* serial port forcing
@ 2003-12-01 11:18 Nikolay Datchev
2003-12-01 11:22 ` Jan-Benedict Glaw
0 siblings, 1 reply; 5+ messages in thread
From: Nikolay Datchev @ 2003-12-01 11:18 UTC (permalink / raw)
To: linux-serial
Hello list,
Here is my problem:
I have one linux machine with one 4xserial PCI board, another one 2xserial
PCI board, and one onboard serial port, total 7 ports. After boot-up, only
6 ports appears (setserial -bg /dev/ttyS? shows ttyS0,1,4,5,6,7, same in
cat /proc/tty/driver/serial). 4 to 7 are the four-port board ports, and
they work fine. 0 and 1 are recognized with standart settings - port
0x3f8/0x2f8, irq 4/3, but i'm able to setserial them with the two board
ports settings, and all works fine. It works also if i setserial only one
of the two board ports, and i let ttyS0 with default settings, which makes
onboard port usable. But i need all the 7 ports.
I'm using linux kernel v. 2.4.22, and i get in /proc/tty/driver/serial
only ports, that has been recognized upon boot-up, hense i'm not able to
force using more than 6 ports (setserial /dev/ttySX port Y irq Z where X
is some serial port not appearing in /proc does not result in anything).
With kernel 2.4.18, i got up to 64 ports in /proc, but those not
recognized in the boot-time has unknown uart and fake recources.
How can i make all 7 ports usable ? In other words, how to force another
serial port recognition ? Kernel downgrade is not an option.
Thanks in advance.
-- Nikolay Datchev
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: serial port forcing
2003-12-01 11:18 serial port forcing Nikolay Datchev
@ 2003-12-01 11:22 ` Jan-Benedict Glaw
2003-12-01 12:02 ` Nikolay Datchev
0 siblings, 1 reply; 5+ messages in thread
From: Jan-Benedict Glaw @ 2003-12-01 11:22 UTC (permalink / raw)
To: linux-serial
[-- Attachment #1: Type: text/plain, Size: 825 bytes --]
On Mon, 2003-12-01 13:18:44 +0200, Nikolay Datchev <nik@varna.net>
wrote in message <Pine.LNX.4.33.0312011257140.11314-100000@ns.varna.net>:
> Hello list,
>
> Here is my problem:
>
> I have one linux machine with one 4xserial PCI board, another one 2xserial
> PCI board, and one onboard serial port, total 7 ports. After boot-up, only
This all sounds like your 2-port serial PCI board isn't recognized. What
kind of board is this *exactly*? Brand name and "lspci -vv" output could
help...
MfG, JBG
--
Jan-Benedict Glaw jbglaw@lug-owl.de . +49-172-7608481
"Eine Freie Meinung in einem Freien Kopf | Gegen Zensur | Gegen Krieg
fuer einen Freien Staat voll Freier Bürger" | im Internet! | im Irak!
ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: serial port forcing
2003-12-01 11:22 ` Jan-Benedict Glaw
@ 2003-12-01 12:02 ` Nikolay Datchev
2003-12-01 12:50 ` Jan-Benedict Glaw
0 siblings, 1 reply; 5+ messages in thread
From: Nikolay Datchev @ 2003-12-01 12:02 UTC (permalink / raw)
To: Jan-Benedict Glaw; +Cc: linux-serial
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: TEXT/PLAIN; charset=X-UNKNOWN, Size: 2049 bytes --]
The board is some taiwan made crap.
When it is not recognized at all, why i got totally 6 ports, 4 recognized
+ 1 onboard must be 5?
And in the instruction manual of the board, in the Linux section, there is
advice to force some unused ttyS with port and irq from lspci. That's
exaclty what i do, but i have not enough ttyS-es to reconfigure =]. At
least one of 7 ports will be unusable in this situation.
I looked in the drivers/char/serial.c (kernel 2.4.22), and in function
line_info() (near line 3268) there is a check:
if (state->type == PORT_UNKNOWN) {
return 0;
}
After i remove this check, in the /proc/tty/driver/serial i got 64 ports,
just like in 2.4.18, but on test machine. Since the 7-port machine is in
heavy production, i cannot try to change the kernel with this check
removed and force all the ports i need. I am not sure that such kind of
forcing will be successful.
-- Nikolay Datchev
On Mon, 1 Dec 2003, Jan-Benedict Glaw wrote:
> On Mon, 2003-12-01 13:18:44 +0200, Nikolay Datchev <nik@varna.net>
> wrote in message <Pine.LNX.4.33.0312011257140.11314-100000@ns.varna.net>:
> > Hello list,
> >
> > Here is my problem:
> >
> > I have one linux machine with one 4xserial PCI board, another one 2xserial
> > PCI board, and one onboard serial port, total 7 ports. After boot-up, only
>
> This all sounds like your 2-port serial PCI board isn't recognized. What
> kind of board is this *exactly*? Brand name and "lspci -vv" output could
> help...
