On Tue, 2024-03-26 at 19:36 -0400, Benjamin Marzinski wrote:
> On Sun, Mar 24, 2024 at 10:12:57PM +0100, Martin Wilck wrote:
> > With the late changes to the device mapper rules, DM_SUSPENDED
> > is not exported any more. Use .DM_SUSPENDED instead.
> >
> > Note that although 11-dm-mpath.rules is not a part of lvm2, it
> > can be considered as part of the device-mapper layer (everything
> > before 13-dm-disk.rules can), and is thus allowed to use
> > .DM_SUSPENDED. In practice .DM_SUSPENDED is equivalent to
> > DM_UDEV_DISABLE_OTHER_RULES_FLAG for multipath devices, but
> > using .DM_SUSPENDED here makes the intention more obvious.
> >
> > Signed-off-by: Martin Wilck <mwilck@xxxxxxxx>
> > ---
> > multipath/11-dm-mpath.rules.in | 13 ++++++++-----
> > 1 file changed, 8 insertions(+), 5 deletions(-)
> >
> > diff --git a/multipath/11-dm-mpath.rules.in b/multipath/11-dm-
> > mpath.rules.in
> > index 8f22954..c4b5685 100644
> > --- a/multipath/11-dm-mpath.rules.in
> > +++ b/multipath/11-dm-mpath.rules.in
> > @@ -4,8 +4,11 @@ ENV{DM_UUID}!="mpath-?*", GOTO="mpath_end"
> >
> > IMPORT{db}="MPATH_DEVICE_READY"
> >
> > +# device-mapper rules v2 compatibility
> > +ENV{.DM_SUSPENDED}!="?*", ENV{.DM_SUSPENDED}="$env{DM_SUSPENDED}"
> > +
> > # Coldplug event while device is suspended (e.g. during a reload)
> > -ACTION=="add", ENV{DM_ACTIVATION}=="1", ENV{DM_SUSPENDED}=="1", \
> > +ACTION=="add", ENV{DM_ACTIVATION}=="1", ENV{.DM_SUSPENDED}=="1", \
> > PROGRAM="/bin/logger -t 11-dm-mpath.rules -p
> > daemon.warning \"Coldplug event for suspended device\"", \
> > ENV{DM_COLDPLUG_SUSPENDED}="1", GOTO="scan_import"
> >
> > @@ -17,7 +20,7 @@ ENV{DM_UDEV_RULES_VSN}=="3",
> > ENV{.DM_SUSPENDED}!="1", GOTO="scan_import"
> > # from DB in 10-dm.rules. If the device is not suspended, clear
> > the flag.
> > # This is safe for multipath where
> > DM_UDEV_DISABLE_OTHER_RULES_FLAG is basically
> > # equivalent to DM_SUSPENDED==1 || DISK_RO==1
> > -ENV{DM_UDEV_RULES_VSN}!="3", ENV{DM_SUSPENDED}!="1",
> > ENV{DISK_RO}!="1", \
> > +ENV{DM_UDEV_RULES_VSN}!="3", ENV{.DM_SUSPENDED}!="1",
> > ENV{DISK_RO}!="1", \
> > ENV{DM_UDEV_DISABLE_OTHER_RULES_FLAG}="",
> > GOTO="scan_import"
> > LABEL="mpath_coldplug_end"
> >
> > @@ -68,7 +71,7 @@ LABEL="mpath_action"
> > # Activation might have been partially skipped. Activate the
> > device now,
> > # i.e. disable the MPATH_UNCHANGED logic and set DM_ACTIVATION=1.
> > IMPORT{db}="DM_COLDPLUG_SUSPENDED"
> > -ENV{DM_COLDPLUG_SUSPENDED}=="1", ENV{DM_SUSPENDED}!="1", \
> > +ENV{DM_COLDPLUG_SUSPENDED}=="1", ENV{.DM_SUSPENDED}!="1", \
>
> When I reviewed patch 0001, I wondered if it was o.k. to not force
> DM_ACTIVATION to "1" if a device had previously been suspended when
> MPATH_DEVICE_READY switched to "1". I thought it might it might be
> o.k.
> because there would be another event when the device resumed, which
> would have DM_ACTIVATION set to "1". But if that's the case then we
> shouldn't need to worry about a device being suspended when a
> coldplug
> happens. We will get an event when the device resumes, and there's
> not
> much we can do until it gets resumed at any rate. So either patch
> 0001
> is wrong or we can remove this DM_COLDPLUG_SUSPENDED forced
> activation
> code or I'm missing some important difference between the two.
I tried to explain this in 02bbc17 ("11-dm-mpath.rules: handle reloads
during coldplug events"). But I just needed to recall the logic again
myself, it's subtle.
We will indeed get another change event when the device is resumed.
But because of our MPATH_UNCHANGED logic, we may see that the
the properties of the device haven't changed, and unset DM_ACTIVATION.
Thus follow-up rules won't get executed, neither in coldplug event
itself (device suspended) nor in the follow-up change event (device
unchanged). I hope this makes sense.
The problem that you pointed out for 0001 for this series is similar,
but it handles a different case where the device switched from 0 to one
usable path while the device was suspended. Patch 0001 of this series
is really broken, and it breaks a patch that I'd just created a month
ago, which is truly embarrassing. udev rules are confusing :-/
Regards
Martin
