Subscribers: ilya-biryukov, MaskRay, jkorous, arphaman, usaxena95, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D78359

Previous patch didn't handle the early return in `emitREXPrefix` correctly,
which causes REX prefix was not emitted for instruction without
operands. This patch includes the fix for that.

This does for G_EXTRACT_VECTOR_ELT what 588bd7be did for G_TRUNC.

Ideally types without a corresponding register class wouldn't reach
here, but we're currently missing some (in particular a 192-bit class
is missing).

These functions don't really do any changes to loop info or
dominator tree. We should state this explicitly using 'const'.

This allows targets to know exactly which operands are contributing to
the dependency, which is required for targets with per-operand
scheduling models.

Differential Revision: https://reviews.llvm.org/D77135

MSVC projects were missing the llvm/Transforms/Vectorize/* headers

Push these includes down to the the writers that actually need them, a number of which were implicitly relying on the MCObjectWriter.h.

Remove unnecessary SmallPtrSet.h/SectionKind.h includes
Remove unused MCAssembler/raw_ostream forward declarations

There is no need to use `--check-prefix` multiple times.
It helps to improve readability/test maintainability.
This patch does it for all tools at once.

Differential revision: https://reviews.llvm.org/D78217

Summary:
This patch remove waymarking and replaces it with storing a pointer to the User in the Use.
here are the results on the measurements for the CTMark tests of the test suite.
```
Metric: instructions_count

Program                                                      baseline      patched       diff
 test-suite :: CTMark/ClamAV/clamscan.test                    72557942065   71733653521  -1.1%
 test-suite :: CTMark/sqlite3/sqlite3.test                    76281422939   75484840636  -1.0%
 test-suite :: CTMark/consumer-typeset/consumer-typeset.test  51364676366   50862185614  -1.0%
 test-suite :: CTMark/SPASS/SPASS.test                        60476106505   59908437767  -0.9%
 test-suite :: CTMark/tramp3d-v4/tramp3d-v4.test              112578442329  111725050856 -0.8%
 test-suite :: CTMark/mafft/pairlocalalign.test               50846133013   50473644539  -0.7%
 test-suite :: CTMark/kimwitu++/kc.test                       54692641250   54349070299  -0.6%
 test-suite :: CTMark/7zip/7zip-benchmark.test                182216614747  181216091230 -0.5%
 test-suite :: CTMark/Bullet/bullet.test                      123459210616  122905866767 -0.4%
 Geomean difference                                                                      -0.8%

Metric: peak_memory_use

Program                                                      baseline  patched   diff
 test-suite :: CTMark/tramp3d-v4/tramp3d-v4.test              326864    338524    3.6%
 test-suite :: CTMark/sqlite3/sqlite3.test                    216412    221240    2.2%
 test-suite :: CTMark/7zip/7zip-benchmark.test                11808284  12022604  1.8%
 test-suite :: CTMark/Bullet/bullet.test                      6831752   6945988   1.7%
 test-suite :: CTMark/SPASS/SPASS.test                        2682552   2721820   1.5%
 test-suite :: CTMark/ClamAV/clamscan.test                    5037256   5107936   1.4%
 test-suite :: CTMark/consumer-typeset/consumer-typeset.test  2752728   2790768   1.4%
 test-suite :: CTMark/mafft/pairlocalalign.test               1517676   1537244   1.3%
 test-suite :: CTMark/kimwitu++/kc.test                       1090748   1103448   1.2%
 Geomean difference                                                               1.8%

Metric: compile_time

Program                                                      baseline patched diff
 test-suite :: CTMark/consumer-typeset/consumer-typeset.test  14.71    14.38  -2.2%
 test-suite :: CTMark/sqlite3/sqlite3.test                    23.18    22.73  -2.0%
 test-suite :: CTMark/7zip/7zip-benchmark.test                57.96    56.99  -1.7%
 test-suite :: CTMark/ClamAV/clamscan.test                    20.75    20.49  -1.2%
 test-suite :: CTMark/kimwitu++/kc.test                       18.35    18.15  -1.1%
 test-suite :: CTMark/SPASS/SPASS.test                        18.72    18.57  -0.8%
 test-suite :: CTMark/mafft/pairlocalalign.test               14.09    14.00  -0.6%
 test-suite :: CTMark/Bullet/bullet.test                      37.38    37.19  -0.5%
 test-suite :: CTMark/tramp3d-v4/tramp3d-v4.test              33.81    33.76  -0.2%
 Geomean difference                                                           -1.1%
```

i believe that it is worth trading +1.8% peak memory use for -1.1% compile time.
also this patch removes waymarking which simplifies the Use and User classes.

Reviewers: nikic, lattner

Reviewed By: lattner

Subscribers: russell.gallop, foad, ggreif, rriddle, ekatz, fhahn, lebedev.ri, mgorny, hiraditya, george.burgess.iv, asbirlea, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D77144

Summary:
Currently, cl::ConsumeAfter only works for the case that has exactly one
positional argument. Without the fix, it skip fulfilling first positional
argument and put that additional positional argument in interpreter arguments.

Reviewers: bkramer, Mordante, rnk, lattner, beanz, craig.topper

Reviewed By: rnk

Subscribers: JosephTremoulet, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D77242

We have no need to include the entire header.

ScoreboardHazardRecognizer.h - remove unnecessary llvm::InstrItineraryData forward declaration. NFC.

We have to include MCInstrItineraries.h where its defined.

