Chandler Carruth [Tue, 4 Dec 2012 09:53:37 +0000 (09:53 +0000)]
Really sort the #include lines in unittests/...
I forgot to re-sort after fixing main module headers.
llvm-svn: 169244
Chandler Carruth [Tue, 4 Dec 2012 09:45:34 +0000 (09:45 +0000)]
Sort the #include lines for unittests/...
I've tried to place sensible headers at the top as main-module headers.
llvm-svn: 169243
Chandler Carruth [Tue, 4 Dec 2012 09:44:38 +0000 (09:44 +0000)]
Teach the include sorting script about the gtest headers; sort them with
the system headers.
llvm-svn: 169242
Chandler Carruth [Tue, 4 Dec 2012 09:37:22 +0000 (09:37 +0000)]
Sort the #include lines for examples/...
llvm-svn: 169241
Chandler Carruth [Tue, 4 Dec 2012 09:25:21 +0000 (09:25 +0000)]
Sort #include lines for tools/...
Completely automated with sort_includes.py
llvm-svn: 169240
Chandler Carruth [Tue, 4 Dec 2012 09:21:50 +0000 (09:21 +0000)]
Update the #include lines of the extra tools.
llvm-svn: 169239
Chandler Carruth [Tue, 4 Dec 2012 09:18:49 +0000 (09:18 +0000)]
Sort #include lines for all files under include/...
This is a simpler sort, entirely automatic with the help of
llvm/utils/sort_includes.py -- no manual edits here.
llvm-svn: 169238
Chandler Carruth [Tue, 4 Dec 2012 09:13:33 +0000 (09:13 +0000)]
Sort all of Clang's files under 'lib', and fix up the broken headers
uncovered.
This required manually correcting all of the incorrect main-module
headers I could find, and running the new llvm/utils/sort_includes.py
script over the files.
I also manually added quite a few missing headers that were uncovered by
shuffling the order or moving headers up to be main-module-headers.
llvm-svn: 169237
NAKAMURA Takumi [Tue, 4 Dec 2012 08:20:41 +0000 (08:20 +0000)]
ASTTests, ASTMatchersTests: Move clangEdit before clangAst in USEDLIB.
llvm-svn: 169236
NAKAMURA Takumi [Tue, 4 Dec 2012 08:20:35 +0000 (08:20 +0000)]
Untabify (in USEDLIBS, Makefile(s)).
llvm-svn: 169235
Kostya Serebryany [Tue, 4 Dec 2012 07:54:41 +0000 (07:54 +0000)]
[tsan] refactoring and comment changes in sanitizer_common/sanitizer_allocator{,64}.h. No functionality change
llvm-svn: 169234
NAKAMURA Takumi [Tue, 4 Dec 2012 07:40:33 +0000 (07:40 +0000)]
clang/Lex: [CMake] Update CMakefiles since r169229.
llvm-svn: 169233
Bill Wendling [Tue, 4 Dec 2012 07:33:40 +0000 (07:33 +0000)]
Don't test for ASM output but for IR output.
llvm-svn: 169232
Dmitry Vyukov [Tue, 4 Dec 2012 07:28:25 +0000 (07:28 +0000)]
tsan: better error message if we fail to intercept some function
currently the message is SIGSEGV
llvm-svn: 169231
Dmitry Vyukov [Tue, 4 Dec 2012 07:27:32 +0000 (07:27 +0000)]
tsan: check if PWD env var is absent
On some programs I see:
failed to open suppressions file '<null>/testing/tsan/v2/tsan.supp'
llvm-svn: 169230
Argyrios Kyrtzidis [Tue, 4 Dec 2012 07:27:05 +0000 (07:27 +0000)]
Refactor recording the preprocessor conditional directive regions out of
PreprocessingRecord and into its own class, PPConditionalDirectiveRecord.
