| Message ID | 20241008094503.368923-1-amir73il@gmail.com (mailing list archive) |
|---|---|
| State | New |
| Headers | show |
| Series | fanotify.7,fanotify_mark.2: update documentation of fanotify w.r.t fsid | expand |
On Tue 08-10-24 11:45:03, Amir Goldstein wrote: > Clarify the conditions for getting the -EXDEV and -ENODEV errors. > > Signed-off-by: Amir Goldstein <amir73il@gmail.com> Looks good. Feel free to add: Reviewed-by: Jan Kara <jack@suse.cz> But I've read somewhere that Alejandro stepped down as manpages maintainer so they are officially unmaintained? Honza > Hi Alejandro, > > This is a followup on fanotify changes from v6.8 > that are forgot to follow up on at the time. > > Thanks, > Amir. > > man/man2/fanotify_mark.2 | 27 +++++++++++++++++++++------ > man/man7/fanotify.7 | 10 ++++++++++ > 2 files changed, 31 insertions(+), 6 deletions(-) > > diff --git a/man/man2/fanotify_mark.2 b/man/man2/fanotify_mark.2 > index fc9b83459..b5e091c25 100644 > --- a/man/man2/fanotify_mark.2 > +++ b/man/man2/fanotify_mark.2 > @@ -659,17 +659,16 @@ The filesystem object indicated by > .I dirfd > and > .I pathname > -is not associated with a filesystem that supports > +is associated with a filesystem that reports zero > .I fsid > (e.g., > .BR fuse (4)). > -.BR tmpfs (5) > -did not support > -.I fsid > -prior to Linux 5.13. > -.\" commit 59cda49ecf6c9a32fae4942420701b6e087204f6 > This error can be returned only with an fanotify group that identifies > filesystem objects by file handles. > +Since Linux 6.8, > +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 > +this error can be returned only when > +trying to add a mount or filesystem mark. > .TP > .B ENOENT > The filesystem object indicated by > @@ -768,6 +767,22 @@ which uses a different > than its root superblock. > This error can be returned only with an fanotify group that identifies > filesystem objects by file handles. > +Since Linux 6.8, > +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 > +this error will be returned > +when trying to add a mount or filesystem mark on a subvolume, > +when trying to add inode marks in different subvolumes, > +or when trying to add inode marks in a > +.BR btrfs (5) > +subvolume and in another filesystem. > +Since Linux 6.8, > +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 > +this error will also be returned > +when trying to add marks in different filesystems, > +where one of the filesystems reports zero > +.I fsid > +(e.g., > +.BR fuse (4)). > .SH STANDARDS > Linux. > .SH HISTORY > diff --git a/man/man7/fanotify.7 b/man/man7/fanotify.7 > index 449af949c..db8fe6c00 100644 > --- a/man/man7/fanotify.7 > +++ b/man/man7/fanotify.7 > @@ -575,6 +575,16 @@ and contains the same value as > .I f_fsid > when calling > .BR statfs (2). > +Note that some filesystems (e.g., > +.BR fuse (4)) > +report zero > +.IR fsid . > +In these cases, it is not possible to use > +.I fsid > +to associate the event with a specific filesystem instance, > +so monitoring different filesystem instances that report zero > +.I fsid > +with the same fanotify group is not supported. > .TP > .I handle > This field contains a variable-length structure of type > -- > 2.34.1 >
On Wed, Oct 9, 2024 at 5:38 PM Jan Kara <jack@suse.cz> wrote: > > On Tue 08-10-24 11:45:03, Amir Goldstein wrote: > > Clarify the conditions for getting the -EXDEV and -ENODEV errors. > > > > Signed-off-by: Amir Goldstein <amir73il@gmail.com> > > Looks good. Feel free to add: > > Reviewed-by: Jan Kara <jack@suse.cz> > > But I've read somewhere that Alejandro stepped down as manpages maintainer > so they are officially unmaintained? Yes, I just caught up with this news. Anyway, it's good to have the patch on the list. I will be maintaining the fanotify man pages queue until manpages are back to maintenance. Thanks, Amir.
