lkml.org 
[lkml]   [2004]   [Nov]   [23]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
Date
From
SubjectRe: [PATCH] fix spurious OOM kills
Thomas Gleixner wrote:
> On Sun, 2004-11-21 at 14:57 +0100, Thomas Gleixner wrote:
>
>>On Sun, 2004-11-21 at 13:17 +0100, Martin MOKREJŠ wrote:
>>
>>>Why can't the algorithm first find the asking for memory now.
>>>When found, kernel should kill first it's children, wait some time,
>>>then kill this process if still exists (it might exit itself when children
>>>get closed).
>>>You have said it's safer to kill that to send ENOMEM as happens
>>>in 2.4, but I still don't undertand why kernel first doesn't send
>>>ENOMEM, and only if that doesn't help it can start after those 5 seconds
>>>OOM killer, and try to kill the very same application.
>>>I don't get the idea why to kill immediately.
>>
>>I see your concern. There are some more changes neccecary to make this
>>reliably work. I'm not sure if it can be done without really big
>>changes. I will look a bit deeper into this.
>
>
> One big problem when killing the requesting process or just sending
> ENOMEM to the requesting process is, that exactly this process might be
> a ssh login, when you try to log into to machine after some application
> went crazy and ate up most of the memory. The result is that you
> _cannot_ log into the machine, because the login is either killed or
> cannot start because it receives ENOMEM.

I believe the application is _first_ who will get ENOMEM. It must be
terrible luck that it would ask exactly for the size of remaining free
memory. Most probably, it will ask for less or more. "Less" in not
a problem in this case, so consider it asks for more. Then, OOM killer
might well expect the application asking for memory is most probably
exactly the application which caused the trouble.

>
> Putting hard coded decisions like "prefer sshd, xyz,...", " don't kill
> a, b, c" are out of discussion.

I'd go for it at least nowadays.

>
> The ideas which were proposed to have a possibility to set a "don't kill
> me" or "yes, I'm a candidate" flag are likely to be a future way to go.
> But at the moment we have no way to make this work in current userlands.

Do you think login or sshd will ever use flag "yes, I'm a candidate"?
I think exactly same bahaviour we get right now with those hard coded decisions
you mention above. Otherwise the hard coded decision is programmed into
every sshd, init instance anyway. I think it's not necessary to put
login and shells on thsi ban list, user will re-login again. ;)

>
> I refined the decision, so it does not longer kill the parent, if there
> were forked child processes available to kill. So it now should keep
> your bash alive.

Yes, it doesn't kill parent bash. I don't understand the _doubled_ output
in syslog, but maybe you do. Is that related to hyperthreading? ;)
Tested on 2.6.10-rc2-mm2.
oom-killer: gfp_mask=0xd2
DMA per-cpu:
cpu 0 hot: low 2, high 6, batch 1
cpu 0 cold: low 0, high 2, batch 1
Normal per-cpu:
cpu 0 hot: low 32, high 96, batch 16
cpu 0 cold: low 0, high 32, batch 16
HighMem per-cpu:
cpu 0 hot: low 14, high 42, batch 7
cpu 0 cold: low 0, high 14, batch 7

