Re: Fix compiler warnings on 64-bit Windows
От | Tom Lane |
---|---|
Тема | Re: Fix compiler warnings on 64-bit Windows |
Дата | |
Msg-id | 15541.1581691930@sss.pgh.pa.us обсуждение исходный текст |
Ответ на | Re: Fix compiler warnings on 64-bit Windows (Peter Eisentraut <peter.eisentraut@2ndquadrant.com>) |
Ответы |
Re: Fix compiler warnings on 64-bit Windows
|
Список | pgsql-hackers |
Peter Eisentraut <peter.eisentraut@2ndquadrant.com> writes: > On 2020-02-13 16:19, Tom Lane wrote: >> According to C99 and POSIX, intptr_t should be provided by <stdint.h> ... >> now that we're requiring C99, can we get away with just #include'ing >> that directly in these test files? > I think in the past we were worried about the C library not being fully > C99. But the build farm indicates that even the trailing edge OS X and > HP-UX members have it, so I'm content to require it. Then we should > probably remove the Autoconf tests altogether. Yeah, I think that the C99 requirement has obsoleted a number of configure tests and related hackery in c.h. We just haven't got round to cleaning that up yet. BTW: I'm still concerned about the possibility of the C library being less than C99. The model that was popular back then, and which still exists on e.g. gaur, was that you could install a C99 *compiler* on a pre-C99 system, and the compiler would bring its own standard header files as necessary. While I don't have the machine booted up to check, I'm pretty sure that gaur's <stdint.h> is being supplied by the gcc installation not directly from /usr/include. On the other hand, that compiler installation is still dependent on the vendor-supplied libc. So the sorts of tests I think we can get away with removing have to do with the presence of C99-required headers, macros, typedefs, etc. Anything that is checking the presence or behavior of code in libc, we probably need to be more careful about. regards, tom lane
В списке pgsql-hackers по дате отправления: