Codegen: Tail-duplicate during placement.
The tail duplication pass uses an assumed layout when making duplication
decisions. This is fine, but passes up duplication opportunities that
may arise when blocks are outlined. Because we want the updated CFG to
affect subsequent placement decisions, this change must occur during
placement.
In order to achieve this goal, TailDuplicationPass is split into a
utility class, TailDuplicator, and the pass itself. The pass delegates
nearly everything to the TailDuplicator object, except for looping over
the blocks in a function. This allows the same code to be used for tail
duplication in both places.
This change, in concert with outlining optional branches, allows
triangle shaped code to perform much better, esepecially when the
taken/untaken branches are correlated, as it creates a second spine when
the tests are small enough.
Issue from previous rollback fixed, and a new test was added for that
case as well. Issue was worklist/scheduling/taildup issue in layout.
Issue from 2nd rollback fixed, with 2 additional tests. Issue was
tail merging/loop info/tail-duplication causing issue with loops that share
a header block.
Issue with early tail-duplication of blocks that branch to a fallthrough
predecessor fixed with test case: tail-dup-branch-to-fallthrough.ll
Differential revision: https://reviews.llvm.org/D18226
llvm-svn: 283934
2016-10-12 04:36:43 +08:00
|
|
|
; RUN: llc -O2 -o - %s | FileCheck %s
|
|
|
|
|
target datalayout = "e-m:e-i64:64-f80:128-n8:16:32:64-S128"
|
|
|
|
|
target triple = "x86_64-unknown-linux-gnu"
|
|
|
|
|
|
|
|
|
|
; Function Attrs: nounwind uwtable
|
|
|
|
|
; CHECK-LABEL: tail_dup_merge_loops
|
|
|
|
|
; CHECK: # %entry
|
|
|
|
|
; CHECK-NOT: # %{{[a-zA-Z_]+}}
|
Codegen: Make chains from trellis-shaped CFGs
Lay out trellis-shaped CFGs optimally.
A trellis of the shape below:
A B
|\ /|
| \ / |
| X |
| / \ |
|/ \|
C D
would be laid out A; B->C ; D by the current layout algorithm. Now we identify
trellises and lay them out either A->C; B->D or A->D; B->C. This scales with an
increasing number of predecessors. A trellis is a a group of 2 or more
predecessor blocks that all have the same successors.
because of this we can tail duplicate to extend existing trellises.
As an example consider the following CFG:
B D F H
/ \ / \ / \ / \
A---C---E---G---Ret
Where A,C,E,G are all small (Currently 2 instructions).
The CFG preserving layout is then A,B,C,D,E,F,G,H,Ret.
The current code will copy C into B, E into D and G into F and yield the layout
A,C,B(C),E,D(E),F(G),G,H,ret
define void @straight_test(i32 %tag) {
entry:
br label %test1
test1: ; A
%tagbit1 = and i32 %tag, 1
%tagbit1eq0 = icmp eq i32 %tagbit1, 0
br i1 %tagbit1eq0, label %test2, label %optional1
optional1: ; B
call void @a()
br label %test2
test2: ; C
%tagbit2 = and i32 %tag, 2
%tagbit2eq0 = icmp eq i32 %tagbit2, 0
br i1 %tagbit2eq0, label %test3, label %optional2
optional2: ; D
call void @b()
br label %test3
test3: ; E
%tagbit3 = and i32 %tag, 4
%tagbit3eq0 = icmp eq i32 %tagbit3, 0
br i1 %tagbit3eq0, label %test4, label %optional3
optional3: ; F
call void @c()
