Previously loop hints such as #pragma loop vectorize_width(#) required a constant. This patch allows a constant expression to be used as well. Such as a non-type template parameter or an expression (2 * c + 1).
Reviewed by Richard Smith
llvm-svn: 219589
This patch improves support for commutative instructions in the x86 memory folding implementation by attempting to fold a commuted version of the instruction if the original folding fails - if that folding fails as well the instruction is 're-commuted' back to its original order before returning.
This mainly helps the stack inliner better fold reloads of 3 (or more) operand instructions (VEX encoded SSE etc.) but by performing this in the lowest foldMemoryOperandImpl implementation it also replaces the X86InstrInfo::optimizeLoadInstr version and is now used by FastISel too.
Differential Revision: http://reviews.llvm.org/D5701
llvm-svn: 219584
A helper routine, MultiplyOverflows, was a less efficient
reimplementation of APInt's smul_ov and umul_ov. While we are here,
clean up the code so it's more uniform.
No functionality change intended.
llvm-svn: 219583
While we ran getUnqualifiedType over the catch type,
it isn't enough for array types. Use getUnqualifiedArrayType instead.
This fixes PR21252.
llvm-svn: 219582
With this change, both local-process llgs and remote-target llgs stdout/stderr
handling from inferior work correctly.
Several log lines have been added around PTY and stdout/stderr redirection
logic on the lldb client side.
Regarding remote llgs execution, see the following:
With these changes, remote llgs with $O now works properly:
$ lldb
(lldb) platform select remote-linux
(lldb) target create ~/some/inferior/exe
(lldb) gdb-remote {some-target}:{port}
(lldb) run
The sequence above will correctly redirect stdout/stderr over gdb-remote $O,
as is needed for remote debugging. That sequence assumes there is a lldb-gdbserver
exe running on the target with {some-host}:{port}.
You can replace the gdb-remote command with a '(lldb) platform connect
connect://{target-ip}:{target-port}'. If you do this and have a
lldb-platform running on the remote end, it will go ahead and launch
llgs for lldb for each target instance that is run/attached.
For local debugging with llgs, the following sequence also works, and
uses local PTYs instead to avoid $O and extra gdb-remote messages:
$ lldb
(lldb) settings set platform.plugin.linux.use-llgs true
(lldb) target create ~/some/inferior/exe
(lldb) run
The above will run the inferior using llgs on the local host, and
will use PTYs rather than $O redirection.
This change also removes the logging that happened after the fork but
before the exec when llgs is launching a new inferior process. Some
aspect of the file handling during that portion of code would not do
the right thing with log handling. We might want to go back later
and have that communicate over a pipe from the child to parent to pass
along any messages that previously were logged in that section of code.
llvm-svn: 219578
On x86_64 this brings it from 80 bytes to 64 bytes. Also make any member
variables private and clean up uses to go through the existing accessors.
NFC.
llvm-svn: 219573
and !=) to support mixed complex and real operand types.
This requires removing an assert from SemaChecking, and adding support
both to the constant evaluator and the code generator to synthesize the
imaginary part when needed. This seemed somewhat cleaner than having
just the comparison operators force real-to-complex conversions.
I've added test cases for these operations. I'm really terrified that
there were *no* tests in-tree which exercised this.
This turned up when trying to build R after my change to the complex
type lowering.
llvm-svn: 219570
Consider the case where X is 2. (2 <<s 31)/s-2147483648 is zero but we
would fold to X. Note that this is valid when we are in the unsigned
domain because we require NUW: 2 <<u 31 results in poison.
This fixes PR21245.
llvm-svn: 219568
consider:
C1 = INT_MIN
C2 = -1
C1 * C2 overflows without a doubt but consider the following:
%x = i32 INT_MIN
This means that (%X /s C1) is 1 and (%X /s C1) /s C2 is -1.
N. B. Move the unsigned version of this transform to InstSimplify, it
doesn't create any new instructions.
This fixes PR21243.
llvm-svn: 219567
consider:
mul i32 nsw %x, -2147483648
this instruction will not result in poison if %x is 1
however, if we transform this into:
shl i32 nsw %x, 31
then we will be generating poison because we just shifted into the sign
bit.
