Update to reflect major changes in vinum kernel module

This commit is contained in:
Greg Lehey 1999-01-21 00:45:11 +00:00
parent 0834a1f2db
commit 52e80fad69
Notes: svn2git 2020-12-20 02:59:44 +00:00
svn path=/head/; revision=42947
4 changed files with 431 additions and 131 deletions

View File

@ -35,7 +35,7 @@
* otherwise) arising in any way out of the use of this software, even if
* advised of the possibility of such damage.
*
* $Id: list.c,v 1.6 1998/12/28 16:32:39 peter Exp $
* $Id: list.c,v 1.17 1999/01/17 02:58:44 grog Exp grog $
*/
#include <ctype.h>
@ -350,56 +350,6 @@ vinum_lpi(int plexno, int recurse)
get_volume_info(&vol, plex.volno);
printf("\t\tPart of volume %s\n", vol.name);
}
if (plex.state == plex_reviving) {
printf("\t\tRevive pointer:\t\t%s (%d%%)\n",
roughlength(plex.revived << DEV_BSHIFT, 0),
(int) (((u_int64_t) (plex.revived * 100)) / plex.length));
printf("\t\tRevive blocksize:\t%s\n"
"\t\tRevive interval:\t%10d seconds\n",
roughlength(plex.revive_blocksize, 0),
plex.revive_interval);
}
if (Verbose) { /* show the unmapped and defective parts */
int re; /* freelist entry */
struct plexregion region;
struct rerq { /* request to pass to ioctl */
int plexno; /* plex for the request */
int re; /* region */
} *rerq = (struct rerq *) &region;
if (plex.unmapped_regions) {
printf("\t\tPlex contains %d unmapped regions:\n\t\t Offset\t Size\n",
plex.unmapped_regions);
for (re = 0; re < plex.unmapped_regions; re++) {
rerq->plexno = plex.plexno;
rerq->re = re;
if (ioctl(superdev, VINUM_GETUNMAPPED, &region) < 0) {
fprintf(stderr,
"Can't get unmapped region %d: %s\n",
re,
strerror(errno));
longjmp(command_fail, -1);
}
printf("\t\t%9qd\t%9qd\n", region.offset, region.length);
}
}
if (plex.defective_regions) {
printf("\t\tPlex contains %d defective regions:\n\t\t Offset\t Size\n",
plex.defective_regions);
for (re = 0; re < plex.defective_regions; re++) {
rerq->plexno = plex.plexno;
rerq->re = re;
if (ioctl(superdev, VINUM_GETDEFECTIVE, &region) < 0) {
fprintf(stderr,
"Can't get defective region %d: %s\n",
re,
strerror(errno));
longjmp(command_fail, -1);
}
printf("\t\t%9qd\t%9qd\n", region.offset, region.length);
}
}
}
} else {
char *org = ""; /* organization */
@ -438,12 +388,15 @@ vinum_lpi(int plexno, int recurse)
if (plex.writes != 0)
printf("\t\tAverage write:\t%16qd bytes\n",
plex.bytes_written / plex.writes);
if ((plex.organization == plex_striped)
|| (plex.organization == plex_raid5))
printf("\t\tMultiblock:\t%16qd\n"
"\t\tMultistripe:\t%16qd\n",
if (((plex.reads + plex.writes) > 0)
&& ((plex.organization == plex_striped)
|| (plex.organization == plex_raid5)))
printf("\t\tMultiblock:\t%16qd (%d%%)\n"
"\t\tMultistripe:\t%16qd (%d%%)\n",
plex.multiblock,
plex.multistripe);
(int) (plex.multiblock * 100 / (plex.reads + plex.writes)),
plex.multistripe,
(int) (plex.multistripe * 100 / (plex.reads + plex.writes)));
}
if (plex.subdisks > 0) {
int sdno;
@ -512,10 +465,16 @@ vinum_lsi(int sdno, int recurse)
if (sd.plexno >= 0) {
get_plex_info(&plex, sd.plexno);
printf("\t\tPlex %s", plex.name);
if (plex.organization == plex_concat)
printf(" at offset %qd\n", (long long) sd.plexoffset * DEV_BSIZE);
else
printf("\n");
printf(" at offset %qd\n", (long long) sd.plexoffset * DEV_BSIZE);
}
if (sd.state == sd_reviving) {
printf("\t\tRevive pointer:\t\t%s (%d%%)\n",
roughlength(sd.revived << DEV_BSHIFT, 0),
(int) (((u_int64_t) (sd.revived * 100)) / sd.sectors));
printf("\t\tRevive blocksize:\t%s\n"
"\t\tRevive interval:\t%10d seconds\n",
roughlength(sd.revive_blocksize, 0),
sd.revive_interval);
}
} else {
printf("S %-21s State: %s\tPO: %s ",
@ -633,7 +592,7 @@ vinum_info(int argc, char *argv[], char *argv0[])
#if VINUMDEBUG
struct rqinfo rq;
#endif
if (ioctl(superdev, VINUM_GETCONFIG, &vinum_conf) < 0) {
perror("Can't get vinum config");
return;
@ -668,7 +627,7 @@ vinum_info(int argc, char *argv[], char *argv0[])
}
#if VINUMDEBUG
if (Verbose) {
printf("\nTime\t\t Event\t Buf\tSD\tDev\tOffset\tBytes\tDoffset\tGoffset\n\n");
printf("\nTime\t\t Event\t Buf\tDev\tOffset\t\tBytes\tSD\tSDoff\tDoffset\tGoffset\n\n");
for (i = RQINFO_SIZE - 1; i >= 0; i--) { /* go through the request list in order */
*((int *) &rq) = i;
if (ioctl(superdev, VINUM_RQINFO, &rq) < 0) {
@ -680,75 +639,77 @@ vinum_info(int argc, char *argv[], char *argv0[])
break;
case loginfo_user_bp: /* this is the bp when strategy is called */
printf("%s 1VS %s %p\t\t0x%x\t0x%x\t%ld\n",
printf("%s 1VS %s %p\t0x%x\t0x%-9x\t%ld\n",
timetext(&rq.timestamp),
rq.info.b.b_flags & B_READ ? "Read " : "Write",
rq.bp,
/* no subdisk */
rq.info.b.b_dev,
rq.info.b.b_blkno,
rq.info.b.b_bcount);
break;
case loginfo_user_bpl: /* and this is the bp at launch time */
printf("%s 2LR %s %p\t\t0x%x\t0x%x\t%ld\n",
printf("%s 2LR %s %p\t0x%x\t0x%-9x\t%ld\n",
timetext(&rq.timestamp),
rq.info.b.b_flags & B_READ ? "Read " : "Write",
rq.bp,
/* no subdisk */
rq.info.b.b_dev,
rq.info.b.b_blkno,
rq.info.b.b_bcount);
break;
case loginfo_rqe: /* user RQE */
printf("%s 3RQ %s %p\t%d\t0x%x\t0x%x\t%ld\t%x\t%x\n",
printf("%s 3RQ %s %p\t0x%x\t0x%-9x\t%ld\t%d\t%x\t%x\t%x\n",
timetext(&rq.timestamp),
rq.info.rqe.b.b_flags & B_READ ? "Read " : "Write",
rq.bp,
rq.info.rqe.sdno,
rq.info.rqe.b.b_dev,
rq.info.rqe.b.b_blkno,
rq.info.rqe.b.b_bcount,
rq.info.rqe.sdno,
rq.info.rqe.sdoffset,
rq.info.rqe.dataoffset,
rq.info.rqe.groupoffset);
break;
case loginfo_iodone: /* iodone called */
printf("%s 4DN %s %p\t%d\t0x%x\t0x%x\t%ld\t%x\t%x\n",
printf("%s 4DN %s %p\t0x%x\t0x%-9x\t%ld\t%d\t%x\t%x\t%x\n",
timetext(&rq.timestamp),
rq.info.rqe.b.b_flags & B_READ ? "Read " : "Write",
rq.bp,
rq.info.rqe.sdno,
rq.info.rqe.b.b_dev,
rq.info.rqe.b.b_blkno,
rq.info.rqe.b.b_bcount,
rq.info.rqe.sdno,
rq.info.rqe.sdoffset,
rq.info.rqe.dataoffset,
rq.info.rqe.groupoffset);
break;
case loginfo_raid5_data: /* RAID-5 write data block */
printf("%s 5RD %s %p\t%d\t0x%x\t0x%x\t%ld\t%x\t%x\n",
printf("%s 5RD %s %p\t0x%x\t0x%-9x\t%ld\t%d\t%x\t%x\t%x\n",
timetext(&rq.timestamp),
rq.info.rqe.b.b_flags & B_READ ? "Read " : "Write",
rq.bp,
rq.info.rqe.sdno,
rq.info.rqe.b.b_dev,
rq.info.rqe.b.b_blkno,
rq.info.rqe.b.b_bcount,
rq.info.rqe.sdno,
rq.info.rqe.sdoffset,
rq.info.rqe.dataoffset,
rq.info.rqe.groupoffset);
break;
case loginfo_raid5_parity: /* RAID-5 write parity block */
printf("%s 6RP %s %p\t%d\t0x%x\t0x%x\t%ld\t%x\t%x\n",
printf("%s 6RP %s %p\t0x%x\t0x%-9x\t%ld\t%d\t%x\t%x\t%x\n",
