Commit Graph

3539 Commits

Author SHA1 Message Date
Sanjoy Das 39c226fdba [STLExtras] Introduce and use llvm::count_if; NFC
(This is split out from was D21115)

llvm-svn: 272435
2016-06-10 21:18:39 +00:00
Alina Sbirlea d665b410c6 Reapply 272328 and 272329 as a single patch.
[cpu-detection] [amdfam10] Return barcelona, and amdfam10 for all other
subtypes. Address Bug 28067.

Along with the refactoring of Host.cpp, getHostCPUName() was modified to
return more precise types for CPUs in amdfam10.
However, callers of getHostCPUName() do string matching on type, so this
cannot be modified.
Currently there is support in the x86 backend for barcelona.
For all other subtypes the assumed return value is amdfam10.

Fix: getHostCPUName() returns barcelona subtype and amdfam10 for all
others. This can be extended further when support for the other subtypes
is added.

Differential revision: http://reviews.llvm.org/D21193

llvm-svn: 272333
2016-06-09 23:04:15 +00:00
Alina Sbirlea e0b7bf42c2 Revert 272328 and 272329 to recommit as a single patch.
llvm-svn: 272332
2016-06-09 23:04:05 +00:00
Alina Sbirlea 4acea0aa22 Keep barcelona subtype for amdfam10
llvm-svn: 272329
2016-06-09 22:47:36 +00:00
Alina Sbirlea e76dcd2b12 [cpu-detection] Return amdfam10 for all subtypes. Address Bug 28067.
Summary: Remove architecture subtype from the string returned by getHostCPUName(). String matching done on type.

Reviewers: llvm-commits, echristo

Subscribers: mehdi_amini

Differential Revision: http://reviews.llvm.org/D21193

llvm-svn: 272328
2016-06-09 22:47:12 +00:00
Richard Smith 2ad6d48b0c Search for llvm-symbolizer binary in the same directory as argv[0], before
looking for it along $PATH. This allows installs of LLVM tools outside of
$PATH to find the symbolizer and produce pretty backtraces if they crash.

llvm-svn: 272232
2016-06-09 00:53:21 +00:00
Alina Sbirlea 080241b75d [cpu-detection] Add missing break statements in outer switches
Summary:
Break on all switch cases for outer and inner switches.
No functionality changed.

Reviewers: llvm-commits, sanjoy

Differential Revision: http://reviews.llvm.org/D21158

llvm-svn: 272228
2016-06-09 00:08:15 +00:00
Reid Kleckner de3d8b500f [DebugInfo] Add calling convention support for DWARF and CodeView
Summary:
Now DISubroutineType has a 'cc' field which should be a DW_CC_ enum.  If
it is present and non-zero, the backend will emit it as a
DW_AT_calling_convention attribute. On the CodeView side, we translate
it to the appropriate enum for the LF_PROCEDURE record.

I added a new LLVM vendor specific enum to the list of DWARF calling
conventions. DWARF does not appear to attempt to standardize these, so I
assume it's OK to do this until we coordinate with GCC on how to emit
vectorcall convention functions.

Reviewers: dexonsmith, majnemer, aaboud, amccarth

Subscribers: mehdi_amini, llvm-commits

Differential Revision: http://reviews.llvm.org/D21114

llvm-svn: 272197
2016-06-08 20:34:29 +00:00
Benjamin Kramer c321e53402 Apply most suggestions of clang-tidy's performance-unnecessary-value-param
Avoids unnecessary copies. All changes audited & pass tests with asan.
No functional change intended.

llvm-svn: 272190
2016-06-08 19:09:22 +00:00
Saleem Abdulrasool 1ef925f0bd Support: correct AArch64 TargetParser implementation
The architecture enumeration is shared across ARM and AArch64.  However, the
data is not.  The code incorrectly would index into the array using the
architecture index which was offset by the ARMv7 architecture enumeration.  We
do not have a marker for indicating the architectural family to which the
enumeration belongs so we cannot be clever about offsetting the index (at least
it is not immediately apparent to me).  Instead, fall back to the tried-and-true
method of slowly iterating the array (its not a large array, so the impact of
this is not too high).

Because of the incorrect indexing, if we were lucky, we would crash, but usually
we would return an invalid StringRef.  We did not have any tests for the AArch64
target parser previously;.  Extend the previous tests I had added for ARM to
cover AArch64 for ensuring that we return expected StringRefs.

