chose is having a non-memcpy/memset use and being larger than any native integer
type. Originally I chose having an access of a size smaller than the total size
of the alloca, but this caused some minor issues on the spirit benchmark where
SRoA runs again after some inlining.
This fixes <rdar://problem/8613163>.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127718 91177308-0d34-0410-b5e6-96231b3b80d8
properties.
Added the self-wrap flag for SCEV::AddRecExpr.
A slew of temporary FIXMEs indicate the intention of the no-self-wrap flag
without changing behavior in this revision.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127590 91177308-0d34-0410-b5e6-96231b3b80d8
Optimize trivial branches in CodeGenPrepare, which often get created from the
lowering of objectsize intrinsics. Unfortunately, a number of tests were relying
on llc not optimizing trivial branches, so I had to add an option to allow them
to continue to test what they originally tested.
This fixes <rdar://problem/8785296> and <rdar://problem/9112893>.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127498 91177308-0d34-0410-b5e6-96231b3b80d8
lowering of objectsize intrinsics. Unfortunately, a number of tests were relying
on llc not optimizing trivial branches, so I had to add an option to allow them
to continue to test what they originally tested.
This fixes <rdar://problem/8785296> and <rdar://problem/9112893>.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127459 91177308-0d34-0410-b5e6-96231b3b80d8
Value, not an Instruction, so casting is not necessary. Also,
it's theoretically possible that the Value is not an
Instruction, since WeakVH follows RAUWs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127427 91177308-0d34-0410-b5e6-96231b3b80d8
after it has finished all of its reassociations, because its
habit of unlinking operands and holding them in a datastructure
while working means that it's not easy to determine when an
instruction is really dead until after all its regular work is
done. rdar://9096268.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127424 91177308-0d34-0410-b5e6-96231b3b80d8
alloca as both integer and floating-point vectors of the same size. Bugpoint is
not cooperating with me, but I'll try to find a manual testcase tomorrow.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127320 91177308-0d34-0410-b5e6-96231b3b80d8
a union of a float, <2 x float>, and <4 x float>. This mostly comes up with the
use of vector intrinsics, especially in NEON when programmers know the layout of
the register file. This enables codegen to eliminate a lot of the subregister
traffic it would otherwise generate.
This commit only enables this for a small number of floating-point cases, but a
lot more integer cases. I assume this is okay for all ports, but I did not do
extensive testing of the quality of code involving i512 vectors and the like. If
there is a use case where this generates worse code than before, let me know and
we can scale it back.
This fixes <rdar://problem/9036264>.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127317 91177308-0d34-0410-b5e6-96231b3b80d8
the percentage of time spent in CodeGenPrepare when llcing 403.gcc from 12.6% to
1.8% of total llc time.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@127069 91177308-0d34-0410-b5e6-96231b3b80d8
and iprintf is available on the target. Currently iprintf is only
marked as being available on the XCore.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126935 91177308-0d34-0410-b5e6-96231b3b80d8
addressing code. On 403.gcc this almost halves CodeGenPrepare time and reduces
total llc time by 9.5%. Unfortunately, getNumUses() is still the hottest function
in llc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126782 91177308-0d34-0410-b5e6-96231b3b80d8
constant, including globals. This makes us generate much more "pretty" pattern
globals as well because it doesn't break it down to an array of bytes all the
time.
This enables us to handle stores of relocatable globals. This kicks in about
48 times in 254.gap, giving us stuff like this:
@.memset_pattern40 = internal constant [2 x %struct.TypHeader* (%struct.TypHeader*, %struct.TypHeader*)*] [%struct.TypHeader* (%struct.TypHeader*, %struct
.TypHeader*)* @IsFalse, %struct.TypHeader* (%struct.TypHeader*, %struct.TypHeader*)* @IsFalse], align 16
...
