[PATCH] Add intrinsic definitions for unary op AVX instructions [x86]

Sanjay Patel spatel at rotateright.com
Thu Jan 22 10:49:05 PST 2015


================
Comment at: lib/Target/X86/X86InstrSSE.td:3810
@@ -3799,5 +3809,3 @@
   def : Pat<(int_x86_sse2_sqrt_sd VR128:$src),
-            (COPY_TO_REGCLASS (VSQRTSDr (f64 (IMPLICIT_DEF)),
-                                        (COPY_TO_REGCLASS VR128:$src, FR64)),
-                              VR128)>;
+            (VSQRTSDr (v2f64 (IMPLICIT_DEF)), VR128:$src)>;
   def : Pat<(int_x86_sse2_sqrt_sd sse_load_f64:$src),
----------------
RKSimon wrote:
> Since you've added it above shouldn't this be VSQRTSDr_Int ?
Good eye - yes, that was a bug in the uploaded patch that I had corrected locally.

http://reviews.llvm.org/D6958

EMAIL PREFERENCES
  http://reviews.llvm.org/settings/panel/emailpreferences/






More information about the llvm-commits mailing list