Greg Clayton [Thu, 1 Dec 2016 18:56:29 +0000 (18:56 +0000)]
This change removes the dependency on DwarfDebug that was used for DW_FORM_ref_addr by making a new DIEUnit class in DIE.cpp.
The DIEUnit class represents a compile or type unit and it owns the unit DIE as an instance variable. This allows anyone with a DIE, to get the unit DIE, and then get back to its DIEUnit without adding any new ivars to the DIE class. Why was this needed? The DIE class has an Offset that is always the CU relative DIE offset, not the "offset in debug info section" as was commented in the header file (the comment has been corrected). This is great for performance because most DIE references are compile unit relative and this means most code that accessed the DIE's offset didn't need to make it into a compile unit relative offset because it already was. When we needed to emit a DW_FORM_ref_addr though, we needed to find the absolute offset of the DIE by finding the DIE's compile/type unit. This class did have the absolute debug info/type offset and could be added to the CU relative offset to compute the absolute offset. With this change we can easily get back to a DIE's DIEUnit which will have this needed offset. Prior to this is required having a DwarfDebug and required calling:
DwarfCompileUnit *DwarfDebug::lookupUnit(const DIE *CU) const;
Now we can use the DIEUnit class to do so without needing DwarfDebug. All clients now use DIEUnit objects (the DwarfDebug stack and the DwarfLinker). A follow on patch for the DWARF generator will also take advantage of this.
Differential Revision: https://reviews.llvm.org/D27170
llvm-svn: 288399
Alexey Bataev [Thu, 1 Dec 2016 18:42:42 +0000 (18:42 +0000)]
[SLP] Fixed cost model for horizontal reduction.
Currently when cost of scalar operations is evaluated the vector type is
used for scalar operations. Patch fixes this issue and fixes evaluation
of the vector operations cost.
Several test showed that vector cost model is too optimistic. It
allowed vectorization of 8 or less add/fadd operations, though scalar
code is faster. Actually, only for 16 or more operations vector code
provides better performance.
Differential Revision: https://reviews.llvm.org/D26277
llvm-svn: 288398
Mandeep Singh Grang [Thu, 1 Dec 2016 18:42:16 +0000 (18:42 +0000)]
[clang] Implement support for -defsym assembler option
Summary:
Adds support for -Wa,-defsym,abc=1 option.
Related llvm patch: https://reviews.llvm.org/D26214
Reviewers: rafael, t.p.northover, colinl, echristo, compnerd, rengolin
Subscribers: mehdi_amini
Tags: #clang-c
Differential Revision: https://reviews.llvm.org/D26213
llvm-svn: 288397
Mandeep Singh Grang [Thu, 1 Dec 2016 18:42:04 +0000 (18:42 +0000)]
[llvm] Implement support for -defsym assembler option
Summary:
Changes to llvm-mc to move common logic to separate function.
Related clang patch: https://reviews.llvm.org/D26213
Reviewers: rafael, t.p.northover, colinl, echristo, rengolin
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D26214
llvm-svn: 288396
Simon Pilgrim [Thu, 1 Dec 2016 18:27:19 +0000 (18:27 +0000)]
[X86][SSE] Moved shuffle mask widening/narrowing helper functions earlier in the file.
Will be necessary for a future patch.
llvm-svn: 288395
Arnold Schwaighofer [Thu, 1 Dec 2016 18:07:38 +0000 (18:07 +0000)]
swiftcc: Add an api to query whether a target ABI stores swifterror in a register
llvm-svn: 288394
Hans Wennborg [Thu, 1 Dec 2016 18:05:41 +0000 (18:05 +0000)]
Fix VS2015 build of clang-format-vsix by using NuGet to pull in required assemblies
Also added a gitignore to help track the right items to commit.
Patch by Antonio Maiorano <amaiorano@gmail.com>!
llvm-svn: 288393
Kostya Serebryany [Thu, 1 Dec 2016 18:02:07 +0000 (18:02 +0000)]
[libFuzzer] add a test for r288389 (-rss_limit_mb=0 means no limit).
llvm-svn: 288392
Ulrich Weigand [Thu, 1 Dec 2016 18:00:50 +0000 (18:00 +0000)]
[SystemZ] Fix fallout from r288374
Avoid undefined behavior due to too-large shift count.
llvm-svn: 288391
Weiming Zhao [Thu, 1 Dec 2016 18:00:36 +0000 (18:00 +0000)]
[AsmParser] Diagnose empty symbol for .set directive
Summary: Diagnose empty symbol to avoid hitting assertion in MCContext::getOrCreateSymbol
Reviewers: eli.friedman, rengolin
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D26728
llvm-svn: 288390
Kostya Serebryany [Thu, 1 Dec 2016 17:56:15 +0000 (17:56 +0000)]
[libFuzzer] treat -rss_limit_mb=0 as no limit
llvm-svn: 288389
George Burgess IV [Thu, 1 Dec 2016 17:52:39 +0000 (17:52 +0000)]
[TableGen] Ignore fake args for parsing-related arg counts.
We should complain about the following:
```
void foo() __attribute__((unavailable("a", "b")));
```
Instead, we currently just ignore "b". (...We also end up ignoring "a",
because we assume elsewhere that this attribute can only have 1 or 0
args.)
