freebsd-nq/sys/geom/geom_subr.c

1546 lines
37 KiB
C
Raw Normal View History

/*-
* 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.
*/
2003-06-11 06:49:16 +00:00
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
#include "opt_ddb.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/devicestat.h>
#include <sys/kernel.h>
#include <sys/malloc.h>
#include <sys/bio.h>
#include <sys/sysctl.h>
#include <sys/proc.h>
#include <sys/kthread.h>
#include <sys/lock.h>
#include <sys/mutex.h>
#include <sys/errno.h>
#include <sys/sbuf.h>
#include <geom/geom.h>
#include <geom/geom_int.h>
#include <machine/stdarg.h>
#ifdef DDB
#include <ddb/ddb.h>
#endif
#ifdef KDB
#include <sys/kdb.h>
#endif
struct class_list_head g_classes = LIST_HEAD_INITIALIZER(g_classes);
static struct g_tailq_head geoms = TAILQ_HEAD_INITIALIZER(geoms);
char *g_wait_event, *g_wait_up, *g_wait_down, *g_wait_sim;
struct g_hh00 {
struct g_class *mp;
struct g_provider *pp;
off_t size;
int error;
int post;
};
/*
* This event offers a new class a chance to taste all preexisting providers.
*/
static void
g_load_class(void *arg, int flag)
{
struct g_hh00 *hh;
struct g_class *mp2, *mp;
struct g_geom *gp;
struct g_provider *pp;
g_topology_assert();
if (flag == EV_CANCEL) /* XXX: can't happen ? */
return;
if (g_shutdown)
return;
hh = arg;
mp = hh->mp;
hh->error = 0;
if (hh->post) {
g_free(hh);
hh = NULL;
}
g_trace(G_T_TOPOLOGY, "g_load_class(%s)", mp->name);
KASSERT(mp->name != NULL && *mp->name != '\0',
("GEOM class has no name"));
LIST_FOREACH(mp2, &g_classes, class) {
if (mp2 == mp) {
printf("The GEOM class %s is already loaded.\n",
mp2->name);
if (hh != NULL)
hh->error = EEXIST;
return;
} else if (strcmp(mp2->name, mp->name) == 0) {
printf("A GEOM class %s is already loaded.\n",
mp2->name);
if (hh != NULL)
hh->error = EEXIST;
return;
}
}
LIST_INIT(&mp->geom);
LIST_INSERT_HEAD(&g_classes, mp, class);
if (mp->init != NULL)
mp->init(mp);
if (mp->taste == NULL)
return;
LIST_FOREACH(mp2, &g_classes, class) {
if (mp == mp2)
continue;
LIST_FOREACH(gp, &mp2->geom, geom) {
LIST_FOREACH(pp, &gp->provider, provider) {
mp->taste(mp, pp, 0);
g_topology_assert();
}
}
}
}
static int
g_unload_class(struct g_class *mp)
{
struct g_geom *gp;
struct g_provider *pp;
struct g_consumer *cp;
int error;
g_topology_lock();
g_trace(G_T_TOPOLOGY, "g_unload_class(%s)", mp->name);
retry:
G_VALID_CLASS(mp);
LIST_FOREACH(gp, &mp->geom, geom) {
/* We refuse to unload if anything is open */
LIST_FOREACH(pp, &gp->provider, provider)
if (pp->acr || pp->acw || pp->ace) {
g_topology_unlock();
return (EBUSY);
}
LIST_FOREACH(cp, &gp->consumer, consumer)
if (cp->acr || cp->acw || cp->ace) {
g_topology_unlock();
return (EBUSY);
}
/* If the geom is withering, wait for it to finish. */
if (gp->flags & G_GEOM_WITHER) {
g_topology_sleep(mp, 1);
goto retry;
}
}
/*
* We allow unloading if we have no geoms, or a class
* method we can use to get rid of them.
*/
if (!LIST_EMPTY(&mp->geom) && mp->destroy_geom == NULL) {
g_topology_unlock();
return (EOPNOTSUPP);
}
/* Bar new entries */
mp->taste = NULL;
mp->config = NULL;
LIST_FOREACH(gp, &mp->geom, geom) {
error = mp->destroy_geom(NULL, mp, gp);
if (error != 0) {
g_topology_unlock();
return (error);
}
}
/* Wait for withering to finish. */
for (;;) {
gp = LIST_FIRST(&mp->geom);
if (gp == NULL)
break;
KASSERT(gp->flags & G_GEOM_WITHER,
("Non-withering geom in class %s", mp->name));
g_topology_sleep(mp, 1);
}
G_VALID_CLASS(mp);
if (mp->fini != NULL)
mp->fini(mp);
LIST_REMOVE(mp, class);
g_topology_unlock();
return (0);
}
int
g_modevent(module_t mod, int type, void *data)
{
struct g_hh00 *hh;
int error;
static int g_ignition;
struct g_class *mp;
mp = data;
if (mp->version != G_VERSION) {
printf("GEOM class %s has Wrong version %x\n",
mp->name, mp->version);
return (EINVAL);
}
if (!g_ignition) {
g_ignition++;
g_init();
}
error = EOPNOTSUPP;
switch (type) {
case MOD_LOAD:
g_trace(G_T_TOPOLOGY, "g_modevent(%s, LOAD)", mp->name);
hh = g_malloc(sizeof *hh, M_WAITOK | M_ZERO);
hh->mp = mp;
/*
* Once the system is not cold, MOD_LOAD calls will be
* from the userland and the g_event thread will be able
* to acknowledge their completion.
*/
if (cold) {
hh->post = 1;
error = g_post_event(g_load_class, hh, M_WAITOK, NULL);
} else {
error = g_waitfor_event(g_load_class, hh, M_WAITOK,
NULL);
if (error == 0)
error = hh->error;
g_free(hh);
}
break;
case MOD_UNLOAD:
g_trace(G_T_TOPOLOGY, "g_modevent(%s, UNLOAD)", mp->name);
error = g_unload_class(mp);
if (error == 0) {
KASSERT(LIST_EMPTY(&mp->geom),
("Unloaded class (%s) still has geom", mp->name));
}
break;
}
return (error);
}
static void
g_retaste_event(void *arg, int flag)
{
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
struct g_class *mp, *mp2;
struct g_geom *gp;
struct g_hh00 *hh;
struct g_provider *pp;
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
struct g_consumer *cp;
g_topology_assert();
if (flag == EV_CANCEL) /* XXX: can't happen ? */
return;
if (g_shutdown || g_notaste)
return;
hh = arg;
mp = hh->mp;
hh->error = 0;
if (hh->post) {
g_free(hh);
hh = NULL;
}
g_trace(G_T_TOPOLOGY, "g_retaste(%s)", mp->name);
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
LIST_FOREACH(mp2, &g_classes, class) {
LIST_FOREACH(gp, &mp2->geom, geom) {
LIST_FOREACH(pp, &gp->provider, provider) {
if (pp->acr || pp->acw || pp->ace)
continue;
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
LIST_FOREACH(cp, &pp->consumers, consumers) {
if (cp->geom->class == mp &&
(cp->flags & G_CF_ORPHAN) == 0)
break;
}
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
if (cp != NULL) {
cp->flags |= G_CF_ORPHAN;
g_wither_geom(cp->geom, ENXIO);
}
mp->taste(mp, pp, 0);
g_topology_assert();
}
}
}
}
int
g_retaste(struct g_class *mp)
{
struct g_hh00 *hh;
int error;
if (mp->taste == NULL)
return (EINVAL);
hh = g_malloc(sizeof *hh, M_WAITOK | M_ZERO);
hh->mp = mp;
if (cold) {
hh->post = 1;
error = g_post_event(g_retaste_event, hh, M_WAITOK, NULL);
} else {
error = g_waitfor_event(g_retaste_event, hh, M_WAITOK, NULL);
if (error == 0)
error = hh->error;
g_free(hh);
}
return (error);
}
struct g_geom *
g_new_geomf(struct g_class *mp, const char *fmt, ...)
