(Bug report by Y.Ohashik).

Too many input files would overflow nfiles
Signed-off-by: NJohn Bowler <jbowler@acm.org>

timepng would could a file as added even if it failed and the assembly file got
rewound.
Signed-off-by: NJohn Bowler <jbowler@acm.org>

Signed-off-by: NJohn Bowler <jbowler@acm.org>

Signed-off-by: NJohn Bowler <jbowler@acm.org>

Signed-off-by: NJohn Bowler <jbowler@acm.org>

Signed-off-by: NJohn Bowler <jbowler@acm.org>

Signed-off-by: NJohn Bowler <jbowler@acm.org>

PNG_FILTER_NONE|PNG_FILTER_SUB|PNG_FILTER_UP).

    - memcpy-free implementations of load3() / store3().
    These should have less variance compiler to compiler.

    - call load3() only when needed at the end of a scanline.
    In the middle, we can use the faster load4(), ignoring that byte.

Although not a bug in this case, but shows up when -Wall is used

This makes tests slightly quicker by allowing the temporary file to be created
beforehand.
Signed-off-by: NJohn Bowler <jbowler@acm.org>

This makes tests slightly quicker by allowing the temporary file to be created
beforehand.
Signed-off-by: NJohn Bowler <jbowler@acm.org>

The code no longer gives up/fails on invalid PNG data, it just skips it (with
error messages).  The code no longer fails on PNG files with data beyond IEND.
Options exist to use png_read_png (reading the whole image, not by row) and, in
that case, to apply any of the supported transforms.  This makes for more
realistic testing; the decoded data actually gets used in a meaningful fashion.
Signed-off-by: NJohn Bowler <jbowler@acm.org>

and added PNG_NO_INTEL_SSE_3BPP macro.

their new location in contrib/intel/* (was intel/*)