linux-kernel - Re: [PATCH v4 01/13] USB: ch341: fix error handling on resume

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Date:	Mon, 2 May 2016 15:45:41 +0200
From:	Johan Hovold <johan@...nel.org>
To:	Grigori Goronzy <greg@...wn.ath.cx>
Cc:	Johan Hovold <johan@...nel.org>,
	Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
	linux-usb@...r.kernel.org, linux-kernel@...r.kernel.org,
	Johan Hovold <jhovold@...il.com>
Subject: Re: [PATCH v4 01/13] USB: ch341: fix error handling on resume

On Fri, Apr 29, 2016 at 05:11:13PM +0200, Grigori Goronzy wrote:
> On 2016-04-29 14:16, Johan Hovold wrote:
> > On Fri, Apr 15, 2016 at 11:14:04PM +0200, Grigori Goronzy wrote:
> >> This may fail, do not assume it always works.
> >> 
> >> Signed-off-by: Grigori Goronzy <greg@...wn.ath.cx>
> >> ---
> >>  drivers/usb/serial/ch341.c | 4 +---
> >>  1 file changed, 1 insertion(+), 3 deletions(-)
> >> 
> >> diff --git a/drivers/usb/serial/ch341.c b/drivers/usb/serial/ch341.c
> >> index c73808f..63df8ce 100644
> >> --- a/drivers/usb/serial/ch341.c
> >> +++ b/drivers/usb/serial/ch341.c
> >> @@ -544,9 +544,7 @@ static int ch341_reset_resume(struct usb_serial 
> >> *serial)
> >>  	priv = usb_get_serial_port_data(serial->port[0]);
> >> 
> >>  	/* reconfigure ch341 serial port after bus-reset */
> >> -	ch341_configure(serial->dev, priv);
> >> -
> >> -	return 0;
> >> +	return ch341_configure(serial->dev, priv);
> > 
> > This is correct, but have noticed that resume is currently broken in
> > that the interrupt urb is never resubmitted on resume in case the port 
> > is
> > already open?
> > 
> > Also ch341_configure must not use GFP_KERNEL either if called from a
> > resume path (use GFP_NOIO).
> > 
> > Care to fix this up as well?
> > 
> 
> Sure. How can I trigger a reset properly? AFAIR, I tried USBDEVFS_RESET 
> and it didn't really do what I wanted, at least the reset_resume 
> callback wasn't invoked.

You should be able to use the USB persistent-device functionality or
temporarily enable the USB_QUIRK_RESET_RESUME quirk for your device to
have reset_resume be called on system resume, but note that a resume
callback also needs to be implemented for normal (system and runtime)
resume.

Also note that reset_resume for usb-serial is currently broken in
mainline (since 4.5).

Thanks,
Johan