call void @memset_pattern16(i8* %scevgep5859, i8* bitcast ([2 x %struct.TypHeader* (%struct.TypHeader*, %struct.TypHeader*)*]* @.memset_pattern40 to i8*
), i64 %tmp75) nounwind
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126044 91177308-0d34-0410-b5e6-96231b3b80d8
unsplatable values into memset_pattern16 when it is available
(recent darwins). This transforms lots of strided loop stores
of ints for example, like 5 in vpr:
Formed memset: call void @memset_pattern16(i8* %4, i8* getelementptr inbounds ([16 x i8]* @.memset_pattern9, i32 0, i32 0), i64 %tmp25)
from store to: {%3,+,4}<%11> at: store i32 3, i32* %scevgep, align 4, !tbaa !4
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@126040 91177308-0d34-0410-b5e6-96231b3b80d8
taken (and used!). This prevents merging the blocks (invalidating
the block addresses) in a case like this:
#define _THIS_IP_ ({ __label__ __here; __here: (unsigned long)&&__here; })
void foo() {
printf("%p\n", _THIS_IP_);
printf("%p\n", _THIS_IP_);
printf("%p\n", _THIS_IP_);
}
which fixes PR4151.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125829 91177308-0d34-0410-b5e6-96231b3b80d8
a loop when unswitching it. It only does this in the complex case, because
everything should be fine already in the simple case.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125369 91177308-0d34-0410-b5e6-96231b3b80d8
Natural Loop Information
Loop Pass Manager
Canonicalize natural loops
Scalar Evolution Analysis
Loop Pass Manager
Induction Variable Users
Canonicalize natural loops
Induction Variable Users
Loop Strength Reduction
into this:
Scalar Evolution Analysis
Loop Pass Manager
Canonicalize natural loops
Induction Variable Users
Loop Strength Reduction
This fixes <rdar://problem/8869639>. I also filed PR9184 on doing this sort of
thing automatically, but it seems easier to just change the ordering of the
passes if this is the only case.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125254 91177308-0d34-0410-b5e6-96231b3b80d8
the active loop. This is generally desirable, and it avoids trouble
in situations such as the testcase in PR9123, though the failure
mode depends on use-list order, so it is infeasible to test.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@125065 91177308-0d34-0410-b5e6-96231b3b80d8
The DEBUG() call at line 606 demands to see raw_ostream's definition. I have no idea why this seems to only break MSVC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124545 91177308-0d34-0410-b5e6-96231b3b80d8
operand being factorized (and erased) could occur several times in Ops,
resulting in freed memory being used when the next occurrence in Ops was
analyzed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124287 91177308-0d34-0410-b5e6-96231b3b80d8
with BasicAA's DecomposeGEPExpression, which recently began
using a TargetData. This fixes PR8968, though the testcase
is awkward to reduce.
Also, update several off GetUnderlyingObject's users
which happen to have a TargetData handy to pass it in.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124134 91177308-0d34-0410-b5e6-96231b3b80d8
occurs because instcombine sinks loads and inserts phis. This kicks in
on such apps as 175.vpr, eon, 403.gcc, xalancbmk and a bunch of times in
spec2006 in some app that uses std::deque.
This resolves the last of rdar://7339113.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124090 91177308-0d34-0410-b5e6-96231b3b80d8
common cases. This triggers a surprising number of times in SPEC2K6
because min/max idioms end up doing this. For example, code from the
STL ends up looking like this to SRoA:
%202 = load i64* %__old_size, align 8, !tbaa !3
%203 = load i64* %__old_size, align 8, !tbaa !3
%204 = load i64* %__n, align 8, !tbaa !3
%205 = icmp ult i64 %203, %204
%storemerge.i = select i1 %205, i64* %__n, i64* %__old_size
%206 = load i64* %storemerge.i, align 8, !tbaa !3
We can now promote both the __n and the __old_size allocas.
This addresses another chunk of rdar://7339113, poor codegen on
stringswitch.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124088 91177308-0d34-0410-b5e6-96231b3b80d8
that have PHI or select uses of their element pointers. This can often happen
when instcombine sinks two loads into a successor, inserting a phi or select.
With this patch, we can scalarize the alloca, but the pinned elements are not
yet promoted. This is still a win for large aggregates where only one element
is used. This fixes rdar://8904039 and part of rdar://7339113 (poor codegen
on stringswitch).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124070 91177308-0d34-0410-b5e6-96231b3b80d8
handle the "Transformation preventing inst" printing,
so that -scalarrepl -debug will always print the rejected
instruction. No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@124066 91177308-0d34-0410-b5e6-96231b3b80d8
without whatever this was trying to do. When/if someone has the time to do some empirical
evaluations, it might be worth it to figure out what this code was trying to do and see if
it's worth resurrecting/fixing.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123684 91177308-0d34-0410-b5e6-96231b3b80d8
checks enabled:
1) Use '<' to compare integers in a comparison function rather than '<='.
2) Use the uniqued set DefBlocks rather than Info.DefiningBlocks to initialize
the priority queue.
The speedup of scalarrepl on test-suite + SPEC2000 + SPEC2006 is a bit less, at
just under 16% rather than 17%.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123662 91177308-0d34-0410-b5e6-96231b3b80d8
eliminating a potentially quadratic data structure, this also gives a 17%
speedup when running -scalarrepl on test-suite + SPEC2000 + SPEC2006. My initial
experiment gave a greater speedup around 25%, but I moved the dominator tree
level computation from dominator tree construction to PromoteMemToReg.
Since this approach to computing IDFs has a much lower overhead than the old
code using precomputed DFs, it is worth looking at using this new code for the
second scalarrepl pass as well.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123609 91177308-0d34-0410-b5e6-96231b3b80d8
then don't try to decimate it into its individual pieces. This will just make a mess of the
IR and is pointless if none of the elements are individually accessed. This was generating
really terrible code for std::bitset (PR8980) because it happens to be lowered by clang
as an {[8 x i8]} structure instead of {i64}.
The testcase now is optimized to:
define i64 @test2(i64 %X) {
br label %L2
L2: ; preds = %0
ret i64 %X
}
before we generated:
define i64 @test2(i64 %X) {
%sroa.store.elt = lshr i64 %X, 56
%1 = trunc i64 %sroa.store.elt to i8
%sroa.store.elt8 = lshr i64 %X, 48
%2 = trunc i64 %sroa.store.elt8 to i8
%sroa.store.elt9 = lshr i64 %X, 40
%3 = trunc i64 %sroa.store.elt9 to i8
%sroa.store.elt10 = lshr i64 %X, 32
%4 = trunc i64 %sroa.store.elt10 to i8
%sroa.store.elt11 = lshr i64 %X, 24
%5 = trunc i64 %sroa.store.elt11 to i8
%sroa.store.elt12 = lshr i64 %X, 16
%6 = trunc i64 %sroa.store.elt12 to i8
%sroa.store.elt13 = lshr i64 %X, 8
%7 = trunc i64 %sroa.store.elt13 to i8
%8 = trunc i64 %X to i8
br label %L2
L2: ; preds = %0
%9 = zext i8 %1 to i64
%10 = shl i64 %9, 56
%11 = zext i8 %2 to i64
%12 = shl i64 %11, 48
%13 = or i64 %12, %10
%14 = zext i8 %3 to i64
%15 = shl i64 %14, 40
%16 = or i64 %15, %13
%17 = zext i8 %4 to i64
%18 = shl i64 %17, 32
%19 = or i64 %18, %16
%20 = zext i8 %5 to i64
%21 = shl i64 %20, 24
%22 = or i64 %21, %19
%23 = zext i8 %6 to i64
%24 = shl i64 %23, 16
%25 = or i64 %24, %22
%26 = zext i8 %7 to i64
%27 = shl i64 %26, 8
%28 = or i64 %27, %25
%29 = zext i8 %8 to i64
%30 = or i64 %29, %28
ret i64 %30
}
In this case, instcombine was able to eliminate the nonsense, but in PR8980 enough
PHIs are in play that instcombine backs off. It's better to not generate this stuff
in the first place.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123571 91177308-0d34-0410-b5e6-96231b3b80d8
multiple uses. In some cases, all the uses are the same operation,
so instcombine can go ahead and promote the phi. In the testcase
this pushes an add out of the loop.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123568 91177308-0d34-0410-b5e6-96231b3b80d8
The basic issue is that isel (very reasonably!) expects conditional branches
to be folded, so CGP leaving around a bunch dead computation feeding
conditional branches isn't such a good idea. Just fold branches on constants
into unconditional branches.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123526 91177308-0d34-0410-b5e6-96231b3b80d8
have objectsize folding recursively simplify away their result when it
folds. It is important to catch this here, because otherwise we won't
eliminate the cross-block values at isel and other times.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123524 91177308-0d34-0410-b5e6-96231b3b80d8
potentially invalidate it (like inline asm lowering) to be sunk into
their proper place, cleaning up a ton of code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123523 91177308-0d34-0410-b5e6-96231b3b80d8
instead of DomTree/DomFrontier. This may be interesting for reducing compile
time. This is currently disabled, but seems to work just fine.
When this is enabled, we eliminate two runs of dominator frontier, one in the
"early per-function" optimizations and one in the "interlaced with inliner"
function passes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123434 91177308-0d34-0410-b5e6-96231b3b80d8
This is a minor extension of SROA to handle a special case that is
important for some ARM NEON operations. Some of the NEON intrinsics
return multiple values, which are handled as struct types containing
multiple elements of the same vector type. The corresponding return
types declared in the arm_neon.h header have equivalent arrays. We
need SROA to recognize that it can split up those arrays and structs
into separate vectors, even though they are not always accessed with
the same type. SROA already handles loads and stores of an entire
alloca by using insertvalue/extractvalue to access the individual
pieces, and that code works the same regardless of whether the type
is a struct or an array. So, all that needs to be done is to check
for compatible arrays and homogeneous structs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123381 91177308-0d34-0410-b5e6-96231b3b80d8
SROA only split up structs and arrays one level at a time, so padding can
only cause trouble if it is located in between the struct or array elements.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123380 91177308-0d34-0410-b5e6-96231b3b80d8
phi nodes. It is called from MergeBlockIntoPredecessor which is
called from GVN, which claims to preserve these.
