Hi, [del] >Here is a little background on this issue. Randolph has done a >considerable amount of work to port libffi to the PA, and then add >unwind and other support support to libjava. This gives us a >working Java implementation. Thanks, Randolph!
Yupp, that's a step forward. >This work is currently being integrated into GCC 3.5. This is working >well with 2.6, but we have problems with unwinding through signal >frames on 2.4. 64-bit support is completely broken. On at least >some 32-bit kernels, we end up with hung processes that can bring >the system to its knees. Well, I could only speak for myself and my little boxes, and I would say only do the absolute minimum required (like security fixes). Getting new gcc 3.5 (or gdb) code to work, I would not count as necessary. >We are having a little debate as to whether we need to support 2.4, >given that GCC 3.5 probably won't be out for a year or so. Just make it clear in the release notes and let the compile fail on 2.4 with an #error, IMHO. Does that sound too lazy ?-) HTH, Uli -- Dipl. Inf. Ulrich Teichert|e-mail: [EMAIL PROTECTED] Stormweg 24 |listening to: Noticable One (Rotten Apples) 24539 Neumuenster, Germany|Obstacle 1 (Interpol) Paranoia (N.Y. Rel-X)
