2014-05-30 18:09:59 +08:00
|
|
|
; RUN: llc < %s -mtriple=armv6-apple-ios5.0 -mattr=+vfp2 -arm-atomic-cfg-tidy=0 | FileCheck %s -check-prefix=CHECKV6
|
|
|
|
; RUN: llc < %s -mtriple=thumbv7-apple-ios5.0 -arm-atomic-cfg-tidy=0 | FileCheck %s -check-prefix=CHECKT2D
|
|
|
|
; RUN: llc < %s -mtriple=armv6-linux-gnueabi -relocation-model=pic -mattr=+vfp2 -arm-atomic-cfg-tidy=0 \
|
2014-03-11 23:09:49 +08:00
|
|
|
; RUN: | FileCheck %s -check-prefix=CHECKELF
|
2011-10-08 01:17:49 +08:00
|
|
|
|
|
|
|
; Enable tailcall optimization for iOS 5.0
|
|
|
|
; rdar://9120031
|
2010-06-19 03:00:18 +08:00
|
|
|
|
|
|
|
@t = weak global i32 ()* null ; <i32 ()**> [#uses=1]
|
|
|
|
|
|
|
|
declare void @g(i32, i32, i32, i32)
|
|
|
|
|
2019-12-25 07:52:21 +08:00
|
|
|
define void @t1() "frame-pointer"="all" {
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKELF-LABEL: t1:
|
2016-06-17 00:09:53 +08:00
|
|
|
; CHECKELF: bl g
|
2010-06-19 03:00:18 +08:00
|
|
|
call void @g( i32 1, i32 2, i32 3, i32 4 )
|
|
|
|
ret void
|
|
|
|
}
|
|
|
|
|
2019-12-25 07:52:21 +08:00
|
|
|
define void @t2() "frame-pointer"="all" {
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKV6-LABEL: t2:
|
2011-07-09 02:50:22 +08:00
|
|
|
; CHECKV6: bx r0
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKT2D-LABEL: t2:
|
2011-05-25 12:45:27 +08:00
|
|
|
; CHECKT2D: ldr
|
|
|
|
; CHECKT2D-NEXT: ldr
|
2011-07-09 02:50:22 +08:00
|
|
|
; CHECKT2D-NEXT: bx r0
|
2015-02-28 05:17:42 +08:00
|
|
|
%tmp = load i32 ()*, i32 ()** @t ; <i32 ()*> [#uses=1]
|
2010-06-19 03:00:18 +08:00
|
|
|
%tmp.upgrd.2 = tail call i32 %tmp( ) ; <i32> [#uses=0]
|
|
|
|
ret void
|
|
|
|
}
|
|
|
|
|
2019-12-25 07:52:21 +08:00
|
|
|
define void @t3() "frame-pointer"="all" {
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKV6-LABEL: t3:
|
2011-07-09 02:50:22 +08:00
|
|
|
; CHECKV6: b _t2
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKELF-LABEL: t3:
|
2016-06-17 00:09:53 +08:00
|
|
|
; CHECKELF: b t2
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKT2D-LABEL: t3:
|
2011-07-09 02:50:22 +08:00
|
|
|
; CHECKT2D: b.w _t2
|
2011-05-23 09:57:17 +08:00
|
|
|
|
2010-07-08 09:18:23 +08:00
|
|
|
tail call void @t2( ) ; <i32> [#uses=0]
|
|
|
|
ret void
|
|
|
|
}
|
2010-12-01 07:55:39 +08:00
|
|
|
|
|
|
|
; Sibcall optimization of expanded libcalls. rdar://8707777
|
2019-12-25 07:52:21 +08:00
|
|
|
define double @t4(double %a) nounwind readonly ssp "frame-pointer"="all" {
|
2010-12-01 07:55:39 +08:00
|
|
|
entry:
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKV6-LABEL: t4:
|
2011-07-09 02:50:22 +08:00
|
|
|
; CHECKV6: b _sin
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKELF-LABEL: t4:
|
2016-06-17 00:09:53 +08:00
|
|
|
; CHECKELF: b sin
|
2010-12-01 07:55:39 +08:00
|
|
|
%0 = tail call double @sin(double %a) nounwind readonly ; <double> [#uses=1]
|
|
|
|
ret double %0
|
|
|
|
}
|
|
|
|
|
2019-12-25 07:52:21 +08:00
|
|
|
define float @t5(float %a) nounwind readonly ssp "frame-pointer"="all" {
|
2010-12-01 07:55:39 +08:00
|
|
|
entry:
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKV6-LABEL: t5:
|
2011-07-09 02:50:22 +08:00
|
|
|
; CHECKV6: b _sinf
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKELF-LABEL: t5:
|
2016-06-17 00:09:53 +08:00
|
|
|
; CHECKELF: b sinf
|
2010-12-01 07:55:39 +08:00
|
|
|
%0 = tail call float @sinf(float %a) nounwind readonly ; <float> [#uses=1]
|
|
|
|