I'm skeptical that this is the actual problem behind PR8954, but
this is a stab in the right direction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123222 91177308-0d34-0410-b5e6-96231b3b80d8
without informing memdep. This could cause nondeterminstic weirdness
based on where instructions happen to get allocated, and will hopefully
breath some life into some broken testers.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123124 91177308-0d34-0410-b5e6-96231b3b80d8
larger memsets. Among other things, this fixes rdar://8760394 and
allows us to handle "Example 2" from http://blog.regehr.org/archives/320,
compiling it into a single 4096-byte memset:
_mad_synth_mute: ## @mad_synth_mute
## BB#0: ## %entry
pushq %rax
movl $4096, %esi ## imm = 0x1000
callq ___bzero
popq %rax
ret
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123089 91177308-0d34-0410-b5e6-96231b3b80d8
that it was leaving in loops after rotation (between the original latch
block and the original header.
With this change, it is possible for rotated loops to have just a single
basic block, which is useful.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123075 91177308-0d34-0410-b5e6-96231b3b80d8
1. Rip out LoopRotate's domfrontier updating code. It isn't
needed now that LICM doesn't use DF and it is super complex
and gross.
2. Make DomTree updating code a lot simpler and faster. The
old loop over all the blocks was just to find a block??
3. Change the code that inserts the new preheader to just use
SplitCriticalEdge instead of doing an overcomplex
reimplementation of it.
No behavior change, except for the name of the inserted preheader.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123072 91177308-0d34-0410-b5e6-96231b3b80d8
they all ready do). This removes two dominator recomputations prior to isel,
which is a 1% improvement in total llc time for 403.gcc.
The only potentially suspect thing is making GCStrategy recompute dominators if
it used a custom lowering strategy.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123064 91177308-0d34-0410-b5e6-96231b3b80d8
them into the loop preheader, eliminating silly instructions like
"icmp i32 0, 100" in fixed tripcount loops. This also better exposes the
bigger problem with loop rotate that I'd like to fix: once this has been
folded, the duplicated conditional branch *often* turns into an uncond branch.
Not aggressively handling this is pessimizing later loop optimizations
somethin' fierce by making "dominates all exit blocks" checks fail.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123060 91177308-0d34-0410-b5e6-96231b3b80d8
1. Take a flags argument instead of a bool. This makes
it more clear to the reader what it is used for.
2. Add a flag that says that "remapping a value not in the
map is ok".
3. Reimplement MapValue to share a bunch of code and be a lot
more efficient. For lookup failures, don't drop null values
into the map.
4. Using the new flag a bunch of code can vaporize in LinkModules
and LoopUnswitch, kill it.
No functionality change.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123058 91177308-0d34-0410-b5e6-96231b3b80d8
map from ValueMapper.h (giving us access to its utilities)
and add a fastpath in the loop rotation code, avoiding expensive
ssa updator manipulation for values with nothing to update.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@123057 91177308-0d34-0410-b5e6-96231b3b80d8
into a separate function, so that it can be called from a loop using a worklist
rather than a loop traversing a whole basic block.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122943 91177308-0d34-0410-b5e6-96231b3b80d8
step is to only process instructions in subloops if they have been modified by
an earlier simplification.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122869 91177308-0d34-0410-b5e6-96231b3b80d8
skipping them, but it should probably use a worklist and only revisit those
instructions in subloops that have actually changed. It should probably also
use a worklist after the first iteration like instsimplify now does. Regardless,
it's only 0.3% of opt -O2 time on 403.gcc if it replaces the instcombine placed
in the middle of the loop passes.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122868 91177308-0d34-0410-b5e6-96231b3b80d8
fewer things into the value numbering maps, but any speedup is beneath the noise threshold on my machine
on 403.gcc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122844 91177308-0d34-0410-b5e6-96231b3b80d8
avoids adding them to the various value numbering tables, resulting in a minor (~3%) speedup for GVN
on 40.gcc.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@122819 91177308-0d34-0410-b5e6-96231b3b80d8