Timeout broken with serial port

Timeout broken with serial port

[Michal Suchanek]

Hello,

I was experiencing a mysterious error with grub. It would not boot
with timeout set.

It turns out that setting serial IO disables timeout.

grub-pc        1.99-11ubuntu1

I think it used to work with earlier grub, I have this file for ages:

# cat /etc/grub.d/03_serial
#!/bin/sh

echo serial --unit=0 --speed=115200
echo terminal_input --append serial\; terminal_output --append serial

Re: Timeout broken with serial port

[Vladimir 'φ-coder/phcoder' Serbinenko]

[-- Attachment #1: Type: text/plain, Size: 805 bytes --]

On 24.08.2011 11:16, Michal Suchanek wrote:
> Hello,
>
> I was experiencing a mysterious error with grub. It would not boot
> with timeout set.
>
> It turns out that setting serial IO disables timeout.
Check your contacts. Bad contacts are known to send noise on I/O which
GRUB interprets as keypresses.
> grub-pc        1.99-11ubuntu1
>
> I think it used to work with earlier grub, I have this file for ages:
>
> # cat /etc/grub.d/03_serial
> #!/bin/sh
>
> echo serial --unit=0 --speed=115200
> echo terminal_input --append serial\; terminal_output --append serial
>
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/grub-devel
>


-- 
Regards
Vladimir 'φ-coder/phcoder' Serbinenko



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 294 bytes --]

Re: Timeout broken with serial port

[Michal Suchanek]

2011/8/24 Vladimir 'φ-coder/phcoder' Serbinenko <phcoder@gmail.com>:
> On 24.08.2011 11:16, Michal Suchanek wrote:
>> Hello,
>>
>> I was experiencing a mysterious error with grub. It would not boot
>> with timeout set.
>>
>> It turns out that setting serial IO disables timeout.
> Check your contacts. Bad contacts are known to send noise on I/O which
> GRUB interprets as keypresses.

I did not have the cable connected at all at the other end. Connecting
it to my notebook resolves the problem.

Still it is definitely a bug in Grub that it interprets noise as keypresses.

Is there not DTR line?

If every switch with serial console failed to boot when there is
nothing connected to it there would be no internet. This is not how a
serial enabled device should behave.

Thanks

Michal

Re: Timeout broken with serial port

[Aleš Nesrsta]

Hi,

I think it is not GRUB related problem, more probably there is some HW
problem on Your serial port. Try to check idle voltage on RxD pin of
serial port (best with oscilloscope... :-) ). Or You can have some
unwanted "leakage" between pins of connector or wires in cable (check
resistance between wires of cable) etc.

Normal serial port should NEVER do anything if there is nothing
connected to it. According to RS232 (V24) specification there should be
voltage in range from -12V to -3V on RxD pin in idle state (in simplest
case there is some internal pull-up resistor connected from RxD to -Vcc
directly inside UART or something else...).
In this normal case nothing is received by serial port and nothing will
be interpreted as keypress in GRUB, i.e. GRUB will boot normally.

In Your case, probably something causes some noise on RxD pin or there
is bad voltage level on RxD pin (e.g. >= +3V or near to 0V etc.) when
nothing is connected - so serial port interprets it as receiving some
character(s).
When You connect Your serial port to another device (notebook in Your
case), RxD is set to correct idle voltage from connected idle TxD, so
the problem is corrected.

From my long praxis with serial ports on different PCs and another
(non-PC) devices, manufacturers very very often don't respect RS232
(V24) specification (mainly for RxD/TxD signals voltage levels). It
often caused bad situations, some of them were in fact the same as Your
case.

Using of DTR line is fine but if You have some HW problem on RxD line,
there could be the same problem also on DTR (RTS, CTS ...) line and
final effect will be the same...
Additionally, lot of devices are working without using any additional
link/modem state signals (i.e., only signals RxD, TxD and GND are used),
it is simplest and almost working solution (and working well - of course
it depends of kind of communication and it also expects that RS232
specification is not violated and all HW is in order...).

