freebsd-dev/sys/geom/geom_kern.c
Alexander Motin 40ea77a036 Merge GEOM direct dispatch changes from the projects/camlock branch.
When safety requirements are met, it allows to avoid passing I/O requests
to GEOM g_up/g_down thread, executing them directly in the caller context.
That allows to avoid CPU bottlenecks in g_up/g_down threads, plus avoid
several context switches per I/O.

The defined now safety requirements are:
 - caller should not hold any locks and should be reenterable;
 - callee should not depend on GEOM dual-threaded concurency semantics;
 - on the way down, if request is unmapped while callee doesn't support it,
   the context should be sleepable;
 - kernel thread stack usage should be below 50%.

To keep compatibility with GEOM classes not meeting above requirements
new provider and consumer flags added:
 - G_CF_DIRECT_SEND -- consumer code meets caller requirements (request);
 - G_CF_DIRECT_RECEIVE -- consumer code meets callee requirements (done);
 - G_PF_DIRECT_SEND -- provider code meets caller requirements (done);
 - G_PF_DIRECT_RECEIVE -- provider code meets callee requirements (request).
Capable GEOM class can set them, allowing direct dispatch in cases where
it is safe.  If any of requirements are not met, request is queued to
g_up or g_down thread same as before.

Such GEOM classes were reviewed and updated to support direct dispatch:
CONCAT, DEV, DISK, GATE, MD, MIRROR, MULTIPATH, NOP, PART, RAID, STRIPE,
VFS, ZERO, ZFS::VDEV, ZFS::ZVOL, all classes based on g_slice KPI (LABEL,
MAP, FLASHMAP, etc).

To declare direct completion capability disk(9) KPI got new flag equivalent
to G_PF_DIRECT_SEND -- DISKFLAG_DIRECT_COMPLETION.  da(4) and ada(4) disk
drivers got it set now thanks to earlier CAM locking work.

This change more then twice increases peak block storage performance on
systems with manu CPUs, together with earlier CAM locking changes reaching
more then 1 million IOPS (512 byte raw reads from 16 SATA SSDs on 4 HBAs to
256 user-level threads).

