similar to: [LLVMdev] Floating Point Register Allocation in X86 backend

On May 25, 2011, at 11:09 AM, aparna kotha wrote: > Hi Guys, > > I was working on some floating point intensive benchmarks and realize that the floating point register allocation in llvm assumes that there are only 7 floating point registers in X86, whereas the hardware has 8. > > Line number > 00266 assert(Reg >= X86::FP0 && Reg <= X86::FP6 &&

Right. But there are 8 registers on the floating point stack from ST0 to ST7 and I think llvm is only using ST0 to ST6 in some code fragments. Could this be because of the assumption that X86::FP registers run from X86::FP0 to X86:FP6 ? --Aparna On Wed, May 25, 2011 at 2:28 PM, Jakob Stoklund Olesen <stoklund at 2pi.dk>wrote: > > On May 25, 2011, at 11:09 AM, aparna kotha wrote:

On May 25, 2011, at 12:08 PM, aparna kotha wrote: > Right. But there are 8 registers on the floating point stack from ST0 to ST7 and I think llvm is only using ST0 to ST6 in some code fragments. Could this be because of the assumption that X86::FP registers run from X86::FP0 to X86:FP6 ? Yes. My guess it that the code converting from FP to ST registers sometimes needs the extra stack slot.

On 20.10.2010 05:00, Jakob Stoklund Olesen wrote: > On Oct 19, 2010, at 6:37 PM, Roland Scheidegger wrote: > >> Thanks for giving it a look! >> >> On 19.10.2010 23:21, Jakob Stoklund Olesen wrote: >>> On Oct 19, 2010, at 11:40 AM, Roland Scheidegger wrote: >>> >>>> So I saw that the code is doing lots of register >>>>

Hi all: I need an instruction that can convert decimal values into floating point numbers. i.e. say I have a decimal number 1110794174 (== 42355FBE in hex ) and (== 45.3435 as a float) essentially the mantissa and exponent representation needs to be used. Is there any way of doing this in llvm? Thanks and Regards -- -- Aparna Kotha Graduate Student Electrical and Computer Engineering

Thanks for giving it a look! On 19.10.2010 23:21, Jakob Stoklund Olesen wrote: > On Oct 19, 2010, at 11:40 AM, Roland Scheidegger wrote: > >> So I saw that the code is doing lots of register >> spilling/reloading. Now I understand that due to calling >> conventions, there's not really a way to avoid this - I tried using >> coldcc but apparently the backend

On Oct 20, 2010, at 7:46 AM, Roland Scheidegger wrote: > On 20.10.2010 05:00, Jakob Stoklund Olesen wrote: >> Look in X86InstrControl.td. The call instructions are all prefixed >> by: >> >> let Defs = [RAX, RCX, RDX, RSI, RDI, R8, R9, R10, R11, FP0, FP1, FP2, >> FP3, FP4, FP5, FP6, ST0, ST1, MM0, MM1, MM2, MM3, MM4, MM5, MM6, MM7, >> XMM0, XMM1, XMM2, XMM3,

(repost with right sender address) On 20.10.2010 18:13, Jakob Stoklund Olesen wrote: > On Oct 20, 2010, at 7:46 AM, Roland Scheidegger wrote: > >> On 20.10.2010 05:00, Jakob Stoklund Olesen wrote: >>> Look in X86InstrControl.td. The call instructions are all prefixed >>> by: >>> >>> let Defs = [RAX, RCX, RDX, RSI, RDI, R8, R9, R10, R11, FP0, FP1,

Hi, I have an ir at the end of this email. Run it with: llc -mcpu=i386 -march=x86-64 -mattr=-sse2 and get assertion below. Changing cpu does not help. I am using llc from the latest svn repository. Any suggestions to work around this? I need to disable sse2 instructions for x86-64. Thanks, -Peng -----error message------ llc: X86FloatingPoint.cpp:332: unsigned int getFPReg(const

