mkdir of /etc/.udev/rules.d/

mkdir of /etc/.udev/rules.d/

[Marco d'Itri]

udev_rules_new() in udev/udev-rules.c unconditionally creates the
directory.
This is a problem because the function is called also by e.g. udevadm
test, and creating /dev/.udev/ when it does not exist is an unacceptable
side effect which will break everything else that checks for its
existence to know if udev is running.
Why is this needed?

-- 
ciao,
Marco

Re: mkdir of /etc/.udev/rules.d/

[Kay Sievers]

On Tue, Nov 3, 2009 at 16:21, Marco d'Itri <md@linux.it> wrote:
> udev_rules_new() in udev/udev-rules.c unconditionally creates the
> directory.
> This is a problem because the function is called also by e.g. udevadm
> test, and creating /dev/.udev/ when it does not exist is an unacceptable
> side effect which will break everything else that checks for its
> existence to know if udev is running.
> Why is this needed?

No specific reason, I remember. It could move into the daemon, before
the inotify watch.

Thanks,
Kay

Re: mkdir of /etc/.udev/rules.d/

[Marco d'Itri]

On Nov 04, Kay Sievers <kay.sievers@vrfy.org> wrote:

> > udev_rules_new() in udev/udev-rules.c unconditionally creates the
> > directory.
> > This is a problem because the function is called also by e.g. udevadm
> > test, and creating /dev/.udev/ when it does not exist is an unacceptable
> > side effect which will break everything else that checks for its
> > existence to know if udev is running.
> > Why is this needed?
> No specific reason, I remember. It could move into the daemon, before
> the inotify watch.
Why create it at all?
(I suspect that this was done to suppress a warning about the missing
directory, but then I'd rather do it properly with a flag...)

-- 
ciao,
Marco

Re: mkdir of /etc/.udev/rules.d/

[Kay Sievers]

On Wed, Nov 4, 2009 at 15:28, Marco d'Itri <md@linux.it> wrote:
> On Nov 04, Kay Sievers <kay.sievers@vrfy.org> wrote:
>
>> > udev_rules_new() in udev/udev-rules.c unconditionally creates the
>> > directory.
>> > This is a problem because the function is called also by e.g. udevadm
>> > test, and creating /dev/.udev/ when it does not exist is an unacceptable
>> > side effect which will break everything else that checks for its
>> > existence to know if udev is running.
>> > Why is this needed?
>> No specific reason, I remember. It could move into the daemon, before
>> the inotify watch.
> Why create it at all?
> (I suspect that this was done to suppress a warning about the missing
> directory, but then I'd rather do it properly with a flag...)

It's to be able to subscribe to changes in that directory with
inotify. We want to create it before we try to watch it, but only the
daemon needs to do this.

Kay

Re: mkdir of /etc/.udev/rules.d/

[Kay Sievers]

On Wed, Nov 4, 2009 at 15:44, Kay Sievers <kay.sievers@vrfy.org> wrote:
> On Wed, Nov 4, 2009 at 15:28, Marco d'Itri <md@linux.it> wrote:
>> On Nov 04, Kay Sievers <kay.sievers@vrfy.org> wrote:
>>
>>> > udev_rules_new() in udev/udev-rules.c unconditionally creates the
>>> > directory.
>>> > This is a problem because the function is called also by e.g. udevadm
>>> > test, and creating /dev/.udev/ when it does not exist is an unacceptable
>>> > side effect which will break everything else that checks for its
>>> > existence to know if udev is running.
>>> > Why is this needed?
>>> No specific reason, I remember. It could move into the daemon, before
>>> the inotify watch.
>> Why create it at all?
>> (I suspect that this was done to suppress a warning about the missing
>> directory, but then I'd rather do it properly with a flag...)
>
> It's to be able to subscribe to changes in that directory with
> inotify. We want to create it before we try to watch it, but only the
> daemon needs to do this.

I have changed this now, and only the daemon creates the directory,
right before it establishes the inotify watch.

Kay