>
> MfG, JBG
>
> --
> Jan-Benedict Glaw jbglaw@lug-owl.de . +49-172-7608481
> "Eine Freie Meinung in einem Freien Kopf | Gegen Zensur | Gegen Krieg
> fuer einen Freien Staat voll Freier Bürger" | im Internet! | im Irak!
> ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));
>
-
To unsubscribe from this list: send the line "unsubscribe linux-serial" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: serial port forcing
2003-12-01 12:02 ` Nikolay Datchev
@ 2003-12-01 12:50 ` Jan-Benedict Glaw
2003-12-01 13:24 ` Nikolay Datchev
0 siblings, 1 reply; 5+ messages in thread
From: Jan-Benedict Glaw @ 2003-12-01 12:50 UTC (permalink / raw)
To: linux-serial
[-- Attachment #1: Type: text/plain, Size: 1050 bytes --]
On Mon, 2003-12-01 14:02:33 +0200, Nikolay Datchev <nik@varna.net>
wrote in message <Pine.LNX.4.33.0312011345470.1388-100000@ns.varna.net>:
> The board is some taiwan made crap.
>
> When it is not recognized at all, why i got totally 6 ports, 4 recognized
> + 1 onboard must be 5?
>
> And in the instruction manual of the board, in the Linux section, there is
> advice to force some unused ttyS with port and irq from lspci. That's
> exaclty what i do, but i have not enough ttyS-es to reconfigure =]. At
You don't have enough nodes? So, then simply create some new! "man
mknod"
> least one of 7 ports will be unusable in this situation.
NO thanks for not sending requested output. Nobody can help, then...
MfG, JBG
--
Jan-Benedict Glaw jbglaw@lug-owl.de . +49-172-7608481
"Eine Freie Meinung in einem Freien Kopf | Gegen Zensur | Gegen Krieg
fuer einen Freien Staat voll Freier Bürger" | im Internet! | im Irak!
ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));
[-- Attachment #2: Digital signature --]
[-- Type: application/pgp-signature, Size: 189 bytes --]
^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: serial port forcing
2003-12-01 12:50 ` Jan-Benedict Glaw
@ 2003-12-01 13:24 ` Nikolay Datchev
0 siblings, 0 replies; 5+ messages in thread
From: Nikolay Datchev @ 2003-12-01 13:24 UTC (permalink / raw)
To: linux-serial
[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #1: Type: TEXT/PLAIN; charset=X-UNKNOWN, Size: 2236 bytes --]
Seems that i'll help myself. There is big difference between filesystem
character devices for ttyS-es and proc (driver) slots. You can have lot of
filesystem /dev entries for serial ports, but if you don't have enough
/proc/tty/driver/serial slots, you're stuck with the number of serial
ports, recognized by the serial driver at boot time.
After removing the check, pointed by my previous email, i have more than
enough slots in the driver (and in the /proc), just like in the older
kernels, and i'm able to force serial ports, even with parameters of
non-existent ones. I hope that some developper is reading this maillist
and will give me good solution, since just removing this check is too
'quick and dirty'. When i have enough time (this week), i'll try to merge
2.4.18 checks for PORT_UNKNOWN with newer kernel to get enough serial
driver slots.
-- Nikolay Datchev
On Mon, 1 Dec 2003, Jan-Benedict Glaw wrote:
> On Mon, 2003-12-01 14:02:33 +0200, Nikolay Datchev <nik@varna.net>
> wrote in message <Pine.LNX.4.33.0312011345470.1388-100000@ns.varna.net>:
> > The board is some taiwan made crap.
> >
> > When it is not recognized at all, why i got totally 6 ports, 4 recognized
> > + 1 onboard must be 5?
> >
> > And in the instruction manual of the board, in the Linux section, there is
> > advice to force some unused ttyS with port and irq from lspci. That's
> > exaclty what i do, but i have not enough ttyS-es to reconfigure =]. At
>
> You don't have enough nodes? So, then simply create some new! "man
> mknod"
>
> > least one of 7 ports will be unusable in this situation.
>
> NO thanks for not sending requested output. Nobody can help, then...
>
> MfG, JBG
>
> --
> Jan-Benedict Glaw jbglaw@lug-owl.de . +49-172-7608481
> "Eine Freie Meinung in einem Freien Kopf | Gegen Zensur | Gegen Krieg
> fuer einen Freien Staat voll Freier Bürger" | im Internet! | im Irak!
> ret = do_actions((curr | FREE_SPEECH) & ~(NEW_COPYRIGHT_LAW | DRM | TCPA));
>
-
To unsubscribe from this list: send the line "unsubscribe linux-serial" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
^ permalink raw reply [flat|nested] 5+ messages in thread
end of thread, other threads:[~2003-12-01 13:24 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2003-12-01 11:18 serial port forcing Nikolay Datchev
2003-12-01 11:22 ` Jan-Benedict Glaw
2003-12-01 12:02 ` Nikolay Datchev
2003-12-01 12:50 ` Jan-Benedict Glaw
2003-12-01 13:24 ` Nikolay Datchev
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox