I incorrectly assumed the log buffer started from the beginning when it
is actually a circular log. Deal with it rolling around. Fortunately, the log area is big and I haven't seen any roll over yet. Update and get rid of the obsolete comment.
This commit is contained in:
parent
edee84f0a7
commit
a41ac17851
@ -774,16 +774,12 @@ mfi_aen_setup(struct mfi_softc *sc, uint32_t seq_start)
|
||||
free(log_state, M_MFIBUF);
|
||||
return (error);
|
||||
}
|
||||
/*
|
||||
* Don't run them yet since we can't parse them.
|
||||
* We can indirectly get the contents from
|
||||
* the AEN mechanism via setting it lower then
|
||||
* current. The firmware will iterate through them.
|
||||
*/
|
||||
/* The message log is a circular buffer */
|
||||
for (seq = log_state->shutdown_seq_num;
|
||||
seq <= log_state->newest_seq_num; seq++) {
|
||||
seq != log_state->newest_seq_num; seq++) {
|
||||
mfi_get_entry(sc, seq);
|
||||
}
|
||||
mfi_get_entry(sc, seq);
|
||||
} else
|
||||
seq = seq_start;
|
||||
mfi_aen_register(sc, seq, class_locale.word);
|
||||
|
Loading…
Reference in New Issue
Block a user