timetext(&rq.timestamp),
rq.info.rqe.b.b_flags & B_READ ? "Read " : "Write",
rq.bp,
rq.info.rqe.sdno,
rq.info.rqe.b.b_dev,
rq.info.rqe.b.b_blkno,
rq.info.rqe.b.b_bcount,
rq.info.rqe.sdno,
rq.info.rqe.sdoffset,
rq.info.rqe.dataoffset,
rq.info.rqe.groupoffset);
}
@ -820,7 +781,7 @@ vinum_printconfig(int argc, char *argv[], char *argv0[])
plex_state(plex.state),
plex_org(plex.organization));
if ((plex.organization == plex_striped)
) {
|| (plex.organization == plex_raid5)) {
fprintf(of, "%db ", (int) plex.stripesize);
}
if (plex.volno >= 0) { /* we have a volume */

View File

@ -36,7 +36,7 @@
*
*/
/* $Id: v.c,v 1.2 1998/12/28 16:32:39 peter Exp $ */
/* $Id: v.c,v 1.24 1999/01/17 02:53:38 grog Exp grog $ */
#include <ctype.h>
#include <errno.h>
@ -57,6 +57,9 @@
#include <sys/wait.h>
#include <readline/history.h>
#include <readline/readline.h>
#include <sys/linker.h>
#include <sys/module.h>
#include <sys/resource.h>
FILE *cf; /* config file handle */
@ -88,6 +91,8 @@ struct drive drive;
jmp_buf command_fail; /* return on a failed command */
int superdev; /* vinum super device */
void start_daemon(void);
#define ofs(x) ((void *) (& ((struct confdata *) 0)->x)) /* offset of x in struct confdata */
/* create description-file
@ -112,6 +117,20 @@ int tokens; /* number of tokens */
int
main(int argc, char *argv[])
{
#if RAID5
#define VINUMMOD "Vinum"
#else
#define VINUMMOD "vinum"
#endif
if (modfind(VINUMMOD) < 0) {
/* need to load the vinum module */
if (kldload(VINUMMOD) < 0 || modfind(VINUMMOD) < 0) {
perror("vinum kernel module not available");
return 1;
}
}
superdev = open(VINUM_SUPERDEV_NAME, O_RDWR); /* open it */
if (superdev < 0) { /* no go */
@ -122,6 +141,10 @@ main(int argc, char *argv[])
return 1;
}
}
/* Check if the dæmon is running. If not, start it in the
* background */
start_daemon();
if (argc > 1) { /* we have a command on the line */
if (setjmp(command_fail) != 0) /* long jumped out */
return -1;
@ -156,6 +179,13 @@ main(int argc, char *argv[])
return 0; /* normal completion */
}
/* stop the hard way */
void
vinum_quit(int argc, char *argv[], char *argv0[])
{
exit(0);
}
#define FUNKEY(x) { kw_##x, &vinum_##x } /* create pair "kw_foo", vinum_foo */
struct funkey {
@ -191,6 +221,10 @@ struct funkey {
FUNKEY(printconfig),
FUNKEY(start),
FUNKEY(stop),
FUNKEY(makedev),
FUNKEY(help),
FUNKEY(quit),
FUNKEY(setdaemon),
FUNKEY(resetstats)
};
@ -348,6 +382,7 @@ make_devices(void)
system("mkdir -p " VINUM_DIR "/drive " /* and make them again */
VINUM_DIR "/plex "
VINUM_DIR "/sd "
VINUM_DIR "/rsd "
VINUM_DIR "/vol "
VINUM_DIR "/rvol "
VINUM_RDIR);
@ -440,9 +475,15 @@ make_devices(void)
if (mknod(filename, S_IRWXU | S_IRGRP | S_IXGRP | S_IROTH | S_IFBLK, sddev) < 0)
fprintf(stderr, "Can't create %s: %s\n", filename, strerror(errno));
/* And /dev/vinum/sd/<sd> */
/* /dev/vinum/sd/<sd> */
sprintf(filename, VINUM_DIR "/sd/%s", sd.name);
if (mknod(filename, S_IRWXU | S_IRGRP | S_IXGRP | S_IROTH | S_IFBLK, sddev) < 0)
fprintf(stderr, "Can't create %s: %s\n", filename, strerror(errno));
/* And /dev/vinum/rsd/<sd> */
sprintf(filename, VINUM_DIR "/rsd/%s", sd.name);
sddev = VINUMCDEV(volno, plexno, sdno, VINUM_SD_TYPE);
