[FFmpeg-devel] [PATCH] MMX2 scaler: Determine "funnyCode" size at runtime.

[Ramiro Polla]

On Mon, Dec 7, 2009 at 1:05 AM, Michael Niedermayer <michaelni at gmx.at> wrote:
> On Sun, Dec 06, 2009 at 11:39:57PM -0200, Ramiro Polla wrote:
>> On Thu, Aug 13, 2009 at 7:12 PM, Michael Niedermayer <michaelni at gmx.at> wrote:
>> > On Thu, Aug 13, 2009 at 03:36:56PM -0300, Ramiro Polla wrote:
>> >> On Thu, Aug 13, 2009 at 12:52 PM, Michael Niedermayer<michaelni at gmx.at> wrote:
>> >> > On Thu, Aug 13, 2009 at 10:16:42AM -0300, Ramiro Polla wrote:
>> >> >> Hi Michael,
>> >> >> On Wed, Aug 12, 2009 at 10:38 PM, Michael Niedermayer<michaelni at gmx.at> wrote:
>> >> >> > also *A/*B are very poor names
>> >> >>
>> >> >> It's your code =). I don't think they're that bad, just that some
>> >> >> documentation would make it much clearer. Otherwise the variable name
>> >> >> would have to grow from fragmentA to something like fragmentRead4 and
>> >> >> it still wouldn't make it as clear as the comment that this is the
>> >> >> only difference between them. I attached a patch with the comment, but
>> >> >> feel free to suggest better variable names.
>> >>
>> >> [...]
>> >>
>> >> >> ?swscale.c | ? ?2 ++
>> >> >> ?1 file changed, 2 insertions(+)
>> >> >> 9cb63462d2b442d2a7768b1a2a15de93d7cb656f ?0004-MMX2-horizontal-scaler-Comment-differences-between.patch
>> >> >> From 7077c874181e2461f2985b28495a7df08ef49522 Mon Sep 17 00:00:00 2001
>> >> >> From: Ramiro Polla <ramiro.polla at gmail.com>
>> >> >> Date: Thu, 13 Aug 2009 09:53:27 -0300
>> >> >> Subject: [PATCH] MMX2 horizontal scaler: Comment differences between fragment A and fragment B.
>> >> >>
>> >> >> ---
>> >> >> ?swscale.c | ? ?2 ++
>> >> >> ?1 files changed, 2 insertions(+), 0 deletions(-)
>> >> >>
>> >> >> diff --git a/swscale.c b/swscale.c
>> >> >> index 4a69ab8..c73039e 100644
>> >> >> --- a/swscale.c
>> >> >> +++ b/swscale.c
>> >> >> @@ -1771,6 +1771,8 @@ static void initMMX2HScaler(int dstW, int xInc, uint8_t *filterCode, int16_t *fi
>> >> >>
>> >> >> ? ? ?//code fragment
>> >> >>
>> >> >> + ? ?/* Fragment A reads 8 samples to perform the filtering, whereas fragment B
>> >> >> + ? ? * reads only 4 samples. Besides that, both fragments are equal. */
>> >> >> ? ? ?__asm__ volatile(
>> >> >
>> >> > this description is poor and not really correct
>> >>
>> >> Hmm, suggestions are welcome. Or else I'll take a closer look later to
>> >> see if I can come up with a better description.
>> >
>> > something with 4 vs 5 input pixels/samples, iam not sure though how to make
>> > a good name out of it though
>>
>> New patch attached.
>
>> ?swscale.c | ? ?6 ++++++
>> ?1 file changed, 6 insertions(+)
>> d28c35330ff653170ad058730a7331e9bc898638 ?mmx2scaler.diff
>
> ok

Applied.