[PATCH] Input: ff-core - Correct magnitude setting for rumble compatibility

[PATCH] Input: ff-core - Correct magnitude setting for rumble compatibility

[Charles Keepax]

When converting a rumble into a periodic effect, for compatibility,
the magnitude is effectively calculated using:

magnitude = max(strong_rubble / 3 + weak_rubble / 6, 0x7fff);

The rumble magnitudes are both u16 and the resulting magnitude is
s16. The max is presumably an attempt to limit the result of the
calculation to the maximum possible magnitude for the s16 result,
and thus should be a min.

However in the case of strong = weak = 0xffff, the result of the first
part of the calculation is 0x7fff, meaning that the min would be
redundant anyway, so simply remove the current max.

Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>
---
 drivers/input/ff-core.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/input/ff-core.c b/drivers/input/ff-core.c
index 1cf5deda06e19..fa8d1a4660142 100644
--- a/drivers/input/ff-core.c
+++ b/drivers/input/ff-core.c
@@ -67,7 +67,7 @@ static int compat_effect(struct ff_device *ff, struct ff_effect *effect)
 		effect->type = FF_PERIODIC;
 		effect->u.periodic.waveform = FF_SINE;
 		effect->u.periodic.period = 50;
-		effect->u.periodic.magnitude = max(magnitude, 0x7fff);
+		effect->u.periodic.magnitude = magnitude;
 		effect->u.periodic.offset = 0;
 		effect->u.periodic.phase = 0;
 		effect->u.periodic.envelope.attack_length = 0;
-- 
2.11.0

Re: [PATCH] Input: ff-core - Correct magnitude setting for rumble compatibility

[Dmitry Torokhov]

On Tue, Nov 30, 2021 at 01:50:39PM +0000, Charles Keepax wrote:
> When converting a rumble into a periodic effect, for compatibility,
> the magnitude is effectively calculated using:
> 
> magnitude = max(strong_rubble / 3 + weak_rubble / 6, 0x7fff);
> 
> The rumble magnitudes are both u16 and the resulting magnitude is
> s16. The max is presumably an attempt to limit the result of the
> calculation to the maximum possible magnitude for the s16 result,
> and thus should be a min.
> 
> However in the case of strong = weak = 0xffff, the result of the first
> part of the calculation is 0x7fff, meaning that the min would be
> redundant anyway, so simply remove the current max.
> 
> Signed-off-by: Charles Keepax <ckeepax@opensource.cirrus.com>

Applied, thank you.

-- 
Dmitry