Re: [PATCH v2 09/13] tools/nolibc: always use 64-bit time types

["Arnd Bergmann" <arnd@arndb.de>]

On Mon, Dec 1, 2025, at 11:35, Willy Tarreau wrote:
> On Mon, Dec 01, 2025 at 08:45:00AM +0100, Arnd Bergmann wrote:
>> On Sun, Nov 30, 2025, at 11:58, Willy Tarreau wrote:
>>
>> #if __TIMESIZE == 64 && __WORDSIZE == 32
>> # define __TIME_T_TYPE          __SQUAD_TYPE
>> # define __SUSECONDS_T_TYPE     __SQUAD_TYPE
>> #else
>> # define __TIME_T_TYPE          __SLONGWORD_TYPE
>> # define __SUSECONDS_T_TYPE     __SLONGWORD_TYPE
>> #endif
>> 
>> so this one is explicitly the same width as tv_sec, which has all
>> the issues you listed, but avoids the need for padding.
>
> Ah we seem to just have checked different versions then,
> as in mine there was still some extra padding left depending
> on the endianness :-)

The padding is definitely there in timespec around tv_nsec,
just not in timeval.

Oddly, the version I quoted is from my arm64 /usr/include/
installation and looks different from what I see in the glibc
history, though that also uses a 64-bit tv_usec:

bits/typesizes.h:#define __SUSECONDS64_T_TYPE   __SQUAD_TYPE
posix/bits/types.h:__STD_TYPE __SUSECONDS64_T_TYPE __suseconds64_t;
struct timeval
{
#ifdef __USE_TIME64_REDIRECTS
  __time64_t tv_sec;            /* Seconds.  */
  __suseconds64_t tv_usec;      /* Microseconds.  */
#else
  __time_t tv_sec;              /* Seconds.  */
  __suseconds_t tv_usec;        /* Microseconds.  */
#endif
};

>> C23 has updated the definition and does allow int64_t tv_nsec.
>
> So it purposely breaks existing apps or does it apply only to those
> compiled with -mstd=c23 ?

Neither, it's just that nolibc with a 64-bit tv_nsec would
be compliant with c23, just not earlier versions.

I expect glibc to stick with 32-bit timespec and padding, which
is still compliant with the new definition of

|   The type of tv_nsec is an implementation-defined signed integer type
|   that can represent integers in [​0​, 999999999]. 

>> I think it makes sense for nolibc to just follow the kernel's
>> definition here.
>
> Given the very narrow range of existing code that can be impacted,
> I'm fine, but in general I try to remain extremely cautious about
> portability: as a general rule, ifdefs needed to address possible
> incompatibilities, if any, should rather be in the libc code itself
> and not in the user application. I just ran a quick check and don't
> have code using &tv_usec nor &tv_nsec so here the risk remains quite
> low.

Ok

     ARnd