USB OHCI problems...

[Julian Elischer]

Barry Bouwsma wrote:

>Julian Elischer asked of me...
>
>  
>
>>>my present 4.x USB stack comes from -current as it is.  Many thanks!
>>>      
>>>
>
>  
>
>>I'm intrigued..
>>    
>>
>
>with good reason, i may add
>
>  
>
>>why do you have a 4.x system with a 5.x USB stack?
>>    
>>
>
>Hysterical raisins, mostly, if you ask me.
>
>Some many months back, Ian Dowse announced his P4 work as a patchset
>to -current, which sounded like it addressed many of the USB problems
>I had been experiencing.
>
>So, I took that and merged it as best I could with my system, borrowing
>from -current when needed.
>
>Sometime around this time, you also did work to bring the 4.x USB
>code more in line with current.
>

well, some of it..

>
>Since your mega-merge, I've seen a few commits to -current USB while
>4.x code has gone mostly untouched.  As I discover more problems
>(mostly OHCI-related), I figure the best thing is for me to try out
>the latest code, as due to my mostly-offline status, I can only
>make large incremental updates, that are not so easy to keep track
>of -- then I can see if the problem is solved there.
>

I can only really concentrate on USB (and usb in 4.x) in short bursts as 
work allows.
I'm currently on that again, looking for some bugs (in but -current and 4.x)

>
>Also, FWIW, it appears that my ehci.c that I compiled was taken
>directly from NetBSD, as I had still experienced data-toggle
>problems with all FBSD source of that time, while they were less
>frequent (though occasionally present) with NetBSD.
>
at WHAT time?

>
>My last source update was made late October, before the latest
>recent updates.  At that time, I didn't notice any 4.x updates
>of interest since May or so.  I'm going to see if I can solve my
>known USB/OHCI issues with the code I have, then if so, I'll set
>this aside and try out clean 4.x source, in order to try and reduce
>the differences between my source and that released, but without
>having to suffer through bugs that may still affect me in original
>4.x code.
>

Is there any possibility that instead of just doing 5.x stacks on 4.x I 
can convince you
to try make that the defautl case (in other words help with MFC-ing 
current code back into 4.x?)
:-)

>
>And in particular, the if_axe code I was trying to get working some
>weeks/months ago was only present in -current, and part of what I
>tried in order to avoid the panics I saw was to try and use as much
>-current code as possible to minimize differences.
>
>Oh, and most importantly, because I don't have the foggiest idea
>what I'm doing, so doing things all wrong is a good way to help me
>learn.
>
>
>next question?
>
>barry bouwsma
>(sheesh, ask a quick question, get a novel in response)
>
>_______________________________________________
>freebsd-usb at freebsd.org mailing list
>http://lists.freebsd.org/mailman/listinfo/freebsd-usb
>To unsubscribe, send any mail to "freebsd-usb-unsubscribe at freebsd.org"
>  
>