similar to: [LLVMdev] [DragonEgg] Strange call to @"\01__isoc99_fscanf"

On Tue, Oct 09, 2018 at 02:28:10PM +0300, Nir Soffer wrote: > > +# Create a background thread running a web server which is > > +# simulating the imageio server. > > > > This functionality should be separated from the fake SDK module, since it is > not part of the SDK, and may be replaced by real imageio server later. Well possibly, but it was very convenient to put it

On Mon, Oct 15, 2018 at 05:13:31PM +0100, Richard W.M. Jones wrote: > On Mon, Oct 15, 2018 at 06:43:10PM +0300, Nir Soffer wrote: > > On Mon, Oct 15, 2018 at 6:21 PM Richard W.M. Jones <rjones@redhat.com> > > wrote: > > > > > On Tue, Oct 09, 2018 at 02:28:10PM +0300, Nir Soffer wrote: > > > > > +# Create a background thread running a web server

On Mon, Oct 15, 2018 at 6:21 PM Richard W.M. Jones <rjones@redhat.com> wrote: > On Tue, Oct 09, 2018 at 02:28:10PM +0300, Nir Soffer wrote: > > > +# Create a background thread running a web server which is > > > +# simulating the imageio server. > > > > > > > This functionality should be separated from the fake SDK module, since > it is > >

On Thu, Sep 20, 2018 at 11:51 AM Richard W.M. Jones <rjones@redhat.com> wrote: > Previously this output method was almost completely untested. > > This commit adds a fake ovirtsdk4 module so we can test the > -o rhv-upload method fairly completely without needing an actual > oVirt instance around. > --- > v2v/Makefile.am | 4 + >

> While testing a File I/O sample program, I've encountered a link failure due > to missing implementation of "\01__isoc99_fscanf" function. I think this > function should be named "__isoc99_fscanf" instead. Please see the program > code and LLVM IR generated by DragonEgg and clang below. It shows that clang > generates "__isoc99_fscanf", while

Dear all, In our compiler we use a modified version LLVM Polly, which is very sensitive to proper code generation. Among the number of limitations, the loop region (enclosed by phi node on induction variable and branch) is required to be free of additional memory-dependent branches. In other words, there must be no conditional "br" instructions below phi nodes. The problem we are facing

I have a bitcode file Y that has a symbol __isoc99_fscanf in it. When I disassemble the bitcode, the symbol is written as "\01__isoc99_fscanf". I am wondering where this symbol is coming from? I grep'ed throughout /usr on my machine for both names, and I only found __isoc99_fscanf, without the special character in front. Regardless of where the symbol is coming from, I am

Hi, Here's another case, different in high-level, but similar in low-level. When Fortran allocatable array is defined in module, its actual dimensions are kept in internal structure. Loads originated from reading these dimensions confuse Polly on any use of this array. Attachments: 1) Sample Fortran source code (to be compiled with and without -DMODULE to see failing and working version,

This patch introduces a the new "linker_private_weak" linkage type. Why a new linkage type? The idea behind the "linker" linkage types is that they are used by the linker and then discarded. I.e., the symbols won't show up in the final linked image. In that regard, all "linker" linkage types are inherently "private". The "linker" linkages are

On Jun 30, 2010, at 1:36 PM, Bill Wendling wrote: > This patch introduces a the new "linker_private_weak" linkage type. > > Why a new linkage type? The idea behind the "linker" linkage types is that they are used by the linker and then discarded. I.e., the symbols won't show up in the final linked image. In that regard, all "linker" linkage types are

---------- Forwarded message ---------- From: charles quarra <charllsnotieneningunputocorreo at gmail.com> Date: 2012/12/13 Subject: error while linking modules with exception handling demo code To: llvmdev at cs.uiuc.edu Hi, I am building a module X with an arithmetic function foo, a module Y with an arithmetic function foo2 that invokes foo. For the invocation be a proper one (being

Hi, I'm trying to understand better how linker_private (and friends) work. From what I can understand from the manual (http://llvm.org/docs/LangRef.html#linkage_linker_private), it seems that a linker_private function cannot be removed by the optimizers. Only the linker can discard the symbol. However, I tested with 'opt -O1' and a function with *no users* is removed. Same

On Fri, Dec 19, 2014 at 12:56 PM, Duncan P. N. Exon Smith < dexonsmith at apple.com> wrote: > > However, I think this would set a bad precedent. There's nowhere else > (that I know of) where we accept two versions of assembly. The > LLParser is relatively easy to work with because it doesn't have that > kind of historical baggage. I can think of two precedents:

Hi useRs, I would like to know what R users are employing to get their images/plots created in R, into Excel sheets. I am aware of the various packages (xlsReadWrite, RODBC) to get data frames into R, but I would like to copy images too. Thank you for any help in solving this problem. Regards, Harsh Singhal

On Jun 30, 2010, at 9:52 PM, Chris Lattner wrote: > On Jun 30, 2010, at 1:36 PM, Bill Wendling wrote: > >> I implemented the new linkage to have its own prefix from linker_private. As it currently stands, the symbol's prefix ("l") is the same for "linker_private" and "linker_private_weak". If this will always be the case, then I can remove the code

I am trying to create a NetCDF file with bathymetry (a matrix) and then to add a grid type (an integer) as variables. This is the relevant part of the code: library(ncdf) wfile="my_file.nc" msv=-10 grdtp=2 # Dimensions d1=dim.def.ncdf("lon","degrees",as.double(lon)) d2=dim.def.ncdf("lat","degrees",as.double(lat)) # Variables

I tried to manually assign each of 3 arrays a unique TBAA node. But it does not seem to help: alias analysis still considers arrays as may-alias, which most likely prevents the desired optimization. Below is the sample code with TBAA metadata inserted. Could you please suggest what might be wrong with it? Many thanks, - D. marcusmae at M17xR4:~/forge/llvm$ opt -time-passes -enable-tbaa -tbaa

what is the best way to handle package conflicts between classes? for example, below, i have the following two classes defined and both include a "package { "perl-LDAP" }", so when used separately they work as expected but when i need to assign both classes to a node i get a conflict. error: "Duplicate definition: Package[perl-LDAP] is already defined" thanks

Dear NVPTX community, I've create a bug http://llvm.org/bugs/show_bug.cgi?id=13521 with reprocase for this issue. Please, help us to fix it. Last 1,5 months we regularly encounter & workaround or fix 1-2 bugs per week in NVPTX backend. This is definitely not the amount of work we can completely serve ourselves... We would really really appreciate some collaboration. Thanks, - D.

Hi, After building out project in release mode, caught an assertion, which we have not seen before: hello_f: /tmp/rpmbuild_debug/BUILD/llvm/build/include/llvm/ADT/DenseMap.h:126: void llvm::DenseMap<KeyT, ValueT, KeyInfoT>::clear() [with KeyT = llvm::MachineBasicBlock*, ValueT = <unnamed>::BlockChain*, KeyInfoT = llvm::DenseMapInfo<llvm::MachineBasicBlock*>]: Assertion