br label %test4
test4: ; G
%tagbit4 = and i32 %tag, 8
%tagbit4eq0 = icmp eq i32 %tagbit4, 0
br i1 %tagbit4eq0, label %exit, label %optional4
optional4: ; H
call void @d()
br label %exit
exit:
ret void
}
here is the layout after D27742:
straight_test: # @straight_test
; ... Prologue elided
; BB#0: # %entry ; A (merged with test1)
; ... More prologue elided
mr 30, 3
andi. 3, 30, 1
bc 12, 1, .LBB0_2
; BB#1: # %test2 ; C
rlwinm. 3, 30, 0, 30, 30
beq 0, .LBB0_3
b .LBB0_4
.LBB0_2: # %optional1 ; B (copy of C)
bl a
nop
rlwinm. 3, 30, 0, 30, 30
bne 0, .LBB0_4
.LBB0_3: # %test3 ; E
rlwinm. 3, 30, 0, 29, 29
beq 0, .LBB0_5
b .LBB0_6
.LBB0_4: # %optional2 ; D (copy of E)
bl b
nop
rlwinm. 3, 30, 0, 29, 29
bne 0, .LBB0_6
.LBB0_5: # %test4 ; G
rlwinm. 3, 30, 0, 28, 28
beq 0, .LBB0_8
b .LBB0_7
.LBB0_6: # %optional3 ; F (copy of G)
bl c
nop
rlwinm. 3, 30, 0, 28, 28
beq 0, .LBB0_8
.LBB0_7: # %optional4 ; H
bl d
nop
.LBB0_8: # %exit ; Ret
ld 30, 96(1) # 8-byte Folded Reload
addi 1, 1, 112
ld 0, 16(1)
mtlr 0
blr
The tail-duplication has produced some benefit, but it has also produced a
trellis which is not laid out optimally. With this patch, we improve the layouts
of such trellises, and decrease the cost calculation for tail-duplication
accordingly.
This patch produces the layout A,C,E,G,B,D,F,H,Ret. This layout does have
back edges, which is a negative, but it has a bigger compensating
positive, which is that it handles the case where there are long strings
of skipped blocks much better than the original layout. Both layouts
handle runs of executed blocks equally well. Branch prediction also
improves if there is any correlation between subsequent optional blocks.
Here is the resulting concrete layout:
straight_test: # @straight_test
; BB#0: # %entry ; A (merged with test1)
mr 30, 3
andi. 3, 30, 1
bc 12, 1, .LBB0_4
; BB#1: # %test2 ; C
rlwinm. 3, 30, 0, 30, 30
bne 0, .LBB0_5
.LBB0_2: # %test3 ; E
rlwinm. 3, 30, 0, 29, 29
bne 0, .LBB0_6
.LBB0_3: # %test4 ; G
rlwinm. 3, 30, 0, 28, 28
bne 0, .LBB0_7
b .LBB0_8
.LBB0_4: # %optional1 ; B (Copy of C)
bl a
nop
rlwinm. 3, 30, 0, 30, 30
beq 0, .LBB0_2
.LBB0_5: # %optional2 ; D (Copy of E)
bl b
nop
rlwinm. 3, 30, 0, 29, 29
beq 0, .LBB0_3
.LBB0_6: # %optional3 ; F (Copy of G)
bl c
nop
rlwinm. 3, 30, 0, 28, 28
beq 0, .LBB0_8
.LBB0_7: # %optional4 ; H
bl d
nop
.LBB0_8: # %exit
Differential Revision: https://reviews.llvm.org/D28522
llvm-svn: 295223
2017-02-16 03:49:14 +08:00
|
|
|
; CHECK: # %exit
|
|
|
|
|
; CHECK-NOT: # %{{[a-zA-Z_]+}}
|
Codegen: Tail-duplicate during placement.
The tail duplication pass uses an assumed layout when making duplication
decisions. This is fine, but passes up duplication opportunities that
may arise when blocks are outlined. Because we want the updated CFG to
affect subsequent placement decisions, this change must occur during
placement.
In order to achieve this goal, TailDuplicationPass is split into a
utility class, TailDuplicator, and the pass itself. The pass delegates
nearly everything to the TailDuplicator object, except for looping over
the blocks in a function. This allows the same code to be used for tail
duplication in both places.
