Re: svn commit: r329458 - head/sbin/devmatch

2018-02-20 Thread John Baldwin
On Saturday, February 17, 2018 08:12:22 AM Warner Losh wrote: > On Sat, Feb 17, 2018 at 8:04 AM, Hans Petter Selasky > wrote: > > > On 02/17/18 15:56, Warner Losh wrote: > > > >> The right fix there, I think, is to load them all at once, in one kldload > >> operation and not loop in /etc/rc.d/dev

Re: svn commit: r329458 - head/sbin/devmatch

2018-02-17 Thread Hans Petter Selasky
On 02/17/18 16:12, Warner Losh wrote: Yes. I know that. Continuing to belabor the obvious won't help make me less grumpy. I'm sorry about that. Implementing a generic solution to match drivers and devices seems more complicated than at first thought :-) OK, I will say away from *devmatch* a

Re: svn commit: r329458 - head/sbin/devmatch

2018-02-17 Thread Hans Petter Selasky
On 02/17/18 16:07, Warner Losh wrote: It would be more helpful if you could turn on devd's logging (by adding !devd *.>=info /var/log/devd.log to /etc/syslog.conf, restarting syslog and then giving me the exact scenario that's failing so I can add it to a tes

Re: svn commit: r329458 - head/sbin/devmatch

2018-02-17 Thread Warner Losh
On Sat, Feb 17, 2018 at 8:04 AM, Hans Petter Selasky wrote: > On 02/17/18 15:56, Warner Losh wrote: > >> The right fix there, I think, is to load them all at once, in one kldload >> operation and not loop in /etc/rc.d/devmatch. >> > > Each driver will invoke the driver loaded device method, so th

Re: svn commit: r329458 - head/sbin/devmatch

2018-02-17 Thread Warner Losh
On Sat, Feb 17, 2018 at 7:58 AM, Hans Petter Selasky wrote: > On 02/17/18 15:56, Warner Losh wrote: > >> I'd request you make no further changes to devmatch without a review. >> > > I don't have a problem with that. > > You can see now what changes are needed. Please make sure you test with > rea

Re: svn commit: r329458 - head/sbin/devmatch

2018-02-17 Thread Hans Petter Selasky
On 02/17/18 15:56, Warner Losh wrote: The right fix there, I think, is to load them all at once, in one kldload operation and not loop in /etc/rc.d/devmatch. Each driver will invoke the driver loaded device method, so this will race aswell! You have a point. But, that means devmatch should r

Re: svn commit: r329458 - head/sbin/devmatch

2018-02-17 Thread Hans Petter Selasky
On 02/17/18 15:56, Warner Losh wrote: I'd request you make no further changes to devmatch without a review. I don't have a problem with that. You can see now what changes are needed. Please make sure you test with real USB devices! Some people are using FreeBSD-current on desktops, and the U

Re: svn commit: r329458 - head/sbin/devmatch

2018-02-17 Thread Warner Losh
On Sat, Feb 17, 2018 at 7:34 AM, Hans Petter Selasky wrote: > Author: hselasky > Date: Sat Feb 17 14:34:47 2018 > New Revision: 329458 > URL: https://svnweb.freebsd.org/changeset/base/329458 > > Log: > Fix USB driver matching in devmatch(8). > > Multiple drivers can match on the same USB devi