C is a constant which can be sign-extended from 8 bits without value loss,
and op is one of: add, sub, imul, and, or, xor.
This allows the JIT to emit the one byte version of the constant instead of
the two or 4 byte version. Because these instructions are very common, this
can save a LOT of code space. For example, I sampled two benchmarks, 176.gcc
and 254.gap.
BM Old New Reduction
176.gcc 2673621 2548962 4.89%
254.gap 498261 475104 4.87%
Note that while the percentage is not spectacular, this did eliminate
124.6 _KILOBYTES_ of codespace from gcc. Not bad.
Note that this doesn't effect the llc version at all, because the assembler
already does this optimization.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9284 91177308-0d34-0410-b5e6-96231b3b80d8
* Implement R1 = R2 * C where R1 and R2 are 32 or 16 bits. This avoids an
extra copy into a register, reducing register pressure.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9278 91177308-0d34-0410-b5e6-96231b3b80d8
getelementptr code path for use by other code paths (like malloc and alloca).
* Optimize comparisons with zero
* Generate neg, not, inc, and dec instructions, when possible.
This gives some code size wins, which might translate into performance. We'll
see tommorow in the nightly tester.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9267 91177308-0d34-0410-b5e6-96231b3b80d8
all, making it pointless for use with the code generator debugger. With it,
it works like a charm.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9245 91177308-0d34-0410-b5e6-96231b3b80d8
* More dense bytecode encoding for varargs calls (like printf)
* Eliminated the extremely old bytecode format. rev #0 is now 1.0
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9220 91177308-0d34-0410-b5e6-96231b3b80d8
ModuleProvider, which has bad consequences in lli::callAsMain() which tries to
access that same Module*.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9205 91177308-0d34-0410-b5e6-96231b3b80d8
Make insertFarJumpAtAddr() return void, because nothing uses its return value.
Remove some commented-out code.
Implement replaceMachineCodeForFunction() for SPARC.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9203 91177308-0d34-0410-b5e6-96231b3b80d8
runJITOnFunction().
Add new method for recompiling and patching in new versions of functions,
recompileAndRelinkFunction().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9199 91177308-0d34-0410-b5e6-96231b3b80d8
Rename SlotCalculator::getValSlot() to SlotCalculator::getSlot(),
SlotCalculator::insertValue() to SlotCalculator::getOrCreateSlot(),
SlotCalculator::insertVal() to SlotCalculator::insertValue(), and
SlotCalculator::doInsertVal() to SlotCalculator::doInsertValue().
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9190 91177308-0d34-0410-b5e6-96231b3b80d8
For now, we translate linkonce into weak linkage in the bytecode format because
we don't have enough bits to represent it. We will rev the bytecode version
soon anyways, so this will be fixed in the near future.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9170 91177308-0d34-0410-b5e6-96231b3b80d8
this list (except use_size()) are constant time. Before the killUse method
(used whenever something stopped using a value) was linear time, and thus
very very slow for large programs.
This speeds GCCAS up _substantially_ on large programs: almost 2x for 176.gcc:
176.gcc: 77.07s -> 37.38s
177.mesa: 7.59s -> 5.57s
252.eon: 21.02s -> 19.52s (*)
253.perlbmk: 11.40s -> 13.05s
254.gap: 7.25s -> 7.42s
252.eon would speed up a whole lot more, but optimization time is being
dominated by the inlining pass, which needs to be fixed.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9160 91177308-0d34-0410-b5e6-96231b3b80d8
* FunctionPassManager ctor now takes in a ModuleProvider
* run() materializes function before running passes on it
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9126 91177308-0d34-0410-b5e6-96231b3b80d8
* Add header comment
* Remove extraneous #includes
* Move the FileType enum into the GCC class
* The GCC class is not virtual.
* Move all of the "constructor" functions into the classes themselves
* Stop using cl::list as arguments, use std::vector instead (which cl::list
derives from)
* Improve comments
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9121 91177308-0d34-0410-b5e6-96231b3b80d8
X86/linux. :( The problem is that a signal delivered while the function
is executing could clobber the functions stack. This is a partial fix
for PR41.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9113 91177308-0d34-0410-b5e6-96231b3b80d8
break dominance relationships, and is otherwise bad. This fixes bug:
Inline/2003-10-13-AllocaDominanceProblem.ll. This also fixes miscompilation
of 3 176.gcc source files (reload1.c, global.c, flow.c)
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9109 91177308-0d34-0410-b5e6-96231b3b80d8
jello statistic for this (just divide #-bytes-of-code-emitted by 4).
Rewrite head-of-file comment.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9098 91177308-0d34-0410-b5e6-96231b3b80d8
multiple times. This reduces the time to construct post-dominance sets a LOT.
For example, optimizing perlbmk goes from taking 12.9894s to 1.4074s.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9091 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix a nasty initializer ordering bug. Any only-CFG passes which registered
themselves before the CFGOnlyAnalysis vector initialized got forgotten and
thus got invalidated and recomputed.
In particular, in my compiled version of gccas, the Loop information pass was
being recomputed unnecessarily.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9074 91177308-0d34-0410-b5e6-96231b3b80d8
Only transform call sites in a setjmp'ing function which are reachable from
the setjmp. If the call dominates the setjmp (for example), the called
function cannot longjmp to the setjmp.
This dramatically reduces the number of invoke instructions created in some
large testcases.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9066 91177308-0d34-0410-b5e6-96231b3b80d8
have a SINGLE backedge. This is useful to, for example, the -indvars pass.
This implements testcase LoopSimplify/single-backedge.ll and closes PR#34
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9065 91177308-0d34-0410-b5e6-96231b3b80d8
* Print floating point values using C99 hexadecimal style FP if possible.
This increases the number of floating point constants that may be emitted
inline, and improves precision for global variable initializers which
can not be emitted in integer form.
This fixes the Olden/Power benchmark with the CBE!!!!
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9052 91177308-0d34-0410-b5e6-96231b3b80d8
* Fix isFPCSafeToPrint to find more constants safe to print, which it was
failing because ftostr was padding with leading space characters.
* Scan the entire module for global constants instead of each function at a
time. This has the advantage of allowing us to emit constants at global
scope instead of function scope. This speeds FP programs quite a bit.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9048 91177308-0d34-0410-b5e6-96231b3b80d8
Linux. This is consistent with what FreeBSD and Solaris both want.
This makes the JIT work on FreeBSD 5.1-RELEASE. Whee.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9045 91177308-0d34-0410-b5e6-96231b3b80d8
In lookupFunction():
Change to use "F" for Function argument instead of ancient "M".
Remove commented-out code.
Change to use GetAddressOfSymbol instead of dlsym.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9013 91177308-0d34-0410-b5e6-96231b3b80d8
are ordered by name, not by slot, so the previous solution wasn't any good.
On a large testcase, this reduces time to parse from 2.17s to 1.58s.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@9002 91177308-0d34-0410-b5e6-96231b3b80d8
changes:
* BytecodeReader::getType(...) used to return a null pointer
on error. This was only checked about half the time. Now we convert
it to throw an exception, and delete the half that checked for error.
This was checked in before, but psmith crashed and lost the change :(
* insertValue no longer returns -1 on error, so callers don't need to
check for it.
* Substantial rewrite of InstructionReader.cpp, to use more efficient,
simpler, data structures. This provides another 5% speedup. This also
makes the code much easier to read and understand.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8984 91177308-0d34-0410-b5e6-96231b3b80d8
simplifies the control flow a bit. This provides a small (~3%) speedup,
but it's primarily a cleanup exercise.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8983 91177308-0d34-0410-b5e6-96231b3b80d8
new, simpler, ForwardReferences data structure. This is just the first
simple replacement, subsequent changes will improve the code more.
This simple change improves the performance of loading a file from HDF5
(contributed by Bill) from 2.36s to 1.93s, a 22% improvement. This
presumably has to do with the fact that we only create ONE placeholder for
a particular forward referenced values, and also may be because the data
structure is much simpler.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8979 91177308-0d34-0410-b5e6-96231b3b80d8
in the bytecode parser. Before we tried to shoehorn basic blocks into the
"getValue" code path with other types of values. For a variety of reasons
this was a bad idea, so this patch separates it out into its own data structure.
This simplifies the code, makes it fit in 80 columns, and is also much faster.
In a testcase provided by Bill, which has lots of PHI nodes, this patch speeds
up bytecode parsing from taking 6.9s to taking 2.32s. More speedups to
follow later.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8977 91177308-0d34-0410-b5e6-96231b3b80d8
to avoid reprocessing PHI nodes needlessly. This speeds up the big bad PHI
testcase 43%: from 104.9826 to 73.5157s
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8964 91177308-0d34-0410-b5e6-96231b3b80d8
of a test that Bill Wendling sent me from 228.5s to 105s. Obviously there is
more improvement to be had, but this is a nice speedup which should be "felt"
by many programs.
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8962 91177308-0d34-0410-b5e6-96231b3b80d8
and TargetInstrDescriptor::ImplicitUses to always point to a null
terminated array and never be null. So there is no need to check for
pointer validity when iterating over those sets. Code that looked
like:
if (const unsigned* AS = TID.ImplicitDefs) {
for (int i = 0; AS[i]; ++i) {
// use AS[i]
}
}
was changed to:
for (const unsigned* AS = TID.ImplicitDefs; *AS; ++AS) {
// use *AS
}
git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@8960 91177308-0d34-0410-b5e6-96231b3b80d8
