Whilst trying to improve the xrun performance of my Intel Core 2 Duo
machine, I discovered that the hardware interrupts were all being serviced
by CPU0. I also have 64studio on an AMD based laptop. Checking its
interrupts, I found they were shared between the two CPU's. Both computers
are running version 3 beta 3. On both machines, the smp_affinity was set to
3. Changing it value to 2 for interrupt 18 on the Intel machine moved the
audio (ICE1712) interrupts to CPU2. Please see the terminal output below. As
I understand, a setting of 3 (0011) in smp_affinity should allow both
processors to service the interrupt whereas a setting of 2 (0010) only
allows CPU2 to service it. My Intel machine does not appear to be behaving
in that manner.
Testing the audio performance in the altered state may have shown a
reduction in xruns, though the random nature of their occurrence makes it
hard to be certain.
The questions are: Is this behaviour to be expected and; Does it matter.
Terminal output:
r...@64studio:/proc/irq/18# ls
ehci_hcd:usb6 ICE1712 smp_affinity spurious uhci_hcd:usb5
r...@64studio:/proc/irq/18# cat smp_affinity
3
r...@64studio:/proc/irq/18# cat /proc/interrupts
CPU0 CPU1
0: 123 0 IO-APIC-edge timer
1: 301 0 IO-APIC-edge i8042
6: 2 0 IO-APIC-edge floppy
8: 1 0 IO-APIC-edge rtc0
9: 0 0 IO-APIC-fasteoi acpi
12: 28720 0 IO-APIC-edge i8042
16: 0 0 IO-APIC-fasteoi uhci_hcd:usb1
17: 5210 0 IO-APIC-fasteoi pata_marvell
18: 120916 0 IO-APIC-fasteoi uhci_hcd:usb5, ehci_hcd:usb6,
ICE1712
19: 10839 0 IO-APIC-fasteoi uhci_hcd:usb4, ata_piix,
ata_piix
21: 0 0 IO-APIC-fasteoi uhci_hcd:usb2
23: 0 0 IO-APIC-fasteoi uhci_hcd:usb3, ehci_hcd:usb7
30: 224 0 PCI-MSI-edge eth0
NMI: 0 0 Non-maskable interrupts
LOC: 73722 71821 Local timer interrupts
CNT: 0 0 Performance counter interrupts
RES: 100027 185719 Rescheduling interrupts
CAL: 27 27 Function call interrupts
TLB: 943 772 TLB shootdowns
TRM: 0 0 Thermal event interrupts
THR: 0 0 Threshold APIC interrupts
SPU: 0 0 Spurious interrupts
ERR: 0
MIS: 0
r...@64studio:/proc/irq/18# echo 2 > smp_affinity
r...@64studio:/proc/irq/18# cat smp_affinity
2
r...@64studio:/proc/irq/18# cat /proc/interrupts
CPU0 CPU1
0: 124 0 IO-APIC-edge timer
1: 391 0 IO-APIC-edge i8042
6: 2 0 IO-APIC-edge floppy
8: 1 0 IO-APIC-edge rtc0
9: 0 0 IO-APIC-fasteoi acpi
12: 32504 0 IO-APIC-edge i8042
16: 0 0 IO-APIC-fasteoi uhci_hcd:usb1
17: 6176 0 IO-APIC-fasteoi pata_marvell
18: 145478 8917 IO-APIC-fasteoi uhci_hcd:usb5, ehci_hcd:usb6,
ICE1712
19: 11326 0 IO-APIC-fasteoi uhci_hcd:usb4, ata_piix,
ata_piix
21: 0 0 IO-APIC-fasteoi uhci_hcd:usb2
23: 0 0 IO-APIC-fasteoi uhci_hcd:usb3, ehci_hcd:usb7
30: 250 0 PCI-MSI-edge eth0
NMI: 0 0 Non-maskable interrupts
LOC: 80364 86472 Local timer interrupts
CNT: 0 0 Performance counter interrupts
RES: 130294 226216 Rescheduling interrupts
CAL: 28 27 Function call interrupts
TLB: 944 773 TLB shootdowns
TRM: 0 0 Thermal event interrupts
THR: 0 0 Threshold APIC interrupts
SPU: 0 0 Spurious interrupts
ERR: 0
MIS: 0
r...@64studio:/proc/irq/18#
Regards
Dick
_______________________________________________
64studio-devel mailing list
[email protected]
http://lists.64studio.com/mailman/listinfo/64studio-devel
