Re: g_file_storage speed issue

[jpollack@xxxxxxxxx]

Okay.  I'm not really familiar with linux kernel code, or inner
details of USB, but I'd like to keep hacking with this.  I think I
might have a vague idea what's going on.  I don't know what you're
referring to by "bottom half", could you elaborate a bit?

Is dummy_timer the timer callback function?

Also, do you know offhand a good resource for documentation for linux
kernel functions?  Things like 'mod_timer' I can try to infer their
meaning from header files, but that game isn't any fun.  I would have
guessed it would be in the "linux-manual" man pages but no such luck.

Thanks,
Joshua

On 3/10/07, Alan Stern <stern@xxxxxxxxxxxxxxxxxxx> wrote:
> The real problem is that dummy-hcd operates from a kernel timer, meant to
> simulate the frames of a real USB bus.  Since you want to maximize the
> transfer rate, you shouldn't use a kernel timer at all; you should use
> sort sort of bottom half instead.
>
> Making the change shouldn't be too difficult, but it isn't entirely
> trivial either.  Maybe you can figure out how to do it.  The idea is to
> change the "kick the scheduler" part at the end of dummy_urb_enqueue() so
> that it enables the bottom half instead of modifying the timer (with
> appropriate adjustments in other places, of course).
>
> Alan Stern
>
>

-------------------------------------------------------------------------
Take Surveys. Earn Cash. Influence the Future of IT
Join SourceForge.net's Techsay panel and you'll get the chance to share your
opinions on IT & business topics through brief surveys-and earn cash
http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________
Linux-usb-users@xxxxxxxxxxxxxxxxxxxxx
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-users