On Oct 19, 2010, at 6:37 PM, Roland Scheidegger wrote: > Thanks for giving it a look! > > On 19.10.2010 23:21, Jakob Stoklund Olesen wrote: >> On Oct 19, 2010, at 11:40 AM, Roland Scheidegger wrote: >> >>> So I saw that the code is doing lots of register >>> spilling/reloading. Now I understand that due to calling >>> conventions, there's not

I am using pthreads. I was also wondering what will the llvm IR be for pthreads ? On Tue, Feb 3, 2009 at 3:51 PM, Bill Wendling <isanbard at gmail.com> wrote: > On Tue, Feb 3, 2009 at 12:33 PM, aparna kotha <kotha.aparna at gmail.com> > wrote: > > Hi all: > > > > I am working on a project using llvm and we need to deal with > multithreaded > >

Thanks a lot. A see the shared library in linux version. I initially tried doing this in cygwin where I dont see the shared library being built. Do u think it could have been cygwin that is the problem? Aparna On Mon, Apr 13, 2009 at 5:16 PM, Milos Puzovic <milos.puzovic at gmail.com>wrote: > In order to generate shared libraries you need to have SHARED_LIBRARY > and

Hi all: I am working on a project using llvm and we need to deal with multithreaded applications. I wanted to know if there was a C front end for llvm that could parse multithreaded applications? I tried llvm-gcc (4.2) and could not get it to work. Is there an extra parameter that I need to pass or something ? Thanks a lot for your help. Regards -- -- Aparna Graduate Student Department

Hi LLVMers, I have finally sorted out licensing issues and found some time, so I'm trying to port my PBQP register allocator to 2.4 in order to contribute it (if you want it). I've run into a bug that has me confused though. I'm currently failing the following assertion: llc: VirtRegMap.cpp:1733: void<unnamed>::LocalSpiller::RewriteMBB(llvm::MachineBasicBlock&,

I am trying to run the LLVM hello world pass. I observe that in the OUPUT_DIR/lib the LLVMHello.a and LLVMHello.la libraries are generated. But I am confused as to how to generate the .so library file from here. I am kinda new to compilers and hence the question I know is basic , but couldnt figure the answer out myself. Thanks a lot for your help. Regards -- -- Aparna Kotha Research

All, I have a ModulePass (A) calling a FunctionPass that inturn calls TargetData (a ModulePass). For reasons of code correctness and modularity I cannot reorganize my passes in any other way . When I use opt to load and run A , it gives the following error. LLVM ERROR: Bad TargetData ctor used. Tool did not specify a TargetData to use? I gather from some previous emails that this was not

On Jan 19, 2012, at 10:12 AM, Joe Groff wrote: > 2012/1/19 Jakob Stoklund Olesen <stoklund at 2pi.dk>: >> How many of these libcalls do you need to implement? What exactly is the calling convention? Which registers are clobbered etc. > > There is only one (that I know about so far). The MSVCRT `_ftol2` > function implements floating-point-to-unsigned conversion for i386

I believe because cygwin behaves like windows (it is just a posix layer over window) it doesn't have dynamic linking hence only static libraries have been built. On 13 Apr 2009, at 23:16, aparna kotha wrote: > Thanks a lot. > > A see the shared library in linux version. I initially tried doing > this in cygwin where I dont see the shared library being built. Do > u

I am using opt , so i cm confused if i need to do this. On the other hand I modified my functionPass not to call TargetData , and I still get this error. My functionPass still calls Alias Analysis though. Aparna On Fri, Aug 13, 2010 at 2:15 PM, John Criswell <criswell at illinois.edu>wrote: > aparna kotha wrote: > >> All, >> I have a ModulePass (A) calling a

On Tue, Feb 3, 2009 at 12:55 PM, aparna kotha <kotha.aparna at gmail.com> wrote: > I am using pthreads. > > I was also wondering what will the llvm IR be for pthreads ? > Okay. Luke gave hints on how to get pthreads to work. LLVM doesn't do anything special for pthreads calls. So they should look like regular calls into a library. -bw