ProfileSummaryInfo already checks whether the module has sample profile
in determining profile counts. This will also be useful in inliner to
clean up threshold updates.
llvm-svn: 303204
Summary:
In SelectionDAG, when a store is immediately chained to another store
to the same address, elide the first store as it has no observable
effects. This is causes small improvements dealing with intrinsics
lowered to stores.
Test notes:
* Many testcases overwrite store addresses multiple times and needed
minor changes, mainly making stores volatile to prevent the
optimization from optimizing the test away.
* Many X86 test cases optimized out instructions associated with
associated with va_start.
* Note that test_splat in CodeGen/AArch64/misched-stp.ll no longer has
dependencies to check and can probably be removed and potentially
replaced with another test.
Reviewers: rnk, john.brawn
Subscribers: aemerson, rengolin, qcolombet, jyknight, nemanjai, nhaehnle, javed.absar, llvm-commits
Differential Revision: https://reviews.llvm.org/D33206
llvm-svn: 303198
Moved slower matchers closer to the end. The total speed up on a large file I
was interested in is not huge, just about 10%, since the check seems to be doing
a lot in the check() method.
llvm-svn: 303191
This inclusion is needed to fix the ARM build. The int_lib.h include is
slightly ugly, but allows us to use the `AEABI_RTABI` macro to decorate
the CC for the functions.
llvm-svn: 303190
These actually may change calling conventions. We cannot simply provide
function aliases as the aliased function may have a different calling
convention. Provide a forwarding function instead to permit the
compiler to synthesize the calling convention adjustment thunk.
Remove the `ARM_EABI_FNALIAS` macro as that is not safe to use.
Resolves PR33030!
llvm-svn: 303188
The referenced tests are derived from:
https://bugs.llvm.org/show_bug.cgi?id=32791
and:
https://reviews.llvm.org/D33172
The motivation for including negative tests may not be clear, so I'm adding an explanatory comment here.
In the post-commit thread for r303133:
http://lists.llvm.org/pipermail/llvm-commits/Week-of-Mon-20170515/453793.html
...it was mentioned that we don't want to add redundant tests. This is a valid point. But in this case,
we have a patch under review (D33172) that demonstrates that no existing regression tests are affected by
a proposed code change, but these are. Therefore, I think these tests have value not visible in any
existing regression tests regardless of whether they show a transform.
Differential Revision: https://reviews.llvm.org/D33242
llvm-svn: 303185
Using LIS can be quite expensive, so caching of calculated region
live-ins and pressure is implemented. It does two things:
1. Caches the info for the second stage when we schedule with
decreased target occupancy.
2. Tracks the basic block from top to bottom thus eliminating the
need to scan whole register file liveness at every region split
in the middle of the block.
The scheduling is now done in 3 stages instead of two, with the first
one being really a no-op and only used to collect scheduling regions
as sent by the scheduler driver.
There is no functional change to the current behavior, only compilation
speed is affected. In general computeBlockPressure() could be simplified
if we switch to backward RP tracker, because scheduler sends regions
within a block starting from the last upward. We could use a natural
order of upward tracker to seamlessly change between regions of the same
block, since live reg set of a previous tracked region would become a
live-out of the next region. That however requires fixing upward tracker
to properly account defs and uses of the same instruction as both are
contributing to the current pressure. When we converge on the produced
pressure we should be able to switch between them back and forth. In
addition, backward tracker is less expensive as it uses LIS in recede
less often than forward uses it in advance.
At the moment the worst known case compilation time has improved from 26
minutes to 8.5.
Differential Revision: https://reviews.llvm.org/D33117
llvm-svn: 303184
According to Intel's Optimization Reference Manual for SNB+:
" For LEA instructions with three source operands and some specific situations, instruction latency has increased to 3 cycles, and must
dispatch via port 1:
- LEA that has all three source operands: base, index, and offset
- LEA that uses base and index registers where the base is EBP, RBP,or R13
- LEA that uses RIP relative addressing mode
- LEA that uses 16-bit addressing mode "
This patch currently handles the first 2 cases only.
Differential Revision: https://reviews.llvm.org/D32277
llvm-svn: 303183
Don't repeat `isInTemplateInstantiation()` and `hasAncestor()` unnecessarily.
This speeds up the check by a factor of up to 3 on some large files.
llvm-svn: 303180
This factors register pressure estimation mechanism from the
GCNSchedStrategy into the forward tracker to unify interface
with other strategies and expose it to other interested phases.
Differential Revision: https://reviews.llvm.org/D33105
llvm-svn: 303179
r276215 made a change to annotate _Unwind_Exception with attribute
"aligned" so that an exception object following field __cxa_exception
is sufficiently aligned. This fix hasn't been incorporated to unwind.h
on Darwin since it is an ABI breaking change.
