IO-APIC in SMP dual Athlon XP1800
Hi follows,
I'm getting in troubles with a A7M266-D motherboard with two Athlon XP 1800 cpus (yes, XP not MP!).
Following the screen shot of my problem:
^-------cut here---------^
EIP: 0010:[<c0111686>] Not tainted EFLAGS: 00011046 <4>CPU: 1 <4>CPU: 1 EIP: 0010:[<c0111686>] Not tainted EFLAGS: 00011046
u!wisuum aeesess 7eg111eg
printing eip:
printing eip:
*pde = 11111010
Stuck ??
CPU #1 not responding - cannot use it.
Error: only one processor found.
ENABLING IO-APIC IRQs
Setting 2 in the phys_id_present_map
...changing IO-APIC physical APIC ID to 2 ... ok.
init IO_APIC IRQs
IO-APIC (apicid-pin) 2-10, 2-11, 2-12, 2-13, 2-16, 2-17, 2-20, 2-21, 2-22,
2-23 not connected.
..TIMER: vector=0x31 pin1=2 pin2=0
^-----cut here-----^
After spend some times put printk's in kernel source like "Reach this point!" I was trying disable IO_APIC in .config file but some link erros ocurred. Has any way to turn IO_APIC disable? Or its extreme necessary?
Thanks,
Ricardo.
Well I'll probably tell you what just about everyone else here will. If you use XP procs on a MP system you are on your own. If you lucky they will work, if you aren't they won't. It appears that you are not lucky.
Steve
Quoted 40 lines · Scorpion
Hi follows,
I'm getting in troubles with a A7M266-D motherboard with two Athlon XP 1800 cpus (yes, XP not MP!).
Following the screen shot of my problem:
^-------cut here---------^EIP: 0010:[<c0111686>] Not tainted EFLAGS: 00011046 <4>CPU: 1 <4>CPU: 1 EIP: 0010:[<c0111686>] Not tainted EFLAGS: 00011046
u!wisuum aeesess 7eg111eg
printing eip:
printing eip:
*pde = 11111010
Stuck ??
CPU #1 not responding - cannot use it.
Error: only one processor found.
ENABLING IO-APIC IRQs
Setting 2 in thephys_id_present_map
...changing IO-APIC physical APIC ID to 2 ... ok.
initIO_APICIRQs
IO-APIC (apicid-pin) 2-10, 2-11, 2-12, 2-13, 2-16, 2-17, 2-20, 2-21, 2-22, 2-23 not connected. ..TIMER: vector=0x31 pin1=2 pin2=0
^-----cut here-----^After spend some times put printk's in kernel source like "Reach this point!"
I was trying disableIO_APICin .config file but some link erros ocurred. Has any way to turnIO_APICdisable? Or its extreme necessary?Thanks,
Ricardo.
-
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/
Quoted 2 lines · Scorpion
I'm getting in troubles with a A7M266-D motherboard with two Athlon XP 1800 cpus (yes, XP not MP!).
Which kernel version ?
Zwane
Signature · 1 line
function.linuxpower.ca
I tried using:
2.4.18-3smp (red hat 7.3) with options: hdc=ide-scsi and using ext3 fs.
2.4.18-5custom (red hat 7.3 recompiled with smp, bigmem and ext3 fs) with
options: hdc=ide-scsi
2.4.18-5smp (red hat 7.3 upgrade of bigmem header problem and ext3 fs problem) with options: hdc=ide-scsi
2.4.18-3 (red hat 7.3 uniprocessor for get up) with options:
hdc=ide-scsi noprobe mem=256m
details about the last:
When changing mem parameter value to 64 or 512 or 128 its stop in
different places (using the "printk debbug technic") in io_apic.c/apic.c parts.
After that I tried the kernel from kernel.org and got the following message:
-----------^^^cut here^^^------------
Total of 2 processors activated (6121.06 BogoMIPS). ENABLING IO-APIC IRQs
Setting 2 in the phys_id_present_map
...changing IO-APIC physical APIC ID to 2 ... ok. ...TIMER: vector=0x31 pin1=2 pin2=0
testing the IO APIC...........................................
......................................................................done. Using local APIC timer interrupts.
calibrating APIC timer ...
..... CPU clock speed is 1533.2911 MHz.
..... host bus clock speed is 266.6593 MHz.
cpu: 0, clocks: 2666593, slice: 888864 CPU0<T0:2666592,T1:1777728,D:0,S:888864,C:2666593> -----------^^^cut here^^^------------
Its a error? Should a try the last last kernel patchs?
I'm not thinking about a hadware problem, cause BIOS detect 2 CPUS and Linux agree with this. Am I wrong?
Regards,
Ricardo.
Quoted 7 lines · The Kernel Developer Zwane Mwaikambo
Quoted 2 lines · Scorpion
I'm getting in troubles with a A7M266-D motherboard with two Athlon XP 1800 cpus (yes, XP not MP!).
Which kernel version ?
Zwane
Hi All,
I got my Dual Athlon XP1800 working now.
Everything gonna well after changed MP 1.4 Support to disable in BIOS, and leaving MP table enabled.
Anyone knows if linux 2.4.19 has not yet a full implemantation of MP 1.4 or if it is just a BIOSes bug?
Thanks in advance,
Scorpion.
Quoted 7 lines · The Kernel Developer Zwane Mwaikambo
Quoted 2 lines · Scorpion
I'm getting in troubles with a A7M266-D motherboard with two Athlon XP 1800 cpus (yes, XP not MP!).
Which kernel version ?
Zwane
Quoted 6 lines · Scorpion
Hi All,
I got my Dual Athlon XP1800 working now.
Everything gonna well after changed MP 1.4 Support to disable in BIOS, and leaving MP table enabled.
Anyone knows if linux 2.4.19 has not yet a full implemantation of MP 1.4 or if it is just a BIOSes bug?
On the 1004 BIOS with the ASUS it seems to be a bios table error. Later BIOSes fixed it, then removed the option, then broke lots of other stuff. I went back to 1004 so I dont know how the newest fare
Quoted 11 lines · Alan Cox
Quoted 6 lines · Scorpion
Hi All,
I got my Dual Athlon XP1800 working now.
Everything gonna well after changed MP 1.4 Support to disable in BIOS, and leaving MP table enabled.
Anyone knows if linux 2.4.19 has not yet a full implemantation of MP 1.4 or if it is just a BIOSes bug?On the 1004 BIOS with the ASUS it seems to be a bios table error. Later BIOSes fixed it, then removed the option, then broke lots of other stuff. I went back to 1004 so I dont know how the newest fare
I am using 1006. I have enabled MP 1.4 in order to get the IRQ numbers 16 and beyond. I am not sure how useful that is since several PCI IRQs are shared and there are lots of free ones. Anyhow, MP 1.1 works just fine too. I am using the Athlon MP 1900+ CPUs.
Note to anyone who is considering this cpu/motherboard, this combo uses *shedloads* of power. 400W is *not* enough (I've seen Alan post to this effect before but it bears repeating). Also, make sure you have good case airflow -- all that power that is being consumed is being turned into heat. Aftermarket CPU heatsink and fan are a good idea. Unfortunately, it is a bit crowded for 80mm CPU fans though.
Things are stable now once I mitigated the heat death problem.
Signature · 1 line
Johan KULLSTAM