RPCS3/llvm-mirror
1
0
Fork 0
mirror of https://github.com/RPCS3/llvm-mirror.git synced 2024-11-26 04:32:44 +01:00
Code Issues Projects Releases Wiki Activity
fca643847d
llvm-mirror/test/CodeGen/SystemZ/memcpy-01.ll
Daniel Neilson f59acc15ad Remove alignment argument from memcpy/memmove/memset in favour of alignment attributes (Step 1) 
Summary:
 This is a resurrection of work first proposed and discussed in Aug 2015:
   http://lists.llvm.org/pipermail/llvm-dev/2015-August/089384.html
and initially landed (but then backed out) in Nov 2015:
   http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20151109/312083.html

 The @llvm.memcpy/memmove/memset intrinsics currently have an explicit argument
which is required to be a constant integer. It represents the alignment of the
dest (and source), and so must be the minimum of the actual alignment of the
two.

 This change is the first in a series that allows source and dest to each
have their own alignments by using the alignment attribute on their arguments.

 In this change we:
1) Remove the alignment argument.
2) Add alignment attributes to the source & dest arguments. We, temporarily,
   require that the alignments for source & dest be equal.

 For example, code which used to read:
  call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 100, i32 4, i1 false)
will now read
  call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 4 %dest, i8* align 4 %src, i32 100, i1 false)

 Downstream users may have to update their lit tests that check for
@llvm.memcpy/memmove/memset call/declaration patterns. The following extended sed script
may help with updating the majority of your tests, but it does not catch all possible
patterns so some manual checking and updating will be required.

s~declare void @llvm\.mem(set|cpy|move)\.p([^(]*)\((.*), i32, i1\)~declare void @llvm.mem\1.p\2(\3, i1)~g
s~call void @llvm\.memset\.p([^(]*)i8\(i8([^*]*)\* (.*), i8 (.*), i8 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.memset.p\1i8(i8\2* \3, i8 \4, i8 \5, i1 \6)~g
s~call void @llvm\.memset\.p([^(]*)i16\(i8([^*]*)\* (.*), i8 (.*), i16 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.memset.p\1i16(i8\2* \3, i8 \4, i16 \5, i1 \6)~g
s~call void @llvm\.memset\.p([^(]*)i32\(i8([^*]*)\* (.*), i8 (.*), i32 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.memset.p\1i32(i8\2* \3, i8 \4, i32 \5, i1 \6)~g
s~call void @llvm\.memset\.p([^(]*)i64\(i8([^*]*)\* (.*), i8 (.*), i64 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.memset.p\1i64(i8\2* \3, i8 \4, i64 \5, i1 \6)~g
s~call void @llvm\.memset\.p([^(]*)i128\(i8([^*]*)\* (.*), i8 (.*), i128 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.memset.p\1i128(i8\2* \3, i8 \4, i128 \5, i1 \6)~g
s~call void @llvm\.memset\.p([^(]*)i8\(i8([^*]*)\* (.*), i8 (.*), i8 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.memset.p\1i8(i8\2* align \6 \3, i8 \4, i8 \5, i1 \7)~g
s~call void @llvm\.memset\.p([^(]*)i16\(i8([^*]*)\* (.*), i8 (.*), i16 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.memset.p\1i16(i8\2* align \6 \3, i8 \4, i16 \5, i1 \7)~g
s~call void @llvm\.memset\.p([^(]*)i32\(i8([^*]*)\* (.*), i8 (.*), i32 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.memset.p\1i32(i8\2* align \6 \3, i8 \4, i32 \5, i1 \7)~g
s~call void @llvm\.memset\.p([^(]*)i64\(i8([^*]*)\* (.*), i8 (.*), i64 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.memset.p\1i64(i8\2* align \6 \3, i8 \4, i64 \5, i1 \7)~g
s~call void @llvm\.memset\.p([^(]*)i128\(i8([^*]*)\* (.*), i8 (.*), i128 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.memset.p\1i128(i8\2* align \6 \3, i8 \4, i128 \5, i1 \7)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i8\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i8 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.mem\1.p\2i8(i8\3* \4, i8\5* \6, i8 \7, i1 \8)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i16\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i16 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.mem\1.p\2i16(i8\3* \4, i8\5* \6, i16 \7, i1 \8)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i32\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i32 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.mem\1.p\2i32(i8\3* \4, i8\5* \6, i32 \7, i1 \8)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i64\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i64 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.mem\1.p\2i64(i8\3* \4, i8\5* \6, i64 \7, i1 \8)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i128\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i128 (.*), i32 [01], i1 ([^)]*)\)~call void @llvm.mem\1.p\2i128(i8\3* \4, i8\5* \6, i128 \7, i1 \8)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i8\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i8 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.mem\1.p\2i8(i8\3* align \8 \4, i8\5* align \8 \6, i8 \7, i1 \9)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i16\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i16 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.mem\1.p\2i16(i8\3* align \8 \4, i8\5* align \8 \6, i16 \7, i1 \9)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i32\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i32 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.mem\1.p\2i32(i8\3* align \8 \4, i8\5* align \8 \6, i32 \7, i1 \9)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i64\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i64 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.mem\1.p\2i64(i8\3* align \8 \4, i8\5* align \8 \6, i64 \7, i1 \9)~g
s~call void @llvm\.mem(cpy|move)\.p([^(]*)i128\(i8([^*]*)\* (.*), i8([^*]*)\* (.*), i128 (.*), i32 ([0-9]*), i1 ([^)]*)\)~call void @llvm.mem\1.p\2i128(i8\3* align \8 \4, i8\5* align \8 \6, i128 \7, i1 \9)~g

