1. Throughput and codesize costs estimations was separated and updated.
2. Updated fdiv cost estimation for different cases.
3. Added scalarization processing for types that are treated as !isSimple() to
improve codesize estimation in getArithmeticInstrCost() and
getArithmeticInstrCost(). The code was borrowed from TCK_RecipThroughput path
of base implementation.
Next step is unify scalarization part in base class that is currently works for
TCK_RecipThroughput path only.
Reviewed By: rampitec
Differential Revision: https://reviews.llvm.org/D89973
This patch introduces the dependencies required to read and manage input files
provided by the command line option. It also adds the infrastructure to create
and write to output files. The output is sent to either stdout or a file
(specified with the `-o` flag).
Separately, in order to be able to test the code for file I/O, it adds
infrastructure to create frontend actions. As a basic testable example, it adds
the `InputOutputTest` FrontendAction. The sole purpose of this action is to
read a file from the command line and print it either to stdout or the output
file. This action is run by using the `-test-io` flag also introduced in this
patch (available for `flang-new` and `flang-new -fc1`). With this patch:
```
flang-new -test-io input-file.f90
```
will read input-file.f90 and print it in the output file.
The `InputOutputTest` frontend action has been introduced primarily to
facilitate testing. It is hidden from users (i.e. it's only displayed with
`--help-hidden`). Currently Clang doesn’t have an equivalent action.
`-test-io` is used to trigger the InputOutputTest action in the Flang frontend
driver. This patch makes sure that “flang-new” forwards it to “flang-new -fc1"
by creating a preprocessor job. However, in Flang.cpp, `-test-io` is passed to
“flang-new -fc1” without `-E`. This way we make sure that the preprocessor is
_not_ run in the frontend driver. This is the desired behaviour: `-test-io`
should only read the input file and print it to the output stream.
co-authored-by: Andrzej Warzynski <andrzej.warzynski@arm.com>
Differential Revision: https://reviews.llvm.org/D87989
The exit blocks of the versioned and non-versioned loops are not dedicated and thus the two loops are not in simplify form.
Insert dummy exit blocks after loop versioning with `formDedicatedExits()` to preserve the simplify form for subsequence passes.
Reviewed By: aeubanks
Differential Revision: https://reviews.llvm.org/D89569
Symbols with special section index SHN_COMMON (0xfff2) haven't been handled so far and caused an invalid section error.
This is a more or less straightforward use of the code commented out at the end of the function. I checked with the ELF spec, that the symbol value gives the alignment.
Reviewed By: lhames
Differential Revision: https://reviews.llvm.org/D89795
The difference is that the former are indirect and go to the GOT while the latter go to the target directly. This info can be used to relax indirect ones that don't need the GOT (because the target is in range). We check for this optimization beforehand. For formal correctness and to avoid confusion, we should only change the relocation kind if we actually apply the relaxation.
As discussed on PR35155, this extends narrowFunnelShift (recently renamed from narrowRotate) to support basic funnel shift patterns.
Unlike matchFunnelShift we don't include the computeKnownBits limitation as extracting the pattern from the zext/trunc layers should be a indicator of reasonable funnel shift codegen, in D89139 we demonstrated how to efficiently promote funnel shifts to wider types.
Differential Revision: https://reviews.llvm.org/D89542
Replace the X86 specific isSplatZeroExtended helper with a generic BuildVectorSDNode method.
I've just used this to simplify the X86ISD::BROADCASTM lowering so far (and remove isSplatZeroExtended), but we should be able to use this in more places to lower to complex broadcast patterns.
Differential Revision: https://reviews.llvm.org/D87930
Rather than performing the cache lookup with both possible orders
for the locations, use the same canonicalization as the other
AliasCache lookups in BasicAA.
Any time we insert a block into VisitedPhiBBs, previously cached
values may no longer be valid for the recursive alias queries. As
such, perform them using an empty AAQueryInfo.
Note that if we recurse to the same phi, the block will already
be inserted, so we reuse the old AAQueryInfo, and thus still
protect against infinite recursion.
This problem can appear with with an without BatchAA, but is more
likely to occur with BatchAA, as more values are cached.
Differential Revision: https://reviews.llvm.org/D90066
This value had the default value of 4 which caused branch relaxation to fail.
Review: Ulrich Weigand
Differential Revision: https://reviews.llvm.org/D90065
Unify the x86 regset API to use XStateRegSet for all FPU registers,
therefore eliminating the legacy API based on FPRegSet. This makes
the code a little bit simpler but most notably, it provides future
compatibility for register caching.
