Start a README.txt of possible optimizations to do in IRgen.

llvm-svn: 59130

1 files changed, 24 insertions, 0 deletions

diff --git a/clang/lib/CodeGen/README.txt b/clang/lib/CodeGen/README.txt
new file mode 100644
index 00000000000..33fbfdb744a
--- /dev/null
+++ b/clang/lib/CodeGen/README.txt
@@ -0,0 +1,24 @@
+IRgen optimization opportunities.
+
+//===---------------------------------------------------------------------===//
+
+The common pattern of
+--
+short x; // or char, etc
+(x == 10)
+--
+generates an zext/sext of x which can easily be avoided.
+
+//===---------------------------------------------------------------------===//
+
+Bitfields accesses can be shifted to simplify masking and sign
+extension. For example, if the bitfield width is 8 and it is
+appropriately aligned then is is a lot shorter to just load the char
+directly.
+
+//===---------------------------------------------------------------------===//
+
+Bitfields should not reload the stored value just to return the
+correct result.
+
+//===---------------------------------------------------------------------===//