Re: Cancel-safe without a thread?
- From: "Don Burn" <burn@xxxxxxxxxxxxxxxx>
- Date: Tue, 5 Sep 2006 08:18:13 -0400
<BubbaGump> wrote in message
news:11jpf2p2le3ot57lci8og6gjklt8lcpr2t@xxxxxxxxxx
Of course, it's not that everything can be understood by everyone. I
love things that involve eliminating race conditions, although I still
find computer science theories like balanced binary trees a bit
mysterious. I get the idea, but the details won't solidify in my mind
to a point where I'm comfortable. So, to a certain degree I trust
some of the idiosyncrasies about algorithms like red-black trees.
Things I don't understand I leave to someone else. Things I do, I
leave to me.
The problem is how do you know which is which? I have seen a number of
drivers that people thought they understood things like cancel, and yes it
worked correctly 99% of the time. They tested didn't see a problem and
thought "gee, this wasn't that hard!". Then it went into the field, and the
bugs started comming in, so now the work is more than doubled, since someone
has to seperate the logic needed from the driver from the broken cancel
logic, then reimplement the cancel logic with a "blessed solution".
Things like this cause crashes, and the rework always under a tight schedule
does not aid quality.
--
Don Burn (MVP, Windows DDK)
Windows 2k/XP/2k3 Filesystem and Driver Consulting
http://www.windrvr.com
Remove StopSpam from the email to reply
.
- References:
- Re: Extra reference?
- From: Don Burn
- Re: Cancel-safe without a thread?
- From: Doron Holan [MS]
- Re: Extra reference?
- Prev by Date: Re: Driver memory footprint/CPU utilization
- Next by Date: Re: Driver memory footprint/CPU utilization
- Previous by thread: Re: Cancel-safe without a thread?
- Next by thread: Re: Cancel-safe without a thread?
- Index(es):
Relevant Pages
|
