This reverts commit r371584. It introduced a dependency from compiler-rt
to llvm/include/ADT, which is problematic for multiple reasons.
One is that it is a novel dependency edge, which needs cross-compliation
machinery for llvm/include/ADT (yes, it is true that right now
compiler-rt included only header-only libraries, however, if we allow
compiler-rt to depend on anything from ADT, other libraries will
eventually get used).
Secondly, depending on ADT from compiler-rt exposes ADT symbols from
compiler-rt, which would cause ODR violations when Clang is built with
the profile library.
llvm-svn: 371598
Currently we only rely on the induction increment to come before the
condition to ensure the required instructions get moved to the new
latch.
This patch duplicates and moves the required instructions to the
newly created latch. We move the condition to the end of the new block,
then process its operands. We stop at operands that are defined
outside the loop, or are the induction PHI.
We duplicate the instructions and update the uses in the moved
instructions, to ensure other users remain intact. See the added
test2 for such an example.
Reviewers: efriedma, mcrosier
Reviewed By: efriedma
Differential Revision: https://reviews.llvm.org/D67367
llvm-svn: 371595
Configure TLI to say that r600/amdgpu does not have any library
functions, such that InstCombine does not do anything like turn sin/cos
into the library function @tan with sufficient fast math flags.
Differential Revision: https://reviews.llvm.org/D67406
Change-Id: I02f907d3e64832117ea9800e9f9285282856e5df
llvm-svn: 371592
TryToSinkInstruction() has a bug: While updating debug info for
sunk instruction, it could clone dbg.declare intrinsic.
That is wrong. There could be only one dbg.declare.
The fix is to not clone dbg.declare intrinsic and to update
it`s arguments, to not to point to sunk instruction.
Differential Revision: https://reviews.llvm.org/D67217
llvm-svn: 371587
This patch contains the basic functionality for reporting potentially
incorrect usage of __builtin_expect() by comparing the developer's
annotation against a collected PGO profile. A more detailed proposal and
discussion appears on the CFE-dev mailing list
(http://lists.llvm.org/pipermail/cfe-dev/2019-July/062971.html) and a
prototype of the initial frontend changes appear here in D65300
We revised the work in D65300 by moving the misexpect check into the
LLVM backend, and adding support for IR and sampling based profiles, in
addition to frontend instrumentation.
We add new misexpect metadata tags to those instructions directly
influenced by the llvm.expect intrinsic (branch, switch, and select)
when lowering the intrinsics. The misexpect metadata contains
information about the expected target of the intrinsic so that we can
check against the correct PGO counter when emitting diagnostics, and the
compiler's values for the LikelyBranchWeight and UnlikelyBranchWeight.
We use these branch weight values to determine when to emit the
diagnostic to the user.
A future patch should address the comment at the top of
LowerExpectIntrisic.cpp to hoist the LikelyBranchWeight and
UnlikelyBranchWeight values into a shared space that can be accessed
outside of the LowerExpectIntrinsic pass. Once that is done, the
misexpect metadata can be updated to be smaller.
In the long term, it is possible to reconstruct portions of the
misexpect metadata from the existing profile data. However, we have
avoided this to keep the code simple, and because some kind of metadata
tag will be required to identify which branch/switch/select instructions
are influenced by the use of llvm.expect
Patch By: paulkirth
Differential Revision: https://reviews.llvm.org/D66324
llvm-svn: 371584
I only want to ensure that %offset is non-zero there,
it doesn't matter how that info is conveyed.
As filed in PR43267, the assumption way does not work.
llvm-svn: 371550
I only want to ensure that %offset is non-zero there,
it doesn't matter how that info is conveyed.
As filed in PR43267, the assumption way does not work.
llvm-svn: 371546
This allows us to fold fma's that multiply with 0.0. Also, the
multiply by 1.0 case is handled there as well. The fneg/fabs cases
are not handled by SimplifyFMulInst, so we need to keep them.
Reviewers: spatel, anemet, lebedev.ri
Reviewed By: spatel
Differential Revision: https://reviews.llvm.org/D67351
llvm-svn: 371518
To prevent AArch64 tests from running when the target is not compiled.
Fixes r371502:
/home/buildslave/ps4-buildslave4/llvm-clang-lld-x86_64-scei-ps4-ubuntu-fast/llvm.src/test/Transforms/ExpandMemCmp/AArch64/memcmp.ll:11:15: error: CHECK-NEXT: expected string not found in input
; CHECK-NEXT: [[TMP0:%.*]] = bitcast i8* [[S1:%.*]] to i64*
llvm-svn: 371503
This patch contains the basic functionality for reporting potentially
incorrect usage of __builtin_expect() by comparing the developer's
annotation against a collected PGO profile. A more detailed proposal and
discussion appears on the CFE-dev mailing list
(http://lists.llvm.org/pipermail/cfe-dev/2019-July/062971.html) and a
prototype of the initial frontend changes appear here in D65300
We revised the work in D65300 by moving the misexpect check into the
LLVM backend, and adding support for IR and sampling based profiles, in
addition to frontend instrumentation.
We add new misexpect metadata tags to those instructions directly
influenced by the llvm.expect intrinsic (branch, switch, and select)
when lowering the intrinsics. The misexpect metadata contains
information about the expected target of the intrinsic so that we can
check against the correct PGO counter when emitting diagnostics, and the
compiler's values for the LikelyBranchWeight and UnlikelyBranchWeight.
We use these branch weight values to determine when to emit the
diagnostic to the user.
A future patch should address the comment at the top of
LowerExpectIntrisic.cpp to hoist the LikelyBranchWeight and
UnlikelyBranchWeight values into a shared space that can be accessed
outside of the LowerExpectIntrinsic pass. Once that is done, the
misexpect metadata can be updated to be smaller.
