[X86][BF16] Add libcall for F80 -> BF16 #109116
Merged
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
This fixes llvm#108936, but the calling convention doesn't match with GCC. I doubt we have such a lib function for now, so leave the calling convention as is.
|
@llvm/pr-subscribers-llvm-ir @llvm/pr-subscribers-backend-x86 Author: Phoebe Wang (phoebewang) ChangesThis fixes #108936, but the calling convention doesn't match with GCC. I doubt we have such a lib function for now, so leave the calling convention as is. Full diff: https://github.com/llvm/llvm-project/pull/109116.diff 3 Files Affected:
diff --git a/llvm/include/llvm/IR/RuntimeLibcalls.def b/llvm/include/llvm/IR/RuntimeLibcalls.def
index c3d5ef9f4e4f82..69cf43140ad4bd 100644
--- a/llvm/include/llvm/IR/RuntimeLibcalls.def
+++ b/llvm/include/llvm/IR/RuntimeLibcalls.def
@@ -367,6 +367,7 @@ HANDLE_LIBCALL(FPROUND_F128_F16, "__trunctfhf2")
HANDLE_LIBCALL(FPROUND_PPCF128_F16, "__trunctfhf2")
HANDLE_LIBCALL(FPROUND_F32_BF16, "__truncsfbf2")
HANDLE_LIBCALL(FPROUND_F64_BF16, "__truncdfbf2")
+HANDLE_LIBCALL(FPROUND_F80_BF16, "__truncxfbf2")
HANDLE_LIBCALL(FPROUND_F64_F32, "__truncdfsf2")
HANDLE_LIBCALL(FPROUND_F80_F32, "__truncxfsf2")
HANDLE_LIBCALL(FPROUND_F128_F32, "__trunctfsf2")
diff --git a/llvm/lib/CodeGen/TargetLoweringBase.cpp b/llvm/lib/CodeGen/TargetLoweringBase.cpp
index eb3190c7cd247a..9fdde454559171 100644
--- a/llvm/lib/CodeGen/TargetLoweringBase.cpp
+++ b/llvm/lib/CodeGen/TargetLoweringBase.cpp
@@ -169,6 +169,8 @@ RTLIB::Libcall RTLIB::getFPROUND(EVT OpVT, EVT RetVT) {
return FPROUND_F32_BF16;
if (OpVT == MVT::f64)
return FPROUND_F64_BF16;
+ if (OpVT == MVT::f80)
+ return FPROUND_F80_BF16;
} else if (RetVT == MVT::f32) {
if (OpVT == MVT::f64)
return FPROUND_F64_F32;
diff --git a/llvm/test/CodeGen/X86/bfloat.ll b/llvm/test/CodeGen/X86/bfloat.ll
index 3759909a2ccc8e..3144fd56d9ccf3 100644
--- a/llvm/test/CodeGen/X86/bfloat.ll
+++ b/llvm/test/CodeGen/X86/bfloat.ll
@@ -1953,3 +1953,25 @@ define void @PR92471(ptr %0, ptr %1) nounwind {
store <7 x float> %4, ptr %1, align 4
ret void
}
+
+define bfloat @PR108936(x86_fp80 %0) nounwind {
+; X86-LABEL: PR108936:
+; X86: # %bb.0:
+; X86-NEXT: subl $12, %esp
+; X86-NEXT: fldt {{[0-9]+}}(%esp)
+; X86-NEXT: fstpt (%esp)
+; X86-NEXT: calll __truncxfbf2
+; X86-NEXT: addl $12, %esp
+; X86-NEXT: retl
+;
+; CHECK-LABEL: PR108936:
+; CHECK: # %bb.0:
+; CHECK-NEXT: subq $24, %rsp
+; CHECK-NEXT: fldt {{[0-9]+}}(%rsp)
+; CHECK-NEXT: fstpt (%rsp)
+; CHECK-NEXT: callq __truncxfbf2@PLT
+; CHECK-NEXT: addq $24, %rsp
+; CHECK-NEXT: retq
+ %2 = fptrunc x86_fp80 %0 to bfloat
+ ret bfloat %2
+}
|
d0k
reviewed
Sep 18, 2024
Sorry, I mixed up. We have the same behavior with GCC. GCC supports it since 13.1 https://godbolt.org/z/dnhaeGvf7 |
d0k
approved these changes
Sep 18, 2024
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/140/builds/6849 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/159/builds/6335 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/12/builds/6066 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/123/builds/5859 Here is the relevant piece of the build log for the reference |
phoebewang
added a commit
that referenced
this pull request
Sep 18, 2024
This reverts commit 76eda76.
phoebewang
added a commit
that referenced
this pull request
Sep 18, 2024
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/174/builds/5337 Here is the relevant piece of the build log for the reference |
phoebewang
added a commit
to phoebewang/llvm-project
that referenced
this pull request
Sep 18, 2024
This reverts commit ababfee78714313a0cad87591b819f0944b90d09.
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/8/builds/4662 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/73/builds/5720 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/11/builds/5213 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/32/builds/5156 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/7/builds/4512 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/153/builds/9289 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/113/builds/2862 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/4/builds/2238 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/143/builds/2242 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/56/builds/7705 Here is the relevant piece of the build log for the reference |
phoebewang
added a commit
that referenced
this pull request
Sep 19, 2024
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/146/builds/1188 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/128/builds/682 Here is the relevant piece of the build log for the reference |
tmsri
pushed a commit
to tmsri/llvm-project
that referenced
this pull request
Sep 19, 2024
This fixes llvm#108936, but the calling convention doesn't match with GCC. I doubt we have such a lib function for now, so leave the calling convention as is.
tmsri
pushed a commit
to tmsri/llvm-project
that referenced
this pull request
Sep 19, 2024
tmsri
pushed a commit
to tmsri/llvm-project
that referenced
this pull request
Sep 19, 2024
…109143) This reverts commit ababfee78714313a0cad87591b819f0944b90d09. Add X86 FP80 check.
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/133/builds/3999 Here is the relevant piece of the build log for the reference |
|
LLVM Buildbot has detected a new failure on builder Full details are available at: https://lab.llvm.org/buildbot/#/builders/81/builds/1960 Here is the relevant piece of the build log for the reference |
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.
This fixes #108936, but the calling convention doesn't match with GCC. I doubt we have such a lib function for now, so leave the calling convention as is.