search for: lipengcheng

...t already. Thanks, - Pengcheng On Nov 5, 2013, at 1:57 PM, Eric Christopher <echristo at gmail.com> wrote: > This is from work that Manman did. I thought she'd fixed this particular issue though. Can you file a bug with a test case please? > > On Nov 5, 2013 10:10 AM, "lipengcheng" <pli at cs.rochester.edu> wrote: > Hello LLVM, > > I encountered the following bugs when I compiled my program in debug version. > Could someone help me with that ? Really appreciate that. Thanks, > > Take the mcf program in SPEC2K6 as example. The compiling options...

Hi, What version of clang are you using? I think r174785 is the fix. Manman On Tue, Nov 5, 2013 at 11:04 AM, lipengcheng <pli at cs.rochester.edu> wrote: > What do you mean by test case please ? Using the following commands to > compile mcf program > can reproduce such issue all the time. > > Probably it is my version's problem. Because I checked out 3.3version from > SVN before published...

Also r174479. Manman On Tue, Nov 5, 2013 at 11:11 AM, Manman Ren <manman.ren at gmail.com> wrote: > Hi, > > What version of clang are you using? I think r174785 is the fix. > > Manman > > > On Tue, Nov 5, 2013 at 11:04 AM, lipengcheng <pli at cs.rochester.edu> wrote: > >> What do you mean by test case please ? Using the following commands to >> compile mcf program >> can reproduce such issue all the time. >> >> Probably it is my version's problem. Because I checked out 3.3version >&g...

Another finding is when I use both -code-model=large and -relocation-model=pic (!!!!!note changed), it still generate the static relocation. So does it mean -relocation-model=pic does not like to work with -code-model=large ? Thanks, -Pengcheng On Nov 6, 2013, at 1:52 PM, lipengcheng <pli at cs.rochester.edu> wrote: > Yep. I tried it. It used a movabsq instruction instead and a static allocation. > Thanks a lot. > > I will read it immediately. Again thanks. > > - Pengcheng > > On Nov 6, 2013, at 1:45 PM, Tim Northover <t.p.northover at gm...

This is from work that Manman did. I thought she'd fixed this particular issue though. Can you file a bug with a test case please? On Nov 5, 2013 10:10 AM, "lipengcheng" <pli at cs.rochester.edu> wrote: > Hello LLVM, > > I encountered the following bugs when I compiled my program in debug > version. > Could someone help me with that ? Really appreciate that. Thanks, > > Take the mcf program in SPEC2K6 as example. The compiling opti...

Hello LLVM, I encountered the following bugs when I compiled my program in debug version. Could someone help me with that ? Really appreciate that. Thanks, Take the mcf program in SPEC2K6 as example. The compiling options are : clang -O0 -c -emit-llvm -o obj/mcf.o -g -DSPEC_CPU mcf.c llvm-link -o mcf.bc obj/mcf.o opt -o mcf.opt.bc mcf.bc -O2 llc -O0 -relocation-model=static -o mcf.s

> If you also give "-code-model=large" then LLVM won't assume that > "addr" is nearby so it'll use a movabsq with a static relocation. Actually, it seems our own Eli has written rather a good description of these code models: http://eli.thegreenplace.net/2012/01/03/understanding-the-x64-code-models/. Well worth the read (the default models he talks about at the

Yep. I tried it. It used a movabsq instruction instead and a static allocation. Thanks a lot. I will read it immediately. Again thanks. - Pengcheng On Nov 6, 2013, at 1:45 PM, Tim Northover <t.p.northover at gmail.com> wrote: >> If you also give "-code-model=large" then LLVM won't assume that >> "addr" is nearby so it'll use a movabsq with a

Hi LLVMDEV, Thanks for your suggestions ahead of time. In the terminator instruction of each basic block, it often appears an br instruction, eg. br i8 cond, label %for.entry, label %for.loop. When this code is translated into assembly codes, it will be a jump instruction with a relative address, like jmp 01. My question is can I do some modification to let the jump instruction be with the

Hi, All, Thanks for your answer ahead of time. I plug in an new pass into LLVM to insert one line of code to each BB for printing current BB number. I insert this pass in the Instrumentation phase. But after code generation, I found the total number of BB is different with the number shown in Assembly code. For example, when I traverse all BBs during the instrumentation phase, there are 30000.

Hi, All, Are there any supports about array region analysis both for serial program and multi-thread programs in LLVM ? I couldn't find them. Thanks, Pengcheng

Hi LLVM, When I compile code using llc with options " -relocation-model=static", the generated code still has relative addressing using *rip* frame pointer. May I know how to get rid of frame pointer and totally using absolute addressing ? E.g. callq *addr(%rip). Here *addr* is a global variable, and I want the global address shows up here, instead of relative addressing.

On Oct 22, 2013, at 1:29 PM, llvmdev-request at cs.uiuc.edu wrote: > Send LLVMdev mailing list submissions to > llvmdev at cs.uiuc.edu > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.cs.uiuc.edu/mailman/listinfo/llvmdev > or, via email, send a message with subject or body 'help' to > llvmdev-request at cs.uiuc.edu > > You can