Hi All!
Still no nmi_watchdog on dual Athlon systems?
Linux version 2.4.22-rc2 on dual
processor : 0
vendor_id : AuthenticAMD
cpu family : 6
model : 6
model name : AMD Athlon(tm) MP 1800+
stepping : 2
cpu MHz : 1533.401
cache size : 256 KB
Total of 2 processors activated (6127.61 BogoMIPS). ENABLING IO-APIC IRQs ..TIMER: vector=0x31 pin1=2 pin2=0 activating NMI Watchdog ... done. testing NMI watchdog ... CPU#0: NMI appears to be stuck! testing the IO APIC....................... .................................... done. Using local APIC timer interrupts. calibrating APIC timer ...
just for curiosity, if I do not give the nmi_watchdog=1
option on the kernel boot line, the nmi seems to work fine?
# cat /proc/interrupts
CPU0 CPU1
0: 4275 2602 IO-APIC-edge timer
1: 0 2 IO-APIC-edge keyboard
2: 0 0 XT-PIC cascade
4: 157 150 IO-APIC-edge serial
8: 1 0 IO-APIC-edge rtc
9: 169 186 IO-APIC-level acpi
15: 12 3 IO-APIC-edge ide1
16: 1195 1155 IO-APIC-level aic7xxx
17: 149 144 IO-APIC-level eth1
19: 1 3 IO-APIC-level eth0
NMI: 4034 3577
LOC: 6782 6777
ERR: 0
MIS: 0
is this intentional, or have I missed something
terrible important? if you need more information,
let me know ...
TIA,
Herbert
It seems NMI watchdog doesn't work on nforce2 (Athlon) chipset boards when APIC
is enabled.
I don't know if it works if APIC is disabled and also don't know
about other boards.
Cl<
On Sun, Aug 17, 2003 at 11:28:24PM +0200, Herbert P?tzl wrote:
>
> Hi All!
>
> Still no nmi_watchdog on dual Athlon systems?
>
> Linux version 2.4.22-rc2 on dual
>
> processor : 0
> vendor_id : AuthenticAMD
> cpu family : 6
> model : 6
> model name : AMD Athlon(tm) MP 1800+
> stepping : 2
> cpu MHz : 1533.401
> cache size : 256 KB
>
> Total of 2 processors activated (6127.61 BogoMIPS). ENABLING IO-APIC IRQs ..TIMER: vector=0x31 pin1=2 pin2=0 activating NMI Watchdog ... done. testing NMI watchdog ... CPU#0: NMI appears to be stuck! testing the IO APIC....................... .................................... done. Using local APIC timer interrupts. calibrating APIC timer ...
>
>
> just for curiosity, if I do not give the nmi_watchdog=1
> option on the kernel boot line, the nmi seems to work fine?
>
> # cat /proc/interrupts
> CPU0 CPU1
> 0: 4275 2602 IO-APIC-edge timer
> 1: 0 2 IO-APIC-edge keyboard
> 2: 0 0 XT-PIC cascade
> 4: 157 150 IO-APIC-edge serial
> 8: 1 0 IO-APIC-edge rtc
> 9: 169 186 IO-APIC-level acpi
> 15: 12 3 IO-APIC-edge ide1
> 16: 1195 1155 IO-APIC-level aic7xxx
> 17: 149 144 IO-APIC-level eth1
> 19: 1 3 IO-APIC-level eth0
> NMI: 4034 3577
> LOC: 6782 6777
> ERR: 0
> MIS: 0
>
> is this intentional, or have I missed something
> terrible important? if you need more information,
> let me know ...
>
> TIA,
> Herbert
>
>
>
> -
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to [email protected]
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
On Sun, Aug 17, 2003 at 11:28:24PM +0200, Herbert P?tzl wrote:
>
> Hi All!
>
> Still no nmi_watchdog on dual Athlon systems?
Hi !
mine works fine only with nmi_watchdog=2. Don't know why. It's an ASUS A7M266D.
Cheers,
Willy
On Mon, Aug 18, 2003 at 01:40:00AM +0200, Karel Kulhav? wrote:
> It seems NMI watchdog doesn't work on nforce2 (Athlon) chipset boards when APIC
> is enabled.
hmm, this is on a
- Tyan Tiger MPX S2466N-4M
- AMD Dual 1.8GHZ Athlon
00:00.0 Host bridge: Advanced Micro Devices [AMD]: Unknown device 700c (rev 11)
00:01.0 PCI bridge: Advanced Micro Devices [AMD]: Unknown device 700d
00:07.0 ISA bridge: Advanced Micro Devices [AMD]: Unknown device 7440 (rev 05)
isn't nForce2 from nVIDIA ?
> I don't know if it works if APIC is disabled and also don't know
> about other boards.
it works perfectly, only the nmi_watchdog cannot be
enabled ...
thanks,
Herbert
> On Sun, Aug 17, 2003 at 11:28:24PM +0200, Herbert P?tzl wrote:
> >
> > Hi All!
> >
> > Still no nmi_watchdog on dual Athlon systems?
> >
> > Linux version 2.4.22-rc2 on dual
> >
> > processor : 0
> > vendor_id : AuthenticAMD
> > cpu family : 6
> > model : 6
> > model name : AMD Athlon(tm) MP 1800+
> > stepping : 2
> > cpu MHz : 1533.401
> > cache size : 256 KB
> >
> > Total of 2 processors activated (6127.61 BogoMIPS). ENABLING IO-APIC IRQs ..TIMER: vector=0x31 pin1=2 pin2=0 activating NMI Watchdog ... done. testing NMI watchdog ... CPU#0: NMI appears to be stuck! testing the IO APIC....................... .................................... done. Using local APIC timer interrupts. calibrating APIC timer ...