Decoupling allows a client to use the functionality of PPConditionalDirectiveRecord
without needing a PreprocessingRecord.
llvm-svn: 169229
Argyrios Kyrtzidis [Tue, 4 Dec 2012 07:26:53 +0000 (07:26 +0000)]
In the PreprocessingRecord, to identify the different conditional directive regions
use the SourceLocation at the start of the respective region, instead of a unique integer.
llvm-svn: 169228
Argyrios Kyrtzidis [Tue, 4 Dec 2012 07:26:48 +0000 (07:26 +0000)]
[libclang] Avoid copying the CompileCommand related strings when wrapping them to a CXString.
llvm-svn: 169227
Argyrios Kyrtzidis [Tue, 4 Dec 2012 07:26:44 +0000 (07:26 +0000)]
Introduce CompilationDatabase::getAllCompileCommands() that returns all
compile commands of the database and expose it via the libclang API.
llvm-svn: 169226
NAKAMURA Takumi [Tue, 4 Dec 2012 07:25:24 +0000 (07:25 +0000)]
VMCoreTests/PassManagerTest.cpp: Appease msvc not to do "using llvm::Pass" in class template.
FIXME: I have not checked whether to be compiled on msvc11.
llvm-svn: 169225
Chandler Carruth [Tue, 4 Dec 2012 07:12:27 +0000 (07:12 +0000)]
Sort includes for all of the .h files under the 'lib' tree. These were
missed in the first pass because the script didn't yet handle include
guards.
Note that the script is now able to handle all of these headers without
manual edits. =]
llvm-svn: 169224
Nadav Rotem [Tue, 4 Dec 2012 07:11:52 +0000 (07:11 +0000)]
Give scalar if-converted blocks half the score because they are not always executed due to CF.
llvm-svn: 169223
Chandler Carruth [Tue, 4 Dec 2012 07:04:58 +0000 (07:04 +0000)]
Address review comments from Matt on the sort_includes.py script.
1) Teach it to handle files with #include on the first line -- these do
actually exist in LLVM.
2) Support llvm-c and clang-c include projects.
3) Nuke some stail imports.
4) Switch to using os.path to split the file extension off.
5) Remove debugging leftovers.
6) Add docstring (a really puny one) for the sort function.
I'm continuing te avoid stripping the whitespace on the RHS to preserve
whatever newline characters happen to be in the original file.
llvm-svn: 169222
Chandler Carruth [Tue, 4 Dec 2012 07:04:57 +0000 (07:04 +0000)]
Add a comment about the requirement that the Windows.h header be last.
This comment has the dual effect of blocking reorderings with the
sort_include script.
llvm-svn: 169221
NAKAMURA Takumi [Tue, 4 Dec 2012 06:58:05 +0000 (06:58 +0000)]
clang/test/CodeGenCXX/debug-info-zero-length-arrays.cpp: Add explicit triple, x86_64-unknown-unknown. It was incompatible to i686.
llvm-svn: 169220
Bill Wendling [Tue, 4 Dec 2012 06:21:27 +0000 (06:21 +0000)]
Add a 'count' field to the DWARF subrange.
The count field is necessary because there isn't a difference between the 'lo'
and 'hi' attributes for a one-element array and a zero-element array. When the
count is '0', we know that this is a zero-element array. When it's >=1, then
it's a normal constant sized array. When it's -1, then the array is unbounded.
llvm-svn: 169219
Bill Wendling [Tue, 4 Dec 2012 06:20:49 +0000 (06:20 +0000)]
Add a 'count' field to the DWARF subrange.
The count field is necessary because there isn't a difference between the 'lo'
and 'hi' attributes for a one-element array and a zero-element array. When the
count is '0', we know that this is a zero-element array. When it's >=1, then
it's a normal constant sized array. When it's -1, then the array is unbounded.
llvm-svn: 169218
Nadav Rotem [Tue, 4 Dec 2012 06:15:11 +0000 (06:15 +0000)]
Add the last part that is needed for vectorization of if-converted code.
Added the code that actually performs the if-conversion during vectorization.
We can now vectorize this code:
for (int i=0; i<n; ++i) {
unsigned k = 0;
if (a[i] > b[i]) <------ IF inside the loop.
k = k * 5 + 3;
a[i] = k; <---- K is a phi node that becomes vector-select.
