even if erroneously annotated with the parallel loop metadata.
Fixes Bug 15794:
"Loop Vectorizer: Crashes with the use of llvm.loop.parallel metadata"
llvm-svn: 180081
This reverts commit r179840 with a fix to test/DebugInfo/two-cus-from-same-file.ll
I'm not sure why that test only failed on ARM & MIPS and not X86 Linux, even
though the debug info was clearly invalid on all of them, but this ought to fix
it.
llvm-svn: 179996
This is an edge case that can happen if we modify a chain of multiple selects.
Update all operands in that case and remove the assert. PR15805.
llvm-svn: 179982
There is the temptation to make this tranform dependent on target information as
it is not going to be beneficial on all (sub)targets. Therefore, we should
probably do this in MI Early-Ifconversion.
This reverts commit r179957. Original commit message:
"SimplifyCFG: If convert single conditional stores
This transformation will transform a conditional store with a preceeding
uncondtional store to the same location:
a[i] =
may-alias with a[i] load
if (cond)
a[i] = Y
into an unconditional store.
a[i] = X
may-alias with a[i] load
tmp = cond ? Y : X;
a[i] = tmp
We assume that on average the cost of a mispredicted branch is going to be
higher than the cost of a second store to the same location, and that the
secondary benefits of creating a bigger basic block for other optimizations to
work on outway the potential case were the branch would be correctly predicted
and the cost of the executing the second store would be noticably reflected in
performance.
hmmer's execution time improves by 30% on an imac12,2 on ref data sets. With
this change we are on par with gcc's performance (gcc also performs this
transformation). There was a 1.2 % performance improvement on a ARM swift chip.
Other tests in the test-suite+external seem to be mostly uninfluenced in my
experiments:
This optimization was triggered on 41 tests such that the executable was
different before/after the patch. Only 1 out of the 40 tests (dealII) was
reproducable below 100% (by about .4%). Given that hmmer benefits so much I
believe this to be a fair trade off.
I am going to watch performance numbers across the builtbots and will revert
this if anything unexpected comes up."
llvm-svn: 179980
Specifically:
1. Added checks that unwind is being properly added to various instructions.
2. Fixed the declaration/calling of objc_release to have a return type of void.
3. Moved all checks to precede the functions and added checks to ensure that the
checks would only match inside the specific function that we are attempting to
check.
llvm-svn: 179973
This transformation will transform a conditional store with a preceeding
uncondtional store to the same location:
a[i] =
may-alias with a[i] load
if (cond)
a[i] = Y
into an unconditional store.
a[i] = X
may-alias with a[i] load
tmp = cond ? Y : X;
a[i] = tmp
We assume that on average the cost of a mispredicted branch is going to be
higher than the cost of a second store to the same location, and that the
secondary benefits of creating a bigger basic block for other optimizations to
work on outway the potential case were the branch would be correctly predicted
and the cost of the executing the second store would be noticably reflected in
performance.
hmmer's execution time improves by 30% on an imac12,2 on ref data sets. With
this change we are on par with gcc's performance (gcc also performs this
transformation). There was a 1.2 % performance improvement on a ARM swift chip.
Other tests in the test-suite+external seem to be mostly uninfluenced in my
experiments:
This optimization was triggered on 41 tests such that the executable was
different before/after the patch. Only 1 out of the 40 tests (dealII) was
reproducable below 100% (by about .4%). Given that hmmer benefits so much I
believe this to be a fair trade off.
I am going to watch performance numbers across the builtbots and will revert
this if anything unexpected comes up.
llvm-svn: 179957
The logic that actually compares the types considers pointers and integers the
same if they are of the same size. This created a strange mismatch between hash
and reality and made the test case for this fail on some platforms (yay,
test cases).
llvm-svn: 179905
Adding another CU-wide list, in this case of imported_modules (since they
should be relatively rare, it seemed better to add a list where each element
had a "context" value, rather than add a (usually empty) list to every scope).
This takes care of DW_TAG_imported_module, but to fully address PR14606 we'll
need to expand this to cover DW_TAG_imported_declaration too.
llvm-svn: 179836
A min/max operation is represented by a select(cmp(lt/le/gt/ge, X, Y), X, Y)
sequence in LLVM. If we see such a sequence we can treat it just as any other
commutative binary instruction and reduce it.
This appears to help bzip2 by about 1.5% on an imac12,2.
radar://12960601
llvm-svn: 179773
This occurs due to an alloca representing a separate ownership from the
original pointer. Thus consider the following pseudo-IR:
objc_retain(%a)
for (...) {
objc_retain(%a)
%block <- %a
F(%block)
objc_release(%block)
}
objc_release(%a)
From the perspective of the optimizer, the %block is a separate
provenance from the original %a. Thus the optimizer pairs up the inner
retain for %a and the outer release from %a, resulting in segfaults.
This is fixed by noting that the signature of a mismatch of
retain/releases inside the for loop is a Use/CanRelease top down with an
None bottom up (since bottom up the Retain-CanRelease-Use-Release
sequence is completed by the inner objc_retain, but top down due to the
differing provenance from the objc_release said sequence is not
completed). In said case in CheckForCFGHazards, we now clear the state
of %a implying that no pairing will occur.
