Add +fp16fml feature for new FP16 instructions, which are a
mandatory part of FP16 from v8.4-A and an optional part of FP16
from v8.2-A. It doesn't seem to be possible to model this in
LLVM, but the relationship between the options is handled by
the related clang patch.
In keeping with what I think is the usual practice, the fp16fml
extension is accepted regardless of base architecture version.
Builds on/replaces Sjoerd Meijer's patch to add these instructions at
https://reviews.llvm.org/D49839.
Differential Revision: https://reviews.llvm.org/D50228
llvm-svn: 340013
Adds some missing tests for the FP16 extension,
fixes an existing test that misnames it.
Differential Revision: https://reviews.llvm.org/D50227
llvm-svn: 340012
Add support for cases where only some c1+c2 results exceed the max bitshift, clamping accordingly.
Differential Revision: https://reviews.llvm.org/D35722
llvm-svn: 340010
Summary:
Looking at the callee argument list, as is done now, might not work if
the function has been typecasted into one that is expected to return
a struct. This change also simplifies the code.
The isFP128ABICall() function can be removed as it is no longer needed.
The test in fp128.ll has been updated to verify this.
Reviewers: jyknight, venkatra
Reviewed By: jyknight
Subscribers: fedor.sergeev, jrtc27, llvm-commits
Differential Revision: https://reviews.llvm.org/D48117
llvm-svn: 340008
Workspace Symbol implementation was introduced in D44882 and should be
complete now.
Reviewed by: ioeric
Differential Revision: https://reviews.llvm.org/D50703
llvm-svn: 340007
Summary:
Sema can only be used for documentation in the current file, other doc
comments should be fetched from the index.
Reviewers: hokein, ioeric, kadircet
Reviewed By: hokein, kadircet
Subscribers: MaskRay, jkorous, mgrang, arphaman, cfe-commits
Differential Revision: https://reviews.llvm.org/D50727
llvm-svn: 340005
Summary:
Previously, clangd was trying to show documentation for the active
parameter instead, which is wrong per LSP specification.
Moreover, the code path that attempts to get parameter comments never
succeds, because no attempt is made to parse function doc comment and
extract parameter-specific parts out of it. So we also remove the code
that claims to fetch parameter comments: it is not used anymore and is
incorrect.
Reviewers: hokein, ioeric, kadircet
Reviewed By: ioeric
Subscribers: MaskRay, jkorous, arphaman, cfe-commits
Differential Revision: https://reviews.llvm.org/D50726
llvm-svn: 340004
Summary: When @llvm.returnaddress is called with a value higher than 0
it needs to read from the call stack to get the return address. This
means that the register windows needs to be flushed to the stack to
guarantee that the data read is valid. For values higher than 1 this
is done indirectly by the call to getFRAMEADDR(), but not for the value 1.
Reviewers: jyknight, venkatra
Reviewed By: jyknight
Subscribers: fedor.sergeev, jrtc27, llvm-commits
Differential Revision: https://reviews.llvm.org/D48636
llvm-svn: 340003
Summary:
Fix an inconsistent behavior of using `LastBuiltPreamble`/`NewPreamble`
in TUScheduler (see the test for details), AST should always use
NewPreamble. This patch makes LastBuiltPreamble always point to
NewPreamble.
Preamble rarely fails to build, even there are errors in headers, so we
assume it would not cause performace issue for code completion.
Reviewers: ilya-biryukov
Reviewed By: ilya-biryukov
Subscribers: javed.absar, ioeric, MaskRay, jkorous, arphaman, cfe-commits
Differential Revision: https://reviews.llvm.org/D50695
llvm-svn: 340001
The description of `isGuaranteedToExecute` does not correspond to its implementation.
According to description, it should return `true` if an instruction is executed under the
assumption that its loop is *entered*. However there is a sophisticated alrogithm inside
that tries to prove that the instruction is executed if the loop is *exited*, which is not the
same thing for infinite loops. There is an attempt to protect from dealing with infinite loops
by prohibiting loops without exit blocks, however an infinite loop can have exit blocks.
As result of that, MustExecute can falsely consider some blocks that are never entered as
mustexec, and LICM can hoist dangerous instructions out of them basing on this fact.
This may introduce UB to programs which did not contain it initially.
This patch removes the problematic algorithm and replaced it with a one which tries to
prove what is required in description.
Differential Revision: https://reviews.llvm.org/D50558
Reviewed By: reames
llvm-svn: 339984
Summary:
Formerly, all timer groups were automatically cleared when printed out. In
https://reviews.llvm.org/rL324788 this behaviour was changed to not-clearing
timers on printout, to allow printing timers more than once, but as a result
clients (specifically Swift) that relied on the clear-on-print behaviour to
inhibit duplicate timer printing on shutdown were broken.
Rather than revert that change, this change adds a new API that enables
clients that _want_ to clear all timers to do so explicitly.
Reviewers: george.karpenkov, thegameg
Reviewed By: george.karpenkov
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D50874
llvm-svn: 339980
"-fno-use-cxa-atexit" was a default provided by the initial
commit offering hexagon support. This is no longer required.
Reviewers: bcahoon, sidneym
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D50816
llvm-svn: 339979
Summary:
We avoid using dynamic memory allocated with the internal allocator in
the profile collection service used by profiling mode. We use aligned
storage for globals and in-struct storage of objects we dynamically
initialize.
We also remove the dependency on `Vector<...>` which also internally
uses the dynamic allocator in sanitizer_common (InternalAlloc) in favour
of the XRay allocator and segmented array implementation.
This change addresses llvm.org/PR38577.
