splassert on some archs (or just sparc64) check that you're not in

an interrupt handler at an ipl level higher than what you're
splasserting you should be at. if you think code should be protected
by IPL_BIO and its entered from an interrupt handler established
at IPL_NET, you have a bug.

add some asserts to gets and puts so we can pick those cases up.

diff --git a/sys/kern/subr_pool.c b/sys/kern/subr_pool.c
index 678e93a..602ff47 100644 (file)
--- a/sys/kern/subr_pool.c
+++ b/sys/kern/subr_pool.c
@@ -1,4 +1,4 @@
-/*     $OpenBSD: subr_pool.c,v 1.176 2015/01/04 08:54:01 dlg Exp $     */
+/*     $OpenBSD: subr_pool.c,v 1.177 2015/01/05 23:54:18 dlg Exp $     */
 /*     $NetBSD: subr_pool.c,v 1.61 2001/09/26 07:14:56 chs Exp $       */
 
 /*-
@@ -536,6 +536,9 @@ pool_do_get(struct pool *pp, int flags, int *slowdown)
 
        MUTEX_ASSERT_LOCKED(&pp->pr_mtx);
 
+       if (pp->pr_ipl != -1)
+               splassert(pp->pr_ipl);
+
        /*
         * Account for this item now to avoid races if we need to give up
         * pr_mtx to allocate a page.
@@ -627,6 +630,9 @@ pool_put(struct pool *pp, void *v)
 
        mtx_enter(&pp->pr_mtx);
 
+       if (pp->pr_ipl != -1)
+               splassert(pp->pr_ipl);
+
        ph = pr_find_pagehead(pp, v);
 
 #ifdef DIAGNOSTIC