Commit Graph

109244 Commits

Author SHA1 Message Date
Colin LeMahieu
8699f5390b [Hexagon] Resubmission of 220427
Modified library structure to deal with circular dependency between HexagonInstPrinter and HexagonMCInst.
Adding encoding bits for add opcode.
Adding llvm-mc tests.
Removing unit tests.

http://reviews.llvm.org/D5624

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220584 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 19:00:32 +00:00
Matt Arsenault
a36665918f Fix copy paste comment
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220581 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 18:13:10 +00:00
Rafael Espindola
68b02dcd54 Don't ever call materializeAllPermanently during LTO.
To do this, change the representation of lazy loaded functions.

The previous representation cannot differentiate between a function whose body
has been removed and one whose body hasn't been read from the .bc file. That
means that in order to drop a function, the entire body had to be read.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220580 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 18:13:04 +00:00
Sanjay Patel
51fa1bcef3 Allow AVX vrsqrtps generation.
This is a follow-on to r220570 that allows a 256-bit (v8f32)
version of vrsqrtps to be generated.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220579 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 17:59:18 +00:00
David Blaikie
7f6ba1758b DebugInfo: Sink DwarfDebug::ScopeVariables down into DwarfFile
(part of refactoring to allow subprogram emission in both the skeleton
and main units to enable -gmlt-like data to be included in the skeleton
for live inlined backtracing purposes)

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220578 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 17:57:34 +00:00
Sanjay Patel
be5d8bfddd remove class/function/variable names from comments; NFC
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220577 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 17:55:59 +00:00
David Blaikie
3ad7d41947 Remove DwarfDebug::FirstCU as it has no use
It was only being used as a flag to identify the lack of debug info from
within endModule - use the section labels for that instead.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220575 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 17:53:38 +00:00
Sanjay Patel
a46f06efe2 Use rsqrt (X86) to speed up reciprocal square root calcs
This is a first step for generating SSE rsqrt instructions for
reciprocal square root calcs when fast-math is allowed.

For now, be conservative and only enable this for AMD btver2
where performance improves significantly - for example, 29%
on llvm/projects/test-suite/SingleSource/Benchmarks/BenchmarkGame/n-body.c
(if we convert the data type to single-precision float).

This patch adds a two constant version of the Newton-Raphson
refinement algorithm to DAGCombiner that can be selected by any target
via a parameter returned by getRsqrtEstimate()..

See PR20900 for more details:
http://llvm.org/bugs/show_bug.cgi?id=20900

Differential Revision: http://reviews.llvm.org/D5658



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220570 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 17:02:16 +00:00
Daniel Sanders
2992ea0cb5 [mips] Replace MipsABIEnum with a MipsABIInfo class.
Summary:
No functional change yet, it's just an object replacement for an enum.
It will allow us to gather ABI information in a single place so that we can
start testing for properties of the ABI's instead of the ABI itself.

For example we will eventually be able to use:
  ABI.MinStackAlignmentInBytes()
instead of:
  (isABI_N32() || isABI_N64()) ? 16 : 8
which is clearer and more maintainable.

Reviewers: matheusalmeida

Reviewed By: matheusalmeida

Differential Revision: http://reviews.llvm.org/D3341

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220568 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 16:15:27 +00:00
Benjamin Kramer
5ea04462dc [Object] Fix MachO's getUuid to return a pointer into the object instead of a dangling ArrayRef.
This works because uuid's are always little endian so it's not swapped.
Fixes use-after-return reported by asan.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220567 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 15:52:05 +00:00
Aaron Ballman
a653bf1b12 These functions are not actually defined for NDEBUG or !LLVM_DUMP_ENABLED, so guarding the declarations as well. NFC, silences MSVC warnings in release builds.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220565 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 15:16:39 +00:00
Daniel Sanders
29e46cd4b7 [mips] Fix >80-column line
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220564 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 14:46:00 +00:00
Daniel Sanders
4d2ff459e9 [mips] Remove redundant code in RetCC_MipsN. NFC.
Summary:
i32 is always promoted to i64 so it no longer makes sense to assign i32 to
registers.

Reviewers: vmedic

Reviewed By: vmedic

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D5964

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220561 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 13:49:54 +00:00
Daniel Sanders
de90aa00ac [mips] For N32/N64, structs must be passed in the upper bits of a register.
Summary:
Most structs were fixed by r218451 but those of between >32-bits and
<64-bits remained broken since they were not marked with [ASZ]ExtUpper.
This patch fixes the remaining cases by using
CCPromoteToUpperBitsInType<i64> on i64's in addition to i32 and smaller.

