Subscribers: lldb-commits

Differential Revision: https://reviews.llvm.org/D48855

llvm-svn: 336149

Summary:
In the individual index files emitted for distributed ThinLTO backends,
the module path ids are not contiguous. Assign slots to module paths in
order to handle this better and also to get contiguous numbering in the
summary assembly.

Reviewers: davidxl, dexonsmith

Subscribers: mehdi_amini, inglorion, eraman, llvm-commits, steven_wu

Differential Revision: https://reviews.llvm.org/D48698

llvm-svn: 336148

Summary:
1) When ReadRegister is called with a null register into on Windows, rather than crashing due to an access violation, simply return false. Not all registers and properties will be read or calculated correctly, but that is consistent with other platforms that also return false in that case
2) Update a couple of tests to reference pr37995 as their reason for failure since it is much more accurate. Support for floating point registers doesn't exist on Windows at all, rather than having issues.

Reviewers: asmith, labath, zturner

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D48844

llvm-svn: 336147

Summary:
This patch refactors the internal completion API. It now takes (as far as possible) a single
CompletionRequest object instead o half a dozen in/out/in-out parameters. The CompletionRequest
contains a common superset of the different parameters as far as it makes sense. This includes
the raw command line string and raw cursor position, which should make the `expr` command
possible to implement (at least without hacks that reconstruct the command line from the args).

This patch is not intended to change the observable behavior of lldb in any way. It's also as
minimal as possible and doesn't attempt to fix all the problems the API has.

Some Q&A:

Q: Why is this not fixing all the problems in the completion API?
A: Because is a blocker for the expr command completion which I want to get in ASAP. This is the
smallest patch that unblocks the expr completion patch and which allows trivial refactoring in the future.
The patch also doesn't really change the internal information flow in the API, so that hopefully
saves us from ever having to revert and resubmit this humongous patch.

Q: Can we merge all the copy-pasted code in the completion methods
(like computing the current incomplete arg) into CompletionRequest class?
A: Yes, but it's out of scope for this patch.

Q: Why the `word_complete = request.GetWordComplete(); ... ` pattern?
A: I don't want to add a getter that returns a reference to the internal integer. So we have
to use a temporary variable and the Getter/Setter instead. We don't throw exceptions
from what I can tell, so the behavior doesn't change.

Q: Why are we not owning the list of matches?
A: Because that's how the previous API works. But that should be fixed too (in another patch).

Q: Can we make the constructor simpler and compute some of the values from the plain command?
A: I think this works, but I rather want to have this in a follow up commit. Especially when making nested
request it's a bit awkward that the parsed arguments behave as both input/output (as we should in theory
propagate the changes on the nested request back to the parent request if we don't want to change the
behavior too much).

Q: Can't we pass one const request object and then just return another result object instead of mixing
them together in one in/out parameter?
A: It's hard to get keep the same behavior with that pattern, but I think we can also get a nice API with just
a single request object. If we make all input parameters read-only, we have a clear separation between what
is actually an input and what an output parameter (and hopefully we get rid of the in-out parameters).

Q: Can we throw out the 'match' variables that are not implemented according to the comment?
A: We currently just forward them as in the old code to the different methods, even though I think
they are really not used. We can easily remove and readd them once every single completion method just
takes a CompletionRequest, but for now I prefer NFC behavior from the perspective of the API user.

Reviewers: davide, jingham, labath

Reviewed By: jingham

Subscribers: mgorny, friss, lldb-commits

Differential Revision: https://reviews.llvm.org/D48796

llvm-svn: 336146

Summary: Add support for atomic store instructions.

Reviewers: dschuff

Subscribers: sbc100, jgravelle-google, sunfish, llvm-commits

Differential Revision: https://reviews.llvm.org/D48839

llvm-svn: 336145

Reviewers: efriedma, rogfer01, javed.absar

Reviewed By: efriedma, rogfer01

Subscribers: kristof.beyls, chrib, llvm-commits

Differential Revision: https://reviews.llvm.org/D48846

llvm-svn: 336144

Differential Revision: https://reviews.llvm.org/D48812

llvm-svn: 336143

[llvm-mca] Clear the content of map VariantDescriptors in InstrBuilder before we start analyzing a new CodeBlock. NFCI.

