Am Mittwoch, 23. Oktober 2002 07:53 schrieb John Tyner:
> > 1. Who can write these settings? How do you control that?
>
> I set permissions to rw-r--r--. Don't these files get owned by root when
> created?
Yes, that's the problem. Consequently device permissions themselves would
have to be 644, which is unacceptable.
Why should shutter speed setting be different from, eg. frame size?
They should be settable by the same method.
Secondly, vicam like most V4L devices, implements exclusive open.
A proc or driverfs file to change a setting violates that model.
> > 2. The module count handling is wrong.
> > These methods don't sleep. Either they are racy or safe with or
> > without manipulating the usage count.
>
> I saw something in the Documentation directory that mentioned something
> about needing to increase the module count so that the module is not
> unloaded while these functions are in use.
That is an issue. But you are not sleeping. Either you can safely handle
the module usage count yourself, which means you are called under BKL,
or you are in danger anyway.
Actually it seems to me that driverfs is buggy in that respect. A patch
will come shortly.
> > 3. vicam_shutter_store races with read_frame.
>
> read_frame races with just about everything at the moment. One thing at a
> time.
Very well, but you should be aware of a few things when you consider
locking changes. An entry point into a driver through driverfs must
be treated like open() in terms of locking. To be precise, you must treat
it like a generic char device open. That means that the alternate driver's
method is invalid, because you cannot assume that video_unregister()
will prevent any further opening already being in progress.
Regards
Oliver
-------------------------------------------------------
This sf.net emial is sponsored by: Influence the future
of Java(TM) technology. Join the Java Community
Process(SM) (JCP(SM)) program now.
http://ads.sourceforge.net/cgi-bin/redirect.pl?sunm0002en
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
