There's a silent bug in our reasoning about zero strides. We assume that having a single static exit implies that if that exit is not taken, then the loop must be infinite. This ignores the potential for abnormal exits via exceptions. Consider the following example:

for (uint_8 i = 0; i < 1; i += 0) {
  throw_on_thousandth_call();
}

Our reasoning is such that we'd conclude this loop can't take the backedge as that would lead to a (presumed) infinite loop.

In practice, this is a silent bug because the loopIsFiniteByAssumption returns false strictly more often than the loopHaNoAbnormalExits property. We could reasonable want to change that in the future, so fixing the codeflow now is worthwhile.

Differential Revision: https://reviews.llvm.org/D109029

Currently, the truncate selection dag node is expanded as a bitwise AND plus compare to 1.  This change enables scalar comparison in the pattern if the truncate node is uniform.

Reviewed By: rampitec

Differential Revision: https://reviews.llvm.org/D108925

As mentioned in D108833, the logic for figuring out if a backedge is dead was somewhat interwoven with the SCEV based logic and the symbolic eval logic. This is my attempt at making the code easier to follow.

Note that this is only NFC after the work done in 29fa37ec.  Thanks to Nikita for catching that case.

Differential Revision: https://reviews.llvm.org/D108848

The commandline flag to specify a particular openmp devicertl library
currently errors like:
```
fatal error: cannot open file
      './runtimes/runtimes-bins/openmp/libomptarget':
      Is a directory
```
CommonArgs successfully appends the directory to the commandline args then
mlink-builtin-bitcode rejects it.

This patch is a point fix to that. If --libomptarget-amdgcn-bc-path=directory
then append the expected name for the current architecture and go on as before.
This is useful for test runners that don't hardcode the architecture.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D109057

With opaque pointers, no actual bitcasts will be present. Instead,
there will be a mismatch between the call FunctionType and the
function ValueType. Change the code to collect CallBases
specifically (rather than general Uses) and compare these types.

RAUW is no longer performed, as there would no longer be any
bitcasts that can be RAUWd.

Differential Revision: https://reviews.llvm.org/D108880

Several instructions with potential TLS relocations were missing.  This
issue was found when building the Canadian LLVM toolchain.

This patch is to add Image Operands in SPIR-V Dialect and also let ImageDrefGather to use Image Operands.

Image Operands are used in many image instructions. "Image Operands encodes what oprands follow, as per Image Operands". And ususally, they are optional to image instructions.

The format of image operands looks like:

    %0 = spv.ImageXXXX %1, ... %3 : f32 ["Bias|Lod"](%4, %5 : f32, f32) -> ...

This patch doesn’t implement all operands (see Section 3.14 in SPIR-V Spec) but provides a skeleton of it. There is TODO in verifyImageOperands function.

Co-authored: Alan Liu <alanliu.yf@gmail.com>

Reviewed by: antiagainst

Differential Revision: https://reviews.llvm.org/D108501

We'd special cased this logic to use pointer types for non-integral pointers, but there's no reason we can't do that for all pointer types.   Doing it this was has a few advantages:
a) The code itself becomes more straight forward, and easier to test.
b) We avoid introducing ptrtoint into programs which didn't have them in the source.
c) The resulting codegen is easier to analyze and simplify (mostly due to lack of ptrtoint).

Note that there are some test diffs, but a) running them through instcombine helps a ton, and b) there's enough missing obvious transforms on both before and after IR that it's clear this isn't performance sensitive.

This is mostly motivated by cleaning up mentions of non-integrals to have a clearer idea of what we actually need to support.

Differential Revision: https://reviews.llvm.org/D104662

Document the CSR AGPRs for GFX90A.

Remove the TODO for gfx908, as the answer is that we don't mark any
AGPRs as callee-saved except for GFX90A, i.e. the docs as-is are correct
for gfx908.

Reviewed By: rampitec

Differential Revision: https://reviews.llvm.org/D109009

Document 'use-external-name' and the various bits of logic that make it
work, to avoid others having to repeat the archival work (given that I
added getFileRefReturnsCorrectNameForDifferentStatPath to
FileManagerTest, seems possible I understood this once before!).

- b59cf679 added 'use-external-name' to
  RedirectingFileSystem. This causes `stat`s to return the external
  name for a redirected file instead of the name it was accessed by,
  leaking it through the VFS.
- d066d4c8 propagated the external name
  further through clang::FileManager.
- 4dc5573a, which added
  clang::FileEntryRef to clang::FileManager, has complicated concession
  to account for this as well (since refactored a bit).

The goal of 'use-external-name' is to enable Clang to report "real" file
paths to users (via diagnostics) and to external tools (such as
debuggers reading debug info and build systems reading `.d` files).

I've added FIXMEs to look at other channels for communicating the
external names, since the current implementation adds complexity to
FileManager and exposes an inconsistent interface to clients.

Besides that, the FileManager logic appears to be kicking in outside of
'use-external-name'. Seems that *some* vfs::FileSystem implementations
canonicalize some paths returned by `stat` in *some* cases (the bug
isn't fully understood yet). Volodymyr Sapsai is investigating, this at
least better documents what *is* understood.

