Re: [RFC PATCH] x86: Add safe_udelay() and safe_msleep()

[Yinghai Lu <yinghai@kernel.org>]

On 01/13/2011 04:44 PM, Greg KH wrote:
> On Thu, Jan 13, 2011 at 04:31:04PM -0800, Yinghai Lu wrote:
>> there are some udelay() in drivers/usb/early/ehci-dbgp.c
>>
>> that is for early USB debug port console support.
>>
>> We should replace them with early version udelay()
> 
> Again, I fail to see why.  They work fine today, right?

did not test that for a while.

from code review, we should not use udelay() that early.

in arch/x86/lib/delay.c we have
inline void __const_udelay(unsigned long xloops)
{
        int d0;

        xloops *= 4;
        asm("mull %%edx"
                :"=d" (xloops), "=&a" (d0)
                :"1" (xloops), "0"
                (this_cpu_read(cpu_info.loops_per_jiffy) * (HZ/4)));

        __delay(++xloops);
}
EXPORT_SYMBOL(__const_udelay);

that percpu cpu_info.loops_per_jiffy only have value after smp_prepare_cpus() is called.

> 
> You _really_ need to prove what you are trying to do here, with lots of
> details and descriptions of what the problem is, as it's getting a bit
> annoying...

sorry for that.

Yinghai