if (mknod(filename, S_IRWXU | S_IRGRP | S_IXGRP | S_IROTH | S_IFCHR, sddev) < 0)
fprintf(stderr, "Can't create %s: %s\n", filename, strerror(errno));
}
}
@ -461,6 +502,13 @@ make_devices(void)
}
}
/* command line interface for the 'makedev' command */
void
vinum_makedev(int argc, char *argv[], char *arg0[])
{
make_devices();
}
/* Find the object "name". Return object type at type,
* and the index as the return value.
* If not found, return -1 and invalid_object.
@ -516,13 +564,13 @@ find_object(const char *name, enum objecttype *type)
return -1;
}
/* Continue reviving a plex in the background */
/* Continue reviving a subdisk in the background */
void
continue_revive(int plexno)
continue_revive(int sdno)
{
struct plex plex;
struct sd sd;
pid_t pid;
get_plex_info(&plex, plexno);
get_sd_info(&sd, sdno);
#if VINUMDEBUG
if (debug)
@ -537,27 +585,59 @@ continue_revive(int plexno)
struct vinum_ioctl_msg *message = (struct vinum_ioctl_msg *) &reply;
openlog("vinum", LOG_CONS | LOG_PERROR | LOG_PID, LOG_KERN);
syslog(LOG_INFO | LOG_KERN, "reviving plex %s", plex.name);
syslog(LOG_INFO | LOG_KERN, "reviving %s", sd.name);
for (reply.error = EAGAIN; reply.error == EAGAIN;) {
message->index = plexno; /* pass plex number */
message->type = plex_object; /* and type of object */
message->index = sdno; /* pass sd number */
message->type = sd_object; /* and type of object */
message->state = object_up;
ioctl(superdev, VINUM_SETSTATE, message);
}
if (reply.error) {
syslog(LOG_ERR | LOG_KERN,
"can't revive plex %s: %s",
plex.name,
"can't revive %s: %s",
sd.name,
reply.msg[0] ? reply.msg : strerror(reply.error));
exit(1);
} else {
get_plex_info(&plex, plexno); /* update the info */
syslog(LOG_INFO | LOG_KERN, "plex %s is %s", plex.name, plex_state(plex.state));
get_sd_info(&sd, sdno); /* update the info */
syslog(LOG_INFO | LOG_KERN, "%s is %s", sd.name, sd_state(sd.state));
exit(0);
}
} else if (pid < 0) /* couldn't fork? */
fprintf(stderr, "Can't continue reviving %s: %s\n", plex.name, strerror(errno));
fprintf(stderr, "Can't continue reviving %s: %s\n", sd.name, strerror(errno));
else
printf("Reviving %s in the background\n", plex.name);
printf("Reviving %s in the background\n", sd.name);
}
/* Check if the daemon is running,
* start it if it isn't. The check itself
* could take a while, so we do it as a separate
* process, which will become the daemon if one isn't
* running already */
void
start_daemon(void)
{
int pid;
int status;
int error;
pid = (int) fork();
if (pid == 0) { /* We're the child, do the work */
error = daemon(0, 0); /* this will fork again, but who's counting? */
if (error != 0) {
fprintf(stderr, "Can't start daemon: %s (%d)\n", strerror(errno), errno);
exit(1);
}
setproctitle("Vinum daemon"); /* show what we're doing */
status = ioctl(superdev, VINUM_FINDDAEMON, NULL);
if (status != 0) { /* no daemon, */
ioctl(superdev, VINUM_DAEMON, &verbose); /* we should hang here */
syslog(LOG_ERR | LOG_KERN, "%s", strerror(errno));
exit(1);
}
exit(0); /* when told to die */
} else if (pid < 0) /* couldn't fork */
printf("Can't fork to check daemon\n");
}

View File

@ -1,6 +1,9 @@
/*
* Copyright (c) 1997 Nan Yang Computer Services Limited
* All rights reserved.
