[Top][All Lists]

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Memory again

From: YAMAMOTO Mitsuharu
Subject: Re: Memory again
Date: Wed, 18 Jan 2012 10:30:24 +0900
User-agent: Wanderlust/2.14.0 (Africa) SEMI/1.14.6 (Maruoka) FLIM/1.14.8 (Shijō) APEL/10.6 Emacs/22.3 (sparc-sun-solaris2.8) MULE/5.0 (SAKAKI)

>>>>> On Tue, 17 Jan 2012 15:14:10 +0200, emacs user <address@hidden> said:

> On Tue, Jan 17, 2012 at 12:58 PM, YAMAMOTO Mitsuharu wrote:
>> Is this also reproducible on other builds such as tty-only or X11?

> if I run the same executable using -nw, it just wont crash.  I can
> bring it to a size of over 1Gb with no buffers at all; with buffers,
> I could bring it to a size of 2.5Gb.  at that point I can use vm and
> other packages that often crash emacs, and it's stable and happy
> apart from the increasing memory.

This would indicate increase in RSS does not necessarily mean
unstableness you originally observed with repetitive invocations of
VM.  It would be worth trying the X11 build to see its behavior is
much like the NS port with graphical frame case (crash in GC?) or tty
case (stable).

> upon killing buffers, the size does go down from 2.5Gb to 1Gb, for
> example, but the minimum memory size used does consistently increase
> with usage.

Do you mean RSS continues to increase on each invocation of VM?  What
do the outputs of the `heap' command look like before and after an
invocation of VM at this stage?  If you find increase in malloced
memory in these outputs, then can you identify the stack traces for
some major differences using `malloc_history'?

                                     YAMAMOTO Mitsuharu

reply via email to

[Prev in Thread] Current Thread [Next in Thread]