Additionally a test case is included.
rdar://12969722
llvm-svn: 179747
If a switch instruction has a case for every possible value of its type,
with the same successor, SimplifyCFG would replace it with an icmp ult,
but the computation of the bound overflows in that case, which inverts
the test.
Patch by Jed Davis!
llvm-svn: 179587
Two return types are not equivalent if one is a pointer and the other is an
integral. This is because we cannot bitcast a pointer to an integral value.
PR15185
llvm-svn: 179569
One performs: (X == 13 | X == 14) -> X-13 <u 2
The other: (A == C1 || A == C2) -> (A & ~(C1 ^ C2)) == C1
The problem is that there are certain values of C1 and C2 that
trigger both transforms but the first one blocks out the second,
this generates suboptimal code.
Reordering the transforms should be better in every case and
allows us to do interesting stuff like turn:
%shr = lshr i32 %X, 4
%and = and i32 %shr, 15
%add = add i32 %and, -14
%tobool = icmp ne i32 %add, 0
into:
%and = and i32 %X, 240
%tobool = icmp ne i32 %and, 224
llvm-svn: 179493
This is basically the same fix in three different places. We use a set to avoid
walking the whole tree of a big ConstantExprs multiple times.
For example: (select cmp, (add big_expr 1), (add big_expr 2))
We don't want to visit big_expr twice here, it may consist of thousands of
nodes.
The testcase exercises this by creating an insanely large ConstantExprs out of
a loop. It's questionable if the optimizer should ever create those, but this
can be triggered with real C code. Fixes PR15714.
llvm-svn: 179458
When trying to collapse sequences of insertelement/extractelement
instructions into single shuffle instructions, there is one specific
case where the Instruction Combiner wrongly updates the resulting
Mask of shuffle indexes.
The problem is in function CollectShuffleElments.
If we have a sequence of insert/extract element instructions
like the one below:
%tmp1 = extractelement <4 x float> %LHS, i32 0
%tmp2 = insertelement <4 x float> %RHS, float %tmp1, i32 1
%tmp3 = extractelement <4 x float> %RHS, i32 2
%tmp4 = insertelement <4 x float> %tmp2, float %tmp3, i32 3
Where:
. %RHS will have a mask of [4,5,6,7]
. %LHS will have a mask of [0,1,2,3]
The Mask of shuffle indexes is wrongly computed to [4,1,6,7]
instead of [4,0,6,7].
When analyzing %tmp2 in order to compute the Mask for the
resulting shuffle instruction, the algorithm forgets to update
the mask index at position 1 with the index associated to the
element extracted from %LHS by instruction %tmp1.
Patch by Andrea DiBiagio!
llvm-svn: 179291
This commit adds the infrastructure for performing bottom-up SLP vectorization (and other optimizations) on parallel computations.
The infrastructure has three potential users:
1. The loop vectorizer needs to be able to vectorize AOS data structures such as (sum += A[i] + A[i+1]).
2. The BB-vectorizer needs this infrastructure for bottom-up SLP vectorization, because bottom-up vectorization is faster to compute.
3. A loop-roller needs to be able to analyze consecutive chains and roll them into a loop, in order to reduce code size. A loop roller does not need to create vector instructions, and this infrastructure separates the chain analysis from the vectorization.
This patch also includes a simple (100 LOC) bottom up SLP vectorizer that uses the infrastructure, and can vectorize this code:
void SAXPY(int *x, int *y, int a, int i) {
x[i] = a * x[i] + y[i];
x[i+1] = a * x[i+1] + y[i+1];
x[i+2] = a * x[i+2] + y[i+2];
x[i+3] = a * x[i+3] + y[i+3];
}
llvm-svn: 179117
The fix for PR14972 in r177055 introduced a real think-o in the *store*
side, likely because I was much more focused on the load side. While we
can arbitrarily widen (or narrow) a loaded value, we can't arbitrarily
widen a value to be stored, as that changes the width of memory access!
Lock down the code path in the store rewriting which would do this to
only handle the intended circumstance.
All of the existing tests continue to pass, and I've added a test from
the PR.
llvm-svn: 178974
The normal dataflow sequence in the ARC optimizer consists of the following
states:
Retain -> CanRelease -> Use -> Release
The optimizer before this patch stored the uses that determine the lifetime of
the retainable object pointer when it bottom up hits a retain or when top down
it hits a release. This is correct for an imprecise lifetime scenario since what
we are trying to do is remove retains/releases while making sure that no
``CanRelease'' (which is usually a call) deallocates the given pointer before we
get to the ``Use'' (since that would cause a segfault).