This happens because `unavailable` has a fake enum arg, and
`AttributeList::{getMinArgs,getMaxArgs}` include fake args in their
counts.
llvm-svn: 288388
Pavel Labath [Thu, 1 Dec 2016 17:48:51 +0000 (17:48 +0000)]
Remove another hack from the android toolchain file
This is no longer an issue with recent versions of the android ndk.
llvm-svn: 288387
Sean Callanan [Thu, 1 Dec 2016 17:46:51 +0000 (17:46 +0000)]
Handle UTF-16 and UTF-32 constant CFStrings
We have a longstanding issue where the expression parser does not handle wide CFStrings (e.g., @"凸凹") correctly, producing the useless error message
Internal error [IRForTarget]: An Objective-C constant string's string initializer is not an array
error: warning: expression result unused
error: The expression could not be prepared to run in the target
This is just a side effect of the fact that we don't handle wide string constants when converting these to CFStringCreateWithBytes. That function takes the string's encoding as an argument, so I made it work and added a testcase.
https://reviews.llvm.org/D27291
<rdar://problem/
13190557>
llvm-svn: 288386
Kuba Mracek [Thu, 1 Dec 2016 17:45:22 +0000 (17:45 +0000)]
Recommit r287403 (reverted in r287804): [lit] When setting SDKROOT on Darwin, use '--sdk macosx' to find the right SDK path.
This shouls now be safe and not break any more bots. It's strictly better to use '--sdk macosx', otherwise xcrun can return weird things for example when you have Command Line Tools or the SDK installed into '/'.
llvm-svn: 288385
Malcolm Parsons [Thu, 1 Dec 2016 17:38:54 +0000 (17:38 +0000)]
Add a blank line to make sphinx happy.
llvm-svn: 288384
Roger Ferrer Ibanez [Thu, 1 Dec 2016 17:36:41 +0000 (17:36 +0000)]
Protect sequences test under libcpp-no-exceptions
Replace throw with TEST_THROW and protect tests that do throw. Also add missing assert(false).
Differential Revision: https://reviews.llvm.org/D27252
llvm-svn: 288383
Roger Ferrer Ibanez [Thu, 1 Dec 2016 17:34:57 +0000 (17:34 +0000)]
Protect futures test under libcpp-no-exceptions
Skip tests that expect an exception be thrown.
Differential Revision: https://reviews.llvm.org/D27253
llvm-svn: 288382
Adam Nemet [Thu, 1 Dec 2016 17:34:50 +0000 (17:34 +0000)]
[GVN, OptDiag] Print the interesting instructions involved in missed load-elimination
[recommitting after the fix in r288307]
This includes the intervening store and the load/store that we're trying
to forward from in the optimization remark for the missed load
elimination.
This is hooked up under a new mode in ORE that allows for compile-time
budget for a bit more analysis to print more insightful messages. This
mode is currently enabled for -fsave-optimization-record (-Rpass is
trickier since it is controlled in the front-end).
With this we can now print the red remark in http://lab.llvm.org:8080/artifacts/opt-view_test-suite/build/SingleSource/Benchmarks/Dhrystone/CMakeFiles/dry.dir/html/_org_test-suite_SingleSource_Benchmarks_Dhrystone_dry.c.html#L446
Differential Revision: https://reviews.llvm.org/D26490
llvm-svn: 288381
Adam Nemet [Thu, 1 Dec 2016 17:34:44 +0000 (17:34 +0000)]
[GVN, OptDiag] Include the value that is forwarded in load elimination
[recommitting after the fix in r288307]
This requires some changes to the opt-diag API. Hal and I have
discussed this at the Dev Meeting and came up with a streaming delimiter
(setExtraArgs) to solve this.
Arguments after this delimiter are only included in the optimization
records and not in the remarks printed in the compiler output. (Note,
how in the test the content of the YAML file changes but the remarks on
the compiler output don't.)
This implements the green GVN message with a bug fix at line
http://lab.llvm.org:8080/artifacts/opt-view_test-suite/build/SingleSource/Benchmarks/Dhrystone/CMakeFiles/dry.dir/html/_org_test-suite_SingleSource_Benchmarks_Dhrystone_dry.c.html#L446
The fix is that now we properly include the constant value in the
message: "load of type i32 eliminated in favor of 7"
Differential Revision: https://reviews.llvm.org/D26489
llvm-svn: 288380
Roger Ferrer Ibanez [Thu, 1 Dec 2016 17:33:36 +0000 (17:33 +0000)]
Protect optional test under libcpp-no-exceptions
Replace throw with TEST_THROW and skip tests that throw exceptions
Differential Revision: https://reviews.llvm.org/D27254
llvm-svn: 288379
Roger Ferrer Ibanez [Thu, 1 Dec 2016 17:31:38 +0000 (17:31 +0000)]
Protect std::ostream::sentry test under libcpp-no-exceptions
Skip test that throws an exception.
Differential Revision: https://reviews.llvm.org/D27255
llvm-svn: 288378
Alexey Bataev [Thu, 1 Dec 2016 17:26:54 +0000 (17:26 +0000)]
[SLP] Additional tests with the cost of vector operations.
llvm-svn: 288377
Eric Liu [Thu, 1 Dec 2016 17:25:55 +0000 (17:25 +0000)]
[change-namespace] don't generate replacements for files that don't match file pattern.