 The remaining changes in the series will:
Step 2) Expand the IRBuilder API to allow creation of memcpy/memmove with differing
   source and dest alignments.
Step 3) Update Clang to use the new IRBuilder API.
Step 4) Update Polly to use the new IRBuilder API.
Step 5) Update LLVM passes that create memcpy/memmove calls to use the new IRBuilder API,
        and those that use use MemIntrinsicInst::[get|set]Alignment() to use
        getDestAlignment() and getSourceAlignment() instead.
Step 6) Remove the single-alignment IRBuilder API for memcpy/memmove, and the
        MemIntrinsicInst::[get|set]Alignment() methods.

Reviewers: pete, hfinkel, lhames, reames, bollu

Reviewed By: reames

Subscribers: niosHD, reames, jholewinski, qcolombet, jfb, sanjoy, arsenm, dschuff, dylanmckay, mehdi_amini, sdardis, nemanjai, david2050, nhaehnle, javed.absar, sbc100, jgravelle-google, eraman, aheejin, kbarton, JDevlieghere, asb, rbar, johnrusso, simoncook, jordy.potman.lists, apazos, sabuasal, llvm-commits

Differential Revision: https://reviews.llvm.org/D41675

llvm-svn: 322965
2018-01-19 17:13:12 +00:00

220 lines
6.8 KiB
LLVM
Raw Blame History

; Test memcpy using MVC.
;
; RUN: llc < %s -mtriple=s390x-linux-gnu | FileCheck %s
declare void @llvm.memcpy.p0i8.p0i8.i32(i8 *nocapture, i8 *nocapture, i32, i1) nounwind
declare void @llvm.memcpy.p0i8.p0i8.i64(i8 *nocapture, i8 *nocapture, i64, i1) nounwind
declare void @foo(i8 *, i8 *)
; Test a no-op move, i32 version.
define void @f1(i8* %dest, i8* %src) {
; CHECK-LABEL: f1:
; CHECK-NOT: %r2
; CHECK-NOT: %r3
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 0, i1 false)
ret void
}
; Test a no-op move, i64 version.
define void @f2(i8* %dest, i8* %src) {
; CHECK-LABEL: f2:
; CHECK-NOT: %r2
; CHECK-NOT: %r3
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 0, i1 false)
ret void
}
; Test a 1-byte move, i32 version.
define void @f3(i8* %dest, i8* %src) {
; CHECK-LABEL: f3:
; CHECK: mvc 0(1,%r2), 0(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 1, i1 false)
ret void
}
; Test a 1-byte move, i64 version.
define void @f4(i8* %dest, i8* %src) {
; CHECK-LABEL: f4:
; CHECK: mvc 0(1,%r2), 0(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 1, i1 false)
ret void
}
; Test the upper range of a single MVC, i32 version.
define void @f5(i8* %dest, i8* %src) {
; CHECK-LABEL: f5:
; CHECK: mvc 0(256,%r2), 0(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 256, i1 false)
ret void
}
; Test the upper range of a single MVC, i64 version.
define void @f6(i8* %dest, i8* %src) {
; CHECK-LABEL: f6:
; CHECK: mvc 0(256,%r2), 0(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 256, i1 false)
ret void
}
; Test the first case that needs two MVCs.
define void @f7(i8* %dest, i8* %src) {
; CHECK-LABEL: f7:
; CHECK: mvc 0(256,%r2), 0(%r3)
; CHECK: mvc 256(1,%r2), 256(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 257, i1 false)
ret void
}
; Test the last-but-one case that needs two MVCs.
define void @f8(i8* %dest, i8* %src) {
; CHECK-LABEL: f8:
; CHECK: mvc 0(256,%r2), 0(%r3)
; CHECK: mvc 256(255,%r2), 256(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 511, i1 false)
ret void
}
; Test the last case that needs two MVCs.
define void @f9(i8* %dest, i8* %src) {
; CHECK-LABEL: f9:
; CHECK: mvc 0(256,%r2), 0(%r3)
; CHECK: mvc 256(256,%r2), 256(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 512, i1 false)
ret void
}
; Test an arbitrary value that uses straight-line code.
define void @f10(i8* %dest, i8* %src) {
; CHECK-LABEL: f10:
; CHECK: mvc 0(256,%r2), 0(%r3)
; CHECK: mvc 256(256,%r2), 256(%r3)
; CHECK: mvc 512(256,%r2), 512(%r3)
; CHECK: mvc 768(256,%r2), 768(%r3)
; CHECK: mvc 1024(255,%r2), 1024(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 1279, i1 false)
ret void
}
; ...and again in cases where not all parts are in range of MVC.
define void @f11(i8* %srcbase, i8* %destbase) {
; CHECK-LABEL: f11:
; CHECK: mvc 4000(256,%r2), 3500(%r3)
; CHECK: lay [[NEWDEST:%r[1-5]]], 4256(%r2)
; CHECK: mvc 0(256,[[NEWDEST]]), 3756(%r3)
; CHECK: mvc 256(256,[[NEWDEST]]), 4012(%r3)
; CHECK: lay [[NEWSRC:%r[1-5]]], 4268(%r3)
; CHECK: mvc 512(256,[[NEWDEST]]), 0([[NEWSRC]])
; CHECK: mvc 768(255,[[NEWDEST]]), 256([[NEWSRC]])
; CHECK: br %r14
%dest = getelementptr i8, i8* %srcbase, i64 4000
%src = getelementptr i8, i8* %destbase, i64 3500
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 1279, i1 false)
ret void
}
; ...and again with a destination frame base that goes out of range.
define void @f12() {
; CHECK-LABEL: f12:
; CHECK: brasl %r14, foo@PLT
; CHECK: mvc 4076(256,%r15), 2100(%r15)
; CHECK: lay [[NEWDEST:%r[1-5]]], 4332(%r15)
; CHECK: mvc 0(256,[[NEWDEST]]), 2356(%r15)
; CHECK: mvc 256(256,[[NEWDEST]]), 2612(%r15)
; CHECK: mvc 512(256,[[NEWDEST]]), 2868(%r15)
; CHECK: mvc 768(255,[[NEWDEST]]), 3124(%r15)
; CHECK: brasl %r14, foo@PLT
; CHECK: br %r14
%arr = alloca [6000 x i8]
%dest = getelementptr [6000 x i8], [6000 x i8] *%arr, i64 0, i64 3900
%src = getelementptr [6000 x i8], [6000 x i8] *%arr, i64 0, i64 1924
call void @foo(i8* %dest, i8* %src)
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 1279, i1 false)