If we are considering the precise lifetime scenario though, this is not
correct. In such a situation, we *DO* care about the previous sequence, but
additionally, we wish to track the uses resulting from the following incomplete
sequences:
Retain -> CanRelease -> Release (TopDown)
Retain <- Use <- Release (BottomUp)
*NOTE* This patch looks large but the most of it consists of updating
test cases. Additionally this fix exposed an additional bug. I removed
the test case that expressed said bug and will recommit it with the fix
in a little bit.
llvm-svn: 178921
This optimization is unstable at this moment; it
1) block us on a very important application
2) PR15200
3) test6 and test7 in test/Transforms/ScalarRepl/dynamic-vector-gep.ll
(the CHECK command compare the output against wrong result)
I personally believe this optimization should not have any impact on the
autovectorized code, as auto-vectorizer is supposed to put gather/scatter
in a "right" way. Although in theory downstream optimizaters might reveal
some gather/scatter optimization opportunities, the chance is quite slim.
For the hand-crafted vectorizing code, in term of redundancy elimination,
load-CSE, copy-propagation and DSE can collectively achieve the same result,
but in much simpler way. On the other hand, these optimizers are able to
improve the code in a incremental way; in contrast, SROA is sort of all-or-none
approach. However, SROA might slighly win in stack size, as it tries to figure
out a stretch of memory tightenly cover the area accessed by the dynamic index.
rdar://13174884
PR15200
llvm-svn: 178912
Pass down the fact that an operand is going to be a vector of constants.
This should bring the performance of MultiSource/Benchmarks/PAQ8p/paq8p on x86
back. It had degraded to scalar performance due to my pervious shift cost change
that made all shifts expensive on x86.
radar://13576547
llvm-svn: 178809
The semantics of ARC implies that a pointer passed into an objc_autorelease
must live until some point (potentially down the stack) where an
autorelease pool is popped. On the other hand, an
objc_autoreleaseReturnValue just signifies that the object must live
until the end of the given function at least.
Thus objc_autorelease is stronger than objc_autoreleaseReturnValue in
terms of the semantics of ARC* implying that performing the given
strength reduction without any knowledge of how this relates to
the autorelease pool pop that is further up the stack violates the
semantics of ARC.
*Even though objc_autoreleaseReturnValue if you know that no RV
optimization will occur is more computationally expensive.
llvm-svn: 178612
The iterator could be invalidated when it's recursively deleting a whole bunch
of constant expressions in a constant initializer.
Note: This was only reproducible if `opt' was run on a `.bc' file. If `opt' was
run on a `.ll' file, it wouldn't crash. This is why the test first pushes the
`.ll' file through `llvm-as' before feeding it to `opt'.
PR15440
llvm-svn: 178531
Specifically, objc-arc-expand will make sure that the
objc_retainAutoreleasedReturnValue, objc_autoreleaseReturnValue, and ret
will all have %call as an argument.
llvm-svn: 178382
clang.arc.used is an interesting call for ARC since ObjCARCContract
needs to run to remove said intrinsic to avoid a linker error (since the
call does not exist).
llvm-svn: 178369
Since we handle optimizable objc_retainBlocks through strength reduction
in OptimizableIndividualCalls, we know that all code after that point
will only see non-optimizable objc_retainBlock calls. IsForwarding is
only called by functions after that point, so it is ok to just classify
objc_retainBlock as non-forwarding.
<rdar://problem/13249661>.
llvm-svn: 178285
If an objc_retainBlock has the copy_on_escape metadata attached to it
AND if the block pointer argument only escapes down the stack, we are
allowed to strength reduce the objc_retainBlock to to an objc_retain and
thus optimize it.
Current there is logic in the ARC data flow analysis to handle
this case which is complicated and involved making distinctions in
between objc_retainBlock and objc_retain in certain places and
considering them the same in others.
This patch simplifies said code by:
1. Performing the strength reduction in the initial ARC peephole
analysis (ObjCARCOpts::OptimizeIndividualCalls).
2. Changes the ARC dataflow analysis (which runs after the peephole
analysis) to consider all objc_retainBlock calls to not be optimizable
(since if the call was optimizable, we would have strength reduced it
already).
This patch leaves in the infrastructure in the ARC dataflow analysis to
handle this case, which due to 2 will just be dead code. I am doing this
on purpose to separate the removal of the old code from the testing of
the new code.
<rdar://problem/13249661>.
llvm-svn: 178284
This reverts commit 342d92c7a0adeabc9ab00f3f0d88d739fe7da4c7.
Turns out we're going with a different schema design to represent
DW_TAG_imported_modules so we won't need this extra field.
llvm-svn: 178215
The test was removed since I had not turned off the test during release
builds. This fails since ARC annotations support is conditionally
compiled out during release builds. I added the proper requires header
to assuage this issue.
llvm-svn: 178101
This is just the basic groundwork for supporting DW_TAG_imported_module but I
wanted to commit this before pushing support further into Clang or LLVM so that
this rather churny change is isolated from the rest of the work. The major
churn here is obviously adding another field (within the common DIScope prefix)
to all DIScopes (files, classes, namespaces, lexical scopes, etc). This should
be the last big churny change needed for DW_TAG_imported_module/using directive
support/PR14606.
llvm-svn: 178099
This will allow for verification and analysis of the merge function of
the data flow analyses in the ARC optimizer.
The actual implementation of this feature is by introducing calls to
the functions llvm.arc.annotation.{bottomup,topdown}.{bbstart,bbend}
which are only declared. Each such call takes in a pointer to a global
with the same name as the pointer whose provenance is being tracked and
a pointer whose name is one of our Sequence states and points to a
string that contains the same name.