Free pages: 3924kB (112kB HighMem)
Active:129272 inactive:124393 dirty:0 writeback:0 unstable:0 free:981 slab:1945 mapped:253410 pagetables:794
DMA free:68kB min:68kB low:84kB high:100kB active:5372kB inactive:5376kB present:16384kB pages_scanned:11867 all_unreclaimable? yes
protections[]: 0 0 0
Normal free:3744kB min:3756kB low:4692kB high:5632kB active:446904kB inactive:427164kB present:901120kB pages_scanned:889018 all_unreclaimable? yes
protections[]: 0 0 0
HighMem free:112kB min:128kB low:160kB high:192kB active:64812kB inactive:65032kB present:131044kB pages_scanned:132293 all_unreclaimable? yes
protections[]: 0 0 0
DMA: 1*4kB 0*8kB 0*16kB 0*32kB 1*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 68kB
Normal: 0*4kB 0*8kB 0*16kB 1*32kB 0*64kB 1*128kB 0*256kB 1*512kB 1*1024kB 1*2048kB 0*4096kB = 3744kB
HighMem: 0*4kB 0*8kB 1*16kB 1*32kB 1*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 112kB
Swap cache: add 293359, delete 293359, find 104/124, race 0+0
Out of Memory: Killed process 6693 (RNAsubopt).
oom-killer: gfp_mask=0x1d2
DMA per-cpu:
cpu 0 hot: low 2, high 6, batch 1
cpu 0 cold: low 0, high 2, batch 1
Normal per-cpu:
cpu 0 hot: low 32, high 96, batch 16
cpu 0 cold: low 0, high 32, batch 16
HighMem per-cpu:
cpu 0 hot: low 14, high 42, batch 7
cpu 0 cold: low 0, high 14, batch 7

Free pages: 3924kB (112kB HighMem)
Active:134740 inactive:118925 dirty:0 writeback:0 unstable:0 free:981 slab:1928 mapped:253410 pagetables:794
DMA free:68kB min:68kB low:84kB high:100kB active:5368kB inactive:5380kB present:16384kB pages_scanned:13467 all_unreclaimable? yes
protections[]: 0 0 0
Normal free:3744kB min:3756kB low:4692kB high:5632kB active:468780kB inactive:405288kB present:901120kB pages_scanned:933980 all_unreclaimable? yes
protections[]: 0 0 0
HighMem free:112kB min:128kB low:160kB high:192kB active:64812kB inactive:65032kB present:131044kB pages_scanned:137605 all_unreclaimable? yes
protections[]: 0 0 0
DMA: 1*4kB 0*8kB 0*16kB 0*32kB 1*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 68kB
Normal: 0*4kB 0*8kB 0*16kB 1*32kB 0*64kB 1*128kB 0*256kB 1*512kB 1*1024kB 1*2048kB 0*4096kB = 3744kB
HighMem: 0*4kB 0*8kB 1*16kB 1*32kB 1*64kB 0*128kB 0*256kB 0*512kB 0*1024kB 0*2048kB 0*4096kB = 112kB
Swap cache: add 293359, delete 293359, find 104/124, race 0+0
Out of Memory: Killed process 6693 (RNAsubopt).
diff -urN --exclude='*~' --exclude='.#*' 2.6.10-rc2-mm2.orig/mm/oom_kill.c 2.6.10-rc2-mm/mm/oom_kill.c
--- 2.6.10-rc2-mm2.orig/mm/oom_kill.c 2004-11-19 14:52:16.000000000 +0100
+++ 2.6.10-rc2-mm/mm/oom_kill.c 2004-11-22 11:53:16.000000000 +0100
@@ -45,8 +45,10 @@
static unsigned long badness(struct task_struct *p, unsigned long uptime)
{
unsigned long points, cpu_time, run_time, s;
+ struct list_head *tsk;

- if (!p->mm)
+ /* Ignore mm-less tasks and init */
+ if (!p->mm || p->pid == 1)
return 0;

if (p->flags & PF_MEMDIE)
@@ -57,6 +59,19 @@
points = p->mm->total_vm;

/*
+ * Processes which fork a lot of child processes are likely
+ * a good choice. We add the vmsize of the childs if they
+ * have an own mm. This prevents forking servers to flood the
+ * machine with an endless amount of childs
+ */
+ list_for_each(tsk, &p->children) {
+ struct task_struct *chld;
+ chld = list_entry(tsk, struct task_struct, sibling);
+ if (chld->mm != p->mm && chld->mm)
+ points += chld->mm->total_vm;
+ }
+
+ /*
* CPU time is in tens of seconds and run time is in thousands
* of seconds. There is no particular reason for this other than
* that it turned out to work very well in practice.
@@ -176,6 +191,27 @@
return mm;
}

