Fix two build issues:
1) Recent commit 348756 accidentally included Unix clang compilers
to use immintrin.h when only clang-cl should be using it leading
to the following error:
openmp-llvm/runtime/src/kmp_lock.cpp:2035:25: error: always_
inline function '_xbegin' requires target feature 'rtm', but would be inlined into function
'__kmp_test_adaptive_lock_only' that is compiled without support for 'rtm'
kmp_uint32 status = _xbegin();
This patch changes the guard to use immintrin.h to only use clang-cl instead of all clang
2) gcc-8 gives a warning about multiline comment in kmp_runtime.cpp:
This patch just changes it to a two line comment
openmp-llvm/runtime/src/kmp_runtime.cpp:7697:8: warning: multi-line comment [-Wcomment]
#endif // KMP_OS_LINUX || KMP_OS_DRAGONFLY || KMP_OS_FREEBSD || KMP_OS_NETBSD \
llvm-svn: 348783
As discussed on D55511, this caused an issue if the inner node deletes a node that the outer node depends upon. As it doesn't affect any lit-tests and I've only been able to expose this with the D55511 change I'm committing this now.
llvm-svn: 348781
Patch from Jordan Soyke (jsoyke@google.com)
Reviewed as D55520
This change adds a new internal class, called __value_func, that adds
a minimal subset of value-type semantics to the internal __func interface.
The change is NFC, and is cleanup for the upcoming ABI v2 function implementation (D55045).
llvm-svn: 348778
Summary:
SSE2 vectorization was added in 2012, but it is 2018 now and I can't
observe any performance boost (testing clang -E [all Sema/* CodeGen/* with proper -I options]) with the existing _mm_movemask_epi8+countTrailingZeros or the following SSE4.2 (compiling with -msse4.2):
__m128i C = _mm_setr_epi8('\r','\n',0,0,0,0,0,0,0,0,0,0,0,0,0,0);
_mm_cmpestri(C, 2, Chunk, 16, _SIDD_UBYTE_OPS | _SIDD_CMP_EQUAL_ANY | _SIDD_POSITIVE_POLARITY | _SIDD_LEAST_SIGNIFICANT)
Delete the vectorization to simplify the code.
Also simplify the code a bit and don't check the line ending sequence \n\r
Reviewers: bkramer, #clang
Reviewed By: bkramer
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D55484
llvm-svn: 348777
This should really be generalized to allow increment and/or
we should replace it by using ISD::matchUnaryPredicate().
See D55515 for context.
llvm-svn: 348776
Summary: Instead use a more reasonable value to start and rely on the fact that SmallString will resize if necessary.
Reviewers: labath, asmith
Reviewed By: labath
Subscribers: lldb-commits
Differential Revision: https://reviews.llvm.org/D55457
llvm-svn: 348775
Refactor the scheduling predicates based on `MCInstPredicate`. In this
case, for the Exynos processors.
Differential revision: https://reviews.llvm.org/D55345
llvm-svn: 348774
This re-commits r348592, which was reverted due to a failing test on
macos.
The issue was that I was passing a null pointer for the
"CreateMemoryInstance" callback when registering ObjectFileBreakpad,
which caused crashes when attemping to load modules from memory. The
correct thing to do is to pass a callback which always returns a null
pointer (as breakpad files are never loaded in inferior memory).
It turns out that there is only one test which exercises this code path,
and it's mac-only, so I've create a new test which should run everywhere
(except windows, as one cannot delete an executable which is being run).
Unfortunately, this test still fails on linux for other reasons, but at
least it gives us something to aim for.
The original commit message was:
This patch adds the scaffolding necessary for lldb to recognise symbol
files generated by breakpad. These (textual) files contain just enough
information to be able to produce a backtrace from a crash
dump. This information includes:
- UUID, architecture and name of the module
- line tables
- list of symbols
- unwind information
A minimal breakpad file could look like this:
MODULE Linux x86_64 0000000024B5D199F0F766FFFFFF5DC30 a.out
INFO CODE_ID 00000000B52499D1F0F766FFFFFF5DC3
FILE 0 /tmp/a.c
FUNC 1010 10 0 _start
1010 4 4 0
1014 5 5 0
1019 5 6 0
101e 2 7 0
PUBLIC 1010 0 _start
STACK CFI INIT 1010 10 .cfa: $rsp 8 + .ra: .cfa -8 + ^
STACK CFI 1011 $rbp: .cfa -16 + ^ .cfa: $rsp 16 +
STACK CFI 1014 .cfa: $rbp 16 +
Even though this data would normally be considered "symbol" information,
in the current lldb infrastructure it is assumed every SymbolFile object
is backed by an ObjectFile instance. So, in order to better interoperate
with the rest of the code (particularly symbol vendors).
In this patch I just parse the breakpad header, which is enough to
populate the UUID and architecture fields of the ObjectFile interface.
The rough plan for followup patches is to expose the individual parts of
the breakpad file as ObjectFile "sections", which can then be used by
other parts of the codebase (SymbolFileBreakpad ?) to vend the necessary
information.