To ensure that the optimizer does not consider these annotations in any
way, I made it so that the annotations are considered to be of IC_None
type.
A test case is included for this commit and the previous
ObjCARCAnnotation commit.
llvm-svn: 177952
The problem is that the code mistakenly took for granted that following constructor
is able to create an APFloat from a *SIGNED* integer:
APFloat::APFloat(const fltSemantics &ourSemantics, integerPart value)
rdar://13486998
llvm-svn: 177906
This simplification happens at 2 places :
- using the nsw attribute when the shl / mul is used by a sign test
- when the shl / mul is compared for (in)equality to zero
llvm-svn: 177856
The original code used i32, and i64 if legal. This introduced unneeded
casts when they aren't legal, or when the index variable i has another
type. In order of preference: try to use i's type; use the smallest
fitting legal type (using an added DataLayout method); default to i32.
A testcase checks that this works when the index gep operand is i16.
Patch by : Ahmed Bougacha <ahmed.bougacha@gmail.com>
Reviewed by : Duncan
llvm-svn: 177712
The simplify-libcalls pass implemented a doInitialization hook to infer
function prototype attributes for well-known functions. Given that the
simplify-libcalls pass is going away *and* that the functionattrs pass
is already in place to deduce function attributes, I am moving this logic
to the functionattrs pass. This approach was discussed during patch
review:
http://lists.cs.uiuc.edu/pipermail/llvm-commits/Week-of-Mon-20121126/157465.html.
llvm-svn: 177619
- it is trivially known to be used inside the loop in a way that can not be optimized away
- there is no use outside of the loop which can take advantage of the computation hoisting
llvm-svn: 177432
This handles the case where we have an inbounds GEP with alloca as the pointer.
This fixes the regression in PR12750 and rdar://13286434.
Note that we can also fix this by handling some GEP cases in isKnownNonNull.
llvm-svn: 177321
*NOTE* I verified that the original bug behind
dont-infinite-loop-during-block-escape-analysis.ll occurs when using opt on
retain-block-escape-analysis.ll.
llvm-svn: 177240
This is the first step to making all DIScopes have a common metadata prefix (so
that things (using directives, for example) that can appear in any scope can be
added to that common prefix). DIFile is itself a DIScope so the common prefix
of all DIScopes cannot be a DIFile - instead it's the raw filename/directory
name pair.
llvm-svn: 177239
This test makes sure that the ObjCARC escape analysis looks at the uses of
instructions which copy the block pointer value by checking all four cases where
that can occur.
llvm-svn: 177232
Rules include:
1)1 x*y +/- x*z => x*(y +/- z)
(the order of operands dosen't matter)
2) y/x +/- z/x => (y +/- z)/x
The transformation is disabled if the new add/sub expr "y +/- z" is a
denormal/naz/inifinity.
rdar://12911472
llvm-svn: 177088
The fundamental problem is that SROA didn't allow for overly wide loads
where the bits past the end of the alloca were masked away and the load
was sufficiently aligned to ensure there is no risk of page fault, or
other trapping behavior. With such widened loads, SROA would delete the
load entirely rather than clamping it to the size of the alloca in order
to allow mem2reg to fire. This was exposed by a test case that neatly
arranged for GVN to run first, widening certain loads, followed by an
inline step, and then SROA which miscompiles the code. However, I see no
reason why this hasn't been plaguing us in other contexts. It seems
deeply broken.
Diagnosing all of the above took all of 10 minutes of debugging. The
really annoying aspect is that fixing this completely breaks the pass.
;] There was an implicit reliance on the fact that no loads or stores
extended past the alloca once we decided to rewrite them in the final
stage of SROA. This was used to encode information about whether the
loads and stores had been split across multiple partitions of the
original alloca. That required threading explicit tracking of whether
a *use* of a partition is split across multiple partitions.
Once that was done, another problem arose: we allowed splitting of
integer loads and stores iff they were loads and stores to the entire
alloca. This is a really arbitrary limitation, and splitting at least
some integer loads and stores is crucial to maximize promotion
opportunities. My first attempt was to start removing the restriction
entirely, but currently that does Very Bad Things by causing *many*
common alloca patterns to be fully decomposed into i8 operations and
lots of or-ing together to produce larger integers on demand. The code
bloat is terrifying. That is still the right end-goal, but substantial
work must be done to either merge partitions or ensure that small i8
values are eagerly merged in some other pass. Sadly, figuring all this
out took essentially all the time and effort here.
So the end result is that we allow splitting only when the load or store
at least covers the alloca. That ensures widened loads and stores don't
hurt SROA, and that we don't rampantly decompose operations more than we
have previously.
All of this was already fairly well tested, and so I've just updated the
tests to cover the wide load behavior. I can add a test that crafts the
pass ordering magic which caused the original PR, but that seems really
brittle and to provide little benefit. The fundamental problem is that
widened loads should Just Work.
llvm-svn: 177055
constructs default arguments. It can now take default arguments from
cl::opt'ions. Add a new -default-gcov-version=... option, and actually test it!
