Summary:
Add and instructions immediately after loads that only have their low
bits used, assuming that the (and (load x) c) will be matched as a
extload and the ands/truncs fed by the extload will be removed by isel.
Reviewers: mcrosier, qcolombet, ab
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D14584
llvm-svn: 253722
This is another step towards allowing SimplifyCFG to speculate harder, but then have
CGP clean things up if the target doesn't like it.
Previous patches in this series:
http://reviews.llvm.org/D12882http://reviews.llvm.org/D13297
D13297 should catch most expensive ops, but speculation of cttz/ctlz requires special
handling because of weirdness in the intrinsic definition for handling a zero input
(that definition can probably be blamed on x86).
For example, if we have the usual speculated-by-select expensive op pattern like this:
%tobool = icmp eq i64 %A, 0
%0 = tail call i64 @llvm.cttz.i64(i64 %A, i1 true) ; is_zero_undef == true
%cond = select i1 %tobool, i64 64, i64 %0
ret i64 %cond
There's an instcombine that will turn it into:
%0 = tail call i64 @llvm.cttz.i64(i64 %A, i1 false) ; is_zero_undef == false
This CGP patch is looking for that case and despeculating it back into:
entry:
%tobool = icmp eq i64 %A, 0
br i1 %tobool, label %cond.end, label %cond.true
cond.true:
%0 = tail call i64 @llvm.cttz.i64(i64 %A, i1 true) ; is_zero_undef == true
br label %cond.end
cond.end:
%cond = phi i64 [ %0, %cond.true ], [ 64, %entry ]
ret i64 %cond
This unfortunately may lead to poorer codegen (see the changes in the existing x86 test),
but if we increase speculation in SimplifyCFG (the next step in this patch series), then
we should avoid those kinds of cases in the first place.
The need for this patch was originally mentioned here:
http://reviews.llvm.org/D7506
with follow-up here:
http://reviews.llvm.org/D7554
Differential Revision: http://reviews.llvm.org/D14630
llvm-svn: 253573
Note, this was reviewed (and more details are in) http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20151109/312083.html
These intrinsics currently have an explicit alignment argument which is
required to be a constant integer. It represents the alignment of the
source and dest, and so must be the minimum of those.
This change allows source and dest to each have their own alignments
by using the alignment attribute on their arguments. The alignment
argument itself is removed.
There are a few places in the code for which the code needs to be
checked by an expert as to whether using only src/dest alignment is
safe. For those places, they currently take the minimum of src/dest
alignments which matches the current behaviour.
For example, code which used to read:
call void @llvm.memcpy.p0i8.p0i8.i32(i8* %dest, i8* %src, i32 500, i32 8, i1 false)
will now read:
call void @llvm.memcpy.p0i8.p0i8.i32(i8* align 8 %dest, i8* align 8 %src, i32 500, i1 false)
For out of tree owners, I was able to strip alignment from calls using sed by replacing:
(call.*llvm\.memset.*)i32\ [0-9]*\,\ i1 false\)
with:
$1i1 false)
and similarly for memmove and memcpy.
I then added back in alignment to test cases which needed it.
A similar commit will be made to clang which actually has many differences in alignment as now
IRBuilder can generate different source/dest alignments on calls.
In IRBuilder itself, a new argument was added. Instead of calling:
CreateMemCpy(Dst, Src, getInt64(Size), DstAlign, /* isVolatile */ false)
you now call
CreateMemCpy(Dst, Src, getInt64(Size), DstAlign, SrcAlign, /* isVolatile */ false)
There is a temporary class (IntegerAlignment) which takes the source alignment and rejects
implicit conversion from bool. This is to prevent isVolatile here from passing its default
parameter to the source alignment.
Note, changes in future can now be made to codegen. I didn't change anything here, but this
change should enable better memcpy code sequences.
Reviewed by Hal Finkel.
llvm-svn: 253511
This is a redo of r251849 except the tests have been split into arch-specific folders
to hopefully make the bots happy.
This is a follow-up from the discussion in D12965. The block-at-a-time limitation of
SelectionDAG also came up in D13297.
Without the InstCombine change from D12965, I don't expect this patch to make any
difference in the real world because InstCombine does not shrink cases like this in
visitSwitchInst(). But we need to have this CGP safety harness in place before
proceeding with any shrinkage in D12965, so we won't generate extra extends for compares.
I've opted for IR regression tests in the patch because that seems like a clearer way to
test the transform, but PowerPC CodeGen for an i16 widening test is shown below. x86
will need more work to solve: https://llvm.org/bugs/show_bug.cgi?id=22473
Before:
BB#0:
mr 4, 3
extsh. 3, 4
ble 0, .LBB0_5
BB#1:
cmpwi 3, 99
bgt 0, .LBB0_9
BB#2:
rlwinm 4, 4, 0, 16, 31 <--- 32-bit mask/extend
li 3, 0
cmplwi 4, 1
beqlr 0
BB#3:
cmplwi 4, 10
bne 0, .LBB0_12
BB#4:
li 3, 1
blr
.LBB0_5:
rlwinm 3, 4, 0, 16, 31 <--- 32-bit mask/extend
cmplwi 3, 65436
beq 0, .LBB0_13
BB#6:
cmplwi 3, 65526
beq 0, .LBB0_15
BB#7:
cmplwi 3, 65535
bne 0, .LBB0_12
BB#8:
li 3, 4
blr
.LBB0_9:
rlwinm 3, 4, 0, 16, 31 <--- 32-bit mask/extend
cmplwi 3, 100
beq 0, .LBB0_14
...
After:
BB#0:
rlwinm 4, 3, 0, 16, 31 <--- mask/extend to 32-bit and then use that for comparisons
cmpwi 4, 999
ble 0, .LBB0_5
BB#1:
lis 3, 0
ori 3, 3, 65525
cmpw 4, 3
bgt 0, .LBB0_9
BB#2:
cmplwi 4, 1000
beq 0, .LBB0_14
BB#3:
cmplwi 4, 65436
bne 0, .LBB0_13
BB#4:
li 3, 6
blr
.LBB0_5:
li 3, 0
cmplwi 4, 1
beqlr 0
BB#6:
cmplwi 4, 10
beq 0, .LBB0_12
BB#7:
cmplwi 4, 100
bne 0, .LBB0_13
BB#8:
li 3, 2
blr
.LBB0_9:
cmplwi 4, 65526
beq 0, .LBB0_15
BB#10:
cmplwi 4, 65535
bne 0, .LBB0_13
...
Differential Revision: http://reviews.llvm.org/D13532
llvm-svn: 251857
This is a follow-up from the discussion in D12965. The block-at-a-time limitation of
SelectionDAG also came up in D13297.
Without the InstCombine change from D12965, I don't expect this patch to make any
difference in the real world because InstCombine does not shrink cases like this in
visitSwitchInst(). But we need to have this CGP safety harness in place before
proceeding with any shrinkage in D12965, so we won't generate extra extends for compares.
I've opted for IR regression tests in the patch because that seems like a clearer way to
test the transform, but PowerPC CodeGen for an i16 widening test is shown below. x86
will need more work to solve: https://llvm.org/bugs/show_bug.cgi?id=22473
Before:
BB#0:
mr 4, 3
extsh. 3, 4
ble 0, .LBB0_5
BB#1:
cmpwi 3, 99
bgt 0, .LBB0_9
BB#2:
rlwinm 4, 4, 0, 16, 31 <--- 32-bit mask/extend
li 3, 0
cmplwi 4, 1
beqlr 0
BB#3:
cmplwi 4, 10
bne 0, .LBB0_12
BB#4:
li 3, 1
blr
.LBB0_5:
rlwinm 3, 4, 0, 16, 31 <--- 32-bit mask/extend
cmplwi 3, 65436
beq 0, .LBB0_13
BB#6:
cmplwi 3, 65526
beq 0, .LBB0_15
BB#7:
cmplwi 3, 65535
bne 0, .LBB0_12
BB#8:
li 3, 4
blr
.LBB0_9:
rlwinm 3, 4, 0, 16, 31 <--- 32-bit mask/extend
cmplwi 3, 100
beq 0, .LBB0_14
...
After:
BB#0:
rlwinm 4, 3, 0, 16, 31 <--- mask/extend to 32-bit and then use that for comparisons
cmpwi 4, 999
ble 0, .LBB0_5
BB#1:
lis 3, 0
ori 3, 3, 65525
cmpw 4, 3
bgt 0, .LBB0_9
BB#2:
cmplwi 4, 1000
beq 0, .LBB0_14
BB#3:
cmplwi 4, 65436
bne 0, .LBB0_13
BB#4:
li 3, 6
blr
.LBB0_5:
li 3, 0
cmplwi 4, 1
beqlr 0
BB#6:
cmplwi 4, 10
beq 0, .LBB0_12
BB#7:
cmplwi 4, 100
bne 0, .LBB0_13
BB#8:
li 3, 2
blr
.LBB0_9:
cmplwi 4, 65526
beq 0, .LBB0_15
BB#10:
cmplwi 4, 65535
bne 0, .LBB0_13
...
Differential Revision: http://reviews.llvm.org/D13532
llvm-svn: 251849
When the target does not support these intrinsics they should be converted to a chain of scalar load or store operations.
If the mask is not constant, the scalarizer will build a chain of conditional basic blocks.
I added isLegalMaskedGather() isLegalMaskedScatter() APIs.
Differential Revision: http://reviews.llvm.org/D13722
llvm-svn: 251237
When we have to convert the masked.load, masked.store to scalar code, we generate a chain of conditional basic blocks.
I added optimization for constant mask vector.
Differential Revision: http://reviews.llvm.org/D13855
llvm-svn: 250893
This was originally checked in at r250527, but reverted at r250570 because of PR25222.
There were at least 2 problems:
1. The cost check was checking for an instruction with an exact cost of TCC_Expensive;
that should have been >=.
2. The cause of the clang stage 1 failures was illegally sinking 'call' instructions;
we can't sink instructions that may have side effects / are not safe to execute speculatively.
Fixed those conditions in sinkSelectOperand() and added test cases.
Original commit message:
This is a follow-up to the discussion in D12882.
Ideally, we would like SimplifyCFG to be able to form select instructions even when the operands
are expensive (as defined by the TTI cost model) because that may expose further optimizations.
However, we would then like a later pass like CodeGenPrepare to undo that transformation if the
target would likely benefit from not speculatively executing an expensive op (this patch).
Once we have this safety mechanism in place, we can adjust SimplifyCFG to restore its
select-formation behavior that changed with r248439.
Differential Revision: http://reviews.llvm.org/D13297
llvm-svn: 250743
Originally I planned to use the same interface for masked gather/scatter and set isConsecutive to "false" in this case.
Now I'm implementing masked gather/scatter and see that the interface is inconvenient. I want to add interfaces isLegalMaskedGather() / isLegalMaskedScatter() instead of using the "Consecutive" parameter in the existing interfaces.
Differential Revision: http://reviews.llvm.org/D13850
llvm-svn: 250686
Ideally, we would like SimplifyCFG to be able to form select instructions even when the operands
are expensive (as defined by the TTI cost model) because that may expose further optimizations.
However, we would then like a later pass like CodeGenPrepare to undo that transformation if the
target would likely benefit from not speculatively executing an expensive op (this patch).
Once we have this safety mechanism in place, we can adjust SimplifyCFG to restore its
select-formation behavior that changed with r248439.
Differential Revision: http://reviews.llvm.org/D13297
llvm-svn: 250527
This patch uses the metadata defined in D12341 to avoid creating an unpredictable branch.
Differential Revision: http://reviews.llvm.org/D12342
llvm-svn: 246692
Create wrapper methods in the Function class for the OptimizeForSize and MinSize
attributes. We want to hide the logic of "or'ing" them together when optimizing
just for size (-Os).
