Re: More on my crash

• Date: Mon, 18 Dec 1995 09:39:04 PST
• From:
  Pavel Curtis <pavel@parc.xerox.com>
• cc: moo-cows.PARC@xerox.com
• In-Reply-to: "Judy Anderson's message of \"Sat, 16 Dec 1995 11:51:05 PST\" <199512161951.LAA18511@Xenon.Stanford.EDU>^?GY"

Judy Anderson writes:
> I tailed the .db.PANIC and found that the last successfully dumped
> thing was a verb.  I looked once the backup was up, and the next verb
> on that object is a verb with 650 lines of code, #103344:@autoport.
> Presumably that's why it was trying to allocate more memory---is
> dumping a large verb like that all that hard?  Should it *not* be
> trying to allocate memory during the dump process?

It's not all that hard to dump a large verb, but it *does* involve allocating a
fair amount of storage; the server does decompilation by converting the opcode
stream into a parse tree and then pretty-printing that tree.

My guess is that we simply ran out of swap space on lambda, which is of course
a real pain.

	Pavel

• More on my crash
• From: Judy Anderson <yduj@cs.stanford.edu>

• Prev: More on my crash
• Next: MPL patch for 1.7.9
• Index: MOO-cows
• Thread: MOO-cows