Reviewers: clayborg, zturner, lemo, amccarth
Subscribers: mgorny, fedor.sergeev, markmentovai, lldb-commits
Differential Revision: https://reviews.llvm.org/D55214
llvm-svn: 348773
Summary:
Use the original shuffle implementation for __kmpc_shuffle_int64 since
default implementation uses the same implementation.
Reviewers: gtbercea
Subscribers: guansong, caomhin, openmp-commits
Differential Revision: https://reviews.llvm.org/D55514
llvm-svn: 348772
This commit changes which l1 flush instruction is used for AMDPAL and
MESA3d workloads to flush the entire l1 cache instead of just the
volatile lines.
Differential Revision: https://reviews.llvm.org/D55367
llvm-svn: 348771
Refactor the scheduling predicates based on `MCInstPredicate`. Augment the
number of helper predicates used by processor specific predicates.
Differential revision: https://reviews.llvm.org/D55375
llvm-svn: 348768
Use zip_longest in two locations that compare iterator ranges.
zip_longest allows the iteration using a range-based for-loop and to be
symmetric over both ranges instead of prioritizing one over the other.
In that latter case code have to handle the case that the first is
longer than the second, the second is longer than the first, and both
are of the same length, which must partially be checked after the loop.
With zip_longest, this becomes an element comparison within the loop
like the comparison of the elements themselves. The symmetry makes it
clearer that neither the first and second iterators are handled
differently. The iterators are not event used directly anymore, just
the ranges.
Differential Revision: https://reviews.llvm.org/D55468
llvm-svn: 348762
Record the stack protector index in MachineFrameInfo when translating
Intrinsic::stackprotector similarly as is done by SelectionDAG when
processing the same intrinsic.
Setting this index allows the Prologue/Epilogue Insertion to recognize
that the stack protection is enabled. The pass can then make sure that
the stack protector comes before local variables on the stack and
assigns potentially vulnerable objects first so they are close to the
stack protector slot.
Differential Revision: https://reviews.llvm.org/D55418
llvm-svn: 348761
When replacing jal with jalr, also emit '.reloc R_MIPS_JALR' (R_MICROMIPS_JALR
for micromips). The linker might then be able to turn jalr into a direct
call.
Add '-mips-jalr-reloc' to enable/disable this feature (default is true).
Differential revision: https://reviews.llvm.org/D55292
llvm-svn: 348760
This triggers an assert when combining concat_vectors of a bitcast of
merge_values.
With asserts disabled, it fails to select:
fatal error: error in backend: Cannot select: 0x7ff19d000e90: i32 = any_extend 0x7ff19d000ae8
0x7ff19d000ae8: f64,ch = CopyFromReg 0x7ff19d000c20:1, Register:f64 %1
0x7ff19d000b50: f64 = Register %1
In function: d
Differential Revision: https://reviews.llvm.org/D55507
llvm-svn: 348759
Summary:
Shuffle on 64bit data is allowed only for CUDA >= 9.0. Also, fixed the
constant for the mask, need one extra L in the end.
Reviewers: gtbercea, kkwli0
Subscribers: guansong, caomhin, openmp-commits
Differential Revision: https://reviews.llvm.org/D55440
llvm-svn: 348758
The list generated in the target parser tests is the
same as the one in the AArch64 target parser.
Use that one instead.
Differential Revision: https://reviews.llvm.org/D55509
llvm-svn: 348757
A new pass to manage the Mode register.
Currently this just manages the floating point double precision
rounding requirements, but is intended to be easily extended to
encompass all Mode register settings.
The immediate motivation comes from the requirement to use the
round-to-zero rounding mode for the 16 bit interpolation
instructions, where the rounding mode setting is shared between
16 and 64 bit operations.
llvm-svn: 348754
Currently, dbg.value's of "nullptr" are dropped when entering a SelectionDAG --
apparently just because of an oversight when recognising Values that are
constant (see PR39787). This patch adds ConstantPointerNull to the list of
constants that can be turned into DBG_VALUEs.
The matter of what bit-value a null pointer constant in LLVM has was raised
in this mailing list thread:
http://lists.llvm.org/pipermail/llvm-dev/2018-December/128234.html
Where it transpires LLVM relies on (IR) null pointers being zero valued,
thus I've baked this assumption into the patch.
Differential Revision: https://reviews.llvm.org/D55227
llvm-svn: 348753
Summary:
If a function argument is byval and RV is located in default or alloca address space
an optimization of creating addrspacecast instead of memcpy is performed. That is
not correct for OpenCL, where that can lead to a situation of address space casting
from __private * to __global *. See an example below:
```
typedef struct {
int x;
} MyStruct;
void foo(MyStruct val) {}
kernel void KernelOneMember(__global MyStruct* x) {
foo (*x);
}
```
for this code clang generated following IR:
...
%0 = load %struct.MyStruct addrspace(1)*, %struct.MyStruct addrspace(1)**
%x.addr, align 4
%1 = addrspacecast %struct.MyStruct addrspace(1)* %0 to %struct.MyStruct*
...
So the optimization was disallowed for OpenCL if RV is located in an address space
different than that of the argument (0).
