Summary:
clang-cl tries to match cl's interface, and treats /U as "Removes a
predefined macro" as cl does. When you feed clang-cl a file that begins with
'/U' (e.g. /Users/xiaobai/foo.c), clang-cl will emit a warning and in some cases
an error, like so:
clang-9: warning: '/Users/xiaobai/foo.c' treated as the '/U' option [-Wslash-u-filename]
clang-9: note: Use '--' to treat subsequent arguments as filenames
clang-9: error: no input files
If you're using clang-cl, make sure '--' is passed before the source file.
Differential Revision: https://reviews.llvm.org/D58860
llvm-svn: 355341
Summary:
The current install-clang-headers target installs clang's resource
directory headers. This is different from the install-llvm-headers
target, which installs LLVM's API headers. We want to introduce the
corresponding target to clang, and the natural name for that new target
would be install-clang-headers. Rename the existing target to
install-clang-resource-headers to free up the install-clang-headers name
for the new target, following the discussion on cfe-dev [1].
I didn't find any bots on zorg referencing install-clang-headers. I'll
send out another PSA to cfe-dev to accompany this rename.
[1] http://lists.llvm.org/pipermail/cfe-dev/2019-February/061365.html
Reviewers: beanz, phosek, tstellar, rnk, dim, serge-sans-paille
Subscribers: mgorny, javed.absar, jdoerfert, #sanitizers, openmp-commits, lldb-commits, cfe-commits, llvm-commits
Tags: #clang, #sanitizers, #lldb, #openmp, #llvm
Differential Revision: https://reviews.llvm.org/D58791
llvm-svn: 355340
I'm not too familiar with this pass, so there might be a better
solution, but this appears to fix the degenerate:
PR40930
PR40931
PR40932
PR40934
...without affecting any real-world code.
As we've seen in several other passes, when we have unreachable blocks,
they can contain semi-bogus IR and/or cause unexpected conditions. We
would not typically expect these patterns to make it this far, but we
have to guard against them anyway.
llvm-svn: 355337
LLVM is dropping support for GCC 4.9. This patch adds
a GCC 5 installation to the buildbot image so we can upgrade
the 4.9 bot to GCC 5.
As a temporary workaround until zorg updates, we install GCC 5
as GCC 4.9.
llvm-svn: 355334
When -forder-file-instrumentation is on, we pass llvm flag to enable the order file instrumentation pass.
https://reviews.llvm.org/D58751
llvm-svn: 355333
Summary:
Currently when we see a built-in we try and import the include location. Instead what we do now is find the buffer like we do for the invalid case and copy that over to the to context.
Differential Revision: https://reviews.llvm.org/D58743
llvm-svn: 355332
Part 1 of CSPGO change in Clang. This includes changes in clang options
and calls to llvm PassManager. Tests will be committed in part2.
This change needs the PassManager change in llvm.
Differential Revision: https://reviews.llvm.org/D54176
llvm-svn: 355331
That patch added a function to HostInfo that returns an instance
of UserIDResolver, but this function was unimplemented on Windows,
leading to linker errors. For now, just return a dummy implementation
that doesn't resolve user ids to get the build green.
llvm-svn: 355329
Apparently older versions of clang like 3.6 require an extra set of curly braces around std::array initializations. I'm told the C++ language was changed regarding this by CWG 1270.
llvm-svn: 355327
The code to materialize a mask from a constant pool load tried to use a 128 bit
LDR to load a 64 bit constant pool entry, which was 8 byte aligned. This resulted
in a link failure in the NEON tests in the test suite since the LDR address was
unaligned. This change fixes that to instead emit a 64 bit LDR if the entry is
64 bit, before converting back to a 128 bit register for the TBL.
llvm-svn: 355326
This patch enables combining integer bitcasts of integer build vectors when the new scalar type is legal. I've avoided floating point because the implementation bitcasts float to int along the way and we would need to check the intermediate types for legality
Differential Revision: https://reviews.llvm.org/D58884
llvm-svn: 355324
Summary:
This creates an abstract base class called "UserIDResolver", which can
be implemented to provide user/group ID resolution capabilities for
various objects. Posix host implement a PosixUserIDResolver, which does
that using posix apis (getpwuid and friends). PlatformGDBRemote
forwards queries over the gdb-remote link, etc. ProcessInstanceInfo
class is refactored to make use of this interface instead of taking a
platform pointer as an argument. The base resolver class already
implements caching and thread-safety, so implementations don't have to
worry about that.
