Added an idea about any_extend for performance tuning.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@24763 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Target/X86/README.txt b/lib/Target/X86/README.txt
index 6e23fca25c62a5b853b8bfd91deb05e01d5686b4..5fdcc696193a05c1e11e88014db61671ca0f02a1 100644 (file)
--- a/lib/Target/X86/README.txt
+++ b/lib/Target/X86/README.txt
@@ -102,3 +102,8 @@ Some isel ideas:
 //===---------------------------------------------------------------------===//
 
 Should we promote i16 to i32 to avoid partial register update stalls?
+
+//===---------------------------------------------------------------------===//
+
+Leave any_extend as pseudo instruction and hint to register
+allocator. Delay codegen until post register allocation.