Subject: Re: lazy mlock?
To: Wojciech Puchar <>
From: Greywolf <>
List: tech-kern
Date: 04/17/2002 22:20:32
Hey, Woj.

I think you can't be possibly farther from the mark.  It's the kernel's
responsibility to limit the abuse from the programmer, because
programmers are either too smart or too stupid for the system's purpose.

[There are those in between who know how to play by the rules.  We call
them nice names like "developers" or "engineers" instead of the much
less magnanimous "application programmer".]

Either way, circumventing protections or not knowing how to take proper
resource precautions are not excuses for the system to say, "screw it, I
give up."

If the kernel doesn't -- or can't -- take the precautions necessary to
keep the system up and going, it's no wonder if it falls over.

I would think that BSD could be smarter than that.  It'd sure put them
ahead of the game.

On Wed, 17 Apr 2002, Wojciech Puchar wrote:

# it's microsoft-like method of solving problems.
# instead of fixing faulty program that overallocate memory you like to add
# feature to kernel that will detect that.

NetBSD: Rock solid!