public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
* TL16C752B DUART: MCR_OUT2 disables interrupts
@ 2006-06-20 20:07 Shaun Jackman
  2006-06-20 20:26 ` Paul Fulghum
  0 siblings, 1 reply; 4+ messages in thread
From: Shaun Jackman @ 2006-06-20 20:07 UTC (permalink / raw)
  To: LKML

The datasheet I have for the TL16C752B labels the MCR_OUT1 bit as
`FIFO Rdy enable' and the MCR_OUT2 bit as `IRQ enable'. The latter bit
concerns me. The 8250.c driver sets MCR_OUT2 by default; however, if
the user space clears MCR_OUT2 (through an ioctl TIOCMSET operation or
similar), it seems to me that interrupts for that UART will stop
working. Can someone confirm my suspicion?

I'd expect that clearing/setting the OUT1 and OUT2 pins from user
space should be an innocuous operation. Disabling interrupts is a
fairly nasty side effect.

Please cc me in your reply. Cheers,
Shaun

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: TL16C752B DUART: MCR_OUT2 disables interrupts
  2006-06-20 20:07 TL16C752B DUART: MCR_OUT2 disables interrupts Shaun Jackman
@ 2006-06-20 20:26 ` Paul Fulghum
  2006-06-20 20:35   ` Shaun Jackman
  0 siblings, 1 reply; 4+ messages in thread
From: Paul Fulghum @ 2006-06-20 20:26 UTC (permalink / raw)
  To: Shaun Jackman; +Cc: LKML

Shaun Jackman wrote:
> The datasheet I have for the TL16C752B labels the MCR_OUT1 bit as
> `FIFO Rdy enable' and the MCR_OUT2 bit as `IRQ enable'. The latter bit
> concerns me. The 8250.c driver sets MCR_OUT2 by default; however, if
> the user space clears MCR_OUT2 (through an ioctl TIOCMSET operation or
> similar), it seems to me that interrupts for that UART will stop
> working. Can someone confirm my suspicion?
> 
> I'd expect that clearing/setting the OUT1 and OUT2 pins from user
> space should be an innocuous operation. Disabling interrupts is a
> fairly nasty side effect.

The old 16550 OUT2 output (MCR:3) used to be a truly
independant control output that was commonly used
with an external gate to connect/disconnect the
interrupt request to the ISA bus.

Reading the TL16C752B datasheet, it seems pretty clear
that this gating is performed internal to the device.
MCR:3 is no longer an independant, general purpose I/O.

The result is the same though: If you dink around with
MCR:3 you could disable interrupts in either case.

So I don't see this as more or less of a problem than
other 16550 type devices.

-- 
Paul Fulghum
Microgate Systems, Ltd.

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: TL16C752B DUART: MCR_OUT2 disables interrupts
  2006-06-20 20:26 ` Paul Fulghum
@ 2006-06-20 20:35   ` Shaun Jackman
  2006-06-20 20:37     ` Paul Fulghum
  0 siblings, 1 reply; 4+ messages in thread
From: Shaun Jackman @ 2006-06-20 20:35 UTC (permalink / raw)
  To: Paul Fulghum; +Cc: LKML

On 6/20/06, Paul Fulghum <paulkf@microgate.com>
> The old 16550 OUT2 output (MCR:3) used to be a truly
> independant control output that was commonly used
> with an external gate to connect/disconnect the
> interrupt request to the ISA bus.
>
> Reading the TL16C752B datasheet, it seems pretty clear
> that this gating is performed internal to the device.
> MCR:3 is no longer an independant, general purpose I/O.
>
> The result is the same though: If you dink around with
> MCR:3 you could disable interrupts in either case.
>
> So I don't see this as more or less of a problem than
> other 16550 type devices.

I didn't know of this historic use of the OUT2 pin to disable
interrupts. I'll put OUT2 on my list of things not to muck with. Is
there any similar historic use of OUT1?

Cheers,
Shaun

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: TL16C752B DUART: MCR_OUT2 disables interrupts
  2006-06-20 20:35   ` Shaun Jackman
@ 2006-06-20 20:37     ` Paul Fulghum
  0 siblings, 0 replies; 4+ messages in thread
From: Paul Fulghum @ 2006-06-20 20:37 UTC (permalink / raw)
  To: Shaun Jackman; +Cc: LKML

Shaun Jackman wrote:
> On 6/20/06, Paul Fulghum <paulkf@microgate.com>
> I didn't know of this historic use of the OUT2 pin to disable
> interrupts. I'll put OUT2 on my list of things not to muck with. Is
> there any similar historic use of OUT1?

Not that I'm aware of. All of the old designs I've
seen left that signal unconnected.

-- 
Paul Fulghum
Microgate Systems, Ltd.

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2006-06-20 20:37 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2006-06-20 20:07 TL16C752B DUART: MCR_OUT2 disables interrupts Shaun Jackman
2006-06-20 20:26 ` Paul Fulghum
2006-06-20 20:35   ` Shaun Jackman
2006-06-20 20:37     ` Paul Fulghum

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox