[PATCH] [AArch64] Customer lowering of CTPOP to SIMD should check for NEON availability

[Tim Northover]

Hi,

The change itself looks reasonable, but I think the tests could be improved:

================
Comment at: test/CodeGen/AArch64/arm64-popcnt.ll:12
@@ -10,2 +11,3 @@
 ; CHECK: ret
+; CHECK-NONEON-NOT: 8b
 }
----------------
Can we check that what actually is produced is sensible? Purely negative tests are discouraged when alternatives are possible.

http://reviews.llvm.org/D6299