{
struct g_geom *gp;
va_list ap;
struct sbuf *sb;
g_topology_assert();
G_VALID_CLASS(mp);
sb = sbuf_new_auto();
va_start(ap, fmt);
sbuf_vprintf(sb, fmt, ap);
va_end(ap);
sbuf_finish(sb);
gp = g_malloc(sizeof *gp, M_WAITOK | M_ZERO);
gp->name = g_malloc(sbuf_len(sb) + 1, M_WAITOK | M_ZERO);
gp->class = mp;
gp->rank = 1;
LIST_INIT(&gp->consumer);
LIST_INIT(&gp->provider);
LIST_INSERT_HEAD(&mp->geom, gp, geom);
TAILQ_INSERT_HEAD(&geoms, gp, geoms);
strcpy(gp->name, sbuf_data(sb));
sbuf_delete(sb);
/* Fill in defaults from class */
gp->start = mp->start;
gp->spoiled = mp->spoiled;
Plumb device physical path reporting from CAM devices, through GEOM and DEVFS, and make it accessible via the diskinfo utility. Extend GEOM's generic attribute query mechanism into generic disk consumers. sys/geom/geom_disk.c: sys/geom/geom_disk.h: sys/cam/scsi/scsi_da.c: sys/cam/ata/ata_da.c: - Allow disk providers to implement a new method which can override the default BIO_GETATTR response, d_getattr(struct bio *). This function returns -1 if not handled, otherwise it returns 0 or an errno to be passed to g_io_deliver(). sys/cam/scsi/scsi_da.c: sys/cam/ata/ata_da.c: - Don't copy the serial number to dp->d_ident anymore, as the CAM XPT is now responsible for returning this information via d_getattr()->(a)dagetattr()->xpt_getatr(). sys/geom/geom_dev.c: - Implement a new ioctl, DIOCGPHYSPATH, which returns the GEOM attribute "GEOM::physpath", if possible. If the attribute request returns a zero-length string, ENOENT is returned. usr.sbin/diskinfo/diskinfo.c: - If the DIOCGPHYSPATH ioctl is successful, report physical path data when diskinfo is executed with the '-v' option. Submitted by: will Reviewed by: gibbs Sponsored by: Spectra Logic Corporation Add generic attribute change notification support to GEOM. sys/sys/geom/geom.h: Add a new attrchanged method field to both g_class and g_geom. sys/sys/geom/geom.h: sys/geom/geom_event.c: - Provide the g_attr_changed() function that providers can use to advertise attribute changes. - Perform delivery of attribute change notifications from a thread context via the standard GEOM event mechanism. sys/geom/geom_subr.c: Inherit the attrchanged method from class to geom (class instance). sys/geom/geom_disk.c: Provide disk_attr_changed() to provide g_attr_changed() access to consumers of the disk API. sys/cam/scsi/scsi_pass.c: sys/cam/scsi/scsi_da.c: sys/geom/geom_dev.c: sys/geom/geom_disk.c: Use attribute changed events to track updates to physical path information. sys/cam/scsi/scsi_da.c: Add AC_ADVINFO_CHANGED to the registered asynchronous CAM events for this driver. When this event occurs, and the updated buffer type references our physical path attribute, emit a GEOM attribute changed event via the disk_attr_changed() API. sys/cam/scsi/scsi_pass.c: Add AC_ADVINFO_CHANGED to the registered asynchronous CAM events for this driver. When this event occurs, update the physical patch devfs alias for this pass instance. Submitted by: gibbs Sponsored by: Spectra Logic Corporation
2011-06-14 17:10:32 +00:00
gp->attrchanged = mp->attrchanged;
Fix a bug which causes a panic in daopen(). The panic is caused by a da(4) instance going away while GEOM is still probing it. In this case, the GEOM disk class instance has been created by disk_create(), and the taste of the disk is queued in the GEOM event queue. While that event is queued, the da(4) instance goes away. When the open call comes into the da(4) driver, it dereferences the freed (but non-NULL) peripheral pointer provided by GEOM, which results in a panic. The solution is to add a callback to the GEOM disk code that is called when all of its resources are cleaned up. This is implemented inside GEOM by adding an optional callback that is called when all consumers have detached from a provider, and the provider is about to be deleted. scsi_cd.c, scsi_da.c: In the register routine for the cd(4) and da(4) routines, acquire a reference to the CAM peripheral instance just before we call disk_create(). Use the new GEOM disk d_gone() callback to register a callback (dadiskgonecb()/cddiskgonecb()) that decrements the peripheral reference count once GEOM has finished cleaning up its resources. In the cd(4) driver, clean up open and close behavior slightly. GEOM makes sure we only get one open() and one close call, so there is no need to set an open flag and decrement the reference count if we are not the first open. In the cd(4) driver, use cam_periph_release_locked() in a couple of error scenarios to avoid extra mutex calls. geom.h: Add a new, optional, providergone callback that is called when a provider is about to be deleted. geom_disk.h: Add a new d_gone() callback to the GEOM disk interface. Bump the DISK_VERSION to version 2. This probably should have been done after a couple of previous changes, especially the addition of the d_getattr() callback. geom_disk.c: Add a providergone callback for the disk class, g_disk_providergone(), that calls the user's d_gone() callback if it exists. Bump the DISK_VERSION to 2. geom_subr.c: In g_destroy_provider(), call the providergone callback if it has been provided. In g_new_geomf(), propagate the class's providergone callback to the new geom instance. blkfront.c: Callers of disk_create() are supposed to pass in DISK_VERSION, not an explicit disk API version number. Update the blkfront driver to do that. disk.9: Update the disk(9) man page to include information on the new d_gone() callback, as well as the previously added d_getattr() callback, d_descr field, and HBA PCI ID fields. MFC after: 5 days
2012-06-24 04:29:03 +00:00
gp->providergone = mp->providergone;
gp->dumpconf = mp->dumpconf;
gp->access = mp->access;
gp->orphan = mp->orphan;
gp->ioctl = mp->ioctl;
gp->resize = mp->resize;
return (gp);
}
void
g_destroy_geom(struct g_geom *gp)
{
g_topology_assert();
G_VALID_GEOM(gp);
g_trace(G_T_TOPOLOGY, "g_destroy_geom(%p(%s))", gp, gp->name);
KASSERT(LIST_EMPTY(&gp->consumer),
("g_destroy_geom(%s) with consumer(s) [%p]",
gp->name, LIST_FIRST(&gp->consumer)));
KASSERT(LIST_EMPTY(&gp->provider),
("g_destroy_geom(%s) with provider(s) [%p]",
gp->name, LIST_FIRST(&gp->provider)));
g_cancel_event(gp);
LIST_REMOVE(gp, geom);
TAILQ_REMOVE(&geoms, gp, geoms);
g_free(gp->name);
g_free(gp);
}
/*
2008-12-08 17:09:02 +00:00
* This function is called (repeatedly) until the geom has withered away.
*/
void
g_wither_geom(struct g_geom *gp, int error)
{
struct g_provider *pp;
g_topology_assert();
G_VALID_GEOM(gp);
g_trace(G_T_TOPOLOGY, "g_wither_geom(%p(%s))", gp, gp->name);
if (!(gp->flags & G_GEOM_WITHER)) {
gp->flags |= G_GEOM_WITHER;
LIST_FOREACH(pp, &gp->provider, provider)
if (!(pp->flags & G_PF_ORPHAN))
g_orphan_provider(pp, error);
}
g_do_wither();
}
/*
* Convenience function to destroy a particular provider.
*/
void
g_wither_provider(struct g_provider *pp, int error)
{
pp->flags |= G_PF_WITHER;
if (!(pp->flags & G_PF_ORPHAN))
g_orphan_provider(pp, error);
}
2004-10-29 09:19:03 +00:00
/*
* This function is called (repeatedly) until the has withered away.
*/
void
g_wither_geom_close(struct g_geom *gp, int error)
{
struct g_consumer *cp;
g_topology_assert();
G_VALID_GEOM(gp);
g_trace(G_T_TOPOLOGY, "g_wither_geom_close(%p(%s))", gp, gp->name);
LIST_FOREACH(cp, &gp->consumer, consumer)
if (cp->acr || cp->acw || cp->ace)
g_access(cp, -cp->acr, -cp->acw, -cp->ace);
g_wither_geom(gp, error);
}
/*
* This function is called (repeatedly) until we cant wash away more
* withered bits at present.
*/
void
g_wither_washer()
{
struct g_class *mp;
struct g_geom *gp, *gp2;
struct g_provider *pp, *pp2;
struct g_consumer *cp, *cp2;
g_topology_assert();
LIST_FOREACH(mp, &g_classes, class) {
LIST_FOREACH_SAFE(gp, &mp->geom, geom, gp2) {
LIST_FOREACH_SAFE(pp, &gp->provider, provider, pp2) {
if (!(pp->flags & G_PF_WITHER))
continue;
if (LIST_EMPTY(&pp->consumers))
g_destroy_provider(pp);
}
if (!(gp->flags & G_GEOM_WITHER))
continue;
LIST_FOREACH_SAFE(pp, &gp->provider, provider, pp2) {
if (LIST_EMPTY(&pp->consumers))
g_destroy_provider(pp);
}
LIST_FOREACH_SAFE(cp, &gp->consumer, consumer, cp2) {
if (cp->acr || cp->acw || cp->ace)
continue;
if (cp->provider != NULL)
g_detach(cp);
g_destroy_consumer(cp);
}
if (LIST_EMPTY(&gp->provider) &&
LIST_EMPTY(&gp->consumer))
g_destroy_geom(gp);
}
}
}
struct g_consumer *
g_new_consumer(struct g_geom *gp)
{
struct g_consumer *cp;
g_topology_assert();
G_VALID_GEOM(gp);
KASSERT(!(gp->flags & G_GEOM_WITHER),
("g_new_consumer on WITHERing geom(%s) (class %s)",
gp->name, gp->class->name));
KASSERT(gp->orphan != NULL,
("g_new_consumer on geom(%s) (class %s) without orphan",
gp->name, gp->class->name));
cp = g_malloc(sizeof *cp, M_WAITOK | M_ZERO);
cp->geom = gp;
cp->stat = devstat_new_entry(cp, -1, 0, DEVSTAT_ALL_SUPPORTED,
DEVSTAT_TYPE_DIRECT, DEVSTAT_PRIORITY_MAX);
LIST_INSERT_HEAD(&gp->consumer, cp, consumer);
return(cp);
}
void
g_destroy_consumer(struct g_consumer *cp)
{
struct g_geom *gp;
g_topology_assert();
G_VALID_CONSUMER(cp);
g_trace(G_T_TOPOLOGY, "g_destroy_consumer(%p)", cp);
KASSERT (cp->provider == NULL, ("g_destroy_consumer but attached"));
KASSERT (cp->acr == 0, ("g_destroy_consumer with acr"));
KASSERT (cp->acw == 0, ("g_destroy_consumer with acw"));
KASSERT (cp->ace == 0, ("g_destroy_consumer with ace"));
g_cancel_event(cp);
gp = cp->geom;
LIST_REMOVE(cp, consumer);
devstat_remove_entry(cp->stat);
g_free(cp);
if (gp->flags & G_GEOM_WITHER)
g_do_wither();
}
static void
g_new_provider_event(void *arg, int flag)
{
struct g_class *mp;
struct g_provider *pp;
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
struct g_consumer *cp, *next_cp;
g_topology_assert();
if (flag == EV_CANCEL)
return;
if (g_shutdown)
return;
pp = arg;
G_VALID_PROVIDER(pp);
KASSERT(!(pp->flags & G_PF_WITHER),
("g_new_provider_event but withered"));
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
LIST_FOREACH_SAFE(cp, &pp->consumers, consumers, next_cp) {
if ((cp->flags & G_CF_ORPHAN) == 0 &&
cp->geom->attrchanged != NULL)
cp->geom->attrchanged(cp, "GEOM::media");
}
if (g_notaste)
return;
LIST_FOREACH(mp, &g_classes, class) {
if (mp->taste == NULL)
continue;
LIST_FOREACH(cp, &pp->consumers, consumers)
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
if (cp->geom->class == mp &&
(cp->flags & G_CF_ORPHAN) == 0)
2009-06-05 23:35:43 +00:00
break;
if (cp != NULL)
continue;
mp->taste(mp, pp, 0);
g_topology_assert();
}
}
struct g_provider *
g_new_providerf(struct g_geom *gp, const char *fmt, ...)
{
struct g_provider *pp;
struct sbuf *sb;
va_list ap;
g_topology_assert();
G_VALID_GEOM(gp);
KASSERT(gp->access != NULL,
("new provider on geom(%s) without ->access (class %s)",
gp->name, gp->class->name));
KASSERT(gp->start != NULL,
("new provider on geom(%s) without ->start (class %s)",
gp->name, gp->class->name));
KASSERT(!(gp->flags & G_GEOM_WITHER),
("new provider on WITHERing geom(%s) (class %s)",
gp->name, gp->class->name));
sb = sbuf_new_auto();
va_start(ap, fmt);
sbuf_vprintf(sb, fmt, ap);
va_end(ap);
sbuf_finish(sb);
pp = g_malloc(sizeof *pp + sbuf_len(sb) + 1, M_WAITOK | M_ZERO);
pp->name = (char *)(pp + 1);
strcpy(pp->name, sbuf_data(sb));
sbuf_delete(sb);
LIST_INIT(&pp->consumers);
pp->error = ENXIO;
pp->geom = gp;
pp->stat = devstat_new_entry(pp, -1, 0, DEVSTAT_ALL_SUPPORTED,
DEVSTAT_TYPE_DIRECT, DEVSTAT_PRIORITY_MAX);
LIST_INSERT_HEAD(&gp->provider, pp, provider);
g_post_event(g_new_provider_event, pp, M_WAITOK, pp, gp, NULL);
return (pp);
}
void
g_error_provider(struct g_provider *pp, int error)
{
/* G_VALID_PROVIDER(pp); We may not have g_topology */
pp->error = error;
}
static void
g_resize_provider_event(void *arg, int flag)
{
struct g_hh00 *hh;
struct g_class *mp;
struct g_geom *gp;
struct g_provider *pp;
struct g_consumer *cp, *cp2;
off_t size;
g_topology_assert();
if (g_shutdown)
return;
hh = arg;
pp = hh->pp;
size = hh->size;
2012-07-18 07:26:20 +00:00
g_free(hh);
G_VALID_PROVIDER(pp);
KASSERT(!(pp->flags & G_PF_WITHER),
("g_resize_provider_event but withered"));
g_trace(G_T_TOPOLOGY, "g_resize_provider_event(%p)", pp);
LIST_FOREACH_SAFE(cp, &pp->consumers, consumers, cp2) {
gp = cp->geom;
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
if (gp->resize == NULL && size < pp->mediasize) {
cp->flags |= G_CF_ORPHAN;
cp->geom->orphan(cp);
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
}
}
pp->mediasize = size;
LIST_FOREACH_SAFE(cp, &pp->consumers, consumers, cp2) {
gp = cp->geom;
if ((gp->flags & G_GEOM_WITHER) == 0 && gp->resize != NULL)
gp->resize(cp);
}
/*
* After resizing, the previously invalid GEOM class metadata
* might become valid. This means we should retaste.
*/
LIST_FOREACH(mp, &g_classes, class) {
if (mp->taste == NULL)
continue;
LIST_FOREACH(cp, &pp->consumers, consumers)
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
if (cp->geom->class == mp &&
(cp->flags & G_CF_ORPHAN) == 0)
break;
if (cp != NULL)
continue;
mp->taste(mp, pp, 0);
g_topology_assert();
}
}
void
g_resize_provider(struct g_provider *pp, off_t size)
{
struct g_hh00 *hh;
G_VALID_PROVIDER(pp);
if (pp->flags & G_PF_WITHER)
return;
if (size == pp->mediasize)
return;
hh = g_malloc(sizeof *hh, M_WAITOK | M_ZERO);
hh->pp = pp;
hh->size = size;
g_post_event(g_resize_provider_event, hh, M_WAITOK, NULL);
}
#ifndef _PATH_DEV
#define _PATH_DEV "/dev/"
#endif
struct g_provider *
g_provider_by_name(char const *arg)
{
struct g_class *cp;
struct g_geom *gp;
struct g_provider *pp, *wpp;
if (strncmp(arg, _PATH_DEV, sizeof(_PATH_DEV) - 1) == 0)
arg += sizeof(_PATH_DEV) - 1;
wpp = NULL;
LIST_FOREACH(cp, &g_classes, class) {
LIST_FOREACH(gp, &cp->geom, geom) {
LIST_FOREACH(pp, &gp->provider, provider) {
if (strcmp(arg, pp->name) != 0)
continue;
if ((gp->flags & G_GEOM_WITHER) == 0 &&
(pp->flags & G_PF_WITHER) == 0)
return (pp);
else
wpp = pp;
}
}
}
return (wpp);
}
void
g_destroy_provider(struct g_provider *pp)
{
struct g_geom *gp;
g_topology_assert();
G_VALID_PROVIDER(pp);
KASSERT(LIST_EMPTY(&pp->consumers),
("g_destroy_provider but attached"));
KASSERT (pp->acr == 0, ("g_destroy_provider with acr"));
KASSERT (pp->acw == 0, ("g_destroy_provider with acw"));
KASSERT (pp->ace == 0, ("g_destroy_provider with ace"));
g_cancel_event(pp);
LIST_REMOVE(pp, provider);
gp = pp->geom;
devstat_remove_entry(pp->stat);
Fix a bug which causes a panic in daopen(). The panic is caused by a da(4) instance going away while GEOM is still probing it. In this case, the GEOM disk class instance has been created by disk_create(), and the taste of the disk is queued in the GEOM event queue. While that event is queued, the da(4) instance goes away. When the open call comes into the da(4) driver, it dereferences the freed (but non-NULL) peripheral pointer provided by GEOM, which results in a panic. The solution is to add a callback to the GEOM disk code that is called when all of its resources are cleaned up. This is implemented inside GEOM by adding an optional callback that is called when all consumers have detached from a provider, and the provider is about to be deleted. scsi_cd.c, scsi_da.c: In the register routine for the cd(4) and da(4) routines, acquire a reference to the CAM peripheral instance just before we call disk_create(). Use the new GEOM disk d_gone() callback to register a callback (dadiskgonecb()/cddiskgonecb()) that decrements the peripheral reference count once GEOM has finished cleaning up its resources. In the cd(4) driver, clean up open and close behavior slightly. GEOM makes sure we only get one open() and one close call, so there is no need to set an open flag and decrement the reference count if we are not the first open. In the cd(4) driver, use cam_periph_release_locked() in a couple of error scenarios to avoid extra mutex calls. geom.h: Add a new, optional, providergone callback that is called when a provider is about to be deleted. geom_disk.h: Add a new d_gone() callback to the GEOM disk interface. Bump the DISK_VERSION to version 2. This probably should have been done after a couple of previous changes, especially the addition of the d_getattr() callback. geom_disk.c: Add a providergone callback for the disk class, g_disk_providergone(), that calls the user's d_gone() callback if it exists. Bump the DISK_VERSION to 2. geom_subr.c: In g_destroy_provider(), call the providergone callback if it has been provided. In g_new_geomf(), propagate the class's providergone callback to the new geom instance. blkfront.c: Callers of disk_create() are supposed to pass in DISK_VERSION, not an explicit disk API version number. Update the blkfront driver to do that. disk.9: Update the disk(9) man page to include information on the new d_gone() callback, as well as the previously added d_getattr() callback, d_descr field, and HBA PCI ID fields. MFC after: 5 days
2012-06-24 04:29:03 +00:00
/*
* If a callback was provided, send notification that the provider
* is now gone.
*/
if (gp->providergone != NULL)
gp->providergone(pp);
g_free(pp);
if ((gp->flags & G_GEOM_WITHER))
g_do_wither();
}
/*
* We keep the "geoms" list sorted by topological order (== increasing
* numerical rank) at all times.
* When an attach is done, the attaching geoms rank is invalidated
* and it is moved to the tail of the list.
* All geoms later in the sequence has their ranks reevaluated in
* sequence. If we cannot assign rank to a geom because it's
* prerequisites do not have rank, we move that element to the tail
* of the sequence with invalid rank as well.
* At some point we encounter our original geom and if we stil fail
* to assign it a rank, there must be a loop and we fail back to
* g_attach() which detach again and calls redo_rank again
* to fix up the damage.
* It would be much simpler code wise to do it recursively, but we
* can't risk that on the kernel stack.
*/
static int
redo_rank(struct g_geom *gp)
{
struct g_consumer *cp;
struct g_geom *gp1, *gp2;
int n, m;
g_topology_assert();
G_VALID_GEOM(gp);
/* Invalidate this geoms rank and move it to the tail */
gp1 = TAILQ_NEXT(gp, geoms);
if (gp1 != NULL) {
gp->rank = 0;
TAILQ_REMOVE(&geoms, gp, geoms);
TAILQ_INSERT_TAIL(&geoms, gp, geoms);
} else {
gp1 = gp;
}
/* re-rank the rest of the sequence */
for (; gp1 != NULL; gp1 = gp2) {
gp1->rank = 0;
m = 1;
LIST_FOREACH(cp, &gp1->consumer, consumer) {
if (cp->provider == NULL)
continue;
n = cp->provider->geom->rank;
if (n == 0) {
m = 0;
break;
} else if (n >= m)
m = n + 1;
}
gp1->rank = m;
gp2 = TAILQ_NEXT(gp1, geoms);
/* got a rank, moving on */
if (m != 0)
continue;
/* no rank to original geom means loop */
if (gp == gp1)
return (ELOOP);
/* no rank, put it at the end move on */
TAILQ_REMOVE(&geoms, gp1, geoms);
TAILQ_INSERT_TAIL(&geoms, gp1, geoms);
}
return (0);
}
int
g_attach(struct g_consumer *cp, struct g_provider *pp)
{
int error;
g_topology_assert();
G_VALID_CONSUMER(cp);
G_VALID_PROVIDER(pp);
2010-02-18 22:27:38 +00:00
g_trace(G_T_TOPOLOGY, "g_attach(%p, %p)", cp, pp);
KASSERT(cp->provider == NULL, ("attach but attached"));
cp->provider = pp;
LIST_INSERT_HEAD(&pp->consumers, cp, consumers);
error = redo_rank(cp->geom);
if (error) {
LIST_REMOVE(cp, consumers);
cp->provider = NULL;
redo_rank(cp->geom);
}
return (error);
}
void
g_detach(struct g_consumer *cp)
{
struct g_provider *pp;
g_topology_assert();
G_VALID_CONSUMER(cp);
g_trace(G_T_TOPOLOGY, "g_detach(%p)", cp);
KASSERT(cp->provider != NULL, ("detach but not attached"));
KASSERT(cp->acr == 0, ("detach but nonzero acr"));
KASSERT(cp->acw == 0, ("detach but nonzero acw"));
KASSERT(cp->ace == 0, ("detach but nonzero ace"));
KASSERT(cp->nstart == cp->nend,
("detach with active requests"));
pp = cp->provider;
LIST_REMOVE(cp, consumers);
cp->provider = NULL;
if ((cp->geom->flags & G_GEOM_WITHER) ||
(pp->geom->flags & G_GEOM_WITHER) ||
(pp->flags & G_PF_WITHER))
g_do_wither();
redo_rank(cp->geom);
}
/*
* g_access()
*
* Access-check with delta values. The question asked is "can provider
* "cp" change the access counters by the relative amounts dc[rwe] ?"
*/
int
g_access(struct g_consumer *cp, int dcr, int dcw, int dce)
{
struct g_provider *pp;
int pr,pw,pe;
int error;
g_topology_assert();
G_VALID_CONSUMER(cp);
pp = cp->provider;
KASSERT(pp != NULL, ("access but not attached"));
G_VALID_PROVIDER(pp);
g_trace(G_T_ACCESS, "g_access(%p(%s), %d, %d, %d)",
cp, pp->name, dcr, dcw, dce);
KASSERT(cp->acr + dcr >= 0, ("access resulting in negative acr"));
KASSERT(cp->acw + dcw >= 0, ("access resulting in negative acw"));
KASSERT(cp->ace + dce >= 0, ("access resulting in negative ace"));
KASSERT(dcr != 0 || dcw != 0 || dce != 0, ("NOP access request"));
KASSERT(pp->geom->access != NULL, ("NULL geom->access"));
/*
* If our class cares about being spoiled, and we have been, we
* are probably just ahead of the event telling us that. Fail
* now rather than having to unravel this later.
*/
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
if (cp->geom->spoiled != NULL && (cp->flags & G_CF_SPOILED) &&
(dcr > 0 || dcw > 0 || dce > 0))
return (ENXIO);
/*
* Figure out what counts the provider would have had, if this
* consumer had (r0w0e0) at this time.
