- unsigned TrueValue = getReg(TrueVal, BB, BB->begin());
+ unsigned TrueValue = getReg(TrueVal);
Fixes the PPC regressions from last night.
The other hunk is just a clarity improvement.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19263 91177308-0d34-0410-b5e6-96231b3b80d8
to get Visual Studio to link in X86.lib to the executables that need it. There
is another way of doing it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19252 91177308-0d34-0410-b5e6-96231b3b80d8
1. Add new instructions for checking parity flags: JP, JNP, SETP, SETNP.
2. Set the isCommutable and isPromotableTo3Address bits on several
instructions.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19246 91177308-0d34-0410-b5e6-96231b3b80d8
1. If we are two-addressing a commutable instruction and the LHS is not the
last use of the variable, see if the instruction is the last use of the
RHS. If so, commute the instruction, allowing us to avoid a
register-register copy in many cases for common instructions like ADD, OR,
AND, etc on X86.
2. If #1 doesn't hold, and if this is an instruction that also existing in
3-address form, promote the instruction to a 3-address instruction to
avoid the register-register copy. We can do this for several common
instructions in X86, including ADDrr, INC, DEC, etc.
This patch implements test/Regression/CodeGen/X86/commute-two-addr.ll,
overlap-add.ll, and overlap-shift.ll when I check in the X86 support for it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19245 91177308-0d34-0410-b5e6-96231b3b80d8
* We can now fold cast instructions into select instructions that
have at least one constant operand.
* We now optimize expressions more aggressively based on bits that are
known to be zero. These optimizations occur a lot in code that uses
bitfields even in simple ways.
* We now turn more cast-cast sequences into AND instructions. Before we
would only do this if it if all types were unsigned. Now only the
middle type needs to be unsigned (guaranteeing a zero extend).
* We transform sign extensions into zero extensions in several cases.
This corresponds to these test/Regression/Transforms/InstCombine testcases:
2004-11-22-Missed-and-fold.ll
and.ll: test28-29
cast.ll: test21-24
and-or-and.ll
cast-cast-to-and.ll
zeroext-and-reduce.ll
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19220 91177308-0d34-0410-b5e6-96231b3b80d8
While we're at it, improve codegen of select instructions. For this
testcase:
int %test(bool %C, int %A, int %B) {
%D = select bool %C, int %A, int %B
ret int %D
}
We used to generate this code:
_test:
cmpwi cr0, r3, 0
bne .LBB_test_2 ;
.LBB_test_1: ;
b .LBB_test_3 ;
.LBB_test_2: ;
or r5, r4, r4
.LBB_test_3: ;
or r3, r5, r5
blr
Now we emit:
_test:
cmpwi cr0, r3, 0
bne .LBB_test_2 ;
.LBB_test_1: ;
or r4, r5, r5
.LBB_test_2: ;
or r3, r4, r4
blr
-Chris
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19214 91177308-0d34-0410-b5e6-96231b3b80d8
SimplifyCFG is one of those passes that we use for final cleanup: it should
not rely on other passes to clean up its garbage. This fixes the "why are
trivially dead setcc's in the output of gccas" problem.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19212 91177308-0d34-0410-b5e6-96231b3b80d8
portability, the --track-space option was inadvertently ignored. This patch
fixes that so that sys::Process::GetMallocUsage() is only invoked if the
--track-spaces option is given. Apparently the mallinfo() call that
GetMallocUsage() uses is *very* slow, especially when processing very large
modules like projects/llvm-test/MultiSource/Applications/kimwitu++.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19163 91177308-0d34-0410-b5e6-96231b3b80d8
* Ensure #includes are wrapped with appropriate HAVE_ guards
* Account for variations in time.h and sys/time.h inclusion.
* Add macros for getting wait(2) exit status correctly (per autoconf sugg.)