Take the opportunity to change some iterator types to references.

This work is needed to support parsing `.arch name` directives in the AArch64
target asm parser.

llvm-svn: 272145
2016-06-08 14:30:00 +00:00
Alina Sbirlea 76c4a855bc [cpu-detection] Substantial refactor of Host CPU detection code (x86)
Summary:
Following D20970 (committed as r271726).
This is a substantial refactoring of the host CPU detection code.

There is no functionality change intended, but the changes are extensive.

Definitions of architecture types and subtypes are by no means exhaustive or
perfectly defined, but a fair starting point.
Suggestions for futher improvements are welcome.

Reviewers: llvm-commits

Differential Revision: http://reviews.llvm.org/D20988

llvm-svn: 271921
2016-06-06 18:29:59 +00:00
Chandler Carruth fe1ffb912d [LPM] Reinstate r271781 which reinstated r271652 to replace the
CALL_ONCE_... macro in the legacy pass manager with the new
llvm::call_once facility.

Nothing changed sicne the last attempt in r271781 which I reverted in
r271788. At least one of the failures I saw was spurious, and I want to
make sure the other failures are real before I work around them -- they
appeared to only effect ppc64le and ppc64be.

Original commit message of r271781:
----
[LPM] Reinstate r271652 to replace the CALL_ONCE_... macro in the legacy
pass manager with the new llvm::call_once facility.

This reverts commit r271657 and re-applies r271652 with a fix to
actually work with arguments. In the original version, we just ended up
directly calling std::call_once via ADL because of the std::once_flag
argument. The llvm::call_once never worked with arguments. Now,
llvm::call_once is a variadic template that perfectly forwards
everything. As a part of this it had to move to the header and we use
a generic functor rather than an explict function pointer. It would be
nice to use std::invoke here but we don't have it yet. That means
pointer to members won't work here, but that seems a tolerable
compromise.

I've also tested this by forcing the fallback path, so hopefully it
sticks this time.
----

Original commit message of r271652:
----
[LPM] Replace the CALL_ONCE_... macro in the legacy pass manager with
the new llvm::call_once facility.

This facility matches the standard APIs and when the platform supports
it actually directly uses the standard provided functionality. This is
both more efficient on some platforms and much more TSan friendly.

The only remaining user of the cas_flag and home-rolled atomics is the
fallback implementation of call_once. I have a patch that removes them
entirely, but it needs a Windows patch to land first.

This alone substantially cleans up the macros for the legacy pass
manager, and should subsume some of the work Mehdi was doing to clear
the path for TSan testing of ThinLTO, a really important step to have
reliable upstream testing of ThinLTO in all forms.
----

llvm-svn: 271800
2016-06-04 19:57:55 +00:00
Chandler Carruth fe9466fe2c [LPM] Revert r271781 which was a re-commit of r271652.
There appears to be a strange exception thrown and crash using call_once
on a PPC build bot, and a *really* weird windows link error for
GCMetadata.obj. Still need to investigate the cause of both problems.

Original change summary:
[LPM] Reinstate r271652 to replace the CALL_ONCE_... macro in the legacy
pass manager with the new llvm::call_once facility.

llvm-svn: 271788
2016-06-04 09:36:40 +00:00
Chandler Carruth fa4890e068 [LPM] Reinstate r271652 to replace the CALL_ONCE_... macro in the legacy
pass manager with the new llvm::call_once facility.

This reverts commit r271657 and re-applies r271652 with a fix to
actually work with arguments. In the original version, we just ended up
directly calling std::call_once via ADL because of the std::once_flag
argument. The llvm::call_once never worked with arguments. Now,
llvm::call_once is a variadic template that perfectly forwards
everything. As a part of this it had to move to the header and we use
a generic functor rather than an explict function pointer. It would be
nice to use std::invoke here but we don't have it yet. That means
pointer to members won't work here, but that seems a tolerable
compromise.

I've also tested this by forcing the fallback path, so hopefully it
sticks this time.

Original commit message:
----
[LPM] Replace the CALL_ONCE_... macro in the legacy pass manager with
the new llvm::call_once facility.

