Re: [PATCH v2 6/6] USB: serial: ch341: Simulate break condition if not supported

From: Johan Hovold <johan@kernel.org>
To: Michael Hanselmann <public@hansmi.ch>
Cc: Johan Hovold <johan@kernel.org>,
	linux-usb@vger.kernel.org, Michael Dreher <michael@5dot1.de>,
	Jonathan Olds <jontio@i4free.co.nz>
Subject: Re: [PATCH v2 6/6] USB: serial: ch341: Simulate break condition if not supported
Date: Mon, 6 Jul 2020 11:31:44 +0200	[thread overview]
Message-ID: <20200706093144.GG3334@localhost> (raw)
In-Reply-To: <f34a9b6e-ec2a-0873-e97b-2d5b2170e2ff@msgid.hansmi.ch>

Hi Michael,

On Sat, Jul 04, 2020 at 08:25:03PM +0200, Michael Hanselmann wrote:
> On 30.06.20 13:39, Johan Hovold wrote:
> > On Thu, May 28, 2020 at 12:21:11AM +0200, Michael Hanselmann wrote:

> Updated patch included below.
> 
> Michael
> 
> ---

When sending patches inline like this, try to avoid adding a (---)
marker like this as it makes git-am discard the commit message when
applying.

> From 41b8b06d343a69541a357d8c9d6d0fe3f22610d6 Mon Sep 17 00:00:00 2001
> Message-Id: <41b8b06d343a69541a357d8c9d6d0fe3f22610d6.1593887001.git.public@hansmi.ch>
> From: Michael Hanselmann <public@hansmi.ch>
> Date: Thu, 5 Mar 2020 01:50:35 +0100
> Subject: [PATCH] USB: serial: ch341: Simulate break condition if not supported
> 
> A subset of all CH341 devices don't support a real break condition. This
> fact is already used in the "ch341_detect_quirks" function. With this
> change a quirk is implemented to simulate a break condition by
> temporarily lowering the baud rate and sending a NUL byte.
> 
> The primary drawbacks of this approach are that the duration of the
> break can't be controlled by userland and that data incoming during
> a simulated break is corrupted.
> 
> The "TTY_DRIVER_HARDWARE_BREAK" serial driver flag was investigated as
> an alternative. It's a driver-wide flag and would've required
> significant changes to the serial and USB-serial driver frameworks to
> expose it for individual USB-serial adapters.
> 
> Tested by sending a break condition and watching the TX pin using an
> oscilloscope.
> 
> Signed-off-by: Michael Hanselmann <public@hansmi.ch>

Now applied with a slightly condensed probe info message.

Thanks,
Johan