Date: Mon, 19 Jul 2004 16:33:55 -0600 (MDT) From: "M. Warner Losh" <imp@bsdimp.com> To: pjd@FreeBSD.ORG Cc: freebsd-current@FreeBSD.ORG Subject: Re: Unloading USB driver while device is attached. Message-ID: <20040719.163355.111547661.imp@bsdimp.com> In-Reply-To: <20040719145745.GH57678@darkness.comp.waw.pl> References: <20040719145745.GH57678@darkness.comp.waw.pl>
next in thread | previous in thread | raw e-mail | index | archive | help
In message: <20040719145745.GH57678@darkness.comp.waw.pl> Pawel Jakub Dawidek <pjd@freebsd.org> writes: : Hello. : : Could someone investigate what's going on for the situation from the : subject? This panic is trivial to reproduce: : : # kldload umass.ko : <plug in your USB bar> : # kldunload umass : # kldload umass : <panic> : : It is not umass related! I think this problem is in our USB implementation. : : One more note with could be helpful. : When device is already inserted and I load USB driver, <drvname>_match : isn't called, but when driver was unloaded while device was in and it is : loaded again (as in my example) <drvname>_match is called and we have : a problem, because there are some stale data, I think (not everything : is cleaned up on kldunload?). : Ok, one more note. : You can also set 'break <drvname>_match' in DDB and try to compare : backtraces when device is removed while driver is loaded and when you : unload USB driver. I do know that loading the driver won't attach umass unless you remove and insert it again. Warner
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20040719.163355.111547661.imp>