search for: prologueepilogueinserter

...due to the C++ runtime containing lots of leaf-like routines that use thread-local storage. I can imagine a number of hackish workarounds, but I think probably the right way to go is to mark routines with thread-local storage accesses in them as non-leaf. I guess that would have to happen in the PrologueEpilogueInserter. Is there an easy way to tell if a MachineFunction uses TLS without doing a full scan of the body? Perhaps SelectionDAG will have to mark the function somehow. -Dave

...ining lots of leaf-like > routines that use thread-local storage. > > I can imagine a number of hackish workarounds, but I think probably > the right way to go is to mark routines with thread-local storage > accesses in them as non-leaf. I guess that would have to happen in > the PrologueEpilogueInserter. > > Is there an easy way to tell if a MachineFunction uses TLS without doing > a full scan of the body? Perhaps SelectionDAG will have to mark the > function somehow. How does a MachineInstr acquire a CallFrameSetupOpcode? I can't find anywhere in the sources or generated files...

...r to find the caller-saved live regs, and the push / pop these. I find that there are no such regs actually live there. It seems that these registers are somehow COPY:d and not coalesced, and thus not live across the call. Is this correct? How does this work? It looked like there was a method in PrologueEpilogueInserter for these regs before, but it has been removed. Please explain. Thanks, Jonas -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20110827/cd5d7f68/attachment.html>

How/where are function prologs and epilogs generated, is it bespoke C++ code or TableGen generated ? If someone could point me in the right direction please. Many thanks in advance, Aaron -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20090409/fb3336e4/attachment.html>

...gt; routines that use thread-local storage. > > > > I can imagine a number of hackish workarounds, but I think probably > > the right way to go is to mark routines with thread-local storage > > accesses in them as non-leaf. I guess that would have to happen in > > the PrologueEpilogueInserter. > > > > Is there an easy way to tell if a MachineFunction uses TLS without doing > > a full scan of the body? Perhaps SelectionDAG will have to mark the > > function somehow. > > How does a MachineInstr acquire a CallFrameSetupOpcode? I can't find > anywhere...

Hi I'm working on the XCore target and am having difficulty building libgcc. Background: If I use a libgcc built by llvm3.0-gcc with my current clang-llvm3.3 compiler, exceptions 'seem' to work. Trying to rebuild libgcc however breaks exception handling - they aren't caught! I thus assumed I needed to focus on the unwind code and particularly functions that call