Currently, we are not consistent about this and rely on a front-end to always set
OptimizeForSize (-Os) if MinSize (-Oz) is on. Thus, there are 18 FIXME changes here
that should be added as follow-on patches with regression tests.
This patch is NFC-intended: it just replaces existing direct accesses of the attributes
by the equivalent wrapper call.
Differential Revision: http://reviews.llvm.org/D11734
llvm-svn: 243994
Summary:
This change is part of a series of commits dedicated to have a single
DataLayout during compilation by using always the one owned by the
module.
Reviewers: echristo
Subscribers: jholewinski, llvm-commits, rafael, yaron.keren
Differential Revision: http://reviews.llvm.org/D11040
From: Mehdi Amini <mehdi.amini@apple.com>
llvm-svn: 241778
Summary:
This change is part of a series of commits dedicated to have a single
DataLayout during compilation by using always the one owned by the
module.
Reviewers: echristo
Subscribers: jholewinski, ted, yaron.keren, rafael, llvm-commits
Differential Revision: http://reviews.llvm.org/D11028
From: Mehdi Amini <mehdi.amini@apple.com>
llvm-svn: 241775
Summary:
SelectionDAG itself is not invoking directly the DataLayout in the
TargetMachine, but the "TargetLowering" class is still using it. I'll
address it in a following commit.
This change is part of a series of commits dedicated to have a single
DataLayout during compilation by using always the one owned by the
module.
Reviewers: echristo
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D11000
From: Mehdi Amini <mehdi.amini@apple.com>
llvm-svn: 241618
Summary:
This change is part of a series of commits dedicated to have a single
DataLayout during compilation by using always the one owned by the
module.
Reviewers: echristo
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D10986
From: Mehdi Amini <mehdi.amini@apple.com>
llvm-svn: 241614
The patch is generated using this command:
tools/clang/tools/extra/clang-tidy/tool/run-clang-tidy.py -fix \
-checks=-*,llvm-namespace-comment -header-filter='llvm/.*|clang/.*' \
llvm/lib/
Thanks to Eugene Kosov for the original patch!
llvm-svn: 240137
It's been used before to avoid infinite loops caused by separate CGP
optimizations undoing one another. We found one more such issue
caused by r238054. To avoid it, generalize the "InsertedTruncs"
set to any inst, and use it to avoid touching those again.
llvm-svn: 239938
The usual CodeGenPrepare trickery, on a target-specific intrinsic.
Without this, the expansion of atomics will usually have the zext
be hoisted out of the loop, defeating the various patterns we have
to catch this precise case.
Differential Revision: http://reviews.llvm.org/D9930
llvm-svn: 238054
We already had a method to iterate over all the incoming values of a PHI. This just changes all eligible code to use it.
Ineligible code included anything which cared about the index, or was also trying to get the i'th incoming BB.
llvm-svn: 237169
Summary:
The original code inserted new instructions by following a
Create->Remove->ReInsert flow. This patch removes the unnecessary
Remove->ReInsert part by setting up the InsertPoint correctly at the
very beginning. This change does not introduce any functionality change.
Patch by Chen Li!
Reviewers: reames, AndyAyers, sanjoy
Reviewed By: sanjoy
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9687
llvm-svn: 237070
Summary:
In RewriteStatepointsForGC pass, we create a gc_relocate intrinsic for
each relocated pointer, and the gc_relocate has the same type with the
pointer. During the creation of gc_relocate intrinsic, llvm requires to
mangle its type. However, llvm does not support mangling of all possible
types. RewriteStatepointsForGC will hit an assertion failure when it
tries to create a gc_relocate for pointer to vector of pointers because
mangling for vector of pointers is not supported.
This patch changes the way RewriteStatepointsForGC pass creates
gc_relocate. For each relocated pointer, we erase the type of pointers
and create an unified gc_relocate of type i8 addrspace(1)*. Then a
bitcast is inserted to convert the gc_relocate to the correct type. In
this way, gc_relocate does not need to deal with different types of
pointers and the unsupported type mangling is no longer a problem. This
change would also ease further merge when LLVM erases types of pointers
and introduces an unified pointer type.
Some minor changes are also introduced to gc_relocate related part in
InstCombineCalls, CodeGenPrepare, and Verifier accordingly.
Patch by Chen Li!
Reviewers: reames, AndyAyers, sanjoy
Reviewed By: sanjoy
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D9592
llvm-svn: 237009
Fill in the TODO in CodeGenPrepare::OptimizeCallInst so that global
variables that are passed to memory intrinsics are aligned in the same
way that allocas are.
Differential Revision: http://reviews.llvm.org/D8421
llvm-svn: 234735
The patch is generated using clang-tidy misc-use-override check.
This command was used:
tools/clang/tools/extra/clang-tidy/tool/run-clang-tidy.py \
-checks='-*,misc-use-override' -header-filter='llvm|clang' \
-j=32 -fix -format
http://reviews.llvm.org/D8925
llvm-svn: 234679
r234638 chained another transform below which was tripping over the
deleted instruction. Use after free found by asan in many regression
tests.
llvm-svn: 234654
Summary:
This change moves creating calls to `llvm.uadd.with.overflow` from
InstCombine to CodeGenPrep. Combining overflow check patterns into
calls to the said intrinsic in InstCombine inhibits optimization because
it introduces an intrinsic call that not all other transforms and
analyses understand.
Depends on D8888.
Reviewers: majnemer, atrick
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D8889
llvm-svn: 234638
The changes to InstCombine do seem a bit silly - it doesn't make
anything obviously better to have the caller access the pointers element
type (the thing I'm trying to remove) than the GEP itself, but it's a
helpful migration step. This will allow me to more obviously lock down
GEP (& Load, etc) API usage, then fix all the code that accesses pointer
element types except the places that need to be removed (most of the
InstCombines) anyway - at which point I'll need to just remove all that
code because it won't be meaningful anymore (there will be no pointer
types, so no bitcasts to combine)
llvm-svn: 233126
Memcpy, and other memory intrinsics, typically tries to use LDM/STM if
the source and target addresses are 4-byte aligned. In CodeGenPrepare
look for calls to memory intrinsics and, if the object is on the
stack, 4-byte align it if it's large enough that we expect that memcpy
would want to use LDM/STM to copy it.
