Re: Improve the granularity of PQsocketPoll's timeout parameter?

On Mon, Jun 10, 2024 at 11:39 PM Tom Lane <tgl(at)sss(dot)pgh(dot)pa(dot)us> wrote:
> The next question is how to spell "int64" in libpq-fe.h.

Hi. Out-of-curiosity, I grep'd for it in my 16.1 libpq:

[ddevienne(at)marsu include]$ grep 'long long' *.h
ecpg_config.h:/* Define to 1 if the system has the type `long long int'. */
ecpg_config.h:/* Define to 1 if `long long int' works and is 64 bits. */
pg_config.h:/* The normal alignment of `long long int', in bytes. */
pg_config.h:/* Define to 1 if `long long int' works and is 64 bits. */
pgtypes_interval.h:typedef long long int int64;

And the relevant snippet of pgtypes_interval.h is:

#ifdef HAVE_LONG_INT_64
#ifndef HAVE_INT64
typedef long int int64;
#endif
#elif defined(HAVE_LONG_LONG_INT_64)
#ifndef HAVE_INT64
typedef long long int int64;
#endif
#else
/* neither HAVE_LONG_INT_64 nor HAVE_LONG_LONG_INT_64 */
#error must have a working 64-bit integer datatype
#endif

Given this precedent, can't the same be done?

And if a 64-bit integer is too troublesome, why not just two 32-bit
parameters instead?
Either a (time_t + int usec), microsecond offset, clamped to [0, 1M),
or (int sec + int usec)?

I'm fine with any portable solution that allows sub-second timeouts, TBH.
Just thinking aloud here. Thanks, --DD