CURRENT Kernel makes the system run very very hot

[Thomas Sparrevohn]

On Saturday 08 December 2007 17:48:14 韓家標 Bill Hacker wrote:
> Thomas Sparrevohn wrote:
> > Hi
> > 
> > There something weird going on - at minimal workloads my system gets very
> > very hot - The system is watercooled, 4 Fan's etc its a quad core QX6700 -
> > make buildkernel - will make the fans run at highest speed (its impossibly to
> > be in the room at the same time).
> > 
> > There are no problems when running other OS'es - Are anybody else having this
> > kind of problems  (PS its a relatively new thing - maybe 2 weeks)? 
> > _______________________________________________ freebsd-current at freebsd.org
> > mailing list http://lists.freebsd.org/mailman/listinfo/freebsd-current To
> > unsubscribe, send any mail to "freebsd-current-unsubscribe at freebsd.org"
> > 
> 
> And 'top -S' reports what?
> 

As I said its relatively new - I am running root on ZFS (and everything else) and 
had a reproduceable deadlock when I was using ZFS of Nvidia ATARAID so I 
dropped the ATA RAID and are now running ZFS on the raw disks - It started occuring some
point after that (last two weeks). 

The funny thing is if I break into DDB and wait app. 30 Sec the fans starts spinning up and 
running at very high speeds - e.g nothing should be running

Top -S -H -c 1

last pid:  1334;  load averages:  0.04,  0.17,  0.13  up 0+00:08:50    17:22:28
207 processes: 6 running, 183 sleeping, 18 waiting

Mem: 251M Active, 21M Inact, 475M Wired, 13M Cache, 592K Buf, 3155M Free
Swap: 32G Total, 32G Free


  PID USERNAME   PRI NICE   SIZE    RES STATE  C   TIME   WCPU COMMAND
   11 root       171 ki31     0K    64K RUN    0  11:12 100.00% {idle: cpu0}
   11 root       171 ki31     0K    64K RUN    2  11:12 100.00% {idle: cpu2}
   11 root       171 ki31     0K    64K CPU1   1  11:12 100.00% {idle: cpu1}
   11 root       171 ki31     0K    64K CPU3   3  11:12 99.37% {idle: cpu3}
 1105 root        46    0   100M 51516K select 3   0:07  1.95% Xorg
   12 root       -24    -     0K   288K WAIT   0   0:12  0.00% {swi6: Giant tas
   12 root       -24    -     0K   288K WAIT   1   0:12  0.00% {swi6: task queu
   12 root       -28    -     0K   288K WAIT   0   0:12  0.00% {swi5: +}
   12 root       -36    -     0K   288K WAIT   0   0:12  0.00% {swi3: vm}
   12 root       -40    -     0K   288K WAIT   1   0:12  0.00% {swi2: cambio}
   12 root       -44    -     0K   288K WAIT   1   0:12  0.00% {swi1: net}
   12 root       -48    -     0K   288K WAIT   0   0:12  0.00% {swi0: sio}
   12 root       -52    -     0K   288K WAIT   0   0:12  0.00% {irq9: acpi0}
   12 root       -60    -     0K   288K WAIT   0   0:12  0.00% {irq1: atkbd0}
   12 root       -60    -     0K   288K WAIT   3   0:12  0.00% {irq12: psm0}
   12 root       -64    -     0K   288K WAIT   0   0:12  0.00% {irq21: ohci0+}
   12 root       -64    -     0K   288K WAIT   0   0:12  0.00% {irq20: atapci2}
   12 root       -64    -     0K   288K WAIT   2   0:12  0.00% {irq23: atapci1}


Sysctl dev.cpu | grep temp - shows

dev.cpu.0.temperature: 70
dev.cpu.1.temperature: 71
dev.cpu.2.temperature: 68
dev.cpu.3.temperature: 69

I was trying to avoid to go back to running "pure UFS" again - but I guess I have to inorder
to eliminate ZFS as the potential source


> FWIW - the opposite experience here.
> Core-2 Quad QX6600 on Asus and Gigabyte, Core-D dual on Tyan.
> 
> Stock Intel cooler & fans, case fans on variable-speed MB connectors, PSU's with 
> 'smart' fans.
> 
> Fans run at nowhere near full-speed even with a long make -j 12.
> 
> Under 'average load', not yet as quiet as my PowerBook G4, but getting close to 
> our C3 MB with constant-speed fans.
> 
> Something in the cooling water?  Bacteria? Goldfish turds?
> 
> Bill
> 
>