[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [Devel] FreeType 2.1.3 segfaults with htst3.ttf
From: |
Owen Taylor |
Subject: |
Re: [Devel] FreeType 2.1.3 segfaults with htst3.ttf |
Date: |
Tue, 7 Jan 2003 15:25:46 -0500 (EST) |
User-agent: |
Gnus/5.0807 (Gnus v5.8.7) Emacs/21.2 |
Anthony Fok <address@hidden> writes:
> Hello,
>
> While I was trying to evaluate a Chinese TrueType font--HuaTian
> Three-step complex TTF song font (GB2312)--which, moreso than MingLiU,
> uses bytecode instructions to shape the strokes and even components of
> Chinese characters (is it the same as "recursive composite" that Owen
> referred to recently?)
No - what I meant by recursive composite there was a font where
a glyph had itself as one of its composite components. That is,
clearly buggy fonts. The characteristic stack backtrace would
be an infinite recursion - thousands of frames of the same
thing.
> The sample font is available for download here:
>
> http://www.htit.com/english/downloads.htm
> http://www.htit.com/download/htst3.ttf
>
> I was able to view this font with ftview1 (freetype 1.3.1) without any
> problem, but with ftview from freetype-2.1.3, a segmentation fault
> would usually occur when a Chinese character is about to be rendered
> (glyph 780 and on). However, I have to admit that It is beyond my skill
> to debug this. :-)
>
> I have attached two gdb backtraces for reference.
If you export the environment variable MALLOC_CHECK_=2, you might
get more meaningful backtraces .... there seems to be some memory
corruption involved.
Regards,
Owen