[llvm-commits] [llvm] r82442 - /llvm/trunk/lib/Target/README.txt

[Chris Lattner]

Author: lattner
Date: Mon Sep 21 01:04:07 2009
New Revision: 82442

URL: http://llvm.org/viewvc/llvm-project?rev=82442&view=rev
Log:
add a note

Modified:
    llvm/trunk/lib/Target/README.txt

Modified: llvm/trunk/lib/Target/README.txt
URL: http://llvm.org/viewvc/llvm-project/llvm/trunk/lib/Target/README.txt?rev=82442&r1=82441&r2=82442&view=diff

==============================================================================
--- llvm/trunk/lib/Target/README.txt (original)
+++ llvm/trunk/lib/Target/README.txt Mon Sep 21 01:04:07 2009
@@ -220,7 +220,20 @@
 ... which would only do one 32-bit XOR per loop iteration instead of two.
 
 It would also be nice to recognize the reg->size doesn't alias reg->node[i], but
-alas...
+alas.
+
+//===---------------------------------------------------------------------===//
+
+This should be optimized to one 'and' and one 'or', from PR4216:
+
+define i32 @test_bitfield(i32 %bf.prev.low) nounwind ssp {
+entry:
+  %bf.prev.lo.cleared10 = or i32 %bf.prev.low, 32962 ; <i32> [#uses=1]
+  %0 = and i32 %bf.prev.low, -65536               ; <i32> [#uses=1]
+  %1 = and i32 %bf.prev.lo.cleared10, 40186       ; <i32> [#uses=1]
+  %2 = or i32 %1, %0                              ; <i32> [#uses=1]
+  ret i32 %2
+}
 
 //===---------------------------------------------------------------------===//