Hi Amir, Jan, On Wed, Oct 09, 2024 at 05:38:36PM +0200, Jan Kara wrote: > On Tue 08-10-24 11:45:03, Amir Goldstein wrote: > > Clarify the conditions for getting the -EXDEV and -ENODEV errors. > > > > Signed-off-by: Amir Goldstein <amir73il@gmail.com> > > Looks good. Feel free to add: Please see some minor inline comments below. > Reviewed-by: Jan Kara <jack@suse.cz> Thanks! > > But I've read somewhere that Alejandro stepped down as manpages maintainer > so they are officially unmaintained? A contract is imminent, and I've started to review/apply old patches today already. I'll probably make an official announcement soon. Maintenance is restored. (As much as I possibly can, since my region has limited electricity, water, and internet, after the worst flooding in centuries.) Have a lovely day! Alex > > Honza > > > Hi Alejandro, > > > > This is a followup on fanotify changes from v6.8 > > that are forgot to follow up on at the time. > > > > Thanks, > > Amir. > > > > man/man2/fanotify_mark.2 | 27 +++++++++++++++++++++------ > > man/man7/fanotify.7 | 10 ++++++++++ > > 2 files changed, 31 insertions(+), 6 deletions(-) > > > > diff --git a/man/man2/fanotify_mark.2 b/man/man2/fanotify_mark.2 > > index fc9b83459..b5e091c25 100644 > > --- a/man/man2/fanotify_mark.2 > > +++ b/man/man2/fanotify_mark.2 > > @@ -659,17 +659,16 @@ The filesystem object indicated by > > .I dirfd > > and > > .I pathname > > -is not associated with a filesystem that supports > > +is associated with a filesystem that reports zero > > .I fsid > > (e.g., > > .BR fuse (4)). > > -.BR tmpfs (5) > > -did not support > > -.I fsid > > -prior to Linux 5.13. > > -.\" commit 59cda49ecf6c9a32fae4942420701b6e087204f6 > > This error can be returned only with an fanotify group that identifies > > filesystem objects by file handles. > > +Since Linux 6.8, > > +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 > > +this error can be returned only when I think "when" is more appropriate in the following line. It also adds some consistency with the rest of the patch below (@@762). > > +trying to add a mount or filesystem mark. > > .TP > > .B ENOENT > > The filesystem object indicated by > > @@ -768,6 +767,22 @@ which uses a different > > than its root superblock. > > This error can be returned only with an fanotify group that identifies > > filesystem objects by file handles. > > +Since Linux 6.8, > > +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 > > +this error will be returned > > +when trying to add a mount or filesystem mark on a subvolume, > > +when trying to add inode marks in different subvolumes, > > +or when trying to add inode marks in a > > +.BR btrfs (5) > > +subvolume and in another filesystem. > > +Since Linux 6.8, > > +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 > > +this error will also be returned > > +when trying to add marks in different filesystems, > > +where one of the filesystems reports zero > > +.I fsid > > +(e.g., > > +.BR fuse (4)). > > .SH STANDARDS > > Linux. > > .SH HISTORY > > diff --git a/man/man7/fanotify.7 b/man/man7/fanotify.7 > > index 449af949c..db8fe6c00 100644 > > --- a/man/man7/fanotify.7 > > +++ b/man/man7/fanotify.7 > > @@ -575,6 +575,16 @@ and contains the same value as > > .I f_fsid > > when calling > > .BR statfs (2). > > +Note that some filesystems (e.g., > > +.BR fuse (4)) > > +report zero > > +.IR fsid . > > +In these cases, it is not possible to use Please break the line after the comma. > > +.I fsid > > +to associate the event with a specific filesystem instance, > > +so monitoring different filesystem instances that report zero > > +.I fsid > > +with the same fanotify group is not supported. > > .TP > > .I handle > > This field contains a variable-length structure of type > > -- > > 2.34.1 > > > -- > Jan Kara <jack@suse.com> > SUSE Labs, CR