> >
> >
> > just for curiosity, if I do not give the nmi_watchdog=1
> > option on the kernel boot line, the nmi seems to work fine?
> >
> > # cat /proc/interrupts
> > CPU0 CPU1
> > 0: 4275 2602 IO-APIC-edge timer
> > 1: 0 2 IO-APIC-edge keyboard
> > 2: 0 0 XT-PIC cascade
> > 4: 157 150 IO-APIC-edge serial
> > 8: 1 0 IO-APIC-edge rtc
> > 9: 169 186 IO-APIC-level acpi
> > 15: 12 3 IO-APIC-edge ide1
> > 16: 1195 1155 IO-APIC-level aic7xxx
> > 17: 149 144 IO-APIC-level eth1
> > 19: 1 3 IO-APIC-level eth0
> > NMI: 4034 3577
> > LOC: 6782 6777
> > ERR: 0
> > MIS: 0
> >
> > is this intentional, or have I missed something
> > terrible important? if you need more information,
> > let me know ...
> >
> > TIA,
> > Herbert
On Mon, Aug 18, 2003 at 12:11:14AM +0200, Willy Tarreau wrote:
> On Sun, Aug 17, 2003 at 11:28:24PM +0200, Herbert P?tzl wrote:
> >
> > Hi All!
> >
> > Still no nmi_watchdog on dual Athlon systems?
>
> Hi !
>
> mine works fine only with nmi_watchdog=2. Don't know why.
> It's an ASUS A7M266D.
hmm, nmi_watchdog=2 on the kernel boot line gives no
difference to booting without, at least according to
the boot messages ...
ENABLING IO-APIC IRQs
..TIMER: vector=0x31 pin1=2 pin2=0
testing the IO APIC.......................
.................................... done.
Using local APIC timer interrupts.
calibrating APIC timer ...
..... CPU clock speed is 1533.4487 MHz.
maybe the nmi_watchdog is always enabled?
maybe it only fails with nmi_watchdog=1 ?
shouldn't there be a message which says that
the NMI watchdog was enabled?
TIA,
Herbert
> Cheers,
> Willy
On Mon, 18 Aug 2003, Herbert P?tzl wrote:
> > mine works fine only with nmi_watchdog=2. Don't know why.
> > It's an ASUS A7M266D.
>
> hmm, nmi_watchdog=2 on the kernel boot line gives no
> difference to booting without, at least according to
> the boot messages ...
nmi_watchdog=2 will work on the majority of i686+ (performance
counters with NMI delivery mode) boxes and you can check whether it's
enabled by doing cat /proc/interrupts and watching if the NMI line ticks
at a decent rate. nmi_watchdog=1 tends to be harder for hardware
manufacturers to get right (for some reason or other).
On Sun, Aug 17, 2003 at 06:28:46PM -0400, Zwane Mwaikambo wrote:
> On Mon, 18 Aug 2003, Herbert P?tzl wrote:
>
> > > mine works fine only with nmi_watchdog=2. Don't know why.
> > > It's an ASUS A7M266D.
> >
> > hmm, nmi_watchdog=2 on the kernel boot line gives no
> > difference to booting without, at least according to
> > the boot messages ...
>
> nmi_watchdog=2 will work on the majority of i686+ (performance
> counters with NMI delivery mode) boxes and you can check whether it's
> enabled by doing cat /proc/interrupts and watching if the NMI line ticks
okay, but this would mean that the nmi_watchdog is
enabled if I do not specify nmi_watchdog at the kernel
boot/command line ... or how should I interpret the
steadily increasing NMI counts in this case?
don't get me wrong, I'm happy if nmi_watchdog is
enabled by default, but I would like to know/verify
that ...
> at a decent rate. nmi_watchdog=1 tends to be harder for hardware
> manufacturers to get right (for some reason or other).
no nmi_watchdog and nmi_watchdog=2, both result in
increasing NMI counts, nmi_watchdog=1, only in LOC
counts ...
# cat /proc/interrupts
CPU0 CPU1
NMI: 75378 74923
LOC: 159900 159896
ERR: 0
MIS: 0
TIA,
Herbert
Zwane Mwaikambo wrote:
> > > mine works fine only with nmi_watchdog=2. Don't know why.
> > > It's an ASUS A7M266D.
> >
> > hmm, nmi_watchdog=2 on the kernel boot line gives no
> > difference to booting without, at least according to
> > the boot messages ...
>
> nmi_watchdog=2 will work on the majority of i686+ (performance
> counters with NMI delivery mode) boxes and you can check whether it's
> enabled by doing cat /proc/interrupts and watching if the NMI line ticks
> at a decent rate. nmi_watchdog=1 tends to be harder for hardware
> manufacturers to get right (for some reason or other).
Is it possible to try both at boot time and pick the one which works?
-- Jamie
On Mon, 18 Aug 2003, Jamie Lokier wrote:
> > nmi_watchdog=2 will work on the majority of i686+ (performance
> > counters with NMI delivery mode) boxes and you can check whether it's
> > enabled by doing cat /proc/interrupts and watching if the NMI line ticks
> > at a decent rate. nmi_watchdog=1 tends to be harder for hardware
> > manufacturers to get right (for some reason or other).
>
> Is it possible to try both at boot time and pick the one which works?
I believe we currently already do that if you do nmi_watchdog=2