Re: ethercons updated for -CURRENT

From: Robert Watson <rwatson_at_FreeBSD.org>
Date: Fri, 9 Jul 2004 10:54:54 -0400 (EDT)
On Fri, 9 Jul 2004, Andrew Gallatin wrote:

> For what its worth, Darwin has this for their KDP ethernet debugging. 
> To receive something, the debugger provides a buffer and specifies a
> timeout.  The driver keeps the same memory mapped for DMA, and just
> copies into the provided void *buffer.  There is also a polled transmit
> routine, where the driver does not return until the outgoing packet has
> been DMA'ed.  In both cases, the driver is not allowed to allocate
> memory, or do anything which can block. 
> 
> One thing that's always concerned me is how do they ensure the driver is
> not in the middle of a "normal" transmit or recv when the debugger is
> entered.. 

I have a lot of concerns about ensuring likely correctness of the model,
including reentrancy, consistent hardware state, packets destined for the
debugger wandering around the stack, etc.  It strikes me that it's likely
the network kernel debugging pieces work most of the time, and likely work
especially well if you're not debugging the network stack :-).  My guess
would be that over time, Apple will start preferring (if they don't
already) firewire debugging simply because it will become available
earlier, work in more sticky situations, etc.

Robert N M Watson             FreeBSD Core Team, TrustedBSD Projects
robert_at_fledge.watson.org      Principal Research Scientist, McAfee Research
Received on Fri Jul 09 2004 - 12:54:58 UTC

This archive was generated by hypermail 2.4.0 : Wed May 19 2021 - 11:38:01 UTC