OK, I think I've figured out my problem here. I was using the wrong UHCI kernel
driver. I hadn't know that
there were two of them. I was using the JE driver for some reason. I switched
to the other one and so far, no
problems. I've copied a few gigs now without any freezing. I am a happy dood.
I don't know if it's worth debugging the problem with the JE driver but I'm
happy to help if necessary. I think
everyone could duplicate the problem by using the JE driver.
Björn, I appreciate the time you spent helping me out with this. Thank you. You
might want to mention this
issue on your web page.
Curtis
Björn Stenberg wrote:
> Curtis Stanford wrote:
>
> > Here's something more interesting. I tried the same sequence of events on
> > my other computer (laptop) and managed to copy over 75M before it hanging.
> > I actually thought it was going to work and than, bang, it hangs after 77M
> > or so.
>
> Ok, time to break out the logs. Enable "USB Mass Storage verbose debug" in
> your kernel and see what is written to the log when the device access hangs.
>
> The actual amount of data will vary, simply because linux buffers very large
> amounts of data before anything actually gets written to the device. I don't
> know the algorithms defining how much data that can be, but I've seen
> instances of over 50 megs on my machine so 75 doesn't sound implausible.
>
> /Björn
âžìn±»¬z»%ŠËl²‹«qçè®§zÔèº{.nÇ+‰·®±ëazV¬µú+™øž•Ö†Ûiÿùb²Û,¢êÜyú+éÞ·ùb²Û?–+-ŠwèþX§»¬në®