cvs commit: src/sys/i386/conf GENERIC src/sys/amd64/conf GENERIC

[Scott Long]

Ian Dowse wrote:
> In message <423AB122.6000508 at portaone.com>, Maxim Sobolev writes:
> 
>>Ian Dowse wrote:
>>
>>>In message <423A21E2.4080000 at FreeBSD.org>, Doug Barton writes:
>>>
>>>
>>>>In short, all of my "issues" seem to have been fixed! I also just updated my
>>
>>>>src tree to try the new code you committed to avoid the detach/reattach 
>>>>problem, and that not only works, but it avoids the weird errors I was 
>>>>getting about my docking station's USB hub and having to reprobe ums0.
>>>>
>>>>You and Julian are to be congratulated on all the hard work you've done in 
>>>>this area. Good luck with hunting down the rest of the corner cases.
>>>>
>>>>Doug
>>>
>>>
>>>Thanks Doug - that's great news - keep the panic reports and stack
>>>traces coming!
>>
>>Can you commit your re-probing fix, so that it gets wider exposure? It 
>>would be nice to have it included into 5.4, since this bug is really 
>>anoying and for example prevents us from turning on ehci on production 
>>machines that use USB boot devices.
> 
> 
> The re-probing fix is committed already - that's what Doug was
> referring to above.
> 
> Ian

There are still quite a few outstanding difference between HEAD and
RELENG_5 in the USB code now, yes?  I'd definitely like to get them
merged for 5.4 once you're comfortable.

Scott