Hello,
I am the person who received the patch. Sorry, but I still have the hang.
The last entries in my syslog:
Jul 16 16:56:16 (none) kernel: pwc Image is mirrored.
Jul 16 16:56:16 (none) kernel: pwc Frame buffer underflow (2 bytes);
discarded.
Jul 16 16:56:16 (none) kernel: pwc Image is normal.
Jul 16 17:03:21 (none) kernel: pwc Frame buffer overflow (flen = 582,
frame_size = 115212).
Jul 16 17:03:30 (none) last message repeated 5 times
Jul 16 17:04:17 (none) last message repeated 7 times
Jul 16 17:07:33 (none) kernel: pwc Frame buffer underflow (3 bytes);
discarded.
Jul 16 17:10:39 (none) kernel: pwc Frame buffer overflow (flen = 582,
frame_size = 115212).
Jul 16 17:10:54 (none) last message repeated 3 times
Jul 16 17:10:59 (none) kernel: pwc Frame buffer overflow (flen = 570,
frame_size = 115212).
Jul 16 17:11:03 (none) last message repeated 3 times
Jul 16 17:11:04 (none) kernel: pwc Frame buffer overflow (flen = 582,
frame_size = 115212).
Jul 16 17:12:08 (none) kernel: pwc Frame buffer overflow (flen = 582,
frame_size = 115212).
Jul 16 17:12:19 (none) last message repeated 15 times
Jul 16 17:12:19 (none) kernel: pwc Frame buffer overflow (flen = 570,
frame_size = 115212).
Jul 16 17:12:21 (none) last message repeated 2 times
Jul 16 17:12:21 (none) kernel: pwc Frame buffer overflow (flen = 582,
frame_size = 115212).
Jul 16 17:12:26 (none) last message repeated 8 times
Jul 16 17:13:49 (none) kernel: pwc Frame buffer underflow (1 bytes);
discarded.
Jul 16 17:26:01 (none) -- MARK --
Jul 16 17:46:01 (none) -- MARK --
It worked about 25 Minutes.
But it does not seem, that the module is down, later I receive again a
message from pwc
Jul 16 18:58:25 (none) kernel: pwc Frame buffer underflow (3 bytes);
discarded.
Jul 16 19:26:02 (none) -- MARK --
but it does not send the frame received anymore to the Java Media Framework.
Can I compile with more messages or is there any other tip how to debug.
Detlef
----- Original Message -----
From: "Oliver Neukum" <[EMAIL PROTECTED]>
To: "Nemosoft Unv." <[EMAIL PROTECTED]>
Cc: <[EMAIL PROTECTED]>
Sent: Thursday, July 17, 2003 7:59 AM
Subject: Re: [linux-usb-devel] pwc stops if the system is loaded
> Am Donnerstag, 17. Juli 2003 01:50 schrieb Nemosoft Unv.:
> > Hello,
> >
> > On Tuesday 15 July 2003 10:43, Oliver Neukum wrote:
> > > > Particularly after taking a quick glance at the code
> > > > and not seeing anything that much cared about nonzero
> > > > fault codes when the urbs are returned.
> > >
> > > Hi nemosoft,
> > >
> > > looking at the 2.5 code error handling seems fishy:
> > >
> > > if (urb->status != -EINPROGRESS && urb->status != 0) {
> > > char *errmsg;
> > >
> > > errmsg = "Unknown";
> > > switch(urb->status) {
> > > case -ENOSR: errmsg = "Buffer error (overrun)"; break;
> > > case -EPIPE: errmsg = "Stalled (device not responding)"; break;
> > > case -EOVERFLOW: errmsg = "Babble (bad cable?)"; break;
> > > case -EPROTO: errmsg = "Bit-stuff error (bad cable?)"; break;
> > > case -EILSEQ: errmsg = "CRC/Timeout"; break;
> > > case -ETIMEDOUT: errmsg = "NAK (device does not respond)"; break;
> > > }
> > > Trace(TRACE_FLOW, "pwc_isoc_handler() called with status %d [%s].\n",
> > > urb->status, errmsg); return;
> > > }
> > >
> > > Judging by this code any such error will mean that the URB is not
> > > resubmitted. Can this be right?
> >
> > Well, it's not resubmitted, period. Even at the end of
pwc_isoc_handler() no
> > resubmission is done, at least on 2.4.20. And that seems to work okay.
>
> For 2.4 this is correct.
>
> > > Shouldn't there be a goto to this piece of code ?
> >
> > Goto??!?! Argh! Yuck! Blegh! :-)
>
> Proper use of 'goto' shows the true master of the art.
>
> > > urb->dev = pdev->udev;
> > > i = usb_submit_urb(urb, GFP_ATOMIC);
> > > if (i != 0)
> > > Err("Error (%d) re-submitting urb in pwc_isoc_handler.\n", i);
> >
> > This is for 2.5, right... So that means the URB submission system got
> > changed back again (Duh...) [I haven't really kept up with all the
changes
> > in 2.5, only checking if the changes made by others don't break my code.
> > But yes, a resubmission would probably help in this case.
>
> I sent one of the people reporting a hang such a patch.
> We'll see.
>
> Regards
> Oliver
>
>
>
> -------------------------------------------------------
> This SF.net email is sponsored by: VM Ware
> With VMware you can run multiple operating systems on a single machine.
> WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the
> same time. Free trial click here: http://www.vmware.com/wl/offer/345/0
> _______________________________________________
> [EMAIL PROTECTED]
> To unsubscribe, use the last form field at:
> https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
>
-------------------------------------------------------
This SF.net email is sponsored by: VM Ware
With VMware you can run multiple operating systems on a single machine.
WITHOUT REBOOTING! Mix Linux / Windows / Novell virtual machines at the
same time. Free trial click here: http://www.vmware.com/wl/offer/345/0
_______________________________________________
[EMAIL PROTECTED]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel
