1e68724d24
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
144 lines
6.0 KiB
LLVM
144 lines
6.0 KiB
LLVM
; RUN: llc < %s -disable-fp-elim -verify-coalescing
|
|
; PR13414
|
|
;
|
|
; During coalescing, remat triggers DCE which deletes the penultimate use of a
|
|
; load. This load should not be folded into the remaining use because it is not
|
|
; safe to move, and it would extend the live range of the address.
|
|
;
|
|
; LiveRangeEdit::foldAsLoad() doesn't extend live ranges, so -verify-coalescing
|
|
; catches the problem.
|
|
|
|
target triple = "i386-unknown-linux-gnu"
|
|
|
|
%type_a = type { %type_a*, %type_b }
|
|
%type_b = type { %type_c, i32 }
|
|
%type_c = type { i32, %type_d }
|
|
%type_d = type { i64 }
|
|
%type_e = type { %type_c, i64 }
|
|
|
|
declare void @llvm.memcpy.p0i8.p0i8.i32(i8* nocapture, i8* nocapture, i32, i1) nounwind
|
|
|
|
define linkonce_odr void @test() nounwind {
|
|
entry:
|
|
br i1 undef, label %while.end.while.end26_crit_edge, label %while.body12.lr.ph
|
|
|
|
while.end.while.end26_crit_edge: ; preds = %entry
|
|
br label %while.end26
|
|
|
|
while.body12.lr.ph: ; preds = %entry
|
|
br label %while.body12
|
|
|
|
while.body12: ; preds = %if.end24, %while.body12.lr.ph
|
|
%tmp = phi %type_a* [ undef, %while.body12.lr.ph ], [ %tmp18, %if.end24 ]
|
|
%ins151154161 = phi i128 [ 0, %while.body12.lr.ph ], [ %phitmp, %if.end24 ]
|
|
%ins135156160 = phi i128 [ 0, %while.body12.lr.ph ], [ %phitmp158, %if.end24 ]
|
|
%ins151 = or i128 0, %ins151154161
|
|
%cmp.i.i.i.i.i67 = icmp sgt i32 undef, 8
|
|
br i1 %cmp.i.i.i.i.i67, label %if.then.i.i.i.i71, label %if.else.i.i.i.i74
|
|
|
|
if.then.i.i.i.i71: ; preds = %while.body12
|
|
%call4.i.i.i.i68 = call noalias i8* @malloc(i32 undef) nounwind
|
|
%tmp1 = getelementptr inbounds %type_a, %type_a* %tmp, i32 0, i32 1, i32 0, i32 1
|
|
%buf_6.i.i.i.i70 = bitcast %type_d* %tmp1 to i8**
|
|
%tmp2 = load i8*, i8** %buf_6.i.i.i.i70, align 4
|
|
call void @llvm.memcpy.p0i8.p0i8.i32(i8* undef, i8* %tmp2, i32 undef, i1 false) nounwind
|
|
unreachable
|
|
|
|
if.else.i.i.i.i74: ; preds = %while.body12
|
|
%i_.i.i.i.i72 = getelementptr inbounds %type_a, %type_a* %tmp, i32 0, i32 1, i32 0, i32 1, i32 0
|
|
%tmp3 = load i64, i64* %i_.i.i.i.i72, align 4
|
|
%tmp4 = zext i64 %tmp3 to i128
|
|
%tmp5 = shl nuw nsw i128 %tmp4, 32
|
|
%ins148 = or i128 %tmp5, %ins151
|
|
%second3.i.i76 = getelementptr inbounds %type_a, %type_a* %tmp, i32 0, i32 1, i32 1
|
|
%tmp6 = load i32, i32* %second3.i.i76, align 4
|
|
%tmp7 = zext i32 %tmp6 to i128
|
|
%tmp8 = shl nuw i128 %tmp7, 96
|
|
%mask144 = and i128 %ins148, 79228162495817593519834398720
|
|
%tmp9 = load %type_e*, %type_e** undef, align 4
|
|
%len_.i.i.i.i86 = getelementptr inbounds %type_e, %type_e* %tmp9, i32 0, i32 0, i32 0
|
|
%tmp10 = load i32, i32* %len_.i.i.i.i86, align 4
|
|
%tmp11 = zext i32 %tmp10 to i128
|
|
%ins135 = or i128 %tmp11, %ins135156160
|
|