Reviewers: hokein
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D27302
llvm-svn: 288376
Malcolm Parsons [Thu, 1 Dec 2016 17:24:42 +0000 (17:24 +0000)]
[clang-tidy] Rename modernize-use-default to modernize-use-equals-default
Reviewers: angelgarcia, aaron.ballman, alexfh
Subscribers: JDevlieghere, Prazek, mgorny, cfe-commits
Differential Revision: https://reviews.llvm.org/D26511
llvm-svn: 288375
Ulrich Weigand [Thu, 1 Dec 2016 17:10:27 +0000 (17:10 +0000)]
[SystemZ] Fix applyFixup for 12-bit fixups
Now that we have fixups that only fill parts of a byte, it turns
out we have to mask off the bits outside the fixup area when
applying them. Failing to do so caused invalid object code to
be emitted for bprp with a negative 12-bit displacement.
llvm-svn: 288374
Rui Ueyama [Thu, 1 Dec 2016 17:09:04 +0000 (17:09 +0000)]
Parallelize ICF to make LLD's ICF really fast.
ICF is short for Identical Code Folding. It is a size optimization to
identify two or more functions that happened to have the same contents
to merges them. It usually reduces output size by a few percent.
ICF is slow because it is computationally intensive process. I tried
to paralellize it before but failed because I couldn't make a
parallelized version produce consistent outputs. Although it didn't
create broken executables, every invocation of the linker generated
slightly different output, and I couldn't figure out why.
I think I now understand what was going on, and also came up with a
simple algorithm to fix it. So is this patch.
The result is very exciting. Chromium for example has 780,662 input
sections in which 20,774 are reducible by ICF. LLD previously took
7.980 seconds for ICF. Now it finishes in 1.065 seconds.
As a result, LLD can now link a Chromium binary (output size 1.59 GB)
in 10.28 seconds on my machine with ICF enabled. Compared to gold
which takes 40.94 seconds to do the same thing, this is an amazing
number.
From here, I'll describe what we are doing for ICF, what was the
previous problem, and what I did in this patch.
In ICF, two sections are considered identical if they have the same
section flags, section data, and relocations. Relocations are tricky,
becuase two relocations are considered the same if they have the same
relocation type, values, and if they point to the same section _in
terms of ICF_.
Here is an example. If foo and bar defined below are compiled to the
same machine instructions, ICF can (and should) merge the two,
although their relocations point to each other.
void foo() { bar(); }
void bar() { foo(); }
This is not an easy problem to solve.
What we are doing in LLD is some sort of coloring algorithm. We color
non-identical sections using different colors repeatedly, and sections
in the same color when the algorithm terminates are considered
identical. Here is the details:
1. First, we color all sections using their hash values of section
types, section contents, and numbers of relocations. At this moment,
relocation targets are not taken into account. We just color
sections that apparently differ in different colors.
2. Next, for each color C, we visit sections having color C to see
if their relocations are the same. Relocations are considered equal
if their targets have the same color. We then recolor sections that
have different relocation targets in new colors.
3. If we recolor some section in step 2, relocations that were
previously pointing to the same color targets may now be pointing to
different colors. Therefore, repeat 2 until a convergence is
obtained.
Step 2 is a heavy operation. For Chromium, the first iteration of step
2 takes 2.882 seconds, and the second iteration takes 1.038 seconds,
and in total it needs 23 iterations.
Parallelizing step 1 is easy because we can color each section
independently. This patch does that.
Parallelizing step 2 is tricky. We could work on each color
independently, but we cannot recolor sections in place, because it
will break the invariance that two possibly-identical sections must
have the same color at any moment.
Consider sections S1, S2, S3, S4 in the same color C, where S1 and S2
are identical, S3 and S4 are identical, but S2 and S3 are not. Thread
A is about to recolor S1 and S2 in C'. After thread A recolor S1 in
C', but before recolor S2 in C', other thread B might observe S1 and
S2. Then thread B will conclude that S1 and S2 are different, and it
will split thread B's sections into smaller groups wrongly. Over-
splitting doesn't produce broken results, but it loses a chance to
merge some identical sections. That was the cause of indeterminism.
To fix the problem, I made sections have two colors, namely current
color and next color. At the beginning of each iteration, both colors
are the same. Each thread reads from current color and writes to next
color. In this way, we can avoid threads from reading partial
results. After each iteration, we flip current and next.
This is a very simple solution and is implemented in less than 50
lines of code.
I tested this patch with Chromium and confirmed that this parallelized
ICF produces the identical output as the non-parallelized one.
Differential Revision: https://reviews.llvm.org/D27247
llvm-svn: 288373
Dominic Chen [Thu, 1 Dec 2016 17:06:39 +0000 (17:06 +0000)]
[analyzer] Drop explicit mention of range constraint solver
Summary: The basic constraint solver was dropped in rL162384, leaving the range constraint solver as the default and only constraint solver. Explicitly specifying it is unnecessary, and makes it difficult to test with other solver backends.
Reviewers: zaks.anna, dcoughlin
Subscribers: cfe-commits
Differential Revision: https://reviews.llvm.org/D26694
llvm-svn: 288372
Alexey Bataev [Thu, 1 Dec 2016 16:45:04 +0000 (16:45 +0000)]
Revert "[SLP] Additional tests with the cost of vector operations."
