[cairo] ghostcript-8.54 causes most cairo-1.2.6 ps surface tests to fail

[Gary V. Vaughan]

Now that I'm running ghostscript-8.54 instead of ghostscript-8.16
on my build machine, I notice that the output of many of the
postscript surface tests are rendered a little differently, which
causes them to fail 11 of 98 tests (with  --disable-pdf to isolate
them from the previous test failures I reported).

Maybe it's just a matter of being a little more lenient with the
number of pixels allowed in the diff?

I can post specific examples if you don't have access to both
ghostscript-8.16 and ghostscript-8.54.

TIA,
	Gary
-- 
Gary V. Vaughan (gary at thewrittenword.com)