similar to: [LLVMdev] Need Advice: X86ShuffleDecode Problem

Hello, I am building LLVM with ninja on Linux environment and I continue to have the error below. I am thinking that the cause might be my PC not having sufficient RAM memory. In this sense, I extended my swap memory with a swap file (90GB), but it didn't solved the problem. Should I add more physical RAM memory to my PC, or is there any software based solution I can try first? Thank you and

Thanks for the reply. Please see my inline responses. On 15/08/13 12:58, Tim Northover wrote: > Hi Dan, > >> It is not stated how the "<ty>" type is used as we are told the >> type of result is type* and sizeof(<type>)*NumElements is the >> amount of memory so from my perspective it looks like <ty> is >> never used which would seem to

Hi! currently instcombine does not detect constant diagonal matrix times vector, for example a.xx * [2 0] + a.yy * [0 3] can be optimized to a * [2 3] I have implemented this for float. I know that this assumes x * 0 = 0 which is not ieee compliant but i post it here in case it is interesting for someone. on my wish list there is still an option for target independent optimizations to have x

Hi, This is quite a simple question so hopefully it's easy to answer. I was looking at the documentation for the alloca instruction ( http://llvm.org/docs/LangRef.html#alloca-instruction ) and something is unclear to me. The given syntax is... <result> = alloca <type>[, <ty> <NumElements>][, align <alignment>] ; yields {type*}:result And the documentation

Hi all, Quoting <http://llvm.org/docs/LangRef.html#i_malloc>: The 'malloc' instruction allocates sizeof(<type>)*NumElements bytes of memory from the operating system and returns a pointer of the appropriate type to the program. If "NumElements" is specified, it is the number of elements allocated. Obviously this does not say that the default for NumElements

Hey, So in Fedora rawhide we are now building llvm 3.7.1 into the lots of little shared libraries format. However I'm running into a major problem with the fact that sometimes dlclose isn't dropping all the LLVM libraries from the address space of the process. We have a sequence like this: a) X server asks mesa gbm library to init, it loads the kms_swrast_dri.so with

On Sat, Jun 5, 2010 at 3:18 AM, Duncan Sands <baldrick at free.fr> wrote: > Hi Tom, > > I'm pretty sure that the LLVMContext constructor is defined in libLLVMCore.a > What is the output of `llvm-config --libs bitwriter`?  Does it include > -lLLVMCore? Hi Duncan, e0082888 at e0082888-laptop:~$ local/bin/llvm-config --libs bitwriter -lLLVMBitWriter -lLLVMCore -lLLVMSupport

I'm trying to do a simple build from the git 8.0.0 sources. The sources seem to build OK but a link step fails from running out of memory. I need some clues how to figure out where the bottleneck might be. The cmake command is: cmake -G Ninja                                          \     -DLLVM_TARGETS_TO_BUILD=X86                         \    

Chris Lattner wrote: > On Apr 19, 2008, at 3:27 PM, Jon Sargeant wrote: >>>> Regarding malloc and alloca, I realized that the size is unsigned, >>>> so a >>>> negative value for NumElements is impossible. I suggest replacing >>>> "it >>>> is the number of elements allocated" with "it is the UNSIGNED number

On Apr 19, 2008, at 3:27 PM, Jon Sargeant wrote: >>> Regarding malloc and alloca, I realized that the size is unsigned, >>> so a >>> negative value for NumElements is impossible. I suggest replacing >>> "it >>> is the number of elements allocated" with "it is the UNSIGNED number >>> of >>> elements allocated".

Thanks for your reply. Chris Lattner wrote: >> Regarding free, I also think your wording isn't clear enough: "If the >> pointer is null, the result is undefined." The free result is void. >> Consider rewording as "If the pointer is null, the operation is valid >> but does not free the pointer." > > It isn't though. free(NULL) could

Hi, I am trying to build a binary of Rquantlib package 0.3.5 under Window 7 64bits. I am using mingw gcc version 4.5.1 (tdm64-1) which is a experimental version of 64 bits ming gcc. (http://tdm-gcc.tdragon.net/). I have successfully compiled boost 1.45 and Quantlib 1.0.1 with this gcc. And then when I tried to compile Rquantlib 0.3.5, it reports some error. The first error is like below.

This is OS X 10.6. I can build llvm-gcc and llvm-g++ just fine, but the gfortran front end cannot find a symbol. Where would it normally be defined ? ==================================================================== /Developer/usr/bin/g++-4.2 -g -O2 -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -Wno-variadic-macros

Hi Justin No i did a test setup after getting to it in real life. In this test setup: - rsync (daemonless) is run - ctrl-c, - results are checked, - dest directory is deleted and so on. in 50 % the file is left over. i can reproduce it on a big fat slow (relative to a small ext4 partition on same lvm) xfs-partition. If i do this on root I even if I manage tp ctrl-c at the right time, I

I tried building boost 1.45 release with clang 2.9 following instructions on http://blog.llvm.org/2010/05/clang-builds-boost.html but got the following result: $ clang++ --version clang version 2.9 (trunk 123420) Target: x86_64-apple-darwin10 Thread model: posix $ ./bjam toolset=clang Building the Boost C++ Libraries. Performing configuration checks - has_icu builds : no

While attempting to build llvm-gcc-4.2 from llvm-2.7 release branch, I am finding that the build fails at... c++ -g -O2 -mdynamic-no-pic -DIN_GCC -W -Wall -Wwrite-strings -Wstrict-prototypes -Wmissing-prototypes -pedantic -Wno-long-long -Wno-variadic-macros -Wno-overlength-strings -Wold-style-definition -Wmissing-format-attribute -mdynamic-no-pic -DHAVE_CONFIG_H -o f951 \ fortran/arith.o

Hi, In another post I was trying to find out how to use libc++ instead of libstdc++ when compiling llvm/clang. I couldnt find the a way to tell cmake to do that. So I switched to using configure to compile llvm/clang. But now I find that the artifacts produced are different. Here are the issues I see: - configure doesnt seem to respect '--prefix' option, it just puts

I use the Hitec Aurora 9 (A9) transmitter for my remote control vehicles. http://www.hitecrcd.com/products/aircraft-radios-receivers-and-accesories/aircraft/aurora9/aurora-9-2-4ghz.html This transmitter is part of a system of devices. One of the devices in this system is the HPP-22, which is a USB device that allows a computer to communicate with the A9 in order to load and save model

ParseDecl.cpp fails to compile with MSVC reporting the following error: ParseDecl.cpp compiler\llvm\tools\clang\lib\Parse\ParseDecl.cpp(2760) : error C2248: 'clang::A STOwningResult<Destroyer>::operator =' : cannot access private member declared i n class 'clang::ASTOwningResult<Destroyer>' with [ Destroyer=::up▲ ]

2015-12-01 1:20 GMT+08:00 Chris Bieneman <beanz at apple.com>: > > On Nov 29, 2015, at 9:09 PM, 慕冬亮 <mudongliangabcd at gmail.com> wrote: > > 2015-11-30 12:58 GMT+08:00 Chris Bieneman <beanz at apple.com>: > > The autotools build system is officially deprecated and will be removed in a > future release. > > CMake is the recommended configuration system,