ret float %0
|
|
|
|
}
|
|
|
|
|
|
|
|
declare float @sinf(float) nounwind readonly
|
|
|
|
|
|
|
|
declare double @sin(double) nounwind readonly
|
|
|
|
|
2019-12-25 07:52:21 +08:00
|
|
|
define i32 @t6(i32 %a, i32 %b) nounwind readnone "frame-pointer"="all" {
|
2010-12-01 07:55:39 +08:00
|
|
|
entry:
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKV6-LABEL: t6:
|
2011-07-09 02:50:22 +08:00
|
|
|
; CHECKV6: b ___divsi3
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKELF-LABEL: t6:
|
2016-06-17 00:09:53 +08:00
|
|
|
; CHECKELF: b __aeabi_idiv
|
2010-12-01 07:55:39 +08:00
|
|
|
%0 = sdiv i32 %a, %b
|
|
|
|
ret i32 %0
|
|
|
|
}
|
2011-01-25 09:28:33 +08:00
|
|
|
|
|
|
|
; Make sure the tail call instruction isn't deleted
|
|
|
|
; rdar://8309338
|
|
|
|
declare void @foo() nounwind
|
|
|
|
|
2019-12-25 07:52:21 +08:00
|
|
|
define void @t7() nounwind "frame-pointer"="all" {
|
2011-01-25 09:28:33 +08:00
|
|
|
entry:
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKT2D-LABEL: t7:
|
2015-11-18 08:40:54 +08:00
|
|
|
; CHECKT2D: it ne
|
|
|
|
; CHECKT2D-NEXT: bne.w _foo
|
|
|
|
; CHECKT2D-NEXT: push
|
|
|
|
; CHECKT2D-NEXT: mov r7, sp
|
2016-05-11 03:17:47 +08:00
|
|
|
; CHECKT2D-NEXT: bl _foo
|
2011-01-25 09:28:33 +08:00
|
|
|
br i1 undef, label %bb, label %bb1.lr.ph
|
|
|
|
|
|
|
|
bb1.lr.ph:
|
|
|
|
tail call void @foo() nounwind
|
|
|
|
unreachable
|
|
|
|
|
|
|
|
bb:
|
|
|
|
tail call void @foo() nounwind
|
|
|
|
ret void
|
|
|
|
}
|
2012-03-30 09:24:39 +08:00
|
|
|
|
|
|
|
; Make sure codegenprep is duplicating ret instructions to enable tail calls.
|
|
|
|
; rdar://11140249
|
2019-12-25 07:52:21 +08:00
|
|
|
define i32 @t8(i32 %x) nounwind ssp "frame-pointer"="all" {
|
2012-03-30 09:24:39 +08:00
|
|
|
entry:
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKT2D-LABEL: t8:
|
2012-03-30 09:24:39 +08:00
|
|
|
; CHECKT2D-NOT: push
|
|
|
|
%and = and i32 %x, 1
|
|
|
|
%tobool = icmp eq i32 %and, 0
|
|
|
|
br i1 %tobool, label %if.end, label %if.then
|
|
|
|
|
|
|
|
if.then: ; preds = %entry
|
|
|
|
; CHECKT2D: bne.w _a
|
|
|
|
%call = tail call i32 @a(i32 %x) nounwind
|
|
|
|
br label %return
|
|
|
|
|
|
|
|
if.end: ; preds = %entry
|
|
|
|
%and1 = and i32 %x, 2
|
|
|
|
%tobool2 = icmp eq i32 %and1, 0
|
|
|
|
br i1 %tobool2, label %if.end5, label %if.then3
|
|
|
|
|
|
|
|
if.then3: ; preds = %if.end
|
[Thumb] Teach ISel how to lower compares of AND bitmasks efficiently
This is essentially a recommit of r285893, but with a correctness fix. The
problem of the original commit was that this:
bic r5, r7, #31
cbz r5, .LBB2_10
got rewritten into:
lsrs r5, r7, #5
beq .LBB2_10
The result in destination register r5 is not the same and this is incorrect
when r5 is not dead. So this fix includes checking the uses of the AND
destination register. And also, compared to the original commit, some regression
tests didn't need changing anymore because of this extra check.
For completeness, this was the original commit message:
For the common pattern (CMPZ (AND x, #bitmask), #0), we can do some more
efficient instruction selection if the bitmask is one consecutive sequence of
set bits (32 - clz(bm) - ctz(bm) == popcount(bm)).
1) If the bitmask touches the LSB, then we can remove all the upper bits and
set the flags by doing one LSLS.
2) If the bitmask touches the MSB, then we can remove all the lower bits and
set the flags with one LSRS.
3) If the bitmask has popcount == 1 (only one set bit), we can shift that bit
into the sign bit with one LSLS and change the condition query from NE/EQ to
MI/PL (we could also implement this by shifting into the carry bit and
branching on BCC/BCS).
4) Otherwise, we can emit a sequence of LSLS+LSRS to remove the upper and lower
zero bits of the mask.
1-3 require only one 16-bit instruction and can elide the CMP. 4 requires two
16-bit instructions but can elide the CMP and doesn't require materializing a
complex immediate, so is also a win.
Differential Revision: https://reviews.llvm.org/D27761
llvm-svn: 289794
2016-12-15 17:38:59 +08:00
|
|
|
; CHECKT2D: bmi.w _b
|
2012-03-30 09:24:39 +08:00
|
|
|
%call4 = tail call i32 @b(i32 %x) nounwind
|
|
|
|
br label %return
|
|
|
|
|
|
|
|
if.end5: ; preds = %if.end
|
|
|
|
; CHECKT2D: b.w _c
|
|
|
|
%call6 = tail call i32 @c(i32 %x) nounwind
|
|
|
|
br label %return
|
|
|
|
|
|
|
|
return: ; preds = %if.end5, %if.then3, %if.then
|
|
|
|
%retval.0 = phi i32 [ %call, %if.then ], [ %call4, %if.then3 ], [ %call6, %if.end5 ]
|
|
|
|
ret i32 %retval.0
|
|
|
|
}
|
|
|
|
|
|
|
|
declare i32 @a(i32)
|
|
|
|
|
|
|
|
declare i32 @b(i32)
|
|
|
|
|
|
|
|
declare i32 @c(i32)
|
2012-04-10 09:51:00 +08:00
|
|
|
|
|
|
|
; PR12419
|
|
|
|
; rdar://11195178
|
|
|
|
; Use the correct input chain for the tailcall node or else the call to
|
|
|
|
; _ZN9MutexLockD1Ev would be lost.
|
|
|
|
%class.MutexLock = type { i8 }
|
|
|
|
|
|
|
|
@x = external global i32, align 4
|
|
|
|
|
2019-12-25 07:52:21 +08:00
|
|
|
define i32 @t9() nounwind "frame-pointer"="all" {
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKT2D-LABEL: t9:
|
2016-05-11 03:17:47 +08:00
|
|
|
; CHECKT2D: bl __ZN9MutexLockC1Ev
|
|
|
|
; CHECKT2D: bl __ZN9MutexLockD1Ev
|
2012-04-10 11:15:42 +08:00
|
|
|
; CHECKT2D: b.w ___divsi3
|
2012-04-10 09:51:00 +08:00
|
|
|
%lock = alloca %class.MutexLock, align 1
|
|
|
|
%1 = call %class.MutexLock* @_ZN9MutexLockC1Ev(%class.MutexLock* %lock)
|
2015-02-28 05:17:42 +08:00
|
|
|
%2 = load i32, i32* @x, align 4
|
2012-04-10 09:51:00 +08:00
|
|
|
%3 = sdiv i32 1000, %2
|
|
|
|
%4 = call %class.MutexLock* @_ZN9MutexLockD1Ev(%class.MutexLock* %lock)
|
|
|
|
ret i32 %3
|
|
|
|
}
|
|
|
|
|
|
|
|
declare %class.MutexLock* @_ZN9MutexLockC1Ev(%class.MutexLock*) unnamed_addr nounwind align 2
|
|
|
|
|
|
|
|
declare %class.MutexLock* @_ZN9MutexLockD1Ev(%class.MutexLock*) unnamed_addr nounwind align 2
|
2013-05-13 18:21:19 +08:00
|
|
|
|
|
|
|
; rdar://13827621
|
|
|
|
; Correctly preserve the input chain for the tailcall node in the bitcast case,
|
|
|
|
; otherwise the call to floorf is lost.
|
2019-12-25 07:52:21 +08:00
|
|
|
define float @libcall_tc_test2(float* nocapture %a, float %b) "frame-pointer"="all" {
|
2013-07-14 14:24:09 +08:00
|
|
|
; CHECKT2D-LABEL: libcall_tc_test2:
|
2016-05-11 03:17:47 +08:00
|
|
|
; CHECKT2D: bl _floorf
|
2013-05-13 18:21:19 +08:00
|
|
|
; CHECKT2D: b.w _truncf
|
2015-02-28 05:17:42 +08:00
|
|
|
%1 = load float, float* %a, align 4
|
2013-05-13 18:21:19 +08:00
|
|
|
%call = tail call float @floorf(float %1)
|
|
|
|
store float %call, float* %a, align 4
|
|
|
|
%call1 = tail call float @truncf(float %b)
|
|
|
|
ret float %call1
|
|
|
|
}
|
|
|
|
|
|
|
|
declare float @floorf(float) readnone
|
|
|
|
declare float @truncf(float) readnone
|