Re: [ft-devel] freetype 2.3.8 warnings

[Sean McBride]

On 1/23/09 7:16 AM, Werner LEMBERG said:

>> WARNING: While resolving call to function 'FTC_GNode_Compare'
>> arguments were dropped!
>
>No idea what this is...

Likewise.  No matter.

>> base/fttrigon.c: In function 'FT_Vector_Rotate':
>> base/fttrigon.c:439: warning: right shift count >= width of type
>> base/fttrigon.c:440: warning: right shift count >= width of type
>
>Hmm.  For me, this warning appears to be completely useless.  Any
>value larger or equal to 32 makes the result consisting of zeros only,
>and there is no datatype which holds only 5 bits.  Does the compiler
>expect a bitmask or what?  Can you dig into the compiler's
>documentation to give more details?

The warning comes from gcc 4.2's -Wall, more specifically from -Wstrict-
overflow=1 (which is part of -Wall).  See:

<http://gcc.gnu.org/onlinedocs/gcc/Warning-Options.html>
<http://www.airs.com/blog/archives/120>

The thing being right-shifted is signed.  IIRC, right-shifting of
negative numbers in C is undefined.  Maybe that's what the warning is
getting at.

--
____________________________________________________________
Sean McBride, B. Eng                 address@hidden
Rogue Research                        www.rogue-research.com
Mac Software Developer              Montréal, Québec, Canada