* Implement ThrowErrno in terms of strerror_r, strerror or none based on
what is available on the platform.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19161 91177308-0d34-0410-b5e6-96231b3b80d8
* Move implementation of sys::PreventCoreFiles function to this file from
the now defunct SysConfig abstraction.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19159 91177308-0d34-0410-b5e6-96231b3b80d8
* Ensure #includes are wrapped with appropriate HAVE_ guards
* Consolidate implementation from operating system specific directory.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19157 91177308-0d34-0410-b5e6-96231b3b80d8
* Consolidate implementation for Unix systems into Unix/TimeValue.cpp
* Avoid use of symbolic link to #include platform-specific implementation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19155 91177308-0d34-0410-b5e6-96231b3b80d8
* Consolidate implementation for Unix systems into Unix/Program.cpp
* Avoid use of symbolic link to #include platform-specific implementation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19154 91177308-0d34-0410-b5e6-96231b3b80d8
* Consolidate implementation for Unix systems into Unix/Signals.cpp
* Avoid use of symbolic link to #include platform-specific implementation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19153 91177308-0d34-0410-b5e6-96231b3b80d8
* Consolidate implementation for Unix systems into Unix/Memory.cpp
* Avoid use of symbolic link to #include platform-specific implementation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19152 91177308-0d34-0410-b5e6-96231b3b80d8
* Consolidate implementation for Unix systems into Unix/MappedFile.cpp
* Avoid use of symbolic link to #include platform-specific implementation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19151 91177308-0d34-0410-b5e6-96231b3b80d8
* Consolidate implementation for Unix systems into Unix/Process.cpp
* Avoid use of symbolic link to #include platform-specific implementation.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19150 91177308-0d34-0410-b5e6-96231b3b80d8
SysConf abstraction was pointless because it had a single function in it that
pertained only to the current process. So merge it into the Process abstraction
and remove the files completely.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19149 91177308-0d34-0410-b5e6-96231b3b80d8
Old versions of the C runtime somehow get loaded into the process. Make
sure they aren't searched for symbols.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19141 91177308-0d34-0410-b5e6-96231b3b80d8
Merge implementations of isValid and GetTemporaryDirectory into this file. There
is not any point having the operating system specific files for such little
variation between the Unix family of systems.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19134 91177308-0d34-0410-b5e6-96231b3b80d8
This implementation is no longer needed, its been merged to Unix/Path.cpp since
there is such little variation between the platforms.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19133 91177308-0d34-0410-b5e6-96231b3b80d8
Use the LLVM_ON_UNIX and LLVM_ON_WIN32 #defines to include the implementation
instead of relying upon the "platform" link
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19132 91177308-0d34-0410-b5e6-96231b3b80d8
a function call at the core of the loop inline and removing unused
stack variables from an often called function. This doesn't improve things
much, the real saving will be by reducing the number of calls to this
function (100K+ when linking kimwitu++).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19119 91177308-0d34-0410-b5e6-96231b3b80d8
Include local FDHandle.h file now that FDHandle is local to this library
and not in lib/System/FileUtilities.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19108 91177308-0d34-0410-b5e6-96231b3b80d8
Remove unix specific code (use of errno and read) from the reader.
Thanks to Jeff Cohen for pointing this out.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19081 91177308-0d34-0410-b5e6-96231b3b80d8
* Implement GetMallocUsage to get usage of malloc heap
* Implement GetMemoryUsage to get total memory usage of process
* Implement GetTimeUsage to get elapsed/user/system time
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19055 91177308-0d34-0410-b5e6-96231b3b80d8
* Move system dependent implementation out of this file.
* Make implementation use sys::Process::GetMallocUsage where necessary.
* Make implementation use sys::Process::GetTimeUsage where necessary.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19053 91177308-0d34-0410-b5e6-96231b3b80d8
Move the functionality of RunProgramWithTimeout into the ExecuteAndWait
function: support a timeout and I/O redirects
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19043 91177308-0d34-0410-b5e6-96231b3b80d8
* Pass sys::Path instead of std::string for paths
* Correct the types of arguments passed to RunProgramWithTimeout due to its
interface using sys::Path instead of std::string
* Replace "/dev/null" (not portable) with empty string which
sys::Program::ExecuteAndWait recognizes as "redirect to bit bucket"
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19041 91177308-0d34-0410-b5e6-96231b3b80d8
* Remove unneeded header files.
