On Mon, 28 Jan 2008 02:17:20 -0400, "bulia byak" wrote:

...

The inkscape folks had run into this problem and I had said we'd have a fix in 1.6. Instead, I'd like to volunteer to help fix the inkscape code to allocate buffers with a properly aligned stride. If some kind inkscape developer were willing to help me with this soon, that would be great.

it was so long ago I completely forgot the context :(

can you please remind me what the problem was, specifically?

The problem was in how inkscape allocates buffers before using them as an A8 mask.

For example, imagine you want to create a 7x7 pixel mask. Apparently the current inkscape code does something like:

width = 7; height = 7; data = malloc (width * height);

The problem would then come in trying to use this data as a cairo image surface:

stride = 7; surface = cairo_image_surface_create_for_data (data, CAIRO_FORMAT_A8, width, height, stride);

In previous releases of cairo, using a stride value like this that is not a multiple of 4 would cause cairo to draw total garbage.

Instead, now, as of cairo 1.5.8, the above would result in an error surface with a status of CAIRO_STATUS_INVALID_STRIDE.

The fixed could would instead look like this:

width = 7; height = 7; stride = cairo_format_stride_for_width (CAIRO_FORMAT_A8, width); data = malloc (stride * width); surface = cairo_image_surface_create_for_data (data, CAIRO_FORMAT_A8, width, height, stride);

Here, the cairo_format_stride_for_width call takes care of any necessary alignment, (which in the current version would return a value of 8 for the stride).

Obviously, that much of a change is trivial.

Beyond that, we'll need to then fixup as much inkscape code as necessary so that all of inkscape's own rendering code properly accounts for the stride not being identical to the width.

Make sense?

Again, I'm glad to help out with this. Just let me know what I can do.

-Carl