audit: don't needlessly reset valid wait time
authorRichard Guy Briggs <rgb@redhat.com>
Wed, 13 Jan 2016 14:15:18 +0000 (09:15 -0500)
committerPaul Moore <paul@paul-moore.com>
Wed, 13 Jan 2016 14:15:18 +0000 (09:15 -0500)
After auditd has recovered from an overflowed queue, the first process
that doesn't use reserves to make it through the queue checks should
reset the audit backlog wait time to the configured value.  After that,
there is no need to keep resetting it.

Signed-off-by: Richard Guy Briggs <rgb@redhat.com>
Signed-off-by: Paul Moore <pmoore@redhat.com>
kernel/audit.c

index 5ffcbd354a520b88781ed2d66c7839a7aaa7f86d..6d00bd1ff249b3d25be863d1e95f97098b65bc69 100644 (file)
@@ -1400,7 +1400,7 @@ struct audit_buffer *audit_log_start(struct audit_context *ctx, gfp_t gfp_mask,
                return NULL;
        }
 
-       if (!reserve)
+       if (!reserve && !audit_backlog_wait_time)
                audit_backlog_wait_time = audit_backlog_wait_time_master;
 
        ab = audit_buffer_alloc(ctx, gfp_mask, type);