similar to: [LLVMdev] Failure handling half type

Here there is the dump: SelectionDAG.cpp:81: static bool llvm::ConstantFPSDNode::isValueValidForType(llvm::EVT, const llvm::APFloat &): Assertion `VT.isFloatingPoint() && "Can only convert between FP types"' failed. 0 libLLVMSupport.so 0x00007f7405022de5 llvm::sys::PrintStackTrace(_IO_FILE*) + 37 1 libLLVMSupport.so 0x00007f74050232e3 2 libpthread.so.0

Make sure you’re running LLVM with assertions enabled. You’ll get more information about what’s going wrong that way. That said, yes, this looks like a bug. Specifically, SelectionDAGLegalize::ExpandConstantFP() assumes that f32 is the smallest floating point constant type it’ll need to handle, and that MVT enum ordering reflects that. while (SVT != MVT::f32) {

Just ran into this as well. Did you ever open a a bug report about it? On Tue, Jun 18, 2013 at 3:43 AM, Alessio Giovanni Baroni < alessiogiovanni.baroni at gmail.com> wrote: > Here there is the dump: > > SelectionDAG.cpp:81: static bool > llvm::ConstantFPSDNode::isValueValidForType(llvm::EVT, const llvm::APFloat > &): Assertion `VT.isFloatingPoint() && "Can

I don't. 2013/6/29 Keno Fischer <kfischer at college.harvard.edu> > Just ran into this as well. Did you ever open a a bug report about it? > > > On Tue, Jun 18, 2013 at 3:43 AM, Alessio Giovanni Baroni < > alessiogiovanni.baroni at gmail.com> wrote: > >> Here there is the dump: >> >> SelectionDAG.cpp:81: static bool >>

Ok, now it's arising another problem. IR code that I obtain is the following: i8* bitcast ([5 x i8] c"hello\00" to i8*) generated from instructions: ConstantExpr::getBitCast (ConstantDataArray::getString (getGlobalContext (), "hello"), PointerType::get (Type::getInt8Ty (getGlobalContext ()), 0)) but the LLC tool says: invalid cast opcode for cast from '[5 x i8]'

On 19/12/12 10:25, Alessio Giovanni Baroni wrote: > Ok, now it's arising another problem. IR code that I obtain is the following: > > i8* bitcast ([5 x i8] c"hello\00" to i8*) You need to put your "hello" in a (constant) global variable. You can then bitcast the address of the global to i8*. I suggest you compile some C code that makes use of the address of a

Because I need to convert an [4 x i8] type to i8* in the instantiation of a GlobalVariable. I have the follow declaration: %xxx = type { i8* } and I must emit the following variable: @yyy = linkonce_odr constant %xxx { [4 x i8] c"hello\00" }, but it's wrong. How do I do? 2012/12/18 Óscar Fuentes <ofv at wanadoo.es> > Alessio Giovanni Baroni <alessiogiovanni.baroni at

Hi to all, I need the class in object, but the file ConstantsContext,h isn't installed. What can I do? Copy that header under include? It's an installation bug? Thanks! -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20121218/e1130d0a/attachment.html>

Hi, I’m on Mac OS X 10.9.2 with all last updates; if I have this source: class xxx explicit { }; int main (int argc, char **argv) { return 0; } and I compile it with clang++ -std=c++11 -Wall -pedantic -c I obtain this error: nn:1: error: expected unqualified-id { ^ 1 error generated. But the C++ 11 standard permits explicit on classes. It’s a bug? -------------- next part

Hello Anton Thu, 3 Aug 2006 15:06:56 +0400 you wrote: > here it is in the attachment :) Ok. Could you also send LibDeps.txt file? It should be in /obj/tools/llvm-config directory -- With best regards, Anton Korobeynikov. Faculty of Mathematics & Mechanics, Saint Petersburg State University.

Hello Anton Fri, 4 Aug 2006 21:45:19 +0400 you wrote: > Written by Mike Haertel and Paul Eggert. > I've updated llvm and llvm-gcc4 ant trying to build it again after > PR845 was resolved. According to Reid's letter this PR coud be the > reason of my problem. Anyway, "sort" call can cause large problems depending, where in your PATH cygwin directory is (before

Alessio Giovanni Baroni <alessiogiovanni.baroni at gmail.com> writes: > I need the class in object, but the file ConstantsContext,h isn't installed. > > What can I do? Copy that header under include? It's an installation bug? That's an internal header, not part of the public API. Why do you need the classes declared on that header?

Alessio Giovanni Baroni <alessiogiovanni.baroni at gmail.com> writes: > Because I need to convert an [4 x i8] type to i8* in the instantiation of a > GlobalVariable. > I have the follow declaration: %xxx = type { i8* } and I must emit the > following variable: > @yyy = linkonce_odr constant %xxx { [4 x i8] c"hello\00" }, but it's wrong. > > How do I do?

> > If you're building llvm-gcc4, you don't need the runtime libraries, so > I'd just stick with the "tools-only" build and declare success. If > you're building llvm-gcc3, I'd suggest you switch to llvm-gcc4 :) I switched to llvm-gcc4 but when I run make from obj folder i run into folowing errors: Can't find a library with no dependencies at

Hello Anton Thu, 3 Aug 2006 12:38:54 +0400 you wrote: > I've updated it yesterday and rebuilt - llvm built fine. But when > building llvm-gcc4 (also updated yesterday from new /trunk > directory) it fails with the same error. You might easily get llvm-gcc4-mingw32 binaries from "prerelease" directory. Since stdcall, fastcall & dllimport stuff is unsupported right now,

Hi, I am currently working on a more or less intrusive patch (D37896), which pulls optimizations on multiplications from some back-ends, e.g., (mul x, 2^N + 1) => (add (shl x, N), x) in AArch64, into the DAGCombiner to have this optimization generic on all targets. However, running the LLVM Tests leads to 67 unexpected results. Am 19.09.2017 um 15:58 schrieb Sanjay Patel: > For the

Hi to all, when I download the clang/llvm officila binaries, I found the libc++ headers under lib directory, but I didn't found libc++.so. Why you ship libc++ headers but no shared library? Thanks! -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20131212/36204500/attachment.html>

There are multiple problems/questions here: 1. Make sure you've updated trunk to the latest rev before running update_llc_test_checks.py on lea-3.ll. Ie, I would only expect the output you're seeing if you're running the script on a version of that test file before r313631. After that commit, each RUN has its own check prefix, so there should be no conflict opportunity. 2. I

Hi Sanjay, thanks for this information. I did get a little bit further with the patch. However, Hexagon gives me headaches. I tried to limit the scope of the patch to the BeforeLegalizeTypes phase and Hexagon still reaches the unreachable. Hexagon tries to split or widen a vector type for a node with custom lowering where the unreachable arises from inside TargetLowering::ReplaceNodeResults

Is VT a legal type on Hexagon? It looks like Hexagon may be setting SHL as Custom for every defined vector type. Try adding TLI.isTypeLegal(VT) too. ~Craig On Thu, Sep 21, 2017 at 10:06 PM, Haidl, Michael < michael.haidl at uni-muenster.de> wrote: > Hi Sanjay, > > thanks for this information. I did get a little bit further with the > patch. However, Hexagon gives me headaches.