My provider is listed by Spamcop - fairly annoying all those bounces today...
--- Begin Message ---Paolo Mantegazza wrote: > Jan Kiszka wrote: >> Jorge Almeida wrote: >> >>> I'm using RTAI 3.4 and RTnet SVN >>> >>> The following error appears compiling Rtnet SVN >>> >>> CC [M] /home/jorge/nfs_mount/rtnet/stack/iovec.o >>> CC [M] /home/jorge/nfs_mount/rtnet/stack/rtdev.o >>> CC [M] /home/jorge/nfs_mount/rtnet/stack/rtdev_mgr.o >>> In file included from >>> /home/jorge/nfs_mount/rtnet/stack/include/rtskb.h:32, >>> from >>> /home/jorge/nfs_mount/rtnet/stack/include/rtdev.h:38, >>> from /home/jorge/nfs_mount/rtnet/stack/rtdev_mgr.c:25: >>> /home/jorge/nfs_mount/rtnet/stack/include/rtnet.h:39: error: >>> conflicting types for 'nanosecs_abs_t' >>> /usr/realtime/include/rtdm/rtdm.h:79: error: previous declaration of >>> 'nanosecs_abs_t' was here >>> /home/jorge/nfs_mount/rtnet/stack/include/rtnet.h:40: error: >>> redefinition of typedef 'nanosecs_rel_t' >>> /usr/realtime/include/rtdm/rtdm.h:85: error: previous declaration of >>> 'nanosecs_rel_t' was here >>> make[4]: *** [/home/jorge/nfs_mount/rtnet/stack/rtdev_mgr.o] Error 1 >>> make[3]: *** [_module_/home/jorge/nfs_mount/rtnet/stack] Error 2 >>> make[3]: Leaving directory `/usr/src/linux-2.6.16.20' >>> make[2]: *** [all-local.ko] Error 2 >>> make[2]: Leaving directory `/home/jorge/nfs_mount/rtnet/stack' >>> make[1]: *** [all-recursive] Error 1 >>> make[1]: Leaving directory `/home/jorge/nfs_mount/rtnet/stack' >>> make: *** [all-recursive] Error 1 >>> >>> >>> maybe something is duplicated. >>> >>> Its already defined in the RTDM and do not need to be defined in >>> RTnet again. >> >> >> Yes, but RTnet takes care to only define it on its own if RTDM doesn't >> support it yet. And that information is derived from RTDM_API_VER. >> >> [Digging deeper...] >> >> Sigh. >> >> Paolo, could you explain why you copied RTDM code from Xenomai trunk >> into a RTAI release right in the middle of our development cycle? We >> haven't released any new Xenomai version yet that includes RTDM revision >> 5, nor was that revision finalised in any other way. I simply pushed the >> number forward to test the required detection code in drivers like RTnet. >> >> The effect is that RTAI 3.4 now contains broken code. Please fix this >> for 3.4.1 or whatever the next release is. I'm offering you to freeze >> the current feature-set to revision 5 and push the number forward for >> the pending patches. But please avoid such chaos in the future. >> >> RTDM is gaining ground in industry, hardware vendor are providing >> drivers against this API. If you want that also RTAI benefits from this, >> you should really take care to keep the interfaces consistent. >> > > Well since I'm not worth being informed of what is going on directly > with RTDM, I just diff now and then, test it with all my in house tests > for a day or so and if it works I commit. In this case I saw changes, > they worked, believed it was a good thing to upgrade and did it. It has > worked so thus far. >It's simple: if there are news, RTDM revisions follow the Xenomai releases. If you want to include latest, yet unreleased changes into a RTAI *release* and you are not sure if RTDM can already be considered stable, feel free to ask me. I will surely not refrain from marking an intermediate step, specifically when it is as easy as incrementing a number. But I cannot follow the RTAI development in every detail, nor do I know in advance when the next release will be rolled out. > In this particular case I must admit I do not even know how to align the > stuff. How and where is RTAI broken? What should I do? > Upgrade to the latest version that is now in trunk, it should contain all you need to get RTnet compiling again. Background: Due to the new time types, this first change to rtdm.h, the user API header, required to push the revision number from rtdm_driver.h to rtdm.h. You hit the race window when this hasn't happened yet, but the new types were already committed. Jan
signature.asc
Description: OpenPGP digital signature
--- End Message ---
signature.asc
Description: OpenPGP digital signature
------------------------------------------------------------------------- Take Surveys. Earn Cash. Influence the Future of IT Join SourceForge.net's Techsay panel and you'll get the chance to share your opinions on IT & business topics through brief surveys -- and earn cash http://www.techsay.com/default.php?page=join.php&p=sourceforge&CID=DEVDEV
_______________________________________________ RTnet-users mailing list RTnet-users@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/rtnet-users