Differential Revision: http://reviews.llvm.org/D7908
llvm-svn: 232627
- Use TargetLowering to check for the actual cost of each extension.
- Provide a factorized method to check for the cost of an extension:
TargetLowering::isExtFree.
- Provide a virtual method TargetLowering::isExtFreeImpl for targets to be able
to tune the cost of non-free extensions.
This refactoring offers a better granularity to model what really happens on
different targets.
No performance changes and very few code differences.
Part of <rdar://problem/19267165>
llvm-svn: 231855
Summary:
Now that the DataLayout is a mandatory part of the module, let's start
cleaning the codebase. This patch is a first attempt at doing that.
This patch is not exactly NFC as for instance some places were passing
a nullptr instead of the DataLayout, possibly just because there was a
default value on the DataLayout argument to many functions in the API.
Even though it is not purely NFC, there is no change in the
validation.
I turned as many pointer to DataLayout to references, this helped
figuring out all the places where a nullptr could come up.
I had initially a local version of this patch broken into over 30
independant, commits but some later commit were cleaning the API and
touching part of the code modified in the previous commits, so it
seemed cleaner without the intermediate state.
Test Plan:
Reviewers: echristo
Subscribers: llvm-commits
From: Mehdi Amini <mehdi.amini@apple.com>
llvm-svn: 231740
that is iterating over it
Inserting elements into a `DenseMap` invalidated iterators pointing
into the `DenseMap` instance.
Differential Revision: http://reviews.llvm.org/D7924
llvm-svn: 230719
a lookup, pass that in rather than use a naked call to getSubtargetImpl.
This involved passing down and around either a TargetMachine or
TargetRegisterInfo. Update all callers/definitions around the targets
and SelectionDAG.
llvm-svn: 230699
Canonicalize access to function attributes to use the simpler API.
getAttributes().getAttribute(AttributeSet::FunctionIndex, Kind)
=> getFnAttribute(Kind)
getAttributes().hasAttribute(AttributeSet::FunctionIndex, Kind)
=> hasFnAttribute(Kind)
Also, add `Function::getFnStackAlignment()`, and canonicalize:
getAttributes().getStackAlignment(AttributeSet::FunctionIndex)
=> getFnStackAlignment()
llvm-svn: 229208
SimplifyCFG now knows how to speculate calls to intrinsic cttz/ctlz that are
'cheap' for the target. Therefore, some of the logic in CodeGenPrepare
that was originally added at revision 224899 can now be removed.
This patch is basically a no functional change. It removes the duplicated
logic in CodeGenPrepare and converts all the existing target specific tests
for cttz/ctlz into SimplifyCFG tests.
Differential Revision: http://reviews.llvm.org/D7608
llvm-svn: 229105
getTTI method used to get an actual TTI object.
No functionality changed. This just threads the argument and ensures
code like the inliner can correctly look up the callee's TTI rather than
using a fixed one.
The next change will use this to implement per-function subtarget usage
by TTI. The changes after that should eliminate the need for FTTI as that
will have become the default.
llvm-svn: 227730
type erased interface and a single analysis pass rather than an
extremely complex analysis group.
The end result is that the TTI analysis can contain a type erased
implementation that supports the polymorphic TTI interface. We can build
one from a target-specific implementation or from a dummy one in the IR.
I've also factored all of the code into "mix-in"-able base classes,
including CRTP base classes to facilitate calling back up to the most
specialized form when delegating horizontally across the surface. These
aren't as clean as I would like and I'm planning to work on cleaning
some of this up, but I wanted to start by putting into the right form.
There are a number of reasons for this change, and this particular
design. The first and foremost reason is that an analysis group is
complete overkill, and the chaining delegation strategy was so opaque,
confusing, and high overhead that TTI was suffering greatly for it.
Several of the TTI functions had failed to be implemented in all places
because of the chaining-based delegation making there be no checking of
this. A few other functions were implemented with incorrect delegation.
The message to me was very clear working on this -- the delegation and
analysis group structure was too confusing to be useful here.
The other reason of course is that this is *much* more natural fit for
the new pass manager. This will lay the ground work for a type-erased
per-function info object that can look up the correct subtarget and even
cache it.
Yet another benefit is that this will significantly simplify the
interaction of the pass managers and the TargetMachine. See the future
work below.
The downside of this change is that it is very, very verbose. I'm going
to work to improve that, but it is somewhat an implementation necessity
in C++ to do type erasure. =/ I discussed this design really extensively
with Eric and Hal prior to going down this path, and afterward showed
them the result. No one was really thrilled with it, but there doesn't
seem to be a substantially better alternative. Using a base class and
virtual method dispatch would make the code much shorter, but as
discussed in the update to the programmer's manual and elsewhere,
a polymorphic interface feels like the more principled approach even if
this is perhaps the least compelling example of it. ;]
Ultimately, there is still a lot more to be done here, but this was the
huge chunk that I couldn't really split things out of because this was
the interface change to TTI. I've tried to minimize all the other parts
of this. The follow up work should include at least:
1) Improving the TargetMachine interface by having it directly return
a TTI object. Because we have a non-pass object with value semantics
and an internal type erasure mechanism, we can narrow the interface
of the TargetMachine to *just* do what we need: build and return
a TTI object that we can then insert into the pass pipeline.
2) Make the TTI object be fully specialized for a particular function.
This will include splitting off a minimal form of it which is
sufficient for the inliner and the old pass manager.
3) Add a new pass manager analysis which produces TTI objects from the
target machine for each function. This may actually be done as part
of #2 in order to use the new analysis to implement #2.
4) Work on narrowing the API between TTI and the targets so that it is
easier to understand and less verbose to type erase.
5) Work on narrowing the API between TTI and its clients so that it is
easier to understand and less verbose to forward.
6) Try to improve the CRTP-based delegation. I feel like this code is
just a bit messy and exacerbating the complexity of implementing
the TTI in each target.
Many thanks to Eric and Hal for their help here. I ended up blocked on
this somewhat more abruptly than I expected, and so I appreciate getting
it sorted out very quickly.
Differential Revision: http://reviews.llvm.org/D7293
llvm-svn: 227669
a DominatorTree argument as that is the analysis that it wants to
update.
This removes the last non-loop utility function in Utils/ which accepts
a raw Pass argument.
llvm-svn: 226537
The pass is really just a means of accessing a cached instance of the
TargetLibraryInfo object, and this way we can re-use that object for the
new pass manager as its result.
Lots of delta, but nothing interesting happening here. This is the
common pattern that is developing to allow analyses to live in both the
old and new pass manager -- a wrapper pass in the old pass manager
emulates the separation intrinsic to the new pass manager between the
result and pass for analyses.
llvm-svn: 226157
While the term "Target" is in the name, it doesn't really have to do
with the LLVM Target library -- this isn't an abstraction which LLVM
targets generally need to implement or extend. It has much more to do
with modeling the various runtime libraries on different OSes and with
different runtime environments. The "target" in this sense is the more
general sense of a target of cross compilation.
This is in preparation for porting this analysis to the new pass
manager.
No functionality changed, and updates inbound for Clang and Polly.
llvm-svn: 226078
The transform is somewhat involved, but the basic idea is simple: find
derived pointers that have been offset from the base pointer using gep
and replace the relocate of the derived pointer with a gep to the
relocated base pointer (with the same offset).
llvm-svn: 226060
type (in addition to the memory type).
The *LoadExt* legalization handling used to only have one type, the
memory type. This forced users to assume that as long as the extload
for the memory type was declared legal, and the result type was legal,
the whole extload was legal.
However, this isn't always the case. For instance, on X86, with AVX,
this is legal:
v4i32 load, zext from v4i8
but this isn't:
v4i64 load, zext from v4i8
Whereas v4i64 is (arguably) legal, even without AVX2.
Note that the same thing was done a while ago for truncstores (r46140),
but I assume no one needed it yet for extloads, so here we go.
Calls to getLoadExtAction were changed to add the value type, found
manually in the surrounding code.
Calls to setLoadExtAction were mechanically changed, by wrapping the
call in a loop, to match previous behavior. The loop iterates over
the MVT subrange corresponding to the memory type (FP vectors, etc...).
I also pulled neighboring setTruncStoreActions into some of the loops;
those shouldn't make a difference, as the additional types are illegal.
(e.g., i128->i1 truncstores on PPC.)
No functional change intended.
Differential Revision: http://reviews.llvm.org/D6532
llvm-svn: 225421
This patch improves the logic added at revision 224899 (see review D6728) that
teaches the backend when it is profitable to speculate calls to cttz/ctlz.
The original algorithm conservatively avoided speculating more than one
instruction from a basic block in a control flow grap modelling an if-statement.
In particular, the only allowed instruction (excluding the terminator) was a
call to cttz/ctlz. However, there are cases where we could be less conservative
and still be able to speculate a call to cttz/ctlz.
With this patch, CodeGenPrepare now tries to speculate a cttz/ctlz if the
result is zero extended/truncated in the same basic block, and the zext/trunc
instruction is "free" for the target.
Added new test cases to CodeGen/X86/cttz-ctlz.ll
Differential Revision: http://reviews.llvm.org/D6853
llvm-svn: 225274
If the control flow is modelling an if-statement where the only instruction in
the 'then' basic block (excluding the terminator) is a call to cttz/ctlz,
CodeGenPrepare can try to speculate the cttz/ctlz call and simplify the control
flow graph.
Example:
\code
entry:
%cmp = icmp eq i64 %val, 0
br i1 %cmp, label %end.bb, label %then.bb
then.bb:
%c = tail call i64 @llvm.cttz.i64(i64 %val, i1 true)
br label %end.bb
end.bb:
%cond = phi i64 [ %c, %then.bb ], [ 64, %entry]
\code
In this example, basic block %then.bb is taken if value %val is not zero.
Also, the phi node in %end.bb would propagate the size-of in bits of %val
only if %val is equal to zero.
With this patch, CodeGenPrepare will try to hoist the call to cttz from %then.bb
into basic block %entry only if cttz is cheap to speculate for the target.
Added two new hooks in TargetLowering.h to let targets customize the behavior
(i.e. decide whether it is cheap or not to speculate calls to cttz/ctlz). The
two new methods are 'isCheapToSpeculateCtlz' and 'isCheapToSpeculateCttz'.
By default, both methods return 'false'.
On X86, method 'isCheapToSpeculateCtlz' returns true only if the target has
LZCNT. Method 'isCheapToSpeculateCttz' only returns true if the target has BMI.
Differential Revision: http://reviews.llvm.org/D6728
llvm-svn: 224899
Masked vector intrinsics are a part of common LLVM IR, but they are really supported on AVX2 and AVX-512 targets. I added a code that translates masked intrinsic for all other targets. The masked vector intrinsic is converted to a chain of scalar operations inside conditional basic blocks.
http://reviews.llvm.org/D6436
llvm-svn: 224897
The type promotion helper does not support vector type, so when make
such it does not kick in in such cases.
Original commit message:
[CodeGenPrepare] Move sign/zero extensions near loads using type promotion.
This patch extends the optimization in CodeGenPrepare that moves a sign/zero
extension near a load when the target can combine them. The optimization may
promote any operations between the extension and the load to make that possible.
Although this optimization may be beneficial for all targets, in particular
AArch64, this is enabled for X86 only as I have not benchmarked it for other
targets yet.
** Context **
Most targets feature extended loads, i.e., loads that perform a zero or sign
extension for free. In that context it is interesting to expose such pattern in
CodeGenPrepare so that the instruction selection pass can form such loads.
Sometimes, this pattern is blocked because of instructions between the load and
the extension. When those instructions are promotable to the extended type, we
can expose this pattern.
** Motivating Example **
Let us consider an example:
define void @foo(i8* %addr1, i32* %addr2, i8 %a, i32 %b) {
%ld = load i8* %addr1
%zextld = zext i8 %ld to i32
%ld2 = load i32* %addr2
%add = add nsw i32 %ld2, %zextld
%sextadd = sext i32 %add to i64
%zexta = zext i8 %a to i32
%addza = add nsw i32 %zexta, %zextld
%sextaddza = sext i32 %addza to i64
%addb = add nsw i32 %b, %zextld
%sextaddb = sext i32 %addb to i64
call void @dummy(i64 %sextadd, i64 %sextaddza, i64 %sextaddb)
ret void
}
As it is, this IR generates the following assembly on x86_64:
[...]
movzbl (%rdi), %eax # zero-extended load
movl (%rsi), %es # plain load
addl %eax, %esi # 32-bit add
movslq %esi, %rdi # sign extend the result of add
movzbl %dl, %edx # zero extend the first argument
addl %eax, %edx # 32-bit add
movslq %edx, %rsi # sign extend the result of add
addl %eax, %ecx # 32-bit add
movslq %ecx, %rdx # sign extend the result of add
[...]
The throughput of this sequence is 7.45 cycles on Ivy Bridge according to IACA.
Now, by promoting the additions to form more extended loads we would generate:
[...]
movzbl (%rdi), %eax # zero-extended load
movslq (%rsi), %rdi # sign-extended load
addq %rax, %rdi # 64-bit add
movzbl %dl, %esi # zero extend the first argument
addq %rax, %rsi # 64-bit add
movslq %ecx, %rdx # sign extend the second argument
addq %rax, %rdx # 64-bit add
[...]
The throughput of this sequence is 6.15 cycles on Ivy Bridge according to IACA.
This kind of sequences happen a lot on code using 32-bit indexes on 64-bit
architectures.
Note: The throughput numbers are similar on Sandy Bridge and Haswell.
** Proposed Solution **
To avoid the penalty of all these sign/zero extensions, we merge them in the
loads at the beginning of the chain of computation by promoting all the chain of
computation on the extended type. The promotion is done if and only if we do not
introduce new extensions, i.e., if we do not degrade the code quality.
To achieve this, we extend the existing “move ext to load” optimization with the
promotion mechanism introduced to match larger patterns for addressing mode
(r200947).
The idea of this extension is to perform the following transformation:
ext(promotableInst1(...(promotableInstN(load))))
=>
promotedInst1(...(promotedInstN(ext(load))))
The promotion mechanism in that optimization is enabled by a new TargetLowering
switch, which is off by default. In other words, by default, the optimization
performs the “move ext to load” optimization as it was before this patch.
** Performance **
Configuration: x86_64: Ivy Bridge fixed at 2900MHz running OS X 10.10.
Tested Optimization Levels: O3/Os
Tests: llvm-testsuite + externals.
Results:
- No regression beside noise.
- Improvements:
CINT2006/473.astar: ~2%
Benchmarks/PAQ8p: ~2%
Misc/perlin: ~3%
The results are consistent for both O3 and Os.
<rdar://problem/18310086>
llvm-svn: 224402
This patch extends the optimization in CodeGenPrepare that moves a sign/zero
extension near a load when the target can combine them. The optimization may
promote any operations between the extension and the load to make that possible.
Although this optimization may be beneficial for all targets, in particular
AArch64, this is enabled for X86 only as I have not benchmarked it for other
targets yet.
** Context **
Most targets feature extended loads, i.e., loads that perform a zero or sign
extension for free. In that context it is interesting to expose such pattern in
CodeGenPrepare so that the instruction selection pass can form such loads.
Sometimes, this pattern is blocked because of instructions between the load and
the extension. When those instructions are promotable to the extended type, we
can expose this pattern.
** Motivating Example **
Let us consider an example:
define void @foo(i8* %addr1, i32* %addr2, i8 %a, i32 %b) {
%ld = load i8* %addr1
%zextld = zext i8 %ld to i32
%ld2 = load i32* %addr2
%add = add nsw i32 %ld2, %zextld
%sextadd = sext i32 %add to i64
%zexta = zext i8 %a to i32
%addza = add nsw i32 %zexta, %zextld
%sextaddza = sext i32 %addza to i64
%addb = add nsw i32 %b, %zextld
%sextaddb = sext i32 %addb to i64
call void @dummy(i64 %sextadd, i64 %sextaddza, i64 %sextaddb)
ret void
}
As it is, this IR generates the following assembly on x86_64:
[...]
movzbl (%rdi), %eax # zero-extended load
movl (%rsi), %es # plain load
addl %eax, %esi # 32-bit add
movslq %esi, %rdi # sign extend the result of add
movzbl %dl, %edx # zero extend the first argument
addl %eax, %edx # 32-bit add
movslq %edx, %rsi # sign extend the result of add
addl %eax, %ecx # 32-bit add
movslq %ecx, %rdx # sign extend the result of add
[...]
The throughput of this sequence is 7.45 cycles on Ivy Bridge according to IACA.
Now, by promoting the additions to form more extended loads we would generate:
[...]
movzbl (%rdi), %eax # zero-extended load
movslq (%rsi), %rdi # sign-extended load
addq %rax, %rdi # 64-bit add
movzbl %dl, %esi # zero extend the first argument
addq %rax, %rsi # 64-bit add
movslq %ecx, %rdx # sign extend the second argument
addq %rax, %rdx # 64-bit add
[...]
The throughput of this sequence is 6.15 cycles on Ivy Bridge according to IACA.