Reviewers: yaxunl, Anastasia
Reviewed By: Anastasia
Subscribers: cfe-commits, asavonic
Differential Revision: https://reviews.llvm.org/D54947
llvm-svn: 348752
This is a fix for PR39896, where dbg.value's of SDNodes that have been
optimised out do not lead to "DBG_VALUE undef" instructions being created.
Such undef instructions are necessary to terminate earlier variable
ranges, otherwise variable values leak past the point where they're valid.
The "invalidated" flag of SDDbgValue is currently being abused to mean two
things:
* The corresponding SDNode is now invalid
* This SDDbgValue should not be emitted
Of which there are several legitimate combinations of meaning:
* The SDNode has been invalidated and we should emit "DBG_VALUE undef"
* The SDNode has been invalidated but the debug data was salvaged, don't
emit anything for this SDDbgValue
* This SDDbgValue has been emitted
This patch introduces distinct "Emitted" and "Invalidated" fields to the
SDDbgValue class, updates users accordingly, and generates "undef"
DBG_VALUEs for invalidated records. Awkwardly, there are circumstances
where we emit SDDbgValue's twice, specifically DebugInfo/X86/dbg-addr-dse.ll
which I've preserved.
Differential Revision: https://reviews.llvm.org/D55372
llvm-svn: 348751
Fixes https://bugs.llvm.org/show_bug.cgi?id=39926.
The size of the first copy was computed as
std::abs(std::abs(LdDisp2) - std::abs(LdDisp1)), which results in
skipped bytes if the signs of LdDisp2 and LdDisp1 differ. As far as
I can see, this should just be LdDisp2 - LdDisp1. The case where
LdDisp1 > LdDisp2 is already handled in the code above, in which case
LdDisp2 is set to LdDisp1 and this subtraction will evaluate to
Size1 = 0, which is the correct value to skip an overlapping copy.
Differential Revision: https://reviews.llvm.org/D55485
llvm-svn: 348750
Summary:
Add new interceptors for a set of functions to reposition a stream:
fgetpos, fseek, fseeko, fsetpos, ftell, ftello, rewind .
Add a dedicated test.
Enable this interface on NetBSD.
Reviewers: joerg, vitalybuka
Reviewed By: vitalybuka
Subscribers: kubamracek, llvm-commits, mgorny, #sanitizers
Tags: #sanitizers
Differential Revision: https://reviews.llvm.org/D55471
llvm-svn: 348743
Summary: This should avoid failing on old CPUs that do not have a cycle counter.
Subscribers: tschuett, llvm-commits
Differential Revision: https://reviews.llvm.org/D55416
llvm-svn: 348740
Summary: The comment says we need 3 extracts and a select at the end. But didn't we just account for the select in the vector cost above. Aren't we just extracting the single element after taking the min/max in the vector register?
Reviewers: RKSimon, spatel, ABataev
Reviewed By: RKSimon
Subscribers: javed.absar, kristof.beyls, llvm-commits
Differential Revision: https://reviews.llvm.org/D55480
llvm-svn: 348739
The addcarry and addcarryx builtins do the same thing. The only difference is that addcarryx previously required adx feature.
This commit removes the adx feature check from addcarryx and removes the addcarry builtin. This matches the builtins that gcc has. We don't guarantee compatibility in builtins, but we generally try to be consistent if its not a burden.
llvm-svn: 348738
Both intrinsics do the exact same thing so we really only need one.
Earlier in the 8.0 cycle we changed the signature of this intrinsic without renaming it. But it looks difficult to get the autoupgrade code to allow me to merge the intrinsics and change the signature at the same time. So I've renamed the intrinsic slightly for the new merged intrinsic. I'm skipping autoupgrading from the previous new to 8.0 signature. I've also renamed the subborrow for consistency.
llvm-svn: 348737
Since TBEHandler doesn't maintain state or otherwise have any need to be
a class right now, the read and write functions have been moved out and
turned into standalone functions. Additionally, the TBE read function
has been updated to return an Expected value for better error handling.
Tests have been updated to reflect these changes.
Differential Revision: https://reviews.llvm.org/D55450
llvm-svn: 348735
Summary:
When bugpoint attempts to find the other executables it needs to run,
such as `opt` or `clang`, it tries searching the user's PATH. However,
in many cases, the 'bugpoint' executable is part of an LLVM build, and
the 'opt' executable it's looking for is in that same directory.
Many LLVM tools handle this case by using the `Paths` parameter of
`llvm::sys::findProgramByName`, passing the parent path of the currently
running executable. Do this same thing for bugpoint. However, to
preserve the current behavior exactly, first search the user's PATH,
and then search for 'opt' in the directory containing 'bugpoint'.
Test Plan:
`check-llvm`. Many of the existing bugpoint tests no longer need to use the
`--opt-command` option as a result of these changes.
Reviewers: MatzeB, silvas, davide
Reviewed By: MatzeB, davide
Subscribers: davide, llvm-commits
Differential Revision: https://reviews.llvm.org/D54884
llvm-svn: 348734
Jonathan Peyton