Author: glebius
Date: Mon Apr 2 05:11:59 2018
New Revision: 331871
URL: https://svnweb.freebsd.org/changeset/base/331871
Log:
Handle a special case when a slab can fit only one allocation,
and zone has a large alignment. With alignment taken into
account uk_rsize will be greater than space in a slab. However,
since we have only one item per slab, it is always naturally
aligned.
Code that will panic before this change with 4k page:
z = uma_zcreate("test", 3984, NULL, NULL, NULL, NULL, 31, 0);
uma_zalloc(z, M_WAITOK);
A practical scenario to hit the panic is a machine with 56 CPUs
and 2 NUMA domains, which yields in zone size of 3984.
PR: 227116
MFC after: 2 weeks
Modified:
head/sys/vm/uma_core.c
Modified: head/sys/vm/uma_core.c
==============================================================================
--- head/sys/vm/uma_core.c Sun Apr 1 22:59:53 2018 (r331870)
+++ head/sys/vm/uma_core.c Mon Apr 2 05:11:59 2018 (r331871)
@@ -1290,7 +1290,15 @@ keg_small_init(uma_keg_t keg)
else
shsize = sizeof(struct uma_slab);
- keg->uk_ipers = (slabsize - shsize) / rsize;
+ if (rsize <= slabsize - shsize)
+ keg->uk_ipers = (slabsize - shsize) / rsize;
+ else {
+ /* Handle special case when we have 1 item per slab, so
+ * alignment requirement can be relaxed. */
+ KASSERT(keg->uk_size <= slabsize - shsize,
+ ("%s: size %u greater than slab", __func__, keg->uk_size));
+ keg->uk_ipers = 1;
+ }
KASSERT(keg->uk_ipers > 0 && keg->uk_ipers <= SLAB_SETSIZE,
("%s: keg->uk_ipers %u", __func__, keg->uk_ipers));
_______________________________________________
svn-src-all@freebsd.org mailing list
https://lists.freebsd.org/mailman/listinfo/svn-src-all
To unsubscribe, send any mail to "svn-src-all-unsubscr...@freebsd.org"
