On Nov 19, 2010, at 6:28 AM, Florian Schwind wrote: > On 19.11.2010 11:12, Florian Schwind wrote: >> Hi all, >> >> I was not able to run the "make ckeck" for GPG 1.4.10 on a SLES 11 >> (i386) successfully (I also tried 1.4.11). Since I'm not sure if there >> is some configuration issue with my server, I'm sending this to the >> bug-address and the user-list as well. >> >> Maybe someone experienced similar problems with GPG on a SLES 11 and >> might help me to figure out what's wrong? (On my SuSE 11.3 x86 it >> compiles without problems) > > Hi, > > I just found out that there seems to be a problem with gcc 4.3.2 and > rijndael.c witch is already known. > (http://lists.gnupg.org/pipermail/gnupg-devel/2008-December/024721.html). > > Option 1 is to updated to gcc 4.5.0 which works or the use CFLAGS="-O1" > instead of O2 and don't optimize the code heavily. > > Perhaps someone wants to look into rijndael.c and find out where the > optimizationproblems with gcc come from?
I was the one who originally tracked this down. Virtually always, when people suspect a problem with gcc or libc, the problem is really in their own code. This was one of the very rare exceptions, and the gcc folks fixed the optimizer issue. I'm not sure I see any benefit in looking at or changing the GPG code in an effort to not trigger a bug in a quite old - and long since replaced - version of gcc. David _______________________________________________ Gnupg-users mailing list Gnupg-users@gnupg.org http://lists.gnupg.org/mailman/listinfo/gnupg-users