In the long term, it is possible to reconstruct portions of the
misexpect metadata from the existing profile data. However, we have
avoided this to keep the code simple, and because some kind of metadata
tag will be required to identify which branch/switch/select instructions
are influenced by the use of llvm.expect
Patch By: paulkirth
Differential Revision: https://reviews.llvm.org/D66324
llvm-svn: 371484
If we're vectorizing a load in a predicated block, check to see if the load can be speculated rather than predicated. This allows us to generate a normal vector load instead of a masked.load.
To do so, we must prove that all bytes accessed on any iteration of the original loop are dereferenceable, and that all loads (across all iterations) are properly aligned. This is equivelent to proving that hoisting the load into the loop header in the original scalar loop is safe.
Note: There are a couple of code motion todos in the code. My intention is to wait about a day - to be sure this sticks - and then perform the NFC motion without furthe review.
Differential Revision: https://reviews.llvm.org/D66688
llvm-svn: 371452
Summary:
This tests inlining size thresholds, but relies on the output of running
the full O2 pipeline, making it brittle against changes in unrelated
passes.
Only run the inlining pass and set thresholds on the test RUN line
instead.
Found while investigating D60318.
Reviewers: RKSimon, qcolombet
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67349
llvm-svn: 371397
Summary:
This is motivated by D67122 sanitizer check enhancement.
That patch seemingly worsens `-fsanitize=pointer-overflow`
overhead from 25% to 50%, which strongly implies missing folds.
In this particular case, given
```
char* test(char& base, unsigned long offset) {
return &base + offset;
}
```
it will end up producing something like
https://godbolt.org/z/LK5-iH
which after optimizations reduces down to roughly
```
define i1 @t0(i8* nonnull %base, i64 %offset) {
%base_int = ptrtoint i8* %base to i64
%adjusted = add i64 %base_int, %offset
%non_null_after_adjustment = icmp ne i64 %adjusted, 0
%no_overflow_during_adjustment = icmp uge i64 %adjusted, %base_int
%res = and i1 %non_null_after_adjustment, %no_overflow_during_adjustment
ret i1 %res
}
```
Without D67122 there was no `%non_null_after_adjustment`,
and in this particular case we can get rid of the overhead:
Here we add some offset to a non-null pointer,
and check that the result does not overflow and is not a null pointer.
But since the base pointer is already non-null, and we check for overflow,
that overflow check will already catch the null pointer,
so the separate null check is redundant and can be dropped.
Alive proofs:
https://rise4fun.com/Alive/WRzq
There are more patterns of "unsigned-add-with-overflow", they are not handled here,
but this is the main pattern, that we currently consider canonical,
so it makes sense to handle it.
https://bugs.llvm.org/show_bug.cgi?id=43246
Reviewers: spatel, nikic, vsk
Reviewed By: spatel
Subscribers: hiraditya, llvm-commits, reames
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67332
llvm-svn: 371349
This is similar to the existing fold for splats added with:
rL365379
If we can adjust the shuffle mask to include another element
in an identity mask (if it changes vector length, that's an
extract/insert subvector operation in the backend), then that
can eliminate extractelement/insertelement pairs in IR.
All targets are expected to lower shuffles with identity masks
efficiently.
llvm-svn: 371340
Summary:
Similar to the previous prefer-256-bit flag. We might want to
enable this by default some CPUs. This just starts the initial
work to implement and prove that it effects TTI's vector width.
Reviewers: RKSimon, echristo, spatel, atdt
Reviewed By: RKSimon
Subscribers: lebedev.ri, hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67311
llvm-svn: 371319
Summary:
Add an intrinsic that takes 2 unsigned integers with
the scale of them provided as the third argument and
performs fixed point multiplication on them. The
result is saturated and clamped between the largest and
smallest representable values of the first 2 operands.
This is a part of implementing fixed point arithmetic
in clang where some of the more complex operations
will be implemented as intrinsics.
Patch by: leonardchan, bjope
Reviewers: RKSimon, craig.topper, bevinh, leonardchan, lebedev.ri, spatel
Reviewed By: leonardchan
Subscribers: ychen, wuzish, nemanjai, MaskRay, jsji, jdoerfert, Ka-Ka, hiraditya, rjmccall, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D57836
llvm-svn: 371308
This addresses the issue mentioned on D19867. When we simplify
with.overflow instructions in CVP, we leave behind extractvalue
of insertvalue sequences that LVI no longer understands. This
means that we can not simplify any instructions based on the
with.overflow anymore (until some over pass like InstCombine
cleans them up).
This patch extends LVI extractvalue handling by calling
SimplifyExtractValueInst (which doesn't do anything more than
constant folding + looking through insertvalue) and using the block
value of the simplification.
A possible alternative would be to do something similar to
SimplifyIndVars, where we instead directly try to replace
extractvalue users of the with.overflow. This would need some
additional structural changes to CVP, as it's currently not legal
to remove anything but the current instruction -- we'd have to
introduce a worklist with instructions scheduled for deletion or similar.
Differential Revision: https://reviews.llvm.org/D67035
llvm-svn: 371306
Summary:
This patch introduces initial `AAValueSimplify` which simplifies a value in a context.
example
- (for function returned) If all the return values are the same and constant, then we can replace callsite returned with the constant.
- If an internal function takes the same value(constant) as an argument in the callsite, then we can replace the argument with that constant.
Reviewers: jdoerfert, sstefan1
Reviewed By: jdoerfert
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66967
llvm-svn: 371291
Summary:
This isn't an important optimization at all... We're already doing:
pow(x, 0.0) -> 1.0
My patch merely teaches instcombine that -0.0 does the same.
However, doing this fixes an AMAZING bug! Compile this program:
extern "C" double pow(double, double);
double boom(double base) {
return pow(base, -0.0);
}
With:
clang++ ~/Desktop/fast-math.cpp -ffast-math -O2 -S
And clang will crash with a signal. Wow, fast math is so fast it ICEs the
compiler! Arguably, the generated math is infinitely fast.
