[PATCH v2] legacy nfs.c: Change filesystem utime_h handler to utimens_h

[Ryan Long]

Reply is below.

From: Joel Sherrill <joel at rtems.org>
Sent: Friday, May 21, 2021 9:44 AM
To: Gedare Bloom <gedare at rtems.org>
Cc: Ryan Long <ryan.long at oarcorp.com>; vijay at rtems.rog; rtems-devel at rtems.org <devel at rtems.org>
Subject: Re: [PATCH v2] legacy nfs.c: Change filesystem utime_h handler to utimens_h

As a warning, this and the corresponding libbsd patch are queued
behind an RSB patch to get UTIME_NOW and UTIME_OMIT in
newlib. That patch was merged into newlib. Then the RTEMS
*utime* patch set and these two will go in at the same time. Otherwise,
there will be breakage.

On Fri, May 21, 2021 at 9:26 AM Gedare Bloom <gedare at rtems.org<mailto:gedare at rtems.org>> wrote:
On Mon, May 3, 2021 at 12:41 PM Ryan Long <ryan.long at oarcorp.com<mailto:ryan.long at oarcorp.com>> wrote:
>
> Changed nfs_utime() to nfs_utimens(), changed the arguments to use
> a timespec array instead of individual variables for access and
> modified time.
>
> Updates #4400
> ---
>  nfsclient/src/nfs.c | 50 +++++++++++++++++++++++++-------------------------
>  1 file changed, 25 insertions(+), 25 deletions(-)
>
> diff --git a/nfsclient/src/nfs.c b/nfsclient/src/nfs.c
> index bc9a2c6..72fefb9 100644
> --- a/nfsclient/src/nfs.c
> +++ b/nfsclient/src/nfs.c
> @@ -2050,19 +2050,19 @@ static int nfs_rmnod(
>         return rv;
>  }
>
> -static int nfs_utime(
> +static int nfs_utimens(
>         const rtems_filesystem_location_info_t  *pathloc, /* IN */
> -       time_t                                   actime,  /* IN */
> -       time_t                                   modtime  /* IN */
> +       struct timespec                          times[2] /* IN */
> +
>  )
>  {
>  sattr  arg;
>
>         /* TODO: add rtems EPOCH - UNIX EPOCH seconds */
> -       arg.atime.seconds  = actime;
> -       arg.atime.useconds = 0;
> -       arg.mtime.seconds  = modtime;
> -       arg.mtime.useconds = 0;
> +       arg.atime.seconds  = times[0].tv_sec;
> +       arg.atime.useconds = times[0].tv_nsec / 1000;
Do we have a macro exposed for this, like in
TOD_NANOSECONDS_PER_MICROSECOND? If not, this is fine I suppose.

This wouldn't be public since it doesn't start with RTEMS_ and I don't
see any of the TOD_ constants redefined in the RTEMS namespace.


> +       arg.mtime.seconds  = times[1].tv_sec;
> +       arg.mtime.useconds = times[1].tv_nsec / 1000;
>
>         return nfs_sattr(pathloc->node_access, &arg, SATTR_ATIME | SATTR_MTIME);
>  }
> @@ -2254,25 +2254,25 @@ sattr   arg;
>  }
>
>  const struct _rtems_filesystem_operations_table nfs_fs_ops = {
> -       .lock_h         = nfs_lock,
> -       .unlock_h       = nfs_unlock,
> -       .eval_path_h    = nfs_eval_path,
> -       .link_h         = nfs_link,
> +       .lock_h            = nfs_lock,
> +       .unlock_h          = nfs_unlock,
> +       .eval_path_h       = nfs_eval_path,
> +       .link_h            = nfs_link,
>         .are_nodes_equal_h = nfs_are_nodes_equal,
> -       .mknod_h        = nfs_mknod,
> -       .rmnod_h        = nfs_rmnod,
> -       .fchmod_h       = nfs_fchmod,
> -       .chown_h        = nfs_chown,
> -       .clonenod_h     = nfs_clonenode,
> -       .freenod_h      = nfs_freenode,
> -       .mount_h        = rtems_filesystem_default_mount,
> -       .unmount_h      = rtems_filesystem_default_unmount,
> -       .fsunmount_me_h = nfs_fsunmount_me,
> -       .utime_h        = nfs_utime,
> -       .symlink_h      = nfs_symlink,
> -       .readlink_h     = nfs_readlink,
> -       .rename_h       = nfs_rename,
> -       .statvfs_h      = rtems_filesystem_default_statvfs
> +       .mknod_h           = nfs_mknod,
> +       .rmnod_h           = nfs_rmnod,
> +       .fchmod_h          = nfs_fchmod,
> +       .chown_h           = nfs_chown,
> +       .clonenod_h        = nfs_clonenode,
> +       .freenod_h         = nfs_freenode,
> +       .mount_h           = rtems_filesystem_default_mount,
> +       .unmount_h         = rtems_filesystem_default_unmount,
> +       .fsunmount_me_h    = nfs_fsunmount_me,
> +       .utimens_h         = nfs_utimens,
> +       .symlink_h         = nfs_symlink,
> +       .readlink_h        = nfs_readlink,
> +       .rename_h          = nfs_rename,
> +       .statvfs_h         = rtems_filesystem_default_statvfs
Why so many additional whitespace changes?
[Ryan Long] When I changed utime_h to utimens_h, it got everything out of alignment, so I moved everything over.


>  };
>
>  /*****************************************
> --
> 1.8.3.1
>
> _______________________________________________
> devel mailing list
> devel at rtems.org<mailto:devel at rtems.org>
> http://lists.rtems.org/mailman/listinfo/devel
_______________________________________________
devel mailing list
devel at rtems.org<mailto:devel at rtems.org>
http://lists.rtems.org/mailman/listinfo/devel
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.rtems.org/pipermail/devel/attachments/20210528/3b7583d6/attachment-0001.html>