On Fri, Jun 11, 2010 at 11:51, Greg Stein <gst...@gmail.com> wrote: > On Fri, Jun 11, 2010 at 09:35, <phi...@apache.org> wrote: >>... >> +++ subversion/trunk/subversion/libsvn_wc/wc_db.c Fri Jun 11 13:35:24 2010 >>... >> @@ -4856,11 +4857,31 @@ svn_wc__db_global_relocate(svn_wc__db_t >> &rb.repos_relpath, &old_repos_root_url, >> &rb.repos_uuid, >> NULL, NULL, NULL, NULL, NULL, NULL, NULL, >> NULL, >> - NULL, NULL, NULL, NULL, NULL, NULL, NULL, >> NULL, >> + NULL, NULL, NULL, NULL, NULL, NULL, NULL, >> + &base_shadowed, >> NULL, NULL, >> db, local_dir_abspath, >> scratch_pool, scratch_pool)); >> >> + if (status == svn_wc__db_status_excluded >> + || status == svn_wc__db_status_incomplete) >> + { >> + svn_sqlite__stmt_t *stmt; >> + >> + SVN_ERR(svn_sqlite__get_statement(&stmt, pdh->wcroot->sdb, >> + STMT_SELECT_BASE_NODE)); >> + SVN_ERR(svn_sqlite__bindf(stmt, "is", >> + pdh->wcroot->wc_id, rb.local_relpath)); >> + SVN_ERR(svn_sqlite__step(&rb.have_base_node, stmt)); >> + SVN_ERR(svn_sqlite__reset(stmt)); >> + } >> + else if (base_shadowed || status == svn_wc__db_status_normal
I also think you should test base_shadowed first, so that you could avoid the extra sql query. >... Cheers, -g