From 5cc750ecee73807fad1870ee226428be05a082f5 Mon Sep 17 00:00:00 2001
From: Matt Pharr <matt.pharr@intel.com>
Date: Wed, 6 Jul 2011 07:37:35 +0100
Subject: [PATCH] Add comment re popcnt on SSE2 target.

---
 stdlib-sse2.ll | 8 ++++++++
 1 file changed, 8 insertions(+)

diff --git a/stdlib-sse2.ll b/stdlib-sse2.ll
index c37fdfb5..d1573e21 100644
--- a/stdlib-sse2.ll
+++ b/stdlib-sse2.ll
@@ -244,6 +244,14 @@ define internal i32 @__max_uniform_uint32(i32, i32) nounwind readonly alwaysinli
 
 ; FIXME: this is very inefficient, loops over all 32 bits...
 
+; we could use the LLVM intrinsic declare i32 @llvm.ctpop.i32(i32),
+; although that currently ends up generating a POPCNT instruction even
+; if we give --target=sse2 on the command line.  We probably need to
+; pipe through the 'sse2' request to LLVM via the 'features' string
+; at codegen time...  (If e.g. --cpu=penryn is also passed along, then
+; it does generate non-POPCNT code and in particular better code than
+; the below does.)
+
 define internal i32 @__popcnt(i32) nounwind readonly alwaysinline {
 entry:
   br label %loop