Reviewers: vmedic

Reviewed By: vmedic

Subscribers: llvm-commits

Differential Revision: http://reviews.llvm.org/D5963

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220556 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 13:09:19 +00:00
Oliver Stannard
9bb3f37aa4 [AArch64] Fix fast-isel of cbz of i1, i8, i16
This fixes a miscompilation in the AArch64 fast-isel which was
triggered when a branch is based on an icmp with condition eq or ne,
and type i1, i8 or i16. The cbz instruction compares the whole 32-bit
register, so values with the bottom 1, 8 or 16 bits clear would cause
the wrong branch to be taken.



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220553 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 09:54:41 +00:00
Timur Iskhodzhanov
806ccfff11 Update test/MC/ARM/coff-debugging-secrel.ll expectations to fix breakage caused by r220544
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220548 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 06:24:07 +00:00
Marcello Maggioni
77348a5703 Added reset of LexicalScope in LiveDebugVariables reset function.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220545 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 02:46:50 +00:00
Timur Iskhodzhanov
92e132a36f Fix PR21189 -- Emit symbol subsection required to debug LLVM-built binaries with VS2012+
Reviewed at http://reviews.llvm.org/D5772

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220544 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 01:27:45 +00:00
David Blaikie
8a5eafb58d DebugInfo: Remove DwarfDebug::addScopeVariable now that it's just a trivial wrapper
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220542 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 00:43:47 +00:00
Ahmed Bougacha
636864d8e7 Make test for r220533 more robust by using GPR pattern.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220541 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 00:03:46 +00:00
Adam Nemet
7f7bf0da6a [AVX512] FMA support for the 231 variants
This is asm/diasm-only support, similar to AVX.

For ISeling the register variant, they are no different from 213 other than
whether the multiplication or the addition operand is destructed.

For ISeling the memory variant, i.e. to fold a load, they are no different
than the 132 variant.  The addition operand (op3) in both cases can come from
memory.  Again the ony difference is which operand is destructed.

There could be a post-RA pass that would convert a 213 or 132 into a 231.

Part of <rdar://problem/17082571>

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220540 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 00:03:00 +00:00
Adam Nemet
97a3cd4383 [AVX512] Introduce fma3p_forms from AVX
This multiclass generates the different forms: 213, 231, 132 in AVX.

132 in AVX512 is a separate class but I am planning to use this same
multiclass to generate 231 relying on the nice the null_frag trick from AVX to
disable codegen pattern for 231.

No functionality change, no change in X86.td.expanded except for the different
instruction definition names.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220539 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-24 00:02:55 +00:00
Nick Lewycky
c08a7201b7 If requested, apply function merging at -O0 too. It's useful there to reduce the time to compile.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220537 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 23:49:31 +00:00
Timur Iskhodzhanov
da45b2bdbd Make getDISubprogram(const Function *F) available in LLVM
Reviewed at http://reviews.llvm.org/D5950

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220536 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 23:46:28 +00:00
Ahmed Bougacha
17c1e34c12 [SelectionDAG] Teach the vector scalarizer about FP conversions.
This adds support for legalization of instructions of the form:

  [fp_conv] <1 x i1> %op to <1 x double>

where fp_conv is one of fpto[us]i, [us]itofp.  This used to assert
because they were simply missing from the vector operand scalarizer.

A similar problem arose in r190830, with trunc instead.

Fixes PR20778.

Differential Revision: http://reviews.llvm.org/D5810


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220533 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 22:49:25 +00:00
Ahmed Bougacha
54d95a7074 Update comment and fix typos in assert message. (NFC)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220531 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 22:40:34 +00:00
Juergen Ributzka
c08387e220 Update llvm.donothing documentation.
llvm.donothing is no longer the only intrinsic that can be invoked.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220530 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 22:36:13 +00:00
Tim Northover
c0ecebb32b ScheduleDAG: record PhysReg dependencies represented by CopyFromReg nodes
x86's CMPXCHG -> EFLAGS consumer wasn't being recorded as a real EFLAGS
dependency because it was represented by a pair of CopyFromReg(EFLAGS) ->
CopyToReg(EFLAGS) nodes. ScheduleDAG was expecting the source to be an
implicit-def on the instruction, where the result numbers in the DAG and the
Uses list in TableGen matched up precisely.

The Copy notation seems much more robust, so this patch extends ScheduleDAG
rather than refactoring x86.