Reviewers: eizan
Reviewed By: eizan
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D50782
llvm-svn: 339978
https://reviews.llvm.org/D50401
Add opcodes for llvm.intrinsic.trunc, round, and update the IRTranslator
for the same.
Reviewed by: dsanders.
llvm-svn: 339977
constructor.
This breaking an old/weird host compiler is my best bet for the current
crashes I'm getting from bots since this functionality was added to this
ADT.
llvm-svn: 339975
Summary:
.rela.debug_info relocations are being done via
ObjectFileELF::ApplyRelocations for aarch64. Currently, the switch case
that iterates over the relocation type is only implemented for a few
different types and `assert(false)`es over the rest.
Implement the relocation for R_AARCH64_ABS32 in ApplyRelocations
Reviewers: sas, xiaobai, javed.absar, espindola
Reviewed By: sas
Subscribers: emaste, arichardson, kristof.beyls
Differential Revision: https://reviews.llvm.org/D50369
Change by Nathan Lanza <lanza@fb.com>
llvm-svn: 339974
Summary:
Since we're casting from double to size_t during input selection, we
really want a discrete distribution over size_t rather than a piecewise
distribution over doubles.
Reviewers: kcc
Reviewed By: kcc
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D50356
llvm-svn: 339973
Summary:
When a seed sequence would lead to having no non-zero significant bits
in the initial state of a `mersenne_twister_engine`, the fallback is to
flip the most significant bit of the first value that appears in the
textual representation of the initial state.
rand.eng.mers describes this as setting the value to be 2 to the power
of one less than w; the previous value encoded in the implementation,
namely one less than "2 to the power of w", is replaced by the correct
value in this patch.
Reviewers: mclow.lists, EricWF, jasonliu
Reviewed By: mclow.lists
Subscribers: mclow.lists, jasonliu, EricWF, christof, ldionne, cfe-commits
Differential Revision: https://reviews.llvm.org/D50736
llvm-svn: 339969
Summary:
This adds support for exception handling to CFGStackify pass. This only
adds TRY / END_TRY markers and DOES NOT yet fix unwind mismatches that
can be created by the linearization of the CFG into the structural wasm
format. The mismatch fix will be added by following patches.
In detail, this patch
- Added support for TRY / END_TRY markers to support EH
- Changed many static functions into class member functions as they take
too many arguments now
- Added several more bookeeping data structures
- Refactored routines that decide where to insert markers, because
without refactoring this got too complicated as we added support for new
kinds of markers (TRY/END_TRY).
- Rewrote rethrow instructions' BB arguments to relative depths in EH
pad stack.
Reviewers: dschuff, sunfish
Subscribers: sbc100, jgravelle-google, llvm-commits
Differential Revision: https://reviews.llvm.org/D48273
llvm-svn: 339967
This reapplies commit r339935 with the following changes:
* make longjmp test C, not C++, to avoid dependency on libc++/libstdc++
* untag pointer in memset interceptor
x86_64 does not have TBI, so hwasan barely works there. Tests must be carefully
written in a way that does not leak tagged pointer to system libraries.
llvm-svn: 339963
well as MIR parsing support for `MCSymbol` `MachineOperand`s.
The only real way to test pre- and post-instruction symbol support is to
use them in operands, so I ended up implementing that within the patch
as well. I can split out the operand support if folks really want but it
doesn't really seem worth it.
The functional implementation of pre- and post-instruction symbols is
now *completely trivial*. Two tiny bits of code in the (misnamed)
AsmPrinter. It should be completely target independent as well. We emit
these exactly the same way as we emit basic block labels. Most of the
code here is to give full dumping, MIR printing, and MIR parsing support
so that we can write useful tests.
The MIR parsing of MC symbol operands still isn't 100%, as it forces the
symbols to be non-temporary and non-local symbols with names. However,
those names often can encode most (if not all) of the special semantics
desired, and unnamed symbols seem especially annoying to serialize and
de-serialize. While this isn't perfect or full support, it seems plenty
to write tests that exercise usage of these kinds of operands.
The MIR support for pre-and post-instruction symbols was quite
straightforward. I chose to print them out in an as-if-operand syntax
similar to debug locations as this seemed the cleanest way and let me
use nice introducer tokens rather than inventing more magic punctuation
like we use for memoperands.
However, supporting MIR-based parsing of these symbols caused me to
change the design of the symbol support to allow setting arbitrary
symbols. Without this, I don't see any reasonable way to test things
with MIR.
Differential Revision: https://reviews.llvm.org/D50833
llvm-svn: 339962
These TODOs were for setting m_type in RegisterValue::SetValueFromString
in the case where reg_info's encoding was eEncodingUint or
eEncodingSint. m_type is set by SetUInt{8,16,32,64.128} during the
SetUInt call.
llvm-svn: 339959
This is a follow-up suggested with rL339604.
For tan(), we don't have a corresponding LLVM
intrinsic -- unlike sin/cos -- so this is the
only way/place that we can do this fold currently.
llvm-svn: 339958
Thread sanitizer instrumentation fails to skip all loads and stores to
profile counters. This can happen if profile counter updates are merged:
%.sink = phi i64* ...
%pgocount5 = load i64, i64* %.sink
%27 = add i64 %pgocount5, 1
%28 = bitcast i64* %.sink to i8*
call void @__tsan_write8(i8* %28)
store i64 %27, i64* %.sink
To suppress TSan diagnostics about racy counter updates, make the
counter updates atomic when TSan is enabled. If there's general interest
in this mode it can be surfaced as a clang/swift driver option.
Testing: check-{llvm,clang,profile}
rdar://40477803
Differential Revision: https://reviews.llvm.org/D50867
llvm-svn: 339955
Bernard Ogden