%cmp.i.i.i.i.i88 = icmp sgt i32 %tmp10, 8
|
|
br i1 %cmp.i.i.i.i.i88, label %if.then.i.i.i.i92, label %if.else.i.i.i.i95
|
|
|
|
if.then.i.i.i.i92: ; preds = %if.else.i.i.i.i74
|
|
%call4.i.i.i.i89 = call noalias i8* @malloc(i32 %tmp10) nounwind
|
|
%ins126 = or i128 0, %ins135
|
|
%tmp12 = getelementptr inbounds %type_e, %type_e* %tmp9, i32 0, i32 0, i32 1
|
|
%buf_6.i.i.i.i91 = bitcast %type_d* %tmp12 to i8**
|
|
%tmp13 = load i8*, i8** %buf_6.i.i.i.i91, align 4
|
|
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %call4.i.i.i.i89, i8* %tmp13, i32 %tmp10, i1 false) nounwind
|
|
br label %A
|
|
|
|
if.else.i.i.i.i95: ; preds = %if.else.i.i.i.i74
|
|
%i_.i.i.i.i93 = getelementptr inbounds %type_e, %type_e* %tmp9, i32 0, i32 0, i32 1, i32 0
|
|
br label %A
|
|
|
|
A: ; preds = %if.else.i.i.i.i95, %if.then.i.i.i.i92
|
|
%ins135157 = phi i128 [ %ins126, %if.then.i.i.i.i92 ], [ undef, %if.else.i.i.i.i95 ]
|
|
%second3.i.i97 = getelementptr inbounds %type_e, %type_e* %tmp9, i32 0, i32 1
|
|
%tmp14 = load i64, i64* %second3.i.i97, align 4
|
|
%tmp15 = trunc i64 %tmp14 to i32
|
|
%cmp.i99 = icmp sgt i32 %tmp6, %tmp15
|
|
%tmp16 = trunc i128 %ins135157 to i32
|
|
%cmp.i.i.i.i.i.i101 = icmp sgt i32 %tmp16, 8
|
|
br i1 %cmp.i.i.i.i.i.i101, label %if.then.i.i.i.i.i103, label %B
|
|
|
|
if.then.i.i.i.i.i103: ; preds = %A
|
|
unreachable
|
|
|
|
B: ; preds = %A
|
|
%tmp17 = trunc i128 %ins148 to i32
|
|
%cmp.i.i.i.i.i.i83 = icmp sgt i32 %tmp17, 8
|
|
br i1 %cmp.i.i.i.i.i.i83, label %if.then.i.i.i.i.i85, label %C
|
|
|
|
if.then.i.i.i.i.i85: ; preds = %B
|
|
unreachable
|
|
|
|
C: ; preds = %B
|
|
br i1 %cmp.i99, label %if.then17, label %if.end24
|
|
|
|
if.then17: ; preds = %C
|
|
br i1 false, label %if.then.i.i.i.i.i43, label %D
|
|
|
|
if.then.i.i.i.i.i43: ; preds = %if.then17
|
|
unreachable
|
|
|
|
D: ; preds = %if.then17
|
|
br i1 undef, label %if.then.i.i.i.i.i, label %E
|
|
|
|
if.then.i.i.i.i.i: ; preds = %D
|
|
unreachable
|
|
|
|
E: ; preds = %D
|
|
br label %if.end24
|
|
|
|
if.end24: ; preds = %E, %C
|
|
%phitmp = or i128 %tmp8, %mask144
|
|
%phitmp158 = or i128 undef, undef
|
|
%tmp18 = load %type_a*, %type_a** undef, align 4
|
|
%tmp19 = load %type_a*, %type_a** undef, align 4
|
|
%cmp.i49 = icmp eq %type_a* %tmp18, %tmp19
|
|
br i1 %cmp.i49, label %while.cond10.while.end26_crit_edge, label %while.body12
|
|
|
|
while.cond10.while.end26_crit_edge: ; preds = %if.end24
|
|
%.pre = load %type_e*, %type_e** undef, align 4
|
|
br label %while.end26
|
|
|
|
while.end26: ; preds = %while.cond10.while.end26_crit_edge, %while.end.while.end26_crit_edge
|
|
br i1 undef, label %while.body.lr.ph.i, label %F
|
|
|
|
while.body.lr.ph.i: ; preds = %while.end26
|
|
br label %while.body.i
|
|
|
|
while.body.i: ; preds = %while.body.i, %while.body.lr.ph.i
|
|
br i1 false, label %while.body.i, label %F
|
|
|
|
F: ; preds = %while.body.i, %while.end26
|
|
ret void
|
|
}
|
|
|
|
declare noalias i8* @malloc(i32) nounwind
|