What's actually happening is that we recurse infinitely in getPow. In debug we
hit its assertion:
assert(Exp != 0 && "Incorrect exponent 0 not handled");
We avoid this entire mess if we instead recognize that an exponent of positive
and negative zero yield 1.0.
A separate commit, r371221, fixed the same problem. This only contains the added
tests.
<rdar://problem/54598300>
Reviewers: scanon
Subscribers: hiraditya, jkorous, dexonsmith, ributzka, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67248
llvm-svn: 371224
This patch sinks add/mul(shufflevector(insertelement())) into the basic block in which they are used so that they can then be selected together.
This is useful for various MVE instructions, such as vmla and others that take R registers.
Loop tests have been added to the vmla test file to make sure vmlas are generated in loops.
Differential revision: https://reviews.llvm.org/D66295
llvm-svn: 371218
Properly check if NewAAInfo conflicts with AAInfo.
Update local variable and alias set that a change occured when a conflict is found.
Resolves PR42969.
llvm-svn: 371139
Summary:
Here we try to avoid issues with "explicit branch" with SimplifyBranchOnICmpChain
which can check on undef. Msan by design reports branches on uninitialized
memory and undefs, so we have false report here.
In general msan does not like when we convert
```
// If at least one of them is true we can MSAN is ok if another is undefs
if (a || b)
return;
```
into
```
// If 'a' is undef MSAN will complain even if 'b' is true
if (a)
return;
if (b)
return;
```
Example
Before optimization we had something like this:
```
while (true) {
bool maybe_undef = doStuff();
while (true) {
char c = getChar();
if (c != 10 && c != 13)
continue
break;
}
// we know that c == 10 || c == 13 if we get here,
// so msan know that branch is not affected by maybe_undef
if (maybe_undef || c == 10 || c == 13)
continue;
return;
}
```
SimplifyBranchOnICmpChain will convert that into
```
while (true) {
bool maybe_undef = doStuff();
while (true) {
char c = getChar();
if (c != 10 && c != 13)
continue;
break;
}
// however msan will complain here:
if (maybe_undef)
continue;
// we know that c == 10 || c == 13, so either way we will get continue
switch(c) {
case 10: continue;
case 13: continue;
}
return;
}
```
Reviewers: eugenis, efriedma
Reviewed By: eugenis, efriedma
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67205
llvm-svn: 371138
A follow-up for r329011.
This may be changed to produce @llvm.sub.with.overflow in a later patch,
but for now just make things more consistent overall.
A few observations stem from this:
* There does not seem to be a similar one-instruction fold for uadd-overflow
* I'm not sure we'll want to canonicalize `B u> A` as `usub.with.overflow`,
so since the `icmp` here no longer refers to `sub`,
reconstructing `usub.with.overflow` will be problematic,
and will likely require standalone pass (similar to DivRemPairs).
https://rise4fun.com/Alive/Zqs
Name: (A - B) u> A --> B u> A
%t0 = sub i8 %A, %B
%r = icmp ugt i8 %t0, %A
=>
%r = icmp ugt i8 %B, %A
Name: (A - B) u<= A --> B u<= A
%t0 = sub i8 %A, %B
%r = icmp ule i8 %t0, %A
=>
%r = icmp ule i8 %B, %A
Name: C u< (C - D) --> C u< D
%t0 = sub i8 %C, %D
%r = icmp ult i8 %C, %t0
=>
%r = icmp ult i8 %C, %D
Name: C u>= (C - D) --> C u>= D
%t0 = sub i8 %C, %D
%r = icmp uge i8 %C, %t0
=>
%r = icmp uge i8 %C, %D
llvm-svn: 371101
If we have:
bb5:
br i1 %arg3, label %bb6, label %bb7
bb6:
%tmp = getelementptr inbounds i32, i32* %arg1, i64 2
store i32 3, i32* %tmp, align 4
br label %bb9
bb7:
%tmp8 = getelementptr inbounds i32, i32* %arg1, i64 2
store i32 3, i32* %tmp8, align 4
br label %bb9
bb9: ; preds = %bb4, %bb6, %bb7
...
We can't sink stores directly into bb9.
This patch creates new BB that is successor of %bb6 and %bb7
and sinks stores into that block.
SplitFooterBB is the parameter to the pass that controls
that behavior.
Change-Id: I7fdf50a772b84633e4b1b860e905bf7e3e29940f
Differential: https://reviews.llvm.org/D66234
llvm-svn: 371089
Summary:
Avoid visiting an instruction more than once by using a map.
This is similar to https://reviews.llvm.org/rL361416.
Reviewers: davidxl
Reviewed By: davidxl
Subscribers: llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67198
llvm-svn: 371086
Summary:
Instead of building attributes for internal functions which we do not
update as long as we assume they are dead, we now do not create
attributes until we assume the internal function to be live. This
improves the number of required iterations, as well as the number of
required updates, in real code. On our tests, the results are mixed.
Reviewers: sstefan1, uenoku
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66914
llvm-svn: 370924
Summary:
We create attributes on-demand so we need to check the white list
on-demand. This also unifies the location at which we create,
initialize, and eventually invalidate new abstract attributes.
The tests show mixed results, a few more call site attributes are
determined which can cause more iterations.
Reviewers: uenoku, sstefan1
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66913
llvm-svn: 370922
Summary:
Before we tried to rule out non-exact definitions early but that lead to
on-demand attributes created for them anyway. As a consequence we needed
to look at the definition in the initialize of each attribute again.
This patch centralized this lookup and tightens the condition under
which we give up on non-exact definitions.
Reviewers: uenoku, sstefan1
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67115
llvm-svn: 370917
SROA pass processes debug info incorrecly if applied twice.
