search for: issafetoconvert

...* rather than i8*). Yep, I did this to beat out all of the assumptions in the clang IRGen that was assuming that the type return by a field access load was sane (they'd blow up when they got an unexpected {}* instead of a function pointer or whatever). Once everything worked, I added the 'isSafeToConvert' logic in clang/lib/CodeGen/CodeGenTypes.cpp. The idea is that when converting the first "g" example, there is no way we can do better than {}*. OTOH, in the second "h" example there is no problem recursively converting X to a nice type, and isSafeToConvert returns true for...

Hi Chris, >> Chris made major changes to the LLVM type system, which requires major changes >> to dragonegg (c.f. the patch Chris applied to llvm-gcc). I'm on holiday which >> is why I haven't taken care of it yet. > > Yeah, sorry about that Duncan. I can't directly hack on the code, but I'm happy to answer questions when you get back. there seems to be

On Jul 18, 2011, at 1:32 PM, Duncan Sands wrote: > On 18/07/11 15:59, Jack Howarth wrote: >> Despite the commit of... > > Chris made major changes to the LLVM type system, which requires major changes > to dragonegg (c.f. the patch Chris applied to llvm-gcc). I'm on holiday which > is why I haven't taken care of it yet. Yeah, sorry about that Duncan. I