Hello, Chris.> It depends on what sort of attribute it is. If it is something specific > to a particular function, adding it to the Function object makes sense > (e.g. see alignment and section info). If it affects the calling > convention, adding it to FunctionType makes sense.I'm wanting some "general" attribute interface which can be applied to concrete Function object. It seems, that just a set of string-string pairs should be enough. Surely, this interface can be used for some specific stuff at first time (e.g. some attributes like visibility nowadays), which can be further moved to its own implementation instead of generic one. -- With best regards, Anton Korobeynikov. Faculty of Mathematics & Mechanics, Saint Petersburg State University.
On Wed, 10 Jan 2007, Anton Korobeynikov wrote:>> It depends on what sort of attribute it is. If it is something specific >> to a particular function, adding it to the Function object makes sense >> (e.g. see alignment and section info). If it affects the calling >> convention, adding it to FunctionType makes sense. > I'm wanting some "general" attribute interface which can be applied to > concrete Function object. It seems, that just a set of string-string > pairs should be enough. > > Surely, this interface can be used for some specific stuff at first time > (e.g. some attributes like visibility nowadays), which can be further > moved to its own implementation instead of generic one.I am currently opposed to a general purpose way to stick random data on functions. With strings, you don't know what to do when linking (what combinations are invalid?), you can't verify them, and the code generator doesn't know how to handle them. -Chris -- http://nondot.org/sabre/ http://llvm.org/
Seemingly Similar Threads
- [LLVMdev] Compile error of latest Dragonegg on Ubuntu with GCC 4.5
- [LLVMdev] Patch: MSIL backend global pointers initialization
- [LLVMdev] General LLVM question
- [LLVMdev] trying to generate a simple inline asm
- [LLVMdev] clang: call extern function using JIT