Sink the reverse-order of the version into GCOVProfiling, hiding it from our
users.
llvm-svn: 177002
This is the next step towards making the metadata for DIScopes have a common
prefix rather than having to delegate based on their tag type.
llvm-svn: 176913
This could be 'null' or the empty string, DIDescriptor::getStringField
coalesces the two cases anyway so it's just a matter of legible/efficient
representation.
The change in behavior of the DICompileUnit::get* functions could be
subsumed by the full verification check - but ideally that should just be an
assertion if we could front-load the actual debug info metadata failure paths.
llvm-svn: 176907
These cases were found by further work to remove support for debug info
versioning. Common cleanups (other than changing the version info in the tag
field) included adding the last parameter to compile_units (recently added for
fission support) and other cases of trailing fields in lexical blocks, compile
units, and subprograms.
llvm-svn: 176834
We want vectorization to happen at -g. Ignore calls to the dbg.value intrinsic
and don't transfer them to the vectorized code.
radar://13378964
llvm-svn: 176768
Summary:
Statistics are still available in Release+Asserts (any +Asserts builds),
and stats can also be turned on with LLVM_ENABLE_STATS.
Move some of the FastISel stats that were moved under DEBUG()
back out of DEBUG(), since stats are disabled across the board now.
Many tests depend on grepping "-stats" output. Move those into
a orig_dir/Stats/. so that they can be marked as unsupported
when building without statistics.
Differential Revision: http://llvm-reviews.chandlerc.com/D486
llvm-svn: 176733
Fixes rdar:13349374.
Volatile loads and stores need to be preserved even if the language
standard says they are undefined. "volatile" in this context means "get
out of the way compiler, let my platform handle it".
Additionally, this is the only way I know of with llvm to write to the
first page (when hardware allows) without dropping to assembly.
llvm-svn: 176599
When considering folding a bitcast of an alloca into the alloca itself,
make sure we don't shrink the amount of memory being allocated, or
things rapidly go sideways.
rdar://13324424
llvm-svn: 176547
This adds minimalistic support for PHI nodes to llvm.objectsize() evaluation
fingers crossed so that it does break clang boostrap again..
llvm-svn: 176408
This matters for example in following matrix multiply:
int **mmult(int rows, int cols, int **m1, int **m2, int **m3) {
int i, j, k, val;
for (i=0; i<rows; i++) {
for (j=0; j<cols; j++) {
val = 0;
for (k=0; k<cols; k++) {
val += m1[i][k] * m2[k][j];
}
m3[i][j] = val;
}
}
return(m3);
}
Taken from the test-suite benchmark Shootout.
We estimate the cost of the multiply to be 2 while we generate 9 instructions
for it and end up being quite a bit slower than the scalar version (48% on my
machine).
Also, properly differentiate between avx1 and avx2. On avx-1 we still split the
vector into 2 128bits and handle the subvector muls like above with 9
instructions.
Only on avx-2 will we have a cost of 9 for v4i64.
I changed the test case in test/Transforms/LoopVectorize/X86/avx1.ll to use an
add instead of a mul because with a mul we now no longer vectorize. I did
verify that the mul would be indeed more expensive when vectorized with 3
kernels:
for (i ...)
r += a[i] * 3;
for (i ...)
m1[i] = m1[i] * 3; // This matches the test case in avx1.ll
and a matrix multiply.
In each case the vectorized version was considerably slower.
radar://13304919
llvm-svn: 176403
The LoopVectorizer often runs multiple times on the same function due to inlining.
When this happens the loop vectorizer often vectorizes the same loops multiple times, increasing code size and adding unneeded branches.
With this patch, the vectorizer during vectorization puts metadata on scalar loops and marks them as 'already vectorized' so that it knows to ignore them when it sees them a second time.
PR14448.
llvm-svn: 176399
The instcombine recognized pattern looks like:
a = b * c
d = a +/- Cst
or
a = b * c
d = Cst +/- a
When creating the new operands for fadd or fsub instruction following the related fmul, the first operand was created with the second original operand (M0 was created with C1) and the second with the first (M1 with Opnd0).
The fix consists in creating the new operands with the appropriate original operand, i.e., M0 with Opnd0 and M1 with C1.
llvm-svn: 176300
This properly asks TargetLibraryInfo if a call is available and if it is, it
can be translated into the corresponding LLVM builtin. We don't vectorize sqrt()
yet because I'm not sure about the semantics for negative numbers. The other
intrinsic should be exact equivalents to the libm functions.
Differential Revision: http://llvm-reviews.chandlerc.com/D465
llvm-svn: 176188
This is a common pattern with dyn_cast and similar constructs, when the
PHI no longer depends on the select it can often be turned into a simpler
construct or even get hoisted out of the loop.
PR15340.
llvm-svn: 175995
Listing all of the attributes for the callee of a call/invoke instruction is way
too much and makes the IR unreadable. Use references to attributes instead.
llvm-svn: 175877
Storing the load/store instructions with the values
and inspect them using Alias Analysis to make sure
they don't alias, since the GEP pointer operand doesn't
take the offset into account.
