With --no-align-segments, there is a bug that the fileoffset may not be
congruent to virtual address modulo page alignment.
This patch fixes the problem.
llvm-svn: 221890
This will become the root of a new class hierarchy separate from
`Value`. As a first step, stick it between `Value` and `MDNode`.
This is part of PR21532.
llvm-svn: 221886
This fixes an assertion when running clang-tidy on a file having
--serialize-diagnostics in compiler options. Committing a regression test
for clang-tidy separately.
Patch by Aaron Wishnick!
llvm-svn: 221884
Summary:
I don't know anything about profiling but it seems to work out of the
box on PowerPC64. At least "make check-profile" works.
A few tests needed tweaking because PowerPC64 IR declares main with
"define signext i32 @main" instead of just "define i32 @main".
This also fixes the asan asan_and_llvm_coverage_test test, which
compiles with -coverage so requires that a profiling version of
libclang_rt has been built.
Reviewers: dexonsmith, kcc, samsonov
Reviewed By: samsonov
Subscribers: samsonov, llvm-commits
Differential Revision: http://reviews.llvm.org/D6233
llvm-svn: 221877
Let's try this again...
This reverts r219432, plus a bug fix.
Description of the bug in r219432 (by Nick):
The bug was using AllPositive to break out of the loop; if the loop break
condition i != e is changed to i != e && AllPositive then the
test_modulo_analysis_with_global test I've added will fail as the Modulo will
be calculated incorrectly (as the last loop iteration is skipped, so Modulo
isn't updated with its Scale).
Nick also adds this comment:
ComputeSignBit is safe to use in loops as it takes into account phi nodes, and
the == EK_ZeroEx check is safe in loops as, no matter how the variable changes
between iterations, zero-extensions will always guarantee a zero sign bit. The
isValueEqualInPotentialCycles check is therefore definitely not needed as all
the variable analysis holds no matter how the variables change between loop
iterations.
And this patch also adds another enhancement to GetLinearExpression - basically
to convert ConstantInts to Offsets (see test_const_eval and
test_const_eval_scaled for the situations this improves).
Original commit message:
This reverts r218944, which reverted r218714, plus a bug fix.
Description of the bug in r218714 (by Nick):
The original patch forgot to check if the Scale in VariableGEPIndex flipped the
sign of the variable. The BasicAA pass iterates over the instructions in the
order they appear in the function, and so BasicAliasAnalysis::aliasGEP is
called with the variable it first comes across as parameter GEP1. Adding a
%reorder label puts the definition of %a after %b so aliasGEP is called with %b
as the first parameter and %a as the second. aliasGEP later calculates that %a
== %b + 1 - %idxprom where %idxprom >= 0 (if %a was passed as the first
parameter it would calculate %b == %a - 1 + %idxprom where %idxprom >= 0) -
ignoring that %idxprom is scaled by -1 here lead the patch to incorrectly
conclude that %a > %b.
Revised patch by Nick White, thanks! Thanks to Lang to isolating the bug.
Slightly modified by me to add an early exit from the loop and avoid
unnecessary, but expensive, function calls.
Original commit message:
Two related things:
1. Fixes a bug when calculating the offset in GetLinearExpression. The code
previously used zext to extend the offset, so negative offsets were converted
to large positive ones.
2. Enhance aliasGEP to deduce that, if the difference between two GEP
allocations is positive and all the variables that govern the offset are also
positive (i.e. the offset is strictly after the higher base pointer), then
locations that fit in the gap between the two base pointers are NoAlias.
Patch by Nick White!
llvm-svn: 221876
Summary: My PowerPC64 Linux box has 64k pages. The test assumes 4k pages. Fix it.
Reviewers: glider, eugenis, samsonov
Reviewed By: samsonov
Subscribers: llvm-commits
Differential Revision: http://reviews.llvm.org/D6228
llvm-svn: 221875
Split getObject's smarts into checkOffset, use this to replace the
handwritten check in getSectionContents. Similarly, replace checks in
section_rel_begin/section_rel_end with getNumberOfRelocations.
No functionality change intended.
llvm-svn: 221873
On error conditions, relocAddressLess might claim that a value is less
than itself. Instead, abort llvm-readobj. No functionality change
intended.
llvm-svn: 221872
lib/Object is supposed to be robust to malformed object files. Don't
assert if we don't have a symbol table. I'll try to come up with a test
case later.
llvm-svn: 221870
getObject didn't consider the case where a pointer came before the start
of the object file. No test is included, trying to come up with
something reasonable.
llvm-svn: 221868
RegisterContextLLDB. I have core files of half a dozen tricky
unwind situations on x86/arm and they're all working pretty much
correctly at this point, but we'll need to keep an eye out for
unwinder regressions for a little while; it's tricky to get these
heuristics completely correct in all unwind situations.
