Re: testing that a Gimple call argument is a string...

[David Malcolm <dmalcolm at redhat dot com>]

On Thu, 2012-10-18 at 22:33 +0200, Basile Starynkevitch wrote:
> Hello
> 
> I'm coding in MELT the ex06/ of https://github.com/bstarynk/melt-examples/
> which should typecheck calls to variadic functions json_pack & json_unpack
> from http://www.digip.org/jansson (a JSON library in C).
> 
> I'm working on a MELT pass on Gimple/SSA after phiopt (maybe that place is wrong?)
> 
> I don't understand well how to check that a given Gimple argument is a string.
> I was thinking of using useless_type_conversion_p or types_compatible_p on the TREE_TYPE of some POINTER_TYPE with char_type_node, but it seems to not work as I would expect
> 
> How would you (in C++ or C for a 4.6) code such a test (that argument 2 of some GIMPLE_CALL is a string, ie. a char* in C parlance)?
> 
> (I'm coding in MELT, but I am not asking a MELT specific question; I have hard time understanding how that should be coded in C++).

FWIW I've written a type-checker for variadic arguments in my Python
plugin; this probably won't help you but the python code can be seen
here:
http://git.fedorahosted.org/cgit/gcc-python-plugin.git/tree/libcpychecker/formatstrings.py
(it checks usage of the CPython API)

In theory I'm looking up the "char" type via integer_types[itk_char],
then getting the pointer type via build_pointer_type() to get "char *",
then comparing for pointer equality.  But it got a *lot* more
complicated than that when handling special cases (e.g. a typedef that's
a char is also acceptable).

> Or where is the typechecking for __attribute__((format(printf))) functions done in the GCC source tree?

I believe it's within gcc/c-family/c-format.c 

Hope this is helpful
Dave