RELENG_4 on flash disk and swap

Dmitry Pryanishnikov dmitry at
Fri Mar 3 22:25:15 UTC 2006


On Sat, 4 Mar 2006, Peter Jeremy wrote:
>> 4.11-RELEASE) such as "make buildwolrd". After successful completion of
>> this procedure I issued "rm -rf /usr/obj/usr" and got the following
>> (single) message from kernel:
>> Mar  3 11:05:32 test3 /kernel: swap_pager_getswapspace: failed
>> Does anybody know whether it's harmless?
> It depends what you mean by "harmless".  The kernel tried to allocate
> swap space for a process and failed.  The kernel tries to recover by
> killing the largest process (which should also be syslog'd).  I'm

  In my case, not a single process has been killed. And I suppose
that I know why...

> surprised that you got this on the "rm" as the buildworld should
> create bigger processes.  If the "rm" was killed, you will need to
> re-issue it to actually delete the files.

... because I think it wasn't a process which requested a page - it
apparently was a softupdates code. Do you really believe that RELENG_4
lacks real memory for "make buildworld" on i386 with 256Mb RAM? I've issued 
"make buildworld" and "make buildkernel" several times, just to be sure. Every 
such a test was successful. My diagnostics (swap_pager_getswapspace: failed) 
occurs only once per OS run (it doesn't repeat until I reboot by box), and 
even with vm.swap_enabled=0! So I think it's harmless, I just want to confirm 
it w/o digging OS vm code.

Sincerely, Dmitry
Atlantis ISP, System Administrator
e-mail:  dmitry at
nic-hdl: LYNX-RIPE

More information about the freebsd-stable mailing list