Sponsored by:	iXsystems, Inc.
MFC after:	2 months
2013-10-22 08:22:19 +00:00

236 lines
6.7 KiB
C

/*-
* Copyright (c) 2002 Poul-Henning Kamp
* Copyright (c) 2002 Networks Associates Technology, Inc.
* All rights reserved.
*
* This software was developed for the FreeBSD Project by Poul-Henning Kamp
* and NAI Labs, the Security Research Division of Network Associates, Inc.
* under DARPA/SPAWAR contract N66001-01-C-8035 ("CBOSS"), as part of the
* DARPA CHATS research program.
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
* are met:
* 1. Redistributions of source code must retain the above copyright
* notice, this list of conditions and the following disclaimer.
* 2. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. The names of the authors may not be used to endorse or promote
* products derived from this software without specific prior written
* permission.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND
* ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
* IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
* ARE DISCLAIMED. IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE
* FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
* DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
* OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
* HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
* LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
* OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
* SUCH DAMAGE.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/eventhandler.h>
#include <sys/malloc.h>
#include <sys/bio.h>
#include <sys/sysctl.h>
#include <sys/proc.h>
#include <sys/unistd.h>
#include <sys/kthread.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/sbuf.h>
#include <sys/sched.h>
#include <sys/sx.h>
#include <geom/geom.h>
#include <geom/geom_int.h>
MALLOC_DEFINE(M_GEOM, "GEOM", "Geom data structures");
struct sx topology_lock;
static struct proc *g_proc;
static struct thread *g_up_td;
static struct thread *g_down_td;
static struct thread *g_event_td;
int g_debugflags;
int g_collectstats = 1;
int g_shutdown;
int g_notaste;
/*
* G_UP and G_DOWN are the two threads which push I/O through the
* stack.
*
* Things are procesed in a FIFO order, but these threads could be
* part of I/O prioritization by deciding which bios/bioqs to service
* in what order.
*
* We have only one thread in each direction, it is belived that until
* a very non-trivial workload in the UP/DOWN path this will be enough,
* but more than one can actually be run without problems.
*
* Holding the "mymutex" is a debugging feature: It prevents people
* from sleeping in the UP/DOWN I/O path by mistake or design (doing
* so almost invariably result in deadlocks since it stalls all I/O
* processing in the given direction.
*/
static void
g_up_procbody(void *arg)
{
mtx_assert(&Giant, MA_NOTOWNED);
thread_lock(g_up_td);
sched_prio(g_up_td, PRIBIO);
thread_unlock(g_up_td);
for(;;) {
g_io_schedule_up(g_up_td);
}
}
static void
g_down_procbody(void *arg)
{
mtx_assert(&Giant, MA_NOTOWNED);
thread_lock(g_down_td);
sched_prio(g_down_td, PRIBIO);
thread_unlock(g_down_td);
for(;;) {
g_io_schedule_down(g_down_td);
}
}
static void
g_event_procbody(void *arg)
{
mtx_assert(&Giant, MA_NOTOWNED);
thread_lock(g_event_td);
sched_prio(g_event_td, PRIBIO);
thread_unlock(g_event_td);
g_run_events();
/* NOTREACHED */
}
int
g_is_geom_thread(struct thread *td)
{
return (td == g_up_td || td == g_down_td || td == g_event_td);
}
static void
geom_shutdown(void *foo __unused)
{
g_shutdown = 1;
}
void
g_init(void)
{
g_trace(G_T_TOPOLOGY, "g_ignition");
sx_init(&topology_lock, "GEOM topology");
g_io_init();
g_event_init();
g_ctl_init();
mtx_lock(&Giant);
kproc_kthread_add(g_event_procbody, NULL, &g_proc, &g_event_td,
RFHIGHPID, 0, "geom", "g_event");
kproc_kthread_add(g_up_procbody, NULL, &g_proc, &g_up_td,
RFHIGHPID, 0, "geom", "g_up");
kproc_kthread_add(g_down_procbody, NULL, &g_proc, &g_down_td,
RFHIGHPID, 0, "geom", "g_down");
mtx_unlock(&Giant);
EVENTHANDLER_REGISTER(shutdown_pre_sync, geom_shutdown, NULL,
SHUTDOWN_PRI_FIRST);
}
static int
sysctl_kern_geom_conftxt(SYSCTL_HANDLER_ARGS)
{
int error;
struct sbuf *sb;
sb = sbuf_new_auto();
g_waitfor_event(g_conftxt, sb, M_WAITOK, NULL);
error = SYSCTL_OUT(req, sbuf_data(sb), sbuf_len(sb) + 1);
sbuf_delete(sb);
return error;
}
static int
sysctl_kern_geom_confdot(SYSCTL_HANDLER_ARGS)
{
int error;
struct sbuf *sb;
sb = sbuf_new_auto();
g_waitfor_event(g_confdot, sb, M_WAITOK, NULL);
error = SYSCTL_OUT(req, sbuf_data(sb), sbuf_len(sb) + 1);
sbuf_delete(sb);
return error;
}
static int
sysctl_kern_geom_confxml(SYSCTL_HANDLER_ARGS)
{
int error;
struct sbuf *sb;
sb = sbuf_new_auto();
g_waitfor_event(g_confxml, sb, M_WAITOK, NULL);
error = SYSCTL_OUT(req, sbuf_data(sb), sbuf_len(sb) + 1);
sbuf_delete(sb);
return error;
}
SYSCTL_NODE(_kern, OID_AUTO, geom, CTLFLAG_RW, 0, "GEOMetry management");
SYSCTL_PROC(_kern_geom, OID_AUTO, confxml, CTLTYPE_STRING|CTLFLAG_RD,
0, 0, sysctl_kern_geom_confxml, "",
"Dump the GEOM config in XML");
SYSCTL_PROC(_kern_geom, OID_AUTO, confdot, CTLTYPE_STRING|CTLFLAG_RD,
0, 0, sysctl_kern_geom_confdot, "",
"Dump the GEOM config in dot");
SYSCTL_PROC(_kern_geom, OID_AUTO, conftxt, CTLTYPE_STRING|CTLFLAG_RD,
0, 0, sysctl_kern_geom_conftxt, "",
"Dump the GEOM config in txt");
TUNABLE_INT("kern.geom.debugflags", &g_debugflags);
SYSCTL_INT(_kern_geom, OID_AUTO, debugflags, CTLFLAG_RW,
&g_debugflags, 0, "Set various trace levels for GEOM debugging");
SYSCTL_INT(_kern_geom, OID_AUTO, notaste, CTLFLAG_RW,
&g_notaste, 0, "Prevent GEOM tasting");
SYSCTL_INT(_kern_geom, OID_AUTO, collectstats, CTLFLAG_RW,
&g_collectstats, 0,
"Control statistics collection on GEOM providers and consumers");
SYSCTL_INT(_debug_sizeof, OID_AUTO, g_class, CTLFLAG_RD,
0, sizeof(struct g_class), "sizeof(struct g_class)");
SYSCTL_INT(_debug_sizeof, OID_AUTO, g_geom, CTLFLAG_RD,
0, sizeof(struct g_geom), "sizeof(struct g_geom)");
SYSCTL_INT(_debug_sizeof, OID_AUTO, g_provider, CTLFLAG_RD,
0, sizeof(struct g_provider), "sizeof(struct g_provider)");
SYSCTL_INT(_debug_sizeof, OID_AUTO, g_consumer, CTLFLAG_RD,
0, sizeof(struct g_consumer), "sizeof(struct g_consumer)");
SYSCTL_INT(_debug_sizeof, OID_AUTO, g_bioq, CTLFLAG_RD,
0, sizeof(struct g_bioq), "sizeof(struct g_bioq)");