Trying hard to not add any extra cost to loads and stores
that don't overlap on global values, AA is *only* calculated
if all of the previous attempts failed.
Using biggest vector register size as the stride for the
vectorization access, as we're being conservative and
the cost model (which calculates the real vectorization
factor) is only run after the legalization phase.
We might re-think this relationship in the future, but
for now, I'd rather be safe than sorry.
llvm-svn: 175818
Profiling tests *do* need a JIT. They'll pass if a cross-compiler targetting
AArch64 by default has been built, but fail if a native AArch64 compiler has
been build. Therefore XFAIL is inappropriate and we mark them unsupported.
ExecutionEngine tests are JIT by definition, they should also be unsupported.
Transforms/LICM only uses the interpreter to check the output is still sane
after optimisation. It can be switched to use an interpreter.
llvm-svn: 175433
This fixes PR15289. This bug was introduced (recently) in r175215; collecting
all std::vector references for candidate pairs to delete at once is invalid
because subsequent lookups in the owning DenseMap could invalidate the
references.
bugpoint was able to reduce a useful test case. Unfortunately, because whether
or not this asserts depends on memory layout, this test case will sometimes
appear to produce valid output. Nevertheless, running under valgrind will
reveal the error.
llvm-svn: 175397
(or (bool?A:B),(bool?C:D)) --> (bool?(or A,C):(or B,D))
By the time the OR is visited, both the SELECTs have been visited and not
optimized and the OR itself hasn't been transformed so we do this transform in
the hopes that the new ORs will be optimized.
The transform is explicitly disabled for vector-selects until "codegen matures
to handle them better".
Patch by Muhammad Tauqir!
llvm-svn: 175380
It enables to work with a smaller constant, which is target friendly for those which can compare to immediates.
It also avoids inserting a shift in favor of a trunc, which can be free on some targets.
This used to work until LLVM-3.1, but regressed with the 3.2 release.
llvm-svn: 175270
isn't using the default calling convention. However, if the transformation is
from a call to inline IR, then the calling convention doesn't matter.
rdar://13157990
llvm-svn: 174724
In the loop vectorizer cost model, we used to ignore stores/loads of a pointer
type when computing the widest type within a loop. This meant that if we had
only stores/loads of pointers in a loop we would return a widest type of 8bits
(instead of 32 or 64 bit) and therefore a vector factor that was too big.
Now, if we see a consecutive store/load of pointers we use the size of a pointer
(from data layout).
This problem occured in SingleSource/Benchmarks/Shootout-C++/hash.cpp (reduced
test case is the first test in vector_ptr_load_store.ll).
radar://13139343
llvm-svn: 174377
The main lists of debug info metadata attached to the compile_unit had an extra
layer of metadata nodes they went through for no apparent reason. This patch
removes that (& still passes just as much of the GDB 7.5 test suite). If anyone
can show evidence as to why these extra metadata nodes are there I'm open to
reverting this patch & documenting why they're there.
llvm-svn: 174266
Prepare it for vectors of pointers and handle simple cases. We don't handle
complicated cases because accumulateConstantOffset bails on pointer vectors.
Fixes selfhost on i386.
llvm-svn: 174179
remaining use of AliasAnalysis concepts such as isIdentifiedObject to
prove pointer inequality.
@external_compare in test/Transforms/InstSimplify/compare.ll shows a simple
case where a noalias argument can be equal to a global variable address, and
while AliasAnalysis can get away with saying that these pointers don't alias,
instsimplify cannot say that they are not equal.
llvm-svn: 174122
The AttrBuilder is for building a collection of attributes. The Attribute object
holds only one attribute. So it's not really useful for the Attribute object to
have a creator which takes an AttrBuilder.
This has two fallouts:
1. The AttrBuilder no longer holds its internal attributes in a bit-mask form.
2. The attributes are now ordered alphabetically (hence why the tests have changed).
llvm-svn: 174110
Changing ARMBaseTargetMachine to return ARMTargetLowering intead of
the generic one (similar to x86 code).
Tests showing which instructions were added to cast when necessary
or cost zero when not. Downcast to 16 bits are not lowered in NEON,
so costs are not there yet.
llvm-svn: 173849
The AttributeSetNode contains all of the attributes. This removes one (hopefully
last) use of the Attribute class as a container of multiple attributes.
llvm-svn: 173761
These tests in particular try to use escaped square brackets as an
argument to grep, which is failing for me with native win32 python. It
appears the backslash is being lost near the CreateProcess*() call.
llvm-svn: 173506
loops over instructions in the basic block or the use-def list of the
value, neither of which are really efficient when repeatedly querying
about values in the same basic block.
What's more, we already know that the CondBB is small, and so we can do
a much more efficient test by counting the uses in CondBB, and seeing if
those account for all of the uses.
Finally, we shouldn't blanket fail on any such instruction, instead we
should conservatively assume that those instructions are part of the
cost.
Note that this actually fixes a bug in the pass because
isUsedInBasicBlock has a really terrible bug in it. I'll fix that in my
next commit, but the fix for it would make this code suddenly take the
compile time hit I thought it already was taking, so I wanted to go
ahead and migrate this code to a faster & better pattern.
