b1d47467e2
This commit removes the artificial types <512 x i1> and <1024 x i1>
from HVX intrinsics, and makes v512i1 and v1024i1 no longer legal on
Hexagon.
It may cause existing bitcode files to become invalid.
* Converting between vector predicates and vector registers must be
done explicitly via vandvrt/vandqrt instructions (their intrinsics),
i.e. (for 64-byte mode):
%Q = call <64 x i1> @llvm.hexagon.V6.vandvrt(<16 x i32> %V, i32 -1)
%V = call <16 x i32> @llvm.hexagon.V6.vandqrt(<64 x i1> %Q, i32 -1)
The conversion intrinsics are:
declare <64 x i1> @llvm.hexagon.V6.vandvrt(<16 x i32>, i32)
declare <128 x i1> @llvm.hexagon.V6.vandvrt.128B(<32 x i32>, i32)
declare <16 x i32> @llvm.hexagon.V6.vandqrt(<64 x i1>, i32)
declare <32 x i32> @llvm.hexagon.V6.vandqrt.128B(<128 x i1>, i32)
They are all pure.
* Vector predicate values cannot be loaded/stored directly. This directly
reflects the architecture restriction. Loading and storing or vector
predicates must be done indirectly via vector registers and explicit
conversions via vandvrt/vandqrt instructions.
23 lines
754 B
LLVM
23 lines
754 B
LLVM
; RUN: llc -march=hexagon < %s | FileCheck %s
|
|
|
|
; Check that constraints q and v are handled correctly.
|
|
; CHECK: q{{.}} = vgtw(v{{.}}.w,v{{.}}.w)
|
|
; CHECK: vand
|
|
; CHECK: vmem
|
|
|
|
target triple = "hexagon"
|
|
|
|
; Function Attrs: nounwind
|
|
define void @foo(<16 x i32> %v0, <16 x i32> %v1, <16 x i32>* nocapture %p) #0 {
|
|
entry:
|
|
%0 = tail call <64 x i1> asm "$0 = vgtw($1.w,$2.w)", "=q,v,v"(<16 x i32> %v0, <16 x i32> %v1) #1
|
|
%1 = tail call <16 x i32> @llvm.hexagon.V6.vandqrt(<64 x i1> %0, i32 -1) #1
|
|
store <16 x i32> %1, <16 x i32>* %p, align 64
|
|
ret void
|
|
}
|
|
|
|
declare <16 x i32> @llvm.hexagon.V6.vandqrt(<64 x i1>, i32) #1
|
|
|
|
attributes #0 = { nounwind "target-cpu"="hexagonv60" "target-features"="+hvxv60,+hvx-length64b" }
|
|
attributes #1 = { nounwind readnone }
|