This change, in concert with outlining optional branches, allows
triangle shaped code to perform much better, esepecially when the
taken/untaken branches are correlated, as it creates a second spine when
the tests are small enough.
Issue from previous rollback fixed, and a new test was added for that
case as well. Issue was worklist/scheduling/taildup issue in layout.
Issue from 2nd rollback fixed, with 2 additional tests. Issue was
tail merging/loop info/tail-duplication causing issue with loops that share
a header block.
Issue with early tail-duplication of blocks that branch to a fallthrough
predecessor fixed with test case: tail-dup-branch-to-fallthrough.ll
Differential revision: https://reviews.llvm.org/D18226
llvm-svn: 283934
2016-10-12 04:36:43 +08:00
|
|
|
; CHECK: # %inner_loop_exit
|
|
|
|
|
; CHECK-NOT: # %{{[a-zA-Z_]+}}
|
|
|
|
|
; CHECK: # %inner_loop_latch
|
|
|
|
|
; CHECK-NOT: # %{{[a-zA-Z_]+}}
|
|
|
|
|
; CHECK: # %inner_loop_test
|
|
|
|
|
define void @tail_dup_merge_loops(i32 %a, i8* %b, i8* %c) local_unnamed_addr #0 {
|
|
|
|
|
entry:
|
|
|
|
|
%notlhs674.i = icmp eq i32 %a, 0
|
|
|
|
|
br label %outer_loop_top
|
|
|
|
|
|
|
|
|
|
outer_loop_top: ; preds = %inner_loop_exit, %entry
|
|
|
|
|
%dst.0.ph.i = phi i8* [ %b, %entry ], [ %scevgep679.i, %inner_loop_exit ]
|
|
|
|
|
br i1 %notlhs674.i, label %exit, label %inner_loop_preheader
|
|
|
|
|
|
|
|
|
|
inner_loop_preheader: ; preds = %outer_loop_top
|
|
|
|
|
br label %inner_loop_top
|
|
|
|
|
|
|
|
|
|
inner_loop_top: ; preds = %inner_loop_latch, %inner_loop_preheader
|
|
|
|
|
%dst.0.i = phi i8* [ %inc, %inner_loop_latch ], [ %dst.0.ph.i, %inner_loop_preheader ]
|
|
|
|
|
%var = load i8, i8* %dst.0.i
|
|
|
|
|
%tobool1.i = icmp slt i8 %var, 0
|
|
|
|
|
br label %inner_loop_test
|
|
|
|
|
|
|
|
|
|
inner_loop_test: ; preds = %inner_loop_top
|
|
|
|
|
br i1 %tobool1.i, label %inner_loop_exit, label %inner_loop_latch
|
|
|
|
|
|
|
|
|
|
inner_loop_exit: ; preds = %inner_loop_test
|
|
|
|
|
%scevgep.i = getelementptr i8, i8* %dst.0.i, i64 1
|
|
|
|
|
%scevgep679.i = getelementptr i8, i8* %scevgep.i, i64 0
|
|
|
|
|
br label %outer_loop_top
|
|
|
|
|
|
|
|
|
|
inner_loop_latch: ; preds = %inner_loop_test
|
|
|
|
|
%cmp75.i = icmp ult i8* %dst.0.i, %c
|
|
|
|
|
%inc = getelementptr i8, i8* %dst.0.i, i64 2
|
|
|
|
|
br label %inner_loop_top
|
|
|
|
|
|
|
|
|
|
exit: ; preds = %outer_loop_top
|
|
|
|
|
ret void
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
@.str.6 = external unnamed_addr constant [23 x i8], align 1
|
|
|
|
|
|
|
|
|
|
; There is an erroneus check in LoopBase::addBasicBlockToLoop(), where it
|
|
|
|
|
; assumes that the header block for a loop is unique.
