Re: scsi cmd slab leak? (Was Re: [ck] Anyone been having OOM killer problems lately?)



On Thursday 19 January 2006 12:13, scientica (GMail) wrote:
> Just out of curiosity, what's the slab? and what is the expected size
> of it? I just checked mine and it seems to eat some 304596 kB
> (2.6.14-ck4, soon 68d uptime). The only problems I've had recently is
> firefox crashing more than it should (but it could simply be me having
> a billion or so windows and tabs open, and it's the
> mozilla-firefox-bin-1.5-r2 from portage which is masked ~amd64, so
> it's probably just buggy - it dies with a segfault after beeing stuck
> at 100% CPU for a while, cant see any OOM-messages anywhere), other
> than that I've been able to both emerge stuff (nice'd though),
> download stuff and burn backups DVD's with out problems -
> simultaneously - and the system was still responsive :)

The slab layer in the kernel is an algorithm that attempts to reserve a sane
amount of memory for a given highly-used data structure in the kernel. By
using the slab layer to keep memory reserved and ready, performance-critical
sections of the kernel code (say, code that receives a packet) doesn't have
to stop and succeed an allocation before continuing.

Cheers,
Chase
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/



Relevant Pages

  • Re: [RFC: 2.6 patch] let GROUP_SCHED depend on BROKEN
    ... kernel developers. ... you've never before written a _single_ Linux kernel feature ... have no desire _at all_ to write true kernel code. ... the level of trivialities. ...
    (Linux-Kernel)
  • [PATCH] Info dump on Oops or panic()
    ... panics. ... I accidentally built my testing kernel without the IDE drivers. ... and informed me that my uptime was 0.38 seconds. ... * Dump out uptime info on oops / panic. ...
    (Linux-Kernel)
  • Re: [patch 4/4] KVM-trace port to tracepoints
    ... Tracepoints allow dormat instrumentation, like the kernel markers, but also ... This patch depends on the "Tracepoints" patch. ... Is it a specific property of KVM-trace that causes this LOC blow-up? ... strings into the kernel code, which, to some, looks like debugging code. ...
    (Linux-Kernel)
  • Re: Just as a matter of interest.
    ... well switch to Windows. ... So the best thing is to go what kernel updates have been ... uptime so my paltry 400 days on a private server seemes rather modest. ...
    (alt.os.linux.suse)
  • netgraph/ng_eiface double panic & turnstile/sio lock order reversal in 5.2-BETA
    ... page fault while in kernel mode ... Uptime: 2m3s ... panic: msleep ... <ACPI PCI bus> on pcib0 ...
    (freebsd-current)