[PATCH] D108844: [MipsISelLowering] avoid emitting libcalls to __mulodi4()

Nick Desaulniers via Phabricator via llvm-commits llvm-commits at lists.llvm.org
Fri Aug 27 14:10:28 PDT 2021


nickdesaulniers created this revision.
nickdesaulniers added reviewers: craig.topper, rengolin, atanasyan.
Herald added subscribers: jrtc27, hiraditya, arichardson, sdardis.
nickdesaulniers requested review of this revision.
Herald added a project: LLVM.
Herald added a subscriber: llvm-commits.

__has_builtin(__builtin_mul_overflow) returns true for 32b MIPS targets,
but Clang is deferring to compiler RT when encountering `long long`
types. This breaks sanitizer builds of the Linux kernel that are using
__builtin_mul_overflow with these types for these targets.

If the semantics of __has_builtin mean "the compiler resolves these,
always" then we shouldn't conditionally emit a libcall.

This will still need to be worked around in the Linux kernel in order to
continue to support malta_defconfig builds of the Linux kernel for this
target with older releases of clang.

Link: https://bugs.llvm.org/show_bug.cgi?id=28629
Link: https://github.com/ClangBuiltLinux/linux/issues/1438


Repository:
  rG LLVM Github Monorepo

https://reviews.llvm.org/D108844

Files:
  llvm/lib/Target/Mips/MipsISelLowering.cpp
  llvm/test/CodeGen/Mips/overflow-intrinsic-optimizations.ll


Index: llvm/test/CodeGen/Mips/overflow-intrinsic-optimizations.ll
===================================================================
--- /dev/null
+++ llvm/test/CodeGen/Mips/overflow-intrinsic-optimizations.ll
@@ -0,0 +1,28 @@
+; RUN: llc %s -mtriple=mipsel -o - | FileCheck %s
+
+define void @no__mulodi4(i32 %a, i64 %b, i32* %c) {
+; CHECK-LABEL: no__mulodi4
+; CHECK-NOT: jal __mulodi4
+entry:
+  %a.addr = alloca i32, align 4
+  %b.addr = alloca i64, align 8
+  %c.addr = alloca i32*, align 4
+  store i32 %a, i32* %a.addr, align 4
+  store i64 %b, i64* %b.addr, align 8
+  store i32* %c, i32** %c.addr, align 4
+  %0 = load i32, i32* %a.addr, align 4
+  %1 = load i64, i64* %b.addr, align 8
+  %2 = load i32*, i32** %c.addr, align 4
+  %3 = sext i32 %0 to i64
+  %4 = call { i64, i1 } @llvm.smul.with.overflow.i64(i64 %3, i64 %1)
+  %5 = extractvalue { i64, i1 } %4, 1
+  %6 = extractvalue { i64, i1 } %4, 0
+  %7 = trunc i64 %6 to i32
+  %8 = sext i32 %7 to i64
+  %9 = icmp ne i64 %6, %8
+  %10 = or i1 %5, %9
+  store i32 %7, i32* %2, align 4
+  ret void
+}
+
+declare { i64, i1 } @llvm.smul.with.overflow.i64(i64, i64)
Index: llvm/lib/Target/Mips/MipsISelLowering.cpp
===================================================================
--- llvm/lib/Target/Mips/MipsISelLowering.cpp
+++ llvm/lib/Target/Mips/MipsISelLowering.cpp
@@ -509,6 +509,7 @@
     setLibcallName(RTLIB::SHL_I128, nullptr);
     setLibcallName(RTLIB::SRL_I128, nullptr);
     setLibcallName(RTLIB::SRA_I128, nullptr);
+    setLibcallName(RTLIB::MULO_I64, nullptr);
   }
 
   setMinFunctionAlignment(Subtarget.isGP64bit() ? Align(8) : Align(4));


-------------- next part --------------
A non-text attachment was scrubbed...
Name: D108844.369182.patch
Type: text/x-patch
Size: 1626 bytes
Desc: not available
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20210827/7839872e/attachment.bin>


More information about the llvm-commits mailing list