Since the NetBSD kernel takes care of providing compatibility with
pre-XSAVE processors, PT_{G,S}ETXSTATE can be used on systems supporting
only FXSAVE or even plain FSAVE (and unlike PT_{G,S}ETXMMREGS, it
clearly indicates that XMM registers are not supported).
Differential Revision: https://reviews.llvm.org/D90034
The class IntrinsicProcTable uses the pimpl idiom and manages its own pointer-to-implementation. However, it violates the rule-of-five and does not implement a move-constructor or assignment-operator. Due to differences between compilers in implementation copy elision, these may or may not be used. Due to the missing user implementation for resource handling, using the results in runtime errors.
Fix my using `std::unique_ptr` instead of custom resource management.
Reviewed By: klausler
Differential Revision: https://reviews.llvm.org/D88794
This patch enables emitting DWARF `DW_OP_implicit_value` opcode when
tuning debug information for LLDB (`-debugger-tune=lldb`).
This will also propagate to Darwin platforms, since they use LLDB tuning
as a default.
rdar://67406059
Differential Revision: https://reviews.llvm.org/D90001
Signed-off-by: Med Ismail Bennani <medismail.bennani@gmail.com>
Patch D88695 introduces a new local variable inside a lambda with the same name as a variable outside of it. In some of the if constexpr regions, msvc prioritizes the outer declaration and emits the error.
```
C:\Users\meinersbur\src\llvm-project\flang\lib\Evaluate\fold-implementation.h(1200): error C3493: 'context' cannot be implicitly captured because no default capture mode has been specified
```
This is fixed by giving the inner variable a different name.
Reviewed By: klausler
Differential Revision: https://reviews.llvm.org/D89367
Remove the assumption that the path separator is `/`. Use functions from `llvm::sys::path` instead.
Reviewed By: isuruf, klausler
Differential Revision: https://reviews.llvm.org/D89369
When all provided source files are filtered out either due to `--ignore-filename-regex` or not part of binary, don't generate coverage reults for all source files. Because if users want to generate coverage results for all source files, they don't even need to provid selected source files or `--ignore-filename-regex`.
Differential Revision: https://reviews.llvm.org/D89359
Simplify `HeaderSearch::LookupFile`. Instead of deconstructing a
`FileEntryRef` into a name and `FileEntry` and then rebuilding it later,
use it as is. This helps to unblock making the constructor of
`FileEntryRef` private to `FileManager`.
Differential Revision:
In `ReplayPreamble::replay`, use `getFileRef` instead of `getFile`, and
then use that `FileEntryRef` later to avoid needing
`FileEntryRef::FileEntryRef`. The latter is going to become private to
`FileManager` in a later commit.
This dependency was already existing indirectly, but is now more direct
since the registration relies on a inline function. This fixes the
link of the tools with BFD.
Duplicated callsites share the same callee profile if the original callsite was inlined. The sharing also causes the profile of callee's callee to be shared. This breaks the assert introduced ealier by D84997 in a tricky way.
To illustrate, I'm using an abstract example. Say we have three functions `A`, `B` and `C`. A calls B twice and B calls C once. Some optimize performed prior to the sample profile loader duplicates first callsite to `B` and the program may look like
```
A()
{
B(); // with nested profile B1 and C1
B(); // duplicated, with nested profile B1 and C1
B(); // with nested profile B2 and C2
}
```
For some reason, the sample profile loader inliner then decides to only inline the first callsite in `A` and transforms `A` into
```
A()
{
C(); // with nested profile C1
B(); // duplicated, with nested profile B1 and C1
B(); // with nested profile B2 and C2.
}
```
Here is what happens next:
1. Failing to inline the callsite `C()` results in `C1`'s samples returned to `C`'s base (outlined) profile. In the meantime, `C1`'s head samples are updated to `C1`'s entry sample. This also affects the profile of the middle callsite which shares `C1` with the first callsite.
2. Failing to inline the middle callsite results in `B1` returned to `B`'s base profile, which in turn will cause `C1` merged into `B`'s base profile. Note that the nest `C` profile in `B`'s base has a non-zero head sample count now. The value actually equals to `C1`'s entry count.
3. Failing to inline last callsite results in `B2` returned to `B`'s base profile. Note that the nested `C` profile in `B`'s base now has an entry count equal to the sum of that of `C1` and `C2`, with the head count equal to that of `C1`. This will trigger the assert later on.
4. Compiling `B` using `B`'s base profile. Failing to inline `C` there triggers the returning of the nested `C` profile. Since the nested `C` profile has a non-zero head count, the returning doesn't go through. Instead, the assert goes off.
