Mailing Lists: Apple Mailing Lists
Image of Mac OS face in stamp
 
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Network driver strangeness

Well... After giving everything a good look over and spent a few days analyzing network traffic, etc., I can't find any differences or inconsistencies. All packets seem to be getting through correctly and with the right checksums, etc. The most I've been able to determine is that it only happens to apps that appear to use OpenTransport as the underlying communications interface. The OpenTransport error message I've seen is kOTStateChangeErr, but I'm at a loss as to what I'm doing inside my driver that might cause OpenTransport to think it's changing state somehow...

From a 3rd machine and a packet sniffer, everything looks good, except that right when it should be sending the first HTTP GET packet, it sends a RST,SYN instead. I can't find any traffic differences between a good signal (either wired or wireless using a different app) and a bad one.

Ideas anyone?

-Rob

On Saturday, February 1, 2003, at 12:46 PM, Andrew Gallatin wrote:


Rob McKeever writes:

The symptom in Explorer is that you can't connect to the internal
web-based admin page on the home router I'm running on. This page
works under the exact same configuration just fine with IE if another
network driver is used and also works with my driver with any other web

Assuming that IE is using only TCP/IP, you may be able to determine
whether the problem is transmit or receive based by using netstat
to look for bad checksums. Eg: run 'netstat -s | grep sum' on the
host your driver is running on.

If you see bad checksums, its likely that your receive handling has
a problem and you should concentrate on that facet of the driver.
If you don't see anything here, then it may be safe to assume that
the problem is transmit based. You might then look for bad checksum
on your home router (if it provides such a mechanism, *nix based ones
sometimes do).

Once you've isolated the problem to transmit or receive, you can stare
at that part of your driver, looking for bugs, adding printfs for
packets which stress error handling paths, etc.

I'm not sure what interfaces IOKit provides to an ethernet driver, as
my driver ignores this part of IOKit, but its conceivable that you're
getting screwed up by getting a very unusual mbuf chain. Eg, a chain
with more segments than you are prepared to handle.

Two tests I've found helpful to isolate bad drivers in the past is to
just telnet to a host using your driver, and do a find . -name '*c'
-exec cat {} \; in some large source tree. Netpipe is also a good
tool.

Good luck!

Drew


--
"Quidquid latine dictum sit, altum viditur."
_______________________________________________
darwin-drivers mailing list | email@hidden
Help/Unsubscribe/Archives: http://www.lists.apple.com/mailman/listinfo/darwin-drivers
Do not post admin requests to the list. They will be ignored.
References: 
 >Re: Network driver strangeness (From: Andrew Gallatin <email@hidden>)



Visit the Apple Store online or at retail locations.
1-800-MY-APPLE

Contact Apple | Terms of Use | Privacy Policy

Copyright © 2007 Apple Inc. All rights reserved.