}
llvm-svn: 169217
Kostya Serebryany [Tue, 4 Dec 2012 06:14:01 +0000 (06:14 +0000)]
[asan] add experimental -asan-realign-stack option (true by default, which does not change the current behavior)
llvm-svn: 169216
Bill Wendling [Tue, 4 Dec 2012 06:12:44 +0000 (06:12 +0000)]
Add a 'getCount' method to get the number of elements in the subrange.
llvm-svn: 169215
Matt Beaumont-Gay [Tue, 4 Dec 2012 05:41:27 +0000 (05:41 +0000)]
Add 'using' declarations to suppress -Woverloaded-virtual warnings.
llvm-svn: 169214
Jyotsna Verma [Tue, 4 Dec 2012 05:00:31 +0000 (05:00 +0000)]
Move all operand definitions into HexagonOperands.td
llvm-svn: 169213
Jyotsna Verma [Tue, 4 Dec 2012 04:29:16 +0000 (04:29 +0000)]
Move generic Hexagon subtarget information into Hexagon.td
llvm-svn: 169212
Sean Silva [Tue, 4 Dec 2012 03:45:27 +0000 (03:45 +0000)]
docs: Fix broken link.
llvm-svn: 169211
Sean Silva [Tue, 4 Dec 2012 03:30:36 +0000 (03:30 +0000)]
docs: Fix dead link.
Apparently Dinkumware are no longer hosting their nice reference
manuals. Thankfully, `cppreference.com` can fill that role well.
llvm-svn: 169210
Shuxin Yang [Tue, 4 Dec 2012 03:28:32 +0000 (03:28 +0000)]
rdar://
12329730 (2nd part, revised)
The type of shirt-right (logical or arithemetic) should remain unchanged
when transforming "X << C1 >> C2" into "X << (C1-C2)"
llvm-svn: 169209
Sean Silva [Tue, 4 Dec 2012 03:20:08 +0000 (03:20 +0000)]
docs: Convert ProgrammersManual to reST.
Patch by Alexander Zinenko!
llvm-svn: 169208
Richard Smith [Tue, 4 Dec 2012 02:48:16 +0000 (02:48 +0000)]
Document the existence of -fsanitize=bounds.
llvm-svn: 169207
Alexander Potapenko [Tue, 4 Dec 2012 02:41:47 +0000 (02:41 +0000)]
[ASan] Do not build the interceptors that use ObjC blocks if the compiler does not support blocks.
Need to define MISSING_BLOCKS_SUPPORT in this case at buildtime.
Patch by Jack Howarth <howarth@bromo.med.uc.edu>
llvm-svn: 169206
Greg Clayton [Tue, 4 Dec 2012 02:22:16 +0000 (02:22 +0000)]
<rdar://problem/
12750060>
Add the ability to get a symbol or symbols by name and type from a SBModule, and also the ability to get all symbols by name and type from SBTarget objects.
llvm-svn: 169205
Jim Ingham [Tue, 4 Dec 2012 01:57:57 +0000 (01:57 +0000)]
Missing #include to pick up def'n of TerminalState.h.
llvm-svn: 169203
Greg Clayton [Tue, 4 Dec 2012 01:57:36 +0000 (01:57 +0000)]
Get rid of a warning where functions in DNB.h were extern "C" when they don't need to be anymore.
llvm-svn: 169202
Alexey Samsonov [Tue, 4 Dec 2012 01:38:15 +0000 (01:38 +0000)]
ASan: add new interface functions - __asan_(un)poison_stack_memory. Calls to these functions are inserted by the instrumentation pass in use-after-scope mode
llvm-svn: 169201
Alexey Samsonov [Tue, 4 Dec 2012 01:34:23 +0000 (01:34 +0000)]
ASan: add initial support for handling llvm.lifetime intrinsics in ASan - emit calls into runtime library that poison memory for local variables when their lifetime is over and unpoison memory when their lifetime begins.
llvm-svn: 169200
NAKAMURA Takumi [Tue, 4 Dec 2012 01:03:31 +0000 (01:03 +0000)]
clang/test/Index/comment-objc-decls.m: Try to fix r169193, to add x86_64-darwin.
llvm-svn: 169199
Jakub Staszak [Tue, 4 Dec 2012 01:00:52 +0000 (01:00 +0000)]
Simplify code. No functionality change.
llvm-svn: 169198
Manman Ren [Tue, 4 Dec 2012 00:52:33 +0000 (00:52 +0000)]
Stack Alignment: when creating stack objects in MachineFrameInfo, make sure
the alignment is clamped to TargetFrameLowering.getStackAlignment if the target
does not support stack realignment or the option "realign-stack" is off.
This will cause miscompile if the address is treated as aligned and add is
replaced with or in DAGCombine.
Added a bool StackRealignable to TargetFrameLowering to check whether stack
realignment is implemented for the target. Also added a bool RealignOption
to MachineFrameInfo to check whether the option "realign-stack" is on.
rdar://
12713765
llvm-svn: 169197
Jakub Staszak [Tue, 4 Dec 2012 00:50:06 +0000 (00:50 +0000)]
Use dyn_cast instead of isa and cast. No functionality change.
llvm-svn: 169196
NAKAMURA Takumi [Tue, 4 Dec 2012 00:49:34 +0000 (00:49 +0000)]
LoopVectorize.cpp: Suppress a warning. [-Wunused-variable]
llvm-svn: 169195
NAKAMURA Takumi [Tue, 4 Dec 2012 00:49:28 +0000 (00:49 +0000)]
Fix whitespace.
llvm-svn: 169194
Fariborz Jahanian [Tue, 4 Dec 2012 00:47:33 +0000 (00:47 +0000)]
Testing objective-C declarations embedded in
<declaration> tag of Comment XML and fixed a
missing declaration of ivars private to @implementation
as result of the testing. // rdar://
12378714
llvm-svn: 169193
Jakob Stoklund Olesen [Tue, 4 Dec 2012 00:46:13 +0000 (00:46 +0000)]
Remove the old TRI::ResolveRegAllocHint() and getRawAllocationOrder() hooks.
These functions have been replaced by TRI::getRegAllocationHints() which
provides the same capabilities.
llvm-svn: 169192
Michael Ilseman [Tue, 4 Dec 2012 00:36:06 +0000 (00:36 +0000)]
Have clang use LLVM IR's fast-math flags when in FastMath or FiniteMathOnly modes. Test cases included.
llvm-svn: 169191
Jakob Stoklund Olesen [Tue, 4 Dec 2012 00:35:59 +0000 (00:35 +0000)]
Remove VirtRegMap::getRegAllocPref().
Now that there can be multiple hint registers from targets, it doesn't
make sense to have a function that returns 'the' preferred register.
llvm-svn: 169190
Greg Clayton [Tue, 4 Dec 2012 00:32:51 +0000 (00:32 +0000)]
<rdar://problem/
12798131>
Cleaned up the option parsing code to always pass around the short options as integers. Previously we cast this down to "char" and lost some information. I recently added an assert that would detect duplicate short character options which was firing during the test suite.
This fix does the following:
- make sure all short options are treated as "int"
- make sure that short options can be non-printable values when a short option is not required or when an option group is mixed into many commands and a short option is not desired
- fix the help printing to "do the right thing" in all cases. Previously if there were duplicate short character options, it would just not emit help for the duplicates
- fix option parsing when there are duplicates to parse options correctly. Previously the option parsing, when done for an OptionGroup, would just start parsing options incorrectly by omitting table entries and it would end up setting the wrong option value
llvm-svn: 169189
Jakob Stoklund Olesen [Tue, 4 Dec 2012 00:30:22 +0000 (00:30 +0000)]
Use MRI::getSimpleHint() instead of getRegAllocPref() in remaining cases.
Targets can provide multiple hints now, so getRegAllocPref() doesn't
make sense any longer because it only returns one preferred register.
Replace it with getSimpleHint() in the remaining heuristics. This
function only
llvm-svn: 169188
Michael Ilseman [Tue, 4 Dec 2012 00:29:55 +0000 (00:29 +0000)]
remove trailing whitespace
llvm-svn: 169187
Manman Ren [Tue, 4 Dec 2012 00:26:44 +0000 (00:26 +0000)]
Stack Alignment: move functions from header file MachineFrameInfo.h.
No functional change for this commit. The follow-up patch will add more stuff to
these functions.
rdar://
12713765
llvm-svn: 169186
Daniel Malea [Tue, 4 Dec 2012 00:23:45 +0000 (00:23 +0000)]
A few more build fixes for gcc 4.6:
- use const char* instead of char* as needed in ObjC language runtime plugin
- use int to iterate through enum (operator++ on enum not defined)
- use initializer list instead of inline initialization of const field
llvm-svn: 169185
NAKAMURA Takumi [Tue, 4 Dec 2012 00:08:14 +0000 (00:08 +0000)]
RuntimeDyld: Fix up r169178. MSVC doesn't like "or".
llvm-svn: 169183
Shuxin Yang [Tue, 4 Dec 2012 00:04:54 +0000 (00:04 +0000)]
rdar://
12329730 (2nd part)
This change tries to simmplify E1 = " X >> C1 << C2" into :
- E2 = "X << (C2 - C1)" if C2 > C1, or
- E2 = "X >> (C1 - C2)" if C1 > C2, or
- E2 = X if C1 == C2.
