[PATCH] [X86] Declare SSE4.1/AVX2 vector extloads covered by the *PMOV*X instructions legal

Quentin Colombet qcolombet at apple.com
Tue Jan 20 14:25:51 PST 2015


Hi Ahmed,

This looks mostly good but I would have expected more test cases for the added patterns.
Moreover, I have one concern on one of the test case. See my inline comment.

Thanks,
-Quentin


================
Comment at: test/CodeGen/X86/widen_load-2.ll:195
@@ +194,3 @@
+; CHECK-NEXT:    movl    (%[[PTR0]]), [[TMP1:%e[abcd]+x]]
+; CHECK-NEXT:    movl    [[TMP1]], [[TMP2:.*]]
+; CHECK-NEXT:    pmovzxbd [[TMP2]], %[[X0:xmm[0-9]+]]
----------------
Why do we need two moves here?

http://reviews.llvm.org/D6533

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






More information about the llvm-commits mailing list