This fixes PR21242.
llvm-svn: 219566
for complex math.
This should fix the windows build bots that started having trouble here
and generally fix complex libcall emission on targets which use sret for
complex data types. It also makes the code a bit simpler (despite
calling into a much more complex bucket of code).
llvm-svn: 219565
to pass in an opt build.
The test case in question does show UBSan catching the error, but it
doesn't then successfully set the exit code of the program. I'll let the
UBSan folks sort out why. It should reproduce trivially with an
optimized build.
llvm-svn: 219563
getSmallConstantTripCount even when it isn't the exiting block.
I missed this in my first audit, very sorry. This was found in LNT and
elsewhere. I don't have a test case, but it was completely obvious from
inspection that this was the problem. I'll see if I can reduce a test
case, but I'm not really hopeful, and the value seems quite low.
llvm-svn: 219562
Reviewed at http://reviews.llvm.org/D5738
This adds an SB API into SBProcess:
bool SBProcess::IsInstrumentationRuntimePresent(InstrumentationRuntimeType type);
which simply tells whether a particular InstrumentationRuntime (read "ASan") plugin is present and active.
llvm-svn: 219560
operators where one type is a C complex type, and to emit both the
efficient and correct implementation for complex arithmetic according to
C11 Annex G using this extra information.
For both multiply and divide the old code was writing a long-hand
reduced version of the math without any of the special handling of inf
and NaN recommended by the standard here. Instead of putting more
complexity here, this change does what GCC does which is to emit
a libcall for the fully general case.
However, the old code also failed to do the proper minimization of the
set of operations when there was a mixed complex and real operation. In
those cases, C provides a spec for much more minimal operations that are
valid. Clang now emits the exact suggested operations. This change isn't
*just* about performance though, without minimizing these operations, we
again lose the correct handling of infinities and NaNs. It is critical
that this happen in the frontend based on assymetric type operands to
complex math operations.
The performance implications of this change aren't trivial either. I've
run a set of benchmarks in Eigen, an open source mathematics library
that makes heavy use of complex. While a few have slowed down due to the
libcall being introduce, most sped up and some by a huge amount: up to
100% and 140%.
In order to make all of this work, also match the algorithm in the
constant evaluator to the one in the runtime library. Currently it is
a broken port of the simplifications from C's Annex G to the long-hand
formulation of the algorithm.
Splitting this patch up is very hard because none of this works without
the AST change to preserve non-complex operands. Sorry for the enormous
change.
Follow-up changes will include support for sinking the libcalls onto
cold paths in common cases and fastmath improvements to allow more
aggressive backend folding.
Differential Revision: http://reviews.llvm.org/D5698
llvm-svn: 219557
Summary: Implement the most basic form of conditional branches in Mips fast-isel.
Test Plan:
br1.ll
run 4 flavors of test-suite. mips32 r1/r2 and at -O0/O2
Reviewers: dsanders
Reviewed By: dsanders
Subscribers: llvm-commits, rfuhler
Differential Revision: http://reviews.llvm.org/D5583
llvm-svn: 219556
Reviewed at http://reviews.llvm.org/D5736
The new test cases for ASan fail if the llvm build that is used with LLDB doesn't have compiler-rt (because the resulting compiler then cannot build with -fsanitize=address). Let's include compiler-rt in build-llvm.pl script and make sure we actually *build* it by removing the NO_RUNTIME_LIBS=1 argument used in the make line. After this, the ASan tests pass on a fresh svn checkout.
llvm-svn: 219555
do that (RunCommandInterpreter, HandleCommands, HandleCommandsFromFile) to gather
the options into an options class. Also expose that to the SB API's.
Change the way the "-o" options to the lldb driver are processed so:
1) They are run synchronously - didn't really make any sense to run the asynchronously.
2) The stop on error
3) "quit" in one of the -o commands will not quit lldb - not the command interpreter
that was running the -o commands.
