Re: divide by zero in serial port driver

Khimenko Victor (khim@sch57.msk.ru)
Mon, 30 Nov 1998 12:13:18 +0300 (MSK)


In <Pine.LNX.3.94.981129184256.9085A-100000@artax.karlin.mff.cuni.cz> Mikulas Patocka (mikulas@artax.karlin.mff.cuni.cz) wrote:
>> 29-Nov-98 17:44 you wrote:
>> > When I was trying to configure serial port using setserial, I got
>> > divide-by-zero kernel error. I can't reproduce it and I don't know where
>> > it happened - modules are loaded each time in different order as I use
>> > devices :-(
>>
>> If you could not reproduce it this could not be fixed. Useless bug report.

MP> So you think that all bugs _must_ be reproducable? Good opinion ;-)

This is ABSOLUTELY MUST if you wish fix for them :-) Read files in
Documentation subdir -- there are info about creating usefull bug-report.

MP> I think the author of serial driver should go through the file,
MP> and check any division (grep '[^\*]\/[^\*]' serial.c). He surely
MP> understand his code better than me and can find it in less time.

Looks like you misunderstood situation even more :-(( Why you are so sure that
error in serial.c ? Just since it was produced by setserial ??? Gosh. You are
joking, right ? Error could be in A LOT OF places in kernel or even in setserial
itself (kernel is not designed to cope with all broken programs started by
root) including all libraries used by setserial and even compiler used to
compiler kernel/or setserial. To find source of error you should be able to
reproduce it or AT LEAST have kernel compilation options handy and decrypted
(via ksymoops) oops-message. Thus you bug report is COMPLETELY useless. All bug
reports where are no usefull info (like your bugreport) are forwarded to
/dev/null by all matainers. One exception: if you just found already well-known
bug.

-
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/