*/
pr = pp->acr - cp->acr;
pw = pp->acw - cp->acw;
pe = pp->ace - cp->ace;
g_trace(G_T_ACCESS,
"open delta:[r%dw%de%d] old:[r%dw%de%d] provider:[r%dw%de%d] %p(%s)",
dcr, dcw, dce,
cp->acr, cp->acw, cp->ace,
pp->acr, pp->acw, pp->ace,
pp, pp->name);
/* If foot-shooting is enabled, any open on rank#1 is OK */
if ((g_debugflags & 16) && pp->geom->rank == 1)
;
/* If we try exclusive but already write: fail */
else if (dce > 0 && pw > 0)
return (EPERM);
/* If we try write but already exclusive: fail */
else if (dcw > 0 && pe > 0)
return (EPERM);
/* If we try to open more but provider is error'ed: fail */
else if ((dcr > 0 || dcw > 0 || dce > 0) && pp->error != 0) {
printf("%s(%d): provider %s has error\n",
__func__, __LINE__, pp->name);
return (pp->error);
}
/* Ok then... */
error = pp->geom->access(pp, dcr, dcw, dce);
KASSERT(dcr > 0 || dcw > 0 || dce > 0 || error == 0,
("Geom provider %s::%s dcr=%d dcw=%d dce=%d error=%d failed "
"closing ->access()", pp->geom->class->name, pp->name, dcr, dcw,
dce, error));
if (!error) {
/*
* If we open first write, spoil any partner consumers.
* If we close last write and provider is not errored,
* trigger re-taste.
*/
if (pp->acw == 0 && dcw != 0)
g_spoil(pp, cp);
else if (pp->acw != 0 && pp->acw == -dcw && pp->error == 0 &&
!(pp->geom->flags & G_GEOM_WITHER))
g_post_event(g_new_provider_event, pp, M_WAITOK,
pp, NULL);
pp->acr += dcr;
pp->acw += dcw;
pp->ace += dce;
cp->acr += dcr;
cp->acw += dcw;
cp->ace += dce;
if (pp->acr != 0 || pp->acw != 0 || pp->ace != 0)
KASSERT(pp->sectorsize > 0,
("Provider %s lacks sectorsize", pp->name));
if ((cp->geom->flags & G_GEOM_WITHER) &&
cp->acr == 0 && cp->acw == 0 && cp->ace == 0)
g_do_wither();
}
return (error);
}
int
g_handleattr_int(struct bio *bp, const char *attribute, int val)
{
return (g_handleattr(bp, attribute, &val, sizeof val));
}
Improve ZFS N-way mirror read performance by using load and locality information. The existing algorithm selects a preferred leaf vdev based on offset of the zio request modulo the number of members in the mirror. It assumes the devices are of equal performance and that spreading the requests randomly over both drives will be sufficient to saturate them. In practice this results in the leaf vdevs being under utilized. The new algorithm takes into the following additional factors: * Load of the vdevs (number outstanding I/O requests) * The locality of last queued I/O vs the new I/O request. Within the locality calculation additional knowledge about the underlying vdev is considered such as; is the device backing the vdev a rotating media device. This results in performance increases across the board as well as significant increases for predominantly streaming loads and for configurations which don't have evenly performing devices. The following are results from a setup with 3 Way Mirror with 2 x HD's and 1 x SSD from a basic test running multiple parrallel dd's. With pre-fetch disabled (vfs.zfs.prefetch_disable=1): == Stripe Balanced (default) == Read 15360MB using bs: 1048576, readers: 3, took 161 seconds @ 95 MB/s == Load Balanced (zfslinux) == Read 15360MB using bs: 1048576, readers: 3, took 297 seconds @ 51 MB/s == Load Balanced (locality freebsd) == Read 15360MB using bs: 1048576, readers: 3, took 54 seconds @ 284 MB/s With pre-fetch enabled (vfs.zfs.prefetch_disable=0): == Stripe Balanced (default) == Read 15360MB using bs: 1048576, readers: 3, took 91 seconds @ 168 MB/s == Load Balanced (zfslinux) == Read 15360MB using bs: 1048576, readers: 3, took 108 seconds @ 142 MB/s == Load Balanced (locality freebsd) == Read 15360MB using bs: 1048576, readers: 3, took 48 seconds @ 320 MB/s In addition to the performance changes the code was also restructured, with the help of Justin Gibbs, to provide a more logical flow which also ensures vdevs loads are only calculated from the set of valid candidates. The following additional sysctls where added to allow the administrator to tune the behaviour of the load algorithm: * vfs.zfs.vdev.mirror.rotating_inc * vfs.zfs.vdev.mirror.rotating_seek_inc * vfs.zfs.vdev.mirror.rotating_seek_offset * vfs.zfs.vdev.mirror.non_rotating_inc * vfs.zfs.vdev.mirror.non_rotating_seek_inc These changes where based on work started by the zfsonlinux developers: https://github.com/zfsonlinux/zfs/pull/1487 Reviewed by: gibbs, mav, will MFC after: 2 weeks Sponsored by: Multiplay
2013-10-23 09:54:58 +00:00
int
g_handleattr_uint16_t(struct bio *bp, const char *attribute, uint16_t val)
{
return (g_handleattr(bp, attribute, &val, sizeof val));
}
int
g_handleattr_off_t(struct bio *bp, const char *attribute, off_t val)
{
return (g_handleattr(bp, attribute, &val, sizeof val));
}
int
g_handleattr_str(struct bio *bp, const char *attribute, const char *str)
{
return (g_handleattr(bp, attribute, str, 0));
}
int
g_handleattr(struct bio *bp, const char *attribute, const void *val, int len)
{
int error = 0;
if (strcmp(bp->bio_attribute, attribute))
return (0);
if (len == 0) {
bzero(bp->bio_data, bp->bio_length);
if (strlcpy(bp->bio_data, val, bp->bio_length) >=
bp->bio_length) {
printf("%s: %s bio_length %jd len %zu -> EFAULT\n",
__func__, bp->bio_to->name,
(intmax_t)bp->bio_length, strlen(val));
error = EFAULT;
}
} else if (bp->bio_length == len) {
bcopy(val, bp->bio_data, len);
} else {
printf("%s: %s bio_length %jd len %d -> EFAULT\n", __func__,
bp->bio_to->name, (intmax_t)bp->bio_length, len);
error = EFAULT;
}
if (error == 0)
bp->bio_completed = bp->bio_length;
g_io_deliver(bp, error);
return (1);
}
int
g_std_access(struct g_provider *pp,
int dr __unused, int dw __unused, int de __unused)
{
g_topology_assert();
G_VALID_PROVIDER(pp);
return (0);
}
void
g_std_done(struct bio *bp)
{
struct bio *bp2;
bp2 = bp->bio_parent;
if (bp2->bio_error == 0)
bp2->bio_error = bp->bio_error;
bp2->bio_completed += bp->bio_completed;
g_destroy_bio(bp);
bp2->bio_inbed++;
if (bp2->bio_children == bp2->bio_inbed)
g_io_deliver(bp2, bp2->bio_error);
}
/* XXX: maybe this is only g_slice_spoiled */
void
g_std_spoiled(struct g_consumer *cp)
{
struct g_geom *gp;
struct g_provider *pp;
g_topology_assert();
G_VALID_CONSUMER(cp);
g_trace(G_T_TOPOLOGY, "g_std_spoiled(%p)", cp);
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
cp->flags |= G_CF_ORPHAN;
g_detach(cp);
gp = cp->geom;
LIST_FOREACH(pp, &gp->provider, provider)
g_orphan_provider(pp, ENXIO);
g_destroy_consumer(cp);
if (LIST_EMPTY(&gp->provider) && LIST_EMPTY(&gp->consumer))
g_destroy_geom(gp);
else
gp->flags |= G_GEOM_WITHER;
}
/*
* Spoiling happens when a provider is opened for writing, but consumers
* which are configured by in-band data are attached (slicers for instance).
* Since the write might potentially change the in-band data, such consumers
* need to re-evaluate their existence after the writing session closes.
* We do this by (offering to) tear them down when the open for write happens
* in return for a re-taste when it closes again.
* Together with the fact that such consumers grab an 'e' bit whenever they
* are open, regardless of mode, this ends up DTRT.