This kind of sequences happen a lot on code using 32-bit indexes on 64-bit
architectures.
Note: The throughput numbers are similar on Sandy Bridge and Haswell.
** Proposed Solution **
To avoid the penalty of all these sign/zero extensions, we merge them in the
loads at the beginning of the chain of computation by promoting all the chain of
computation on the extended type. The promotion is done if and only if we do not
introduce new extensions, i.e., if we do not degrade the code quality.
To achieve this, we extend the existing “move ext to load” optimization with the
promotion mechanism introduced to match larger patterns for addressing mode
(r200947).
The idea of this extension is to perform the following transformation:
ext(promotableInst1(...(promotableInstN(load))))
=>
promotedInst1(...(promotedInstN(ext(load))))
The promotion mechanism in that optimization is enabled by a new TargetLowering
switch, which is off by default. In other words, by default, the optimization
performs the “move ext to load” optimization as it was before this patch.
** Performance **
Configuration: x86_64: Ivy Bridge fixed at 2900MHz running OS X 10.10.
Tested Optimization Levels: O3/Os
Tests: llvm-testsuite + externals.
Results:
- No regression beside noise.
- Improvements:
CINT2006/473.astar: ~2%
Benchmarks/PAQ8p: ~2%
Misc/perlin: ~3%
The results are consistent for both O3 and Os.
<rdar://problem/18310086>
llvm-svn: 224351
Split `Metadata` away from the `Value` class hierarchy, as part of
PR21532. Assembly and bitcode changes are in the wings, but this is the
bulk of the change for the IR C++ API.
I have a follow-up patch prepared for `clang`. If this breaks other
sub-projects, I apologize in advance :(. Help me compile it on Darwin
I'll try to fix it. FWIW, the errors should be easy to fix, so it may
be simpler to just fix it yourself.
This breaks the build for all metadata-related code that's out-of-tree.
Rest assured the transition is mechanical and the compiler should catch
almost all of the problems.
Here's a quick guide for updating your code:
- `Metadata` is the root of a class hierarchy with three main classes:
`MDNode`, `MDString`, and `ValueAsMetadata`. It is distinct from
the `Value` class hierarchy. It is typeless -- i.e., instances do
*not* have a `Type`.
- `MDNode`'s operands are all `Metadata *` (instead of `Value *`).
- `TrackingVH<MDNode>` and `WeakVH` referring to metadata can be
replaced with `TrackingMDNodeRef` and `TrackingMDRef`, respectively.
If you're referring solely to resolved `MDNode`s -- post graph
construction -- just use `MDNode*`.
- `MDNode` (and the rest of `Metadata`) have only limited support for
`replaceAllUsesWith()`.
As long as an `MDNode` is pointing at a forward declaration -- the
result of `MDNode::getTemporary()` -- it maintains a side map of its
uses and can RAUW itself. Once the forward declarations are fully
resolved RAUW support is dropped on the ground. This means that
uniquing collisions on changing operands cause nodes to become
"distinct". (This already happened fairly commonly, whenever an
operand went to null.)
If you're constructing complex (non self-reference) `MDNode` cycles,
you need to call `MDNode::resolveCycles()` on each node (or on a
top-level node that somehow references all of the nodes). Also,
don't do that. Metadata cycles (and the RAUW machinery needed to
construct them) are expensive.
- An `MDNode` can only refer to a `Constant` through a bridge called
`ConstantAsMetadata` (one of the subclasses of `ValueAsMetadata`).
As a side effect, accessing an operand of an `MDNode` that is known
to be, e.g., `ConstantInt`, takes three steps: first, cast from
`Metadata` to `ConstantAsMetadata`; second, extract the `Constant`;
third, cast down to `ConstantInt`.
The eventual goal is to introduce `MDInt`/`MDFloat`/etc. and have
metadata schema owners transition away from using `Constant`s when
the type isn't important (and they don't care about referring to
`GlobalValue`s).
In the meantime, I've added transitional API to the `mdconst`
namespace that matches semantics with the old code, in order to
avoid adding the error-prone three-step equivalent to every call
site. If your old code was:
MDNode *N = foo();
bar(isa <ConstantInt>(N->getOperand(0)));
baz(cast <ConstantInt>(N->getOperand(1)));
bak(cast_or_null <ConstantInt>(N->getOperand(2)));
bat(dyn_cast <ConstantInt>(N->getOperand(3)));
bay(dyn_cast_or_null<ConstantInt>(N->getOperand(4)));
you can trivially match its semantics with:
MDNode *N = foo();
bar(mdconst::hasa <ConstantInt>(N->getOperand(0)));
baz(mdconst::extract <ConstantInt>(N->getOperand(1)));
bak(mdconst::extract_or_null <ConstantInt>(N->getOperand(2)));
bat(mdconst::dyn_extract <ConstantInt>(N->getOperand(3)));
bay(mdconst::dyn_extract_or_null<ConstantInt>(N->getOperand(4)));
and when you transition your metadata schema to `MDInt`:
MDNode *N = foo();
bar(isa <MDInt>(N->getOperand(0)));
baz(cast <MDInt>(N->getOperand(1)));
bak(cast_or_null <MDInt>(N->getOperand(2)));
bat(dyn_cast <MDInt>(N->getOperand(3)));
bay(dyn_cast_or_null<MDInt>(N->getOperand(4)));
- A `CallInst` -- specifically, intrinsic instructions -- can refer to
metadata through a bridge called `MetadataAsValue`. This is a
subclass of `Value` where `getType()->isMetadataTy()`.
`MetadataAsValue` is the *only* class that can legally refer to a
`LocalAsMetadata`, which is a bridged form of non-`Constant` values
like `Argument` and `Instruction`. It can also refer to any other
`Metadata` subclass.
