cvs commit: src/sys/vm vm_kern.c
silby at silby.com
Mon Feb 16 12:01:35 PST 2004
On Mon, 16 Feb 2004, Robert Watson wrote:
> getting somewhere. I'm not sure what the right answer in procfs is, but I
> think this isn't it. Maybe we need a new M_flag that says "And fail if
> it's rediculous", but I'm very concerned that we just substituted memory
> allocation semantics throughout the kernel and the impact it will have...
> It could be harmless, but it's also not a change to make without a lot of
> hard cogitation.
> Robert N M Watson FreeBSD Core Team, TrustedBSD Projects
> robert at fledge.watson.org Senior Research Scientist, McAfee Research
If maintaining the old semantics is important, then adding a third flag
sounds like the way to go. Unfortunately, I can't think of something
appropriate that would fit into the form M_XXXXXX.
I'm only jumping in because I had considered making a change similar to
DES's back when I was investigating why the old pipe code could panic the
machine; the current situation of panic vs NOWAIT is very frustrating.
Mike "Silby" Silbersack
More information about the cvs-src