Reviewed by Nadav. Thanks!
llvm-svn: 169182
NAKAMURA Takumi [Mon, 3 Dec 2012 23:29:49 +0000 (23:29 +0000)]
clang/test/Driver/fsanitize-blacklist.c: Mark it as XFAIL:cygming, due to PR12920.
llvm-svn: 169180
Jakob Stoklund Olesen [Mon, 3 Dec 2012 23:23:50 +0000 (23:23 +0000)]
Add VirtRegMap::hasKnownPreference().
Virtual registers with a known preferred register are prioritized by
RAGreedy. This function makes the condition explicit without depending
on getRegAllocPref().
llvm-svn: 169179
Akira Hatanaka [Mon, 3 Dec 2012 23:12:19 +0000 (23:12 +0000)]
Runtime dynamic linker for MCJIT should support MIPS BigEndian architecture.
This small change adds support for that. It will make all MCJIT tests pass
in make-check on BigEndian platforms.
Patch by Petar Jovanovic.
llvm-svn: 169178
Akira Hatanaka [Mon, 3 Dec 2012 23:11:12 +0000 (23:11 +0000)]
Classic JIT is still being supported by MIPS, along with MCJIT.
This change adds endian-awareness to MipsJITInfo and emitWordLE in
MipsCodeEmitter has become emitWord now to support both endianness.
Patch by Petar Jovanovic.
llvm-svn: 169177
Michael Ilseman [Mon, 3 Dec 2012 22:57:47 +0000 (22:57 +0000)]
Minor tweaking to SmallVector static size.
llvm-svn: 169176
Nadav Rotem [Mon, 3 Dec 2012 22:57:09 +0000 (22:57 +0000)]
minor renaming, documentation and cleanups.
llvm-svn: 169175
Akira Hatanaka [Mon, 3 Dec 2012 22:51:22 +0000 (22:51 +0000)]
Functions in MipsCodeEmitter.cpp that expand unaligned loads/stores are dead
code. Removing it.
Patch by Petar Jovanovic.
llvm-svn: 169174
Jakob Stoklund Olesen [Mon, 3 Dec 2012 22:51:04 +0000 (22:51 +0000)]
Use the new getRegAllocationHints() hook from AllocationOrder.
This simplifies the hinting code quite a bit while making the targets
easier to write at the same time.
llvm-svn: 169173
Nadav Rotem [Mon, 3 Dec 2012 22:47:12 +0000 (22:47 +0000)]
constify the cost API
llvm-svn: 169172
Nadav Rotem [Mon, 3 Dec 2012 22:46:31 +0000 (22:46 +0000)]
IF-conversion: teach the cost-model how to grade if-converted loops.
llvm-svn: 169171
Richard Smith [Mon, 3 Dec 2012 22:39:14 +0000 (22:39 +0000)]
Fix test failure when building Clang with g++4.7 -- don't use a Twine temporary
after its lifetime has ended!
llvm-svn: 169170
Jakob Stoklund Olesen [Mon, 3 Dec 2012 22:35:35 +0000 (22:35 +0000)]
Implement ARMBaseRegisterInfo::getRegAllocationHints().
This provides the same functionality as getRawAllocationOrder() for the
even/odd hints, but without the many constant register arrays.
llvm-svn: 169169
Jyotsna Verma [Mon, 3 Dec 2012 22:26:28 +0000 (22:26 +0000)]
Define store instructions with base+immediate offset addressing mode
using multiclass.
llvm-svn: 169168
Michael J. Spencer [Mon, 3 Dec 2012 22:09:52 +0000 (22:09 +0000)]
[Support] Make FileOutputBuffer work on Windows.
llvm-svn: 169167
Michael J. Spencer [Mon, 3 Dec 2012 22:09:31 +0000 (22:09 +0000)]
[Support][FileSystem] Fix open mode in resize_file on Windows.
llvm-svn: 169166
Michael J. Spencer [Mon, 3 Dec 2012 22:07:00 +0000 (22:07 +0000)]
Revert the header sort on this file.