* Move RedirectFD static function to lib/System/Unix/Program.cpp
* Delete RunProgramWithTimeout, now implemented by
sys::Program::ExecuteAndWait. RunProgramWithTimeout is now a convenience func.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19040 91177308-0d34-0410-b5e6-96231b3b80d8
Regression/Analysis/GlobalsModRef/purecse.ll. Isn't this what the
-Woverload-whatever flag would warn about :)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19018 91177308-0d34-0410-b5e6-96231b3b80d8
save small amounts of time for functions that don't call llvm.returnaddress
or llvm.frameaddress (which is almost all functions).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19006 91177308-0d34-0410-b5e6-96231b3b80d8
* Make the OutputC and OutputAsm functions work with sys::Path for the output
file name instead of using std::string.
* Get rid of extraneous "toString" calls.
* Change "removeFile" to sys::Path::destroyFile()
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@19000 91177308-0d34-0410-b5e6-96231b3b80d8
* Remove the "removeFile" function, now implemented by the
sys::Path::destroyFile method.
* Make the FileRemove work with a sys::Path instead of a std::string
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18999 91177308-0d34-0410-b5e6-96231b3b80d8
createTemporaryFile semantics where it doesn't create a fully unique name
if the basename doesn't exist. This functionality is now optionally
provided by the boolean reuse_current parameter to createTemporaryFile and
makeUnique. The default values differ because of the way these functions
are used in LLVM.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18961 91177308-0d34-0410-b5e6-96231b3b80d8
* Remove #inclusion of FileUtilities.h, not needed any more.
* Convert getUniqueFilename -> sys::Pat::makeUnique()
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18948 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix implementation and documentation about LLVMGCCDIR/bytecode-libs
* Add the makeUnique method, replacement for getUniqueFilename in Support.
* Add the sys::CopyFile function, replacement for CopyFile in Support.
* Move GetLLVMConfigDir() into generic code area since its generic.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18947 91177308-0d34-0410-b5e6-96231b3b80d8
Remove getUniqueFilename and CopyFile. These are now implemented by
sys::Path::makeUnique and sys::CopyFile, respectively.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18946 91177308-0d34-0410-b5e6-96231b3b80d8
Adjust to changes in the interface of FindExecutable, getting ToolRunner
ready for bigger things to come.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18919 91177308-0d34-0410-b5e6-96231b3b80d8
* Remove isExecutable as its now implemented by sys::Path::executable
* Make FindExecutable a thin veneer over sys::Program::FindProgramByName.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18918 91177308-0d34-0410-b5e6-96231b3b80d8
Remove AllocateRWXMemory as it is not used any more in LLVM. The function
has been replaced with sys::Memory::AllocateRWX several months ago.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18912 91177308-0d34-0410-b5e6-96231b3b80d8
Remove the MakeFileReadable and MakeFileExecutable functions which are no
longer present in LLVM. They have been replaced with the sys::Path methods
makeReadable and makeExecutable, respectively.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18910 91177308-0d34-0410-b5e6-96231b3b80d8
do not insert a prototype for malloc of: void* malloc(uint): on 64-bit u
targets this is not correct. Instead of prototype it as void *malloc(...),
and pass the correct intptr_t through the "...".