*/
static void
g_spoil_event(void *arg, int flag)
{
struct g_provider *pp;
struct g_consumer *cp, *cp2;
g_topology_assert();
if (flag == EV_CANCEL)
return;
pp = arg;
G_VALID_PROVIDER(pp);
g_trace(G_T_TOPOLOGY, "%s %p(%s:%s:%s)", __func__, pp,
pp->geom->class->name, pp->geom->name, pp->name);
for (cp = LIST_FIRST(&pp->consumers); cp != NULL; cp = cp2) {
cp2 = LIST_NEXT(cp, consumers);
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
if ((cp->flags & G_CF_SPOILED) == 0)
continue;
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
cp->flags &= ~G_CF_SPOILED;
if (cp->geom->spoiled == NULL)
continue;
cp->geom->spoiled(cp);
g_topology_assert();
}
}
void
g_spoil(struct g_provider *pp, struct g_consumer *cp)
{
struct g_consumer *cp2;
g_topology_assert();
G_VALID_PROVIDER(pp);
G_VALID_CONSUMER(cp);
LIST_FOREACH(cp2, &pp->consumers, consumers) {
if (cp2 == cp)
continue;
/*
KASSERT(cp2->acr == 0, ("spoiling cp->acr = %d", cp2->acr));
KASSERT(cp2->acw == 0, ("spoiling cp->acw = %d", cp2->acw));
*/
KASSERT(cp2->ace == 0, ("spoiling cp->ace = %d", cp2->ace));
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
cp2->flags |= G_CF_SPOILED;
}
g_post_event(g_spoil_event, pp, M_WAITOK, pp, NULL);
}
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
static void
g_media_changed_event(void *arg, int flag)
{
struct g_provider *pp;
int retaste;
g_topology_assert();
if (flag == EV_CANCEL)
return;
pp = arg;
G_VALID_PROVIDER(pp);
/*
* If provider was not open for writing, queue retaste after spoiling.
* If it was, retaste will happen automatically on close.
*/
retaste = (pp->acw == 0 && pp->error == 0 &&
!(pp->geom->flags & G_GEOM_WITHER));
g_spoil_event(arg, flag);
if (retaste)
g_post_event(g_new_provider_event, pp, M_WAITOK, pp, NULL);
}
int
g_media_changed(struct g_provider *pp, int flag)
{
struct g_consumer *cp;
LIST_FOREACH(cp, &pp->consumers, consumers)
cp->flags |= G_CF_SPOILED;
return (g_post_event(g_media_changed_event, pp, flag, pp, NULL));
}
int
g_media_gone(struct g_provider *pp, int flag)
{
struct g_consumer *cp;
LIST_FOREACH(cp, &pp->consumers, consumers)
cp->flags |= G_CF_SPOILED;
return (g_post_event(g_spoil_event, pp, flag, pp, NULL));
}
int
g_getattr__(const char *attr, struct g_consumer *cp, void *var, int len)
{
int error, i;
i = len;
error = g_io_getattr(attr, cp, &i, var);
if (error)
return (error);
if (i != len)
return (EINVAL);
return (0);
}
static int
g_get_device_prefix_len(const char *name)
{
int len;
if (strncmp(name, "ada", 3) == 0)
len = 3;
else if (strncmp(name, "ad", 2) == 0)
len = 2;
else
return (0);
if (name[len] < '0' || name[len] > '9')
return (0);
do {
len++;
} while (name[len] >= '0' && name[len] <= '9');
return (len);
}
int
g_compare_names(const char *namea, const char *nameb)
{
int deva, devb;
if (strcmp(namea, nameb) == 0)
return (1);
deva = g_get_device_prefix_len(namea);
if (deva == 0)
return (0);
devb = g_get_device_prefix_len(nameb);
if (devb == 0)
return (0);
if (strcmp(namea + deva, nameb + devb) == 0)
return (1);
return (0);
}
#if defined(DIAGNOSTIC) || defined(DDB)
/*
* This function walks the mesh and returns a non-zero integer if it
* finds the argument pointer is an object. The return value indicates
* which type of object it is believed to be. If topology is not locked,
* this function is potentially dangerous, but we don't assert that the
* topology lock is held when called from debugger.
*/
int
g_valid_obj(void const *ptr)
{
struct g_class *mp;
struct g_geom *gp;
struct g_consumer *cp;
struct g_provider *pp;
#ifdef KDB
if (kdb_active == 0)
#endif
g_topology_assert();
LIST_FOREACH(mp, &g_classes, class) {
if (ptr == mp)
return (1);
LIST_FOREACH(gp, &mp->geom, geom) {
if (ptr == gp)
return (2);
LIST_FOREACH(cp, &gp->consumer, consumer)
if (ptr == cp)
return (3);
LIST_FOREACH(pp, &gp->provider, provider)
if (ptr == pp)
return (4);
}
}
return(0);
}
#endif
#ifdef DDB
#define gprintf(...) do { \
db_printf("%*s", indent, ""); \
db_printf(__VA_ARGS__); \
} while (0)
#define gprintln(...) do { \
gprintf(__VA_ARGS__); \
db_printf("\n"); \
} while (0)
#define ADDFLAG(obj, flag, sflag) do { \
if ((obj)->flags & (flag)) { \
if (comma) \
strlcat(str, ",", size); \
strlcat(str, (sflag), size); \
comma = 1; \
} \
} while (0)
static char *
provider_flags_to_string(struct g_provider *pp, char *str, size_t size)
{
int comma = 0;
bzero(str, size);
if (pp->flags == 0) {
strlcpy(str, "NONE", size);
return (str);
}
ADDFLAG(pp, G_PF_WITHER, "G_PF_WITHER");
ADDFLAG(pp, G_PF_ORPHAN, "G_PF_ORPHAN");
return (str);
}
static char *
geom_flags_to_string(struct g_geom *gp, char *str, size_t size)
{
int comma = 0;
bzero(str, size);
if (gp->flags == 0) {
strlcpy(str, "NONE", size);
return (str);
}
ADDFLAG(gp, G_GEOM_WITHER, "G_GEOM_WITHER");
return (str);
}
static void
db_show_geom_consumer(int indent, struct g_consumer *cp)
{
if (indent == 0) {
gprintln("consumer: %p", cp);
gprintln(" class: %s (%p)", cp->geom->class->name,
cp->geom->class);
gprintln(" geom: %s (%p)", cp->geom->name, cp->geom);
if (cp->provider == NULL)
gprintln(" provider: none");
else {
gprintln(" provider: %s (%p)", cp->provider->name,
cp->provider);
}
gprintln(" access: r%dw%de%d", cp->acr, cp->acw, cp->ace);
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
gprintln(" flags: 0x%04x", cp->flags);
gprintln(" nstart: %u", cp->nstart);
gprintln(" nend: %u", cp->nend);
} else {
gprintf("consumer: %p (%s), access=r%dw%de%d", cp,
cp->provider != NULL ? cp->provider->name : "none",
cp->acr, cp->acw, cp->ace);
Implement media change notification for DA and CD removable media devices. It includes three parts: 1) Modifications to CAM to detect media media changes and report them to disk(9) layer. For modern SATA (and potentially UAS) devices it utilizes Asynchronous Notification mechanism to receive events from hardware. Active polling with TEST UNIT READY commands with 3 seconds period is used for incapable hardware. After that both CD and DA drivers work the same way, detecting two conditions: "NOT READY: Medium not present" after medium was detected previously, and "UNIT ATTENTION: Not ready to ready change, medium may have changed". First one reported to disk(9) as media removal, second as media insert/change. To reliably receive second event new AC_UNIT_ATTENTION async added to make UAs broadcasted to all periphs by generic error handling code in cam_periph_error(). 2) Modifications to GEOM core to handle media remove and change events. Media removal handled by spoiling all consumers attached to the provider. Media change event also schedules provider retaste after spoiling to probe new media. New flag G_CF_ORPHAN was added to consumers to reflect that consumer is in process of destruction. It allows retaste to create new geom instance of the same class, while previous one is still dying. 3) Modifications to some GEOM classes: DEV -- to report media change events to devd; VFS -- to handle spoiling same as orphan to prevent accessing replaced media. PART class already handles spoiling alike to orphan. Reviewed by: silence on geom@ and scsi@ Tested by: avg Sponsored by: iXsystems, Inc. / PC-BSD MFC after: 2 months
2012-07-29 11:51:48 +00:00
if (cp->flags)
db_printf(", flags=0x%04x", cp->flags);
db_printf("\n");
}
}
static void
db_show_geom_provider(int indent, struct g_provider *pp)
{
struct g_consumer *cp;
char flags[64];
if (indent == 0) {
gprintln("provider: %s (%p)", pp->name, pp);
gprintln(" class: %s (%p)", pp->geom->class->name,
pp->geom->class);
gprintln(" geom: %s (%p)", pp->geom->name, pp->geom);
