Add a note about efficient codegen for binary log.

llvm-svn: 135996

1 files changed, 48 insertions, 0 deletions

diff --git a/llvm/lib/Target/X86/README.txt b/llvm/lib/Target/X86/README.txt
index 560947a4a04..d6ceaa7ed4b 100644
--- a/llvm/lib/Target/X86/README.txt
+++ b/llvm/lib/Target/X86/README.txt
@@ -2066,3 +2066,51 @@ The trick is to match "fetch_and_add(X, -C) == C".
 
 //===---------------------------------------------------------------------===//
 
+unsigned log2(unsigned x) {
+  return x > 1 ? 32-__builtin_clz(x-1) : 0;
+}
+
+generates (x86_64):
+	xorl	%eax, %eax
+	cmpl	$2, %edi
+	jb	LBB0_2
+## BB#1:
+	decl	%edi
+	movl	$63, %eax
+	bsrl	%edi, %ecx
+	cmovel	%eax, %ecx
+	xorl	$31, %ecx
+	movl	$32, %eax
+	subl	%ecx, %eax
+LBB0_2:
+	ret
+
+The cmov and the early test are redundant:
+	xorl	%eax, %eax
+	cmpl	$2, %edi
+	jb	LBB0_2
+## BB#1:
+	decl	%edi
+	bsrl	%edi, %ecx
+	xorl	$31, %ecx
+	movl	$32, %eax
+	subl	%ecx, %eax
+LBB0_2:
+	ret
+
+If we want to get really fancy we could use some two's complement magic:
+	xorl	%eax, %eax
+	cmpl	$2, %edi
+	jb	LBB0_2
+## BB#1:
+	decl	%edi
+	bsrl	%edi, %ecx
+	xorl	$-32, %ecx
+	leal    33(%ecx), %eax
+LBB0_2:
+	ret
+
+This is only useful on targets that can't encode the first operand of a sub
+directly.  The rule is C1 - (X^C2) -> (C1+1) + (X^~C2).
+
+//===---------------------------------------------------------------------===//