similar to: [LLVMdev] gcc 4.2 to llvm-gcc 4.2 transition

On Jul 24, 2010, at 4:41 AM, Alexandre Colucci wrote: > I am currently studying the possibility to make the transition from gcc 4.2 to llvm-gcc 4.2 for the projects I am working on. > These projects are huge projects, most of the source code is written in C++ but we have objc, c and assembly code. We are targeting MacOSX 10.5 and MacOSX 10.6. Ok, great! > Our build system is currently

On Sat, 24 Jul 2010 13:41:42 +0200 Alexandre Colucci <timac at timac.org> wrote: > Hi, > > I am currently studying the possibility to make the transition from > gcc 4.2 to llvm-gcc 4.2 for the projects I am working on. Since you are switching compilers, why not switch to clang instead of llvm-gcc? Best regards, --Edwin

There are several reasons why I am currently not considering the transition to Clang (although I would love to switch to it): - Clang (LLVM compiler 2.x) is only integrated by default in Xcode 4 which has not been released yet. I guess it might be possible to use Clang in Xcode 3.2 and Xcode 3.1. - Clang is a new compiler and the C++ support is really new and might contain bugs. Moreover Apple

Hey, My projects are not compiling on 10.5 using llvm-gcc 4.2 in Xcode 3.1.4. These are caused by some llvm bugs that are now fixed. Since Apple will not release a Xcode update for 10.5, I am trying to compile a recent version of llvm-gcc 4.2 (from the LLVM 2.7 branch - but I could use ToT if this helps) to use on 10.5 intel. I easily built llvm-gcc-4.2 on 10.5 intel by following the README.LLVM

Hi, llvm-gcc 4.2 generates warnings when I compile inline assembly code that contains used labels with -Wunused-label. The generated code seems to work yet. gcc 4.2 doesn't generate those warnings. I haven't found any bugs regarding this issue in the llvm bug database. Does anyone know if this is a known llvm issue? Is it a warning that I can ignore and does not affect the generated

Hello > Did someone manage to update llvm-gcc 4.2 contained in Xcode 3.1.4? Is there a trick? Does anyone know how to enable the -arch option? You have to build llvm-gcc in so-called "Apple-way". Read README.Apple for more information. Note, however, that it describes the non-LLVM-enabled build, so you should tune the cmdlines mentioned there accordingly. -- With best regards,

On 10.10.2011, at 16:10, Alexandre Colucci wrote: > Hi, > > I am currently making the transition from gcc 4.2 to clang for the projects (mostly C++) I am working on. > I think I discovered a major optimization bug in Clang++. I managed to create a simple (60 lines of code) test case which exhibits the issue. > > When you compile this file for i386 with optimizations turned on

Hi, I am currently making the transition from gcc 4.2 to clang for the projects (mostly C++) I am working on. I think I discovered a major optimization bug in Clang++. I managed to create a simple (60 lines of code) test case which exhibits the issue. When you compile this file for i386 with optimizations turned on (O2, O3 or Os), you get an unexpected result. When you compile it for x86_64, or

Hi, I think I discovered a major armv7 optimization bug in Clang. I create a simple test case which exhibits the issue. When you compile the attached file for armv7 with optimizations turned on (O2, O3 or Os), the binary generated led to a crash. The issue can't be reproduced when using GCC 4.2. It can't be reproduced with Clang when the optimization is turned off (O0). This issue can be

This is fixed in TOT (rev 107051 for C, 108732 for C++) On Jul 23, 2010, at 2:20 AMPDT, Alexandre Colucci wrote: > Hi, > > llvm-gcc 4.2 generates warnings when I compile inline assembly code > that contains used labels with -Wunused-label. The generated code > seems to work yet. gcc 4.2 doesn't generate those warnings. I > haven't found any bugs regarding this

In practice all Apple hardwares support misaligned accesses for single-register loads and stores. If a pointer is not aligned, LLVM should not use the double-register loads and stores. It should keep the two single-register loads instead of trying to optimize them as one unsupported double-register load. Note that this code compiled with GCC 4.2 runs perfectly whereas LLVM will produce a binary

Hi...Frustrated!! I have been trying to install via the directions on http://wiki.winehq.org/MacOSX/Building I am trying to install with "minimum fuss" I am using a macbook pro with Jaguar 10.5.8, xcode 3.1 and quartz x11. Downloads and configures fine, but when I "make" It ends with the following x11drv_main.c: In function ?process_attach?: x11drv_main.c:527: error:

On Jul 24, 2010, at 10:15 AM, Alexandre Colucci wrote: > > There are several reasons why I am currently not considering the transition to Clang (although I would love to switch to it): > > - Clang (LLVM compiler 2.x) is only integrated by default in Xcode 4 which has not been released yet. I guess it might be possible to use Clang in Xcode 3.2 and Xcode 3.1. Right. > - Clang

The problem is in your code, not the compiler. You're casting an unaligned char* to an int*, even though an int* pointer must be 4-byte aligned in every ARM ABI that I've ever seen. On Jan 23, 2012, at 6:14 AM, Alexandre Colucci wrote: > Hi, > > I think I discovered a major armv7 optimization bug in Clang. I create a simple test case which exhibits the issue. > When you

I was wondering if Wine lost 10.5 compatibility, or if any of the few people out there with it have noticed a problem. Up through 1.3.25 it seems to compile and work fine. On 1.3.26 and 1.3.27 the compile will fail with errors about a bunch of symbols in msxml3.dll.so that are not found. Building 1.3.26 and 1.3.27 on 10.6 or 10.7 using a 10.5 SDK it will build fine and still work on 10.6 and

On 24.07.2010, at 21:27, Thomas Clement wrote: > On 24 juil. 2010, at 20:29, Chris Lattner wrote: > >>> - I am not sure if clang can generate code that runs on 10.5. >> >> It can, but it does not support powerpc (llvm-gcc does). > > Really? I couldn't find any Apple documentation that mentions this. > Also we have built powerpc code with the clang

Looking through the documentation, I discovered that atomic loads and stores are only supported for integer types. Can anyone provide some background on this? Why is this true? Currently, given code: std::atomic<float> aFloat; void foo() { float f = atomic_load(&aFloat); .. } Clang generates code like:|| %"struct.std::atomic.2" = type { float } @aFloat = global

Yo fellas! I'm having some problems with WineBottler. I'm trying to connect to a ventrilo server with a gsm codec (I think) which isn't supported on mac os. A friend who is on the same server uses regular wine and it works for him, but are there any way to make it work with WineBottler as well? Or should I just do it the "hard" wy and install Wine? Couldn't find XCode

On 24 juil. 2010, at 20:29, Chris Lattner wrote: >> - I am not sure if clang can generate code that runs on 10.5. > > It can, but it does not support powerpc (llvm-gcc does). Really? I couldn't find any Apple documentation that mentions this. Also we have built powerpc code with the clang "LLVM compiler 1.5" and it does run on powerpc. Should we use the llvm-gcc

Hello everyone, I'm not sure whether this is the best place to send the question, but since I'm mostly interested in FLAC and there didn't appear to be a better place for it.. here goes. I pulled the latest source code from the SVN repo along with all the external dependencies including FLAC. I've been able to compile the FLAC.framework file using the 10.5 SDK in Xcode 3.2 with a