The main motivating factor for this was to remove external dependencies
from the ProcessInstanceInfo class (so it can be put next to
ProcessLaunchInfo and friends), but it has other benefits too:
- ability to test the user name caching code
- ability to test ProcessInstanceInfo dumping code
- consistent interface for user/group resolution between Platform and
Host classes.
Reviewers: zturner, clayborg, jingham
Subscribers: mgorny, lldb-commits
Differential Revision: https://reviews.llvm.org/D58167
llvm-svn: 355323
The above builtins are currently implemented for MSVC mode, however GCC
also implements these. This patch enables them for all platforms.
Additionally, this corrects the type for these builtins to always be
'long int' to match the specification in the Intel Intrinsics Guide.
Change-Id: Ida34be98078709584ef5136c8761783435ec02b1
llvm-svn: 355322
Summary:
This is quite minimal so far, introduce them with .section,
fill them with .int8 or .asciz, end with .size
Reviewers: dschuff, sbc100, aheejin
Subscribers: jgravelle-google, sunfish, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D58660
llvm-svn: 355321
Add a convenience 'expectedFailureNetBSD' decorator and mark all tests
currently failing on NetBSD with it. Also skip a few tests that hang
the test suite. This should establish a baseline for the test suite
and get us closer to enabling tests on buildbot. This will help us
catch regressions while we still have a lot of work to do to get tests
working.
It seems that there are also some flaky tests. I am going to address
them later on.
Differential Revision: https://reviews.llvm.org/D58527
llvm-svn: 355320
This patch introduces non-lockfree atomics that do not require using
an external libatomic. This work is done with the long-term goal of
allowing the use of <atomic> in freestanding environments.
Thanks to Olivier Giroux for the patch.
Differential Revision: https://reviews.llvm.org/D56913
llvm-svn: 355318
On SPIR targets, the default calling convention is SpirFunction.
However, operator new/delete and builtins were being created with CC_C.
The result is indirect references to new/delete (or builtins that are permitted
to be called indirectly have a mismatched type, as well as questionable codegen
in some cases.
This patch sets both to the default calling convention, so that it
properly matches the calling convention of the target.
Differential Revision: https://reviews.llvm.org/D58844
Change-Id: I52065bb00bc2655945caea8f29c409ba1e0ac24a
llvm-svn: 355317
This patch adds a new flag named -bottleneck-analysis to print out information
about throughput bottlenecks.
MCA knows how to identify and classify dynamic dispatch stalls. However, it
doesn't know how to analyze and highlight kernel bottlenecks. The goal of this
patch is to teach MCA how to correlate increases in backend pressure to backend
stalls (and therefore, the loss of throughput).
From a Scheduler point of view, backend pressure is a function of the scheduler
buffer usage (i.e. how the number of uOps in the scheduler buffers changes over
time). Backend pressure increases (or decreases) when there is a mismatch
between the number of opcodes dispatched, and the number of opcodes issued in
the same cycle. Since buffer resources are limited, continuous increases in
backend pressure would eventually leads to dispatch stalls. So, there is a
strong correlation between dispatch stalls, and how backpressure changed over
time.
This patch teaches how to identify situations where backend pressure increases
due to:
- unavailable pipeline resources.
- data dependencies.
Data dependencies may delay execution of instructions and therefore increase the
time that uOps have to spend in the scheduler buffers. That often translates to
an increase in backend pressure which may eventually lead to a bottleneck.
Contention on pipeline resources may also delay execution of instructions, and
lead to a temporary increase in backend pressure.
Internally, the Scheduler classifies instructions based on whether register /
memory operands are available or not.
An instruction is marked as "ready to execute" only if data dependencies are
fully resolved.
Every cycle, the Scheduler attempts to execute all instructions that are ready
to execute. If an instruction cannot execute because of unavailable pipeline
resources, then the Scheduler internally updates a BusyResourceUnits mask with
the ID of each unavailable resource.
ExecuteStage is responsible for tracking changes in backend pressure. If backend
pressure increases during a cycle because of contention on pipeline resources,
then ExecuteStage sends a "backend pressure" event to the listeners.
That event would contain information about instructions delayed by resource
pressure, as well as the BusyResourceUnits mask.
