Use rsqrt (X86) to speed up reciprocal square root calcs

[oota-llvm.git] / lib / Target / X86 / README-UNIMPLEMENTED.txt

diff --git a/lib/Target/X86/README-UNIMPLEMENTED.txt b/lib/Target/X86/README-UNIMPLEMENTED.txt
index b9ba4937792c82fa465e8e917e5cfbb92558cdae..c26c75ab951ca348593786a4ab905e8037da2dda 100644 (file)
--- a/lib/Target/X86/README-UNIMPLEMENTED.txt
+++ b/lib/Target/X86/README-UNIMPLEMENTED.txt
@@ -9,6 +9,6 @@ which would be great.
 
 1) vector shifts
 2) vector comparisons
-3) vector fp<->int comparisons: PR2683, PR2684, PR2685, PR2686, PR2688
-
-
+3) vector fp<->int conversions: PR2683, PR2684, PR2685, PR2686, PR2688
+4) bitcasts from vectors to scalars: PR2804
+5) llvm.atomic.cmp.swap.i128.p0i128: PR3462