r357018 rlibby Jan. 23, 2020, 4:56 a.m.
Previously UMA had some false negatives in the leak report at keg
destruction time, where it only reported leaks if there were free items
in the slab layer (rather than allocated items), which notably would not
be true for single-item slabs (large items).  Now, report a leak if
there are any allocated pages, and calculate and report the number of
allocated items rather than free items.

Reviewed by:	jeff, markj
Sponsored by:	Dell EMC Isilon
Differential Revision:	https://reviews.freebsd.org/D23275
ViewVC
r357017 jeff Jan. 23, 2020, 4:54 a.m.
directly.  This improves API compliance, asserts, etc.

Reviewed by:	kib, markj
Differential Revision:	https://reviews.freebsd.org/D23283
ViewVC
r357016 cy Jan. 23, 2020, 3:38 a.m.
r357015 jeff Jan. 23, 2020, 3:37 a.m.
claiming kva in uma_startup2() to handle this.

Reported by:	bdragon
ViewVC
r357014 jeff Jan. 23, 2020, 3:36 a.m.
it.  The introduction of lockless switch in r355784 created a race to
re-use the exiting thread that was only possible to hit on a hypervisor.

Reported/Tested by:	rlibby
Discussed with:	rlibby, jhb
ViewVC
r357013 glebius Jan. 23, 2020, 1:49 a.m.
epoch, stop entering it in pfil_run_hooks().  Assert the epoch there.
ViewVC
r357012 glebius Jan. 23, 2020, 1:47 a.m.
is marked with IFF_NEEDSEPOCH.
ViewVC
r357011 glebius Jan. 23, 2020, 1:46 a.m.
r357010 glebius Jan. 23, 2020, 1:41 a.m.
supposedly may call into ether_input() without network epoch.

They all need to be reviewed before 13.0-RELEASE.  Some may need
be fixed.  The flag is not planned to be used in the kernel for
a long time.
ViewVC
r357009 glebius Jan. 23, 2020, 1:38 a.m.
epoch here.

The tun(4) side doesn't need this, as netisr code will take care.
ViewVC
r357008 glebius Jan. 23, 2020, 1:35 a.m.
r357007 glebius Jan. 23, 2020, 1:30 a.m.
needs to enter epoch.  Assert that in the netisr_poll() and do
the work for the idle poll routine.
ViewVC
r357006 glebius Jan. 23, 2020, 1:27 a.m.
In upcoming changes ether_input() is going to be changed not
to enter the network epoch.  It is going to be responsibility
of network interrupt.  In case of iflib - its taskqueue.
ViewVC
r357005 glebius Jan. 23, 2020, 1:25 a.m.
r357004 glebius Jan. 23, 2020, 1:24 a.m.
Provide tunable to limit how many times handlers may be executed
without reentering epoch.

Differential Revision:	https://reviews.freebsd.org/D23242
ViewVC