Windows separator problem. Fixing that broke another regex.
This reverts commit 0173e024.

This reverts commit 7a228f87.
Failing test case under CI

This patch adds a skeleton as a preparatory step for the next patch which
adds the actual implementations of the condition variable functions.

Reviewed By: michaelrj

Differential Revision: https://reviews.llvm.org/D108947

If the true and false values are the same, we don't need a SELECT_CC.

This would normally be folded before a select is legalized to
select_cc. The test case exploits the late legalization of vscale
to trigger a case where they become identical after legalization.

This works around an issue found on a test case in D107957. In that
case the true/false values were both eventually 0 and the select was
used by a vector AVL operand. The select_cc got expanded to control
flow and a phi, but the phi inputs were both copies from X0. MachineIR
optimizations simplified this to a single copy from X0 going into the
vector instruction. This became the input of a vsetvli after vsetvli
insertion. Then register coalescing folded the copy into the vsetvli.
X0 as the source of a vsetvli is a special encoding and should not be
created by coalesing. We need to fix our vsetvli handling to make sure
this can never happen any other way, but removing the unneeded select
is still a worthwhile optimization.

[ubsan] warn inside the sigaction interceptor if static linking is suspected, and continue instead of crashing on null deref

[ubsan] warn inside the sigaction interceptor if static linking is suspected, and continue instead of crashing on null deref

Reviewed By: kostik

Differential Revision: https://reviews.llvm.org/D109081

Given D109057, change test runner to use the libomptarget-x-bc-path
argument instead of the LIBRARY_PATH environment variable to find the device
library.

Also drop the use of LIBRARY_PATH environment variable as it is far
too easy to pull in the device library from an unrelated toolchain by accident
with the current setup. No loss in flexibility to developers as the clang
commandline used here is still available.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D109061

With the context split work, the context-based (an array of strings) sorting performed at profile load time is way more expansive than single-string-based sorting. This is likely due to auxiliary operations done on each array element, such as indirect references, std::min operations, also likely cache misses. In this change I'm presorting profiles during profile generation time to avoid sorting at compile time.

Compared to the previous context-split work, this effectively cuts down compile time by 20% for one of our large services and brings us closer to non-CS build, with still a small gap in build time.

Reviewed By: wenlei, wmi

Differential Revision: https://reviews.llvm.org/D109036

Store the used element type in the InductionDescriptor. For typed
pointers, it remains the pointer element type. For opaque pointers,
we always use an i8 element type, such that the step is a simple
offset.

A previous version of this patch instead tried to guess the element
type from an induction GEP, but this is not reliable, as the GEP
may be hidden (see @both in iv_outside_user.ll).

Differential Revision: https://reviews.llvm.org/D104795

Reviewed By: michaelrj

Differential Revision: https://reviews.llvm.org/D109045

In D104421, we changed the API for pass registration.
Before you would write:

      void registerPass("my-pass", "My Pass Description.",
                        [] { return createMyPass(); });
while now you’d only write:

      void registerPass([] { return createMyPass(); });

If you’re using TableGen to define your pass registration, you shouldn’t have anything to do. If you’re using directly the C++ API here are some changes.
Your project may also be broken even if you use TableGen and you call the
generated registration API in case your pass implementation didn’t inherit from
the MyPassBase class generated by TableGen.

If you don't use TableGen, the "my-pass" and "My Pass Description." fields must
be provided by overriding methods on the pass itself:

  llvm::StringRef getArgument() const final { return "my-pass"; }
  llvm::StringRef getDescription() const final {
    return "My Pass Description.";
  }

Reviewed By: rriddle

Differential Revision: https://reviews.llvm.org/D104429

This extends D108921 into a generic rule applied to constructing ExitLimits along all paths. The remaining paths (primarily howFarToZero) don't have the same reasoning about UB sensitivity as the howManyLessThan ones did. Instead, the remain cause for max counts being more precise than exact counts is that we apply context sensitive loop guards on the max path, and not on the exact path. That choice is mildly suspect, but out of scope of this patch.

The MVETailPredication.cpp change deserves a bit of explanation. We were previously figuring out that two SCEVs happened to be equal because the happened to be identical. When we optimized one with context sensitive information, but not the other, we lost the ability to prove them equal. So, cover this case by subtracting and then applying loop guards again. Without this, we see changes in test/CodeGen/Thumb2/mve-blockplacement.ll

Differential Revision: https://reviews.llvm.org/D109015

In https://reviews.llvm.org/D47480 I complained that there's no positive
form of this flag, so let's add one :)

https://gcc.gnu.org/PR29931 also has a pending patch to add the positive
form to gcc (but there's admittedly not a lot of movement on that bug).

This doesn't change any defaults.

Differential Revision: https://reviews.llvm.org/D108818

This is used by BOLT to do patching of DebugInfo section, and Line Table. Directly by using find, and through getAttrFieldOffsetForUnit.

