Summary:
`// taze: ... from ...` comments are used help tools where a
specific global symbol comes from.
Before:
// taze: many, different, symbols from
// 'some_long_location_here'
After:
// taze: many, different, symbols from 'some_long_location_here'
Reviewers: djasper
Subscribers: klimek, cfe-commits
Differential Revision: https://reviews.llvm.org/D24477
llvm-svn: 281857
Summary:
Before when a semicolon was missing after a boolean literal:
a = true
return 1;
clang-format would parse this as one line and format as:
a = true return 1;
It turns out that C++ does not consider `true` and `false` to be literals, we
have to check for that explicitly.
Reviewers: djasper
Subscribers: klimek, cfe-commits
Differential Revision: https://reviews.llvm.org/D24574
llvm-svn: 281856
These are all emitted into a section with a cstring_literal attribute. The
attribute permits the linker to coalesce the string contents. The address of
the strings are not important.
llvm-svn: 281855
These strings are constants, mark them as such. This doesn't matter too much in
practice on MachO since the constants are placed into a special section and not
referred to directly.
llvm-svn: 281854
With D24253 we can now use SelectionDAG::SignBitIsZero with vector operations.
This patch uses SelectionDAG::SignBitIsZero to recognise that a zero sign bit means that we can use a sitofp instead of a uitofp (which is not directly support on pre-AVX512 hardware).
While AVX512 does provide support for uitofp, the conversion to sitofp should not cause any regressions.
Differential Revision: https://reviews.llvm.org/D24343
llvm-svn: 281852
Simplified GEP cloning in vectorizeMemoryInstruction().
Added an assertion that checks consecutive GEP, which should have only one loop-variant operand.
Differential Revision: https://reviews.llvm.org/D24557
llvm-svn: 281851
In case sequential kernels are found deeper in the loop tree than any parallel
kernel, the overall scop is probably mostly sequential. Hence, run it on the
CPU.
llvm-svn: 281849
Offloading to a GPU is only beneficial if there is a sufficient amount of
compute that can be accelerated. Many kernels just have a very small number
of dynamic compute, which means GPU acceleration is not beneficial. We
compute at run-time an approximation of how many dynamic instructions will be
executed and fall back to CPU code in case this number is not sufficiently
large. To keep the run-time checking code simple, we over-approximate the
number of instructions executed in each statement by computing the volume of
the rectangular hull of its iteration space.
llvm-svn: 281848
Summary:
This fixes an issue when files are compiled with -flto=thin
at default -O0. We need to rename anonymous globals before attempting
to write the module summary because all values need names for
the summary. This was happening at -O1 and above, but not before
the early exit when constructing the pipeline for -O0.
Also add an internal -prepare-for-thinlto option to enable this
to be tested via opt.
Fixes PR30419.
Reviewers: mehdi_amini
Subscribers: probinson, llvm-commits, mehdi_amini
Differential Revision: https://reviews.llvm.org/D24701
llvm-svn: 281840
We may generate GPU kernels that store into scalars in case we run some
sequential code on the GPU because the remaining data is expected to already be
on the GPU. For these kernels it is important to not keep the scalar values
in thread-local registers, but to store them back to the corresponding device
memory objects that backs them up.
We currently only store scalars back at the end of a kernel. This is only
correct if precisely one thread is executed. In case more than one thread may
be run, we currently invalidate the scop. To support such cases correctly,
we would need to always load and store back from a corresponding global
memory slot instead of a thread-local alloca slot.
llvm-svn: 281838
Previously it used llvm-readobj -s, now changed to llvm-objdump -section-headers,
that reduced the output significantly and helps to read/support it.
llvm-svn: 281832
Our implementation supported integer value previously.
ld can use expression,
for example, it is OK to write
. = SEGMENT_START("foobar", .);
Patch implements that.
llvm-svn: 281831
Fix the output of clang-rename for the files without modifications.
Update the code in clang-reorder-fields/tool/ClangReorderFields.cpp
to avoid inconsistency.
Example:
a.h:
struct A {};
a.cpp:
#include "a.h"
int main() { return 0; }
Before the changes the output looks like this:
clang-rename -qualified-name=A -new-name=B a.cpp
<<<<<INVALID SOURCE LOCATION>>>>>
Test plan: make -j8 check-clang-tools
Differential revision: https://reviews.llvm.org/D24634
llvm-svn: 281826
We were trying to avoid using a FrameIndex operand in non-pointer
operands in a convoluted way, and would break because of
using TargetFrameIndex. The TargetFrameIndex should only be used
in the case where it makes sense to fold it as part of the addressing
mode, otherwise it requires materialization like a normal constant.
This wasn't working reliably and failed in the added testcase, hitting
the assert when processing the frame index.
The TargetFrameIndex was coming from trying to produce an AssertZext
limiting the maximum stack size. I'm not sure this was correct to begin
with, because it is apparently possible to have a single workitem
dispatch that requires all 4G of private memory.
llvm-svn: 281824
This reduces the number of copies and reg_sequences
when using fp constant vectors. This significantly
reduces the code size in local-stack-alloc-bug.ll
llvm-svn: 281822
Summary:
No behavioral change intended. The change makes iterating the replacements set more intuitive in Replacements class implementation. Previously, insertion is ordered before an deletion/replacement with the same offset, which is counter-intuitive for implementation, especially for a followup patch to support adding insertions around replacements.
With the current ordering, we only need to make `applyAllReplacements` iterate the replacements set reversely when applying them so that deletion/replacement is still applied before insertion with the same offset.
Reviewers: klimek, djasper
Subscribers: klimek, cfe-commits
Differential Revision: https://reviews.llvm.org/D24663
llvm-svn: 281819
It was possible situation about some commands just were not processed
(were skipped) because of a bug appeared when constraint checking used.
Testcase is attached.
llvm-svn: 281818
Simon Pilgrim