Remove broken patterns matching it. This was matching the
unsafe math pattern and expanding the fix for the buggy instruction
from the pattern. The problems are also on CI. Remove the workarounds
and only use fract with unsafe math or from the intrinsic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271078 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Unroll factor (Count) calculations moved to a new function.
Early exits on pragma and "-unroll-count" defined factor added.
New type of unrolling "Force" introduced (previously used implicitly).
New unroll preference "AllowRemainder" introduced and set "true" by default.
(should be set to false for architectures that suffers from it).
Reviewers: hfinkel, mzolotukhin, zzheng
Differential Revision: http://reviews.llvm.org/D19553
From: Evgeny Stupachenko <evstupac@gmail.com>
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271071 91177308-0d34-0410-b5e6-96231b3b80d8
warning: cast from type ‘const void*’ to type
‘llvm::PrettyStackTraceEntry*’ casts away qualifiers [-Wcast-qual]
PrettyStackTraceHead = (PrettyStackTraceEntry*)Top;
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271069 91177308-0d34-0410-b5e6-96231b3b80d8
DynamicNoPIC was only every used on darwin. This maps it to static on
ELF. It matches what is done on X86.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271052 91177308-0d34-0410-b5e6-96231b3b80d8
When running mir tests, a pass created in that constructor would not be
freed, leading to memory leaks.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271043 91177308-0d34-0410-b5e6-96231b3b80d8
We were accidentally bounds checking the read against the output
ArrayRef instead of against the size of the read.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271040 91177308-0d34-0410-b5e6-96231b3b80d8
Due to differences in template instantiation rules, it is not
portable to static_assert(false) inside of an invalid specialization
of a template. Instead I just =delete the method so that it can't
be used, and leave a comment that it must be explicitly specialized.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271027 91177308-0d34-0410-b5e6-96231b3b80d8
This reverts commit r271024 due to error: static_assert failed
"You must either provide a specialization of VarStreamArrayExtractor
or a custom extractor"
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271026 91177308-0d34-0410-b5e6-96231b3b80d8
r231483 taught ConstantRange::multiply to be clever about signed vs unsigned ranges. For example, an unsigned range could be full-set while the signed range is more specific than that.
In looking at the allocations trace for LTO'ing verify-uselistorder (see r236629 for details), millions of allocations are from APInt, many of which come from ConstantRange's.
This change tries to avoid some (3.2 million) allocations by returning the unsigned range if its suitable. The checks here are that it should not be a wrapping range, and should be positive. That should be enough to check for ranges such as [1, 10) which the signed range will be equal to, if we were to calculate it.
Differential Revision: http://reviews.llvm.org/D20723
Reviewed by James Molloy
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271020 91177308-0d34-0410-b5e6-96231b3b80d8
When we traced through a phi node looking for floating-point reductions, we
forgot whether we'd ever seen an instruction without fast-math flags (that
would block vectorization). This propagates it through to the end.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271015 91177308-0d34-0410-b5e6-96231b3b80d8
Summary:
Turn off lifetime-start-on-first-use enhancement for the moment
pending a fix for bug 27903.
Bug: 27903
Reviewers: tejohnson, wmi, qcolombet, gbiv
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D20731
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@271003 91177308-0d34-0410-b5e6-96231b3b80d8
The isMemWithSimmOffset predicate rejects relocations which is incorrect
behaviour. Linkers and other tools should handle|warn|error when the
field overflows.
Reviewers: dsanders, vkalintiris
Differential Revision: http://reviews.llvm.org/D20727
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270995 91177308-0d34-0410-b5e6-96231b3b80d8
Currently we consider that each constant has itself as a base value. I.e "base(const) = const".
This introduces couple of problems when we are trying to avoid reporting constants in statepoint live sets:
1. When querying "base( phi(const1, const2) )" we will get "phi(const1, const2)" as a base pointer. Since
it's not a constant we will record it in a stack map. However on practice we don't want this to happen
(constant are never relocated).
2. base( phi(const, gc ptr) ) = phi( const, base(gc ptr) ). This particular case imposes challenge on our
runtime - we don't expect to see constant base pointers other than null. This problems can be avoided
by treating all constant as if they were derived from null pointer base. I.e in a first case we will
not include constant pointer in a stack map at all. In a second case we will get "phi(null, base(gc ptr))"
as a base pointer which is a lot more convenient.
