VFS "stat light" fstatat, inconsistent *at() function interfaces

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

 



Oleg Drokin wrote:
> Also it adds AT_STAT_SELF flag to mean that the stat should be  
> executed not on a path, but on the passed filedecriptor itself.

Ok.  See below.

> Jamie Lokier wrote:
>>    ret = fstatat(fd, 0, &stat, AT_STAT_TYPE|AT_STAT_SIZE);
>> Where 0 in the path automatically means AT_STAT_SELF.
>> Or this?
>>    ret = fstatat(fd, AT_STAT_SELF, &stat, AT_STAT_TYPE|AT_STAT_SIZE);
>> Where AT_STAT_SELF is defined as (const char *)-1 in system headers,
>> to better catch null pointer application bugs.
> 
> While I think the last one might be attractive, we also need to think
> about compatibility, unaware kernels would return -EFAULT if you pass  
> some nonsense as a path, which is much less useful than say nice EINVAL.

About compatibility, you might look at this in "man utimensat":

       On Linux, futimens() is a library function implemented on  top  of  the
       utimensat() system call.  To support this, the Linux utimensat() system
       call implements a non-standard feature: if pathname is NULL,  then  the
       call  modifies  the  timestamps  of  the  file  referred to by the file
       descriptor dirfd (which may refer to any type  of  file).   Using  this
       feature, the call futimens(fd, times) is implemented as:

           utimensat(fd, NULL, times, 0);

It's inconsistent to have AT_STAT_SELF for the stat-light call, but
NULL path for the utimensat() call (system call; Glibc hides it)
to mean the same thing, which is act on the passed file descriptor.

I think your scheme is better :-) And safer when an application passes
NULL due to programming eror.  But it is inconsistent.

It would be nice if they weren't, or if they used a common flag, in
the same way that other flags like AT_SYMLINK_NOFOLLOW have the same
meaning for all the *at() calls.

Then again, the *at() system calls are somewhat inconsistent anyway.
fchmodat() doesn't accept AT_SYMLINK_NOFOLLOW according to the man
page I have here.

-- Jamie
--
To unsubscribe from this list: send the line "unsubscribe linux-fsdevel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[Index of Archives]     [Linux Ext4 Filesystem]     [Union Filesystem]     [Filesystem Testing]     [Ceph Users]     [Ecryptfs]     [AutoFS]     [Kernel Newbies]     [Share Photos]     [Security]     [Netfilter]     [Bugtraq]     [Yosemite News]     [MIPS Linux]     [ARM Linux]     [Linux Security]     [Linux Cachefs]     [Reiser Filesystem]     [Linux RAID]     [Samba]     [Device Mapper]     [CEPH Development]
  Powered by Linux