/*-
* Copyright (c) 1997, 1998
* Nan Yang Computer Services Limited. All rights reserved.
*
* This software is distributed under the so-called ``Berkeley
* License'':
*
* Redistribution and use in source and binary forms, with or without
* modification, are permitted provided that the following conditions
@ -12,28 +15,30 @@
* documentation and/or other materials provided with the distribution.
* 3. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed for the NetBSD Project
* by Jason R. Thorpe.
* 4. The name of the author may not be used to endorse or promote products
* derived from this software without specific prior written permission.
* This product includes software developed by Nan Yang Computer
* Services Limited.
* 4. Neither the name of the Company nor the names of its contributors
* may be used to endorse or promote products derived from this software
* without specific prior written permission.
*
* This software is provided ``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 company 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.
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``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 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.
*/
/* $Id: vext.h,v 1.1.1.1 1998/09/16 05:57:36 grog Exp $ */
/* $Id: vext.h,v 1.10 1999/01/12 04:31:45 grog Exp grog $ */
#define MAXARGS 64 /* maximum number of args on a line */
#define PLEXINITSIZE MAXPHYS /* block size to write when initializing */
#define PLEXINITSIZE 61440 /* this is what the system does somewhere */
enum {
KILOBYTE = 1024,
@ -61,6 +66,10 @@ void vinum_resetconfig(int argc, char *argv[], char *arg0[]);
void vinum_start(int argc, char *argv[], char *arg0[]);
void continue_revive(int plexno);
void vinum_stop(int argc, char *argv[], char *arg0[]);
void vinum_makedev(int argc, char *argv[], char *arg0[]);
void vinum_help(int argc, char *argv[], char *arg0[]);
void vinum_quit(int argc, char *argv[], char *arg0[]);
void vinum_setdaemon(int argc, char *argv[], char *arg0[]);
void reset_volume_stats(int volno, int recurse);
void reset_plex_stats(int plexno, int recurse);
void reset_sd_stats(int sdno, int recurse);
@ -77,6 +86,7 @@ void vinum_ld(int argc, char *argv[], char *arg0[]);
void vinum_ls(int argc, char *argv[], char *arg0[]);
void vinum_lp(int argc, char *argv[], char *arg0[]);
void vinum_lv(int argc, char *argv[], char *arg0[]);
void start_daemon(void);
#ifdef VINUMDEBUG
void vinum_debug(int argc, char *argv[], char *arg0[]);
#endif

View File

@ -1,6 +1,6 @@
.\" Hey, Emacs, edit this file in -*- nroff-fill -*- mode
.\"
.Dd 11 July 1998
.Dd 15 January 1999
.Dt vinum 8
.Os FreeBSD
.Sh NAME
@ -30,9 +30,14 @@ Attach a plex to a volume, or a subdisk to a plex.
.in +1i
Cause the volume manager to enter the kernel debugger.
.in
.Cd debug
.Ar flags
.in +1i
Set debugging flags.
.in
.Cd detach
.Op Ar plex | subdisk
.in +1
.in +1i
Detach a plex or subdisk from the volume or plex to which it is attached.
.in
.Cd info
@ -112,12 +117,21 @@ List information about volumes
Remake the device nodes in
.Ar /dev/vinum .
.in
.Cd quit
.in +1i
Exit the
.Nm
program when running in interactive mode. Normally this would be done by
entering the
.Ar EOF
character.
.in
.Cd read
.Ar disk-partition
.Ar disk Op disk...
.in +1i
Read the
.Nm
configuration from the specified disk partition.
configuration from the specified disks.
.in
.Cd rename Op Fl r
.Ar [ drive | subdisk | plex | volume ]
@ -164,6 +178,11 @@ XXX
Set the state of the object to \fIstate\fP\|
.in
..
.Cd setdaemon
.Op value
.in +1i
Set dæmon configuration.
.in
.Cd start
.Op volume | plex | subdisk
.in +1i
@ -184,7 +203,10 @@ for more information about the volume manager.
.Xr vinum 8
is designed either for interactive use, when started without a command, or to
execute a single command if the command is supplied as arguments to
.Nm vinum.
.Nm vinum .
In interactive mode,
.Nm
maintains a command line history.
.Ss OPTIONS
.Nm
commands may optionally be followed by an option. Any of the following options
@ -245,7 +267,8 @@ commands perform the following functions:
.Bl -hang
.It Nm attach Ar plex Ar volume
.Op Nm rename
.sp -1v
.if n .sp -1v
.if t .sp -.6v
.It Nm attach Ar subdisk Ar plex Ar [offset]
.Op Nm rename
.sp
@ -300,13 +323,49 @@ CONFIGURATION FILE below for more information.
.Ar debug
is used to enter the remote kernel debugger. It is only activated if
.Nm
is built with the DEBUG option. This option will stop the execution of the
operating system until the kernel debugger is exited. If there is no remote
is built with the
.Ar VINUMDEBUG
option. This option will stop the execution of the operating system until the
kernel debugger is exited. If remote debugging is set and there is no remote
connection for a kernel debugger, it will be necessary to reset the system and
reboot in order to leave the debugger.
.It Nm debug
.Ar flags
.Pp
Set a bit mask of internal debugging flags. These will change without warning
as the product matures; to be certain, read the header file
.Pa sys/dev/vinumvar.h .
The bit mask is composed of the following values:
.Bl -hang
.It DEBUG_ADDRESSES (1)
.br
Show buffer information during requests
.It DEBUG_NUMOUTPUT (2)
.br
Show the value of
.Dv vp->v_numoutput.
.It DEBUG_RESID (4)
.br
Go into debugger in
.Fd complete_rqe.
.It DEBUG_LASTREQS (8)
.br
Keep a circular buffer of last requests.
.It DEBUG_REVIVECONFLICT (16)
.br
Print info about revive conflicts.
.It DEBUG_REMOTEGDB (256)
.br
Go into remote
.Ic gdb
when the
.Nm debug
command is issued.
.El
.It Nm detach Op Fl f
.Ar plex
.sp -1v
.if n .sp -1v
.if t .sp -.6v
.It Nm detach Op Fl f
.Ar subdisk
.sp
@ -327,7 +386,7 @@ the subdisk is marked absent, and can later be replaced with the
.Nm attach
command.
.It Nm info
.Pp
.br
.Nm
.Ar info
displays information about
@ -335,6 +394,117 @@ displays information about
memory usage. This is intended primarily for debugging. With the
.Fl v
option, it will give detailed information about the memory areas in use.
.Pp
With the
.Fl V
option,
.Ar info
displays information about the last up to 64 I/O requests handled by the
.Nm
driver. This information is only collected if debug flag 8 is set. The format
looks like:
.Pp
.Bd -literal
vinum -> info -V
Flags: 0x200 1 opens
Total of 38 blocks malloced, total memory: 16460
Maximum allocs: 56, malloc table at 0xf0f72dbc
Time Event Buf Dev Offset Bytes SD SDoff Doffset Goffset
14:40:00.637758 1VS Write 0xf2361f40 0x5b03 0x10 16384
14:40:00.639280 2LR Write 0xf2361f40 0x5b03 0x10 16384
14:40:00.639294 3RQ Read 0xf2361f40 0x427 0x104109 8192 19 0 0 0
14:40:00.639455 3RQ Read 0xf2361f40 0x417 0xd2109 8192 17 0 0 0
14:40:00.639529 3RQ Read 0xf2361f40 0x40f 0x6e109 8192 16 0 0 0
14:40:00.652978 4DN Read 0xf2361f40 0x427 0x104109 8192 19 0 0 0
14:40:00.667040 4DN Read 0xf2361f40 0x40f 0x6e109 8192 16 0 0 0
14:40:00.668556 4DN Read 0xf2361f40 0x417 0xd2109 8192 17 0 0 0
14:40:00.669777 6RP Write 0xf2361f40 0x427 0x104109 8192 19 0 0 0
14:40:00.685547 4DN Write 0xf2361f40 0x427 0x104109 8192 19 0 0 0
.Ed
.Pp
The
.Ar Buf
field always contains the address of the user buffer header. This can be used
to identify the requests associated with a user request, though this is not 100%
reliable: theoretically two requests in sequence could use the same buffer
header, though this is not common. The beginning of a request can be identified
by the event
.Ar 1VS .
