Hi Alan !
>From user reports, it seems that other model of Huawei modems are still
broken in a way that seems very similar to what we observed and fixed
with the E169 but not fixed by the request sense size fix we did.
IE. The initial fix that "enabled" usb-storage to work in 2.6.31.1 broke
them causing them to get into a reset loop due to some issues with
usb-storage.
I've obtained a usbmon log from one of those users with a E620, I would
appreciate if you could have a look as you are a lot better than I am at
parsing those things :-) I attached it to the existing BZ
http://bugzilla.kernel.org/show_bug.cgi?id=14355 .
Of course it's always possible that a whole lot of users have been
testing the wrong kernel and thus don't have the fix :-) But that might
also show a different problem with those modems usb storage
implementation.
Cheers,
Ben.
On Fri, 2009-11-06 at 11:55 +1100, Benjamin Herrenschmidt wrote:
> Of course it's always possible that a whole lot of users have been
> testing the wrong kernel and thus don't have the fix :-) But that might
> also show a different problem with those modems usb storage
> implementation.
And indeed, the user -does- have the wrong kernel...
Ben.
On Fri, 6 Nov 2009, Benjamin Herrenschmidt wrote:
> On Fri, 2009-11-06 at 11:55 +1100, Benjamin Herrenschmidt wrote:
>
> > Of course it's always possible that a whole lot of users have been
> > testing the wrong kernel and thus don't have the fix :-) But that might
> > also show a different problem with those modems usb storage
> > implementation.
>
> And indeed, the user -does- have the wrong kernel...
And indeed, the log shows exactly the same problem as your modem had.
So presumably the same code change will fix it.
Alan Stern