The bug in isUsedInBasicBlock was also causing other tests to test the
wrong thing entirely: for example we weren't actually disabling
speculation for floating point operations as intended (and tested), but
the test passed because we failed to speculate them due to the
isUsedInBasicBlock failure.
llvm-svn: 173417
Original commit message:
Plug TTI into the speculation logic, giving it a real cost interface
that can be specialized by targets.
The goal here is not to be more aggressive, but to just be more accurate
with very obvious cases. There are instructions which are known to be
truly free and which were not being modeled as such in this code -- see
the regression test which is distilled from an inner loop of zlib.
Everywhere the TTI cost model is insufficiently conservative I've added
explicit checks with FIXME comments to go add proper modelling of these
cost factors.
If this causes regressions, the likely solution is to make TTI even more
conservative in its cost estimates, but test cases will help here.
llvm-svn: 173357
We use constant folding to see if an intrinsic evaluates to the same value as a
constant that we know. If we don't take the undefinedness into account we get a
value that doesn't match the actual implementation, and miscompiled code.
This was uncovered by Chandler's simplifycfg changes.
llvm-svn: 173356
that can be specialized by targets.
The goal here is not to be more aggressive, but to just be more accurate
with very obvious cases. There are instructions which are known to be
truly free and which were not being modeled as such in this code -- see
the regression test which is distilled from an inner loop of zlib.
Everywhere the TTI cost model is insufficiently conservative I've added
explicit checks with FIXME comments to go add proper modelling of these
cost factors.
If this causes regressions, the likely solution is to make TTI even more
conservative in its cost estimates, but test cases will help here.
llvm-svn: 173342
a cost fuction that seems both a bit ad-hoc and also poorly suited to
evaluating constant expressions.
Notably, it is missing any support for trivial expressions such as
'inttoptr'. I could fix this routine, but it isn't clear to me all of
the constraints its other users are operating under.
The core protection that seems relevant here is avoiding the formation
of a select instruction wich a further chain of select operations in
a constant expression operand. Just explicitly encode that constraint.
Also, update the comments and organization here to make it clear where
this needs to go -- this should be driven off of real cost measurements
which take into account the number of constants expressions and the
depth of the constant expression tree.
llvm-svn: 173340
SSPStrong applies a heuristic to insert stack protectors in these situations:
* A Protector is required for functions which contain an array, regardless of
type or length.
* A Protector is required for functions which contain a structure/union which
contains an array, regardless of type or length. Note, there is no limit to
the depth of nesting.
* A protector is required when the address of a local variable (i.e., stack
based variable) is exposed. (E.g., such as through a local whose address is
taken as part of the RHS of an assignment or a local whose address is taken as
part of a function argument.)
This patch implements the SSPString attribute to be equivalent to
SSPRequired. This will change in a subsequent patch.
llvm-svn: 173230
We ignore the cpu frontend and focus on pipeline utilization. We do this because we
don't have a good way to estimate the loop body size at the IR level.
llvm-svn: 172964
This separates the check for "too few elements to run the vector loop" from the
"memory overlap" check, giving a lot nicer code and allowing to skip the memory
checks when we're not going to execute the vector code anyways. We still leave
the decision of whether to emit the memory checks as branches or setccs, but it
seems to be doing a good job. If ugly code pops up we may want to emit them as
separate blocks too. Small speedup on MultiSource/Benchmarks/MallocBench/espresso.
Most of this is legwork to allow multiple bypass blocks while updating PHIs,
dominators and loop info.
llvm-svn: 172902
Okay, here's how to reproduce the problem:
1) Build a Release (or Release+Asserts) version of clang in the normal way.
2) Using the clang & clang++ binaries from (1), build a Release (or
Release+Asserts) version of the same sources, but this time enable LTO ---
specify the `-flto' flag on the command line.
3) Run the ARC migrator tests:
$ arcmt-test --args -triple x86_64-apple-darwin10 -fsyntax-only -x objective-c++ ./src/tools/clang/test/ARCMT/cxx-rewrite.mm
You'll see that the output isn't correct (the whitespace is off).
The mis-compile is in the function `RewriteBuffer::RemoveText' in the
clang/lib/Rewrite/Core/Rewriter.cpp file. When that function and RewriteRope.cpp
are compiled with LTO and the `arcmt-test' executable is regenerated, you'll see
the error. When those files are not LTO'ed, then the output of the `arcmt-test'
is fine.
It is *really* hard to get a testcase out of this. I'll file a PR with what I
have currently.
--- Reverse-merging r172363 into '.':
U include/llvm/Analysis/MemoryBuiltins.h
U lib/Analysis/MemoryBuiltins.cpp
--- Reverse-merging r171325 into '.':
U test/Transforms/InstCombine/objsize.ll
G include/llvm/Analysis/MemoryBuiltins.h
G lib/Analysis/MemoryBuiltins.cpp
llvm-svn: 172756
some optimization opportunities (in the enclosing supper-expressions).
rule 1. (-0.0 - X ) * Y => -0.0 - (X * Y)
if expression "-0.0 - X" has only one reference.
rule 2. (0.0 - X ) * Y => -0.0 - (X * Y)
if expression "0.0 - X" has only one reference, and
the instruction is marked "noSignedZero".
