GNU bug report logs - #2685
Emergency (alloc): Warning: past 85% of memory limit

Previous Next

Package: emacs;

Reported by: Laurent GUERBY <laurent <at> guerby.net>

Date: Mon, 16 Mar 2009 11:25:04 UTC

Severity: normal

Done: Chong Yidong <cyd <at> stupidchicken.com>

Bug is archived. No further changes may be made.

To add a comment to this bug, you must first unarchive it, by sending
a message to control AT debbugs.gnu.org, with unarchive 2685 in the body.
You can then email your comments to 2685 AT debbugs.gnu.org in the normal way.

Toggle the display of automated, internal messages from the tracker.

View this report as an mbox folder, status mbox, maintainer mbox


Report forwarded to bug-submit-list <at> lists.donarmstrong.com, Emacs Bugs <bug-gnu-emacs <at> gnu.org>:
bug#2685; Package emacs. (Mon, 16 Mar 2009 11:25:04 GMT) Full text and rfc822 format available.

Acknowledgement sent to Laurent GUERBY <laurent <at> guerby.net>:
New bug report received and forwarded. Copy sent to Emacs Bugs <bug-gnu-emacs <at> gnu.org>. (Mon, 16 Mar 2009 11:25:04 GMT) Full text and rfc822 format available.

Message #5 received at submit <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Laurent GUERBY <laurent <at> guerby.net>
To: emacs-pretest-bug <at> gnu.org
Subject: Emergency (alloc): Warning: past 85% of memory limit
Date: Mon, 16 Mar 2009 12:13:55 +0100
Hi,

I compiled using GCC 4.3.2 from debian lenny (abi=32):

GNU Emacs 23.0.91.1 (mips64el-unknown-linux-gnu, X toolkit) of
2009-03-15 on gcc51   

When launching emacs (in text mode) I get a *Warnings* window
with the following message:

Emergency (alloc): Warning: past 85% of memory limit

The machine gcc51 has 1GB of RAM so this warning seems surprising.

Sincerely,

Laurent






Information forwarded to bug-submit-list <at> lists.donarmstrong.com, Emacs Bugs <bug-gnu-emacs <at> gnu.org>:
bug#2685; Package emacs. (Wed, 15 Apr 2009 20:40:05 GMT) Full text and rfc822 format available.

Acknowledgement sent to John Sullivan <johns <at> fsf.org>:
Extra info received and forwarded to list. Copy sent to Emacs Bugs <bug-gnu-emacs <at> gnu.org>. (Wed, 15 Apr 2009 20:40:06 GMT) Full text and rfc822 format available.

Message #10 received at 2685 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: John Sullivan <johns <at> fsf.org>
To: 2685 <at> debbugs.gnu.org
Subject: Also seeing this
Date: Wed, 15 Apr 2009 16:32:39 -0400
I'm also seeing this on pretest with emacs compiled on mipsel (mips64),
using gcc-4.3. I have 1GB of RAM, and ulimit returns unlimited.

It's a Debian Lenny system, Linux 2.6.27.

I've verified that it happens with emacs -Q -nw.




Information forwarded to bug-submit-list <at> lists.donarmstrong.com, Emacs Bugs <bug-gnu-emacs <at> gnu.org>:
bug#2685; Package emacs. (Wed, 20 May 2009 23:40:06 GMT) Full text and rfc822 format available.

Acknowledgement sent to Chong Yidong <cyd <at> stupidchicken.com>:
Extra info received and forwarded to list. Copy sent to Emacs Bugs <bug-gnu-emacs <at> gnu.org>. (Wed, 20 May 2009 23:40:06 GMT) Full text and rfc822 format available.

Message #15 received at 2685 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Chong Yidong <cyd <at> stupidchicken.com>
To: Laurent GUERBY <laurent <at> guerby.net>, John Sullivan <johns <at> fsf.org>
Cc: 2685 <at> debbugs.gnu.org
Subject: Re: Emergency (alloc): Warning: past 85% of memory limit
Date: Wed, 20 May 2009 19:32:35 -0400
> I compiled using GCC 4.3.2 from debian lenny (abi=32):
>
> GNU Emacs 23.0.91.1 (mips64el-unknown-linux-gnu, X toolkit) of
> 2009-03-15 on gcc51   
>
> When launching emacs (in text mode) I get a *Warnings* window
> with the following message:
>
> Emergency (alloc): Warning: past 85% of memory limit
>
> The machine gcc51 has 1GB of RAM so this warning seems surprising.

I can't reproduce this on x86.  I don't have access to any mipsel
machine; could either of you give me a temporary account so that I can
try to debug this (if so, please contact me by private email).




Information forwarded to bug-submit-list <at> lists.donarmstrong.com, Emacs Bugs <bug-gnu-emacs <at> gnu.org>:
bug#2685; Package emacs. (Fri, 22 May 2009 04:45:07 GMT) Full text and rfc822 format available.

Acknowledgement sent to Chong Yidong <cyd <at> stupidchicken.com>:
Extra info received and forwarded to list. Copy sent to Emacs Bugs <bug-gnu-emacs <at> gnu.org>. (Fri, 22 May 2009 04:45:07 GMT) Full text and rfc822 format available.

