>Travis Vitek wrote: > >So now I'm asking myself the question. Should I just revert this change >and then 'fix' all of the existing code that uses static buffers to use >dynamic ones, or should I be fixing the code that is expecting >rw_asnprintf() to deallocate the buffer. I can't really use the >asprintf() as a reference because rw_asnprintf() behave very >differently in this respect. The asprintf() function always allocates >a buffer of the appropriate size and never looks at the contents of the >provided pointer. > >In case it matters. I think I can fix the offending test driver/suite >code in about 3 hours. > >Travis >
Fixing this problem is likely to require more much more discussion and implementation time. Being that the 4.2.1 release target is so close I think it would be best to revert my recent changes completely and to workaround this issue in the one place that it is necessary. I realize that this is fixing a symptom instead of addressing the actual problem, but at this point in time I think it is the best thing to do. Travis