Differential Revision: http://reviews.llvm.org/D20584
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270993 91177308-0d34-0410-b5e6-96231b3b80d8
Register numbers may be specified as assembly-time expressions.
This feature can be useful in macros and alike. However, expressions
are supported within sqare braces only.
Sqare braces were initially intended to support specifying of multiple
(pairs/quads...) registers. Syntax like v[8:8] which specifies single register
is also supported. That allows expressions but looks a bit unnatural.
This change supports syntax REG[EXPR].
Tests added.
Differential Revision: http://reviews.llvm.org/D20588
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270990 91177308-0d34-0410-b5e6-96231b3b80d8
clang-tidy's performance-unnecessary-copy-initialization with some manual
fixes. No functional changes intended.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270988 91177308-0d34-0410-b5e6-96231b3b80d8
Fix: updated clang code which was not updated by mistake.
Original commit message:
[llvm-mc] - Teach llvm-mc to generate zlib styled compression sections.
This patch is strongly based on previously reverted D20331.
(because of gnuutils < 2.26 does not support compressed debug sections in non zlib-gnu style)
Difference that this patch supports both zlib and zlib-gnu styles.
-compress-debug-sections option now supports next values:
-compress-debug-sections=zlib-gnu
-compress-debug-sections=zlib
-compress-debug-sections=none
Previously specifying -compress-debug-sections enabled zlib-gnu compression,
so anyone can put "-compress-debug-sections=zlib-gnu" to restore the behavior
that was before this patch for case when compression was enabled.
Differential revision: http://reviews.llvm.org/D20676
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270987 91177308-0d34-0410-b5e6-96231b3b80d8
It broke buildbot:
http://lab.llvm.org:8011/builders/llvm-clang-lld-x86_64-scei-ps4-ubuntu-fast/builds/13585/steps/build/logs/stdio
Initial commit message:
[llvm-mc] - Teach llvm-mc to generate zlib styled compression sections.
This patch is strongly based on previously reverted D20331.
(because of gnuutils < 2.26 does not support compressed debug sections in non zlib-gnu style)
Difference that this patch supports both zlib and zlib-gnu styles.
-compress-debug-sections option now supports next values:
-compress-debug-sections=zlib-gnu
-compress-debug-sections=zlib
-compress-debug-sections=none
Previously specifying -compress-debug-sections enabled zlib-gnu compression,
so anyone can put "-compress-debug-sections=zlib-gnu" to restore the behavior
that was before this patch for case when compression was enabled.
Differential revision: http://reviews.llvm.org/D20676
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270978 91177308-0d34-0410-b5e6-96231b3b80d8
This patch is strongly based on previously reverted D20331.
(because of gnuutils < 2.26 does not support compressed debug sections in non zlib-gnu style)
Difference that this patch supports both zlib and zlib-gnu styles.
-compress-debug-sections option now supports next values:
-compress-debug-sections=zlib-gnu
-compress-debug-sections=zlib
-compress-debug-sections=none
Previously specifying -compress-debug-sections enabled zlib-gnu compression,
so anyone can put "-compress-debug-sections=zlib-gnu" to restore the behavior
that was before this patch for case when compression was enabled.
Differential revision: http://reviews.llvm.org/D20676
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270977 91177308-0d34-0410-b5e6-96231b3b80d8
This patch removes the llvm intrinsics VPMOVSX and (V)PMOVZX sign/zero extension intrinsics and auto-upgrades to SEXT/ZEXT calls instead. We already did this for SSE41 PMOVSX sometime ago so much of that implementation can be reused.
A companion patch (D20684) removes/auto-upgrade the clang intrinsics.
Differential Revision: http://reviews.llvm.org/D20686
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270973 91177308-0d34-0410-b5e6-96231b3b80d8
APInt::operator+(uint64_t) just forwarded to operator+(const APInt&).
Constructing the APInt for the RHS takes an allocation which isn't
required. Also, for APInt's in the slow path, operator+ would
call add() internally which iterates over both arrays of values. Instead
we can use add_1 and sub_1 which only iterate while there is something to do.
Using the memory for 'opt -O2 verify-uselistorder.lto.opt.bc -o opt.bc'
(see r236629 for details), this reduces the number of allocations from
23.9M to 22.7M.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270959 91177308-0d34-0410-b5e6-96231b3b80d8
Add a new CMake function (``add_libfuzzer_test()``) to simplify
declaration of executables for testing LibFuzzer and use it to
reorganise how tests are declared.
Note that configuration of the lit configuration files has been moved
as late as possible because we are going to need to disable some tests
for some platforms and we will need to propagate this information into
the lit configuration.
Note the code for custom mains was removed because no tests are
currently written for this and Kostya seems happy to remove this.
Differential Revision: http://reviews.llvm.org/D20706
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270958 91177308-0d34-0410-b5e6-96231b3b80d8
objc_storeStrong can be formed from a sequence such as
%0 = tail call i8* @objc_retain(i8* %p) nounwind
%tmp = load i8*, i8** @x, align 8
store i8* %0, i8** @x, align 8
tail call void @objc_release(i8* %tmp) nounwind
The code was already looking through bitcasts for most of the values
involved, but had missed one case where the pointer operand for the
store was a bitcast. Ultimately the pointer for the load and store
have to be the same value, after stripping casts.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270955 91177308-0d34-0410-b5e6-96231b3b80d8
PDBs can be extremely large. We're already mapping the entire
PDB into the process's address space, but to make matters worse
the blocks of the PDB are not arranged contiguously. So, when
we have something like an array or a string embedded into the
stream, we have to make a copy. Since it's convenient to use
traditional data structures to iterate and manipulate these
records, we need the memory to be contiguous.
As a result of this, we were using roughly twice as much memory
as the file size of the PDB, because every stream was copied
out and re-stitched together contiguously.
This patch addresses this by improving the MappedBlockStream
to allocate from a BumpPtrAllocator only when a read requires
a discontiguous read. Furthermore, it introduces some data
structures backed by a stream which can iterate over both
fixed and variable length records of a PDB. Since everything
is backed by a stream and not a buffer, we can read almost
everything from the PDB with zero copies.
Differential Revision: http://reviews.llvm.org/D20654
Reviewed By: ruiu
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270951 91177308-0d34-0410-b5e6-96231b3b80d8
StringError can be used to represent Errors that aren't recoverable based on
the error type, but that have a useful error message that can be reported to
the user or logged.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270948 91177308-0d34-0410-b5e6-96231b3b80d8
If every operands of a constant are mapping to themselves, and the
type does not change, we have an early exit as acknowledged in the
comment:
// Otherwise, we have some other constant to remap. Start by checking to see
// if all operands have an identity remapping.
However instead of checking for identity the code was checking if the
operands were mapped to the constant itself, which is rarely true.
As a consequence, the coverage report showed that the early exit was
never taken.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270944 91177308-0d34-0410-b5e6-96231b3b80d8
CriticalAntiDepBreaker was not correctly tracking defs of the high X86 byte
registers, leading to incorrect use of a busy register to break an
antidependence.
Fixes pr27681, and its duplicates pr27580, pr27804.
Differential Revision: http://reviews.llvm.org/D20456
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270935 91177308-0d34-0410-b5e6-96231b3b80d8
Condition might be simplified to a Constant, but it doesn't have to be
ConstantInt, so we should dyn_cast, instead of cast.
This fixes PR27886.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270924 91177308-0d34-0410-b5e6-96231b3b80d8
This patch builds upon r270776 and speeds up
LiveDebugValues::transferDebugValue() by adding an index that maps each
DebugVariable to its open VarLoc.
The transferDebugValue() function needs to close all open ranges for a
given DebugVariable. Iterating over the set bits of OpenRanges is
prohibitively slow in practice. I experimented with using the sorted map
of VarLocs in the UniqueVector to iterate only over the range of VarLocs
with a given DebugVariable, but the binary search turned out to be even
more expensive than just iterating over the set bits in OpenRanges.
Instead, this patch exploits the fact that there can only be one open
location for each DebugVariable and redundantly stores this location in a
DenseMap.
