similar to: [LLVMdev] Problem linking class to a pass in a loadable module

Dear LLVMers, I have a class X that represents a set of Value* in a specific way. That class is stored in two files: X.cpp and X.h The class X works perfectly in a pass that I already have. However, I need to use that class in some other passes, but I don't want to have a copy of X.cpp and X.h in every pass folder. I've tried to do the following: - put X.h in the folder

Hello, I'm having a little problem with passes that use LoopInfo. Their statistics aren't the same when I execute the same pass multiple times in the same program. After some investigation, I've found out that LoopInfo is the root of the problem. I wrote a little pass that counts the loop headers (number of basic blocks for which LoopInfo.isLoopHeader(BB) is true). I've picked

[Please CC the mailing list] Eli Gottlieb <eligottlieb at gmail.com> writes: > OK, I'm just going to paste in my CMakeLists.txt file. Like I said, > I'm building an LLVM install myself by untarring llvm, mkdir build in > the root dir of the source, cd build/, cmake .., make. > >> cmake_minimum_required (VERSION 2.6) >> project (libjllvm) >>

Hi, The LLVM has a good test suite. I recommend it. You can test your pass with single-source programs and multi-source programs. You can also configure the test-suite to run your test with external benchmarks within the LLVM test suite. You can also run some programs after you use your pass, to verify if your pass don't affect the behavior of the compiled program. You can do it with the

I compiled and installed it to the prefix /usr, but that's not the issue. Once I actually compile and install LLVM with CMake by hand, I get the share/llvm/cmake stuff installed correctly (can those files be included in "normal" builds, or will LLVM switch to CMake as its primary build system?). Now I'm running into the problem of cflags or includes or something not being

Hi again, I'm trying to upgrade my LLVM bindings in Java from 2.9 to 3.1. To do so, I regenerated the JNI bindings from fresh LLVM 3.1 headers, and did a slight rewrite of my CMakeLists.txt file for building the C code. Problem is, cmake no longer finishes at all. I receive the following output, and then it just runs forever (while still responding to a CTRL-C): > eli at

Dear all, I want to build an LLVM pass by using CMake. After reading the ralevant part of the documentation ( http://llvm.org/releases/3.1/docs/CMake.html#passdev), I copied the files in the llvm/lib/Transforms/Hello to another folder in order to give a try. Then I renamed the folder as Hello2. Then I have changed CMakeLists.txt as following: cmake_minimum_required(VERSION 2.8) # A

Dear all, I am trying to move a project from Linux to OSX, with bad results at my first attempt. First, I am using LLVM 3.2, the latest version, compiled from sources (configured with just the install prefix, vanilla for other options). By the way, I found it annoying to reinstall LLVM, I expected a complete and working LLVM environment with Apple's move to CLANG, but maybe I didn't

On Tue, Oct 9, 2012 at 9:25 PM, Jun Koi <junkoi2004 at gmail.com> wrote: > On Fri, Oct 5, 2012 at 3:45 PM, Jun Koi <junkoi2004 at gmail.com> wrote: >> On Fri, Oct 5, 2012 at 3:39 PM, David Chisnall >> <David.Chisnall at cl.cam.ac.uk> wrote: >>> On 5 Oct 2012, at 08:34, Jun Koi wrote: >>> >>>> any idea on how to fix the problem?

Thanks for responding, Duncan, and clarifying that y'all need more info to help. I'm trying to compile binaries on os x 10.5.8 intel hardware that are compatible on ppc os x 10.4. When I include various flags to llvm-gcc, including: -m32 -arch ppc -isysroot /Developer/SDKs/MacOS10.4u.sdk -mmacosx-version-min=10.4 I am seeing errors when compiling using llvm-gcc 4.2. If I leave out

Thanks for the warning! I checked and the executables take up about 2/3 of the space (380MB) vs. about 180MB for the libs. Since we are usually only debugging one executable at a time (if any), I modified Makefile.common to strip each executable in tools/Debug by default. You can define KEEP_SYMBOLS in the Makefile of any particular tool in order to avoid stripping that executable. I have

sam lee <skynare at gmail.com> writes: > Hi, > > I built LLVM on Windows XP through cmake and MinGW make. That was a breeze! > Now, I wanted to start a project. > It seems like I need Makefile.rules and Makefile.config to start a new > project (from http://llvm.org/docs/Projects.html). Those files are unrelated to cmake. > And llvm/projects/sample uses autoconf (or

Dear llvmers, I've noticed that since the version 2.7 we don't have a pass named gnvpre anymore. Is there any pass that does PRE in the current version of LLVM? Regards, Raphael Ernani

Ah, thank you so much. How did you get the list of target_link_libraries() ? Did you use llvm-config? If you did not have access to llvm-config, how would you get list of libraries to link to? Thanks. Sam On Sun, Jun 13, 2010 at 2:44 PM, Nathan Jeffords <blunted2night at gmail.com>wrote: > This isn't meant to be a sample, but I can share with you the cmake files I > use for my

This isn't meant to be a sample, but I can share with you the cmake files I use for my compiler. It builds in both windows user MSVC, and linux using G++, but should work with MinGW. the structure is src/ CMakeLists.txt # root make file front/ CMakeLists.txt # library middle/ CMakeLists.txt # library back/ CMakeLists.txt # library npp/ CMakeLists.txt # executable

erkan diken <erkandiken at gmail.com> writes: [snip] > It seems that cmake is ok. When I run make install, it gives errors (see > below) and i think the reason is that it can not find the directory to llvm > header files. > I could not figure it out which variable to set and how to use it in order > to point the required directory ? > ( as in the make pass build which

> Vladimir, > > I took a *very quick* look at your patch. I won't be able to look at it in > more detail until this weekend, but perhaps you can address a couple of > concerns: > > 1. I don't understand the need for LLVMToolDirSlash. This seems like a > gratuitous change to me. If LLVM_OBJ_ROOT (with LLVM_SRC_ROOT) and LLVM_ROOT both not set without

I needed the following patch to be able to use gprof with profiled build. regards, Benoit --- a/Makefile.rules (revision 37946) +++ b/Makefile.rules (working copy) @@ -212,6 +212,7 @@ CXX.Flags := $(OPTIMIZE_OPTION) -pg -g C.Flags := $(OPTIMIZE_OPTION) -pg -g LD.Flags := $(OPTIMIZE_OPTION) -pg -g + KEEP_SYMBOLS := 1 else ifeq ($(ENABLE_OPTIMIZED),1) BuildMode := Release

I built the list by hand though trial and error. I'm not sure if and how llvm-config works with cmake. Its seems now that I got the right dependencies in the right order, there have been no ongoing maintenance issues. On Sun, Jun 13, 2010 at 4:32 PM, sam lee <skynare at gmail.com> wrote: > Ah, thank you so much. > How did you get the list of target_link_libraries() ? > Did you

Eli Gottlieb <eligottlieb at gmail.com> writes: > The two lines >> set(CMAKE_MODULE_PATH ${CMAKE_MODULE_PATH} >> "${LLVM_ROOT}/share/llvm/cmake") >> include(LLVM) > don't seem to work. I'm checking with installing LLVM from source to > see if that makes it work, but the LLVM install from the Ubuntu repos > doesn't seem to have a