Specifically, after SROA works first time, instcombine converts dbg.declare
intrinsics into dbg.value. Inlining creates new opportunities for SROA,
so it is called again. This time it does not handle correctly previously
inserted dbg.value intrinsics.
Differential Revision: https://reviews.llvm.org/D64595
llvm-svn: 370906
Add the no-capture argument attribute deduction to the Attributor
fixpoint framework.
The new string attributed "no-capture-maybe-returned" is introduced to
allow deduction of no-capture through functions that "capture" an
argument but only by "returning" it. It is only used by the Attributor
for testing.
Differential Revision: https://reviews.llvm.org/D59922
llvm-svn: 370817
This extends the existing logic for propagating constant expressions in an analogous manner for what we do across basic blocks. The core point is that we chose some order of operands, and canonicalize uses towards that one.
The heuristic used is inspired by the one used across blocks; in a follow up change, I'd plan to common them so that the cross block version uses the slightly stronger ordering herein.
As noted by the TODOs in the code, there's a good amount of room for improving the existing code and making it more powerful. Some follow up work planned.
Differential Revision: https://reviews.llvm.org/D66977
llvm-svn: 370791
Summary:
Fold-tail currently supports reduction last-vector-value live-out's,
but has yet to support last-scalar-value live-outs, including
non-header phi's. As it relies on AllowedExit in order to detect
them and bail out we need to add the non-header PHI nodes to
AllowedExit, otherwise we end up with miscompiles.
Solves https://bugs.llvm.org/show_bug.cgi?id=43166
Reviewers: fhahn, Ayal
Reviewed By: fhahn, Ayal
Subscribers: anna, hiraditya, rkruppe, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D67074
llvm-svn: 370721
bitcast <N x i8> (shuf X, undef, <N, N-1,...0>) to i{N*8} --> bswap (bitcast X to i{N*8})
In PR43146:
https://bugs.llvm.org/show_bug.cgi?id=43146
...we have a more complicated case where SLP is making a mess of bswap. This patch won't
do anything for that currently, but we need to improve bswap recognition in instcombine,
SLP, and/or a standalone pass to avoid that problem.
This is limited using the data-layout so we don't try to do this transform with actual
vector types. The backend does not appear to have folds to convert in either direction,
so we don't want to mess up something that is actually better lowered as a shuffle.
On x86, we're trading something like this:
vmovd %edi, %xmm0
vpshufb LCPI0_0(%rip), %xmm0, %xmm0 ## xmm0 = xmm0[3,2,1,0,u,u,u,u,u,u,u,u,u,u,u,u]
vmovd %xmm0, %eax
For:
movl %edi, %eax
bswapl %eax
Differential Revision: https://reviews.llvm.org/D66965
llvm-svn: 370659
Summary:
Back-end currently expands mempcpy, but middle-end should work with memcpy instead of mempcpy to enable more memcpy-optimization.
GCC backend emits mempcpy, so LLVM backend could form it too, if we know mempcpy libcall is better than memcpy + n.
https://godbolt.org/z/dOCG96
Reviewers: efriedma, spatel, craig.topper, RKSimon, jdoerfert
Reviewed By: efriedma
Subscribers: hjl.tools, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D65737
llvm-svn: 370593
Use a { iN undef, i1 false } struct as the base, and only insert
the first operand, instead of using { iN undef, i1 undef } as the
base and inserting both. This is the same as what we do in InstCombine.
Differential Revision: https://reviews.llvm.org/D67034
llvm-svn: 370573
cold versus function being newly added.
This is the second half of https://reviews.llvm.org/D66374.
Profile symbol list is the collection of function symbols showing up in
the binary which generates the current profile. It is used to discriminate
function being cold versus function being newly added. Profile symbol list
is only added for profile with ExtBinary format.
During profile use compilation, when profile-sample-accurate is enabled,
a function without profile will be regarded as cold only when it is
contained in that list.
Differential Revision: https://reviews.llvm.org/D66766
llvm-svn: 370563
This is an updated version of https://reviews.llvm.org/D66909 to fix PR42605.
Basically, current phi translatation translates an old value number to an new
value number for a call instruction based on the literal equality of call
expression, without verifying there is no clobber in between. This is incorrect.
To get a finegrain check, use MachineDependence analysis to do the job. However,
this is still not ideal. Although given a call instruction,
`MemoryDependenceResults::getCallDependencyFrom` returns identical call
instructions without clobber in between using MemDepResult with its DepType to
be `Def`. However, identical is too strict here and we want it to be relaxed a
little to consider phi-translation -- callee is the same, param operands can be
different. That means changing the semantic of `MemDepResult::Def` and I don't
know the potential impact.
So currently the patch is still conservative to only handle
MemDepResult::NonFuncLocal, which means the current call has no function local
clobber. If there is clobber, even if the clobber doesn't stand in between the
current call and the call with the new value, we won't do phi-translate.
Differential Revision: https://reviews.llvm.org/D67013
llvm-svn: 370547
Summary:
Instead of recomputing information for call sites we now use the
function information directly. This is always valid and once we have
call site specific information we can improve here.
This patch also bootstraps attributes that are created on-demand through
an initial update call. Information that is known will then directly be
available in the new attribute without causing an iteration delay.
The tests show how this improves the iteration count.
Reviewers: sstefan1, uenoku
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66781
llvm-svn: 370480
Summary:
Any pointer could have load/store users not only floating ones so we
move the manifest logic for alignment into the AAAlignImpl class.
Reviewers: uenoku, sstefan1
Subscribers: hiraditya, bollu, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66922
llvm-svn: 370479
Summary:
@mclow.lists brought up this issue up in IRC.
It is a reasonably common problem to compare some two values for equality.
Those may be just some integers, strings or arrays of integers.
In C, there is `memcmp()`, `bcmp()` functions.
In C++, there exists `std::equal()` algorithm.