Best regards
Ales
 

Michal Suchanek wrote:
> 2011/8/24 Vladimir 'φ-coder/phcoder' Serbinenko <phcoder@gmail.com>:
> > On 24.08.2011 11:16, Michal Suchanek wrote:
> >> Hello,
> >>
> >> I was experiencing a mysterious error with grub. It would not boot
> >> with timeout set.
> >>
> >> It turns out that setting serial IO disables timeout.
> > Check your contacts. Bad contacts are known to send noise on I/O which
> > GRUB interprets as keypresses.
> 
> I did not have the cable connected at all at the other end. Connecting
> it to my notebook resolves the problem.
> 
> Still it is definitely a bug in Grub that it interprets noise as keypresses.
> 
> Is there not DTR line?
> 
> If every switch with serial console failed to boot when there is
> nothing connected to it there would be no internet. This is not how a
> serial enabled device should behave.
> 
> Thanks
> 
> Michal
> 
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/grub-devel

Re: Timeout broken with serial port

[Michal Suchanek]

On 5 September 2011 22:51, Aleš Nesrsta <starous@volny.cz> wrote:
> Hi,
>
> I think it is not GRUB related problem, more probably there is some HW
> problem on Your serial port. Try to check idle voltage on RxD pin of
> serial port (best with oscilloscope... :-) ). Or You can have some
> unwanted "leakage" between pins of connector or wires in cable (check
> resistance between wires of cable) etc.
>
> Normal serial port should NEVER do anything if there is nothing
> connected to it. According to RS232 (V24) specification there should be
> voltage in range from -12V to -3V on RxD pin in idle state (in simplest
> case there is some internal pull-up resistor connected from RxD to -Vcc
> directly inside UART or something else...).
> In this normal case nothing is received by serial port and nothing will
> be interpreted as keypress in GRUB, i.e. GRUB will boot normally.
>
> In Your case, probably something causes some noise on RxD pin or there
> is bad voltage level on RxD pin (e.g. >= +3V or near to 0V etc.) when
> nothing is connected - so serial port interprets it as receiving some
> character(s).
> When You connect Your serial port to another device (notebook in Your
> case), RxD is set to correct idle voltage from connected idle TxD, so
> the problem is corrected.
>
> From my long praxis with serial ports on different PCs and another
> (non-PC) devices, manufacturers very very often don't respect RS232
> (V24) specification (mainly for RxD/TxD signals voltage levels). It
> often caused bad situations, some of them were in fact the same as Your
> case.
>
> Using of DTR line is fine but if You have some HW problem on RxD line,
> there could be the same problem also on DTR (RTS, CTS ...) line and
> final effect will be the same...
> Additionally, lot of devices are working without using any additional
> link/modem state signals (i.e., only signals RxD, TxD and GND are used),
> it is simplest and almost working solution (and working well - of course
> it depends of kind of communication and it also expects that RS232
> specification is not violated and all HW is in order...).
>

FWIW I am using

       description: Motherboard
       product: DQ45CB
       vendor: Intel Corporation
       physical id: 0
       version: AAE30148-302

Since the BIOS does not support serial boot it is not tested for this
I guess. I don't have an oscilloscope so I can't tell what is going on
on the wires.

Is there some alternative to using the onboard port? Does Grub support
some serial addon cards?

Also since this seems to be a common problem with PC hardware perhaps
a warning in the manual and the configuration file should be added.