This patch brings the time spent in the LiveDebugValues pass down to an
almost neglectiable amount.
http://llvm.org/bugs/show_bug.cgi?id=26055http://reviews.llvm.org/D20636
rdar://problem/24091200
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270923 91177308-0d34-0410-b5e6-96231b3b80d8
This adds support for YAML round tripping dyld info lazy bindings. The storage and format of these is the same as regular bind opcodes, they are just interpreted differently by dyld, and can have DONE opcodes in the middle of the opcode lists.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270920 91177308-0d34-0410-b5e6-96231b3b80d8
Previously the flags were only being set correctly when the
build type was "Release". Now the build should work properly
for all the supported build types. When building libFuzzer
the optimization level respects whatever is used for the
rest of LLVM but for the LibFuzzer tests we force -O0.
Differential Revision: http://reviews.llvm.org/D20558
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270912 91177308-0d34-0410-b5e6-96231b3b80d8
This adds support for YAML round tripping dyld info weak bindings. The storage and format of these is the same as regular bind opcodes, they are just interpreted differently by dyld.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270911 91177308-0d34-0410-b5e6-96231b3b80d8
We were creating a weak external that tried to reference a static symbol. That
would always fail to link with link.exe.
We now create an external symbol in the same position as the local and refer
to that. This works with link.exe and matches what gas does.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270906 91177308-0d34-0410-b5e6-96231b3b80d8
This adds support for YAML round tripping dyld info bind opcodes. Bind opcodes can have signed or unsigned LEB128 data, and they can have symbols associated with them.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270901 91177308-0d34-0410-b5e6-96231b3b80d8
The aggressive anti-dependency breaker can rename the restored callee-
saved registers. To prevent this, mark these registers are live on all
paths to the return/tail-call instructions, and add implicit use operands
for them to these instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270898 91177308-0d34-0410-b5e6-96231b3b80d8
Canonicalize (srl (bswap i32 x), 16) to (rotr (bswap i32 x), 16), if the high
16-bits of x are zero. Similarly, canonicalize (srl (bswap i64 x), 32) to
(rotr (bswap i64 x), 32), if the high 32-bits of x are zero.
test_rev_w_srl16: test_rev_w_srl16:
and w8, w0, #0xffff and w8, w0, #0xffff
rev w8, w8 ---> rev16 w0, w8
lsr w0, w8, #16
test_rev_x_srl32: test_rev_x_srl32:
rev x8, x8 ---> rev32 x0, x8
lsr x0, x8, #32
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270896 91177308-0d34-0410-b5e6-96231b3b80d8
r270777 improved the precision of alloca vs. inbounbds GEP alias queries: if
we have (a) an inbounds GEP and (b) a pointer based on an alloca, and the
beginning of the object the GEP points to would have a negative offset with
respect to the alloca, then the GEP can not alias pointer (b).
This makes the same logic fire when (b) is based on a GlobalVariable instead
of an alloca.
Differential Revision: http://reviews.llvm.org/D20652
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270893 91177308-0d34-0410-b5e6-96231b3b80d8
An exception could prevent a store from occurring but MemCpyOpt's
callslot optimization would fire anyway, causing the store to occur.
This fixes PR27849.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270892 91177308-0d34-0410-b5e6-96231b3b80d8
APInt::slt was copying the LHS and RHS in to temporaries then making
them unsigned so that it could use an unsigned comparision. It did
this even on the paths which were trivial to give results for, such
as the sign bit of the LHS being set while RHS was not set.
This changes the logic to return out immediately in the trivial cases,
and use an unsigned comparison in the remaining cases. But this time,
just use the unsigned comparison directly without creating any temporaries.
This works because, for example:
true = (-2 slt -1) = (0xFE ult 0xFF)
Also added some tests explicitly for slt with APInt's larger than 64-bits
so that this new code is tested.
Using the memory for 'opt -O2 verify-uselistorder.lto.opt.bc -o opt.bc'
(see r236629 for details), this reduces the number of allocations from
26.8M to 23.9M.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270881 91177308-0d34-0410-b5e6-96231b3b80d8
Fixes build error on windows where MSVC does not
support list initialization inside member initializer list.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@270877 91177308-0d34-0410-b5e6-96231b3b80d8
