Summary:
TargetBaseAlign is no longer required since LSV checks if target allows misaligned accesses.
A constant defining a base alignment is still needed for stack accesses where alignment can be adjusted.
Previous patch (D22936) was reverted because tests were failing. This patch also fixes the cause of those failures:
- x86 failing tests either did not have the right target, or the right alignment.
- NVPTX failing tests did not have the right alignment.
- AMDGPU failing test (merge-stores) should allow vectorization with the given alignment but the target info
considers <3xi32> a non-standard type and gives up early. This patch removes the condition and only checks
for a maximum size allowed and relies on the next condition checking for %4 for correctness.
This should be revisited to include 3xi32 as a MVT type (on arsenm's non-immediate todo list).
Note that checking the sizeInBits for a MVT is undefined (leads to an assertion failure),
so we need to create an EVT, hence the interface change in allowsMisaligned to include the Context.
Reviewers: arsenm, jlebar, tstellarAMD
Subscribers: jholewinski, arsenm, mzolotukhin, llvm-commits
Differential Revision: https://reviews.llvm.org/D23068
llvm-svn: 277735
Summary:
The misc-argument-comment check now ignores leading and trailing underscores and
case. The new `StrictMode` local/global option can be used to switch back to
strict checking.
Add getLocalOrGlobal version for integral types, minor cleanups.
Reviewers: hokein, aaron.ballman
Subscribers: aaron.ballman, Prazek, cfe-commits
Differential Revision: https://reviews.llvm.org/D23135
llvm-svn: 277729
On modern Intel processors hardware SQRT in many cases is faster than RSQRT
followed by Newton-Raphson refinement. The patch introduces a simple heuristic
to choose between hardware SQRT instruction and Newton-Raphson software
estimation.
The patch treats scalars and vectors differently. The heuristic is that for
scalars the compiler should optimize for latency while for vectors it should
optimize for throughput. It is based on the assumption that throughput bound
code is likely to be vectorized.
Basically, the patch disables scalar NR for big cores and disables NR completely
for Skylake. Firstly, scalar SQRT has shorter latency than NR code in big cores.
Secondly, vector SQRT has been greatly improved in Skylake and has better
throughput compared to NR.
Differential Revision: https://reviews.llvm.org/D21379
llvm-svn: 277725
node->dn.task is only filled after the dependencies are already processed.
This currently leads to unhelpful output from KA_TRACE or even a crash
if one enables KMP_SUPPORT_GRAPH_OUTPUT.
llvm-svn: 277717
When using orbis-llvm-cov.exe to generate the HTML report, the HTML report
can look quite different to the source file if it includes tabs.The default
tab size is 2 spaces instead of 8 spaces. A command line switch is
be added to set the tab size.
Differential Revision: https://reviews.llvm.org/D23087
llvm-svn: 277715
ASSERT(exp, message)
Ensure that exp is non-zero. If it is zero, then exit the linker with an error
code, and print message.
ASSERT is useful and was seen in few projects in the wild.
Differential revision: https://reviews.llvm.org/D22912
llvm-svn: 277710
r277702 introduced clang-format changes so that later commits wouldn't introduce
non-functional changes while running clang-format before commiting. Though,
few changes by clang-format weren't in the patch.
llvm-svn: 277709
Enable tail calls by default for (micro)MIPS(64).
microMIPS is slightly more tricky than doing it for MIPS(R6) or microMIPSR6.
microMIPS has two instruction encodings: 16bit and 32bit along with some
restrictions on the size of the instruction that can fill the delay slot.
For safe tail calls for microMIPS, the delay slot filler attempts to find
a correct size instruction for the delay slot of TAILCALL pseudos.
Reviewers: dsanders, vkalintris
Subscribers: jfb, dsanders, sdardis, llvm-commits
Differential Revision: https://reviews.llvm.org/D21138
llvm-svn: 277708
According to spec:
"SORT_BY_ALIGNMENT will sort sections into descending order by
alignment before placing them in the output file"
Previously they were sorted into ascending order.
llvm-svn: 277706
So that later commits don't introduce non-functional changes when
running clang-format before committing.
Reviewers: klimek
Differential Revision: https://reviews.llvm.org/D23153
llvm-svn: 277702
This should ensure that we can atomically write two bytes (on top of the
retq and the one past it) and have those two bytes not straddle cache
lines.
We also move the label past the alignment instruction so that we can refer
to the actual first instruction, as opposed to potential padding before the
aligned instruction.
Update the tests to allow us to reflect the new order of assembly.
Reviewers: rSerge, echristo, majnemer
Subscribers: llvm-commits, mehdi_amini
Differential Revision: https://reviews.llvm.org/D23101
llvm-svn: 277701
This just tests that the register limit isn't exceeded,
so the regisetr allocation doesn't need to be great.'
The critically slow part is all in greedy RA, so
switch to basic.
llvm-svn: 277700
Pass the content of scalar array references to the alloca on the kernel side
and do not pass them additional as normal LLVM scalar value.
llvm-svn: 277699
Since the directory is empty on Darwin, disable the inclusion and avoid
the warning below. Exclude on Android as well to match the behavior from
lib/interception/tests/CMakeLists.txt
lit.py:
/Users/buildslave/jenkins/sharedspace/clang-R_master@2/llvm/utils/lit/lit/discovery.py:224:
warning: input
'/Users/buildslave/jenkins/sharedspace/clang-R_master@2/clang-build/Build/tools/clang/runtime/compiler-rt-bins/test/interception/Unit'
contained no tests
This fixes the above warning in some of public bots, like
http://lab.llvm.org:8080/green/job/clang-stage2-configure-Rlto_check/8686
Differential Revision: https://reviews.llvm.org/D23128
rdar://problem/27581108
llvm-svn: 277692
overloaded (and simpler).
Sean rightly pointed out in code review that we've started using
"wrapper pass" as a specific part of the old pass manager, and in fact
it is more applicable there. Here, we really have a pass *template* to
build a repeated pass, so call it that.
llvm-svn: 277689
With the previous change, it is now obvious that readProvide in
this context appended new commands to a wrong command list.
It was mistakenly adding new commands to the top level.
Thus, all commands inside output section descriptions were
interpreted as they were written on top level.
PROVIDE command naturally requires symbol assignment support
in the output section description. We don't have that one yet.
I removed the implementation because there's no way to fix it now.
We can resurrect the test once we support the symbol assignment
(with a modification to detect errors that we failed to find as
described.)
llvm-svn: 277687
Previously, many read* functions created new command objects and
add them directly to the top-level data structure. This is not
work for some commands because some commands, such as the assignment,
can appear inside and outside of the output section description.
This patch is to not append objects to the top-level data structure.
Callers are now responsible to do that.
llvm-svn: 277686
Alina Sbirlea