<rdar://problem/18937193>
llvm-svn: 221866
MIPS ELF symbols might contain some additional MIPS-specific flags
in the st_other field besides visibility ones. These flags indicate
code properties like microMIPS / MIPS16 encoding, position independent
code etc. We need to transfer the flags from input objects to the
output linked file to write them into the symbol table, adjust symbols
addresses etc.
I add new attribute CodeModel to the DefinedAtom class to hold target
specific flag and to get over YAML/Native format conversion barrier.
Other architectures/targets can extend CodeModel enumeration by their
own flags.
MIPS specific part of this patch adds support for STO_MIPS_MICROMIPS
flag. This flag marks microMIPS symbols. Such symbol should:
a) Has STO_MIPS_MICROMIPS in the corresponding .symtab record.
b) Has adjusted (odd) address in the corresponding .symtab
and .dynsym records.
llvm-svn: 221864
The segment alignment for PT_LOAD segments is set to page size by default, but
if any of the sections require an alignment more than the page size, the segment
alignment property is set to the maximum alignment of the sections that are part
of the segment.
llvm-svn: 221862
between splitting a vector into 128-bit lanes and recombining them vs.
decomposing things into single-input shuffles and a final blend.
This handles a large number of cases in AVX1 where the cross-lane
shuffles would be much more expensive to represent even though we end up
with a fast blend at the root. Instead, we can do a better job of
shuffling in a single lane and then inserting it into the other lanes.
This fixes the remaining bits of Halide's regression captured in PR21281
for AVX1. However, the bug persists in AVX2 because I've made this
change reasonably conservative. The cases where it makes sense in AVX2
to split into 128-bit lanes are much more rare because we can often do
full permutations across all elements of the 256-bit vector. However,
the particular test case in PR21281 is an example of one of the rare
cases where it is *always* better to work in a single 128-bit lane. I'm
going to try to teach the logic to detect and form the good code even in
AVX2 next, but it will need to use a separate heuristic.
Finally, there is one pesky regression here where we previously would
craftily use vpermilps in AVX1 to shuffle both high and low halves at
the same time. We no longer pull that off, and not for any really good
reason. Ultimately, I think this is just another missing nuance to the
selection heuristic that I'll try to add in afterward, but this change
already seems strictly worth doing considering the magnitude of the
improvements in common matrix math shuffle patterns.
As always, please let me know if this causes a surprising regression for
you.
llvm-svn: 221861
Part of TestConcurrentEvents starts threads that are supposed to be
delayed by one second.
Test was adding "delay" threads to the "actions" thread list instead
of the "delay_actions" list, which caused them to be started without
delay.
llvm-svn: 221859
The user can use the max-page-size option and set the maximum page size. Dont
check for maximum allowed values for page size, as its what the kernel is
configured with.
Fix the test as well.
llvm-svn: 221858
Each entry in the delay-import address table had a wrong alignment
requirement if 32 bit. As a result it got wrong delay-import table.
Because llvm-readobj doesn't print out that field, we don't have a
test for that. I'll submit a test that would catch this bug after
improving llvm-readobj.
llvm-svn: 221853
re-combining shuffles because nothing was available in the wider vector
type.
The key observation (which I've put in the comments for future
maintainers) is that at this point, no further combining is really
possible. And so even though these shuffles trivially could be combined,
we need to actually do that as we produce them when producing them this
late in the lowering.
This fixes another (huge) part of the Halide vector shuffle regressions.
As it happens, this was already well covered by the tests, but I hadn't
noticed how bad some of these got. The specific patterns that turn
directly into unpckl/h patterns were occurring *many* times in common
vector processing code.
There are still more problems here sadly, but trying to incrementally
tease them apart and it looks like this is the core of the problem in
the splitting logic.
There is some chance of regression here, you can see it in the test
changes. Specifically, where we stop forming pshufb in some cases, it is
possible that pshufb was in fact faster. Intel "says" that pshufb is
slower than the instruction sequences replacing it.
llvm-svn: 221852
Prior to this patch the TypePromotionHelper was promoting only sign extensions.
Supporting zero extensions changes:
- How constants are extended.
- How sign extensions, zero extensions, and truncate are composed together.
- How the type of the extended operation is recorded. Now we need to know the
kind of the extension as well as its type.
Each change is fairly small, unlike the diff.
Most of the diff are comments/variable renaming to say "extension" instead of
"sign extension".
The performance improvements on the test suite are within the noise.
Related to <rdar://problem/18310086>.
llvm-svn: 221851
This folds the compare emission into the select emission when possible, so we
can directly use the flags and don't have to emit a separate compare.
Related to rdar://problem/18960150.
llvm-svn: 221847
The Autoconf build already does this, but it was never ported to
CMake. The host linker version affects the flags that Clang pass
to the linker, notably whether it passes -demangle or not.
http://reviews.llvm.org/D6239
llvm-svn: 221844
Rafael Espindola