Re: [PATCH] drm/i915: use udelay for very short delays

From: Jani Nikula
Date: Thu Dec 15 2016 - 04:09:03 EST


On Thu, 15 Dec 2016, Nicholas Mc Guire <hofrat@xxxxxxxxx> wrote:
> Even on fast systems a 2 microsecond delay is most likely more efficient
> as a busy-wait loop. The overhead of a hrtimer does not seem warranted -
> change this to a udelay(2).

Similar concerns as in [1]. We don't need the accuracy of udelay() here,
so this boils down to which is the better use of CPU. We could probably
relax the max delay more if that was helpful. But I'm not immediately
sold on "is most likely more efficient" which sounds like a gut feeling.

I'm sorry it's not clear in my other reply that I do appreciate
addressing incorrect/silly use of usleep_range(); I'm just not (yet)
convinced udelay() is the answer.

BR,
Jani.


[1] http://lkml.kernel.org/r/8737hpr32a.fsf@xxxxxxxxx


>
> Signed-off-by: Nicholas Mc Guire <hofrat@xxxxxxxxx>
> ---
>
> Problem found by coccinelle:
>
> Patch was compile tested with: x86_64_defconfig (implies CONFIG_DRM_I915)
>
> Patch is against 4.9.0 (localversion-next is next-20161214)
>
> drivers/gpu/drm/i915/intel_dsi.c | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/drivers/gpu/drm/i915/intel_dsi.c b/drivers/gpu/drm/i915/intel_dsi.c
> index 5b72c50..19fe86b 100644
> --- a/drivers/gpu/drm/i915/intel_dsi.c
> +++ b/drivers/gpu/drm/i915/intel_dsi.c
> @@ -379,7 +379,7 @@ static void bxt_dsi_device_ready(struct intel_encoder *encoder)
> val &= ~ULPS_STATE_MASK;
> val |= (ULPS_STATE_ENTER | DEVICE_READY);
> I915_WRITE(MIPI_DEVICE_READY(port), val);
> - usleep_range(2, 3);
> + udelay(2);
>
> /* 3. Exit ULPS */
> val = I915_READ(MIPI_DEVICE_READY(port));

--
Jani Nikula, Intel Open Source Technology Center