One can also write that function manually.
libstdc++'s `std::equal()` is specialized to directly call `memcmp()` for
various types, but not `std::byte` from C++2a. https://godbolt.org/z/mx2ejJ
libc++ does not do anything like that, it simply relies on simple C++'s
`operator==()`. https://godbolt.org/z/er0Zwf (GOOD!)
So likely, there exists a certain performance opportunities.
Let's compare performance of naive `std::equal()` (no `memcmp()`) with one that
is using `memcmp()` (in this case, compiled with modified compiler). {F8768213}
```
#include <algorithm>
#include <cmath>
#include <cstdint>
#include <iterator>
#include <limits>
#include <random>
#include <type_traits>
#include <utility>
#include <vector>
#include "benchmark/benchmark.h"
template <class T>
bool equal(T* a, T* a_end, T* b) noexcept {
for (; a != a_end; ++a, ++b) {
if (*a != *b) return false;
}
return true;
}
template <typename T>
std::vector<T> getVectorOfRandomNumbers(size_t count) {
std::random_device rd;
std::mt19937 gen(rd());
std::uniform_int_distribution<T> dis(std::numeric_limits<T>::min(),
std::numeric_limits<T>::max());
std::vector<T> v;
v.reserve(count);
std::generate_n(std::back_inserter(v), count,
[&dis, &gen]() { return dis(gen); });
assert(v.size() == count);
return v;
}
struct Identical {
template <typename T>
static std::pair<std::vector<T>, std::vector<T>> Gen(size_t count) {
auto Tmp = getVectorOfRandomNumbers<T>(count);
return std::make_pair(Tmp, std::move(Tmp));
}
};
struct InequalHalfway {
template <typename T>
static std::pair<std::vector<T>, std::vector<T>> Gen(size_t count) {
auto V0 = getVectorOfRandomNumbers<T>(count);
auto V1 = V0;
V1[V1.size() / size_t(2)]++; // just change the value.
return std::make_pair(std::move(V0), std::move(V1));
}
};
template <class T, class Gen>
void BM_bcmp(benchmark::State& state) {
const size_t Length = state.range(0);
const std::pair<std::vector<T>, std::vector<T>> Data =
Gen::template Gen<T>(Length);
const std::vector<T>& a = Data.first;
const std::vector<T>& b = Data.second;
assert(a.size() == Length && b.size() == a.size());
benchmark::ClobberMemory();
benchmark::DoNotOptimize(a);
benchmark::DoNotOptimize(a.data());
benchmark::DoNotOptimize(b);
benchmark::DoNotOptimize(b.data());
for (auto _ : state) {
const bool is_equal = equal(a.data(), a.data() + a.size(), b.data());
benchmark::DoNotOptimize(is_equal);
}
state.SetComplexityN(Length);
state.counters["eltcnt"] =
benchmark::Counter(Length, benchmark::Counter::kIsIterationInvariant);
state.counters["eltcnt/sec"] =
benchmark::Counter(Length, benchmark::Counter::kIsIterationInvariantRate);
const size_t BytesRead = 2 * sizeof(T) * Length;
state.counters["bytes_read/iteration"] =
benchmark::Counter(BytesRead, benchmark::Counter::kDefaults,
benchmark::Counter::OneK::kIs1024);
state.counters["bytes_read/sec"] = benchmark::Counter(
BytesRead, benchmark::Counter::kIsIterationInvariantRate,
benchmark::Counter::OneK::kIs1024);
}
template <typename T>
static void CustomArguments(benchmark::internal::Benchmark* b) {
const size_t L2SizeBytes = []() {
for (const benchmark::CPUInfo::CacheInfo& I :
benchmark::CPUInfo::Get().caches) {
if (I.level == 2) return I.size;
}
return 0;
}();
// What is the largest range we can check to always fit within given L2 cache?
const size_t MaxLen = L2SizeBytes / /*total bufs*/ 2 /
/*maximal elt size*/ sizeof(T) / /*safety margin*/ 2;
b->RangeMultiplier(2)->Range(1, MaxLen)->Complexity(benchmark::oN);
}
BENCHMARK_TEMPLATE(BM_bcmp, uint8_t, Identical)
->Apply(CustomArguments<uint8_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint16_t, Identical)
->Apply(CustomArguments<uint16_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint32_t, Identical)
->Apply(CustomArguments<uint32_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint64_t, Identical)
->Apply(CustomArguments<uint64_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint8_t, InequalHalfway)
->Apply(CustomArguments<uint8_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint16_t, InequalHalfway)
->Apply(CustomArguments<uint16_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint32_t, InequalHalfway)
->Apply(CustomArguments<uint32_t>);
BENCHMARK_TEMPLATE(BM_bcmp, uint64_t, InequalHalfway)
->Apply(CustomArguments<uint64_t>);
```
{F8768210}
```
$ ~/src/googlebenchmark/tools/compare.py --no-utest benchmarks build-{old,new}/test/llvm-bcmp-bench
RUNNING: build-old/test/llvm-bcmp-bench --benchmark_out=/tmp/tmpb6PEUx
2019-04-25 21:17:11
Running build-old/test/llvm-bcmp-bench
Run on (8 X 4000 MHz CPU s)
CPU Caches:
L1 Data 16K (x8)
L1 Instruction 64K (x4)
L2 Unified 2048K (x4)
L3 Unified 8192K (x1)
Load Average: 0.65, 3.90, 4.14
---------------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations UserCounters...