This facility matches the standard APIs and when the platform supports
it actually directly uses the standard provided functionality. This is
both more efficient on some platforms and much more TSan friendly.

The only remaining user of the cas_flag and home-rolled atomics is the
fallback implementation of call_once. I have a patch that removes them
entirely, but it needs a Windows patch to land first.

This alone substantially cleans up the macros for the legacy pass
manager, and should subsume some of the work Mehdi was doing to clear
the path for TSan testing of ThinLTO, a really important step to have
reliable upstream testing of ThinLTO in all forms.

llvm-svn: 271781
2016-06-04 07:25:44 +00:00
Taewook Oh 99497fdebd Revert commit r271704, a patch that enables warnings for non-portable #include and #import paths (Corresponding clang patch has been reverted by r271761). Patches are reverted because they generate lots of unadressable warnings for windows and fail tests under ASAN.
llvm-svn: 271764
2016-06-04 03:36:12 +00:00
Bruno Cardoso Lopes 42b1f65f03 [LockFileManager] Improve error output by using better error messages
This is currently used by clang to lock access to modules; improve the
error message so that clang can use better output messages from locking
error issues.

rdar://problem/26529101

Differential Review: http://reviews.llvm.org/D20942

llvm-svn: 271755
2016-06-04 00:34:00 +00:00
Alina Sbirlea 400eb02775 [cpu-detection] Naming convention
Summary:
    Follow-up to D20926 (committed as r271595, r271596).
    This patch is in preparation for a substantial refactoring of the code.

    No functionality changed.

Differential Revision: http://reviews.llvm.org/D20970

llvm-svn: 271726
2016-06-03 20:27:50 +00:00
Alina Sbirlea 8c9ad10a7f Test commit. Removes some spaces. No functionality changed.
Summary:
Test commit. Removes some spaces.
No functionality changed.

Reviewers: llvm-commits

Differential Revision: http://reviews.llvm.org/D20972

llvm-svn: 271711
2016-06-03 19:20:37 +00:00
Taewook Oh dfec58e80c In openFileForRead, attempt to fetch the actual name of the file on disk -- including case -- so that clang can later warn about non-portable #include and #import directives.
Differential Revision: http://reviews.llvm.org/D19842

Patch by Eric Niebler

llvm-svn: 271704
2016-06-03 18:38:39 +00:00
Ben Craig 60adb9229c Adding reserve and capacity methods to FoldingSet
http://reviews.llvm.org/D20930

llvm-svn: 271669
2016-06-03 13:54:48 +00:00
Eric Christopher b820edd58a 80-column fixup after last formatting change.
llvm-svn: 271598
2016-06-02 21:32:30 +00:00
Eric Christopher d9804bb696 Fix a couple of misformatted comments spotted in post-commit review.
llvm-svn: 271596
2016-06-02 21:09:17 +00:00
Eric Christopher 5db9d661f0 This patch is in preparation for a substantial refactoring of the
code. To make the diffs easier to read, clang-format everything first.

No functionality changed.

Patch by Alina Sbirlea!

http://reviews.llvm.org/D20926

llvm-svn: 271595
2016-06-02 21:03:19 +00:00
Chandler Carruth d6c5bc2c81 Fix the use of sys::MemoryFence after including WindowsSupport.h that
r271558 introduced.

llvm-svn: 271563
2016-06-02 18:42:23 +00:00
Chandler Carruth dd1463823a This is yet another attempt to re-instate r220932 as discussed in
D19271.

Previous attempt was broken by NetBSD, so in this version I've made the
fallback path generic rather than Windows specific and sent both Windows
and NetBSD to it.

I've also re-formatted the code some, and used an exact clone of the
code in PassSupport.h for doing manual call-once using our atomics
rather than rolling a new one.

If this sticks, we can replace the fallback path for Windows with
a Windows-specific implementation that is more reliable.

Original commit message:
This patch adds an llvm_call_once which is a wrapper around
std::call_once on platforms where it is available and devoid
of bugs. The patch also migrates the ManagedStatic mutex to
be allocated using llvm_call_once.

These changes are philosophically equivalent to the changes
added in r219638, which were reverted due to a hang on Win32
which was the result of a bug in the Windows implementation
of std::call_once.

