Typo in stat(2)

Typo in stat(2)

[Christoph Kalchreuter]

Hello,

I'm under the impression that I found three typing errors in stat(2).

Used distribution: "Ubuntu 18.04.5 LTS"
Also appears in: "https://man7.org/linux/man-pages/man2/stat.2.html"

There are three Letters "e" missing in Section "DESCRIPTION", Subsection "The 
stat structure":

struct timespec st_atim;  /* Time of last access */
struct timespec st_mtim;  /* Time of last modification */
sruct timespec st_ctim;  /* Time of last status change */

should possibly be:

struct timespec st_atime;  /* Time of last access */
struct timespec st_mtime;  /* Time of last modification */
sruct timespec st_ctime;  /* Time of last status change */

Re: Typo in stat(2)

[Florian Weimer]

* Christoph Kalchreuter:

> There are three Letters "e" missing in Section "DESCRIPTION",
> Subsection "The stat structure":
>
> struct timespec st_atim;  /* Time of last access */
> struct timespec st_mtim;  /* Time of last modification */
> sruct timespec st_ctim;  /* Time of last status change */
>
> should possibly be:
>
> struct timespec st_atime;  /* Time of last access */
> struct timespec st_mtime;  /* Time of last modification */
> sruct timespec st_ctime;  /* Time of last status change */

This typo is also present in POSIX, so we cannot change it (like the
creat function).

Thanks,
Florian
-- 
Red Hat GmbH, https://de.redhat.com/ , Registered seat: Grasbrunn,
Commercial register: Amtsgericht Muenchen, HRB 153243,
Managing Directors: Charles Cachera, Brian Klemm, Laurie Krebs, Michael O'Neill

Re: Typo in stat(2)

[Michael Kerrisk (man-pages)]

On 9/30/20 2:22 PM, Florian Weimer wrote:
> * Christoph Kalchreuter:
> 
>> There are three Letters "e" missing in Section "DESCRIPTION",
>> Subsection "The stat structure":
>>
>> struct timespec st_atim;  /* Time of last access */
>> struct timespec st_mtim;  /* Time of last modification */
>> sruct timespec st_ctim;  /* Time of last status change */
>>
>> should possibly be:
>>
>> struct timespec st_atime;  /* Time of last access */
>> struct timespec st_mtime;  /* Time of last modification */
>> sruct timespec st_ctime;  /* Time of last status change */
> 
> This typo is also present in POSIX, so we cannot change it (like the
> creat function).

Hi Christoph,

Florian was being very deadpan in his humor! The names 
really are correct. In the header files, one can find

[[  # /usr/include/bits/stat.h

    struct timespec st_atim;            /* Time of last access.  */
    struct timespec st_mtim;            /* Time of last modification.  */
    struct timespec st_ctim;            /* Time of last status change.  */
# define st_atime st_atim.tv_sec        /* Backward compatibility.  */
# define st_mtime st_mtim.tv_sec
# define st_ctime st_ctim.tv_sec
]]

And struct timespec is defined as

              struct timespec {
                  time_t  tv_sec;  /* Seconds */
                  long    tv_nsec; /* Nanoseconds */
              };


Thanks,

Michael

-- 
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Linux/UNIX System Programming Training: http://man7.org/training/

Re: Typo in stat(2)

[Christoph Kalchreuter]

Hi Michael,

you're right, st_ctim is in struct stat if it's read via lstat and I missed the 
NOTES-section where it's clearly stated ("The nanosecond components of each 
timestamp are available via names of the form st_atim.tv_nsec").

I originally came from nftw(3) where I concluded from the sentence "sb is a 
pointer to the stat structure returned by a call to stat(2) for fpath." that I 
can find the description for struct stat in stat(2).

In stat(2) I found that "struct timespec st_mtim" exists in struct stat, so I 
tried to read st_mtim.tv_sec from struct stat that nftw returned.

But the compiler complained: "‘const struct stat’ has no member named ‘st_mtim’; 
did you mean ‘st_mtime’" and I concluded that there is no st_mtim in struct stat 
and as there isn't any description of st_atim, st_mtim and st_ctim in the 
explanation "The fields in the stat structure are as follows:" I assumed that 
it's an typing error in the manual.

It isn't and I surely did something wrong somewhere (I've to admit that I'm not 
that experienced programmer).

So, sorry for false alarm.

Best,
Christoph



On 9/30/20 9:38 PM, Michael Kerrisk (man-pages) wrote:
> On 9/30/20 2:22 PM, Florian Weimer wrote:
>> * Christoph Kalchreuter:
>>
>>> There are three Letters "e" missing in Section "DESCRIPTION",
>>> Subsection "The stat structure":
>>>
>>> struct timespec st_atim;  /* Time of last access */
>>> struct timespec st_mtim;  /* Time of last modification */
>>> sruct timespec st_ctim;  /* Time of last status change */
>>>
>>> should possibly be:
>>>
>>> struct timespec st_atime;  /* Time of last access */
>>> struct timespec st_mtime;  /* Time of last modification */
>>> sruct timespec st_ctime;  /* Time of last status change */
>>
>> This typo is also present in POSIX, so we cannot change it (like the
>> creat function).
> 
> Hi Christoph,
> 
> Florian was being very deadpan in his humor! The names
> really are correct. In the header files, one can find
> 
> [[  # /usr/include/bits/stat.h
> 
>      struct timespec st_atim;            /* Time of last access.  */
>      struct timespec st_mtim;            /* Time of last modification.  */
>      struct timespec st_ctim;            /* Time of last status change.  */
> # define st_atime st_atim.tv_sec        /* Backward compatibility.  */
> # define st_mtime st_mtim.tv_sec
> # define st_ctime st_ctim.tv_sec
> ]]
> 
> And struct timespec is defined as
> 
>                struct timespec {
>                    time_t  tv_sec;  /* Seconds */
>                    long    tv_nsec; /* Nanoseconds */
>                };
> 
> 
> Thanks,
> 
> Michael
> 

Re: Typo in stat(2)

[Jakub Wilk]

* Christoph Kalchreuter <christoph@kalchreuter.de>, 2020-10-01, 08:20:
>But the compiler complained: "‘const struct stat’ has no member named 
>‘st_mtim’

glibc doesn't always expose the st_mtim member (see the NOTES section 
for details), but it should be available by default.

Did you use any compiler options, or define any feature test macros?

-- 
Jakub Wilk