Comment 105 for bug 553745

Review of attachment 36279:

::: mountall-2.14/src/mountall.c
@@ +2742,2 @@
    nih_assert_not_reached ();
   }

This stuff seems to be unrelated to plymouth

::: plymouth-0.8.2/src/libply/ply-event-loop.c
@@ +1193,3 @@

+ if (watch == NULL)
+ {

This shouldn't even happen. If this is happening there must be a bug some where else in the code.

@@ +1248,3 @@
+ number_of_received_events = epoll_wait (loop->epoll_fd,
+ events, PLY_EVENT_LOOP_NUM_EVENT_HANDLERS,
+ timeout);

Ah, this is a nice catch! Hard to believe this bug went unnoticed for so long. Unfortunately, it just means plymouth was processing less events per iteration than its configured upper limit (8 instead of 64). I don't think this is related to your crash. I've committed this part though: http://cgit.freedesktop.org/plymouth/commit/?id=26bc6f87c0507d0b457b23548094c475e00e75c9

::: plymouth-0.8.2/src/libply/ply-list.c
@@ +225,2 @@
   list->number_of_nodes--;
+ // assert (ply_list_find_node (list, node->data) != node);

again, if this is failing there is a bug somewhere else in the code.