix=CHECK --check-prefix=FTZ
// RUN: %clang_cc1 -x cuda %s -emit-llvm -mlink-builtin-bitcode %t.bc \
// RUN: -fdenormal-fp-math-f32=preserve-sign -o - \
// RUN: -fcuda-is-device -funsafe-math-optimizations -triple nvptx-unknown-unknown \
-// RUN: | FileCheck %s --check-prefix=CHECK --check-prefix=FAST
+// RUN: | FileCheck %s --check-prefix=CHECK
#ifndef LIB
#include "Inputs/cuda.h"
@@ -65,9 +64,6 @@ __global__ void kernel() { lib_fn(); }
// CHECK-SAME: "no-trapping-math"="true"
-// FAST-SAME: "unsafe-fp-math"="true"
-// NOFAST-NOT: "unsafe-fp-math"="true"
-
// Check the attribute list for lib_fn.
// CHECK: attributes [[fattr]] = {
@@ -81,6 +77,3 @@ __global__ void kernel() { lib_fn(); }
// NOFTZ-NOT: "denormal-fp-math-f32"
// CHECK-SAME: "no-trapping-math"="true"
-
-// FAST-SAME: "unsafe-fp-math"="true"
-// NOFAST-NOT: "unsafe-fp-math"="true"
diff --git a/clang/test/CodeGenOpenCL/relaxed-fpmath.cl b/clang/test/CodeGenOpenCL/relaxed-fpmath.cl
index a5f0019dbc1e5..c113d23c3028d 100644
--- a/clang/test/CodeGenOpenCL/relaxed-fpmath.cl
+++ b/clang/test/CodeGenOpenCL/relaxed-fpmath.cl
@@ -33,37 +33,31 @@ float spscalardiv(float a, float b) {
// NORMAL-NOT: "no-infs-fp-math"
// NORMAL-NOT: "no-nans-fp-math"
// NORMAL-NOT: "no-signed-zeros-fp-math"
-// NORMAL-NOT: "unsafe-fp-math"
// FAST: "less-precise-fpmad"="true"
// FAST: "no-infs-fp-math"="true"
// FAST: "no-nans-fp-math"="true"
// FAST: "no-signed-zeros-fp-math"="true"
-// FAST: "unsafe-fp-math"="true"
// FINITE-NOT: "less-precise-fpmad"
// FINITE: "no-infs-fp-math"="true"
// FINITE: "no-nans-fp-math"="true"
// FINITE-NOT: "no-signed-zeros-fp-math"
-// FINITE-NOT: "unsafe-fp-math"
// UNSAFE: "less-precise-fpmad"="true"
// UNSAFE-NOT: "no-infs-fp-math"
// UNSAFE-NOT: "no-nans-fp-math"
// UNSAFE: "no-signed-zeros-fp-math"="true"
-// UNSAFE: "unsafe-fp-math"="true"
// MAD: "less-precise-fpmad"="true"
// MAD-NOT: "no-infs-fp-math"
// MAD-NOT: "no-nans-fp-math"
// MAD-NOT: "no-signed-zeros-fp-math"
-// MAD-NOT: "unsafe-fp-math"
// NOSIGNED-NOT: "less-precise-fpmad"
// NOSIGNED-NOT: "no-infs-fp-math"
// NOSIGNED-NOT: "no-nans-fp-math"
// NOSIGNED: "no-signed-zeros-fp-math"="true"
-// NOSIGNED-NOT: "unsafe-fp-math"
#else
// Undefine this to avoid putting it in the PCH.
|
arsenm
approved these changes
Oct 10, 2025
|
I'm very happy to see progress being made on transitioning away from the function attributes for fast-math control. That said, it looks like there are still some things in the PowerPC ISel code that were using this attribute (by way of the UnsafeFPMath setting in TargetOptions). Do you have plans to remove the backend handling of this attribute? |
|
@RolandF77 Are you a good person to ask about the uses of |
PowerPC part is in #154901, but it needs fast-math flags from uitofp, which is impossible in current implementation, because of nneg flag. |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 participants
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.
These global flags block furthur improvements for clang, users should always use fast-math flags
see also https://discourse.llvm.org/t/rfc-honor-pragmas-with-ffp-contract-fast/80797
Remove them incrementally, this is the clang part.