On Fri 12 Dec 2008 01:34, [email protected] pondered: > Log Message > > Add anomaly 05000412: TESTSET L2 data and there is data in external memory > which is cached in writeback mode can cause either or both data corrupt >
Doing this sort of thing is bad (IMHO). If you want to update things, great - but there are some rules to follow to make sure everything works. - anomaly.h in the kernel should match exactly the anomaly sheet versions noted in the top of the file. Now they do not. Why weren't the other anomalies added(05000416, 05000425, 05000426,05000428, 05000443 were added to Rev Q of the 561 anomaly sheet)? Why set everything to zero in the other mach files when there was no workaround needed in common code (which is the only time we do that). - anomaly.h in the kernel should be a copy from the toolchain's toolchain/proc-defs/header-frags/*/anomaly.h files. These are the master location, since anomaly.h is used in multiple projects. Now they do not. (and next time they are synced - any changes may be lost). _______________________________________________ Linux-kernel-commits mailing list [email protected] http://blackfin.uclinux.org/mailman/listinfo/linux-kernel-commits
