[PATCH] inptl: Use 64 bit time_t stat internally for sem_open (BZ #28880)

[PATCH] inptl: Use 64 bit time_t stat internally for sem_open (BZ #28880)

[Adhemerval Zanella]

The __sem_check_add_mapping internal stat calls fails with
EOVERFLOW if system time is larger than 32 bit.

It is a missing spot from 52a5fe70a2c fix to use 64 bit stat
internally.

Checked on x86_64-linux-gnu and i686-linux-gnu.
---
 sysdeps/pthread/sem_routines.c | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/sysdeps/pthread/sem_routines.c b/sysdeps/pthread/sem_routines.c
index 9e1924398d..877e1e8556 100644
--- a/sysdeps/pthread/sem_routines.c
+++ b/sysdeps/pthread/sem_routines.c
@@ -25,7 +25,7 @@
 struct inuse_sem
 {
   dev_t dev;
-  ino_t ino;
+  ino64_t ino;
   int refcnt;
   sem_t *sem;
   char name[];
@@ -34,7 +34,7 @@ struct inuse_sem
 struct search_sem
 {
   dev_t dev;
-  ino_t ino;
+  ino64_t ino;
   int refcnt;
   sem_t *sem;
   char name[NAME_MAX + 1];
@@ -77,8 +77,8 @@ __sem_check_add_mapping (const char *name, int fd, sem_t *existing)
   sem_t *result = SEM_FAILED;
 
   /* Get the information about the file.  */
-  struct stat64 st;
-  if (__fstat64 (fd, &st) == 0)
+  struct __stat64_t64 st;
+  if (__fstat64_time64 (fd, &st) == 0)
     {
       /* Get the lock.  */
       lll_lock (sem_mappings_lock, LLL_PRIVATE);
-- 
2.32.0

Re: [PATCH] inptl: Use 64 bit time_t stat internally for sem_open (BZ #28880)

[Florian Weimer]

* Adhemerval Zanella via Libc-alpha:

> The __sem_check_add_mapping internal stat calls fails with
> EOVERFLOW if system time is larger than 32 bit.
>
> It is a missing spot from 52a5fe70a2c fix to use 64 bit stat
> internally.

Typo in the subject, and this isn't actually nptl-specific, so you could
use pthread.  But I think you will need to come up with caustom solution
for Hurd.

Thanks,
Florian

Re: [PATCH] inptl: Use 64 bit time_t stat internally for sem_open (BZ #28880)

[Adhemerval Zanella]

On 11/02/2022 11:52, Florian Weimer wrote:
> * Adhemerval Zanella via Libc-alpha:
> 
>> The __sem_check_add_mapping internal stat calls fails with
>> EOVERFLOW if system time is larger than 32 bit.
>>
>> It is a missing spot from 52a5fe70a2c fix to use 64 bit stat
>> internally.
> 
> Typo in the subject, and this isn't actually nptl-specific, so you could
> use pthread.  But I think you will need to come up with caustom solution
> for Hurd.

Hurd will continue to use the code as-is, __fstat64_time64 will continue
to be __fstat64.