On 12/17/16 7:57 PM, Deepa Dinamani wrote:
> Add the utimes command to provide a way to utilize
> the futimens C library call. This is the
> interface to the utimensat system call, which updates
> the mtime and atime of a file.
>
> Signed-off-by: Deepa Dinamani <[email protected]>
...
> +static int
> +utimes_f(
> + int argc,
> + char **argv)
> +{
> + struct timespec t[2];
> + int result;
> +
> + if (argc != 5)
> + return command_usage(&utimes_cmd);
Because you set argsmin & argsmax to 4, it should be impossible
to get here with anything other than argc=5 - it's caught
elsewhere:
xfs_io> utimes
bad argument count 0 to utimes, expected 4 arguments
xfs_io> utimes 1 2 3 4 5
bad argument count 5 to utimes, expected 4 arguments
> +
> + /* Get the timestamps */
> + result = timespec_from_string(argv[1], argv[2], &t[0]);
> + if (result) {
> + fprintf(stderr, "Bad value for atime\n");
> + return 1;
> + }
> + result = timespec_from_string(argv[3], argv[4], &t[1]);
> + if (result) {
> + fprintf(stderr, "Bad value for mtime\n");
> + return 1;
> + }
> +
> + /* Call futimens to update time. */
> + if (futimens(file->fd, t)) {
> + perror("futimens");
> + return 1;
> + }
Most xfs_io functions return 0 even on errors, possibly after
setting exit_code = 1 to change the ultimate exit code;
returning 1 will cause all processing to stop, and/or kick you
out of the interactive shell:
$ xfs_io file
xfs_io> utimes a b c d
Bad value for atime
$
This needs some attention across all of xfs_io, but you might want
to return 0 for now for consistency with other commands.
-Eric
> +
> + return 0;
> +}
> +
> +void
> +utimes_init(void)
> +{
> + utimes_cmd.name = "utimes";
> + utimes_cmd.cfunc = utimes_f;
> + utimes_cmd.argmin = 4;
> + utimes_cmd.argmax = 4;
> + utimes_cmd.flags = CMD_NOMAP_OK | CMD_FOREIGN_OK;
> + utimes_cmd.args = _("atime_sec atime_nsec mtime_sec mtime_nsec");
> + utimes_cmd.oneline = _("Update file times of the current file");
> + utimes_cmd.help = utimes_help;
> +
> + add_command(&utimes_cmd);
> +}
> diff --git a/libxcmd/input.c b/libxcmd/input.c
> index 5a7dce3..2fdb3e8 100644
> --- a/libxcmd/input.c
> +++ b/libxcmd/input.c
> @@ -327,6 +327,28 @@ timestr(
> }
>
> /*
> + * Convert from a pair of arbitrary user strings into a timespec.
> + */
> +
> +int
> +timespec_from_string(
> + const char * secs,
> + const char * nsecs,
> + struct timespec * ts)
> +{
> + char* p;
> + if (!secs || !nsecs || !ts)
> + return -1;
> + ts->tv_sec = strtoull(secs, &p, 0);
> + if (*p)
> + return -1;
> + ts->tv_nsec = strtoull(nsecs, &p, 0);
> + if (*p)
> + return -1;
> + return 0;
I'd return 1/0 not -1/0 - not that big a deal, but the reason
the i.e. prid_from_string() functions return -1 on error
is because they actually return an ID, which is >= 0, so
it detects "== -1" as an error, and can't simply test
1/0.
> +}
> +
> +/*
> * Convert from arbitrary user strings into a numeric ID.
> * If it's all numeric, we convert that inplace, else we do
> * the name lookup, and return the found identifier.
> diff --git a/man/man8/xfs_io.8 b/man/man8/xfs_io.8
> index 2c56f09..3ffe439 100644
> --- a/man/man8/xfs_io.8
> +++ b/man/man8/xfs_io.8
> @@ -589,6 +589,17 @@ Copy data into the open file beginning at
> Copy up to
> .I length
> bytes of data.
> +.RE
> +.PD
> +.TP
> +.TP
don't need two .TPs, a patch to remove the others is pending.
thanks,
-Eric
> +.BI utimes " atime_sec atime_nsec mtime_sec mtime_nsec"
> +The utimes command changes the atime and mtime of the current file.
> +sec uses UNIX timestamp notation and is the seconds elapsed since
> +1970-01-01 00:00:00 UTC.
> +nsec is the nanoseconds since the sec. This value needs to be in
> +the range 0-999999999 with UTIME_NOW and UTIME_OMIT being exceptions.
> +Each (sec, nsec) pair constitutes a single timestamp value.
>
> .SH MEMORY MAPPED I/O COMMANDS
> .TP
> @@ -875,6 +886,7 @@ verbose output will be printed.
> .BR fstatfs (2),
> .BR fsync (2),
> .BR ftruncate (2),
> +.BR futimens (3),
> .BR mmap (2),
> .BR msync (2),
> .BR open (2),
>
_______________________________________________
Y2038 mailing list
[email protected]
https://lists.linaro.org/mailman/listinfo/y2038