---------------------------------------------------------------------------------------------------
<...>
BM_bcmp<uint8_t, Identical>/512000 432131 ns 432101 ns 1613 bytes_read/iteration=1000k bytes_read/sec=2.20706G/s eltcnt=825.856M eltcnt/sec=1.18491G/s
BM_bcmp<uint8_t, Identical>_BigO 0.86 N 0.86 N
BM_bcmp<uint8_t, Identical>_RMS 8 % 8 %
<...>
BM_bcmp<uint16_t, Identical>/256000 161408 ns 161409 ns 4027 bytes_read/iteration=1000k bytes_read/sec=5.90843G/s eltcnt=1030.91M eltcnt/sec=1.58603G/s
BM_bcmp<uint16_t, Identical>_BigO 0.67 N 0.67 N
BM_bcmp<uint16_t, Identical>_RMS 25 % 25 %
<...>
BM_bcmp<uint32_t, Identical>/128000 81497 ns 81488 ns 8415 bytes_read/iteration=1000k bytes_read/sec=11.7032G/s eltcnt=1077.12M eltcnt/sec=1.57078G/s
BM_bcmp<uint32_t, Identical>_BigO 0.71 N 0.71 N
BM_bcmp<uint32_t, Identical>_RMS 42 % 42 %
<...>
BM_bcmp<uint64_t, Identical>/64000 50138 ns 50138 ns 10909 bytes_read/iteration=1000k bytes_read/sec=19.0209G/s eltcnt=698.176M eltcnt/sec=1.27647G/s
BM_bcmp<uint64_t, Identical>_BigO 0.84 N 0.84 N
BM_bcmp<uint64_t, Identical>_RMS 27 % 27 %
<...>
BM_bcmp<uint8_t, InequalHalfway>/512000 192405 ns 192392 ns 3638 bytes_read/iteration=1000k bytes_read/sec=4.95694G/s eltcnt=1.86266G eltcnt/sec=2.66124G/s
BM_bcmp<uint8_t, InequalHalfway>_BigO 0.38 N 0.38 N
BM_bcmp<uint8_t, InequalHalfway>_RMS 3 % 3 %
<...>
BM_bcmp<uint16_t, InequalHalfway>/256000 127858 ns 127860 ns 5477 bytes_read/iteration=1000k bytes_read/sec=7.45873G/s eltcnt=1.40211G eltcnt/sec=2.00219G/s
BM_bcmp<uint16_t, InequalHalfway>_BigO 0.50 N 0.50 N
BM_bcmp<uint16_t, InequalHalfway>_RMS 0 % 0 %
<...>
BM_bcmp<uint32_t, InequalHalfway>/128000 49140 ns 49140 ns 14281 bytes_read/iteration=1000k bytes_read/sec=19.4072G/s eltcnt=1.82797G eltcnt/sec=2.60478G/s
BM_bcmp<uint32_t, InequalHalfway>_BigO 0.40 N 0.40 N
BM_bcmp<uint32_t, InequalHalfway>_RMS 18 % 18 %
<...>
BM_bcmp<uint64_t, InequalHalfway>/64000 32101 ns 32099 ns 21786 bytes_read/iteration=1000k bytes_read/sec=29.7101G/s eltcnt=1.3943G eltcnt/sec=1.99381G/s
BM_bcmp<uint64_t, InequalHalfway>_BigO 0.50 N 0.50 N
BM_bcmp<uint64_t, InequalHalfway>_RMS 1 % 1 %
RUNNING: build-new/test/llvm-bcmp-bench --benchmark_out=/tmp/tmpQ46PP0
2019-04-25 21:19:29
Running build-new/test/llvm-bcmp-bench
Run on (8 X 4000 MHz CPU s)
CPU Caches:
L1 Data 16K (x8)
L1 Instruction 64K (x4)
L2 Unified 2048K (x4)
L3 Unified 8192K (x1)
Load Average: 1.01, 2.85, 3.71
---------------------------------------------------------------------------------------------------
Benchmark Time CPU Iterations UserCounters...
---------------------------------------------------------------------------------------------------
<...>
BM_bcmp<uint8_t, Identical>/512000 18593 ns 18590 ns 37565 bytes_read/iteration=1000k bytes_read/sec=51.2991G/s eltcnt=19.2333G eltcnt/sec=27.541G/s
BM_bcmp<uint8_t, Identical>_BigO 0.04 N 0.04 N
BM_bcmp<uint8_t, Identical>_RMS 37 % 37 %
<...>
BM_bcmp<uint16_t, Identical>/256000 18950 ns 18948 ns 37223 bytes_read/iteration=1000k bytes_read/sec=50.3324G/s eltcnt=9.52909G eltcnt/sec=13.511G/s
BM_bcmp<uint16_t, Identical>_BigO 0.08 N 0.08 N
BM_bcmp<uint16_t, Identical>_RMS 34 % 34 %
<...>
BM_bcmp<uint32_t, Identical>/128000 18627 ns 18627 ns 37895 bytes_read/iteration=1000k bytes_read/sec=51.198G/s eltcnt=4.85056G eltcnt/sec=6.87168G/s
BM_bcmp<uint32_t, Identical>_BigO 0.16 N 0.16 N
BM_bcmp<uint32_t, Identical>_RMS 35 % 35 %
<...>
BM_bcmp<uint64_t, Identical>/64000 18855 ns 18855 ns 37458 bytes_read/iteration=1000k bytes_read/sec=50.5791G/s eltcnt=2.39731G eltcnt/sec=3.3943G/s
BM_bcmp<uint64_t, Identical>_BigO 0.32 N 0.32 N
BM_bcmp<uint64_t, Identical>_RMS 33 % 33 %
<...>
BM_bcmp<uint8_t, InequalHalfway>/512000 9570 ns 9569 ns 73500 bytes_read/iteration=1000k bytes_read/sec=99.6601G/s eltcnt=37.632G eltcnt/sec=53.5046G/s
BM_bcmp<uint8_t, InequalHalfway>_BigO 0.02 N 0.02 N
BM_bcmp<uint8_t, InequalHalfway>_RMS 29 % 29 %
<...>
BM_bcmp<uint16_t, InequalHalfway>/256000 9547 ns 9547 ns 74343 bytes_read/iteration=1000k bytes_read/sec=99.8971G/s eltcnt=19.0318G eltcnt/sec=26.8159G/s
BM_bcmp<uint16_t, InequalHalfway>_BigO 0.04 N 0.04 N
BM_bcmp<uint16_t, InequalHalfway>_RMS 29 % 29 %
<...>
BM_bcmp<uint32_t, InequalHalfway>/128000 9396 ns 9394 ns 73521 bytes_read/iteration=1000k bytes_read/sec=101.518G/s eltcnt=9.41069G eltcnt/sec=13.6255G/s
BM_bcmp<uint32_t, InequalHalfway>_BigO 0.08 N 0.08 N
BM_bcmp<uint32_t, InequalHalfway>_RMS 30 % 30 %
<...>
BM_bcmp<uint64_t, InequalHalfway>/64000 9499 ns 9498 ns 73802 bytes_read/iteration=1000k bytes_read/sec=100.405G/s eltcnt=4.72333G eltcnt/sec=6.73808G/s
