diff options
| author | Simon Thum <simon.thum@gmx.de> | 2009-04-30 12:58:48 +0200 |
|---|---|---|
| committer | Peter Hutterer <peter.hutterer@who-t.net> | 2009-05-07 16:49:02 +1000 |
| commit | 35fce4e5c7fa34f98b3e4010c6cb09ce38a9205c (patch) | |
| tree | dbc654d2c9e950b1fe5f664665a7ae38bfee7e17 | |
| parent | 1b1b20d6e3e696e4437a9ef56128cde70a485f66 (diff) | |
dix: fix warning in pointer acceleration
newer gcc's warn against how this cast is done (though it eludes me why),
and lrintf() is also faster especially on insane processors like the P4
(http://www.mega-nerd.com/FPcast).
Signed-off-by: Peter Hutterer <peter.hutterer@who-t.net>
| -rw-r--r-- | dix/ptrveloc.c | 10 |
1 files changed, 8 insertions, 2 deletions
diff --git a/dix/ptrveloc.c b/dix/ptrveloc.c index 99efc794c..0a9a22ec2 100644 --- a/dix/ptrveloc.c +++ b/dix/ptrveloc.c @@ -994,14 +994,20 @@ acceleratePointerPredictable( if (dx) { tmp = mult * fdx + pDev->last.remainder[0]; - *px = (int)roundf(tmp); + /* Since it may not be apparent: lrintf() does not offer + * strong statements about rounding; however because we + * process each axis conditionally, there's no danger + * of a toggling remainder. Its lack of guarantees hopefully + * makes it faster on the average target. */ + *px = lrintf(tmp); pDev->last.remainder[0] = tmp - (float)*px; } if (dy) { tmp = mult * fdy + pDev->last.remainder[1]; - *py = (int)roundf(tmp); + *py = lrintf(tmp); pDev->last.remainder[1] = tmp - (float)*py; } + DebugAccelF("pos (%i | %i) remainders x: %.3f y: %.3f delta x:%.3f y:%.3f\n", *px, *py, pDev->last.remainder[0], pDev->last.remainder[1], fdx, fdy); } } } |