Summary:
Before this PR, `modernize-use-using` would transform the typedef in
```

template <typename a> class TemplateKeyword {
  typedef typename a::template f<> e;
  typedef typename a::template f<>::d e2;
};
```
into
```
template <typename a> class TemplateKeyword {
  using d = typename a::b<>;
  using d2 = typename a::template a::b<>::c;
};
```
The first one is missing the `template` keyword,
the second one has an extra `a::` scope. Both result
in compilation errors.

Reviewers: aaron.ballman, alexfh, hokein, njames93

Subscribers: xazax.hun, cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D78139

Summary:
We can simplify the LHSTy correction for
fixed-point compassign by moving it below
the point where we know we have a compound
assignment.

Also, we shouldn't look at the LHS and RHS
separately; look at the computation result
type instead.

Looking at the LHS and RHS is also wrong
for compassigns with fixed and floating
point (though this does not work upstream
yet).

Reviewers: leonardchan

Subscribers: cfe-commits

Tags: #clang

Differential Revision: https://reviews.llvm.org/D78294

Now that write data continously into the memory mapping, we don't need
to keep the VMO handle around after it has been mapped. This change also
ensures that the VMO is always closed on error.

Differential Revision: https://reviews.llvm.org/D76963

Add patterns that use a normal, non-wrapping, add and sub nodes along
with an arm vshr imm node.

Differential Revision: https://reviews.llvm.org/D77065

These two tests utilize pre-generated opt-viewer output to diff against
a run of opt-viewer over a known yaml file.

In commit 4b428e8f (D76126), the escape function used for rendering was changed
from cgi.escape to html.escape. This modification causes a behavioral
difference with regards to quote characters.

cgi will not escape quotes by default, but html will.

Therefore, these tests were failing because they expected the old behavior
of "string", but was instead seeing "string&quot.

This solution modifies the known test outputs to use the escaped quotes
rather than not escaping quotes during rendering for no particular reason.

It is notable that when testing the optimization records generated by
LLVM, there was never quotes in the remarks I could find, specifically in
the Callee field where they exist in the pre-generated yaml for testing.

Differential Revision: https://reviews.llvm.org/D78241

It causes some encoding fails. Plan to recommit it after fixing that.

This reverts commit 3017580c.

Summary:
We determine the REX prefix used by instruction in `determineREXPrefix`,
and this value is used in `emitMemModRMByte' and used as the return
value of `emitOpcodePrefix`.

Before this patch, REX was passed as reference to `emitPrefixImpl`, it
is strange and not necessary, e.g, we have to write
```
bool Rex = false;
emitPrefixImpl(CurOp, CurByte, Rex, MI, STI, OS);
```
in `emitPrefix` even if `Rex` will not be used.

So we let HasREX be the return value of `emitPrefixImpl`. The HasREX is passed
from `emitREXPrefix` to  `emitOpcodePrefix` and then to
`emitPrefixImpl`. This makes sense since REX is a kind of opcode prefix
and of course is a prefix.

Reviewers: craig.topper, pengfei

Reviewed By: craig.topper

Subscribers: annita.zhang, craig.topper, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D78276

If we are and the constant like 0xFFFFFFC00000, for now, we are using several
instructions to generate this 48bit constant and final an "and". However, we
could exploit it with two rotate instructions.

       MB          ME               MB+63-ME
+----------------------+     +----------------------+
|0000001111111111111000| ->  |0000000001111111111111|
+----------------------+     +----------------------+
 0                    63      0                    63
Rotate left ME + 1 bit first, and then, mask it with (MB + 63 - ME, 63),
finally, rotate back. Notice that, we need to round it with 64 bit for the
wrapping case.

Reviewed by: ChenZheng, Nemanjai

Differential Revision: https://reviews.llvm.org/D71831

[SampleProfile] Use CallBase in function arguments and data structures to reduce the number of explicit casts. NFCI

Removing CallSite left us with a bunch of explicit casts from
Instruction to CallBase. This moves the casts earlier so that
function arguments and data structure types are CallBase so
we don't have to cast when we use them.

Differential Revision: https://reviews.llvm.org/D78246

The header magic value is held in the native endianness, so the method used in
cc0ec3fd. Use MachOReader / MachOWriter's existing endianness tests instead.

Summary:
Rather than having a full, recursive, lowering of vector.broadcast
to LLVM IR, it is much more elegant to have a progressive lowering
of each vector.broadcast into a lower dimensional vector.broadcast,
until only elementary vector operations remain. This results
in more elegant, step-wise code, that is easier to understand.
Also makes some optimizations in the generated code.

Reviewers: nicolasvasilache, mehdi_amini, andydavis1, grosul1

Reviewed By: nicolasvasilache

Subscribers: mehdi_amini, rriddle, jpienaar, burmako, shauheen, antiagainst, nicolasvasilache, arpith-jacob, mgester, lucyrfox, liufengdb, Joonsoo, grosul1, frgossen, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D78071

Use shift/mask operations to access r_symbolnum rather than relying on
MachO::relocation_info. This should fix the big-endian bot failures that were
caused by 386f1c11.

Remove the llvm/Support/StringPool.h file and related support now that it has no clients.  A plain old StringSet<> is a better replacement.

Subscribers: mgorny, hiraditya, llvm-commits

Tags: #llvm

Differential Revision: https://reviews.llvm.org/D78336

Temporarily reverts commit d10386e1
because it breaks the Windows build. MSVC complains about an ambiguous
call to an overloaded function.