BM_bcmp<uint64_t, InequalHalfway>_BigO 0.16 N 0.16 N
BM_bcmp<uint64_t, InequalHalfway>_RMS 28 % 28 %
Comparing build-old/test/llvm-bcmp-bench to build-new/test/llvm-bcmp-bench
Benchmark Time CPU Time Old Time New CPU Old CPU New
---------------------------------------------------------------------------------------------------------------------------------------
<...>
BM_bcmp<uint8_t, Identical>/512000 -0.9570 -0.9570 432131 18593 432101 18590
<...>
BM_bcmp<uint16_t, Identical>/256000 -0.8826 -0.8826 161408 18950 161409 18948
<...>
BM_bcmp<uint32_t, Identical>/128000 -0.7714 -0.7714 81497 18627 81488 18627
<...>
BM_bcmp<uint64_t, Identical>/64000 -0.6239 -0.6239 50138 18855 50138 18855
<...>
BM_bcmp<uint8_t, InequalHalfway>/512000 -0.9503 -0.9503 192405 9570 192392 9569
<...>
BM_bcmp<uint16_t, InequalHalfway>/256000 -0.9253 -0.9253 127858 9547 127860 9547
<...>
BM_bcmp<uint32_t, InequalHalfway>/128000 -0.8088 -0.8088 49140 9396 49140 9394
<...>
BM_bcmp<uint64_t, InequalHalfway>/64000 -0.7041 -0.7041 32101 9499 32099 9498
```
What can we tell from the benchmark?
* Performance of naive equality check somewhat improves with element size,
maxing out at eltcnt/sec=1.58603G/s for uint16_t, or bytes_read/sec=19.0209G/s
for uint64_t. I think, that instability implies performance problems.
* Performance of `memcmp()`-aware benchmark always maxes out at around
bytes_read/sec=51.2991G/s for every type. That is 2.6x the throughput of the
naive variant!
* eltcnt/sec metric for the `memcmp()`-aware benchmark maxes out at
eltcnt/sec=27.541G/s for uint8_t (was: eltcnt/sec=1.18491G/s, so 24x) and
linearly decreases with element size.
For uint64_t, it's ~4x+ the elements/second.
* The call obvious is more pricey than the loop, with small element count.
As it can be seen from the full output {F8768210}, the `memcmp()` is almost
universally worse, independent of the element size (and thus buffer size) when
element count is less than 8.
So all in all, bcmp idiom does indeed pose untapped performance headroom.
This diff does implement said idiom recognition. I think a reasonable test
coverage is present, but do tell if there is anything obvious missing.
Now, quality. This does succeed to build and pass the test-suite, at least
without any non-bundled elements. {F8768216} {F8768217}
This transform fires 91 times:
```
$ /build/test-suite/utils/compare.py -m loop-idiom.NumBCmp result-new.json
Tests: 1149
Metric: loop-idiom.NumBCmp
Program result-new
MultiSourc...Benchmarks/7zip/7zip-benchmark 79.00
MultiSource/Applications/d/make_dparser 3.00
SingleSource/UnitTests/vla 2.00
MultiSource/Applications/Burg/burg 1.00
MultiSourc.../Applications/JM/lencod/lencod 1.00
MultiSource/Applications/lemon/lemon 1.00
MultiSource/Benchmarks/Bullet/bullet 1.00
MultiSourc...e/Benchmarks/MallocBench/gs/gs 1.00
MultiSourc...gs-C/TimberWolfMC/timberwolfmc 1.00
MultiSourc...Prolangs-C/simulator/simulator 1.00
```
The size changes are:
I'm not sure what's going on with SingleSource/UnitTests/vla.test yet, did not look.
```
$ /build/test-suite/utils/compare.py -m size..text result-{old,new}.json --filter-hash
Tests: 1149
Same hash: 907 (filtered out)
Remaining: 242
Metric: size..text
Program result-old result-new diff
test-suite...ingleSource/UnitTests/vla.test 753.00 833.00 10.6%
test-suite...marks/7zip/7zip-benchmark.test 1001697.00 966657.00 -3.5%
test-suite...ngs-C/simulator/simulator.test 32369.00 32321.00 -0.1%
test-suite...plications/d/make_dparser.test 89585.00 89505.00 -0.1%
test-suite...ce/Applications/Burg/burg.test 40817.00 40785.00 -0.1%
test-suite.../Applications/lemon/lemon.test 47281.00 47249.00 -0.1%
test-suite...TimberWolfMC/timberwolfmc.test 250065.00 250113.00 0.0%
test-suite...chmarks/MallocBench/gs/gs.test 149889.00 149873.00 -0.0%
test-suite...ications/JM/lencod/lencod.test 769585.00 769569.00 -0.0%
test-suite.../Benchmarks/Bullet/bullet.test 770049.00 770049.00 0.0%
test-suite...HMARK_ANISTROPIC_DIFFUSION/128 NaN NaN nan%
test-suite...HMARK_ANISTROPIC_DIFFUSION/256 NaN NaN nan%
test-suite...CHMARK_ANISTROPIC_DIFFUSION/64 NaN NaN nan%
test-suite...CHMARK_ANISTROPIC_DIFFUSION/32 NaN NaN nan%
test-suite...ENCHMARK_BILATERAL_FILTER/64/4 NaN NaN nan%
Geomean difference nan%
result-old result-new diff
count 1.000000e+01 10.00000 10.000000
mean 3.152090e+05 311695.40000 0.006749
std 3.790398e+05 372091.42232 0.036605
min 7.530000e+02 833.00000 -0.034981
25% 4.243300e+04 42401.00000 -0.000866
50% 1.197370e+05 119689.00000 -0.000392
75% 6.397050e+05 639705.00000 -0.000005
max 1.001697e+06 966657.00000 0.106242
```
I don't have timings though.