Differential Revision: http://reviews.llvm.org/D5922

llvm-svn: 271558
2016-06-02 18:22:12 +00:00
Chandler Carruth 49471dfb31 Remove all of the legacy home-grown atomic operations LLVM provided
except for CompareAndSwap. That is the only one still being used
anywhere now that statistics have been moved onto std::atomic.

Also, add a warning to the header that we shouldn't introduce more uses
of these old style atomics and instead should be using C++11's
std::atomic facilities.

Really hoping that we can hammer out the last couple of users here and
replace them with something more localized and/or principled, but
figured this was a pretty good start. =]

Note that this patch will need to be reverted if r271504 needs to be
reverted as that removes the last user of these. However, the biggest
risk for that patch was MSVC 2013 and at least one bot has already
passed where it would have failed there. I've tested MSVC 2015 using
their web interfaces and other platforms seem fine, so I'm optimistic.

Differential Revision: http://reviews.llvm.org/D20901

llvm-svn: 271540
2016-06-02 17:11:11 +00:00
Sanjoy Das ebeee960bd Use C++ casts to avoid a warning
warning: cast from type ‘const void*’ to type
‘llvm::PrettyStackTraceEntry*’ casts away qualifiers [-Wcast-qual]
   PrettyStackTraceHead = (PrettyStackTraceEntry*)Top;

llvm-svn: 271069
2016-05-27 23:04:28 +00:00
Benjamin Kramer 4fed928f53 Avoid some copies by using const references.
clang-tidy's performance-unnecessary-copy-initialization with some manual
fixes. No functional changes intended.

llvm-svn: 270988
2016-05-27 12:30:51 +00:00
Peter Collingbourne 1eaa97f439 Linker: teach the IR mover to return llvm::Error.
This will be needed in order to consistently return an Error
to clients of the API being developed in D20268.

Differential Revision: http://reviews.llvm.org/D20550

llvm-svn: 270967
2016-05-27 05:21:35 +00:00
Pete Cooper 18e9102a85 Don't allocate unnecessarily in APInt::operator[+-]. NFC.
APInt::operator+(uint64_t) just forwarded to operator+(const APInt&).

Constructing the APInt for the RHS takes an allocation which isn't
required.  Also, for APInt's in the slow path, operator+ would
call add() internally which iterates over both arrays of values.  Instead
we can use add_1 and sub_1 which only iterate while there is something to do.

Using the memory for 'opt -O2 verify-uselistorder.lto.opt.bc -o opt.bc'
(see r236629 for details), this reduces the number of allocations from
23.9M to 22.7M.

llvm-svn: 270959
2016-05-27 03:42:17 +00:00
Lang Hames bd8e954216 [Support] Rename unconvertibleErrorCode to inconvertibleErrorCode.
Based on a totally scientific, 30 second google search "in-" appears to be the
preferred prefix.

llvm-svn: 270950
2016-05-27 01:54:25 +00:00
Lang Hames c5e0bbd781 [Support] Add a StringError convenience class to Error.h
StringError can be used to represent Errors that aren't recoverable based on
the error type, but that have a useful error message that can be reported to
the user or logged.

llvm-svn: 270948
2016-05-27 01:37:32 +00:00
Richard Smith 3de58a5e42 Don't use recursion to print out the PrettyStackTrace after a crash. If the
crash was due to a stack overflow, chances are good that this would also cause
a stack overflow.

llvm-svn: 270903
2016-05-26 20:21:55 +00:00
Pete Cooper d6e6bf1808 Don't allocate in APInt::slt. NFC.
APInt::slt was copying the LHS and RHS in to temporaries then making
them unsigned so that it could use an unsigned comparision.  It did
this even on the paths which were trivial to give results for, such
as the sign bit of the LHS being set while RHS was not set.

This changes the logic to return out immediately in the trivial cases,
and use an unsigned comparison in the remaining cases.  But this time,
just use the unsigned comparison directly without creating any temporaries.

This works because, for example:
  true = (-2 slt -1) = (0xFE ult 0xFF)

Also added some tests explicitly for slt with APInt's larger than 64-bits
so that this new code is tested.