"Windows.h" includes <Windows.h> which defines a bunch of stuff it shouldn't
(even with all the restriction macros). We have no control over this file, so
make it's scope as small as possible.
llvm-svn: 169165
Pedro Artigas [Mon, 3 Dec 2012 21:56:57 +0000 (21:56 +0000)]
moves doInitialization and doFinalization to the Pass class and removes some unreachable code in MachineModuleInfo
reviewed by Evan Cheng <evan.cheng@apple.com>
llvm-svn: 169164
Argyrios Kyrtzidis [Mon, 3 Dec 2012 21:46:21 +0000 (21:46 +0000)]
Add a getMemorySize() function for DenseSet.
llvm-svn: 169163
Sebastian Pop [Mon, 3 Dec 2012 21:34:09 +0000 (21:34 +0000)]
adapt cloog codegen testcases to isl
llvm-svn: 169161
Sebastian Pop [Mon, 3 Dec 2012 21:33:55 +0000 (21:33 +0000)]
use -polly-ast instead of -polly-cloog
llvm-svn: 169160
Sebastian Pop [Mon, 3 Dec 2012 21:33:40 +0000 (21:33 +0000)]
execute cloog specific testcases only with CLOOG_FOUND
llvm-svn: 169159
Nadav Rotem [Mon, 3 Dec 2012 21:33:08 +0000 (21:33 +0000)]
Now that we have a basic if-conversion infrastructure we can rename the
"single basic block loop vectorizer" to "innermost loop vectorizer".
llvm-svn: 169158
Michael Ilseman [Mon, 3 Dec 2012 21:29:36 +0000 (21:29 +0000)]
Since this SmallVector immediately grows on the next line, don't waste stack space. SmallVector is still needed due to existing APIs growing their arguments
llvm-svn: 169157
Sean Callanan [Mon, 3 Dec 2012 21:28:37 +0000 (21:28 +0000)]
Fixed a crash in which we examined the extension of
a file name, whether the file name had an extension
or not.
<rdar://problem/
12793152>
llvm-svn: 169156
Alexander Potapenko [Mon, 3 Dec 2012 21:21:22 +0000 (21:21 +0000)]
Use a struct to hold the /proc/self/maps buffer on Linux.
llvm-svn: 169155
Jakob Stoklund Olesen [Mon, 3 Dec 2012 21:17:00 +0000 (21:17 +0000)]
Add a new hook for providing register allocator hints more flexibly.
The TargetRegisterInfo::getRegAllocationHints() function is going to
replace the existing mechanisms for providing target-dependent hints to
the register allocator: ResolveRegAllocHint() and
getRawAllocationOrder().
The new hook is more flexible because it allows the target to provide
multiple preferred candidate registers for each virtual register, and it
is easier to use because targets are not required to return a reference
to a constant array like getRawAllocationOrder().
An optional VirtRegMap argument can be used to provide target-dependent
hints that depend on the provisional assignments of other virtual
registers.
llvm-svn: 169154
Jyotsna Verma [Mon, 3 Dec 2012 21:13:13 +0000 (21:13 +0000)]
Define load instructions with base+immediate offset addressing mode
using multiclass.
llvm-svn: 169153
Nadav Rotem [Mon, 3 Dec 2012 21:06:35 +0000 (21:06 +0000)]
Add initial support for IF-conversion. This patch implements the first 1/3,
which is the legality of the if-conversion transformation. The next step is to
implement the cost-model for the if-converted code as well as the
vectorization itself.
llvm-svn: 169152
Manuel Klimek [Mon, 3 Dec 2012 20:55:42 +0000 (20:55 +0000)]
Fixes a compile warning and crash in the tests.
The necessity of this fix points to a problem with the design
of the addToken during the optimiation phase, which we need to address
in a much more principled way.
llvm-svn: 169151
Tobias Grosser [Mon, 3 Dec 2012 20:47:57 +0000 (20:47 +0000)]
Update CLooG
This cloog versions passes all tests with isl 0.11.
llvm-svn: 169150
Jyotsna Verma [Mon, 3 Dec 2012 20:39:45 +0000 (20:39 +0000)]
Define unsigned const-ext predicates.
llvm-svn: 169149
Jyotsna Verma [Mon, 3 Dec 2012 20:14:38 +0000 (20:14 +0000)]
Removing unnecessary 'else' statement from the predicates defined in HexagonOperards.td.
llvm-svn: 169148
Argyrios Kyrtzidis [Mon, 3 Dec 2012 19:59:23 +0000 (19:59 +0000)]
Eliminate redundant bitwise operations when using a llvm/ADT/PointerUnion.
