linux-kernel - Re: [PATCH 1/2] tty/hvc_console: Add DTR/RTS callback to handle HUPCL control

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:	Thu, 17 Oct 2013 10:16:42 +0200
From:	Hendrik Brueckner <brueckner@...ux.vnet.ibm.com>
To:	Benjamin Herrenschmidt <benh@....ibm.com>
Cc:	Hendrik Brueckner <brueckner@...ux.vnet.ibm.com>, jslaby@...e.cz,
	gregkh@...uxfoundation.org, brueckner@...ux.vnet.com,
	heiko.carstens@...ibm.com, schwidefsky@...ibm.com,
	linuxppc-dev@...ts.ozlabs.org, linux-kernel@...r.kernel.org,
	linux-s390@...r.kernel.org
Subject: Re: [PATCH 1/2] tty/hvc_console: Add DTR/RTS callback to handle
 HUPCL control

On Wed, Oct 16, 2013 at 06:21:12PM -0500, Benjamin Herrenschmidt wrote:
> On Wed, 2013-10-16 at 11:04 +0200, Hendrik Brueckner wrote:
> > Indeed, two callbacks change the DTR line.  The main difference is that
> > tiocmget/tiocmset can be called from user space by ioctl.  That's not the case
> > for the dtr_cts callback.  Also, tiocmget/tiocmset provide more flags that can
> > be changed (ST, SR, CTS, CD, RNG, RI,  ...)
> > 
> > Assume we would like to unify them have a single callback to change DTR, then
> > we have to take care of these differences.  So the question to you now is
> > whether you plan for a) other modem flags to be changed and b) if changing the
> > DTR line (or other control flags) through an ioctl?
> > 
> > Depending on your results, I could work on sth that helps us both and reduces
> > the callbacks.
> 
> Can we not just have the users of dtr_cts just call the backend's tiocmset ?

That's possible.  The only concern is that the tiocmset() callback could be
triggered from within the hvc_console() layer as well as from user space via
ioctl.  For the hvc_iucv driver, I do not want to introduce this ioctl.

One option would be to add parameter to the hvc_callbacks that indicate the
origin so that a backend could filter.

> If they need to filter or clamp bits, we could handle all that in hvc_console
> by caching the user intended value and passing a cooked value down to the backend..

Sure the hvc_console layer could do as much as possible.

> 
> None of that is urgent or anything, it's just odd and would be nice to cleanup.

Thanks and kind regards,
  Hendrik

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/