Possible bug in serial port driver in 2.2.3

Anthony Barbachan (barbacha@Hinako.AMBusiness.com)
Fri, 26 Mar 1999 04:42:32 -0500


While haggling with my pppd configuration files due to changes in the
pppd configuration files between 2.2.0 and 2.3.5 I think I may have come
across a bug in the serial driver included with the 2.2.3 kernel. During my
tests of pppd and uugetty (which answered the modem) trying to get
everything to work under the new kernel I came across an unexpected error
message, rarely but it did appear several times. To the best of my memory
this message seamed to especially appear when I manually killed uugetty
because it appeared stuck. The message said either corruption or
inconsistancy and then printed two structure variable followed by their
values. I've looked through the serial driver source and think that I've
found the printk message related to error message that I got several times.
I can't be absolutely certain that this is the one as that machine is at a
different location and I left their late after fixing the pppd problems
without any time to reinvestigate the problem. However I think this message
is too familiar for me to be worng about this.

Line 2256 in drivers/char/serial.c

printk("rs_close: bad serial port count; tty->count is 1, "
"state->count is %d\n", state->count);

As this seams to have appeared only when I killed uugetty its possible
that the error was somehow related to the occasional times that uugetty was
unresponsive to dialins, however I cannot be certain that that was the only
times that it appeared. If its of any help the modem is a PNP Cirrus Logic
33.6 in PNP mode (the jumpers did not work on this modem for some reason).
The modem has also been working consistantly for several months before now
except for a couple of extremely rare occasion that the modem became
unresponsive because for some reason the PNP on the modem had deconfigured
itself. Those two times the modem was brought back into working order by
just rerunning isapnp. However the problems that appeared this time are not
the same and did not have to be resolved in the same manner. This modem is
the only internal and I did not notice any similar warning messages emerging
from any of the three external modems, however those modems did not receive
anything close to the same amount of use that the internal one did due to
the number of trial dialins I made while trying to get PPP to function
correctly.

- A. Barbachan

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