Displaying 15 results from an estimated 15 matches for "bugspy".
Did you mean:
bugs
2010 Mar 18
2
[LLVMdev] JIT : Does it cache the compiled code of executed functions upon runFunction(..)?
...- Does section 4 above makes the EE recompile the function
1000 times? If no, why is the performance is so bad? what am I doing
wrong ? Maybe I should somehow cache the function's binary code?
llvm-version 2.7
OS: Linux 2.6.31 64 bits (intel dual core, 4GB memory)
--
Regards,
Gabi
http://bugspy.net
2010 Mar 26
1
[LLVMdev] Using GetElementPtr to traverse large arrays
...eate(intVars, idx,
"ElemPtr", someBlock);
new StoreInst(idx, gep, entryBlock); //Store some value in it
}
Apart from the size issue, What is the idiomatic way to traverse small
arrays ? The above loop ? I tried to use CreateInBounds but got
totally confused..
--
Regards,
Gabi
http://bugspy.net
2010 Mar 19
2
[LLVMdev] JIT : Does it cache the compiled code of executed functions upon runFunction(..)?
...;getOrInsertFunction(name,
Type::getFloatTy(ctx), Type::getFloatTy(ctx), (Type*)NULL);
And the orginal slow execution was:
std::vector<GenericValue> args(1);
args[0].FloatVal = 8.0f;
GenericValue retVal = EE -> runFunction(f, args);
return retVal.FloatVal;
--
Regards,
Gabi
http://bugspy.net
2010 Mar 18
0
[LLVMdev] JIT : Does it cache the compiled code of executed functions upon runFunction(..)?
...nce yourself :
runFunction -> getPointerToFunction ->getPointerToGlobalIfAvailable which
returns the native address of the jitted function.
You can even try to measure time needed by each runFunction to run to
double-convince yourself. :)
Olivier.
On Thu, Mar 18, 2010 at 9:21 AM, Gabi <bugspynet at gmail.com> wrote:
> Hello
> I have the following scenario, and I am not sure why the performance
> is so bad (take 30 minutes to complete with very simple generated
> functions):
>
> 1. Create module
> 2. Do something like EE =
> EngineBuilder(theModule).setEngineK...
2010 Mar 23
3
[LLVMdev] How to avoid memory leaks
...gen/Debug/vgen)
==17474== by 0x88423B: llvm::ConstantInt::get(llvm::LLVMContext&,
llvm::APInt const&) (in /home/gabi/vgen/Debug/vgen)
==17474== by 0x8843B4: llvm::ConstantInt::get(llvm::IntegerType
const*, unsigned long, bool) (in /home/gabi/vgen/Debug/vgen)
--
Regards,
Gabi
http://bugspy.net
2010 Mar 18
0
[LLVMdev] Does it cache the compiled code of executed functions upon runFunction(..)?
...the cause. It appears that FP operations and functions
(I used combinations of CreateFMulL CreateFAdd and CreateFSub ) is
MUCH slower than their integer equivalents.
I mean really slower (20x slower or even more)
Can It be the cause ? Does the FP binary ops are so slow ?
--
Regards,
Gabi
http://bugspy.net
2010 Mar 23
0
[LLVMdev] How to avoid memory leaks
...e`. (We
need to reproduce your problem locally in order to fix it.) Note that
if the leaks appear on
http://google1.osuosl.org:8011/builders/llvm-x86_64-linux-vg_leak/builds/5/steps/test-llvm/logs/stdio
we already know about them and are working on fixes.
On Tue, Mar 23, 2010 at 12:42 PM, Gabi <bugspynet at gmail.com> wrote:
> Hi Jeffrey,
> Listed below the Full valgrind report (using latest revision r99309)
> The program creates many thousands of instructions and values as you
> can see from the report below
>
> ...
2010 Apr 06
1
[LLVMdev] PrintModule: How to make it show floats in decimal instead of hex?
Is there a way to make the printer to display float values in decimal
form instead of the cryptic 0x... ?
--
Regards,
Gabi
http://bugspy.net
2010 Apr 06
2
[LLVMdev] PrintModule: How to make it show floats in decimal instead of hex?
> PS: Maybe not for x86 long double though.
Right, so how to make x86 double displayed in decimal ?
--
Regards,
Gabi
http://bugspy.net
2010 Apr 14
1
[LLVMdev] Most efficient way to count # of intrcutions in a module ?
...or f = mod->begin(), fe = mod->end(); f != fe; ++f)
for (Function::iterator b = f->begin(), be = f->end(); b != be; ++b)
{
BasicBlock* bb = (BasicBlock*) b;
size+=bb->getInstList().size();
}
return size;
}
--
Regards,
Gabi
http://bugspy.net
2010 Apr 24
1
[LLVMdev] LLVM ERROR: Cannot yet select: 0x2625340: f64 = ConstantFP<3.540000e+02> :What is it?
I get this message once in a while and llvm exits the program.
What might be the cause ?
llvm version: svn
host: Linux x86_64
--
Regards,
Gabi
http://bugspy.net
2010 Apr 25
2
[LLVMdev] LLVM ERROR: Cannot yet select: 0x2625340: f64 = ConstantFP<3.540000e+02> :What is it?
Hi Christoph
I am compiling for x86-64. This error happens randomly (at least it
appears that way) and rarely. About every few thousand runs
--
Regards,
Gabi
http://bugspy.net
2010 Mar 19
0
[LLVMdev] JIT : Does it cache the compiled code of executed functions upon runFunction(..)?
...is not cached and lives as long as
the JIT, so if you use runFunction repeatedly you will be generating
many snippets of machine code and leaking memory.
The leak could be fixed by freeing the stub before returning its return value.
Hope that helps,
Reid
On Fri, Mar 19, 2010 at 5:27 AM, Gabi <bugspynet at gmail.com> wrote:
> Reid,
>
> Thanks! You were right!
> Changing the code to:
>
> float (*theF)(float) = (float (*)(float)) EE -> getPointerToFunction(f);
> float retVal = theF(arg1);
>
> made the difference. Now it is dozens of times faster!
>
> I don...
2010 Apr 04
1
[LLVMdev] Code generators (both llvmc and Jit) get stuck when dealing circular CFG
...s=1]
%"BrCounter++8" = add i32 %28, 1 ; <i32> [#uses=2]
store i32 %"BrCounter++8", i32* %BranchCounter
%29 = icmp ult i32 %"BrCounter++8", 10 ; <i1> [#uses=1]
br i1 %29, label %Body6, label %Exit
}
--
Regards,
Gabi
http://bugspy.net
2010 Mar 23
2
[LLVMdev] How to avoid memory leaks
...ich a pointer was found) are not shown.
==20504== To see them, rerun with: --leak-check=full --show-reachable=yes
==20504==
==20504== For counts of detected and suppressed errors, rerun with: -v
==20504== ERROR SUMMARY: 72 errors from 72 contexts (suppressed: 4 from 4)
--
Regards,
Gabi
http://bugspy.net