Reviewed By: dblaikie

Differential Revision: https://reviews.llvm.org/D107874

The commandline flag to specify a particular openmp devicertl library
currently errors like:
```
fatal error: cannot open file
      './runtimes/runtimes-bins/openmp/libomptarget':
      Is a directory
```
CommonArgs successfully appends the directory to the commandline args then
mlink-builtin-bitcode rejects it.

This patch is a point fix to that. If --libomptarget-amdgcn-bc-path=directory
then append the expected name for the current architecture and go on as before.
This is useful for test runners that don't hardcode the architecture.

Reviewed By: jdoerfert

Differential Revision: https://reviews.llvm.org/D109057

isFreeToInvert allows min/max with 'not' on both operands,
so easing the argument restriction catches the case where
that operand has one use.

We already handle the sub-patterns when there are less uses:
https://alive2.llvm.org/ce/z/8Jatm_

...but this is another step towards parity with the
equivalent icmp+select idioms ( D98152 ).

Differential Revision: https://reviews.llvm.org/D109059

This mimics the code for the corresponding cmp-select idiom.

This also prevents an infinite loop because isFreeToInvert
does not match constant expressions.

So this patch solves the same problem as D108814 and obsoletes
it, but my main motivation is to enhance the pattern matching
to allow more invertible ops. That change will be a follow-up
patch on top of this one.

Differential Revision: https://reviews.llvm.org/D109058

Introduced in D108788. These are clearer.

The OutermostLoad condition is supposed to strip the outermost
DW_OP_deref operation because dbg.declares are implicitly
indirect. This patch makes sure the heuristic is only applied to
dbg.declare intrinsics and only if the outermost instruction is a
load.

This was found while qualifying the latest Swift compiler rebranch.

rdar://82037764

See description in test.

Reviewed By: frasercrmck

Differential Revision: https://reviews.llvm.org/D108965

Using rpath instead of LD_LIBRARY_PATH to find libomp.so and
libomptarget.so lets one rerun the already built test executables without
setting environment variables and removes the risk of the test runner picking
up different libraries to the developer debugging the failure.

rpath usually means runpath, which is not transitive, so set runpath on
libomptarget itself so that it can find the plugins located next to it,
spelled $ORIGIN. This provides sufficient functionality to drop D102043

Reviewed By: tianshilei1992

Differential Revision: https://reviews.llvm.org/D109071

libdevice bitcode provided by NVIDIA is linked with clang/LLVM-generated IR
which uses nvptx*-nvidia-cuda triple. We need to mark them as compatible.

Differential Revision: https://reviews.llvm.org/D108835

This implements the struct `__format_arg_store` and its dependencies:
* the class basic_format_arg,
* the class basic_format_args,
* the class basic_format_context,
* the function make_format_args,
* the function wmake_format_args,
* the function visit_format_arg,
* several Standard required typedefs.

The following parts will be implemented in a later patch:

* the child class `basic_format_arg::handle`,
* the function `basic_format_arg::basic_format_arg(const T* p)`.

The following extension has been implemented:
* the class basic_format_arg supports `__[u]int128_t` on platform where libc++ supports 128 bit integrals.

Implements parts of:
* P0645 Text Formatting

Completes:
* LWG3371 visit_format_arg and make_format_args are not hidden friends
* LWG3542 basic_format_arg mishandles basic_string_view with custom traits

Note https://mordante.github.io/blog/2021/06/05/format.html gives a bit more information about the goals and non-goals of this initial patch serie...

This is part one of a couple of patches to fully rename these methods.

I've made the mistake of assuming that these indexes are for parameters
multiple times, but actually they're based off of a weird indexing
scheme AttributeList::AttrIndex where 0 is the return value and ~0 is
the function. Hopefully renaming these methods will make this clearer.
Ideally users should use more specific methods like
AttributeList::getFnAttr().

This patch simply adds the name that we want in the end. This is so the
removal of the methods with the original names happens in a separate
change to make it easier for downstream users.

This touches all relevant methods in AttributeList, CallBase, and Function.

Reviewed By: rnk

Differential Revision: https://reviews.llvm.org/D108788

Previously extra wide v4f32 to v4f64 extending loads would be legalized to v2f32
to v2f64 extending loads, which would then be scalarized by legalization. (v2f32
to v2f64 extending loads not produced by legalization were already being emitted
correctly.) Instead, mark v2f32 to v2f64 extending loads as legal and explicitly
lower them using promote_low. This regresses the addressing modes supported for
the extloads not produced by legalization, but that's a fine trade off for now.

Differential Revision: https://reviews.llvm.org/D108496

std::vector pointers can be invalided while growing. Using std::list instead.

When we have an any-extending FPR bank load, none of the tablegen patterns
match and we fall back to the C++ selector. Like with the truncating stores
that were fixed recently, the C++ wasn't able to handle it and ended up
generating invalid copies between different size regclasses.

This change adds handling for this case, splitting the load into a regular
load and a SUBREG_TO_REG to extend it into the original wide destination reg.