Hi folks, A couple of quick fixes - the new verifier code leaks EFSBADCRC to userspace, so any code that is checking for EFSCORRUPTED is not going to be working correctly. The first patch fixes this regression. The second patch fixes a long standing issues with vmapped RAM potentially causing deadlocks due to GFP_KERNEL allocations being done for page tables inside GFP_NOFS contexts. This will fix many of the lockdep reports about it, but won't get rid of them all. It will, however, guarantee that we won't deadlock due to the problem which is a big step up. Cheers, Dave. _______________________________________________ xfs mailing list xfs@xxxxxxxxxxx http://oss.sgi.com/mailman/listinfo/xfs
