Network Slowdowns?

[Matthew Dillon]

:I Got a crash again. I've uploaded it to crash/pf. I reorganized the
:directory a little bit, so the core you checked is vmcore.old and the
:new core is vmcore.new, with corresponding asf-files asf.old and asf.new.
:
:Because of stupidity the previous asf-file contained paths to /modules,
:so if you used it, the trace could be wrong. I'm very sorry about that.
:
:Anyway, I've checked that the cores produces the same traces as on my box
:with the provided asf-files.
:
:Setting kern.mmxopt=0 _might_ have actually helped against one problem,
:because this crash has a different trace from my old crashes. However, I
:recognize this trace, so it has occured sometime before too.
:
:Hope this helps, and apologies for the bad asf-file.
:
:        -Richard

    The error is very similar.  Somehow the instruction pointer is
    winding up pointing to the stack, which implies stack corruption,
    but the error occurs in a procedure where that simply should not
    be possible.

    Are you compiling your kernel with any particular optimizations?

					-Matt
					Matthew Dillon 
					<dillon at xxxxxxxxxxxxx>