This reverts commit
a61718435fc4118c82f8aa6133fd81f803789c1e.
llvm-svn: 288371
Adam Nemet [Thu, 1 Dec 2016 16:40:32 +0000 (16:40 +0000)]
[GVN] Basic optimization remark support
[recommitting after the fix in r288307]
Follow-on patches will add more interesting cases.
The goal of this patch-set is to get the GVN messages printed in
opt-viewer from Dhrystone as was presented in my Dev Meeting talk. This
is the optimization view for the function (the last remark in the
function has a bug which is fixed in this series):
http://lab.llvm.org:8080/artifacts/opt-view_test-suite/build/SingleSource/Benchmarks/Dhrystone/CMakeFiles/dry.dir/html/_org_test-suite_SingleSource_Benchmarks_Dhrystone_dry.c.html#L430
Differential Revision: https://reviews.llvm.org/D26488
llvm-svn: 288370
Alexey Bataev [Thu, 1 Dec 2016 16:11:48 +0000 (16:11 +0000)]
[SLP] Additional tests with the cost of vector operations.
llvm-svn: 288369
Andrey Churbanov [Thu, 1 Dec 2016 16:08:52 +0000 (16:08 +0000)]
fixed type in Windows-specific code
llvm-svn: 288368
Simon Pilgrim [Thu, 1 Dec 2016 16:00:14 +0000 (16:00 +0000)]
[X86][SSE] Classify AND bitmasks as variable shuffle masks
They are loading the bitmasks from the constant pool so the cost is similar to loading a shuffle mask.
llvm-svn: 288367
Manuel Klimek [Thu, 1 Dec 2016 15:45:06 +0000 (15:45 +0000)]
Adds hasUnqualifiedDesugaredType to allow matching through type sugar.
Differential Revision: https://reviews.llvm.org/D27207
llvm-svn: 288366
Simon Pilgrim [Thu, 1 Dec 2016 15:41:40 +0000 (15:41 +0000)]
[X86][SSE] Add support for combining AND bitmasks to shuffles.
llvm-svn: 288365
Pavel Labath [Thu, 1 Dec 2016 15:20:34 +0000 (15:20 +0000)]
Remove iostream include from WasmObjectFile
The file does not seems to use c++ iostreams (and is is llvm policy to avoid
that). Committing as obvious.
llvm-svn: 288364
Asaf Badouh [Thu, 1 Dec 2016 15:19:10 +0000 (15:19 +0000)]
[LMT] Restrict nop length to one
not all lakemont MCU support long nop.
we can't assume we can generate long nop by default for MCU.
Differential Revision: https://reviews.llvm.org/D26895
llvm-svn: 288363
Simon Pilgrim [Thu, 1 Dec 2016 14:56:48 +0000 (14:56 +0000)]
[X86][SSE] Tidied up filecheck prefixes for uitofp fast-math tests.
They should be in 'narrowing' order from common to more specific test prefixes.
llvm-svn: 288338
Daniel Jasper [Thu, 1 Dec 2016 14:33:50 +0000 (14:33 +0000)]
Silence GCC's -Wenum-compare after r288335 in the same way it is done
in X86FastISel.cpp.
llvm-svn: 288337
Nicolai Haehnle [Thu, 1 Dec 2016 14:04:13 +0000 (14:04 +0000)]
[SelectionDAG] Rename and clarify visitFMULForFMADCombine (NFC)
Summary: Suggested by @spatel in D26602.
Reviewers: spatel, hfinkel
Subscribers: spatel, llvm-commits
Differential Revision: https://reviews.llvm.org/D27260
llvm-svn: 288336
Simon Pilgrim [Thu, 1 Dec 2016 13:47:02 +0000 (13:47 +0000)]
[X86][SSE] Add support for combining target shuffles to AND bitmasks.
llvm-svn: 288335
Alex Lorenz [Thu, 1 Dec 2016 12:14:38 +0000 (12:14 +0000)]
[ObjC] Avoid a @try/@finally/@autoreleasepool fixit when parsing an expression
This patch ensures that the typo fixit for the @try/@finally/@autoreleasepool {}
directive is shown only when we're parsing an actual statement where such
directives can actually be present.
rdar://
19669565
Differential Revision: https://reviews.llvm.org/D26916
llvm-svn: 288334
Simon Pilgrim [Thu, 1 Dec 2016 11:52:37 +0000 (11:52 +0000)]
[X86][SSE] Add support for combining ISD::AND with shuffles.
Attempts to convert an AND with a vector of 255 or 0 values into a shuffle (blend) mask.
llvm-svn: 288333
Joey Gouly [Thu, 1 Dec 2016 11:30:49 +0000 (11:30 +0000)]
[OpenCL] Refactor read_only/write_only pipes.
This adds the access qualifier to the Pipe Type, rather than using a class
hierarchy.
It also fixes mergeTypes for Pipes, by disallowing merges. Only identical
pipe types can be merged. The test case in invalid-pipes-cl2.0.cl is added
to check that.
llvm-svn: 288332
Pavel Labath [Thu, 1 Dec 2016 11:30:08 +0000 (11:30 +0000)]
Remove a hack from the Android toolchain file
Summary:
The fix is to make sure llvm does not pull in dlopen() in configurations where it
is not available.
