search for: d3240

...a ret which uses the bitcast. Importantly, LLVM must perform TCO regardless of the computation necessary to compute the arguments for the callee, which is not the case today. I sent a patch to llvm-commits, but I'd like to hear high-level feedback on llvmdev: http://llvm-reviews.chandlerc.com/D3240 Thanks! -------------- next part -------------- An HTML attachment was scrubbed... URL: <http://lists.llvm.org/pipermail/llvm-dev/attachments/20140331/71db259e/attachment.html>

...easonable to require the code generator to emit the body of a function containing only a 'musttail' function call as a single branch instruction. Then you could emit such functions in the jump table section. Reid, what do you think? Thanks, -- Peter [1] http://llvm-reviews.chandlerc.com/D3240

...sttail' > > function call as a single branch instruction. Then you could emit such > > functions in the jump table section. > > > > Reid, what do you think? > > > > Thanks, > > -- > > Peter > > > > [1] http://llvm-reviews.chandlerc.com/D3240 > > -- Peter

On Fri, Mar 21, 2014 at 12:54:07PM -0700, Tom Roeder wrote: > On Fri, Mar 21, 2014 at 12:15 PM, Peter Collingbourne <peter at pcc.me.uk> wrote: > >> The way I've implemented it (see the patch I sent to llvm-commits > >> yesterday), it's not just metadata: the intrinsic lowers to the > >> jumptable entry code given above. The CFI pass then generates a