I added an entry to the run options to stop-on-crash, but I haven't implemented that yet.
llvm-svn: 219553
declaration in the instantiation if the previous declaration came from another
definition of the class template that got merged into the pattern definition.
llvm-svn: 219552
Summary: This fixes PR21235.
Test Plan: Includes an automated test.
Reviewers: hansw
Subscribers: cfe-commits
Differential Revision: http://reviews.llvm.org/D5718
llvm-svn: 219551
routines and fix all of the bugs they expose.
I hit a test case that crashed even without these asserts due to passing
a non-exiting latch to the ExitingBlock parameter of the trip count
computation machinery. However, when I add the nice asserts, it turns
out we have plenty of coverage of these bugs, they just didn't manifest
in crashers.
The core problem seems to stem from an assumption that the latch *is*
the exiting block. While this is often true, and somewhat the "normal"
way to think about loops, it isn't necessarily true. The correct way to
call the trip count routines in a *generic* fashion (that is, without
a particular exit in mind) is to just use the loop's single exiting
block if it has one. The trip count can't be computed generically unless
it does. This works great for the loop vectorizer. The loop unroller
actually *wants* to select the latch when it has to chose between
multiple exits because for unrolling it is the latch trips that matter.
But if this is the desire, it needs to explicitly guard for non-exiting
latches and check for the generic trip count in that case.
I've added the asserts, and added convenience APIs for querying the trip
count generically that check for a single exit block. I've kept the APIs
consistent between computing trip count and trip multiples.
Thansk to Mark for the help debugging and tracking down the *right* fix
here!
llvm-svn: 219550
Reviewed at http://reviews.llvm.org/D5592
This patch gives LLDB some ability to interact with AddressSanitizer runtime library, on top of what we already have (historical memory stack traces provided by ASan). Namely, that's the ability to stop on an error caught by ASan, and access the report information that are associated with it. The report information is also exposed into SB API.
More precisely this patch...
adds a new plugin type, InstrumentationRuntime, which should serve as a generic superclass for other instrumentation runtime libraries, these plugins get notified when modules are loaded, so they get a chance to "activate" when a specific dynamic library is loaded
an instance of this plugin type, AddressSanitizerRuntime, which activates itself when it sees the ASan dynamic library or founds ASan statically linked in the executable
adds a collection of these plugins into the Process class
AddressSanitizerRuntime sets an internal breakpoint on __asan::AsanDie(), and when this breakpoint gets hit, it retrieves the report information from ASan
this breakpoint is then exposed as a new StopReason, eStopReasonInstrumentation, with a new StopInfo subclass, InstrumentationRuntimeStopInfo
the StopInfo superclass is extended with a m_extended_info field (it's a StructuredData::ObjectSP), that can hold arbitrary JSON-like data, which is the way the new plugin provides the report data
the "thread info" command now accepts a "-s" flag that prints out the JSON data of a stop reason (same way the "-j" flag works now)
SBThread has a new API, GetStopReasonExtendedInfoAsJSON, which dumps the JSON string into a SBStream
adds a test case for all of this
I plan to also get rid of the original ASan plugin (memory history stack traces) and use an instance of AddressSanitizerRuntime for that purpose.
Kuba
llvm-svn: 219546
We can safely rely on the architecture to distinguish iOS device builds from
iOS simulator builds. We already have code to do that, in fact. This simplifies
some of the error checking for the option handling.
llvm-svn: 219545
output style can be customized. Change the built-in default to be
more similar to gdb's disassembly formatting.
The disassembly-format for a gdb-like output is
${addr-file-or-load} <${function.name-without-args}${function.concrete-only-addr-offset-no-padding}>:
The disassembly-format for the lldb style output is
{${function.initial-function}{${module.file.basename}`}{${function.name-without-args}}:\n}{${function.changed}\n{${module.file.basename}`}{${function.name-without-args}}:\n}{${current-pc-arrow} }{${addr-file-or-load}}:
The two backticks in the lldb style formatter triggers the sub-expression evaluation in
CommandInterpreter::PreprocessCommand() so you can't use that one as-is ... changing to
use ' characters instead of ` would work around that.
<rdar://problem/9885398>
llvm-svn: 219544
Richard Smith