similar to: [LLVMdev] Problem migrating from llvm 2.9 to 3.0

Hi Juan, On 07/11/12 13:35, Juan Manuel Martinez Caamaño wrote: > Hello, I'm moving a compiler pass from llvm 2.9 to 3.0, and I'm getting some > strange extra store instructions when optimizing the code. > > For example, in the code that is in the attachments (On block *for.body111* for > 2.9 and block *for.body67 *for 3.0), the number of stores goes from 1 to 10. And >

Hello, We have a tool which reads in bitcode, processes it, and re-emits it. We use this tool as a flexible way to integrate our tool into the Xcode, Android NDK, Chromium, and Linux build process. The problem we face is that bitcode changes, and when it does… future versions can read it, but past versions are left in the lurch. For instance LLVM 3.2svn can BitcodeReader from LLVM 3.1, but

Hello My name is Guy Quantrell. I work for an organisation called WA Consultants. We are a recruitment agency based in the UK, specialising in supplying to mobile handset providers. I am currently recruiting for a major handset manufacturer based near London who is looking for an LLVM developer to lead its open source activities in this area. I am very keen to have a chat with anybody wishing to

Hello, I apologize if this has already been fixed or reported. I believe there is a bug in the way the optimizer deals with thread_local variables. The attached program, test.c, has a thread-local variable "int Foo" and a global variable "int *Ptr". The program takes the following steps: 1) The main thread spawns a new thread and waits 2) The new thread writes Foo = 50 and

LLVM Weekly - #216, Feb 19th 2018 ================================= If you prefer, you can read a HTML version of this email at <http://llvmweekly.org/issue/216>. Welcome to the two hundred and sixteenth issue of LLVM Weekly, a weekly newsletter (published every Monday) covering developments in LLVM, Clang, and related projects. LLVM Weekly is brought to you by [Alex

Hi, I've been writing a 80's era-flavor of QuickBASIC compiler based on LLVM, and it's come surprisingly far very quickly, LLVM is working great. GOSUB's, GOTO's, FIELD statements. I'm using Visual Studio 2010. Here's my setup: I use LLVM to turn my AST into code, which, during debugging I then execute natively using "getPointerToFunction" and the runtime

I'm starting to work with custom types and wondering how I can pass/return them via the ExecutionEngine? runFunction returns only a GenericValue, so it appears somehow I must make use of its pointer member. For example, how do call this function/type: %struct.mine = type { i32, i8, float, double } define void @make_struct(%struct.mine* noalias sret %agg.result, i32 %a, i8 signext %b) --

I can't find a lot of information about using structures directly as parameters, returns, and in registers. Is this fully supported on all platforms? Does it always convert to creating a hidden parameter when too large? For example (assume very.large is too big to fit in the target machine registers): define %very.large @get_struct() { %m1 = insertvalue %very.large undef, i32 10, 0 ...

We have been using LLVM 3.1 to support JITing of AVX. From dumping the MC generating by the MCJIT I noticed it always emits 'VZEROUPPER' to clear the high 128 bit before calling another function. In some cases I know the function called either only use AVX or does not use SSE. I will like to inform the backend it is safe not to emit that instruction. Have not been able to figure out how

I have a const array of structs (struct is made up of two int). This code: ArrayRef<Constant*> AR; CA->get(AT, AR); where CA = ConstantArray* and AT* = ArrayType gives AR.size() of 0 I just want to iterate through the ConstantArray and find the values of the integers inside the structs. Thanks. -------------- next part -------------- An HTML attachment was scrubbed... URL:

On 6 November 2012 14:53, Duncan Sands <baldrick at free.fr> wrote: > Hi Anitha, > > > On 05/11/12 10:29, Anitha Boyapati wrote: > >> >> >> On 5 November 2012 14:32, Duncan Sands <baldrick at free.fr >> <mailto:baldrick at free.fr>> wrote: >> >> Hi Anitha, >> >> >>

On Wed, Nov 7, 2012 at 12:39 AM, Renato Golin <rengolin at systemcall.org>wrote: > On 6 November 2012 22:28, Daniel Dunbar <daniel at zuster.org> wrote: > > Is it possible instead to refactor the tests so that each binary > corresponds > > to one test? For example, look at how Hal went about integrating TSVC: > > It should be possible. I'll have to

On 7 November 2012 15:29, Duncan Sands <baldrick at free.fr> wrote: > > That way the output should be exactly the same as the output dragonegg > would > normally run the LLVM optimizers on, e.g. GCC constant folding and other > such > optimizations which get turned on at -O2 will still have happened > (dragonegg > turns off almost all GCC optimizations by default,

I just installed a TE820 octal span T1 card, and it's not showing up in dahdi_hardware output. This was installed into a test machine that already has a TDM800P card in it, and that one is showing up and working fine. Is there some kernel module that I'm missing? Lspci: 05:04.0 Ethernet controller: Digium, Inc. Wildcard TDM800P 8-port analog card (rev 11) 21:08.0 Communication

Dear list, we would really like to be able to "invite a third and fourth party" to our current one-on-one call. At the moment, we have to agree to dial into MeetMe 10 minutes later, then make calls to the third parties, and hope it all works out. I have found a couple of examples on the Internet for converting channels into conferences, but I could not get any of them working. Does

Hi Susan, Sorry for the delayed response. Thanks for the test cases - I'm looking in to this now. - Lang. On Mon, Nov 5, 2012 at 2:58 PM, Susan Horwitz <horwitz at cs.wisc.edu> wrote: > Hi Lang, > > I looked more into one of the problems I'm now having, and I've attached 3 > files: > > Gcra.cpp is like your version except that for two specific vregs it uses

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 ATI 6.8 pre-release This is the long awaited radeon driver with randr 1.2 support. radeon 6.7 was pretty much a dead end, so it's been branched for those that still want to play with it. In addition to randr support there are quite a few other goodies: zero copy tfp, improved connector table parsing, TV-out support, and much more. Changes since

Hi Anitha, > http://llvm.org/bugs/show_bug.____cgi?id=14185 > <http://llvm.org/bugs/show_bug.__cgi?id=14185> > > <http://llvm.org/bugs/show___bug.cgi?id=14185 > <http://llvm.org/bugs/show_bug.cgi?id=14185>> > I am stuck on analysis. Does any one have alternate suggestions > on debugging >

On 6 November 2012 22:34, Daniel Dunbar <daniel at zuster.org> wrote: > You are right, the failure on compile_time indicates that the test isn't > even building. As provided, the tests don't actually define the cpuida() or > calculateMHz() functions so that seems expected to me. I defined both functions as NOPs. I got what it was. The original makefile had a "-o

On Wed, Nov 7, 2012 at 8:28 AM, John Slagel <john.slagel at gmail.com> wrote: > Hi, > > I've been writing a 80's era-flavor of QuickBASIC compiler based on LLVM, > and it's come surprisingly far very quickly, LLVM is working great. GOSUB's, > GOTO's, FIELD statements. > I'm using Visual Studio 2010. > Here's my setup: I use LLVM to turn my AST