Using taskset -c 1 with opencl program,it runs at the same speed that without use of it in four core machine. But the load with taskset is 13% without it 50%.
I am using AMD drivers 13.12 in fedora 20, after too much problems (amd r9 opencl not supported by fedora and AMD drivers does not compile, requires third part patch).
The system has 3 GPUs and I trying to assign interrupts of one GPU to one CPU core. The problem is that I can not echo anything to smp_affinity in /proc/irq/60/
CPU0 CPU1 CPU2 CPU3
0: 125 0 0 0 IO-APIC-edge timer
1: 3 3 2 2 IO-APIC-edge i8042
7: 1 0 0 0 IO-APIC-edge
8: 0 0 1 0 IO-APIC-edge rtc0
9: 0 0 0 0 IO-APIC-fasteoi acpi
14: 20553 19951 18934 23545 IO-APIC-edge pata_atiixp
15: 0 0 0 0 IO-APIC-edge pata_atiixp
16: 209 210 333 254 IO-APIC-fasteoi snd_hda_intel
17: 0 0 0 0 IO-APIC-fasteoi ehci_hcd:usb1, ehci_hcd:usb2
18: 1 1 1 0 IO-APIC-fasteoi ohci_hcd:usb3, ohci_hcd:usb4, ohci_hcd:usb5
40: 0 0 0 0 PCI-MSI-edge ahci
41: 4 4 3 5 PCI-MSI-edge xhci_hcd
42: 0 0 0 0 PCI-MSI-edge xhci_hcd
43: 0 0 0 0 PCI-MSI-edge xhci_hcd
44: 0 0 0 0 PCI-MSI-edge xhci_hcd
45: 0 0 0 0 PCI-MSI-edge xhci_hcd
46: 0 0 0 0 PCI-MSI-edge xhci_hcd
47: 0 0 0 0 PCI-MSI-edge xhci_hcd
48: 0 0 0 0 PCI-MSI-edge xhci_hcd
49: 0 0 0 0 PCI-MSI-edge xhci_hcd
50: 0 0 0 0 PCI-MSI-edge xhci_hcd
51: 0 0 0 0 PCI-MSI-edge xhci_hcd
52: 0 0 0 0 PCI-MSI-edge xhci_hcd
53: 0 0 0 0 PCI-MSI-edge xhci_hcd
54: 0 0 0 0 PCI-MSI-edge xhci_hcd
55: 0 0 0 0 PCI-MSI-edge xhci_hcd
56: 260453 27 39686 10222 PCI-MSI-edge p19p1
57: 30 29 28 31 PCI-MSI-edge snd_hda_intel
58: 30 29 29 30 PCI-MSI-edge snd_hda_intel
59: 30 30 29 29 PCI-MSI-edge snd_hda_intel
60: 376820 387077 422262 414046 PCI-MSI-edge fglrx[0]@PCI:1:0:0
61: 233781 260237 282034 273928 PCI-MSI-edge fglrx[1]@PCI:2:0:0
62: 325068 337322 385699 381124 PCI-MSI-edge fglrx[2]@PCI:3:0:0
NMI: 109 466 134 132 Non-maskable interrupts
LOC: 1455844 4537211 1861268 1830786 Local timer interrupts
SPU: 0 0 0 0 Spurious interrupts
PMI: 109 466 134 132 Performance monitoring interrupts
IWI: 48127 24519 17582 13174 IRQ work interrupts
RTR: 0 0 0 0 APIC ICR read retries
RES: 33436070 33287609 27497110 27413290 Rescheduling interrupts
CAL: 1727 1523 1401 1314 Function call interrupts
TLB: 416983 448396 436569 455603 TLB shootdowns
TRM: 0 0 0 0 Thermal event interrupts
THR: 0 0 0 0 Threshold APIC interrupts
MCE: 0 0 0 0 Machine check exceptions
MCP: 18 18 18 18 Machine check polls
Linux * 3.12.6-300.fc20.x86_64 #1 SMP Mon Dec 23 16:44:31 UTC 2013 x86_64 x86_64 x86_64 GNU/Linux
display: :0.0 screen: 0 OpenGL vendor string: Advanced Micro Devices, Inc. OpenGL renderer string: AMD Radeon R9 200 Series OpenGL version string: 4.3.12618 Compatibility Profile Context 13.251 display: :0.0 screen: 1 OpenGL vendor string: Advanced Micro Devices, Inc. OpenGL renderer string: AMD Radeon R9 200 Series OpenGL version string: 4.3.12618 Compatibility Profile Context 13.251 display: :0.0 screen: 2 OpenGL vendor string: Advanced Micro Devices, Inc. OpenGL renderer string: AMD Radeon R9 200 Series OpenGL version string: 4.3.12618 Compatibility Profile Context 13.251
It's possible to assign IRQ 60 to core 0, 61 to 1 and 62 to 2 ? Or reduce CPU load.