Message #20 received at 2685 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Chong Yidong <cyd <at> stupidchicken.com>
To: laurent <at> guerby.net
Cc: John Sullivan <johns <at> fsf.org>, 2685 <at> debbugs.gnu.org
Subject: Re: Emergency (alloc): Warning: past 85% of memory limit
Date: Fri, 22 May 2009 00:39:27 -0400
The problem is that Emacs is not determining the start of the data
segment properly on MIPS machines.  In src/m/mips.h, we define
DATA_START as 0x10000000.  There must be a reason for this, but it's
wrong on the machine you provided for testing; there, __morecore reports
the memory break (which must lie inside the data segment) as a value
less than 0x10000000.

If I comment out DATA_START, so that Emacs uses another method to
determine the start of the data segment size, everything seems to work
OK.  But I don't know if this may lead to a different bug elsewhere.

*** trunk/src/m/mips.h.~1.42.~	2009-03-12 09:42:54.000000000 -0400
--- trunk/src/m/mips.h	2009-05-22 00:34:03.000000000 -0400
***************
*** 49,55 ****
  /* Describe layout of the address space in an executing process.  */
  
  #define TEXT_START      0x00400000
! #define DATA_START      0x10000000
  
  
  #if 0 /* These definitions were advantageous when not using
--- 49,55 ----
  /* Describe layout of the address space in an executing process.  */
  
  #define TEXT_START      0x00400000
! /* #define DATA_START      0x10000000 */
  
  
  #if 0 /* These definitions were advantageous when not using





Reply sent to Chong Yidong <cyd <at> stupidchicken.com>:
You have taken responsibility. (Fri, 22 May 2009 18:50:05 GMT) Full text and rfc822 format available.

Notification sent to Laurent GUERBY <laurent <at> guerby.net>:
bug acknowledged by developer. (Fri, 22 May 2009 18:50:05 GMT) Full text and rfc822 format available.

Message #25 received at 2685-done <at> emacsbugs.donarmstrong.com (full text, mbox):

From: Chong Yidong <cyd <at> stupidchicken.com>
To: Laurent GUERBY <laurent <at> guerby.net>
Cc: John Sullivan <johns <at> fsf.org>, 2685-done <at> debbugs.gnu.org
Subject: Re: Emergency (alloc): Warning: past 85% of memory limit
Date: Fri, 22 May 2009 14:46:56 -0400
> The problem is that Emacs is not determining the start of the data
> segment properly on MIPS machines.  In src/m/mips.h, we define
> DATA_START as 0x10000000

On further research, I think this is wrong for GNU/Linux.  It was
probably written for other operating systems in mind.  Undefining
DATA_START on GNU/Linux is pretty harmless; as I said, this causes Emacs
to fall back on another heuristic method for determining the start of
the data segment.  (Anyway, this change only affects this particular
platform).  So, I've checked in the change.




Information forwarded to bug-submit-list <at> lists.donarmstrong.com, Emacs Bugs <bug-gnu-emacs <at> gnu.org>:
bug#2685; Package emacs. (Wed, 27 May 2009 03:10:05 GMT) Full text and rfc822 format available.

Acknowledgement sent to John Sullivan <johns <at> fsf.org>:
Extra info received and forwarded to list. Copy sent to Emacs Bugs <bug-gnu-emacs <at> gnu.org>. (Wed, 27 May 2009 03:10:05 GMT) Full text and rfc822 format available.

Message #30 received at 2685 <at> emacsbugs.donarmstrong.com (full text, mbox):

From: John Sullivan <johns <at> fsf.org>
To: Chong Yidong <cyd <at> stupidchicken.com>
Cc: Laurent GUERBY <laurent <at> guerby.net>, John Sullivan <johns <at> fsf.org>,
        2685 <at> debbugs.gnu.org
Subject: Re: Emergency (alloc): Warning: past 85% of memory limit
Date: Tue, 26 May 2009 23:06:31 -0400
Chong Yidong <cyd <at> stupidchicken.com> writes:

>> The problem is that Emacs is not determining the start of the data
>> segment properly on MIPS machines.  In src/m/mips.h, we define
>> DATA_START as 0x10000000
>
> On further research, I think this is wrong for GNU/Linux.  It was
> probably written for other operating systems in mind.  Undefining
> DATA_START on GNU/Linux is pretty harmless; as I said, this causes Emacs
> to fall back on another heuristic method for determining the start of
> the data segment.  (Anyway, this change only affects this particular
> platform).  So, I've checked in the change.
>

I just built the pretest and the error is gone. Thanks!



bug archived. Request was from Debbugs Internal Request <help-debbugs <at> gnu.org> to internal_control <at> emacsbugs.donarmstrong.com. (Wed, 24 Jun 2009 14:24:10 GMT) Full text and rfc822 format available.

This bug report was last modified 15 years and 363 days ago.

Previous Next


GNU bug tracking system
Copyright (C) 1999 Darren O. Benham, 1997,2003 nCipher Corporation Ltd, 1994-97 Ian Jackson.