gprintln(" mediasize: %jd", (intmax_t)pp->mediasize);
gprintln(" sectorsize: %u", pp->sectorsize);
gprintln(" stripesize: %u", pp->stripesize);
gprintln(" stripeoffset: %u", pp->stripeoffset);
gprintln(" access: r%dw%de%d", pp->acr, pp->acw,
pp->ace);
gprintln(" flags: %s (0x%04x)",
provider_flags_to_string(pp, flags, sizeof(flags)),
pp->flags);
gprintln(" error: %d", pp->error);
gprintln(" nstart: %u", pp->nstart);
gprintln(" nend: %u", pp->nend);
if (LIST_EMPTY(&pp->consumers))
gprintln(" consumers: none");
} else {
gprintf("provider: %s (%p), access=r%dw%de%d",
pp->name, pp, pp->acr, pp->acw, pp->ace);
if (pp->flags != 0) {
db_printf(", flags=%s (0x%04x)",
provider_flags_to_string(pp, flags, sizeof(flags)),
pp->flags);
}
db_printf("\n");
}
if (!LIST_EMPTY(&pp->consumers)) {
LIST_FOREACH(cp, &pp->consumers, consumers) {
db_show_geom_consumer(indent + 2, cp);
if (db_pager_quit)
break;
}
}
}
static void
db_show_geom_geom(int indent, struct g_geom *gp)
{
struct g_provider *pp;
struct g_consumer *cp;
char flags[64];
if (indent == 0) {
gprintln("geom: %s (%p)", gp->name, gp);
gprintln(" class: %s (%p)", gp->class->name, gp->class);
gprintln(" flags: %s (0x%04x)",
geom_flags_to_string(gp, flags, sizeof(flags)), gp->flags);
gprintln(" rank: %d", gp->rank);
if (LIST_EMPTY(&gp->provider))
gprintln(" providers: none");
if (LIST_EMPTY(&gp->consumer))
gprintln(" consumers: none");
} else {
gprintf("geom: %s (%p), rank=%d", gp->name, gp, gp->rank);
if (gp->flags != 0) {
db_printf(", flags=%s (0x%04x)",
geom_flags_to_string(gp, flags, sizeof(flags)),
gp->flags);
}
db_printf("\n");
}
if (!LIST_EMPTY(&gp->provider)) {
LIST_FOREACH(pp, &gp->provider, provider) {
db_show_geom_provider(indent + 2, pp);
if (db_pager_quit)
break;
}
}
if (!LIST_EMPTY(&gp->consumer)) {
LIST_FOREACH(cp, &gp->consumer, consumer) {
db_show_geom_consumer(indent + 2, cp);
if (db_pager_quit)
break;
}
}
}
static void
db_show_geom_class(struct g_class *mp)
{
struct g_geom *gp;
db_printf("class: %s (%p)\n", mp->name, mp);
LIST_FOREACH(gp, &mp->geom, geom) {
db_show_geom_geom(2, gp);
if (db_pager_quit)
break;
}
}
/*
* Print the GEOM topology or the given object.
*/
DB_SHOW_COMMAND(geom, db_show_geom)
{
struct g_class *mp;
if (!have_addr) {
/* No address given, print the entire topology. */
LIST_FOREACH(mp, &g_classes, class) {
db_show_geom_class(mp);
db_printf("\n");
if (db_pager_quit)
break;
}
} else {
switch (g_valid_obj((void *)addr)) {
case 1:
db_show_geom_class((struct g_class *)addr);
break;
case 2:
db_show_geom_geom(0, (struct g_geom *)addr);
break;
case 3:
db_show_geom_consumer(0, (struct g_consumer *)addr);
break;
case 4:
db_show_geom_provider(0, (struct g_provider *)addr);
break;
default:
db_printf("Not a GEOM object.\n");
break;
}
}
}
static void
db_print_bio_cmd(struct bio *bp)
{
db_printf(" cmd: ");
switch (bp->bio_cmd) {
case BIO_READ: db_printf("BIO_READ"); break;
case BIO_WRITE: db_printf("BIO_WRITE"); break;
case BIO_DELETE: db_printf("BIO_DELETE"); break;
case BIO_GETATTR: db_printf("BIO_GETATTR"); break;
case BIO_FLUSH: db_printf("BIO_FLUSH"); break;
case BIO_CMD0: db_printf("BIO_CMD0"); break;
case BIO_CMD1: db_printf("BIO_CMD1"); break;
case BIO_CMD2: db_printf("BIO_CMD2"); break;
Add support for managing Shingled Magnetic Recording (SMR) drives. This change includes support for SCSI SMR drives (which conform to the Zoned Block Commands or ZBC spec) and ATA SMR drives (which conform to the Zoned ATA Command Set or ZAC spec) behind SAS expanders. This includes full management support through the GEOM BIO interface, and through a new userland utility, zonectl(8), and through camcontrol(8). This is now ready for filesystems to use to detect and manage zoned drives. (There is no work in progress that I know of to use this for ZFS or UFS, if anyone is interested, let me know and I may have some suggestions.) Also, improve ATA command passthrough and dispatch support, both via ATA and ATA passthrough over SCSI. Also, add support to camcontrol(8) for the ATA Extended Power Conditions feature set. You can now manage ATA device power states, and set various idle time thresholds for a drive to enter lower power states. Note that this change cannot be MFCed in full, because it depends on changes to the struct bio API that break compatilibity. In order to avoid breaking the stable API, only changes that don't touch or depend on the struct bio changes can be merged. For example, the camcontrol(8) changes don't depend on the new bio API, but zonectl(8) and the probe changes to the da(4) and ada(4) drivers do depend on it. Also note that the SMR changes have not yet been tested with an actual SCSI ZBC device, or a SCSI to ATA translation layer (SAT) that supports ZBC to ZAC translation. I have not yet gotten a suitable drive or SAT layer, so any testing help would be appreciated. These changes have been tested with Seagate Host Aware SATA drives attached to both SAS and SATA controllers. Also, I do not have any SATA Host Managed devices, and I suspect that it may take additional (hopefully minor) changes to support them. Thanks to Seagate for supplying the test hardware and answering questions. sbin/camcontrol/Makefile: Add epc.c and zone.c. sbin/camcontrol/camcontrol.8: Document the zone and epc subcommands. sbin/camcontrol/camcontrol.c: Add the zone and epc subcommands. Add auxiliary register support to build_ata_cmd(). Make sure to set the CAM_ATAIO_NEEDRESULT, CAM_ATAIO_DMA, and CAM_ATAIO_FPDMA flags as appropriate for ATA commands. Add a new get_ata_status() function to parse ATA result from SCSI sense descriptors (for ATA passthrough over SCSI) and ATA I/O requests. sbin/camcontrol/camcontrol.h: Update the build_ata_cmd() prototype Add get_ata_status(), zone(), and epc(). sbin/camcontrol/epc.c: Support for ATA Extended Power Conditions features. This includes support for all features documented in the ACS-4 Revision 12 specification from t13.org (dated February 18, 2016). The EPC feature set allows putting a drive into a power power mode immediately, or setting timeouts so that the drive will automatically enter progressively lower power states after various idle times. sbin/camcontrol/fwdownload.c: Update the firmware download code for the new build_ata_cmd() arguments. sbin/camcontrol/zone.c: Implement support for Shingled Magnetic Recording (SMR) drives via SCSI Zoned Block Commands (ZBC) and ATA Zoned Device ATA Command Set (ZAC). These specs were developed in concert, and are functionally identical. The primary differences are due to SCSI and ATA differences. (SCSI is big endian, ATA is little endian, for example.) This includes support for all commands defined in the ZBC and ZAC specs. sys/cam/ata/ata_all.c: Decode a number of additional ATA command names in ata_op_string(). Add a new CCB building function, ata_read_log(). Add ata_zac_mgmt_in() and ata_zac_mgmt_out() CCB building functions. These support both DMA and NCQ encapsulation. sys/cam/ata/ata_all.h: Add prototypes for ata_read_log(), ata_zac_mgmt_out(), and ata_zac_mgmt_in(). sys/cam/ata/ata_da.c: Revamp the ada(4) driver to support zoned devices. Add four new probe states to gather information needed for zone support. Add a new adasetflags() function to avoid duplication of large blocks of flag setting between the async handler and register functions. Add new sysctl variables that describe zone support and paramters. Add support for the new BIO_ZONE bio, and all of its subcommands: DISK_ZONE_OPEN, DISK_ZONE_CLOSE, DISK_ZONE_FINISH, DISK_ZONE_RWP, DISK_ZONE_REPORT_ZONES, and DISK_ZONE_GET_PARAMS. sys/cam/scsi/scsi_all.c: Add command descriptions for the ZBC IN/OUT commands. Add descriptions for ZBC Host Managed devices. Add a new function, scsi_ata_pass() to do ATA passthrough over SCSI. This will eventually replace scsi_ata_pass_16() -- it can create the 12, 16, and 32-byte variants of the ATA PASS-THROUGH command, and supports setting all of the registers defined as of SAT-4, Revision 5 (March 11, 2016). Change scsi_ata_identify() to use scsi_ata_pass() instead of scsi_ata_pass_16(). Add a new scsi_ata_read_log() function to facilitate reading ATA logs via SCSI. sys/cam/scsi/scsi_all.h: Add the new ATA PASS-THROUGH(32) command CDB. Add extended and variable CDB opcodes. Add Zoned Block Device Characteristics VPD page. Add ATA Return SCSI sense descriptor. Add prototypes for scsi_ata_read_log() and scsi_ata_pass(). sys/cam/scsi/scsi_da.c: Revamp the da(4) driver to support zoned devices. Add five new probe states, four of which are needed for ATA devices. Add five new sysctl variables that describe zone support and parameters. The da(4) driver supports SCSI ZBC devices, as well as ATA ZAC devices when they are attached via a SCSI to ATA Translation (SAT) layer. Since ZBC -> ZAC translation is a new feature in the T10 SAT-4 spec, most SATA drives will be supported via ATA commands sent via the SCSI ATA PASS-THROUGH command. The da(4) driver will prefer the ZBC interface, if it is available, for performance reasons, but will use the ATA PASS-THROUGH interface to the ZAC command set if the SAT layer doesn't support translation yet. As I mentioned above, ZBC command support is untested. Add support for the new BIO_ZONE bio, and all of its subcommands: DISK_ZONE_OPEN, DISK_ZONE_CLOSE, DISK_ZONE_FINISH, DISK_ZONE_RWP, DISK_ZONE_REPORT_ZONES, and DISK_ZONE_GET_PARAMS. Add scsi_zbc_in() and scsi_zbc_out() CCB building functions. Add scsi_ata_zac_mgmt_out() and scsi_ata_zac_mgmt_in() CCB/CDB building functions. Note that these have return values, unlike almost all other CCB building functions in CAM. The reason is that they can fail, depending upon the particular combination of input parameters. The primary failure case is if the user wants NCQ, but fails to specify additional CDB storage. NCQ requires using the 32-byte version of the SCSI ATA PASS-THROUGH command, and the current CAM CDB size is 16 bytes. sys/cam/scsi/scsi_da.h: Add ZBC IN and ZBC OUT CDBs and opcodes. Add SCSI Report Zones data structures. Add scsi_zbc_in(), scsi_zbc_out(), scsi_ata_zac_mgmt_out(), and scsi_ata_zac_mgmt_in() prototypes. sys/dev/ahci/ahci.c: Fix SEND / RECEIVE FPDMA QUEUED in the ahci(4) driver. ahci_setup_fis() previously set the top bits of the sector count register in the FIS to 0 for FPDMA commands. This is okay for read and write, because the PRIO field is in the only thing in those bits, and we don't implement that further up the stack. But, for SEND and RECEIVE FPDMA QUEUED, the subcommand is in that byte, so it needs to be transmitted to the drive. In ahci_setup_fis(), always set the the top 8 bits of the sector count register. We need it in both the standard and NCQ / FPDMA cases. sys/geom/eli/g_eli.c: Pass BIO_ZONE commands through the GELI class. sys/geom/geom.h: Add g_io_zonecmd() prototype. sys/geom/geom_dev.c: Add new DIOCZONECMD ioctl, which allows sending zone commands to disks. sys/geom/geom_disk.c: Add support for BIO_ZONE commands. sys/geom/geom_disk.h: Add a new flag, DISKFLAG_CANZONE, that indicates that a given GEOM disk client can handle BIO_ZONE commands. sys/geom/geom_io.c: Add a new function, g_io_zonecmd(), that handles execution of BIO_ZONE commands. Add permissions check for BIO_ZONE commands. Add command decoding for BIO_ZONE commands. sys/geom/geom_subr.c: Add DDB command decoding for BIO_ZONE commands. sys/kern/subr_devstat.c: Record statistics for REPORT ZONES commands. Note that the number of bytes transferred for REPORT ZONES won't quite match what is received from the harware. This is because we're necessarily counting bytes coming from the da(4) / ada(4) drivers, which are using the disk_zone.h interface to communicate up the stack. The structure sizes it uses are slightly different than the SCSI and ATA structure sizes. sys/sys/ata.h: Add many bit and structure definitions for ZAC, NCQ, and EPC command support. sys/sys/bio.h: Convert the bio_cmd field to a straight enumeration. This will yield more space for additional commands in the future. After change r297955 and other related changes, this is now possible. Converting to an enumeration will also prevent use as a bitmask in the future. sys/sys/disk.h: Define the DIOCZONECMD ioctl. sys/sys/disk_zone.h: Add a new API for managing zoned disks. This is very close to the SCSI ZBC and ATA ZAC standards, but uses integers in native byte order instead of big endian (SCSI) or little endian (ATA) byte arrays. This is intended to offer to the complete feature set of the ZBC and ZAC disk management without requiring the application developer to include SCSI or ATA headers. We also use one set of headers for ioctl consumers and kernel bio-level consumers. sys/sys/param.h: Bump __FreeBSD_version for sys/bio.h command changes, and inclusion of SMR support. usr.sbin/Makefile: Add the zonectl utility. usr.sbin/diskinfo/diskinfo.c Add disk zoning capability to the 'diskinfo -v' output. usr.sbin/zonectl/Makefile: Add zonectl makefile. usr.sbin/zonectl/zonectl.8 zonectl(8) man page. usr.sbin/zonectl/zonectl.c The zonectl(8) utility. This allows managing SCSI or ATA zoned disks via the disk_zone.h API. You can report zones, reset write pointers, get parameters, etc. Sponsored by: Spectra Logic Differential Revision: https://reviews.freebsd.org/D6147 Reviewed by: wblock (documentation)
2016-05-19 14:08:36 +00:00
case BIO_ZONE: db_printf("BIO_ZONE"); break;
default: db_printf("UNKNOWN"); break;
}
db_printf("\n");
}
static void
db_print_bio_flags(struct bio *bp)
{
int comma;
comma = 0;
db_printf(" flags: ");
if (bp->bio_flags & BIO_ERROR) {
db_printf("BIO_ERROR");
comma = 1;
}
if (bp->bio_flags & BIO_DONE) {
db_printf("%sBIO_DONE", (comma ? ", " : ""));
comma = 1;
}
if (bp->bio_flags & BIO_ONQUEUE)
db_printf("%sBIO_ONQUEUE", (comma ? ", " : ""));
db_printf("\n");
}
/*
* Print useful information in a BIO
*/
DB_SHOW_COMMAND(bio, db_show_bio)
{
struct bio *bp;
if (have_addr) {
bp = (struct bio *)addr;
db_printf("BIO %p\n", bp);
db_print_bio_cmd(bp);
db_print_bio_flags(bp);
db_printf(" cflags: 0x%hx\n", bp->bio_cflags);
db_printf(" pflags: 0x%hx\n", bp->bio_pflags);
db_printf(" offset: %jd\n", (intmax_t)bp->bio_offset);
db_printf(" length: %jd\n", (intmax_t)bp->bio_length);
db_printf(" bcount: %ld\n", bp->bio_bcount);
db_printf(" resid: %ld\n", bp->bio_resid);
db_printf(" completed: %jd\n", (intmax_t)bp->bio_completed);
db_printf(" children: %u\n", bp->bio_children);
db_printf(" inbed: %u\n", bp->bio_inbed);
db_printf(" error: %d\n", bp->bio_error);
db_printf(" parent: %p\n", bp->bio_parent);
db_printf(" driver1: %p\n", bp->bio_driver1);
db_printf(" driver2: %p\n", bp->bio_driver2);
db_printf(" caller1: %p\n", bp->bio_caller1);
db_printf(" caller2: %p\n", bp->bio_caller2);
db_printf(" bio_from: %p\n", bp->bio_from);
db_printf(" bio_to: %p\n", bp->bio_to);
#if defined(BUF_TRACKING) || defined(FULL_BUF_TRACKING)
db_printf(" bio_track_bp: %p\n", bp->bio_track_bp);
#endif
}
}
#undef gprintf
#undef gprintln
#undef ADDFLAG
#endif /* DDB */