It's good that `C1` is only returned once, based on using a non-zero head count to ensure an inline profile is only returned once. However C2 is never returned. While it seems hard to solve this perfectly within the current framework, I'm just removing the broken assert. This should be reasonably fixed by the upcoming CSSPGO work where counts returning is based on context-sensitivity and a distribution factor for callsite probes.
The simple example is extracted from one of our internal services. In reality, why the original callsite `B()` and duplicate one having different inline behavior is a magic. It has to do with imperfect counts in profile and extra complicated inlining that makes the hotness for them different.
Reviewed By: wenlei
Differential Revision: https://reviews.llvm.org/D90056
As mentioned in the comment inside the code, the Intel documentation
states that the internal CPU buffer is flushed out to RAM only when tracing is
disabled. Otherwise, the buffer on RAM might be stale.
This diff disables tracing when the trace buffer is going to be read. This is a
quite safe operation, as the reading is done when the inferior is paused at a
breakpoint, so we are not losing any packets because there's no code being
executed.
After the reading is finished, tracing is enabled back.
It's a bit hard to write a test for this now, but Greg Clayton and I will
refactor the PT support and writing tests for it will be easier. However
I tested it manually by doing a script that automates
the following flow
```
(lldb) b main
Breakpoint 1: where = a.out`main + 15 at main.cpp:4:7, address = 0x000000000040050f
(lldb) r
Process 3078226 stopped
* thread #1, name = 'a.out', stop reason = breakpoint 1.1
frame #0: 0x000000000040050f a.out`main at main.cpp:4:7
(lldb) processor-trace start
(lldb) b 5
Breakpoint 2: where = a.out`main + 22 at main.cpp:5:12, address = 0x0000000000400516
(lldb) c
Process 3078226 resuming
Process 3078226 stopped
* thread #1, name = 'a.out', stop reason = breakpoint 2.1
frame #0: 0x0000000000400516 a.out`main at main.cpp:5:12
(lldb) processor-trace show-instr-log
thread #1: tid=3078226
0x40050f <+15>: movl $0x0, -0x8(%rbp)
>>> Before, some runs of the script up to this point lead to empty traces
(lldb) b 6
Breakpoint 3: where = a.out`main + 42 at main.cpp:6:14, address = 0x000000000040052a
(lldb) c
Process 3092991 resuming
Process 3092991 stopped
* thread #1, name = 'a.out', stop reason = breakpoint 3.1
frame #0: 0x000000000040052a a.out`main at main.cpp:6:14
(lldb) processor-trace show-instr-log thread #1: tid=3092991
0x40050f <+15>: movl $0x0, -0x8(%rbp)
0x400516 <+22>: movl $0x0, -0xc(%rbp)
0x40051d <+29>: cmpl $0x2710, -0xc(%rbp) ; imm = 0x2710
0x400524 <+36>: jge 0x400546 ; <+70> at main.cpp
0x400524 <+36>: jge 0x400546 ; <+70> at main.cpp
>>> The trace was re-enabled correctly and includes the instruction of the
first reading.
```
Those instructions correspond to these lines
```
3 int main() {
4 int z = 0;
5 for (int i = 0; i < 10000; i++) {
6 z += fun(z)
...
```
Differential Revision: https://reviews.llvm.org/D85241
lambda-expression's captures.
The built-in structured binding rules for classes require that all
fields can be accessed by name, and the fields introduced for lambda
captures are unnamed, so decomposing a capturing lambda is ill-formed.
-print_full_coverage=1 produces a detailed branch coverage dump when run on a single file.
Uses same infrastructure as -print_coverage flag, but prints all branches (regardless of coverage status) in an easy-to-parse format.
Usage: For internal use with machine learning fuzzing models which require detailed coverage information on seed files to generate mutations.
Differential Revision: https://reviews.llvm.org/D85928
Reverts the XFAIL added in b67a2aef8a,
which had no effect.
Adjust the test to make sure all output is dumped to stderr, so that
hopefully I can get a better idea of where/why this is failing.
Remove some redundant checking while here.
This doesn't support -structurizecfg-skip-uniform-regions since that
would require porting LegacyDivergenceAnalysis.
The NPM doesn't support adding a non-analysis pass as a dependency of
another, so I had to add -lowerswitch to some tests or pin them to the
legacy PM.
This is the only RegionPass in tree, so I simply copied the logic for
finding all Regions from the legacy PM's RGManager into
StructurizeCFG::run().
Reviewed By: arsenm
Differential Revision: https://reviews.llvm.org/D89026
Benjamin Kramer