* Wrong Vmalloc numbers in /proc/meminfo
@ 2009-09-28 0:09 Frans Pop
2009-09-28 2:32 ` KAMEZAWA Hiroyuki
0 siblings, 1 reply; 4+ messages in thread
From: Frans Pop @ 2009-09-28 0:09 UTC (permalink / raw)
To: linux-kernel
With 2.6.31 (on x86_64) I'm seeing the following in /proc/meminfo.
$ cat /proc/meminfo
MemTotal: 2030968 kB
MemFree: 539408 kB
Buffers: 94984 kB
Cached: 308564 kB
SwapCached: 16048 kB
Active: 652236 kB
Inactive: 428808 kB
Active(anon): 436808 kB
Inactive(anon): 246996 kB
Active(file): 215428 kB
Inactive(file): 181812 kB
Unevictable: 1632 kB
Mlocked: 1632 kB
SwapTotal: 2097144 kB
SwapFree: 1979524 kB
Dirty: 12 kB
Writeback: 0 kB
AnonPages: 671276 kB
Mapped: 57684 kB
Slab: 177692 kB
SReclaimable: 160516 kB
SUnreclaim: 17176 kB
PageTables: 14812 kB
NFS_Unstable: 0 kB
Bounce: 0 kB
WritebackTmp: 0 kB
CommitLimit: 3112628 kB
Committed_AS: 988780 kB
VmallocTotal: 34359738367 kB
VmallocUsed: 340084 kB
VmallocChunk: 34359387131 kB
DirectMap4k: 2008768 kB
DirectMap2M: 63488 kB
Is it me or are VmallocTotal and VmallocChunk off by a factor 10,000 or so?
Cheers,
FJP
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Wrong Vmalloc numbers in /proc/meminfo
2009-09-28 0:09 Wrong Vmalloc numbers in /proc/meminfo Frans Pop
@ 2009-09-28 2:32 ` KAMEZAWA Hiroyuki
2009-09-28 16:56 ` Frans Pop
0 siblings, 1 reply; 4+ messages in thread
From: KAMEZAWA Hiroyuki @ 2009-09-28 2:32 UTC (permalink / raw)
To: Frans Pop; +Cc: linux-kernel
On Mon, 28 Sep 2009 02:09:30 +0200
Frans Pop <elendil@planet.nl> wrote:
> With 2.6.31 (on x86_64) I'm seeing the following in /proc/meminfo.
>
> $ cat /proc/meminfo
> MemTotal: 2030968 kB
> MemFree: 539408 kB
> Buffers: 94984 kB
> Cached: 308564 kB
> SwapCached: 16048 kB
> Active: 652236 kB
> Inactive: 428808 kB
> Active(anon): 436808 kB
> Inactive(anon): 246996 kB
> Active(file): 215428 kB
> Inactive(file): 181812 kB
> Unevictable: 1632 kB
> Mlocked: 1632 kB
> SwapTotal: 2097144 kB
> SwapFree: 1979524 kB
> Dirty: 12 kB
> Writeback: 0 kB
> AnonPages: 671276 kB
> Mapped: 57684 kB
> Slab: 177692 kB
> SReclaimable: 160516 kB
> SUnreclaim: 17176 kB
> PageTables: 14812 kB
> NFS_Unstable: 0 kB
> Bounce: 0 kB
> WritebackTmp: 0 kB
> CommitLimit: 3112628 kB
> Committed_AS: 988780 kB
> VmallocTotal: 34359738367 kB
> VmallocUsed: 340084 kB
> VmallocChunk: 34359387131 kB
> DirectMap4k: 2008768 kB
> DirectMap2M: 63488 kB
>
> Is it me or are VmallocTotal and VmallocChunk off by a factor 10,000 or so?
>
I'm sorry I misunderstand your 10,000 implies.
On my x86-64 box.
VmallocTotal: 34359738367 kB
VmallocUsed: 100856 kB
VmallocChunk: 34359583863 kB
Then,
34359738367 - 34359583863
154504
Then difference is 13412 (pages)
Hmm. in /proc/vmallocinfo
==
..