Reviewers: tberghammer, beanz
Subscribers: danalbert, srhines, lldb-commits, mgorny
Differential Revision: https://reviews.llvm.org/D26505
llvm-svn: 288331
Simon Pilgrim [Thu, 1 Dec 2016 11:26:07 +0000 (11:26 +0000)]
[X86][SSE] Added tests showing missed combines of shuffles with ANDs.
llvm-svn: 288330
Johannes Doerfert [Thu, 1 Dec 2016 11:12:14 +0000 (11:12 +0000)]
[NFC] Check for feasibility prior to the profitability check
Feasibility is checked late on its own but early it is hidden behind
the "PollyProcessUnprofitable" guard. This change will make sure we opt
out early if the runtime context is infeasible anyway.
llvm-svn: 288329
Johannes Doerfert [Thu, 1 Dec 2016 11:10:45 +0000 (11:10 +0000)]
[FIX] Do not try to hoist obviously overwritten loads
llvm-svn: 288328
Florian Hahn [Thu, 1 Dec 2016 11:02:59 +0000 (11:02 +0000)]
Fix crash with unsupported architectures in Linux/Gnu target triples.
Summary: This patch adds a check and an error message to gnutools::Linker::ConstructJob in case the architecture is not supported. For most other operating systems, the error message is created in lib/Basic/Targets.cpp:AllocateTarget, but when construction the linker arguments for the gnutools linker a supported architecture is required.
Reviewers: rafael, joerg, echristo
Subscribers: mehdi_amini, joerg, dschuff, cfe-commits
Differential Revision: https://reviews.llvm.org/D27066
llvm-svn: 288327
Pavel Labath [Thu, 1 Dec 2016 10:57:30 +0000 (10:57 +0000)]
Use Timeout<> in Process::RunThreadPlan
Summary:
Since the function is way too big already, I tried at least to factor out the
timeout computation stuff into a separate function. I've tried to make the new
code semantically equivalent, and it also makes sense when I look at it as a done
deal.
Reviewers: jingham
Subscribers: lldb-commits
Differential Revision: https://reviews.llvm.org/D27258
llvm-svn: 288326
Davide Italiano [Thu, 1 Dec 2016 08:48:14 +0000 (08:48 +0000)]
[SCCP] Switch over to DEBUG() and drop an #ifdef.
llvm-svn: 288325
Davide Italiano [Thu, 1 Dec 2016 08:36:12 +0000 (08:36 +0000)]
[SCCP] Prefer `auto` when the type is obvious. NFCI.
llvm-svn: 288324
Tobias Grosser [Thu, 1 Dec 2016 08:08:47 +0000 (08:08 +0000)]
Add newline at end of debug print
In '[DBG] Allow to emit the RTC value at runtime' the diagnostics were printed
without a newline at the end of each diagnostic. We add such a newline to
improve readability.
llvm-svn: 288323
Eric Christopher [Thu, 1 Dec 2016 07:50:12 +0000 (07:50 +0000)]
Temporarily Revert "Move most EH from MachineModuleInfo to MachineFunction"
This apprears to have broken the global isel bot:
http://lab.llvm.org:8080/green/job/clang-stage1-cmake-RA-globalisel_build/5174/console
This reverts commit r288293.
llvm-svn: 288322
Peter Collingbourne [Thu, 1 Dec 2016 07:00:35 +0000 (07:00 +0000)]
Object: Set SF_Indirect in ModuleSymbolTable.
This lets us remove the last use of IRObjectFile::getSymbolGV() in llvm-nm.
Differential Revision: https://reviews.llvm.org/D27076
llvm-svn: 288321
Peter Collingbourne [Thu, 1 Dec 2016 06:53:47 +0000 (06:53 +0000)]
Object: Add SF_Executable symbol flag.
This allows us to remove a few uses of IRObjectFile::getSymbolGV() in
llvm-nm.
While here change host-dependent logic in llvm-nm to target-dependent
logic.
Differential Revision: https://reviews.llvm.org/D27075
llvm-svn: 288320
Peter Collingbourne [Thu, 1 Dec 2016 06:51:47 +0000 (06:51 +0000)]
Object: Extract a ModuleSymbolTable class from IRObjectFile.
This class represents a symbol table built from in-memory IR. It provides
access to GlobalValues and should only be used if such access is required
(e.g. in the LTO implementation). We will eventually change IRObjectFile
to read from a bitcode symbol table rather than using ModuleSymbolTable,
so it would not be able to expose the module.
Differential Revision: https://reviews.llvm.org/D27073
llvm-svn: 288319
Peter Collingbourne [Thu, 1 Dec 2016 06:21:08 +0000 (06:21 +0000)]
Bitcode: The index used by ModuleSummaryIndexBitcodeReader is now required, so make it a reference. NFCI.
llvm-svn: 288318
Peter Collingbourne [Thu, 1 Dec 2016 06:00:53 +0000 (06:00 +0000)]
Bitcode: Introduce BitcodeModule::{has,get}Summary().
These are equivalent to hasGlobalValueSummary() and getModuleSummaryIndex().