+static struct mm_struct *oom_kill_process(task_t *p)
+{
+ struct mm_struct *mm;
+ struct task_struct *g, *q;
+
+ mm = oom_kill_task(p);
+ if (!mm)
+ return NULL;
+ /*
+ * kill all processes that share the ->mm (i.e. all threads),
+ * but are in a different thread group
+ */
+ do_each_thread(g, q)
+ if (q->mm == mm && q->tgid != p->tgid)
+ __oom_kill_task(q);
+
+ while_each_thread(g, q);
+ if (!p->mm)
+ printk(KERN_INFO "Fixed up OOM kill of mm-less task\n");
+ return mm;
+}

/**
* oom_kill - kill the "best" process when we run out of memory
@@ -188,7 +224,9 @@
void oom_kill(void)
{
struct mm_struct *mm;
- struct task_struct *g, *p, *q;
+ struct task_struct *c, *p;
+ struct list_head *tsk;
+ int mmcnt = 0;

read_lock(&tasklist_lock);
retry:
@@ -200,21 +238,32 @@
panic("Out of memory and no killable processes...\n");
}

- mm = oom_kill_task(p);
- if (!mm)
- goto retry;
/*
- * kill all processes that share the ->mm (i.e. all threads),
- * but are in a different thread group
+ * Kill the forked child processes first
*/
- do_each_thread(g, q)
- if (q->mm == mm && q->tgid != p->tgid)
- __oom_kill_task(q);
- while_each_thread(g, q);
- if (!p->mm)
- printk(KERN_INFO "Fixed up OOM kill of mm-less task\n");
+ list_for_each(tsk, &p->children) {
+ c = list_entry(tsk, struct task_struct, sibling);
+ /* Do not touch threads, as they get killed later */
+ if (c->mm == p->mm)
+ continue;
+ mm = oom_kill_process(c);
+ if (mm) {
+ mmcnt ++;
+ mmput(mm);
+ }
+ }
+
+ /*
+ * If we managed to kill one or more child processes
+ * then let the parent live for now
+ */
+ if (!mmcnt) {
+ mm = oom_kill_process(p);
+ if (!mm)
+ goto retry;
+ mmput(mm);
+ }
read_unlock(&tasklist_lock);
- mmput(mm);
return;
}

@@ -224,14 +273,22 @@
void out_of_memory(int gfp_mask)
{
/*
- * oom_lock protects out_of_memory()'s static variables.
- * It's a global lock; this is not performance-critical.
- */
- static DEFINE_SPINLOCK(oom_lock);
+ * inprogress protects out_of_memory()'s static variables.
+ * We don't use a spin_lock here, as spinlocks are
+ * nops on UP systems.
+ */
+ static unsigned long inprogress;
+ static unsigned int freepages = 1000000;
static unsigned long first, last, count, lastkill;
unsigned long now, since;

- spin_lock(&oom_lock);
+ if (test_and_set_bit(0, &inprogress))
+ return;
+
+ /* Check, if memory was freed since the last oom kill */
+ if (freepages < nr_free_pages())
+ goto out_unlock;
+
now = jiffies;
since = now - last;
last = now;
@@ -271,12 +328,10 @@
* Ok, really out of memory. Kill something.
*/
lastkill = now;
-
printk("oom-killer: gfp_mask=0x%x\n", gfp_mask);
show_free_areas();
-
- /* oom_kill() sleeps */
- spin_unlock(&oom_lock);
+ /* Store free pages * 2 for the check above */
+ freepages = (nr_free_pages() << 1);
oom_kill();
/*
* Make kswapd go out of the way, so "p" has a good chance of
@@ -284,17 +339,11 @@
* for more memory.
*/
yield();
- spin_lock(&oom_lock);

reset:
- /*
- * We dropped the lock above, so check to be sure the variable
- * first only ever increases to prevent false OOM's.
- */
- if (time_after(now, first))
- first = now;
+ first = jiffies;
count = 0;

out_unlock:
- spin_unlock(&oom_lock);
+ clear_bit(0, &inprogress);
}
\
 
 \ /
  Last update: 2005-03-22 14:08    [W:0.168 / U:0.156 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site