Instead of annotating struct _Unwind_Exception with the attribute, this
commit annotates field unwindHeader of __cxa_exception. This ensures the
exception object is sufficiently aligned without breaking the ABI.
This recommits r302978 and r302981, which were reverted in r303016
because a libcxx test was failing on an AArch64 bot. I also modified the
libcxxabi test case to check the alignment of the pointer returned by
__cxa_allocate_exception rather than compiling the test with -O1 and
checking whether it segfaults.
rdar://problem/25364625
Differential Revision: https://reviews.llvm.org/D33030
llvm-svn: 303175
Summary:
A refactoring to decouple ProtocolHandlers and Language Server input parsing
loop from the ClangdLSPServer.
The input parsing was extracted from `main` to a function(runLanguageServerLoop).
ProtocolHandlers now provide an interface to handle various LSP methods,
this interface is used by ClangdLSPServer.
Methods for code formatting were moved from ProtocolHandlers to ClangdServer.
ClangdLSPServer now provides a cleaner interface that only runs Language Server
input loop.
Reviewers: bkramer, krasimir
Reviewed By: krasimir
Subscribers: cfe-commits, klimek
Tags: #clang-tools-extra
Differential Revision: https://reviews.llvm.org/D33201
llvm-svn: 303173
Summary:
RewritePHIs algorithm used in building of CoroFrame inserts a placeholder
```
%placeholder = phi [%val]
```
on every edge leading to a block starting with PHI node with multiple incoming edges,
so that if one of the incoming values was spilled and need to be reloaded, we have a
place to insert a reload. We use SplitEdge helper function to split the incoming edge.
SplitEdge function does not deal with unwind edges comping into a block with an EHPad.
This patch adds an ehAwareSplitEdge function that can correctly split the unwind edge.
For landing pads, we clone the landing pad into every edge block and replace the original
landing pad with a PHI collection the values from all incoming landing pads.
For WinEH pads, we keep the original EHPad in place and insert cleanuppad/cleapret in the
edge blocks.
Reviewers: majnemer, rnk
Reviewed By: majnemer
Subscribers: EricWF, llvm-commits
Differential Revision: https://reviews.llvm.org/D31845
llvm-svn: 303172
Recommit of r303159 "[DWARF] - Use DWARFAddressRange struct instead of uint64_t pair for DWARFAddressRangesVector"
All places were shitched to use DWARFAddressRange now.
Suggested during review of D33184.
llvm-svn: 303163
Even though the shouldInlineCall function returns true, it can happen that the
function is not going to be inlined (as it can be seen at line 913 and below).
Moved the bumpNumTimesInlined(D) (the counter increaser) call to the inlineCall
function where it logically belongs.
Differential Revision: https://reviews.llvm.org/D32179
llvm-svn: 303158
This commit fixes a bug that's tracked by PR 10758 and duplicates like PR 30343.
The bug causes clang to crash with a stack overflow while recursing infinitely
trying to perform copy-initialization on a type without a copy constructor but
with a constructor that accepts another type that can be constructed using the
original type.
The commit fixes this bug by detecting the recursive behavior and failing
correctly with an appropriate error message. It also tries to provide a
meaningful diagnostic note about the constructor which leads to this behavior.
rdar://28483944
Differential Revision: https://reviews.llvm.org/D25051
llvm-svn: 303156
The information collected when requested by -time-passes is only printed when
llvm_shutdown is called at the moment. This means that when linking against the LTO
library dynamically and using the C interface, it is not possible to see the timing
information, because llvm_shutdown cannot be called. This change modifies the LTO
code generation functions for both regular LTO and thin LTO to explicitly print and
reset the timing information.
I have tested that this works with our proprietary linker. However, as this relies
on a specific method of building and linking against the LTO library, I'm not sure
how or if this can be tested in the LLVM testsuite.
Reviewed by: mehdi_amini
Differential Revision: https://reviews.llvm.org/D32803
llvm-svn: 303152
Summary:
This commit restores r303067(reverted by r303094) and fixes the 'formatting.test'
failure.
The failure is due to destructors of `ClangdLSPServer`'s fields(`FixItsMap` and
`FixItsMutex`) being called before destructor of `Server`. It led to the worker
thread calling `consumeDiagnostics` after `FixItsMutex` and `FixItsMap`
destructors were called.
Also, clangd is now run with '-run-synchronously' flag in 'formatting.test'.
Reviewers: bkramer, krasimir
Reviewed By: krasimir
Subscribers: mgorny, cfe-commits, klimek
Differential Revision: https://reviews.llvm.org/D33233
llvm-svn: 303151
Easwaran Raman