Should fix PR20376.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220529 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 22:31:48 +00:00
David Blaikie
991327143f DebugInfo: Remove DwarfDebug::CurrentFnArguments since we have to handle argument ordering of other arguments (abstract arguments) in the same way and already have code for that too.
While refactoring this code I was confused by both the name I had
introduced (addNonArgumentVariable... but it has all this logic to
handle argument numbering and keep things in order?) and by the
redundancy. Seems when I fixed the misordered inlined argument handling,
I didn't realize it was mostly redundant with the argument ordering code
(which I may've also written, I'm not sure). So let's just rely on the
more general case.

The only oddity in output this produces is that it means when we emit
all the variables for the current function, we don't track when we've
finished the argument variables and are about to start the local
variables and insert DW_AT_unspecified_parameters (for varargs
functions) there. Instead it ends up after the local variables, scopes,
etc. But this isn't invalid and doesn't cause DWARF consumers problems
that I know of... so we'll just go with that because it makes the code
nice & simple.

(though, let's see what the buildbots have to say about this - *crosses
fingers*)

There will be some cleanup commits to follow to remove the now trivial
wrappers, etc.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220527 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 22:27:50 +00:00
Timur Iskhodzhanov
13535f412a PR21189: Teach llvm-readobj to dump bits of COFF symbol subsections required to debug using VS2012+
Reviewed at http://reviews.llvm.org/D5755
Thanks to Andrey Guskov for his help investigating this!



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220526 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 22:25:31 +00:00
David Blaikie
48ba51a8cd DebugInfo: Sink DwarfDebug::addNonArgumentScopeVariable into DwarfFile.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220520 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 22:04:30 +00:00
Hans Wennborg
e4e2ea6839 MachODump.cpp: fix MSVC build
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220518 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 21:59:17 +00:00
Ahmed Bougacha
4d962b05ca [X86] Improve mul w/ overflow codegen, to MUL8+SETO.
Currently, @llvm.smul.with.overflow.i8 expands to 9 instructions, where
3 are really needed.

This adds X86ISD::UMUL8/SMUL8 SD nodes, and custom lowers them to
MUL8/IMUL8 + SETO.

i8 is a special case because there is no two/three operand variants of
(I)MUL8, so the first operand and return value need to go in AL/AX.

Also, we can't write patterns for these instructions: TableGen refuses
patterns where output operands don't match SDNode results. In this case,
instructions where the output operand is an implicitly defined register.

A related special case (and FIXME) exists for MUL8 (X86InstrArith.td):

  // FIXME: Used for 8-bit mul, ignore result upper 8 bits.
  // This probably ought to be moved to a def : Pat<> if the
  // syntax can be accepted.
  [(set AL, (mul AL, GR8:$src)), (implicit EFLAGS)]

Ideally, these go away with UMUL8, but we still need to improve TableGen
support of implicit operands in patterns.

Before this change:
  movsbl  %sil, %eax
  movsbl  %dil, %ecx
  imull   %eax, %ecx
  movb    %cl, %al
  sarb    $7, %al
  movzbl  %al, %eax
  movzbl  %ch, %esi
  cmpl    %eax, %esi
  setne   %al

After:
  movb    %dil, %al
  imulb   %sil
  seto    %al

Also, remove a made-redundant testcase for PR19858, and enable more FastISel
ALU-overflow tests for SelectionDAG too.

Differential Revision: http://reviews.llvm.org/D5809


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220516 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 21:55:31 +00:00
David Blaikie
78f043e90a DebugInfo: Remove DwarfDebug::addCurrentFnArgument declaration now that it's moved to DwarfFile.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220515 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 21:53:17 +00:00
Sanjay Patel
d2153694e0 Handle sqrt() shrinking in SimplifyLibCalls like any other call
This patch removes a chunk of special case logic for folding 
(float)sqrt((double)x) -> sqrtf(x)
in InstCombineCasts and handles it in the mainstream path of SimplifyLibCalls.

No functional change intended, but I loosened the restriction on the existing
sqrt testcases to allow for this optimization even without unsafe-fp-math because
that's the existing behavior.

I also added a missing test case for not shrinking the llvm.sqrt.f64 intrinsic
in case the result is used as a double.