|
|
|
|
|
; For most of compilation this assumption is true, but during layout we allow
|
|
|
|
|
; this assumption to be violated. The following code will trigger the bug:
|
|
|
|
|
|
|
|
|
|
; The loops in question is eventually headed by the block shared_loop_header
|
|
|
|
|
;
|
|
|
|
|
; During layout The block labeled outer_loop_header gets tail-duplicated into
|
|
|
|
|
; outer_loop_latch, and into shared_preheader, and then removed. This leaves
|
|
|
|
|
; shared_loop_header as the header of both loops. The end result
|
|
|
|
|
; is that there are 2 valid loops, and that they share a header. If we re-ran
|
|
|
|
|
; the loop analysis, it would classify this as a single loop.
|
|
|
|
|
; So far this is fine as far as layout is concerned.
|
|
|
|
|
; After layout we tail merge blocks merge_other and merge_predecessor_split.
|
|
|
|
|
; We do this even though they share only a single instruction, because
|
|
|
|
|
; merge_predecessor_split falls through to their shared successor:
|
|
|
|
|
; outer_loop_latch.
|
|
|
|
|
; The rest of the blocks in the function are noise unfortunately. Bugpoint
|
|
|
|
|
; couldn't shrink the test any further.
|
|
|
|
|
|
|
|
|
|
; CHECK-LABEL: loop_shared_header
|
|
|
|
|
; CHECK: # %entry
|
|
|
|
|
; CHECK: # %shared_preheader
|
|
|
|
|
; CHECK: # %shared_loop_header
|
|
|
|
|
; CHECK: # %inner_loop_body
|
|
|
|
|
; CHECK: # %outer_loop_latch
|
2016-10-13 04:39:33 +08:00
|
|
|
; CHECK: # %merge_predecessor_split
|
Codegen: Tail-duplicate during placement.
The tail duplication pass uses an assumed layout when making duplication
decisions. This is fine, but passes up duplication opportunities that
may arise when blocks are outlined. Because we want the updated CFG to
affect subsequent placement decisions, this change must occur during
placement.
In order to achieve this goal, TailDuplicationPass is split into a
utility class, TailDuplicator, and the pass itself. The pass delegates
nearly everything to the TailDuplicator object, except for looping over
the blocks in a function. This allows the same code to be used for tail
duplication in both places.
This change, in concert with outlining optional branches, allows
triangle shaped code to perform much better, esepecially when the
taken/untaken branches are correlated, as it creates a second spine when
the tests are small enough.
Issue from previous rollback fixed, and a new test was added for that
case as well. Issue was worklist/scheduling/taildup issue in layout.
Issue from 2nd rollback fixed, with 2 additional tests. Issue was
tail merging/loop info/tail-duplication causing issue with loops that share
a header block.
Issue with early tail-duplication of blocks that branch to a fallthrough
