similar to: Vectorizing structure reads, writes, etc on X86-64 AVX

Thank you for your reply. FWIW, I wrote the .ll by hand after taking the C program, using clang to emit the llvm and seeing the memcpy. The memcpy version that clang generates gets compiled into assembly that uses the large sequence of movs and does not use the vector hardware at all. When I started debugging, I took that clang produced .ll and started to write it different ways trying to get

----- Original Message ----- > From: "Sanjay Patel via llvm-dev" <llvm-dev at lists.llvm.org> > To: "Jay McCarthy" <jay.mccarthy at gmail.com> > Cc: "llvm-dev" <llvm-dev at lists.llvm.org> > Sent: Tuesday, November 3, 2015 12:30:51 PM > Subject: Re: [llvm-dev] Vectorizing structure reads, writes, etc on X86-64 AVX > > If the

Hi Jay - I see the slow, small accesses using an older clang [Apple LLVM version 7.0.0 (clang-700.1.76)], but this looks fixed on trunk. I made a change that comes into play if you don't specify a particular CPU: http://llvm.org/viewvc/llvm-project?view=revision&revision=245950 $ ./clang -O1 -mavx copy.c -S -o - ... movslq %edi, %rax movq _spr_dynamic at GOTPCREL(%rip),

Hi, I have a vector of strings (class character) with 6 elements (length 6). I call it 'names'. "Graham Chapman" "John Cleese" "Terry Gilliam" "Eric Idle" "Terry Jones" "Michael Palin" And I want to turn it into another vector of strings called 'shortnames' with the same length. The new vector should look like:

I've tried to wrap my head around SELinux several times, and so far, all I've managed to do is rap my head, and then somewhat painfully. Is there a detailed "SELinux Unleashed" style book anybody can recommend? Anybody read Oreilly's book care to comment? Link:

I am utterly stumped. I need libcurl.so.3 for CentOS 6.4. I have Googled everything and I have been going all over the Internet looking for a solution. I give up; I need help. The error message I get when I try to run a program is: error while loading shared libraries: libcurl.so.3: cannot open shared object file: No such file or directory I have: Package libcurl-7.19.7-36.el6_4.i686 already

Well, I'm a recent convert from WBEL. My biggest concern with CentOS is that the community here seems to want to be more than a recompile of RHEL. But WBEL is floundering, what with Katrina and Rita, and there really being only 1 developer behind it, etc. I offer an automated shell script to switch from WBEL4 to CentOS4 (easy, it's hosted on my home DSL line!) It assumes that

Hi list, I tried to convert a data.frame into a matrix using data.matrix. Unfortunately my matrix contains missing values (NA) wherefore all columns including NA's were changed into factors. I thought that many people stumbled across that problem already wherefore there must be a simple solution. But it seems there isn't. I tried lapply, replace and other things in combination but still

Hi there, I have a cablemodem, ARRIS brand, model tm502G. It has two FXS ports. I was wondering if anyone has details about the correct signalling of these FXS ports when connected to original X100p. Tests: fxsks on the zapata.conf and zaptel.conf files. From my cellphone I call the ARRIS, it starts ringing but the zap channel sees no call coming in. fxsls on the zapata.conf and zaptel.conf

Tonight, I tried to roll out fuse on my CentOS 4 production system. (in order to use GlusterFS) I have two identical servers, and one took, the other didn't. How simple could this be? # yum install yum-plugin-priorities # yum install rpmforge-release # yum install fuse dkms-fuse both of these seem to work. Yet I run [root at kepler drivers]# modprobe fuse FATAL: Module fuse not

Is there anyway to achieve inheritance of specifications? Suppose I had an action like def index filter = params[:filter] ? params[:filter] : ''%'' order = params[:order] ? params[:order] : ''created_at DESC'' @posts = Post.find(:conditions => "title LIKE #{filter}", :order => order) ... end Now I might have half a dozen specs for this

I'll explain what we see in the code. 1. The caller saves XMM registers across the call if needed (according to DEFS definition). YMMs are not in the set, so caller does not take care. 2. The callee preserves XMMs but works with YMMs and clobbering them. 3. So after the call, the upper part of YMM is gone. - Elena -----Original Message----- From: Bruno Cardoso Lopes [mailto:bruno.cardoso at

Hi, What is the calling conventions for YMM. According to documents I saw till now, the YMMs are scratch and not saved in callee. This is also the default behavior of the Intel Compiler. In X86InstrControl.td the YMMs are not in "defs" set of call. - Elena --------------------------------------------------------------------- Intel Israel (74) Limited This e-mail and any attachments

Hi, > What is the calling conventions for YMM. According to documents I saw till now, the YMMs are scratch and not saved in callee. > This is also the default behavior of the Intel Compiler. x86_64 Non-windows targets use the rules defined in the x86_64 abi! > In X86InstrControl.td the YMMs are not in "defs" set of call. The XMMs are subregisters of YMMs, and they are in the

On Jan 8, 2012, at 11:18 PM, Demikhovsky, Elena wrote: > I'll explain what we see in the code. > 1. The caller saves XMM registers across the call if needed (according to DEFS definition). > YMMs are not in the set, so caller does not take care. This is not how the register allocator works. It saves the registers holding values, it doesn't care which alias is clobbered. Are you

On Jan 9, 2012, at 10:00 AM, Jakob Stoklund Olesen wrote: > > On Jan 8, 2012, at 11:18 PM, Demikhovsky, Elena wrote: > >> I'll explain what we see in the code. >> 1. The caller saves XMM registers across the call if needed (according to DEFS definition). >> YMMs are not in the set, so caller does not take care. > > This is not how the register allocator

This is the wrong code: declare <16 x float> @foo(<16 x float>) define <16 x float> @test(<16 x float> %x, <16 x float> %y) nounwind { entry: %x1 = fadd <16 x float> %x, %y %call = call <16 x float> @foo(<16 x float> %x1) nounwind %y1 = fsub <16 x float> %call, %y ret <16 x float> %y1 } ./llc -mattr=+avx

Clone the existing movq emulation to also support the most fundamental SIMD FP moves. Extend the testing code to also exercise these instructions. Signed-off-by: Jan Beulich <jbeulich@suse.com> --- a/tools/tests/x86_emulator/test_x86_emulator.c +++ b/tools/tests/x86_emulator/test_x86_emulator.c @@ -629,6 +629,60 @@ int main(int argc, char **argv) else

Hello, LLVM generates two additional instructions for 128->256 bit typecasts (e.g. _mm256_castsi128_si256()) to clear out the upper 128 bits of YMM register corresponding to source XMM register. vxorps xmm2,xmm2,xmm2 vinsertf128 ymm0,ymm2,xmm0,0x0 Most of the industry-standard C/C++ compilers (GCC, Intel's compiler, Visual Studio compiler) don't generate any extra moves

I've been working on adding AVX to LLVM and have run across a number of questions. Here's the first one. In some ways AVX is "just another" SSE level. Having AVX implies you have SSE1-SSE4.2. However AVX is very different from SSE and there are a number of sub-features which may or may not be available on various implementations. So right now I've done this: def