PNG encoder ignores upper 16-bits of width and height values

[DEF7]

This means that PNG images generated using this code can only have dimensions that are less than 2^16.

miniz/miniz_tdef.c

Line 1544 in 293d4db

pnghdr[18] = (mz_uint8)(w >> 8);

The PNG file format specification includes 32-bits for the width and height values, each, in the header. This implementation only uses the lowest 16-bits of the width and height values passed into the function and leaves the upper two bytes for them in the header initialized to zero.

It can easily be modified to include the entire width and height values (though they are signed ints when passed into the function) but I just thought it was worth pointing out for anyone else who is trying to output huge PNG files with this encoder. :)

[uroni]

Could you create a PR to fix this? Looks easy to do (at least for 8 more bits).