predecessor fixed with test case: tail-dup-branch-to-fallthrough.ll
Differential revision: https://reviews.llvm.org/D18226
llvm-svn: 283934
2016-10-12 04:36:43 +08:00
|
|
|
; CHECK: # %outer_loop_latch
|
|
|
|
|
; CHECK: # %cleanup
|
|
|
|
|
define i32 @loop_shared_header(i8* %exe, i32 %exesz, i32 %headsize, i32 %min, i32 %wwprva, i32 %e_lfanew, i8* readonly %wwp, i32 %wwpsz, i16 zeroext %sects) local_unnamed_addr #0 {
|
|
|
|
|
entry:
|
|
|
|
|
%0 = load i32, i32* undef, align 4
|
|
|
|
|
%mul = shl nsw i32 %0, 2
|
|
|
|
|
br i1 undef, label %if.end19, label %cleanup
|
|
|
|
|
|
|
|
|
|
if.end19: ; preds = %entry
|
|
|
|
|
%conv = zext i32 %mul to i64
|
|
|
|
|
%call = tail call i8* @cli_calloc(i64 %conv, i64 1)
|
|
|
|
|
%1 = icmp eq i32 %exesz, 0
|
|
|
|
|
%notrhs = icmp eq i32 %0, 0
|
|
|
|
|
%or.cond117.not = or i1 %1, %notrhs
|
|
|
|
|
%or.cond202 = or i1 %or.cond117.not, undef
|
|
|
|
|
%cmp35 = icmp ult i8* undef, %exe
|
|
|
|
|
%or.cond203 = or i1 %or.cond202, %cmp35
|
|
|
|
|
br i1 %or.cond203, label %cleanup, label %if.end50
|
|
|
|
|
|
|
|
|
|
if.end50: ; preds = %if.end19
|
|
|
|
|
tail call void @llvm.memcpy.p0i8.p0i8.i64(i8* nonnull %call, i8* undef, i64 %conv, i32 1, i1 false)
|
|
|
|
|
%cmp1.i.i = icmp ugt i32 %mul, 3
|
|
|
|
|
br i1 %cmp1.i.i, label %shared_preheader, label %wunpsect.exit.thread.loopexit391
|
|
|
|
|
|
|
|
|
|
shared_preheader: ; preds = %if.end50
|
|
|
|
|
br label %outer_loop_header
|
|
|
|
|
|
|
|
|
|
outer_loop_header: ; preds = %outer_loop_latch, %shared_preheader
|
|
|
|
|
%bits.1.i = phi i8 [ 32, %shared_preheader ], [ %bits.43.i, %outer_loop_latch ]
|
|
|
|
|
%dst.0.ph.i = phi i8* [ undef, %shared_preheader ], [ %scevgep679.i, %outer_loop_latch ]
|
|
|
|
|
%2 = icmp eq i32 undef, 0
|
|
|
|
|
br i1 %2, label %while.cond.us1412.i, label %shared_loop_header
|
|
|
|
|
|
|
|
|
|
while.cond.us1412.i: ; preds = %outer_loop_header
|
|
|
|
|
%.pre.i = add i8 %bits.1.i, -1
|
|
|
|
|
%tobool2.us1420.i = icmp eq i8 %.pre.i, 0
|
|
|
|
|
%or.cond.us1421.i = or i1 undef, %tobool2.us1420.i
|
|
|
|
|
br i1 %or.cond.us1421.i, label %if.end41.us1436.i, label %cleanup
|
|
|
|
|
|
|
|
|
|
if.end41.us1436.i: ; preds = %while.cond.us1412.i
|
|
|
|
|
unreachable
|
|
|
|
|
|
|
|
|
|
shared_loop_header: ; preds = %dup_early2, %dup_early1
|
|
|
|
|
%dst.0.i = phi i8* [ undef, %inner_loop_body ], [ %dst.0.ph.i, %outer_loop_header ], [ undef, %dead_block ]
|
|
|
|
|
%cmp3.i1172.i = icmp ult i8* undef, %call
|
|
|
|
|
br i1 %cmp3.i1172.i, label %wunpsect.exit.thread.loopexit389, label %inner_loop_body
|
|
|
|
|
|
|
|
|
|
inner_loop_body: ; preds = %shared_loop_header
|
|
|
|
|
%3 = icmp slt i32 undef, 0
|
|
|
|
|
br i1 %3, label %if.end96.i, label %shared_loop_header
|
|
|
|
|
|
|
|
|
|
dead_block: ; preds = %inner_loop_body
