Re: Memory leak?

[Damian Lubosch <dl@xxxxxxxx>]

Am 05.09.2010 um 11:33 schrieb Damian Lubosch:

> Hello!
> 
> I just checked the memory status of my new Dragonfly-based server.
> 
> the top command prints out:
> load averages:  0.00,  0.00,  0.00;               up 19+12:15:40                                                                          09:29:28
> 38 processes: 38 running
> 95653tates:  0.0% user,  0.0% nice,  0.0% system,  0.8% interrupt, 99.2% idle
> Memory: 6598M Active, 3152M Inact, 1628M Wired, 186M Cache, 399M Buf, 283M Free

I uploaded the src-directory to http://pick.xiqit.de/src.tgz (so you can see the changes that were made to the sources, but it's about 450MB) 

The last dmesg logs show: (complete dmesg http://pick.xiqit.de/dmesg.log the first lines can be ignored, it was a stresstest made by stress2)
. ..
twa0: tw_cli_scsi_complete: req_id = 0x2f, status = 0x2
twa0: tw_cli_scsi_complete: req_id = 0x30, status = 0x2
twa0: tw_cli_scsi_complete: req_id = 0x31, status = 0x2
CAM: finished configuring all busses (0 left)
no B_DEVMAGIC (bootdev=0)
da0 at twa0 bus 0 target 0 lun 0
da0: <AMCC 9650SE-4LP DISK 4.08> Fixed Direct Access SCSI-5 device 
da0: Serial Number 9VS3WZHT482E2100BCC3
da0: 100.000MB/s transfers
da0: 1430501MB (2929666048 512 byte sectors: 255H 63S/T 182363C)
Mounting root from hammer:serno/9VS3WZHT482E2100BCC3.s1d
tryroot serno/9VS3WZHT482E2100BCC3.s1d
HAMMER(ROOT) recovery check seqno=0596b141
HAMMER(ROOT) recovery range 3000000005569098-3000000005569098
HAMMER(ROOT) recovery nexto 3000000005569098 endseqno=0596b142
HAMMER(ROOT) mounted clean, no recovery needed
Mounting devfs
twa0: INFO: (0x04: 0x0029): Verify started: unit=0
twa0: INFO: (0x04: 0x002B): Verify completed: unit=0
Warning: pmap_interlock 0xfffffffe638e0ee0 0001000d
twa0: INFO: (0x04: 0x0029): Verify started: unit=0
twa0: INFO: (0x04: 0x002B): Verify completed: unit=0
twa0: INFO: (0x04: 0x0029): Verify started: unit=0
twa0: INFO: (0x04: 0x002B): Verify completed: unit=0

I am wondering about the pmap_interlock warning...

Many thanks,
Damian