Problem fixed.  See thread "Config File for Linux 2.6.22".

Brad


-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] On
Behalf Of Ali Saidi
Sent: Wednesday, June 04, 2008 6:40 AM
To: M5 users mailing list
Subject: Re: [m5-users] No IPI error with >4p systems

You're probably using the wrong PAL code (ts_osfpal instead of  
tsb_osfpal). I modified the latter to add different configuration  
addresses that support more than 4 processors (the original limit of  
the Tsunami platform).

Ali


On Jun 4, 2008, at 2:40 AM, Beckmann, Brad wrote:

> Hi,
>
> I'm encountering a repeated warning message from M5 and I was hoping  
> you could provide some detail of how serious of a warning message  
> this is.  In particular, I'm running the Netperf stream workload  
> with Linux 2.6.22 and a drive system containing 5 processors.  Soon  
> after starting the simulation, a continuous stream of "warn: clear  
> IPI for CPU=0, but NO IPI" warning messages are printed to the  
> screen.  As you probably know, this message comes from the  
> clearIPI() function of tsunami_cchip.cc.  This warning message only  
> occurs when the system is greater than 4 cpus.  By calling the  
> remote kernel debugger from the user-level debugger on M5, I  
> obtained the following kernel call stack.  However, I don't think it  
> is very useful.  It appears to be a simple delay (idle?) loop.
>
> Could you please provide me some more information of what this  
> warning message means?
>
> Thanks!  I appreciate the help.
>
> Brad
>
>
>
> #0  0xfffffc00004edef4 in __delay (loops=Variable "loops" is not  
> available.
> ) at arch/alpha/lib/udelay.c:25
> #1  0xfffffc000031c13c in __cpu_up (cpu=Variable "cpu" is not  
> available.
> ) at arch/alpha/kernel/smp.c:397
> #2  0xfffffc00007b1838 in cpu_up (cpu=4) at kernel/cpu.c:232
> #3  0xfffffc00007a2eb4 in kernel_init (unused=Variable "unused" is  
> not available.
> ) at init/main.c:393
> #4  0xfffffc0000311508 in kernel_thread ()
> #5  0xfffffc0000311508 in kernel_thread ()
> Previous frame identical to this frame (corrupt stack?)
> _______________________________________________
> m5-users mailing list
> [email protected]
> http://m5sim.org/cgi-bin/mailman/listinfo/m5-users

_______________________________________________
m5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users


_______________________________________________
m5-users mailing list
[email protected]
http://m5sim.org/cgi-bin/mailman/listinfo/m5-users

Reply via email to