Daniel Stenberg wrote:
> On Mon, 14 Apr 2008, Michal Marek wrote:
>> I mean, it's a pity that passing a bitfield to the curl_easy_setopt
>> macro generates code that gcc doesn't accept, but I don't see a way to
>> work around this.
>
> Right, and I think we should add an #ifdef or similar on a define (that isn't
> __CURL_TYPECHECK_GCC_H) to allow applications to specificly disable the type
> check. I mean, this code probably still works fine built without the type
> check and it seems silly to not provide a work-around for this problem then.

OK, this is something I wanted to avoid, but this issue probably
justifies it. Is the attached patch OK? Or should there be an #ifdef
around the whole block, so that the #define
curl_easy_setopt(handle,opt,param) curl_easy_setopt(handle,opt,param)
macros are also disabled?

Michal