2. Eliminate negation (The compiler was already able to handle these
opt if the 0.0s are replaced with -0.0.)
rule 3: (0.0 - X) * (0.0 - Y) => X * Y
rule 4: (0.0 - X) * C => X * -C
if the expr is flagged "noSignedZero".
3.
Rule 5: (X*Y) * X => (X*X) * Y
if X!=Y and the expression is flagged with "UnsafeAlgebra".
The purpose of this transformation is two-fold:
a) to form a power expression (of X).
b) potentially shorten the critical path: After transformation, the
latency of the instruction Y is amortized by the expression of X*X,
and therefore Y is in a "less critical" position compared to what it
was before the transformation.
4. Remove the InstCombine code about simplifiying "X * select".
The reasons are following:
a) The "select" is somewhat architecture-dependent, therefore the
higher level optimizers are not able to precisely predict if
the simplification really yields any performance improvement
or not.
b) The "select" operator is bit complicate, and tends to obscure
optimization opportunities. It is btter to keep it as low as
possible in expr tree, and let CodeGen to tackle the optimization.
llvm-svn: 172551
Test was failing for clang-native-arm-cortex-a9 build-bot configuration.
The reason for the failure was the test was using hardcoded names.
The attached patch fixes this failure by replacing the hard-coded variables
names with pattern-matched variable names.
Patch by Manish Verma, ARM
llvm-svn: 172534
---------------------------------------------------------------------------
C_A: reassociation is allowed
C_R: reciprocal of a constant C is appropriate, which means
- 1/C is exact, or
- reciprocal is allowed and 1/C is neither a special value nor a denormal.
-----------------------------------------------------------------------------
rule1: (X/C1) / C2 => X / (C2*C1) (if C_A)
=> X * (1/(C2*C1)) (if C_A && C_R)
rule 2: X*C1 / C2 => X * (C1/C2) if C_A
rule 3: (X/Y)/Z = > X/(Y*Z) (if C_A && at least one of Y and Z is symbolic value)
rule 4: Z/(X/Y) = > (Z*Y)/X (similar to rule3)
rule 5: C1/(X*C2) => (C1/C2) / X (if C_A)
rule 6: C1/(X/C2) => (C1*C2) / X (if C_A)
rule 7: C1/(C2/X) => (C1/C2) * X (if C_A)
llvm-svn: 172488
The reason that this occurs is that tail calling objc_autorelease eventually
tail calls -[NSObject autorelease] which supports fast autorelease. This can
cause us to violate the semantic gaurantees of __autoreleasing variables that
assignment to an __autoreleasing variables always yields an object that is
placed into the innermost autorelease pool.
The fix included in this patch works by:
1. In the peephole optimization function OptimizeIndividualFunctions, always
remove tail call from objc_autorelease.
2. Whenever we convert to/from an objc_autorelease, set/unset the tail call
keyword as appropriate.
*NOTE* I also handled the case where objc_autorelease is converted in
OptimizeReturns to an autoreleaseRV which still violates the ARC semantics. I
will be removing that in a later patch and I wanted to make sure that the tree
is in a consistent state vis-a-vis ARC always.
Additionally some test cases are provided and all tests that have tail call marked
objc_autorelease keywords have been modified so that tail call has been removed.
*NOTE* One test fails due to a separate bug that I am going to commit soon. Thus
I marked the check line TMP: instead of CHECK: so make check does not fail.
llvm-svn: 172287
the target if it supports the different CAST types. We didn't do this
on X86 because of the different register sizes and types, but on ARM
this makes sense.
llvm-svn: 172245
We don't have a detailed analysis on which values are vectorized and which stay scalars in the vectorized loop so we use
another method. We look at reduction variables, loads and stores, which are the only ways to get information in and out
of loop iterations. If the data types are extended and truncated then the cost model will catch the cost of the vector
zext/sext/trunc operations.
llvm-svn: 172178
Messages:
Converted test case trivial_codegen_tailcall.ll to use FileCheck.
Converted test return_constant.ll to use FileCheck instead of grep.
Converted test reorder_load.ll to use FileCheck instead of grep.
Converted test intervening-inst.ll to use FileCheck instead of grep.
llvm-svn: 172171
small loops. On small loops post-loop that handles scalars (and runs slower) can take more time to execute than the
rest of the loop. This patch disables widening of loops with a small static trip count.
llvm-svn: 171798
o. X/C1 * C2 => X * (C2/C1) (if C2/C1 is neither special FP nor denormal)
o. X/C1 * C2 -> X/(C1/C2) (if C2/C1 is either specical FP or denormal, but C1/C2 is a normal Fp)
Let MDC denote multiplication or dividion with one & only one operand being a constant
o. (MDC ± C1) * C2 => (MDC * C2) ± (C1 * C2)
(so long as the constant-folding doesn't yield any denormal or special value)
llvm-svn: 171793
Pekka Jaaskelainen