[RFC,v2,04/20] include/time.h: Fix conflicting timespec types on 32-bit
diff mbox series

Message ID 79521bb1530c313aad9ff019769a88ee10c5b135.1561421042.git.alistair.francis@wdc.com
State New
Headers show
Series
  • [RFC,v2,01/20] y2038: Introduce internal for glibc struct __timespec64
Related show

Commit Message

Alistair Francis June 25, 2019, 12:08 a.m. UTC
On 32-bit systems with 64-bit time_t the timespec variable will be set
to struct __timespec64 which doesn't match the expected function pointer
for __clock_settime64.

Signed-off-by: Alistair Francis <alistair.francis@wdc.com>
---
 include/time.h | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

Comments

Arnd Bergmann June 25, 2019, 11:17 a.m. UTC | #1
On Tue, Jun 25, 2019 at 2:11 AM Alistair Francis
<alistair.francis@wdc.com> wrote:

> --- a/include/time.h
> +++ b/include/time.h
> @@ -53,7 +53,8 @@ extern void __tz_compute (__time64_t timer, struct tm *tm, int use_localtime)
>    __THROW attribute_hidden;
>
>  #if __WORDSIZE == 64 \
> -  || (defined __SYSCALL_WORDSIZE && __SYSCALL_WORDSIZE == 64)
> +  || (defined __SYSCALL_WORDSIZE && __SYSCALL_WORDSIZE == 64) \
> +  || __TIMESIZE == 64
>  # define __timespec64 timespec
>  #else
>  /* The glibc Y2038-proof struct __timespec64 structure for a time value.

I think you need a bit more here: on 32-bit architectures with 64-bit time_t,
we require padding in 'struct timespec' next to tv_nsec (before or after,
depending on endianess), and I don't see  a patch in your series that
changes 'timespec' accordingly.

See also https://sourceware.org/glibc/wiki/Y2038ProofnessDesign#struct_timespec

       Arnd
Alistair Francis June 25, 2019, 10:20 p.m. UTC | #2
On Tue, Jun 25, 2019 at 4:17 AM Arnd Bergmann <arnd@arndb.de> wrote:
>
> On Tue, Jun 25, 2019 at 2:11 AM Alistair Francis
> <alistair.francis@wdc.com> wrote:
>
> > --- a/include/time.h
> > +++ b/include/time.h
> > @@ -53,7 +53,8 @@ extern void __tz_compute (__time64_t timer, struct tm *tm, int use_localtime)
> >    __THROW attribute_hidden;
> >
> >  #if __WORDSIZE == 64 \
> > -  || (defined __SYSCALL_WORDSIZE && __SYSCALL_WORDSIZE == 64)
> > +  || (defined __SYSCALL_WORDSIZE && __SYSCALL_WORDSIZE == 64) \
> > +  || __TIMESIZE == 64
> >  # define __timespec64 timespec
> >  #else
> >  /* The glibc Y2038-proof struct __timespec64 structure for a time value.
>
> I think you need a bit more here: on 32-bit architectures with 64-bit time_t,
> we require padding in 'struct timespec' next to tv_nsec (before or after,
> depending on endianess), and I don't see  a patch in your series that
> changes 'timespec' accordingly.
>
> See also https://sourceware.org/glibc/wiki/Y2038ProofnessDesign#struct_timespec

Good point, I have updated this.

Alistair

>
>        Arnd

Patch
diff mbox series

diff --git a/include/time.h b/include/time.h
index 7155b2e4db..d62fc7b2c3 100644
--- a/include/time.h
+++ b/include/time.h
@@ -53,7 +53,8 @@  extern void __tz_compute (__time64_t timer, struct tm *tm, int use_localtime)
   __THROW attribute_hidden;
 
 #if __WORDSIZE == 64 \
-  || (defined __SYSCALL_WORDSIZE && __SYSCALL_WORDSIZE == 64)
+  || (defined __SYSCALL_WORDSIZE && __SYSCALL_WORDSIZE == 64) \
+  || __TIMESIZE == 64
 # define __timespec64 timespec
 #else
 /* The glibc Y2038-proof struct __timespec64 structure for a time value.