Differential Revision: https://reviews.llvm.org/D27242
llvm-svn: 288317
Peter Collingbourne [Thu, 1 Dec 2016 05:52:32 +0000 (05:52 +0000)]
LTO: Remove ModuleLoader, make loadModuleFromBuffer static and move into its only client, ThinLTOCodeGenerator.
This is no longer the recommended way to load modules for importing, so it should not be public API.
Differential Revision: https://reviews.llvm.org/D27292
llvm-svn: 288316
Peter Collingbourne [Thu, 1 Dec 2016 05:47:58 +0000 (05:47 +0000)]
Bitcode: Correctly handle Fixed and VBR arrays in BitstreamCursor::skipRecord().
The assertions were wrong; we need to call getEncodingData() on the element,
not the array. While here, simplify the skipRecord() implementation for Fixed
and Char6 arrays. This is tested by the code I added to llvm-bcanalyzer
which makes sure that we can skip any record.
Differential Revision: https://reviews.llvm.org/D27241
llvm-svn: 288315
Sean Silva [Thu, 1 Dec 2016 05:43:48 +0000 (05:43 +0000)]
Add `isRelExprOneOf` helper
In various places in LLD's hot loops, we have expressions of the form
"E == R_FOO || E == R_BAR || ..." (E is a RelExpr).
Some of these expressions are quite long, and even though they usually go just
a very small number of ways and so should be well predicted, they can still
occupy branch predictor resources harming other parts of the code, or they
won't be predicted well if they overflow branch predictor resources or if the
branches are too dense and the branch predictor can't track them all (the
compiler can in theory avoid this, at a cost in text size). And some of these
expressions are so large and executed so frequently that even when
well-predicted they probably still have a nontrivial cost.
This speedup should be pretty portable. The cost of these simple bit tests is
independent of:
- the target we are linking for
- the distribution of RelExpr's for a given link (which can depend on how the
input files were compiled)
- what compiler was used to compile LLD (it is just a simple bit test;
hopefully the compiler gets it right!)
- adding new target-dependent relocations (e.g. needsPlt doesn't pay any extra
cost checking R_PPC_PLT_OPD on x86-64 builds)
I did some rough measurements on clang-fsds and this patch gives over about 4%
speedup for a regular -O1 link, about 2.5% for -O3 --gc-sections and over 5%
for -O0. Sorry, I don't have my current machine set up for doing really
accurate measurements right now.
This also is just a bit cleaner. Thanks for Joerg for suggesting for
this approach.
Differential Revision: https://reviews.llvm.org/D27156
llvm-svn: 288314
John McCall [Thu, 1 Dec 2016 05:33:30 +0000 (05:33 +0000)]
Teach ConstantBuilder how to emit a reference to the current position
that will be filled in when the initializer is set.
llvm-svn: 288313
Philip Reames [Thu, 1 Dec 2016 04:37:35 +0000 (04:37 +0000)]
Revert previous whitespace change
llvm-svn: 288312
Philip Reames [Thu, 1 Dec 2016 04:37:09 +0000 (04:37 +0000)]
Test commit of whitespace to check permissions.
llvm-svn: 288311
Rui Ueyama [Thu, 1 Dec 2016 04:36:54 +0000 (04:36 +0000)]
Simplify ScriptParser.
- Rename currentBuffer -> getCurrentMB to start it with verb.
- Simplify containsString.
- Add llvm_unreachable at end of getCurrentMB.
llvm-svn: 288310
Rui Ueyama [Thu, 1 Dec 2016 04:36:51 +0000 (04:36 +0000)]
Do not name a variable Ret which is not a return value.
llvm-svn: 288309
Rui Ueyama [Thu, 1 Dec 2016 04:36:49 +0000 (04:36 +0000)]
Make get{Line,Column}Number members of StringParser.
This patch also renames currentLocation getCurrentLocation.
llvm-svn: 288308
Adam Nemet [Thu, 1 Dec 2016 03:56:43 +0000 (03:56 +0000)]
[GVN] When merging blocks update LoopInfo if it's available
If LoopInfo is available during GVN, BasicAA will use it. However
MergeBlockIntoPredecessor does not update LI as it merges blocks.
This didn't use to cause problems because LI was freed before
GVN/BasicAA. Now with OptimizationRemarkEmitter, the lifetime of LI is
extended so LI needs to be kept up-to-date during GVN.
Differential Revision: https://reviews.llvm.org/D27288
llvm-svn: 288307
Rui Ueyama [Thu, 1 Dec 2016 03:56:27 +0000 (03:56 +0000)]
Split getPos into getLineNumber and getColumnNumber.
llvm-svn: 288306
Richard Smith [Thu, 1 Dec 2016 03:32:42 +0000 (03:32 +0000)]
P0012R1: add Itanium ABI support for throwing non-noexcept function pointers and catching as noexcept.
llvm-svn: 288305
Richard Smith [Thu, 1 Dec 2016 03:04:07 +0000 (03:04 +0000)]
Revert r285664, cxx-abi-dev chose to go in a different direction for the ABI here.
llvm-svn: 288304
Ivan Krasin [Thu, 1 Dec 2016 02:54:54 +0000 (02:54 +0000)]
Use trigrams to speed up SpecialCaseList.
