Verifed. work on both x86 and sparc. laca or Dermot, can you open the branch for me? Tom already approve this. thanks.
Brian Cameron wrote: > >>> However, it's ugly and makes it impossible to upstream the patch, >>> so please fix the patch instead so that it can be used on >>> sparc and x86. >> Good point. I think Joerg has come up with a proposed fix in this >> thread :) > > The gst-plugins-good-02-cdda.diff patch is not Sparc only so %ifarch is > not a possibility. Also, Joerg's patch is not a short-term fix. Joerg > provided a patch to cdda2wav. It would take several weeks, at the > earliest to coordinate with the team who owns cdda2wav to integrate this > fix. Once Joerg's fix gets integrated into Nevada we could use it, > though. > > For now, I recommend the attached patch. The attached gst-diff.txt > shows the difference between the previous version of the patch. > Basically, this simply does the following and sets CDDA2WAV_BYTE_ORDER > to the right argument to use on each platform, and passes that along > to cdda2wav. I'd think this should work okay, though I haven't tested > it. Jerry could you test this patch on both x86 and Sparc and commit > it into spec-files if it resolves the problem? > > Thanks, > > Brian > > ---- > > ++#if G_BYTE_ORDER == G_LITTLE_ENDIAN > ++#define CDDA2WAV_BYTE_ORDER "little" > ++#elif G_BYTE_ORDER == G_BIG_ENDIAN > ++#define CDDA2WAV_BYTE_ORDER "big" > ++#else > ++#error "G_BYTE_ORDER should be big or little endian." > ++#endif > > [...] > > -+ "/usr/bin/cdda2wav -q dev=%s -Oraw %s -interactive -", > -+ src->current_device, do_paranoia ? "-paranoia > -paraopts=minoverlap=2":""); > ++ "/usr/bin/cdda2wav -q --output-endianess=%s dev=%s -Oraw %s > -interactive -", > ++ CDDA2WAV_BYTE_ORDER, src->current_device, > ++ do_paranoia ? "-paranoia -paraopts=minoverlap=2":"");