(I'll break all your testcases in a follow-up commit, when I propagate
this change to assembly.)
llvm-svn: 223802
Rewrite the pattern match code to work also with Values instead with
Instructions only. Also remove the no longer need matcher (m_Instruction).
llvm-svn: 223797
This optimization transforms code like:
bb1:
%0 = icmp ne i32 %a, 0
%1 = icmp ne i32 %b, 0
%or.cond = or i1 %0, %1
br i1 %or.cond, label %TrueBB, label %FalseBB
into a multiple branch instructions like:
bb1:
%0 = icmp ne i32 %a, 0
br i1 %0, label %TrueBB, label %bb2
bb2:
%1 = icmp ne i32 %b, 0
br i1 %1, label %TrueBB, label %FalseBB
This optimization is already performed by SelectionDAG, but not by FastISel.
FastISel cannot perform this optimization, because it cannot generate new
MachineBasicBlocks.
Performing this optimization at CodeGenPrepare time makes it available to both -
SelectionDAG and FastISel - and the implementation in SelectiuonDAG could be
removed. There are currenty a few differences in codegen for X86 and PPC, so
this commmit only enables it for FastISel.
Reviewed by Jim Grosbach
This fixes rdar://problem/19034919.
llvm-svn: 223786
This is to be consistent with StringSet and ultimately with the standard
library's associative container insert function.
This lead to updating SmallSet::insert to return pair<iterator, bool>,
and then to update SmallPtrSet::insert to return pair<iterator, bool>,
and then to update all the existing users of those functions...
llvm-svn: 222334
Prior to this patch the TypePromotionHelper was promoting only sign extensions.
Supporting zero extensions changes:
- How constants are extended.
- How sign extensions, zero extensions, and truncate are composed together.
- How the type of the extended operation is recorded. Now we need to know the
kind of the extension as well as its type.
Each change is fairly small, unlike the diff.
Most of the diff are comments/variable renaming to say "extension" instead of
"sign extension".
The performance improvements on the test suite are within the noise.
Related to <rdar://problem/18310086>.
llvm-svn: 221851
r221820 fixed a problem (PR21548) where an iPTR was used in TLI legality checks,
which isn't valid and resulted in a failed assertion.
The solution was to lower pointer types into the correct target's VT, by
using TL::getValueType instead of EVT::getEVT.
This commit changes 3 other uses of EVT::getEVT, but without any tests:
- One of these non-lowered EVTs is passed to allowsMisalignedMemoryAccesses,
which goes into target's TL implementation and doesn't cause any problem (yet.)
- Two others are passed to TLI.isOperationLegalOrCustom:
- one only looks at extensions, so doesn't concern pointers.
- one only looks at binary operators, so also isn't a problem.
The latter might some day be exposed to pointers and cause the same assert as
the original PR, because there's a comment hinting at also supporting cast ops.
For consistency, update all of them and be done with it.
llvm-svn: 221827
This patch adds an optimization in CodeGenPrepare to move an extractelement
right before a store when the target can combine them.
The optimization may promote any scalar operations to vector operations in the
way to make that possible.
** Context **
Some targets use different register files for both vector and scalar operations.
This means that transitioning from one domain to another may incur copy from one
register file to another. These copies are not coalescable and may be expensive.
For example, according to the scheduling model, on cortex-A8 a vector to GPR
move is 20 cycles.
** Motivating Example **
Let us consider an example:
define void @foo(<2 x i32>* %addr1, i32* %dest) {
%in1 = load <2 x i32>* %addr1, align 8
%extract = extractelement <2 x i32> %in1, i32 1
%out = or i32 %extract, 1
store i32 %out, i32* %dest, align 4
ret void
}
As it is, this IR generates the following assembly on armv7:
vldr d16, [r0] @vector load
vmov.32 r0, d16[1] @ cross-register-file copy: 20 cycles
orr r0, r0, #1 @ scalar bitwise or
str r0, [r1] @ scalar store
bx lr
Whereas we could generate much faster code:
vldr d16, [r0] @ vector load
vorr.i32 d16, #0x1 @ vector bitwise or
vst1.32 {d16[1]}, [r1:32] @ vector extract + store
bx lr
Half of the computation made in the vector is useless, but this allows to get
rid of the expensive cross-register-file copy.
** Proposed Solution **
To avoid this cross-register-copy penalty, we promote the scalar operations to
vector operations. The penalty will be removed if we manage to promote the whole
chain of computation in the vector domain.
Currently, we do that only when the chain of computation ends by a store and the
target is able to combine an extract with a store.
Stores are the most likely candidates, because other instructions produce values
that would need to be promoted and so, extracted as some point[1]. Moreover,
this is customary that targets feature stores that perform a vector extract (see
AArch64 and X86 for instance).
The proposed implementation relies on the TargetTransformInfo to decide whether
or not it is beneficial to promote a chain of computation in the vector domain.
Unfortunately, this interface is rather inaccurate for this level of details and
although this optimization may be beneficial for X86 and AArch64, the inaccuracy
will lead to the optimization being too aggressive.
Basically in TargetTransformInfo, everything that is legal has a cost of 1,
whereas, even if a vector type is legal, usually a vector operation is slightly
more expensive than its scalar counterpart. That will lead to too many
promotions that may not be counter balanced by the saving of the
cross-register-file copy. For instance, on AArch64 this penalty is just 4
cycles.
For now, the optimization is just enabled for ARM prior than v8, since those
processors have a larger penalty on cross-register-file copies, and the scope is
limited to basic blocks. Because of these two factors, we limit the effects of
the inaccuracy. Indeed, I did not want to build up a fancy cost model with block
frequency and everything on top of that.
[1] We can imagine targets that can combine an extractelement with other
instructions than just stores. If we want to go into that direction, the current
interfaces must be augmented and, moreover, I think this becomes a global isel
problem.
Differential Revision: http://reviews.llvm.org/D5921
<rdar://problem/14170854>
llvm-svn: 220978
introduced in r217629.
We were returning the old sext instead of the new zext as the promoted instruction!
Thanks Joerg Sonnenberger for the test case.
llvm-svn: 217800
