stall during boot on x86-64.

stall during boot on x86-64.

[Dave Jones]

When I boot my EM64T, I get a slight noticable pause
really early on in boot.  Booting with 'time'
shows an interesting artifact.

Kernel command line: ro root=/dev/VolGroup00/LogVol00 console=ttyS0,38400 console=tty0 time
kernel profiling enabled (shift: 1)
[    0.000000] Initializing CPU#0
[    0.000000] PID hash table entries: 4096 (order: 12, 131072 bytes)
[    0.000000] time.c: Using 14.318180 MHz HPET timer.
[    0.000000] time.c: Detected 2793.081 MHz processor.
[   27.449661] Console: colour VGA+ 80x25
[   28.484309] Dentry cache hash table entries: 131072 (order: 8, 1048576 bytes)
[   28.506519] Inode-cache hash table entries: 65536 (order: 7, 524288 bytes)
[   28.539543] Memory: 1014240k/1047080k available (2490k kernel code, 32456k reserved, 1664k data, 236k init)

Note the jump in the time value..
I'm not sure this is actually where the pause I see is, as the text
is buffered, but it's something I can't explain.

		Dave

RE: stall during boot on x86-64.

[Pallipadi, Venkatesh]

 

>-----Original Message-----
>From: linux-kernel-owner@vger.kernel.org 
>[mailto:linux-kernel-owner@vger.kernel.org] On Behalf Of Dave Jones
>Sent: Tuesday, December 13, 2005 7:22 PM
>To: Linux Kernel
>Subject: stall during boot on x86-64.
>
>When I boot my EM64T, I get a slight noticable pause
>really early on in boot.  Booting with 'time'
>shows an interesting artifact.
>
>Kernel command line: ro root=/dev/VolGroup00/LogVol00 
>console=ttyS0,38400 console=tty0 time
>kernel profiling enabled (shift: 1)
>[    0.000000] Initializing CPU#0
>[    0.000000] PID hash table entries: 4096 (order: 12, 131072 bytes)
>[    0.000000] time.c: Using 14.318180 MHz HPET timer.
>[    0.000000] time.c: Detected 2793.081 MHz processor.
>[   27.449661] Console: colour VGA+ 80x25
>[   28.484309] Dentry cache hash table entries: 131072 (order: 
>8, 1048576 bytes)
>[   28.506519] Inode-cache hash table entries: 65536 (order: 
>7, 524288 bytes)
>[   28.539543] Memory: 1014240k/1047080k available (2490k 
>kernel code, 32456k reserved, 1664k data, 236k init)
>
>Note the jump in the time value..

May be this is just the origin of time as far as kernel is concerned.
No?

Thanks,
Venki

Re: stall during boot on x86-64.

[Andi Kleen]

"Pallipadi, Venkatesh" <venkatesh.pallipadi@intel.com> writes:

> >[    0.000000] time.c: Detected 2793.081 MHz processor.
> >[   27.449661] Console: colour VGA+ 80x25
> >[   28.484309] Dentry cache hash table entries: 131072 (order: 
> >8, 1048576 bytes)
> >[   28.506519] Inode-cache hash table entries: 65536 (order: 
> >7, 524288 bytes)
> >[   28.539543] Memory: 1014240k/1047080k available (2490k 
> >kernel code, 32456k reserved, 1664k data, 236k init)
> >
> >Note the jump in the time value..
> 
> May be this is just the origin of time as far as kernel is concerned.
> No?

It is. Before that the timer interrupt doesn't run and jiffies won't 
increase.

-Andi

Re: stall during boot on x86-64.

[Dave Jones]

On Wed, Dec 14, 2005 at 08:23:29AM +0100, Andi Kleen wrote:
 > "Pallipadi, Venkatesh" <venkatesh.pallipadi@intel.com> writes:
 > 
 > > >[    0.000000] time.c: Detected 2793.081 MHz processor.
 > > >[   27.449661] Console: colour VGA+ 80x25
 > > >[   28.484309] Dentry cache hash table entries: 131072 (order: 
 > > >8, 1048576 bytes)
 > > >[   28.506519] Inode-cache hash table entries: 65536 (order: 
 > > >7, 524288 bytes)
 > > >[   28.539543] Memory: 1014240k/1047080k available (2490k 
 > > >kernel code, 32456k reserved, 1664k data, 236k init)
 > > >
 > > >Note the jump in the time value..
 > > 
 > > May be this is just the origin of time as far as kernel is concerned.
 > > No?
 > 
 > It is. Before that the timer interrupt doesn't run and jiffies won't 
 > increase.

Makes sense now. After sleeping on it, I think the stall I see
is caused by framebuffer console. I'll poke some more at it
this evening.

		Dave