On Sat, 13 Nov 2010 14:49:53 +0100 Benoit Izac <ben...@izac.org> wrote:
> Stevan Bajić wrote: > > >>> On my local 32/64 bit safe version I use: > >>> header = (hash_drv_header_t)((uintptr_t)map.addr + filepos); > >> > >> #include <stdint.h> is not present in the git version I have. > > > > Ahh. Yes. It's not in GIT on Source Forge. I started a while a go with > > a fork of DSPAM 3.9.1 RC1 with the intention to include into DSPAM > > some things that I need. One of those changes was addressing the issue > > with the Hash driver not being 32/64 Bit safe (current DSPAM version > > will break terribly if you use the CSS files on 32 Bit AND on the same > > time on 64 Bit systems or you move from 32 Bit to 64 Bit or vice > > versa). The other issue with the Hash driver is that it has an issue > > with endianness (aka: it does not distinquish between big-endian or > > little-endian). And I was anyway dealing with changes in the > > PostgreSQL driver (switching to binary instead of ASCII) and > > PostgreSQL required me to send the data in network order so I anyway > > had to add stdint.h (for *nix. Windows needs other headers) on > > a global lever in order to be able to create proper functions for 64 > > Bit values (aka: hton64, ntoh64) for the various platforms on which > > DSPAM is running. Anyway... such small task like making DSPAM as much > > as possible Bit-independent and endianness aware ended in a massive > > change in the source tree of DSPAM. And I have not commited any of > > those changes to GIT. So this is one of the reasons current GIT does > > not use stdint.h. > > I see the problem (I have only 64 bits PC). Would Something like > <http://c-faq.com/stdio/extconform.html> not be sufficient? > Yes. It would be. I mean: the direction is right, but you would need to code much more then just the transformation from one side to the other. You need to implement it in every part of the code where binary data is read or writen. > >>> I could commit your code if you want me to do so. > >> > >> Yes, please as it avoids a segmentation fault. > > > > Those two lines just avoid one of those segmentation faults. There are > > others which are still not fixed with those two lines. > > Which (assuming that I stay on the same machine so not 32/64 bits or > endianness relative)? > I don't understand that sentence. Can you rephraze? > > Anyway.... I will commit your patch. But not right now. Will do that > > in about 12 to 18 hours from now on. > > Thank you. > No problem. It is already in GIT. > Regards. > -- > Benoit Izac > -- Kind Regards from Switzerland, Stevan Bajić ------------------------------------------------------------------------------ Centralized Desktop Delivery: Dell and VMware Reference Architecture Simplifying enterprise desktop deployment and management using Dell EqualLogic storage and VMware View: A highly scalable, end-to-end client virtualization framework. Read more! http://p.sf.net/sfu/dell-eql-dev2dev _______________________________________________ Dspam-devel mailing list Dspam-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/dspam-devel