For comparison, with this code sample:
PointerUnion<int *, char *> Data;
PointerUnion<int *, char *> foo1() {
Data = new int;
return new int;
}
PointerUnion<int *, char *> foo2() {
Data = new char;
return new char;
}
Before this patch we would get:
define i64 @_Z4foo1v() uwtable ssp {
%1 = tail call noalias i8* @_Znwm(i64 4)
%2 = ptrtoint i8* %1 to i64
%3 = load i64* getelementptr inbounds (%"class.llvm::PointerUnion"* @Data, i64 0, i32 0, i32 0), align 8
%4 = and i64 %3, 1
%.masked.i = and i64 %2, -3
%5 = or i64 %4, %.masked.i
store i64 %5, i64* getelementptr inbounds (%"class.llvm::PointerUnion"* @Data, i64 0, i32 0, i32 0), align 8
%6 = tail call noalias i8* @_Znwm(i64 4)
%7 = ptrtoint i8* %6 to i64
%8 = and i64 %7, -3
ret i64 %8
}
define i64 @_Z4foo2v() uwtable ssp {
%1 = tail call noalias i8* @_Znwm(i64 1)
%2 = ptrtoint i8* %1 to i64
%3 = load i64* getelementptr inbounds (%"class.llvm::PointerUnion"* @Data, i64 0, i32 0, i32 0), align 8
%4 = and i64 %3, 1
%5 = or i64 %2, %4
%6 = or i64 %5, 2
store i64 %6, i64* getelementptr inbounds (%"class.llvm::PointerUnion"* @Data, i64 0, i32 0, i32 0), align 8
%7 = tail call noalias i8* @_Znwm(i64 1)
%8 = ptrtoint i8* %7 to i64
%9 = or i64 %8, 2
ret i64 %9
}
After the patch:
define i64 @_Z4foo1v() uwtable ssp {
%1 = tail call noalias i8* @_Znwm(i64 4)
%2 = ptrtoint i8* %1 to i64
store i64 %2, i64* getelementptr inbounds (%"class.llvm::PointerUnion"* @Data, i64 0, i32 0, i32 0), align 8
%3 = tail call noalias i8* @_Znwm(i64 4)
%4 = ptrtoint i8* %3 to i64
ret i64 %4
}
declare noalias i8* @_Znwm(i64)
define i64 @_Z4foo2v() uwtable ssp {
%1 = tail call noalias i8* @_Znwm(i64 1)
%2 = ptrtoint i8* %1 to i64
%3 = or i64 %2, 2
store i64 %3, i64* getelementptr inbounds (%"class.llvm::PointerUnion"* @Data, i64 0, i32 0, i32 0), align 8
%4 = tail call noalias i8* @_Znwm(i64 1)
%5 = ptrtoint i8* %4 to i64
%6 = or i64 %5, 2
ret i64 %6
}
llvm-svn: 169147
Eli Bendersky [Mon, 3 Dec 2012 19:58:12 +0000 (19:58 +0000)]
Simplify this test a bit because DWARF emission/dumping on some platforms
is not yet good enough for more sophistication. The important goal of this
test is to make sure llc doesn't crash on this IR like it used to.
llvm-svn: 169146
Bill Wendling [Mon, 3 Dec 2012 19:44:25 +0000 (19:44 +0000)]
Add 'getInt64Field()' method to get the signed integer instead of unsigned.
llvm-svn: 169145
Alexey Samsonov [Mon, 3 Dec 2012 19:12:58 +0000 (19:12 +0000)]
Add Clang flags -fsanitize-blacklist and -fno-sanitize-blacklist. Make this flag usable for ASan. Blacklisting can be used to disable sanitizer checks for particular file/function/object.
llvm-svn: 169144
Alexey Samsonov [Mon, 3 Dec 2012 19:09:26 +0000 (19:09 +0000)]
ASan: add blacklist file to ASan pass options. Clang patch for this will follow.
llvm-svn: 169143
Eli Bendersky [Mon, 3 Dec 2012 18:45:45 +0000 (18:45 +0000)]
Fix PR12942: Allow two CUs to be generated from the same source file.
Thanks Eric for the review.
llvm-svn: 169142
Kostya Serebryany [Mon, 3 Dec 2012 18:39:21 +0000 (18:39 +0000)]
[asan] two more bits for __sparc__
llvm-svn: 169141