0xffffc9000652b000-0xffffc9000672c000 2101248 sys_swapon+0x6a2/0xcd0 pages=512 vmalloc N0=512
0xffffc90007efd000-0xffffc90008dde000 15601664 pcpu_alloc+0x3ad/0x4e0 vmalloc
0xffffc900096d9000-0xffffc900096dd000 16384 e1000e_setup_tx_resources+0x39/0xe0 [e1000e] pages=3 vmalloc N0=3
0xffffc900096de000-0xffffc900096e2000 16384 e1000e_setup_rx_resources+0x37/0x160 [e1000e] pages=3 vmalloc N0=3
==
It seems some amount of memory holes (rather than PAGESIZE) between vmalloc area
info.(especially around pcpu_alloc..) please check /proc/vmallocinfo if you
feel something strange.
Thanks,
-Kame
> Cheers,
> FJP
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Wrong Vmalloc numbers in /proc/meminfo
2009-09-28 2:32 ` KAMEZAWA Hiroyuki
@ 2009-09-28 16:56 ` Frans Pop
2009-09-28 23:59 ` KAMEZAWA Hiroyuki
0 siblings, 1 reply; 4+ messages in thread
From: Frans Pop @ 2009-09-28 16:56 UTC (permalink / raw)
To: KAMEZAWA Hiroyuki; +Cc: linux-kernel
On Monday 28 September 2009, KAMEZAWA Hiroyuki wrote:
> > VmallocTotal: 34359738367 kB
> > VmallocUsed: 340084 kB
> > VmallocChunk: 34359387131 kB
> >
> > Is it me or are VmallocTotal and VmallocChunk off by a factor 10,000
> > or so?
>
> I'm sorry I misunderstand your 10,000 implies.
What I meant is: is the Vmalloc area really 32 *terra*bytes in size?
Seems rather big for a system with only 2GB RAM.
I'd never noticed it before and it looks strange to me, but I guess it's
just the result of having 64-bit addressing: a theoretically addressable
area most of which will never actually be used?
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: Wrong Vmalloc numbers in /proc/meminfo
2009-09-28 16:56 ` Frans Pop
@ 2009-09-28 23:59 ` KAMEZAWA Hiroyuki
0 siblings, 0 replies; 4+ messages in thread
From: KAMEZAWA Hiroyuki @ 2009-09-28 23:59 UTC (permalink / raw)
To: Frans Pop; +Cc: linux-kernel
On Mon, 28 Sep 2009 18:56:08 +0200
Frans Pop <elendil@planet.nl> wrote:
> On Monday 28 September 2009, KAMEZAWA Hiroyuki wrote:
> > > VmallocTotal: 34359738367 kB
> > > VmallocUsed: 340084 kB
> > > VmallocChunk: 34359387131 kB
> > >
> > > Is it me or are VmallocTotal and VmallocChunk off by a factor 10,000
> > > or so?
> >
> > I'm sorry I misunderstand your 10,000 implies.
>
> What I meant is: is the Vmalloc area really 32 *terra*bytes in size?
> Seems rather big for a system with only 2GB RAM.
>
VmallocTotal means "available address space size for VMALLOC"
It's fixed size as VMALLOC_END - VMALLOC_START.
VmallocChunk just means "free space in VMALLOC area".
Both of them are not related to size of RAM.
> I'd never noticed it before and it looks strange to me, but I guess it's
> just the result of having 64-bit addressing: a theoretically addressable
> area most of which will never actually be used?
>
you're right. Most of space will never used.
Regards,
-Kame
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2009-09-29 0:01 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2009-09-28 0:09 Wrong Vmalloc numbers in /proc/meminfo Frans Pop
2009-09-28 2:32 ` KAMEZAWA Hiroyuki
2009-09-28 16:56 ` Frans Pop
2009-09-28 23:59 ` KAMEZAWA Hiroyuki
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.