Summary:
it's often the case when the rules in the SpecialCaseList
are of the form hel.o*bar. That gives us a chance to build
trigram index to quickly discard 99% of inputs without
running a full regex. A similar idea was used in Google Code Search
as described in the blog post:
https://swtch.com/~rsc/regexp/regexp4.html
The check is defeated, if there's at least one regex
more complicated than that. In this case, all inputs
will go through the regex. That said, the real-world
rules are often simple or can be simplied. That considerably
speeds up compiling Chromium with CFI and UBSan.
As measured on Chromium's content_message_generator.cc:
before, CFI: 44 s
after, CFI: 23 s
after, CFI, no blacklist: 23 s (~1% slower, but 3 runs were unable to show the difference)
after, regular compilation to bitcode: 23 s
Reviewers: pcc
Subscribers: mgorny, llvm-commits
Differential Revision: https://reviews.llvm.org/D27188
llvm-svn: 288303
Peter Collingbourne [Thu, 1 Dec 2016 02:51:12 +0000 (02:51 +0000)]
LTO: Remove Symbol::getIRName().
Its only use was in the LTO implementation. Also document
Symbol::getName().
llvm-svn: 288302
Richard Smith [Thu, 1 Dec 2016 02:11:49 +0000 (02:11 +0000)]
PR31081: ignore exception specifications when deducing function template
arguments from a declaration; despite what the standard says, this form of
deduction should not be considering exception specifications.
llvm-svn: 288301
Kostya Serebryany [Thu, 1 Dec 2016 02:06:56 +0000 (02:06 +0000)]
revert r288283 as it causes debug info (line numbers) to be lost in instrumented code. also revert r288299 which was a workaround for the problem.
llvm-svn: 288300
Kostya Serebryany [Thu, 1 Dec 2016 01:33:44 +0000 (01:33 +0000)]
[libFuzzer] temporary disable a part of the test broken by r288283
llvm-svn: 288299
Rui Ueyama [Thu, 1 Dec 2016 01:22:48 +0000 (01:22 +0000)]
Dump Codeview type information correctly.
llvm-svn: 288298
George Burgess IV [Thu, 1 Dec 2016 00:13:18 +0000 (00:13 +0000)]
[TableGen] Minor clean-ups. NFC.
Primarily: try to use DenseSet<StringRef> instead of
std::set<std::string>, and use pretty range algos where we can.
Small sizes were arbitrarily chosen.
llvm-svn: 288297
Derek Schuff [Thu, 1 Dec 2016 00:11:15 +0000 (00:11 +0000)]
[WebAssembly] Emit .import_global assembler directives
Support a new assembler directive, .import_global, to declare imported
global variables (i.e. those with external linkage and no
initializer). The linker turns these into wasm imports.
Patch by Jacob Gravelle
Differential Revision: https://reviews.llvm.org/D26875
llvm-svn: 288296
John McCall [Wed, 30 Nov 2016 23:54:50 +0000 (23:54 +0000)]
Clean up CGObjCMac's APIs for deriving class references. NFC.
llvm-svn: 288295
Kelvin Li [Wed, 30 Nov 2016 23:51:03 +0000 (23:51 +0000)]
[OpenMP] Sema and parsing for 'teams distribute parallel for simd' pragma
This patch is to implement sema and parsing for 'teams distribute parallel for simd' pragma.
Differential Revision: https://reviews.llvm.org/D27084
llvm-svn: 288294
Matthias Braun [Wed, 30 Nov 2016 23:49:01 +0000 (23:49 +0000)]
Move most EH from MachineModuleInfo to MachineFunction
Most of the exception handling members in MachineModuleInfo is actually
per function data (talks about the "current function") so it is better
to keep it at the function instead of the module.
This is a necessary step to have machine module passes work properly.
Also:
- Rename TidyLandingPads() to tidyLandingPads()
- Use doxygen member groups instead of "//===- EH ---"... so it is clear
where a group ends.
- I had to add an ugly const_cast at two places in the AsmPrinter
because the available MachineFunction pointers are const, but the code
wants to call tidyLandingPads() in between
(markFunctionEnd()/endFunction()).
Differential Revision: https://reviews.llvm.org/D27227
llvm-svn: 288293
Matthias Braun [Wed, 30 Nov 2016 23:48:50 +0000 (23:48 +0000)]
Move VariableDbgInfo from MachineModuleInfo to MachineFunction
VariableDbgInfo is per function data, so it makes sense to have it with
the function instead of the module.
This is a necessary step to have machine module passes work properly.
Differential Revision: https://reviews.llvm.org/D27186
llvm-svn: 288292
Matthias Braun [Wed, 30 Nov 2016 23:48:42 +0000 (23:48 +0000)]
Move FrameInstructions from MachineModuleInfo to MachineFunction
This is per function data so it is better kept at the function instead
of the module.
This is a necessary step to have machine module passes work properly.
Differential Revision: https://reviews.llvm.org/D27185
llvm-svn: 288291
Matthias Braun [Wed, 30 Nov 2016 23:48:26 +0000 (23:48 +0000)]
MCStreamer: Use "cfi" for CFI related temp labels.
Choosing a "cfi" name makes the intend a bit clearer in an assembly dump
and more importantly the assembly dumps are slightly more stable as the
numbers don't move around anymore when unrelated code calls
createTempSymbol() more or less often.
As they are temp labels the name doesn't influence the generated object
code.