Finally, fix Regression/CodeGen/SparcV9/2004-12-13-MallocCrash.ll, by not
forming constantexpr casts from pointer to uint.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18908 91177308-0d34-0410-b5e6-96231b3b80d8
Implement three new functions to allow setting access/permission bits on
the file referenced by a path. The makeReadable and makeExecutable methods
replace the FileUtilities MakeFileReadable and MakeFileExecutable
functions. The makeWritable function is new and provided for consistency
since Path has a writable() method.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18907 91177308-0d34-0410-b5e6-96231b3b80d8
The ReadFileIntoAddressSpace and UnmapFileFromAddressSpace functions are no
longer used by LLVM. Remove them. Replacement functionality for both
functions is now encapsulated in the sys::MappedFile class.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18903 91177308-0d34-0410-b5e6-96231b3b80d8
Use sys::MappedFile instead of ReadFileIntoAddressSpace and
UnmapFileFromAddressSpace. sys::MappedFile has the nice property that it
cleans up after itself so exception handling can be removed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18902 91177308-0d34-0410-b5e6-96231b3b80d8
don't support long double anyway, and this gives us FP results closer to
other targets.
This also speeds up 179.art from 41.4s to 18.32s, by eliminating a problem
with extra precision that causes an FP == comparison to fail (leading to
extra loop iterations).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18895 91177308-0d34-0410-b5e6-96231b3b80d8
The getFileTimestamp and getFileSize functions have been removed from \
FileUtilities.{h,cpp}. They are replaced by Path::getTimestamp and \
Path::getSize,respectively.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18892 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix loop style per standards
* Don't create a new Module when the Linker's module is released.
* Add/fix function comments.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18871 91177308-0d34-0410-b5e6-96231b3b80d8
* Remove IsLibrary and GetLibraryPath, replaced by FindLibrary in ../Path.cpp \
* Implement GetSystemLibraryPaths and GetBytecodeLibraryPaths, instead of the \
GetSystemLibraryPath1 and GetSystemLibraryPath2 methods
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18862 91177308-0d34-0410-b5e6-96231b3b80d8
in SPEC, the subsequent optimziations that we are after don't play with
with FP values, so disable this xform for them. Really we just don't want
stuff like:
double G; (always 0 or 412312.312)
= G;
turning into:
bool G_b;
= G_b ? 412312.312 : 0;
We'd rather just do the load.
-Chris
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18819 91177308-0d34-0410-b5e6-96231b3b80d8
down to actually BE a bool. This allows simple value range propagation
stuff work harder, deleting comparisons in bzip2 in some hot loops.
This implements GlobalOpt/integer-bool.ll, which is the essence of the
loop condition distilled into a testcase.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18817 91177308-0d34-0410-b5e6-96231b3b80d8
if the other side is overdefined.
This allows us to fold conditions like: if (X < Y || Y > Z) in some cases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18807 91177308-0d34-0410-b5e6-96231b3b80d8
1. Actually increment the Statistic for the GV elim optzn
2. When resolving undef branches, only resolve branches in executable blocks,
avoiding marking a bunch of completely dead blocks live. This has a big
impact on the quality of the generated code.
With this patch, we positively rip up vortex, compiling Ut_MoveBytes to a
single memcpy call. In vortex we get this:
12 ipsccp - Number of globals found to be constant
986 ipsccp - Number of arguments constant propagated
1378 ipsccp - Number of basic blocks unreachable
8919 ipsccp - Number of instructions removed
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18796 91177308-0d34-0410-b5e6-96231b3b80d8
ctor parameters can be defaulted.
Print the transformed llvm code input to the instruction selector
when -print-machineinstrs is on, just like V9.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18794 91177308-0d34-0410-b5e6-96231b3b80d8
* Don't emit the Index * ElementSize multiply if Index is a constant.
* Use a shift, not a multiply, if ElementSize is 1/2/4/8.
* If ElementSize fits in the immediate field of SMUL, then put it there.
Fix a bug where struct offsets might be truncated (ConstantSInt::get is
now used instead of ConstantInt::get).
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18792 91177308-0d34-0410-b5e6-96231b3b80d8
In functions where we fully constant prop the return value, replace all
ret instructions with 'ret undef'.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18786 91177308-0d34-0410-b5e6-96231b3b80d8
This implements SCCP/ipsccp-basic.ll, rips apart Olden/mst (as described in
PR415), and does other nice things.
