Re: SMP interrupt wierdness in Linux 2.6.8.1 - Interrupt Load Balancing not working

NOULARD Eric <[email protected]> Sun, 07 Nov 2004 14:12:03 +0100
Newsgroups org.kernel.vger.linux-smp
Message-ID <[email protected]>
I just compiled and booted 2.6.9
the result is almost identical interrupt seems to 
goes only to CPU0 (excepted some of them ??)

My kernel config says:
CONFIG_IRQBALANCE=y


> cat /proc/interrupts
           CPU0       CPU1
  0:    6350285         14    IO-APIC-edge  timer
  1:       8647          0    IO-APIC-edge  i8042
  2:          0          0          XT-PIC  cascade
  5:     166414          0    IO-APIC-edge  SoundBlaster
  7:          1          1    IO-APIC-edge  parport0
  8:          0          1    IO-APIC-edge  rtc
 12:     176710          0    IO-APIC-edge  i8042
 16:     375735          0   IO-APIC-level  r128@PCI:1:0:0
 17:         29          1   IO-APIC-level  aic7xxx
 18:     110899          1   IO-APIC-level  sym53c8xx, sym53c8xx
 19:      26444          0   IO-APIC-level  uhci_hcd, eth0
NMI:          0          0
LOC:    6350405    6350423
ERR:          0
MIS:          0


all /proc/irq/<irqno>/smp_affinity mask are set to 'ffffffff'?

I do not have any idea why this happen?


SideNote: My preceding irq sharing trouble disappeared when I switch
          from MPS 1.1 to MPS 1.4 support in BIOS.

Le mar 02/11/2004 à 20:47, NOULARD Eric a écrit :
> I just have almost the same config ASUS-P2BD with 2xPIII 450MHz
> with the same symptoms, minus the running processes problem,
> plus one more suspect IRQ9 shared by most of PCI devices???
> 
> I should give 2.6.9 a try with more careful config.
> 
> Did you tried 2.6.9?
> Did you have the shared IRQ trouble?
> 
> Note that in my case I have effectively
> 2 runnings processes at a time.
> Using top did you see the load of the 2 distinct CPU like this.
> 
> My favourite test to check if I am effectively using both CPU
> is to run a simple multi-threaded C program that try to put
> one thread to SCHED_FIFO with max priority if your machine
> booted with one proc or could only use one as soon as you run
> the program your box will apparently freeze until the 
> SCHED_FIFO thread terminates.
> 
> If your box is not freezing you will be able to see
> (with top for example) that 1 processor is 100% used
> by the FIFO process.
> 
> 
> ----------------------
> top - 20:41:54 up  1:14,  8 users,  load average: 0.39, 0.36, 0.32
> Tasks: 106 total,   3 running, 102 sleeping,   0 stopped,   1 zombie
>  Cpu0 : 22.0% us,  3.6% sy,  0.0% ni, 74.0% id,  0.0% wa,  0.4% hi, 0.0%
> si
>  Cpu1 : 22.5% us,  3.8% sy,  0.0% ni, 73.8% id,  0.0% wa,  0.0% hi, 
> 0.0% si
> Mem:    449300k total,   416836k used,    32464k free,    45004k buffers
> 
> -----------------------
> 
> cat /proc/interrupts
>            CPU0       CPU1
>   0:    4068279         12    IO-APIC-edge  timer
>   1:       9892          0    IO-APIC-edge  i8042
>   2:          0          0          XT-PIC  cascade
>   5:       5344          1    IO-APIC-edge  SoundBlaster
>   7:          1          1    IO-APIC-edge  parport0
>   8:          1          0    IO-APIC-edge  rtc
>   9:     229563          1   IO-APIC-level  sym53c8xx, sym53c8xx,
> uhci_hcd, aic7xxx, eth0 11:     238198          1   IO-APIC-level  
> r128@PCI:1:0:0
>  12:     119151          0    IO-APIC-edge  i8042
> NMI:          0          0
> LOC:    4068215    4068222
> ERR:          0
> MIS:          0
> 
> 
> Le ven 29/10/2004 à 10:07, Lars Michael Jogbäck a écrit :
> > Hello,
> > 
> > I've recently upgraded my SMP machine (2xPIII-500 on Asus P2B-D)
> > from Linux 2.4 to Linux 2.6.8.1 and now it seem's that the machine
> > only routes interrupts to one CPU. When loading the machine and
> > running top, I still only have seen 1 process in "Running" state
> > at the time. Should this be 2?
> > 
> > I have CONFIG_IRQBALANCE=y and of course CONFIG_SMP=y
> > 
> > lm@cleopatra:~$ cat /proc/interrupts
> >            CPU0       CPU1
> >   0:   36751142         16    IO-APIC-edge  timer
> >   1:          9          1    IO-APIC-edge  i8042
> >   2:          0          0          XT-PIC  cascade
> >   8:          3          0    IO-APIC-edge  rtc
> >  14:     273037          1    IO-APIC-edge  ide0
> >  15:       1189          0    IO-APIC-edge  ide1
> >  16:       8561          1   IO-APIC-level  sym53c8xx
> >  17:      40698          1   IO-APIC-level  sym53c8xx
> >  18:     110871          8   IO-APIC-level  ide2, ide3, ide4, ide5
> >  19:     219831          0   IO-APIC-level  eth0
> > NMI:          0          0
> > LOC:   36752415   36752416
> > ERR:          0
> > MIS:        245
> > 
> > When running 2.4 the interrupts was distributed between the two CPUs.
> > 
> > The CPU's are indentical:
> > lm@cleopatra:~$ cat /proc/cpuinfo
> > processor       : 0
> > vendor_id       : GenuineIntel
> > cpu family      : 6
> > model           : 7
> > model name      : Pentium III (Katmai)
> > stepping        : 3
> > cpu MHz         : 501.328
> > cache size      : 512 KB
> > fdiv_bug        : no
> > hlt_bug         : no
> > f00f_bug        : no
> > coma_bug        : no
> > fpu             : yes
> > fpu_exception   : yes
> > cpuid level     : 2
> > wp              : yes
> > flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
> > mca cmov pat pse36 mmx fxsr sse
> > bogomips        : 989.18
> > 
> > processor       : 1
> > vendor_id       : GenuineIntel
> > cpu family      : 6
> > model           : 7
> > model name      : Pentium III (Katmai)
> > stepping        : 3
> > cpu MHz         : 501.328
> > cache size      : 512 KB
> > fdiv_bug        : no
> > hlt_bug         : no
> > f00f_bug        : no
> > coma_bug        : no
> > fpu             : yes
> > fpu_exception   : yes
> > cpuid level     : 2
> > wp              : yes
> > flags           : fpu vme de pse tsc msr pae mce cx8 apic sep mtrr pge
> > mca cmov pat pse36 mmx fxsr sse
> > bogomips        : 999.42
> > 
> > Is there anyone out there who have an idea on where to start looking?
> > 
> > Best Regards,
> > /LM
> > 
> > 
> > 
> > -
> > To unsubscribe from this list: send the line "unsubscribe linux-smp" in
> > the body of a message to [email protected]
> > More majordomo info at  http://vger.kernel.org/majordomo-info.html
-- 
Eric NOULARD
E-mail: [email protected]