Re: [PATCH] powerpc: Replaced msleep with usleep_range

[Michael Ellerman]

Hi Daniel,

Daniel Klamt <eleon@xxxxxxxx> writes:
> Replaced msleep for less than 10ms with usleep_range because will
> often sleep longer than intended.
> For original explanation see:
> Documentation/timers/timers-howto.txt
>
> Signed-off-by: Daniel Klamt <eleon@xxxxxxxx>
> Signed-off-by: Bjoern Noetel <bjoern@xxxxxxxxxx>
> ---
>  arch/powerpc/sysdev/xive/native.c | 24 ++++++++++++------------
>  1 file changed, 12 insertions(+), 12 deletions(-)
>
> diff --git a/arch/powerpc/sysdev/xive/native.c b/arch/powerpc/sysdev/xive/native.c
> index 311185b9960a..b164b1cdf4d6 100644
> --- a/arch/powerpc/sysdev/xive/native.c
> +++ b/arch/powerpc/sysdev/xive/native.c
> @@ -109,7 +109,7 @@ int xive_native_configure_irq(u32 hw_irq, u32 target, u8 prio, u32 sw_irq)
>  		rc = opal_xive_set_irq_config(hw_irq, target, prio, sw_irq);
>  		if (rc != OPAL_BUSY)
>  			break;
> -		msleep(1);
> +		usleep_range(1000, 1100)
>  	}

We have actually recently added OPAL_BUSY_DELAY_MS (see opal.h) and
should convert these to use that. eg:

		msleep(OPAL_BUSY_DELAY_MS);


That has the effect of documenting that this is a short sleep to wait
for opal (our firmware), and we aren't concerned about rounding errors
in msleep().

If you send me a patch doing that I'd be happy to merge it.

cheers