Date: Wed, 28 Sep 2016 09:08:48 -0700 From: John Baldwin <jhb@freebsd.org> To: Hans Petter Selasky <hps@selasky.org> Cc: Jan Henrik Sylvester <me@janh.de>, FreeBSD-USB Mailing List <freebsd-usb@freebsd.org> Subject: Re: Status of PCIe Hotplug? Message-ID: <1635480.htdnZkDgDb@ralph.baldwin.cx> In-Reply-To: <b192f906-7e2e-339c-675c-5d6a01e6b7cd@selasky.org> References: <D7E110DB-EF6D-41DD-8D69-A021422916FA@sarenet.es> <3c6ab189-aad0-2593-1d3c-f221de294174@janh.de> <b192f906-7e2e-339c-675c-5d6a01e6b7cd@selasky.org>
next in thread | previous in thread | raw e-mail | index | archive | help
On Wednesday, September 28, 2016 11:55:47 AM Hans Petter Selasky wrote: > On 09/28/16 11:31, Jan Henrik Sylvester wrote: > > On 09/28/2016 10:32, Hans Petter Selasky wrote: > >> Can you compile your kernel with debugging enabled so that we get the > >> sourcelines in the dump below. Also please include some parts of the > > > > The kernel is from the yet-to-be-released 11.0-RELEASE installed by > > freebsd-update. Now I have extracted kernel-dbg.txz, base-dbg.txz, and > > lib32-dbg.txz from the 11.0-RELEAESE distribution, but the backtrace > > still looks similar. Do have have to do anything else to use the debug > > symbols or do I really have to compile a new kernel? > > > > Hi, > > Can you try the attached patch. Does it make any difference? > > device_delete_child() will delete devices starting at the leaves, while > USB expects parents to be detached first. > > jhb: Any comments? This is fine, though I feel like we should fix device_delete_child instead as this has broken several places. That is, I think device_delete_child should do the detach first and then delete any dangling grandchildren after device_detach() of the child. -- John Baldwin
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?1635480.htdnZkDgDb>