|
|
|
|
|
%cmp75.i = icmp ult i8* %dst.0.i, undef
|
|
|
|
|
br label %shared_loop_header
|
|
|
|
|
|
|
|
|
|
if.end96.i: ; preds = %inner_loop_body
|
|
|
|
|
%cmp97.i = icmp ugt i32 undef, 2
|
|
|
|
|
br i1 %cmp97.i, label %if.then99.i, label %if.end287.i
|
|
|
|
|
|
|
|
|
|
if.then99.i: ; preds = %if.end96.i
|
|
|
|
|
tail call void (i8*, ...) @cli_dbgmsg(i8* getelementptr inbounds ([23 x i8], [23 x i8]* @.str.6, i64 0, i64 0), i32 undef)
|
|
|
|
|
br label %cleanup
|
|
|
|
|
|
|
|
|
|
if.end287.i: ; preds = %if.end96.i
|
|
|
|
|
%cmp291.i = icmp ne i32 undef, 1
|
|
|
|
|
%conv294.i = select i1 %cmp291.i, i16 4, i16 3
|
|
|
|
|
br i1 undef, label %if.end308.i, label %outer_loop_latch
|
|
|
|
|
|
|
|
|
|
if.end308.i: ; preds = %if.end287.i
|
|
|
|
|
br i1 undef, label %if.end335.i, label %merge_predecessor_split
|
|
|
|
|
|
|
|
|
|
merge_predecessor_split: ; preds = %if.end308.i
|
|
|
|
|
%4 = bitcast i8* undef to i32*
|
|
|
|
|
br label %outer_loop_latch
|
|
|
|
|
|
|
|
|
|
if.end335.i: ; preds = %if.end308.i
|
|
|
|
|
br i1 undef, label %outer_loop_latch, label %merge_other
|
|
|
|
|
|
|
|
|
|
merge_other: ; preds = %if.end335.i
|
|
|
|
|
br label %outer_loop_latch
|
|
|
|
|
|
|
|
|
|
outer_loop_latch: ; preds = %merge_other, %if.end335.i, %merge_predecessor_split, %if.end287.i
|
|
|
|
|
%bits.43.i = phi i8 [ undef, %if.end287.i ], [ undef, %merge_other ], [ 32, %merge_predecessor_split ], [ 0, %if.end335.i ]
|
|
|
|
|
%backsize.0.i = phi i16 [ %conv294.i, %if.end287.i ], [ 0, %merge_other ], [ 0, %merge_predecessor_split ], [ 0, %if.end335.i ]
|
|
|
|
|
%5 = add i16 %backsize.0.i, -1
|
|
|
|
|
%6 = zext i16 %5 to i64
|
|
|
|
|
%scevgep.i = getelementptr i8, i8* %dst.0.ph.i, i64 1
|
|
|
|
|
%scevgep679.i = getelementptr i8, i8* %scevgep.i, i64 %6
|
|
|
|
|
br label %outer_loop_header
|
|
|
|
|
|
|
|
|
|
wunpsect.exit.thread.loopexit389: ; preds = %shared_loop_header
|
|
|
|
|
unreachable
|
|
|
|
|
|
|
|
|
|
wunpsect.exit.thread.loopexit391: ; preds = %if.end50
|
|
|
|
|
unreachable
|
|
|
|
|
|
|
|
|
|
cleanup: ; preds = %if.then99.i, %while.cond.us1412.i, %if.end19, %entry
|
|
|
|
|
%retval.0 = phi i32 [ 0, %if.then99.i ], [ 1, %entry ], [ 1, %if.end19 ], [ 1, %while.cond.us1412.i ]
|
|
|
|
|
ret i32 %retval.0
|
|
|
|
|
}
|
|
|
|
|
|
|
|
|
|
; Function Attrs: nounwind
|
|
|
|
|
declare void @cli_dbgmsg(i8*, ...) local_unnamed_addr #0
|
|
|
|
|
|
|
|
|
|
; Function Attrs: nounwind
|
|
|
|
|
declare i8* @cli_calloc(i64, i64) local_unnamed_addr #0
|
|
|
|
|
|
|
|
|
|
; Function Attrs: argmemonly nounwind
|
|
|
|
|
declare void @llvm.memcpy.p0i8.p0i8.i64(i8* nocapture writeonly, i8* nocapture readonly, i64, i32, i1) #1
|
|
|
|
|
attributes #0 = { nounwind }
|
|
|
|
|
attributes #1 = { argmemonly nounwind }
|