NetBSD-Users archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: intrctl(8)



On 2019/06/01 0:07, Sad Clouds wrote:
> Hi, I have a rather slow Intel Atom CPU, so when running NFS copy over
> 1GbE, one CPU is quite busy servicing interrupts. So with NetBSD-8
> there is intrctl command:
> 
> # intrctl list
> interrupt id      CPU0  CPU1  CPU2  CPU3  device name(s)
> ioapic0 pin 9        0*    0     0     0  acpi SCI
> ioapic0 pin 1        0*    0     0     0  pckbc1 kbd
> ioapic0 pin 12       0*    0     0     0  pckbc2 aux
> ioapic0 pin 16       0*    0     0     0  i915, uhci3
> msi0 vec 0           0*    0     0     0  hdaudio0
> ioapic0 pin 18 1538867*    0     0     0  unknown, uhci2
> ioapic0 pin 23      50*    0     0     0  uhci0, ehci0
> ioapic0 pin 19       0*    0     0     0  uhci1, ichsmb0
> ioapic0 pin 14  276896*    0     0     0  piixide0 primary
> ioapic0 pin 15       0*    0     0     0  piixide0 secondary
> 
> but trying to use it:
> 
> # intrctl affinity -c 1 -i 18
> intrctl: sysctl kern.intr.affinity: No such file or directory

Use:

	intrctl affinity -c 1 -i "ioapic0 pin 18"

> Any idea? Also is it possible to assign a single interrupt ID to
> multiple CPUs, i.e. load balance it?
> 
> Another issue I'm seeing is that NFS quite quickly grinds to a halt,
> but I've seen this before. Doing "ls" on the local NetBSD file system
> where NFS server is writing, also hangs for some seconds, but
> eventually returns.
> 

-- 
-----------------------------------------------
                SAITOH Masanobu (msaitoh%execsw.org@localhost
                                 msaitoh%netbsd.org@localhost)


Home | Main Index | Thread Index | Old Index