llvm.noalias - Look through noalias for shouldAlignPointerArgs

[hfinkel]

Summary

This is part of the series started by D9375. When determining whether or not we should align memory likely used by memcpy, etc., look through llvm.noalias.

[john.brawn]

LGTM, assuming that the conclusion of D9383 is that stripAndAccumulateInBoundsConstantOffsets takes a LookThroughNoAlias argument and that it doesn't default to true.

[rengolin]

Hal,

This doesn't seem to have been committed. Just checking for stale reviews on my list...

--renato

[hfinkel]

I'm in the process of updating these locally, I'll post new versions soon.

Thanks again!

Sent from my Verizon Wireless 4G LTE DROID

[rengolin]

Still stale... abandon?

[hfinkel]

In D9385#467821, @rengolin wrote:

Still stale... abandon?

I'll rebase these again soon.

[hfinkel]

This change should not be needed if noalias intrinsics are stripped by default.

lib/CodeGen/CodeGenPrepare.cpp

for (auto &Arg : CI->arg_operands()) {
       // We want to align both objects whose address is used directly and
       // objects whose address is used in casts and GEPs, though it only makes
       // sense for GEPs if the offset is a multiple of the desired alignment and
       // if size - offset meets the size threshold.
       if (!Arg->getType()->isPointerTy())
         continue;
       APInt Offset(TD->getPointerSizeInBits(
                      cast<PointerType>(Arg->getType())->getAddressSpace()), 0);
-      Value *Val = Arg->stripAndAccumulateInBoundsConstantOffsets(*TD, Offset);
+      Value *Val = Arg->stripAndAccumulateInBoundsConstantOffsets(*TD, Offset,
+                          /*LookThroughNoAlias*/true);
       uint64_t Offset2 = Offset.getLimitedValue();
       if ((Offset2 & (PrefAlign-1)) != 0)
         continue;
       AllocaInst *AI;
       if ((AI = dyn_cast<AllocaInst>(Val)) &&
           AI->getAlignment() < PrefAlign &&
           TD->getTypeAllocSize(AI->getAllocatedType()) >= MinSize + Offset2)
         AI->setAlignment(PrefAlign);

test/CodeGen/ARM/memfunc.ll

entry:
   ; CHECK-EABI: __aeabi_memset
   %arr2 = alloca [9 x i8], align 1
   %2 = bitcast [9 x i8]* %arr2 to i8*
   call void @llvm.memset.p0i8.i32(i8* %2, i8 0, i32 %n, i32 0, i1 false)
 
   unreachable
 }
 
+; From @f2 above, but make sure that a noalias call does not prevent the
+; optimization.
+define void @f2a(i8* %dest, i32 %n) {
+entry:
+  ; CHECK-LABEL: f2a
+
+  ; IOS (ARMv7) should 8-byte align, others should 4-byte align
+  ; CHECK-IOS: add r1, sp, #32
+  ; CHECK-IOS: memmove
+  ; CHECK-DARWIN: add r1, sp, #28
+  ; CHECK-DARWIN: memmove
+  ; CHECK-EABI: add r1, sp, #28
+  ; CHECK-EABI: __aeabi_memmove
+  %arr0 = alloca [9 x i8], align 1
+  %0 = bitcast [9 x i8]* %arr0 to i8*
+  %q = call i8* @llvm.noalias.p0i8(i8* %0, metadata !1)
+  call void @llvm.memmove.p0i8.p0i8.i32(i8* %dest, i8* %q, i32 %n, i32 0, i1 false)
+
+  ; CHECK: add r1, sp, #16
+  ; CHECK-IOS: memcpy
+  ; CHECK-DARWIN: memcpy
+  ; CHECK-EABI: __aeabi_memcpy
+  %arr1 = alloca [9 x i8], align 1
+  %1 = bitcast [9 x i8]* %arr1 to i8*
+  %r = call i8* @llvm.noalias.p0i8(i8* %1, metadata !1)
+  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %r, i32 %n, i32 0, i1 false)
+
+  ; CHECK-IOS: mov r0, sp
+  ; CHECK-IOS: mov r1, #0
+  ; CHECK-IOS: memset
+  ; CHECK-DARINW: add r0, sp, #4
+  ; CHECK-DARWIN: movs r1, #0
+  ; CHECK-DARWIN: memset
+  ; CHECK-EABI: add r0, sp, #4
+  ; CHECK-EABI: mov r2, #0
+  ; CHECK-EABI: __aeabi_memset
+  %arr2 = alloca [9 x i8], align 1
+  %2 = bitcast [9 x i8]* %arr2 to i8*
+  %s = call i8* @llvm.noalias.p0i8(i8* %2, metadata !1)
+  call void @llvm.memset.p0i8.i32(i8* %s, i8 0, i32 %n, i32 0, i1 false)
+
+  unreachable
+}
+
 ; Check that alloca arguments are not aligned if less than 8 bytes in size
 define void @f3(i8* %dest, i32 %n) {
 entry:
   ; CHECK-LABEL: f3
 
   ; CHECK: {{add(.w)? r1, sp, #17|sub(.w)? r1, r7, #15}}
   ; CHECK-IOS: memmove
   ; CHECK-DARWIN: memmove

 ; CHECK: arr5:
 ; CHECK-NOT: .align
 ; CHECK: arr6:
 ; CHECK-NOT: arr7:
 
 declare void @llvm.memmove.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i32, i1) nounwind
 declare void @llvm.memset.p0i8.i32(i8* nocapture, i8, i32, i32, i1) nounwind
+
+declare i8* @llvm.noalias.p0i8(i8*, metadata) nounwind
+
+!0 = !{!0, !"some domain"}
+!1 = !{!1, !0, !"some scope"}
+