Re: [PATCH] x86: printk_time to use tsc before cpu_clock is ready

["Yinghai Lu" <yhlu.kernel@gmail.com>]

On Tue, Jul 29, 2008 at 3:13 AM, Peter Zijlstra <peterz@infradead.org> wrote:
> On Tue, 2008-07-29 at 03:09 -0700, Yinghai Lu wrote:
>> so can get tsc value on printk
>>
>> need to apply after
>>       [PATCH] printk_time: prepare stub for using other than cpu_clock
>
>
> You failed to mention why we want this... ;-)
>

"
otherwise will get
[    0.000000]....
.....
[    0.000000] TSC calibrated against PM_TIMER
[    0.000000] Detected 2493.378 MHz processor.
[    0.008000] spurious 8259A interrupt: IRQ7.
[    0.008000] Console: colour VGA+ 80x25
[    0.008000] console handover: boot [uart0] -> real [ttyS0]
[    0.008000] Checking aperture...
[    0.008000] No AGP bridge found
[    0.008000] Node 0: aperture @ dc000000 size 64 MB
[    0.008000] Node 1: aperture @ dc000000 size 64 MB
[    0.008000] Node 2: aperture @ dc000000 size 64 MB
[    0.008000] Node 3: aperture @ dc000000 size 64 MB
[    0.008000] Node 4: aperture @ dc000000 size 64 MB
[    0.008000] Node 5: aperture @ dc000000 size 64 MB
[    0.008000] Node 6: aperture @ dc000000 size 64 MB
[    0.008000] Node 7: aperture @ dc000000 size 64 MB
[    0.008000] Memory: 264718324k/269090816k available (7532k kernel
code, 3716316k reserved, 5099k data, 484k init)
[    0.008000] CPA: page pool initialized 1 of 1 pages preallocated
[    0.008000] SLUB: Genslabs=13, HWalign=64, Order=0-3, MinObjects=0,
CPUs=32, Nodes=8
[    0.008000] hpet clockevent registered
[    0.008000] Calibrating delay loop (skipped), value calculated
using timer frequency.. 4986.72 BogoMIPS (lpj=9973456)
[    0.060003] Dentry cache hash table entries: 33554432 (order: 16,
268435456 bytes)

with this patch could get tsc to replace those [0,000000] and
[0.0080000], so could debug possible delay on big system on early
stage.
"

YH