Differential Revision: https://reviews.llvm.org/D27244
llvm-svn: 288290
John McCall [Wed, 30 Nov 2016 23:25:13 +0000 (23:25 +0000)]
Introduce a type-safe enum for ForDefinition.
llvm-svn: 288289
Peter Collingbourne [Wed, 30 Nov 2016 23:19:05 +0000 (23:19 +0000)]
llvm-lto2: Simpler workaround for PR30396.
Maintain the command line resolutions as a map to a list of resolutions
rather than a single resolution, and apply the resolutions in the order
observed. This is not only simpler but allows us to test the scenario where
the two symbols have different resolutions.
Differential Revision: https://reviews.llvm.org/D27285
llvm-svn: 288288
John McCall [Wed, 30 Nov 2016 23:15:55 +0000 (23:15 +0000)]
Fix some layering violations where CGObjCMac's NSString emission was
performed at the CodeGenModule level.
Would be NFC except we now also use a different uniquing structure so
that we don't get spurious conflicts if you ask for both an NSString
and a CFString for the same content (which is possible with builtins).
llvm-svn: 288287
Paul Robinson [Wed, 30 Nov 2016 23:14:27 +0000 (23:14 +0000)]
[PS4] Tighten up a triple check.
llvm-svn: 288286
Eugene Zelenko [Wed, 30 Nov 2016 23:10:42 +0000 (23:10 +0000)]
Fix some Clang-tidy modernize-use-default and Include What You Use warnings; other minor fixes (NFC).
llvm-svn: 288285
Jason Molenda [Wed, 30 Nov 2016 23:00:52 +0000 (23:00 +0000)]
Add another address to look for the kernel load addr in
DynamicLoaderDarwinKernel::SearchForKernelWithDebugHints
on 32-bit devices.
<rdar://problem/
29423290>
llvm-svn: 288284
Paul Robinson [Wed, 30 Nov 2016 22:49:55 +0000 (22:49 +0000)]
Recommit r288212: Emit 'no line' information for interesting 'orphan' instructions.
The LLDB tests are now ready for this patch.
DWARF specifies that "line 0" really means "no appropriate source
location" in the line table. Use this for branch targets and some
other cases that have no specified source location, to prevent
inheriting unfortunate line numbers from physically preceding
instructions (which might be from completely unrelated source).
Differential Revision: http://reviews.llvm.org/D24180
llvm-svn: 288283
Paul Robinson [Wed, 30 Nov 2016 22:47:25 +0000 (22:47 +0000)]
PR31214: Make a test tolerate "line 0" when stepping by instruction.
Differential Revision: http://reviews.llvm.org/D27282
llvm-svn: 288282
Kostya Serebryany [Wed, 30 Nov 2016 22:39:35 +0000 (22:39 +0000)]
[libFuzzer] extend -rss_limit_mb to crash instantly on a single malloc that exceeds the limit
llvm-svn: 288281
David Callahan [Wed, 30 Nov 2016 22:32:58 +0000 (22:32 +0000)]
Only computeRelativePath() on new members
Summary:
When using thin archives, and processing the same archive multiple times, we were mangling existing entries. The root cause is that we were calling computeRelativePath() more than once. Here, we only call it when adding new members to an archive.
Note that D27218 changes the way thin archives are printed, and will break the new unit test included here. Depending on which one lands first, the other will need to be slightly modified.
Reviewers: rafael, davide
Subscribers: llvm-commits
Differential Revision: https://reviews.llvm.org/D27217
llvm-svn: 288280
Joel Jones [Wed, 30 Nov 2016 22:25:24 +0000 (22:25 +0000)]
[AArch64] Refactor LSE support as feature separate from V8.1a support.
Summary:
This is preparation for ThunderX processors that have Large
System Extension (LSE) atomic instructions, but not the
other instructions introduced by V8.1a.
This will mimic changes to GCC as described here:
https://gcc.gnu.org/ml/gcc-patches/2015-06/msg00388.html
LSE instructions are: LD/ST<op>, CAS*, SWP
Reviewers: t.p.northover, echristo, jmolloy, rengolin
Subscribers: aemerson, mehdi_amini
Differential Revision: https://reviews.llvm.org/D26621
llvm-svn: 288279
Evgeny Stupachenko [Wed, 30 Nov 2016 22:23:51 +0000 (22:23 +0000)]
Fix LSR best register search algorithm.
Summary:
Fix a case when first register in a search has maximum
RegUses.getUsedByIndices(Reg).count()
Reviewers: qcolombet
Differential Revision: http://reviews.llvm.org/D26877
From: Evgeny Stupachenko <evstupac@gmail.com>
llvm-svn: 288278
Matthias Braun [Wed, 30 Nov 2016 22:17:10 +0000 (22:17 +0000)]
Clarify rules for reserved regs, fix aarch64 ones.
No test case necessary as the problematic condition is checked with the
newly introduced assertAllSuperRegsMarked() function.
Differential Revision: https://reviews.llvm.org/D26648
llvm-svn: 288277
Kostya Serebryany [Wed, 30 Nov 2016 21:53:32 +0000 (21:53 +0000)]
[libFuzzer] extend -print_coverage to print the comma-separated list of covered dirs. Note: the Windows stub for DirName is left unimplemented
llvm-svn: 288276