Hi all, first thanks to all who provided me with hints towards this issue; lacking the appropriate hardware I would have never found out myself ;-)
> however, will build without sse2 then in future.
This is an appropriate workaround for dvdisaster 0.7x, since it does not gain much from SSE2 yet. However looking a bit into the future I see hand-crafted SSE2 code appearing in the RS codec of V0.80. Then SSE2 should not be disabled globally on 686 archs since it gives a nice performance boost on machines which do support it. There is still enough time figuring out how to test for it properly, though.
I would, however, also be highly interested as to why configure thinks we do have SSE2 ...
It is incorrectly assuming that a SSE2-able machine can be determined at compile time by testing whether gcc supports -msse2. The fallacy behind this reasoning is obvious considering the way debian (and other GNU/Linux) packages work ;-) A corrected upstream version will be available on Monday. Cheers, Carsten "cos(0.0) = 1" G. -- Carsten Gnörlich * Project homepage http://www.dvdisaster.org dvdisaster: Additional error correction for CD and DVD media. GnuPG FP: 12B3 1535 AF90 3ADE 9E73 BA7E 5A59 0EFE F5F6 C46C -- To UNSUBSCRIBE, email to debian-bugs-dist-requ...@lists.debian.org with a subject of "unsubscribe". Trouble? Contact listmas...@lists.debian.org
