RE: Fwd: LiS-2.18 and driver open failure due to q->q_ptr being not Null.

 Hi Brian,

Thanks for the reponse.

This works fine now on 2.4 kernel with LiS-2.18 after adding qprocson(q)
& qprocsoff(q) in open() and close() respectively but not with 2.6
kernel.

Close() takes care of setting q_ptr to Null as you pointed out.

I am using simple test program just to open the device and this is the
first open()that fails, which shouldn't have to do anything with the
close().

I am manually doing mknod /dev/xyz c clone_drvr xyz_drvr. Is this
changed in 2.6? Since it enters my entry point I am assuming this should
be okay.

I will try with simple driver, if still an issue post the code.


Ed



-----Original Message-----
From: Brian F. G. Bidulock [mailto:[email protected]] 
Sent: Wednesday, February 02, 2005 12:43 PM
To: Rodrigues, Edward
Cc: LiS Mailing List
Subject: Re: [Linux-streams] Fwd: LiS-2.18 and driver open failure due
to q->q_ptr being not Null.

Edward,

The only reason I can think that that would happen is if you forgot
to do:

	q->q_ptr = WR(q)->q_ptr = NULL;

in your qclose(9) procedure.


Otherwise, I think you will have to share your qopen(9) and qclose(9)
procedure
to help you further.

--brian

On Tue, 01 Feb 2005, Dave Grothe wrote:

> 
>      Subject: LiS-2.18 and driver open failure due to q->q_ptr being
not
>      Null.
>      Date: Tue, 1 Feb 2005 13:47:23 -0500
>      From: "Rodrigues, Edward" <[email protected]>
>      To: <[email protected]>
>      Hi Dave,
>      I  was  trying  LiS-2.18 and 2.6 kernel with our drivers and
bumped
>      into
>      this  problem  where  my open fails because q->q_ptr is not null.
I
>      traced
>      this in lis_qopen() -> clone_open() and drvr_open(). It is fine
in
>      lis_qopen()  as  well  as clone_open() when it gets to the driver
I
>      have
>      wrong q.
>      I  tried  the  same with 2.4 kernel and ran into same issue. I
have
>      made
>      sure I use sys/LiS/module.h. I am sure I must be missing
something
>      basic.
>      Any help certainly appreciated.
>      Thanks
>      Ed

-- 
Brian F. G. Bidulock    | The reasonable man adapts himself to the |
[email protected]    | world; the unreasonable one persists in  |
http://www.openss7.org/ | trying  to adapt the  world  to himself. |
                        | Therefore  all  progress  depends on the |
                        | unreasonable man. -- George Bernard Shaw |