> If I use the debugger to manually set target->node_revision to NULL
> inside svn_fs_fs__dag_increment_mergeinfo_count then the commit works.
> I'm not exactly sure how all the FSFS caching layers are supposed to
> interact. Is tree.c:update_ancestry supposed to update the in-memory
> predecessor_count? Should there be a svn_fs_fs__dag_xxx function to
> change the predecessor count? Should target->node_revision be set to
> NULL soemwehere? Something else?
Moving update_ancestry from tree.c to dag.c is one way to fix the
problem. Daniel also suggested removing the node_revision member of
dag_node_t altogether and relying on new 1.7 caching to give us the
performance. I suppose we would still need a patch like this for 1.6.