And now to the code. The basic idea is to completely replace the whole loop.
If we can't fully kill it, don't transform.
I have left one or two comments in the code, so hopefully it can be understood.
Also, there is a few TODO's that i have left for follow-ups:
* widening of `memcmp()`/`bcmp()`
* step smaller than the comparison size
* Metadata propagation
* more than two blocks as long as there is still a single backedge?
* ???
Reviewers: reames, fhahn, mkazantsev, chandlerc, craig.topper, courbet
Reviewed By: courbet
Subscribers: hiraditya, xbolva00, nikic, jfb, gchatelet, courbet, llvm-commits, mclow.lists
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D61144
llvm-svn: 370454
Summary:
- Similar to the workaround in fix of PR30188, skip sinking common
lifetime markers of `alloca`. They are mostly left there after
inlining functions in branches.
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D66950
llvm-svn: 370376
This is the naive implementation of x86 BZHI/BEXTR instruction:
it takes input and bit count, and extracts low nbits up to bit width.
I.e. unlike shift it does not have any UB when nbits >= bitwidth.
Which means we don't need a while PHI here, simple select will do.
And if it's a select, it should then be trivial to fix codegen
to select it to BEXTR/BZHI.
See https://bugs.llvm.org/show_bug.cgi?id=34704
llvm-svn: 370369
Summary:
Now that with D65143/D65144 we've produce `@llvm.umul.with.overflow`,
and with D65147 we've flattened the CFG, we now can see that
the guard may have been there to prevent division by zero is redundant.
We can simply drop it:
```
----------------------------------------
Name: no overflow or zero
%iszero = icmp eq i4 %y, 0
%umul = smul_overflow i4 %x, %y
%umul.ov = extractvalue {i4, i1} %umul, 1
%umul.ov.not = xor %umul.ov, -1
%retval.0 = or i1 %iszero, %umul.ov.not
ret i1 %retval.0
=>
%iszero = icmp eq i4 %y, 0
%umul = smul_overflow i4 %x, %y
%umul.ov = extractvalue {i4, i1} %umul, 1
%umul.ov.not = xor %umul.ov, -1
%retval.0 = or i1 %iszero, %umul.ov.not
ret i1 %umul.ov.not
Done: 1
Optimization is correct!
```
Note that this is inverted from what we have in a previous patch,
here we are looking for the inverted overflow bit.
And that inversion is kinda problematic - given this particular
pattern we neither hoist that `not` closer to `ret` (then the pattern
would have been identical to the one without inversion,
and would have been handled by the previous patch), neither
do the opposite transform. But regardless, we should handle this too.
I've filled [[ https://bugs.llvm.org/show_bug.cgi?id=42720 | PR42720 ]].
Reviewers: nikic, spatel, xbolva00, RKSimon
Reviewed By: spatel
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D65151
llvm-svn: 370351
Summary:
Now that with D65143/D65144 we've produce `@llvm.umul.with.overflow`,
and with D65147 we've flattened the CFG, we now can see that
the guard may have been there to prevent division by zero is redundant.
We can simply drop it:
```
----------------------------------------
Name: no overflow and not zero
%iszero = icmp ne i4 %y, 0
%umul = umul_overflow i4 %x, %y
%umul.ov = extractvalue {i4, i1} %umul, 1
%retval.0 = and i1 %iszero, %umul.ov
ret i1 %retval.0
=>
%iszero = icmp ne i4 %y, 0
%umul = umul_overflow i4 %x, %y
%umul.ov = extractvalue {i4, i1} %umul, 1
%retval.0 = and i1 %iszero, %umul.ov
ret %umul.ov
Done: 1
Optimization is correct!
```
Reviewers: nikic, spatel, xbolva00
Reviewed By: spatel
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D65150
llvm-svn: 370350
Summary:
As it can be seen in the tests in D65143/D65144, even though we have formed an '@llvm.umul.with.overflow'
and got rid of potential for division-by-zero, the control flow remains, we still have that branch.
We have this condition:
```
// Don't fold i1 branches on PHIs which contain binary operators
// These can often be turned into switches and other things.
if (PN->getType()->isIntegerTy(1) &&
(isa<BinaryOperator>(PN->getIncomingValue(0)) ||
isa<BinaryOperator>(PN->getIncomingValue(1)) ||
isa<BinaryOperator>(IfCond)))
return false;
```
which was added back in rL121764 to help with `select` formation i think?
That check prevents us to flatten the CFG here, even though we know
we no longer need that guard and will be able to drop everything
but the '@llvm.umul.with.overflow' + `not`.
As it can be seen from tests, we end here because the `not` is being
sinked into the PHI's incoming values by InstCombine,
so we can't workaround this by hoisting it to after PHI.
Thus i suggest that we relax that check to not bailout if we'd get to hoist the `not`.
Reviewers: craig.topper, spatel, fhahn, nikic
Reviewed By: spatel
Subscribers: hiraditya, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D65147
llvm-svn: 370349
Dmitri Gribenko