Hi, Here's what we've got on 32bit systems: sizeof(char) = 1 sizeof(short) = 2 sizeof(int) = 4 sizeof(long) = 4 sizeof(void *) = 4 On 64bit we've got: sizeof(char) = 1 sizeof(short) = 2 sizeof(int) = 4 sizeof(long) = 8 sizeof(void *) = 8 i.e., int is always 4 bytes, long is the size of a pointer and is either 4 bytes or 8 bytes. In the past we've had some confusion or paranoia around int's and long's. They've been fixed in 4.5 although the code hasn't been scrubbed top to bottom yet nor have the configure scripts and Makefile's been scrubbed for optimal 64bit compile. And, I know of one confusion of a pointer with int in the thread code for Win32 which isn't fixed so 64bit Windows may not work and has never been tried. In addition, the code hasn't been fully scrubbed to deal with large objects, e.g., big files beyond 4gig. In practice, AOLserver 4.5 compiles and runs on 64bit Unix. A more extensive 64bit cleaning effort is planned for the next release. We don't want to hold up 4.5 any longer -- it's got tons of new features. -Jim On Jun 22, 2006, at 9:28 AM, Rick Gutleber wrote: In strict C terms, "int" is supposed to be the most convenient size for the platform and sizeof( long ) >= sizeof( int ) >= sizeof( short ). It's my understanding that sizeof( int ) and sizeof( $TYPE * ) would, under most, if not all circumstances, to be the same.
|
- [AOLSERVER] New release? Daniël Mantione
- Re: [AOLSERVER] New release? Nathan Folkman
- Re: [AOLSERVER] New release? Jon Griffin
- Re: [AOLSERVER] New release? Nathan Folkman
- Re: [AOLSERVER] New release? Jon Griffin
- Re: [AOLSERVER] New release? Nathan Folkman
- Re: [AOLSERVER] New release? Jim Davidson
- Re: [AOLSERVER] New release? aT
- Re: [AOLSERVER] New release? Daniël Mantione
- Re: [AOLSERVER] New release? Rick Gutleber
- Re: [AOLSERVER] New release? Jim Davidson
- Re: [AOLSERVER] New release? Daniël Mantione