Hi Alejandro! On Fri 01-11-24 14:07:32, Alejandro Colomar wrote: > On Wed, Oct 09, 2024 at 05:38:36PM +0200, Jan Kara wrote: > > On Tue 08-10-24 11:45:03, Amir Goldstein wrote: > > > Clarify the conditions for getting the -EXDEV and -ENODEV errors. > > > > > > Signed-off-by: Amir Goldstein <amir73il@gmail.com> > > > > Looks good. Feel free to add: > > Please see some minor inline comments below. > > > Reviewed-by: Jan Kara <jack@suse.cz> > > Thanks! > > > > > But I've read somewhere that Alejandro stepped down as manpages maintainer > > so they are officially unmaintained? > > A contract is imminent, and I've started to review/apply old patches > today already. I'll probably make an official announcement soon. I'm happy to hear that! > Maintenance is restored. (As much as I possibly can, since my region > has limited electricity, water, and internet, after the worst flooding > in centuries.) I've heard about huge floods in Spain. We had pretty bad ones a month and half ago in Czech republic as well. But my area was only lightly affected. Stay safe! Honza
diff --git a/man/man2/fanotify_mark.2 b/man/man2/fanotify_mark.2 index fc9b83459..b5e091c25 100644 --- a/man/man2/fanotify_mark.2 +++ b/man/man2/fanotify_mark.2 @@ -659,17 +659,16 @@ The filesystem object indicated by .I dirfd and .I pathname -is not associated with a filesystem that supports +is associated with a filesystem that reports zero .I fsid (e.g., .BR fuse (4)). -.BR tmpfs (5) -did not support -.I fsid -prior to Linux 5.13. -.\" commit 59cda49ecf6c9a32fae4942420701b6e087204f6 This error can be returned only with an fanotify group that identifies filesystem objects by file handles. +Since Linux 6.8, +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 +this error can be returned only when +trying to add a mount or filesystem mark. .TP .B ENOENT The filesystem object indicated by @@ -768,6 +767,22 @@ which uses a different than its root superblock. This error can be returned only with an fanotify group that identifies filesystem objects by file handles. +Since Linux 6.8, +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 +this error will be returned +when trying to add a mount or filesystem mark on a subvolume, +when trying to add inode marks in different subvolumes, +or when trying to add inode marks in a +.BR btrfs (5) +subvolume and in another filesystem. +Since Linux 6.8, +.\" commit 30ad1938326bf9303ca38090339d948975a626f5 +this error will also be returned +when trying to add marks in different filesystems, +where one of the filesystems reports zero +.I fsid +(e.g., +.BR fuse (4)). .SH STANDARDS Linux. .SH HISTORY diff --git a/man/man7/fanotify.7 b/man/man7/fanotify.7 index 449af949c..db8fe6c00 100644 --- a/man/man7/fanotify.7 +++ b/man/man7/fanotify.7 @@ -575,6 +575,16 @@ and contains the same value as .I f_fsid when calling .BR statfs (2). +Note that some filesystems (e.g., +.BR fuse (4)) +report zero +.IR fsid . +In these cases, it is not possible to use +.I fsid +to associate the event with a specific filesystem instance, +so monitoring different filesystem instances that report zero +.I fsid +with the same fanotify group is not supported. .TP .I handle This field contains a variable-length structure of type
Clarify the conditions for getting the -EXDEV and -ENODEV errors. Signed-off-by: Amir Goldstein <amir73il@gmail.com> --- Hi Alejandro, This is a followup on fanotify changes from v6.8 that are forgot to follow up on at the time. Thanks, Amir. man/man2/fanotify_mark.2 | 27 +++++++++++++++++++++------ man/man7/fanotify.7 | 10 ++++++++++ 2 files changed, 31 insertions(+), 6 deletions(-)