call void @foo(i8* %dest, i8* %src)
ret void
}
; ...and again with a source frame base that goes out of range.
define void @f13() {
; CHECK-LABEL: f13:
; CHECK: brasl %r14, foo@PLT
; CHECK: mvc 200(256,%r15), 3826(%r15)
; CHECK: mvc 456(256,%r15), 4082(%r15)
; CHECK: lay [[NEWSRC:%r[1-5]]], 4338(%r15)
; CHECK: mvc 712(256,%r15), 0([[NEWSRC]])
; CHECK: mvc 968(256,%r15), 256([[NEWSRC]])
; CHECK: mvc 1224(255,%r15), 512([[NEWSRC]])
; CHECK: brasl %r14, foo@PLT
; CHECK: br %r14
%arr = alloca [6000 x i8]
%dest = getelementptr [6000 x i8], [6000 x i8] *%arr, i64 0, i64 24
%src = getelementptr [6000 x i8], [6000 x i8] *%arr, i64 0, i64 3650
call void @foo(i8* %dest, i8* %src)
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 1279, i1 false)
call void @foo(i8* %dest, i8* %src)
ret void
}
; Test the last case that is done using straight-line code.
define void @f14(i8* %dest, i8* %src) {
; CHECK-LABEL: f14:
; CHECK: mvc 0(256,%r2), 0(%r3)
; CHECK: mvc 256(256,%r2), 256(%r3)
; CHECK: mvc 512(256,%r2), 512(%r3)
; CHECK: mvc 768(256,%r2), 768(%r3)
; CHECK: mvc 1024(256,%r2), 1024(%r3)
; CHECK: mvc 1280(256,%r2), 1280(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 1536, i1 false)
ret void
}
; Test the first case that is done using a loop.
define void @f15(i8* %dest, i8* %src) {
; CHECK-LABEL: f15:
; CHECK: lghi [[COUNT:%r[0-5]]], 6
; CHECK: [[LABEL:\.L[^:]*]]:
; CHECK: pfd 2, 768(%r2)
; CHECK: mvc 0(256,%r2), 0(%r3)
; CHECK: la %r2, 256(%r2)
; CHECK: la %r3, 256(%r3)
; CHECK: brctg [[COUNT]], [[LABEL]]
; CHECK: mvc 0(1,%r2), 0(%r3)
; CHECK: br %r14
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 1537, i1 false)
ret void
}
; ...and again with frame bases, where the base must be loaded into a
; register before the loop.
define void @f16() {
; CHECK-LABEL: f16:
; CHECK: brasl %r14, foo@PLT
; CHECK-DAG: lghi [[COUNT:%r[0-5]]], 6
; CHECK-DAG: la [[BASE:%r[0-5]]], 160(%r15)
; CHECK: [[LABEL:\.L[^:]*]]:
; CHECK: pfd 2, 2368([[BASE]])
; CHECK: mvc 1600(256,[[BASE]]), 0([[BASE]])
; CHECK: la [[BASE]], 256([[BASE]])
; CHECK: brctg [[COUNT]], [[LABEL]]
; CHECK: mvc 1600(1,[[BASE]]), 0([[BASE]])
; CHECK: brasl %r14, foo@PLT
; CHECK: br %r14
%arr = alloca [3200 x i8]
%dest = getelementptr [3200 x i8], [3200 x i8] *%arr, i64 0, i64 1600
%src = getelementptr [3200 x i8], [3200 x i8] *%arr, i64 0, i64 0
call void @foo(i8* %dest, i8* %src)
call void @llvm.memcpy.p0i8.p0i8.i64(i8* %dest, i8* %src, i64 1537, i1 false)
call void @foo(i8* %dest, i8* %src)
ret void
}
Reference in New Issue View Git Blame Copy Permalink