diff options
author | Chris Lattner <sabre@nondot.org> | 2008-04-27 00:37:18 +0000 |
---|---|---|
committer | Chris Lattner <sabre@nondot.org> | 2008-04-27 00:37:18 +0000 |
commit | 5a09690446a36f94f990db7d18d9b9ac8587888a (patch) | |
tree | 4ea3e1952e67a7ec106e63b63ad6996a7f3d6ab6 /test/CodeGen/X86/2008-04-26-Asm-Optimize-Imm.ll | |
parent | 507ffd2423cfd0f5204c69e44c50aa3c759609d8 (diff) | |
download | llvm-5a09690446a36f94f990db7d18d9b9ac8587888a.tar.gz llvm-5a09690446a36f94f990db7d18d9b9ac8587888a.tar.bz2 llvm-5a09690446a36f94f990db7d18d9b9ac8587888a.tar.xz |
Implement a signficant optimization for inline asm:
When choosing between constraints with multiple options,
like "ir", test to see if we can use the 'i' constraint and
go with that if possible. This produces more optimal ASM in
all cases (sparing a register and an instruction to load it),
and fixes inline asm like this:
void test () {
asm volatile (" %c0 %1 " : : "imr" (42), "imr"(14));
}
Previously we would dump "42" into a memory location (which
is ok for the 'm' constraint) which would cause a problem
because the 'c' modifier is not valid on memory operands.
Isn't it great how inline asm turns 'missed optimization'
into 'compile failed'??
Incidentally, this was the todo in
PowerPC/2007-04-24-InlineAsm-I-Modifier.ll
Please do NOT pull this into Tak.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@50315 91177308-0d34-0410-b5e6-96231b3b80d8
Diffstat (limited to 'test/CodeGen/X86/2008-04-26-Asm-Optimize-Imm.ll')
-rw-r--r-- | test/CodeGen/X86/2008-04-26-Asm-Optimize-Imm.ll | 11 |
1 files changed, 11 insertions, 0 deletions
diff --git a/test/CodeGen/X86/2008-04-26-Asm-Optimize-Imm.ll b/test/CodeGen/X86/2008-04-26-Asm-Optimize-Imm.ll new file mode 100644 index 0000000000..82721a53b8 --- /dev/null +++ b/test/CodeGen/X86/2008-04-26-Asm-Optimize-Imm.ll @@ -0,0 +1,11 @@ +; RUN: llvm-as < %s | llc | grep {1 \$2 3} +; rdar://5720231 +target datalayout = "e-p:32:32:32-i1:8:8-i8:8:8-i16:16:16-i32:32:32-i64:32:64-f32:32:32-f64:32:64-v64:64:64-v128:128:128-a0:0:64-f80:128:128" +target triple = "i386-apple-darwin8" + +define void @test() nounwind { +entry: + tail call void asm sideeffect " ${0:c} $1 ${2:c} ", "imr,imr,i,~{dirflag},~{fpsr},~{flags}"( i32 1, i32 2, i32 3 ) nounwind + ret void +} + |