Different CodeBlocks don't overlap. The same MCInst cannot appear in more than
one code block because all blocks are instantiated before the simulation is run.

We should always clear the content of map VariantDescriptors before every
simulation, since VariantDescriptors cannot possibly store useful information
for the next blocks. It is also "safer" to clear its content because `MCInst*`
is used as the key type for map VariantDescriptors.

llvm-svn: 336142

Summary:
_is_chartype_l (needed for isxdigit_l) in MinGW compares locale_t and NULL.
NULL is 'long long' for 64-bit, and this results in ambiguous overloads when
compiled with Clang.  Define a concrete overload for the operators to fix the
ambiguity.

Reviewers: mstorsjo, EricWF, srhines, danalbert

Subscribers: christof, cfe-commits, ldionne

Differential Revision: https://reviews.llvm.org/D48749

llvm-svn: 336141

Summary:
Comment on Transforms/LoopVersioning/incorrect-phi.ll: With the change
SCEV is able to prove that the loop doesn't wrap-self (due to zext i16
to i64), disabling the entire loop versioning pass. Removed the zext and
just use i64.

Reviewers: sanjoy

Subscribers: jlebar, hiraditya, javed.absar, bixia, llvm-commits

Differential Revision: https://reviews.llvm.org/D48409

llvm-svn: 336140

Summary:
    It would be useful for Flutter apps, especially, to be able to use
    malloc hooks to debug memory leaks on Fuchsia. They're not able to do
    this right now, so it'd be a nice bonus to throw in with the Scudo
    switchover.


Reviewers: cryptoad, alekseyshl

Reviewed By: cryptoad

Differential Revision: https://reviews.llvm.org/D48618

llvm-svn: 336139

LLVM doesn't guarantee anything about the high bits of a register holding
an i1 value at the IR level, so don't translate LLVM IR i1 values directly
into WebAssembly conditional branch operands. WebAssembly's conditional
branches do demand all 32 bits be valid.

Fixes PR38019.

llvm-svn: 336138

Summary:
Emmiting new intrinsic that strips invariant.groups to make
devirtulization sound, as described in RFC: Devirtualization v2.

Reviewers: rjmccall, rsmith, amharc, kuhar

Subscribers: llvm-commits, cfe-commits

Differential Revision: https://reviews.llvm.org/D47299



Co-authored-by: Krzysztof Pszeniczny <krzysztof.pszeniczny@gmail.com>
llvm-svn: 336137

llvm-svn: 336136

llvm-svn: 336135

Add registers still missing after r328016 (D43353):
- for bits 15-8  of SI, DI, BP, SP (*H), and R8-R15 (*BH),
- for bits 31-16 of R8-R15 (*WH).

Thanks to Craig Topper for pointing it out.

llvm-svn: 336134

llvm-svn: 336133

llvm-svn: 336132

Similar to rLLD336129

llvm-svn: 336131

Summary: It is common to have the following min/max pattern during the intermediate stages of SLP since we only optimize at the end. This patch tries to catch such patterns and allow more vectorization.

         %1 = extractelement <2 x i32> %a, i32 0
         %2 = extractelement <2 x i32> %a, i32 1
         %cond = icmp sgt i32 %1, %2
         %3 = extractelement <2 x i32> %a, i32 0
         %4 = extractelement <2 x i32> %a, i32 1
         %select = select i1 %cond, i32 %3, i32 %4

Author: FarhanaAleen

Reviewed By: ABataev, RKSimon, spatel

Differential Revision: https://reviews.llvm.org/D47608

llvm-svn: 336130

Post commit review at rLLD335992

llvm-svn: 336129

This extends D48485 to allow another pair of binops (add/or) to be combined either
with or without a leading shuffle:
or X, C --> add X, C (when X and C have no common bits set)

Here, we need value tracking to determine that the 'or' can be reversed into an 'add',
and we've added general infrastructure to allow extending to other opcodes or moving 
to where other passes could use that functionality.

Differential Revision: https://reviews.llvm.org/D48662

llvm-svn: 336128

On darwin, all virtual sections have zerofill type, and having a
.zerofill directive in a non-virtual section is not allowed. Instead of
asserting, show a nicer error.

In order to use the equivalent of .zerofill in a non-virtual section,
the usage of .zero of .space is required.

This patch replaces the assert with an error.