Using the memory for 'opt -O2 verify-uselistorder.lto.opt.bc -o opt.bc'
(see r236629 for details), this reduces the number of allocations from
26.8M to 23.9M.

llvm-svn: 270881
2016-05-26 17:40:07 +00:00
Chris Bieneman 1bc0f4395c [CMake] Restrict libxar linkage to just llvm-objdump
At some point we're going to need libObject to have this dependency, but as it is now this is causing too many headaches. This commit will reduce the linkage to just llvm-objdump where it is strictly needed, and we'll cross the libObject bridge later when we need it.

llvm-svn: 270866
2016-05-26 16:32:40 +00:00
Renato Golin dfc5d1ffab [AArch64] Adding a TargetParser for AArch64
There's already a ARMTargetParser,now adding a similar one for aarch64.
so we can use it to do ARCH/CPU/FPU parsing in clang and llvm, instead of
string comparison.

Patch by Jojo Ma.

llvm-svn: 270687
2016-05-25 12:02:33 +00:00
Gerolf Hoflehner 613e704190 [Support] Reapply cleanup r270643
llvm-svn: 270674
2016-05-25 06:23:45 +00:00
Gerolf Hoflehner 7c1841a55e [Support] revert previous commit r270643
llvm-svn: 270670
2016-05-25 05:51:05 +00:00
Gerolf Hoflehner 1ac739b2b5 [Support] Cleanup of an ancient Darwin work-around in Signals.inc (PR26174)
Patch by Jeremy Huddleston Sequoia

llvm-svn: 270643
2016-05-25 00:54:39 +00:00
Chris Bieneman 96e44e7c8a Add libxar to LLVM_SYSTEM_LIBS if available
This should fix PR27855. We have some terrible hacks in the CMake to add linking SYSTEM_LIBS to all tools. I think we need a better way to do this in the future.

llvm-svn: 270605
2016-05-24 20:19:05 +00:00
Peter Collingbourne 4718f8b5f1 Add FIXMEs to all derived classes of std::error_category.
This helps make clear that we're moving away from std::error_code.

Differential Revision: http://reviews.llvm.org/D20592

llvm-svn: 270604
2016-05-24 20:13:46 +00:00
Richard Smith b8b0be93ef Enable use of sigaltstack for signal handlers when available. With this,
backtraces from the signal handler on stack overflow now work reliably (on my
system at least...).

llvm-svn: 270395
2016-05-23 06:47:37 +00:00
Chris Bieneman f236347f54 Fix implicit type conversion. NFC.
llvm-svn: 270299
2016-05-21 00:36:47 +00:00
Richard Smith 4b735c54ad Switch from the linux-specific 'struct sigaltstack' to POSIX's 'stack_t'. This
is what I get for trusting my system's man pages I suppose.

llvm-svn: 270280
2016-05-20 21:38:15 +00:00
Richard Smith abab5d236d Add a configure-time check for the existence of sigaltstack. It seems that some
systems provide a <signal.h> that doesn't declare it.

llvm-svn: 270278
2016-05-20 21:26:00 +00:00
Richard Smith 14d965166c Reinstate r269992 (reverting r270267), but restricted to cases where glibc is
the C standard library implementation in use.

This works around a glibc bug in the backtrace() function where it fails to
produce a backtrace on x86_64 if libgcc / libunwind is statically linked.

llvm-svn: 270276
2016-05-20 21:18:12 +00:00
Richard Smith e88113414b Create a sigaltstack when we register our signal handlers. Otherwise we'd very
likely fail to produce a backtrace if we crash due to stack overflow.

llvm-svn: 270273
2016-05-20 21:07:41 +00:00
Chris Bieneman 6b58c4723a Revert "Work around a glibc bug: backtrace() spuriously fails if..."
This commit has been breaking the FreeBSD bots:
http://lab.llvm.org:8011/builders/lld-x86_64-freebsd

This reverts commit r269992.

llvm-svn: 270267
2016-05-20 20:15:17 +00:00
Richard Smith 61b41e0737 Work around a glibc bug: backtrace() spuriously fails if
- glibc is dynamically linked, and
 - libgcc_s is unavailable (for instance, another library is being used to
   provide the compiler runtime or libgcc is statically linked), and
 - the target is x86_64.

If we run backtrace() and it fails to find any stack frames, try using
_Unwind_Backtrace instead if available.

llvm-svn: 269992
2016-05-18 22:26:36 +00:00
Aaron Ballman 54269226ba Removing an unused variable introduced in r269911; NFC.
llvm-svn: 269915
2016-05-18 12:52:04 +00:00