similar to: [LLVMdev] Shared libs?

On Sun, Jun 8, 2008 at 7:57 PM, Thomas Hudson <hudson at speakeasy.net> wrote: > I've been playing with LLVM for a couple of months now and really am > enjoying myself. I finally got around to doing something useful and > had started implementing bindings for Ruby. I've been working off the > tip of trunk and realized my build didn't have shared libs. I've spent

Eli Friedman wrote: > This isn't first-hand, but from what I remember hearing on IRC, > putting llvm into shared libraries caused a ridiculous explosion in > dynamic linking (and therefore startup) times. So there is no option > to make shared libraries, at least at the moment. Well, by tweaking configure and make options, I've managed to build LLVM 2.2 shared libraries on

On Monday 09 June 2008, Albert Graef wrote: > Unfortunately, that approach doesn't work on x86-64 with LLVM 2.2, > since some parts of the LLVM JIT apparently contain non-relocatable > code; I hope that this will be fixed in the forthcoming LLVM 2.3. Unfortunately it's not fixed in 2.3 :( I made a patch ([1]) for 2.2 and gave it to one of the developer, I guess he forgot about

Cyrille Berger wrote: > Unfortunately it's not fixed in 2.3 :( That's indeed unfortunate. On x86-64 the Pure interpreter currently is a 7MB behemoth, and most of that is LLVM. ;-) On 32 bit I have all that stuff in a separate runtime library, resulting in a 27K interpreter executable. It goes without saying that this makes a world of a difference. I don't care if LLVM is a shared

Hi, > Cool, many thanks! That's exactly what I've been looking for. :) If you > do get around updating the patch for 2.3, it would be very kind if you > could let me know. I have updated the patch for llvm 2.3, and open a bug report to follow the status on this subject: http://llvm.org/bugs/show_bug.cgi?id=2514 -- Cyrille Berger

On Jun 8, 2008, at 9:36 PM, Eli Friedman wrote: > > If you'd like to elaborate on why exactly you need shared libs, you > might get a better answer. > Because I'm writing an extension to a scripting language interpreter. Perl/Python/Ruby all need to be able to dynamically load the library when the user 'requires' or 'imports' the extension. Obviously I can

Hi Cyrille, concerning your patch, > [1] http://www.opengtl.org/download/X86JITInfo.cpp.pic.patch there are two minor glitches: s/#ifdef/#if/, s/#end/#endif/. Otherwise it works great over here! Pure builds fine with a shared runtime lib on 64 bit now, so I'm a happy camper. :) I would like to make this patch available on the Pure website, ok with you? Albert -- Dr. Albert Gr"af

Hi! I'm a new LLVM user. I want to start using the LLVM System v2.3. However, I don't want to spend all my time hunting, pecking, downgrading, and/or upgrading packages to get LLVM v2.3 running. I was wondering which Linux distrubutions(FC7, Ubuntu, NetBSD, etc) have required the least effort out of the box to start running LLVM v2.3. tia, Bernardo Elayda -------------- next part

hi, this is a very basic question -- sorry for posing it: how can i force R to print 0.0001 instead of 1e-04??? .--------------------. | > 0.0001 | | [1] 1e-04 | `--------------------' i tried the functions format, formatC, ... and changed options()$digits with no success! thanks for advice, tomy -- no signature

Hi, When compiling the cfrontend for Linux/PPC architecture on a powermac G5, i got the following error : -------------------------------Compilation error listing if [ -f stmp-dirs ]; then true; else touch stmp-dirs; fi make GCC_FOR_TARGET="/usr/local/home/cyrille/project/llvm/cfrontend/build/gcc/xgcc -B/usr/local/home/cyrille/project/llvm/cfrontend/build/gcc/

Hi, how about to provide a simple way to forward raw file descriptors through ssh tunnels. something which may provide a way to write something like : (echo 3; read > out3) |& exec 3<&p 4>&p echo 5 >| out5 exec 5<> out5 echo 1 | ssh -d 3:rd -d 4:wr -d 5:rw ' read <&3; echo $REPLY >&4 read; echo $REPLY read <&5; echo $REPLY >&5

Albert, > The problem with using -fPIC on x86 is not only the speed of the > generated code, I also get JIT-related segfaults in the Pure interpreter > (on Linux). See, e.g.: Please fill a PR in the LLVM bugzilla, so this issue won't get lost. -- With best regards, Anton Korobeynikov Faculty of Mathematics and Mechanics, Saint Petersburg State University

Hi all, I noticed that now that --enable-pic is the default in svn, there's also some logic in Makefiles.rules to eliminate -fPIC for the case of mingw and cygwin: ifeq ($(ENABLE_PIC),1) ifeq ($(OS), $(filter $(OS), Cygwin MingW)) # Nothing. Win32 defaults to PIC and warns when given -fPIC else ... I would suggest that it should be done this way (i.e., eliminating -fPIC)

> Okay, I took a closer look at your output. > > I was correct in that gccas is attempting to assemble a native PPC > assembly language file. However, this appears to be happening because > the assembly file is generated by the GCC Makefiles for libgcc2 (in > llvm-gcc/gcc/config/rs6000/t-ppccomm, I think). > > The first thing I would try is Marco's suggestion: add

On Sun, Nov 29, 2009 at 2:19 AM, Albert Graef <Dr.Graef at t-online.de> wrote: > Jon Harrop wrote: >> I've come up with the following minimal repro that segfaults on my machine: > > Jon, were you able to resolve this? > > FWIW, TOT is causing all kinds of weird segfaults related to tail calls > in my Pure interpreter, too (at least on x86-64). In my case these

Jon Harrop wrote: > I've come up with the following minimal repro that segfaults on my machine: Jon, were you able to resolve this? FWIW, TOT is causing all kinds of weird segfaults related to tail calls in my Pure interpreter, too (at least on x86-64). In my case these seem to be limited to the JIT, however (batch-compiled Pure programs via opt+llc all work fine, even with TCO), so

On Mon, Jun 13, 2005 at 11:02:10AM -0500, John Criswell wrote: > Cyrille Mescam wrote: > >Hi, > > > > > >When compiling the cfrontend for Linux/PPC architecture on a powermac > >G5, i got > >the following error : > > I hate to ask a silly question, but are you using "make bootstrap" to > build the cfrontend? Your output looks like gccas

I am using speex in a voip project.I use an ARM 9 based cpu. all is ok but on problem exists. after I pause my speaking a while, for example, 2 minutes, the other peer hearsthe voice delay. I have made many test to lookup its reason, it\'s seems that decode process needs more CPU power in this case. I use mono 8KHz , encoding without VAD and preprocessor. who can tell me how to adjust some

Cyrille Mescam wrote: > On Mon, Jun 13, 2005 at 11:02:10AM -0500, John Criswell wrote: > >>Cyrille Mescam wrote: >> [snip] > > In fact, i am using the CFE build procedures. I am on instructions 5. > > cyrille > Okay, I took a closer look at your output. I was correct in that gccas is attempting to assemble a native PPC assembly language file. However,

Dear All, I am considering a possibility of using LLVM JIT for an algebraic modelling language. I have already done some prototyping following the Kaleidoscope tutorial and currently thinking of how to connect the jitted code to a runtime library (for this language) which I would like to code in C++. If it was *NIX I would use g++ possibly with '-rdynamic' option as suggested in the