Re: More USB_QUIRK_RESET_RESUME devices

On Tue, 11 Sep 2007 [email protected] wrote:

> Sorry, I was thinking about usb persist and USB_QUIRK_RESET_RESUME together.
> Anyway if /sys/block/sdX/removable is set incorrectly sometimes then it will not
> help for usb persist.

Maybe...  But remember that USB persist applies to _all_ USB devices,
whereas only USB disk-type devices will have entries in
/sys/block/sdX/removable.

> Another question, ClassmatePC uses Ralink's rt73 usb wireless card. I asked
> Ralink's engineers to implement .suspend() and .resume() callbacks in their
> driver. They did it but they have one doubt that I could not answer:
> 
> "We have several success trial for the suspend and resume now. One thing I don't
> quite understand is the resume process of USB.
> When I doing the suspend to RAM, the suspend call back function is called and
> work well.
> When resume, usb subsystem will call disconnect call back (remove module) then
> call the close call back function. Seem unreasonable. Since resume and our
> driver in the kernel space why remove the module and do the probe again? Any
> setting we can configure for the CMPC?"

Tell them to try testing by doing

	echo suspend >/sys/bus/usb/device/.../power/level

and

	echo auto >/sys/bus/usb/device/.../power/level

where "..." should be replaced with the device ID of the wireless card.  
If that works then the driver probably implements suspend and resume 
correctly.

> I found this mailling list thread
> http://sourceforge.net/mailarchive/message.php?msg_id=1188828970.8195.39.camel%40localhost
> where you answered to someone that has a similar problem.

In that thread it looked like the problem was caused by a bug in the
BIOS or the motherboard's ACPI implementation.  The people found that
things worked better on a different computer.

> In this thread
> http://rt2x00.serialmonkey.com/phpBB2/viewtopic.php?p=21976&sid=14bd99cbf92c3d7d90838d1ba3e06ae9
> someone suggest that it is a kernel bug. Do you think it is a kernel?

I do not think it is a bug in the kernel.  However, in that thread 
people mentioned that attempts to do "rmmod rt73" would hang; that 
could be caused by a bug in the rt73 driver or by a bug in the EHCI 
controller hardware.

> Do you
> know where Ralink can get documentation about Linux USB subsystem in order to
> make it work with suspend?

The easiest way for them to get answers is to post their questions on 
linux-usb-devel.

Alan Stern


-------------------------------------------------------------------------
This SF.net email is sponsored by: Microsoft
Defy all challenges. Microsoft(R) Visual Studio 2005.
http://clk.atdmt.com/MRT/go/vse0120000070mrt/direct/01/
_______________________________________________
[email protected]
To unsubscribe, use the last form field at:
https://lists.sourceforge.net/lists/listinfo/linux-usb-devel