Note that ExecuteStage also knows how to identify situations where backpressure
increased because of delays introduced by data dependencies.
The SummaryView observes "backend pressure" events and prints out a "bottleneck
report".
Example of bottleneck report:
```
Cycles with backend pressure increase [ 99.89% ]
Throughput Bottlenecks:
Resource Pressure [ 0.00% ]
Data Dependencies: [ 99.89% ]
- Register Dependencies [ 0.00% ]
- Memory Dependencies [ 99.89% ]
```
A bottleneck report is printed out only if increases in backend pressure
eventually caused backend stalls.
About the time complexity:
Time complexity is linear in the number of instructions in the
Scheduler::PendingSet.
The average slowdown tends to be in the range of ~5-6%.
For memory intensive kernels, the slowdown can be significant if flag
-noalias=false is specified. In the worst case scenario I have observed a
slowdown of ~30% when flag -noalias=false was specified.
We can definitely recover part of that slowdown if we optimize class LSUnit (by
doing extra bookkeeping to speedup queries). For now, this new analysis is
disabled by default, and it can be enabled via flag -bottleneck-analysis. Users
of MCA as a library can enable the generation of pressure events through the
constructor of ExecuteStage.
This patch partially addresses https://bugs.llvm.org/show_bug.cgi?id=37494
Differential Revision: https://reviews.llvm.org/D58728
llvm-svn: 355308
X86TargetLowering::EmitLoweredSelect presently detects sequences of CMOV pseudo
instructions without accounting for debug intrinsics. This leads to different
codegen with and without option -g, if a DBG_VALUE instruction lands in the
middle of several lowered selects.
Work around this by skipping over debug instructions when looking for CMOV
sequences, and sinking those debug insts into the EmitLoweredSelect sunk block.
This might slightly shift where variables appear in the instruction sequence,
but won't re-order assignments.
Differential Revision: https://reviews.llvm.org/D58672
llvm-svn: 355307
The isScaledConstantInRange function takes upper and lower bounds which are
checked after dividing by the scale, so the bounds checks for half, single and
double precision should all be the same. Previously, we had wrong bounds checks
for half precision, so selected an immediate the instructions can't actually
represent.
Differential revision: https://reviews.llvm.org/D58822
llvm-svn: 355305
1) GCC complains that KnownValid is set but not used.
2) In ARMInstructionSelector::selectGlobal() the code is mixing "enumeral
and non-enumeral type in conditional expression". Solve this by casting
to unsigned which is the final type anyway.
Differential Revision: https://reviews.llvm.org/D58834
llvm-svn: 355304
Under the term "subchecker", I mean checkers that do not have a checker class on
their own, like unix.MallocChecker to unix.DynamicMemoryModeling.
Since a checker object was required in order to retrieve checker options,
subcheckers couldn't possess options on their own.
This patch is also an excuse to change the argument order of getChecker*Option,
it always bothered me, now it resembles the actual command line argument
(checkername:option=value).
Differential Revision: https://reviews.llvm.org/D57579
llvm-svn: 355297
Summary:
Before when we implemented the first EH proposal, 'catch <tag>'
instruction may not catch an exception so there were multiple EH pads an
exception can unwind to. That means a BB could have multiple EH pad
successors.
Now after we switched to the new proposal, every 'catch' instruction
catches an exception, and there is only one catchpad per catchswitch, so
we at most have one EH pad successor, making `ThrowUnwindDest` map in
`WasmEHInfo` unnecessary.
Keeping `ThrowUnwindDest` map in `WasmEHInfo` has its own problems,
because other optimization passes can split a BB that contains possibly
throwing calls (previously invokes), and we have to update the map every
time that happens, which is not easy for common CodeGen passes.
This also correctly updates successor info in LateEHPrepare when we add
a rethrow instruction.
Reviewers: dschuff
Subscribers: sbc100, jgravelle-google, sunfish, llvm-commits
Tags: #llvm
Differential Revision: https://reviews.llvm.org/D58486
llvm-svn: 355296
There are no tests for this case, and I'm not sure how it could ever work,
so I'm just removing this option from the matcher. This should fix PR40940:
https://bugs.llvm.org/show_bug.cgi?id=40940
llvm-svn: 355292
InputIsKnownDead check is shared by all operands. Compute it once.
For non-integer instructions, use Visited.insert(I).second to replace a
find() and an insert().
llvm-svn: 355290
Alex Langford