Differential Revision: https://reviews.llvm.org/D48517

llvm-svn: 336127

Summary:
This adds a new -no-weak flag to nm to hide weak symbols in its output.
This also adds a -W alias for this which is analogous to -U.

Patch by Keith Smiley

Reviewers: kastiglione, enderby, compnerd

Reviewed By: kastiglione

Subscribers: llvm-commits

Differential Revision: https://reviews.llvm.org/D48751

llvm-svn: 336126

Alternate opcode handling only supports binary operators, these tests demonstrate a missed opportunity to vectorize ceil/floor calls

llvm-svn: 336125

llvm-svn: 336124

Use an -implicit-check-not to make sure an error which should not occur
in fact does not occur before the first CHECK line.

Suggested by Paul Robinson in post-commit feedback for r335897.

llvm-svn: 336123

Add cost tests for fp ceil, floor, nearbyint, rint and trunc.

llvm-svn: 336122

Similarily, don't fold fp128 loads into SSE instructions if the load isn't aligned. Unless we're targeting an AMD CPU that doesn't check alignment on arithmetic instructions.

Should fix PR38001

llvm-svn: 336121

We currently don't any-extend vararg parameters before storing them to the stack
locations on Darwin. However, SelectionDAG however does this, and so user code
is in the wild which inadvertently relies on this extension. This can manifest
in cases where the value stored is (int)0, but the actual parameter is interpreted
by va_arg as a pointer, and so not extending to 64 bits causes the callee to
load additional undefined bits.

llvm-svn: 336120

Summary:
This allows hovering on keywords that refer to deduced types.
This should cover most useful cases. Not covered:
- auto template parameters: Since this can be instantiated with many types,
it would not be practical to show the types.
- Structured binding: This could be done later to show multiple deduced types
in the hover.
- auto:: (part of concepts): Outside the scope of this patch.

Signed-off-by: Marc-Andre Laperle <marc-andre.laperle@ericsson.com>

Subscribers: ilya-biryukov, ioeric, MaskRay, jkorous, cfe-commits

Differential Revision: https://reviews.llvm.org/D48159

llvm-svn: 336119

Subscribers: dschuff, mehdi_amini, inglorion, jgravelle-google, aheejin, sunfish, steven_wu, llvm-commits

Differential Revision: https://reviews.llvm.org/D48689

llvm-svn: 336118

Temporary revert because of a failing test on some buildbots.

This reverts commit r336114.

llvm-svn: 336117

Differential Revision: https://reviews.llvm.org/D48748

llvm-svn: 336116

Imply dotprod for armv8.4-a, because it is mandatory from v8.4.

llvm-svn: 336115

Summary:
This patch is the first in a series of patches related to the [[ http://lists.llvm.org/pipermail/llvm-dev/2018-June/123883.html | RFC - A new dominator tree updater for LLVM ]].

This patch introduces the DomTreeUpdater class, which provides a cleaner API to perform updates on available dominator trees (none, only DomTree, only PostDomTree, both) using different update strategies (eagerly or lazily) to simplify the updating process.

—Prior to the patch—

   - Directly calling update functions of DominatorTree updates the data structure eagerly while DeferredDominance does updates lazily.
   - DeferredDominance class cannot be used when a PostDominatorTree also needs to be updated.
   - Functions receiving DT/DDT need to branch a lot which is currently necessary.
   - Functions using both DomTree and PostDomTree need to call the update function separately on both trees.
   - People need to construct an additional DeferredDominance class to use functions only receiving DDT.

—After the patch—

Patch by Chijun Sima <simachijun@gmail.com>.

Reviewers: kuhar, brzycki, dmgreen, grosser, davide

Reviewed By: kuhar, brzycki

Subscribers: vsk, mgorny, llvm-commits

Author: NutshellySima

Differential Revision: https://reviews.llvm.org/D48383

llvm-svn: 336114

We were only doing this for basic blends, despite shuffle lowering now being good enough to handle more complex blends. This means that the two v8i16 splat shifts are performed in parallel instead of serially as the general shift case.

llvm-svn: 336113

We have special case support for 2 shift values for basic blends, but irregular shift patterns end up using the generic lowering, despite shuffle lowering being good enough to handle more complex blends.

llvm-svn: 336112

llvm-svn: 336111

llvm-svn: 336110