The vertical scaler handles 2 rows at a time and thus requires
alignment by 2, or else it'll read invalid memory and result in
corrupt output.

Signed-off-by: Janne Grunau <janne-libav@jannau.net>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Reimar Döffinger <Reimar.Doeffinger@gmx.de>

SSE-optimized hScale() scales up to 4 pixels at once, so we need to
allocate up to 3 padding pixels to prevent overreads. This fixes
valgrind errors in various swscale-tests on fate.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

The logged information is possibly false, and it tends to be outdated
after each change since the logging code needs to be manually updated.

Simplify and prevent confusing wrong debug messages.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

Also remove the unnecessary isSupportedIn/Out macros.

Make the code more compact/readable, and simplify the access to
lsws-specific pixel format information.

Signed-off-by: Luca Barbato <lu_zero@gentoo.org>

This reverts commit ac0fb59348. It
causes valgrind errors which I'll want to investigate before
resubmitting this.

Directly call sws_isSupportedInput/Output.

Signed-off-by: Mans Rullgard <mans@mansr.com>

Fixes a crash when forcing libc to strictly adhere to malloc sizes.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

We would use the second half of the U plane buffer, rather than the
V plane buffer, to output the V plane pixels.

For 9/10bit, it means we don't have to upscale to 16bit before
actual scaling or pixel format conversion, and thus a performance
gain.

We would use the second half of the U plane buffer, rather than the
V plane buffer, to output the V plane pixels.

This functionality is only implemented for RGB24/32 and causes crashes
otherwise.

For 9/10bit, it means we don't have to upscale to 16bit before
actual scaling or pixel format conversion, and thus a performance
gain.

This means that precision is retained when scaling between sample
formats with >8 bits per component (48bit RGB, 16bit grayscale,
9/10/16bit YUV).

This means that precision is retained when scaling between sample
formats with >8 bits per component (48bit RGB, 16bit grayscale,
9/10/16bit YUV).

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>

This works through some non-obvious hacks in utils.c.

Signed-off-by: Diego Biurrun <diego@biurrun.de>

Also add missing glue code for recently added YUV422P10 formats
to swscale.

Signed-off-by: Ronald S. Bultje <rsbultje@gmail.com>

Inline functions are easier to read, maintain, modify and test,
which justifies the slightly increased source size. This patch
also adds support for non-native endianness RGB15/16 and fixes
isSupportedOutput() to no longer claim that we support writing
non-native RGB565/555/444.

Reintroduce the internal symbol which was removed in:

commit e1197b9e17
Author: Stefano Sabatini <stefano.sabatini-lala@poste.it>
Date:   Sun May 29 17:57:40 2011 +0200

    swscale: remove sws_format_name()

    Use av_get_pix_fmt_name() instead.

The symbol is used by some external libs (hi libx264!), this gives
time to them to use the recently added av_get_pix_fmt_name() rather
than an internal symbol.

Altivec sws code intentionally overreads buffers for better performance,
so we need to allocate larger buffers to handle that.

Use av_get_pix_fmt_name() instead.

commit 5a5a0f1613
Author: Diego Biurrun <diego@biurrun.de>
Date:   Fri May 27 19:46:39 2011 +0200

    swscale: Remove unused variables in x86 code.

    libswscale/x86/swscale_template.c:2072: warning: unused variable ‘canMMX2BeUsed’
    libswscale/x86/swscale_template.c:2145: warning: unused variable ‘canMMX2BeUsed’
    libswscale/x86/swscale_template.c:2209: warning: unused variable ‘chrVPixBuf’
    libswscale/x86/swscale_template.c:2237: warning: unused variable ‘chrVSrcPtr’

commit 389e2000eb
Author: Ronald S. Bultje <rsbultje@gmail.com>
Date:   Fri May 27 12:23:32 2011 -0400

    swscale: delay allocation of formatConvBuffer().

    That means it won't be allocated when not needed. Alongside
    this, it fixes valgrind/fate-detected memory leaks.

commit f327bfa6dc
Author: Ronald S. Bultje <rsbultje@gmail.com>
Date:   Fri May 27 11:36:43 2011 -0400

    swscale: fix build with --disable-swscale-alpha.

commit 9f5d45025e
Author: Ronald S. Bultje <rsbultje@gmail.com>
Date:   Fri May 27 09:28:38 2011 -0400

    swscale: fix non-bitexact yuv2yuv[X2]() MMX/MMX2 functions.

commit 93681fbd50
    Author: Ronald S. Bultje <rsbultje@gmail.com>
    Date:   Thu May 26 11:32:32 2011 -0400

        swscale: fix compile on ppc.

    commit e758573a88
    Author: Ronald S. Bultje <rsbultje@gmail.com>
    Date:   Thu May 26 10:36:47 2011 -0400

        swscale: fix compile on x86-32.

    commit 0f4eb8b043
    Author: Ronald S. Bultje <rsbultje@gmail.com>
    Date:   Thu May 26 09:17:52 2011 -0400

        swscale: remove VOF/VOFW.

    commit b4a224c5e4
    Author: Ronald S. Bultje <rsbultje@gmail.com>
    Date:   Wed May 25 14:30:09 2011 -0400

        swscale: split chroma buffers into separate U/V planes.

        Preparatory step to implement support for sizes > VOFW.

This allows to convert between formats of arbitrary width,
regardless of the value of VOF/VOFW.

That means it won't be allocated when not needed. Alongside
this, it fixes valgrind/fate-detected memory leaks.

Signed-off-by: Michael Niedermayer <michaelni@gmx.at>