Date: Tue, 9 May 2006 05:30:24 GMT From: Anish Mistry <amistry@am-productions.biz> To: freebsd-usb@FreeBSD.org Subject: Re: usb/93949 : ugen(4)-related repeatable kernel panic in 6.1-PRERELEASE Message-ID: <200605090530.k495UOpJ082900@freefall.freebsd.org>
next in thread | raw e-mail | index | archive | help
The following reply was made to PR usb/93949; it has been noted by GNATS. From: Anish Mistry <amistry@am-productions.biz> To: Eugene Grosbein <eugen@grosbein.pp.ru> Cc: bug-followup@freebsd.org Subject: Re: usb/93949 : ugen(4)-related repeatable kernel panic in 6.1-PRERELEASE Date: Tue, 9 May 2006 01:20:26 -0400 --nextPart5356158.N9tKWMFHiX Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: quoted-printable Content-Disposition: inline Alright. I've been able to reproduce this panic. The problem is that=20 the thread that called select() is gone (since the device is=20 detached) when clear_selinfo_list is called in kern_select(). The=20 easiest way to reproduce this is to plug in a usb mouse and have it=20 attached as ugen and then have a small program call select() on the=20 0.1 endpoint with a long timeout (ie. 10 seconds). While the select=20 call is waiting pull the plug on the device and when the operation=20 times out you'll get the panic. I'm not sure how to handle this case. It seems that=20 clear_selinfo_list should be able to handle this case. =2D-=20 Anish Mistry amistry@am-productions.biz AM Productions http://am-productions.biz/ --nextPart5356158.N9tKWMFHiX Content-Type: application/pgp-signature -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.3 (FreeBSD) iD8DBQBEYCajxqA5ziudZT0RAl0IAKCOwwuHgsePVk8ylVUgh3zsbycYYACfeNQ5 eDXgrF7NNC6i/c1PwpJG+7A= =kpiQ -----END PGP SIGNATURE----- --nextPart5356158.N9tKWMFHiX--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200605090530.k495UOpJ082900>