similar to: [LLVMdev] SDNode enum

Hi Misha,Chris It will be really great if you update this information on the mentioned document. Thanks, Ankur

Hi Ahmad, Yes, merging works good. However, my problem is like this - I have a C library which consists of 1000's of functions spread through various files. The functions do not have dependency amoung each other. I want to link only relavant files( files which have functions called from my application). Since ar has a global symbol table, I believe it should be faster to look for a symol in

Hi Rail, Below 2 revisions might be of your interest which Detect SAD patterns and emit psadbw instructions on X86.: http://reviews.llvm.org/D14840 http://reviews.llvm.org/D14897 Intrinsics related to absdiff revisons : http://reviews.llvm.org/D10867 http://reviews.llvm.org/D11678 Hope this helps. Regards, Suyog On Sat, May 28, 2016 at 4:20 AM, Rail Shafigulin via llvm-dev < llvm-dev at

I'm having a great deal of trouble figuring out how to write instruction patterns which actually match the DAG produced by the compiler. I can't seem to find any documentation on both what the various nodes represent or on what the syntax accepted by TableGen is. The backends I have access to all seem to do this in different (and obscure) ways. And when things go wrong the compiler seems

Hi Ankur, Why do you need archive in this case? The other way of doing this is to merge all bitcode files into single file: $ clang -c -emit-llvm abc.c -o abc.bc $ clang -c -emit-llvm bcd.c -o bcd.bc llvm-link bcd.bc abc.bc -o merged.bc Cheers, Ahmad From: llvmdev-bounces at cs.uiuc.edu [mailto:llvmdev-bounces at cs.uiuc.edu] On Behalf Of ankur deshwal Sent: 21 February 2013 17:54 To:

Hi Guy, Thanks a lot for the clarification. I tried using the triple for SPIR as $ clang -x cl -fno-builtin -emit-llvm -c -Xclang -triple -Xclang spir-unknown-unknown Simple_Kernel.cl However I get the following error. error: unknown target triple 'spir-unknown-unknown', please use -triple or -arch I also tried with triple nvptx-unknown-unknown clang -x cl -fno-builtin -emit-llvm -S

Hi Ankur, For a MachineBasicBlock you can call getBasicBlock() to find the IR level BasicBlock from which it was generated (if there isn't a one, it will return NULL). I don't know of a way to get a finer granularity than that, though. -Jim On Jul 1, 2011, at 1:06 AM, ankur deshwal wrote: > Hi, > > Is there any way to get a pointer to the LLVM Instr from which a >

On Wed, Jun 6, 2012 at 3:09 AM, Joey Gouly <joel.gouly at gmail.com> wrote: > There are plans to upload, when Chandler has the time. The YouTube account > is LLVMProject. > Should get them up this week. Sorry I didn't even remember to do it after they were posted! > Joey > On Jun 6, 2012 8:45 AM, "ankur deshwal" <a.s.deshwal at gmail.com> wrote: >

Hello everyone, Thanks a lot for the help. -target nvptx worked great with llvm-3.2. However it gives error with -target spir. LLVM/Clang trunk although works good for both options ( probably because spir is still work in progress ). Thanks again. - Ankur On Thu, Jan 24, 2013 at 12:43 PM, Benyei, Guy <guy.benyei at intel.com> wrote: > Hi Jordan,**** > > You’re right, and

Hi, I tried to build an llvm archive and link it against an llvm bc file. However, it fails. Following is the procedure I followed ( abc.c is file which calls a function whose definition is present in bcd.c) $ clang -c -emit-llvm abc.c $ clang -c -emit-llvm bcd.c $ llvm-ar cr bsd.ar bcd.o $ llvm-link abc.o bsd.ar llvm-link: bsd.ar:1:2: error: expected integer !<arch> ^ llvm-link:

There are plans to upload, when Chandler has the time. The YouTube account is LLVMProject. Joey On Jun 6, 2012 8:45 AM, "ankur deshwal" <a.s.deshwal at gmail.com> wrote: > Hi, > > I am unable to download videos from the site > > http://llvm.org/devmtg/2012-04-12/. > > The response I always get is > > Proxy request sent, awaiting response... 503 Service

Suyog, Thanks for the reply. Do you know if it is possible to add a new intrinsic without actually modifying core code (ISDOpcodes.h is an example of core code)? I'd like to add this intrinsic with as little code change as possible. On Fri, May 27, 2016 at 8:59 PM, suyog sarda <sardask01 at gmail.com> wrote: > Hi Rail, > > Below 2 revisions might be of your interest which

Hi, I am unable to download videos from the site http://llvm.org/devmtg/2012-04-12/. The response I always get is Proxy request sent, awaiting response... 503 Service Unavailable 2012-06-06 12:00:14 ERROR 503: Service Unavailable. If possible, please share the videos on other medium ( youtube etc.) too. This will spare LLVM web server from heavy traffic due to video downloads too. Thanks and

Hi, Is there any way to get a pointer to the LLVM Instr from which a particular MachineInstr is generated ? Although I understand it goes through many optimization after a LLVM instruction is converted to SDAG node, so maintaining such a mapping will be hard. Still is the mapping stored some where which can give some sane results at least in case of -O0? Thanks and regards, Ankur

The videos are OK now. Except the video of "Autovectorization with LLVM" that have a little mistake with "_" and "-" in the filenames. Thank you. 2012/6/6 Chandler Carruth <chandlerc at google.com>: > On Wed, Jun 6, 2012 at 3:09 AM, Joey Gouly <joel.gouly at gmail.com> wrote: >> >> There are plans to upload, when Chandler has the time.

Hi, I was trying to use the following code in a if condition I->getOperand(0)->hasName() where I is an instruction with following condition already true - I->getOpcode() == Instruction::Call When I compiled a C code with llvm-2.9, I see a function call with no operand name (for operand 0) . This happens for a function with varArgs. The interesting part is that the same C code when

Hi Ankur, Since you use -Xclang, the clang executable passes multiple triples to "clang -cc1". You can see that if you add the -v option. I'm sure there is someone here who can explain it better than I... Anyhow, I think you better use clang -cc1. Make sure -cc1 is the first command line option you use. $ clang -cc1 -fno-builtin -emit-llvm-bc -triple spir-unknown-unknown

On Fri, Jul 1, 2011 at 12:35 AM, ankur deshwal <a.s.deshwal at gmail.com> wrote: > Hi, > > I was trying to use the following code in a if condition > I->getOperand(0)->hasName() > > where I is an instruction with following condition already true - > I->getOpcode() == Instruction::Call > > When I compiled a C code with llvm-2.9, I see a function call with

Hi Jordan, You're right, and the driver -target option works perfectly with the SPIR triples. Thanks Guy Benyei [email_signature_guy_new2] From: Jordan Rose [mailto:jordan_rose at apple.com] Sent: Wednesday, January 23, 2013 19:47 To: Benyei, Guy Cc: ankur deshwal; cfe-dev at cs.uiuc.edu; llvmdev at cs.uiuc.edu Subject: Re: [cfe-dev] OpenCL SPIR/NVPTX code generation On Jan 23, 2013, at

I guess I miss-interpreted the text in [2]. It talks about the optimizations are hindered if the compiler driver invokes link time optimizer "separately". I found that all files compiled are in bc format. Also in library archives, the embedded files were in bitcode format. So the gold linker and LLVMgold plugin are indeed working fine. However, there is still question over only 6% gain