[PATCH xf86-input-synaptics 2/2] Ensure history delta computations are signed

[Chase Douglas]

On 02/08/2012 03:54 PM, Peter Hutterer wrote:
> On Tue, Feb 07, 2012 at 01:07:08PM -0800, Chase Douglas wrote:
>> Signed-off-by: Chase Douglas <chase.douglas at canonical.com>
>> ---
>>  src/synaptics.c |    2 +-
>>  1 files changed, 1 insertions(+), 1 deletions(-)
>>
>> diff --git a/src/synaptics.c b/src/synaptics.c
>> index a4c1e5a..106b5ee 100644
>> --- a/src/synaptics.c
>> +++ b/src/synaptics.c
>> @@ -1791,7 +1791,7 @@ HandleTapProcessing(SynapticsPrivate *priv, struct SynapticsHwState *hw,
>>  }
>>  
>>  #define HIST(a) (priv->move_hist[((priv->hist_index - (a) + SYNAPTICS_MOVE_HISTORY) % SYNAPTICS_MOVE_HISTORY)])
>> -#define HIST_DELTA(a, b, e) ((HIST((a)).e) - (HIST((b)).e))
>> +#define HIST_DELTA(a, b, e) ((int)((HIST((a)).e) - (HIST((b)).e)))
> 
> this only casts the result to a signed int, the actual computation is still
> undefined for b > a. With the revert, HIST_DELTA is only used in one place
> so we could either do it there or just write a helper function like this

Why is it undefined? We can have a negative time difference, which is
what the calculation was assuming, IIUC.

-- Chase