Thanks

Michal

Re: Timeout broken with serial port

[starous]

<Michal Suchanek"> wrote:
> On 5 September 2011 22:51, Aleš Nesrsta
> <starous@volny.cz> wrote:
> > Hi,
> >
> > I think it is not GRUB related problem, more
> > probably there is some HW
> > > problem on Your serial port. Try to check idle
> > voltage on RxD pin of
> > > serial port (best with oscilloscope... :-) ). Or
> > You can have some
> > > unwanted "leakage" between pins of connector or
> > wires in cable (check
> > > resistance between wires of cable) etc.
> >
> > Normal serial port should NEVER do anything if
> > there is nothing
> > > connected to it. According to RS232 (V24)
> > specification there should be
> > > voltage in range from -12V to -3V on RxD pin in
> > idle state (in simplest
> > > case there is some internal pull-up resistor
> > connected from RxD to -Vcc
> > > directly inside UART or something else...).
> > In this normal case nothing is received by
> > serial port and nothing will
> > > be interpreted as keypress in GRUB, i.e. GRUB
> > will boot normally.
> > >
> > In Your case, probably something causes some
> > noise on RxD pin or there
> > > is bad voltage level on RxD pin (e.g. >= +3V or
> > near to 0V etc.) when
> > > nothing is connected - so serial port interprets
> > it as receiving some
> > > character(s).
> > When You connect Your serial port to another
> > device (notebook in Your
> > > case), RxD is set to correct idle voltage from
> > connected idle TxD, so
> > > the problem is corrected.
> >
> > From my long praxis with serial ports on
> > different PCs and another
> > > (non-PC) devices, manufacturers very very often
> > don't respect RS232
> > > (V24) specification (mainly for RxD/TxD signals
> > voltage levels). It
> > > often caused bad situations, some of them were
> > in fact the same as Your
> > > case.
> >
> > Using of DTR line is fine but if You have some
> > HW problem on RxD line,
> > > there could be the same problem also on DTR
> > (RTS, CTS ...) line and
> > > final effect will be the same...
> > Additionally, lot of devices are working without
> > using any additional
> > > link/modem state signals (i.e., only signals
> > RxD, TxD and GND are used),
> > > it is simplest and almost working solution (and
> > working well - of course
> > > it depends of kind of communication and it also
> > expects that RS232
> > > specification is not violated and all HW is in
> > order...).
> > >
> 
> FWIW I am using
> 
> description: Motherboard
> product: DQ45CB
> vendor: Intel Corporation
> physical id: 0
> version: AAE30148-302
> 
> Since the BIOS does not support serial boot it is
> not tested for this
> I guess. I don't have an oscilloscope so I can't
> tell what is going on
> on the wires.

If You can, try at least simply measure voltage between pins 2 (RxD)
and 5 (GND) - if You have commonly used 9 pins connector on PC.
Common ("minus" - often black color) measuring wire should be
connected to pin 5 (GND). If You will measure voltage  outside the
range approx. <-12..-3V> then most probably there is something wrong
with Your serial port - from HW point of view.

There is also question (but maybe I don't understand Your original
post) - why do You use serial module in GRUB when You don't need
serial port ?
I think, if serial module will not be loaded in GRUB, You should
don't care about serial port even if it is broken or buggy...

> 
> Is there some alternative to using the onboard
> port? Does Grub support
> some serial addon cards?

AFAIK, GRUB supports at least USB serial port converter.
Or probably any addon card which looks from HW side as standard COM
port...

> 
> Also since this seems to be a common problem with
> PC hardware perhaps
> a warning in the manual and the configuration file
> should be added.

Hm, why not, it is maybe not bad idea - but I am not maintainer of
related GRUB parts (serial port nor configuration etc.), maybe
somebody else will discuse it with You...
But, generally, there could be many partially or fully broken HW
parts on PC and it is impossible to warn everybody against all of
them... :-)

BR,
Ales

> 
> Thanks
> 
> Michal


> 
> _______________________________________________
> Grub-devel mailing list
> Grub-devel@gnu.org
> https://lists.gnu.org/mailman/listinfo/grub-devel
> 


-- 
Užitečné rady a tipy pro studenty i jejich rodiče a blízké čtěte ve
speciálu portálu VOLNÝ.cz Škola volá.
http://web.volny.cz/data/click.php?id=1267

Re: Timeout broken with serial port

[Vladimir 'φ-coder/phcoder' Serbinenko]

[-- Attachment #1: Type: text/plain, Size: 719 bytes --]

