Allow ext3 pages to be written back by VM pressure
PATCH 5/10 Allow ext3 pages to be written back by VM pressure
When ext3 refiles journalled buffers for writeback, it is better to set the page dirty as well as the buffer. So the page will be written out by VM pressure rather than by kjournald alone.
=====================================
fs/jbd/transaction.c renamed from fs/jbd/transaction.c~ext3-mark_buffer_dirty +1 -1
@@ -1548,7 +1548,7 @@ void __journal_unfile_buffer(struct jour__blist_del_buffer (list, jh); jh->b_jlist = BJ_None;if (test_and_clear_bit (BH_JBDDirty, &jh2bh (jh)->b_state)) -set_buffer_dirty (jh2bh (jh)); +mark_buffer_dirty (jh2bh (jh));/* Expose it to the VM */ }void journal_unfile_buffer (struct journal_head *jh)
-