Differential Revision: http://reviews.llvm.org/D5919



git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220514 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 21:52:45 +00:00
Kevin Enderby
4d529076b6 Change the name of the field BindTable to bindtable to not over lap the type.
Should fix the build bot issues from commit r220500.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220504 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 19:53:12 +00:00
Peter Collingbourne
1ea06a0ead Make llvm-go test dependency optional.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220503 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 19:51:40 +00:00
Rafael Espindola
e37b00e50b Make llvm-link behave a bit more like LTO.
* Read modules lazily
* Don't treat the first file specially, instead merge all inputs into an empty
  module.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220501 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 19:40:45 +00:00
Kevin Enderby
a1afbd6421 Update llvm-objdump’s Mach-O symbolizer code for Objective-C references.
This prints disassembly comments for Objective-C references to CFStrings,
Selectors, Classes and method calls.


git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220500 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 19:37:31 +00:00
Rafael Espindola
a023f50940 Cleanup this test a bit.
Use simpler names and remove unnecessary fields.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220499 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 19:36:21 +00:00
Rafael Espindola
c32d7489b0 Cleanup this test a bit.
Use simpler names and remove unnecessary fields.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220498 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 19:23:42 +00:00
David Blaikie
129ab6c6c5 DebugInfo: Simplify/tidy/correct global variable decl/def emission handling.
This fixes a bug (introduced by fixing the IR emitted from Clang where
the definition of a static member would be scoped within the class,
rather than within its lexical decl context) where the definition of a
static variable would be placed inside a class.

It also improves source fidelity by scoping static class member
definitions inside the lexical decl context in which tehy are written
(eg: namespace n { class foo { static int i; } int foo::i; } - the
definition of 'i' will be within the namespace 'n' in the DWARF output
now).

Lastly, and the original goal, this reduces debug info size slightly
(and makes debug info easier to read, etc) by placing the definitions of
non-member global variables within their namespace, rather than using a
separate namespace-scoped declaration along with a definition at global
scope.

Based on patches and discussion with Frédéric.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220497 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 19:12:43 +00:00
Rafael Espindola
b11d9944d9 Make this test a bit stricter.
This now:
* Forces the linker to include the internal definition.
* Checks the full output.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220495 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 18:52:46 +00:00
Rafael Espindola
74ff8164d7 Make this test a bit stricter.
This now:
* Forces the linker to include the internal definition.
* Checks the full output.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220494 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 18:44:07 +00:00
Reid Kleckner
e852ac7772 Revert "Don't count inreg params when mangling fastcall functions"
This reverts commit r214981.

I'm not sure what I was thinking when I wrote this. Testing with MSVC
shows that this function is mangled to '@f@8':
  int __fastcall f(int a, int b);

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220492 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 17:50:42 +00:00
Chris Bieneman
71926e73ba Adding llvm-shlib to CMake build system with a few new bells and whistles
Summary:
This patch adds a new CMake build setting LLVM_BUILD_LLVM_DYLIB, which defaults to OFF. When set to ON, this will generate a shared library containing most of LLVM. The contents of the shared library can be overriden by specifying LLVM_DYLIB_COMPONENTS. LLVM_DYLIB_COMPONENTS can be set to a semi-colon delimited list of any LLVM components that you llvm-config can resolve.

On Windows, unless you are using Cygwin, you must specify an explicit symbol export file using LLVM_EXPORTED_SYMBOL_FILE. On Cygwin and all unix-like platforms if you do not specify LLVM_EXPORTED_SYMBOL_FILE, an export file containing only the LLVM C API will be auto-generated from the list of LLVM components specified in LLVM_DYLIB_COMPONENTS.

Reviewers: rnk

Reviewed By: rnk

Subscribers: rnk, llvm-commits

Differential Revision: http://reviews.llvm.org/D5890

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220490 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 17:22:14 +00:00
David Blaikie
3d71678283 Remove explicit (void) use of DwarfFile::DD that was accidentally left in r220452.
Caught in post-commit review by Frédéric.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220487 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 16:12:58 +00:00
Renato Golin
06b11e36e5 Do not emit intermediate register for zero FP immediate
This updates check for double precision zero floating point constant to allow
use of instruction with immediate value rather than temporary register.
Currently "a == 0.0", where "a" is of "double" type generates:

vmov.i32        d16, #0x0
vcmpe.f64       d0, d16

With this change it becomes:

vcmpe.f64        d0, #0

Patch by Sergey Dmitrouk.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220486 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 15:31:50 +00:00
Rafael Espindola
8b8e2b2f81 clang-format two code snippets to make the next patch easy to read.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220484 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 15:20:05 +00:00
Rafael Espindola
08aeb166fd Add unittest for extreme alignments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@220483 91177308-0d34-0410-b5e6-96231b3b80d8
2014-10-23 14:45:19 +00:00