There is still more to come with this, but it's a start.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18752 91177308-0d34-0410-b5e6-96231b3b80d8
each pair. I think this fixes that.
One of these days, I swear I'm going to get the hang of C++ iterators.
Really.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18734 91177308-0d34-0410-b5e6-96231b3b80d8
a different algorithm that was extremely inefficient for instructions with
many operands.
This reduces the time of this code snippet from .23s for 176.gcc to 0.03s
in a debug build, which speeds up total llvm-dis time just barely.
It's more of a code cleanup than a speedup.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18685 91177308-0d34-0410-b5e6-96231b3b80d8
successor block. This turns cases like this:
x = a op b
if (c) {
use x
}
into:
if (c) {
x = a op b
use x
}
This triggers 3965 times in spec, and is tested by
Regression/Transforms/InstCombine/sink_instruction.ll
This appears to expose a bug in the X86 backend for 177.mesa, which I'm
looking in to.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18677 91177308-0d34-0410-b5e6-96231b3b80d8
Make only one print method to avoid overloaded virtual warnings when \
compiled with -Woverloaded-virtual
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18589 91177308-0d34-0410-b5e6-96231b3b80d8
Test the range of float constants to ensure we are not attempting to create a \
float constant using a double value that is out of range for a float
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18585 91177308-0d34-0410-b5e6-96231b3b80d8
Make sure to check isValueValidForType on floating point constants and give \
an error if the value is not valid, otherwise it would assert in the VMCore
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18584 91177308-0d34-0410-b5e6-96231b3b80d8
intrinsic lowering ever introduces constants.
Rename local symbols before printing function bodies, fixing 255.vortex
with the CBE!!!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18534 91177308-0d34-0410-b5e6-96231b3b80d8
All SPEC CFP 95 programs now work, though the JIT isn't loading -lf2c right
so they aren't testing correctly.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18499 91177308-0d34-0410-b5e6-96231b3b80d8
it. If a block has a terminator then it is certainly non-empty so the
verifier will not crash on it.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18484 91177308-0d34-0410-b5e6-96231b3b80d8
correctly link globals whose LLVM types do not match.
This fixes several of the F2C SPEC FP benchmarks, which were failing this
due to the implementation of common blocks used by f2c.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18465 91177308-0d34-0410-b5e6-96231b3b80d8
This fixes the ugly darwin "cannot find symbols starting with __" issue.
Thanks for Owen/resistor for testing this out for me.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18454 91177308-0d34-0410-b5e6-96231b3b80d8
Including alloca.h on Solaris brings in the prototype of strftime(), which
breaks compilation of CBE generated code.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18435 91177308-0d34-0410-b5e6-96231b3b80d8
instead of 80-bits of precision. This fixes PR467.
This change speeds up fldry on X86 with LLC from 7.32s on apoc to 4.68s.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18433 91177308-0d34-0410-b5e6-96231b3b80d8
* Make the linker find lib*.bca files now instead of lib*.bc since those
are what the makefiles now generate for bytecode archives.
* Make sure the linker only links archives when LinkLibraries is called.
Previously if it found a lib*.bc file and that file was a bytecode file,
it would link in the entire bytecode. This could make -lc -lc fail with
duplicate symbols error but it shouldn't as searching multiple libraries,
even the same one more than once, is permitted.
* Now that the above problems are corrected, implement the dependent libs
feature. After the module is linked with all specified libraries, the
LinkLibraries function will obtain the set of dependent libraries from
the linked modules and attemp to find and link against those libraries.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18428 91177308-0d34-0410-b5e6-96231b3b80d8
in scary and unknown ways before we promote it. This fixes the miscompilation
of 188.ammp that has been plauging us since a globalopt patch went in.
Thanks a ton to Tanya for helping me diagnose the problem!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@18418 91177308-0d34-0410-b5e6-96231b3b80d8