> But, generally, there could be many partially or fully broken HW
> parts on PC and it is impossible to warn everybody against all of
> them... :-)
>
Why not? "Your computer is manufactured by Sony/Apple/..., it may have
hardware defects and exhibit any kind of known or unknown, predictable
or unpredictable, explainable or mysterious misbehavior subject to laws
of physics as we know them or yet undiscovered proprietary extensions"

> BR,
> Ales
>
>> Thanks
>>
>> Michal
>
>> _______________________________________________
>> Grub-devel mailing list
>> Grub-devel@gnu.org
>> https://lists.gnu.org/mailman/listinfo/grub-devel
>>
>


-- 
Regards
Vladimir 'φ-coder/phcoder' Serbinenko



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 294 bytes --]

Re: Timeout broken with serial port

[richardvoigt]

On Mon, Sep 5, 2011 at 3:51 PM, Aleš Nesrsta <starous@volny.cz> wrote:
> Hi,
>
> I think it is not GRUB related problem, more probably there is some HW
> problem on Your serial port. Try to check idle voltage on RxD pin of
> serial port (best with oscilloscope... :-) ). Or You can have some
> unwanted "leakage" between pins of connector or wires in cable (check
> resistance between wires of cable) etc.
>
> Normal serial port should NEVER do anything if there is nothing
> connected to it. According to RS232 (V24) specification there should be
> voltage in range from -12V to -3V on RxD pin in idle state (in simplest
> case there is some internal pull-up resistor connected from RxD to -Vcc
> directly inside UART or something else...).
> In this normal case nothing is received by serial port and nothing will
> be interpreted as keypress in GRUB, i.e. GRUB will boot normally.
>
> In Your case, probably something causes some noise on RxD pin or there
> is bad voltage level on RxD pin (e.g. >= +3V or near to 0V etc.) when
> nothing is connected - so serial port interprets it as receiving some
> character(s).


An idle transmitter should be in the range -3V...-12V.  But here there
is no transmitter.  I would not be at all surprised if some hardware
implements "cable detect" using a weak pull to 0V.

Still, data shouldn't be reported by the receiver unless the voltage
is in the valid range, and the input should not be left floating
either.

Re: Timeout broken with serial port

[Lennart Sorensen]

On Mon, Sep 05, 2011 at 09:40:42PM +0200, Michal Suchanek wrote:
> I did not have the cable connected at all at the other end. Connecting
> it to my notebook resolves the problem.
> 
> Still it is definitely a bug in Grub that it interprets noise as keypresses.
> 
> Is there not DTR line?
> 
> If every switch with serial console failed to boot when there is
> nothing connected to it there would be no internet. This is not how a
> serial enabled device should behave.

If the UART (not grub) interprets noise as valid data, then it will pass
that to grub which legitimately sees that as keypresses.

Not a bug in software, just bad hardware.

DTR is used for modems and such, generally not for consoles, and besides
who is the terminal?  The serial port on many devices can be the terminal
if you connect it to a modem, but might just as well be the client device
connected to some other terminal (like when used as the grub interface).

-- 
Len Sorensen

Re: Timeout broken with serial port

[Vladimir 'φ-coder/phcoder' Serbinenko]

[-- Attachment #1: Type: text/plain, Size: 716 bytes --]

On 06.09.2011 02:05, Lennart Sorensen wrote:
> If the UART (not grub) interprets noise as valid data, then it will pass
> that to grub which legitimately sees that as keypresses.
I agree with both sides of dispute. It's a hardware bug but GRUB has
reliability and unattended boot as one of its goals. And this noise can
be detected. Noise is likely to have two neighbouring bits equal (due to
noise not being on the same frequency as serial speed) and unlikely to
produce a lot of valid characters. I propose to buffer serial input for
200/speed seconds and after this period do a small analysis and in case
of bad data discard the whole buffer.

-- 
Regards
Vladimir 'φ-coder/phcoder' Serbinenko



[-- Attachment #2: OpenPGP digital signature --]
[-- Type: application/pgp-signature, Size: 294 bytes --]