Skip to contentSkip to conversation

swap/memory patches

DISCUSSION Rik van Riel 3 participants 3 messages Oct 22, 1998 – Oct 26, 1998 1 message not in the archive Memory management

lore
Share & tools
lore

A message that is not in the archive (sent off-list or not yet received): [email protected]

#1 lore raw
Quoted 4 lines · Kurt Garloff

There have been quite some patches on MM/swap/OOM of Linux by you, Andrea and Stephen during the last time. Is there a final version? Are there recent (2.1.125+) patches for these problems? Did anything get into the kernel?

There are no final versions and nothing got into the kernel. Personally, I think Andrea's patch (minus the kswapd early give-up) should go into the kernel now.

My part of the patch (OOM killing when kswapd can't cope any more _and_ we're really out of memory) is running well at a lot of people's places, but it really is an added feature and we're in feature freeze right now.

This feature freeze means that I'm not going to drive Linus crazy in order to include a feature, not even if 100 sysadmins have asked me to. If you really really want it in you'll either have to drive Linus crazy yourself or you'll have to apply the patch to the kernel yourself -- I will make sure that it's available...

Quoted 5 lines

What I have in mind, when writing these lines, is:
(1) cow-swapin: I often observed that after compiling a large C++ program (which needs some swap), the shell keeps swapping something in on every <Enter> keypress. This is cured by swapoff -a; swapon -a.
If I correctly understood, this is what cow-swapin was supposed to cure.

Wasn't this fixed -- Andrea, Stephen?

Quoted 5 lines

(2) somebody claimed having found swap bugs
(3) better oom behaviour

I consider (1) to be a serious bug. I don't know much about (2). (3) is obviously hard to solve cleanly.

Several people are running with my solution to (3) in their kernel and have reported that it kills the right task 9 out of 10 times. This indicates that my patch is working rather good...

Rik.

+-------------------------------------------------------------------+
| Linux memory management tour guide.        [email protected] |
| Scouting Vries cubscout leader.      http://www.phys.uu.nl/~riel/ |
+-------------------------------------------------------------------+

--
This is a majordomo managed list. To unsubscribe, send a message with the body 'unsubscribe linux-mm me@address' to: [email protected]

#2 lore raw
Quoted 6 lines · Rik van Riel
Quoted 4 lines

(1) cow-swapin: I often observed that after compiling a large C++ program (which needs some swap), the shell keeps swapping something in on every <Enter> keypress. This is cured by swapoff -a; swapon -a.
If I correctly understood, this is what cow-swapin was supposed to cure.

Wasn't this fixed -- Andrea, Stephen?

Just fixed in Linus's tree.

Andrea Arcangeli

--
This is a majordomo managed list. To unsubscribe, send a message with the body 'unsubscribe linux-mm me@address' to: [email protected]

#3 lore raw

Hi,

In article <[email protected]>,

Quoted 9 lines · Andrea Arcangeli
Quoted 6 lines · Rik van Riel
Quoted 4 lines

(1) cow-swapin: I often observed that after compiling a large C++ program (which needs some swap), the shell keeps swapping something in on every <Enter> keypress. This is cured by swapoff -a; swapon -a.
If I correctly understood, this is what cow-swapin was supposed to cure.

Wasn't this fixed -- Andrea, Stephen?

Just fixed in Linus's tree.

Definitely should be! The fix went in in 2.1.123, so if you can reproduce the problem on any later kernel, we'd like to hear about it.

--Stephen
--
This is a majordomo managed list. To unsubscribe, send a message with the body 'unsubscribe linux-mm me@address' to: [email protected]