The example above shows the requests involved in a single user request.
.Pp
The
.Ar Event
field contains information related to the sequence of events in the request
chain. The digit
.Ar 1
to
.Ar 6
indicates the approximate sequence of events, and the two-letter abbreviation is
a mnemonic for the location
.Bl -hang
.It 1VS
(vinumstrategy) shows information about the user request on entry to
.Fd vinumstrategy .
The device number is the
.Nm
device, and offset and length are the user parameters. This is always the
beginning of a request sequence.
.It 2LR
(launch_requests) shows the user request just prior to launching the low-level
.Nm
requests in the function
.Fd launch_requests.
The parameters should be the same as in the
.Ar 1VS
information.
.Pp
In the following requests,
.Ar Dev
is the device number of the associated disk partition,
.Ar Offset
is the offset from the beginning of the partition,
.Ar SD
is the subdisk index in
.Dv vinum_conf ,
.Ar SDoff
is the offset from the beginning of the subdisk,
.Ar Doffset
is the offset of the associated data request, and
.Ar Goffset
is the offset of the associated group request, where applicable.
.It 3RQ
(request) shows one of possibly several low-level
.Nm
requests which are launched to satisfy the high-level request. This information
is also logged in
.Fd launch_requests.
.It 4DN
(done) is called from
.Fd complete_rqe,
showing the completion of a request. This completion should match a request
launched either at stage
.Ar 4DN
from
.Fd launch_requests,
or from
.Fd complete_raid5_write
at stage
.Ar 5RD
or
.Ar 6RP .
.It 5RD
(RAID-5 data) is called from
.Fd complete_raid5_write
and represents the data written to a RAID-5 data stripe after calculating
parity.
.It 6RP
(RAID-5 parity) is called from
.Fd complete_raid5_write
and represents the data written to a RAID-5 parity stripe after calculating
parity.
.El
.\" XXX
.It Nm init Ar plex
.Pp
@ -374,33 +544,38 @@ This command is deprecated.
.Op Fl r
.Op Fl V
.Op volume | plex | subdisk
.sp -1
.if n .sp -1v
.if t .sp -.6v
.It Nm l
.Op Fl r
.Op Fl V
.Op volume | plex | subdisk
.sp -1
.if n .sp -1v
.if t .sp -.6v
.It Nm ld
.Op Fl r
.Op Fl s
.Op Fl v
.Op Fl V
.Op volume
.sp -1
.if n .sp -1v
.if t .sp -.6v
.It Nm ls
.Op Fl r
.Op Fl s
.Op Fl v
.Op Fl V
.Op subdisk
.sp -1
.if n .sp -1v
.if t .sp -.6v
.It Nm lp
.Op Fl r
.Op Fl s
.Op Fl v
.Op Fl V
.Op plex
.sp -1
.if n .sp -1v
.if t .sp -.6v
.It Nm lv
.Op Fl r
.Op Fl s
@ -447,18 +622,34 @@ command removes the directory /dev/vinum and recreates it with device nodes
which reflect the current configuration. This command is not intended for
general use, and is provided for emergency use only.
.Pp
.It Nm quit
Exit the
.Nm
program when running in interactive mode. Normally this would be done by
entering the
.Ar EOF
character.
.It Nm read
.Ar disk-partition
.Ar disk Op disk...
.Pp
The
.Nm read
command reads a previously created
.Nm
configuration from the specified disk partition.
command scans the specified disks for
.Nm
partitions containing previously created configuration information. It reads
the configuration in order from the most recently updated to least recently
updated configuration. This is the normal method used to start
.Nm vinum .
.Nm
maintains an up-to-date copy of all configuration information on each disk
partition. You can specify any of the partitions in a configuration as the
partition. You must specify all of the partitions in a configuration as the
parameter to this command.
.Nm
.Nm read
will accept the names of non-Vinum partitions, so you can specify all partitions
upon which a
.Nm
partition might be resident to this command.
.It Nm rename
.Op Fl r
.Ar [ drive | subdisk | plex | volume ]
@ -568,6 +759,15 @@ care: it can result in total loss of data on a volume.
.\"XXX
.Nm This command has not yet been implemented.
..
.It Nm setdaemon
.Op value
.Pp
.Nm setdaemon
sets a variable bitmask for the
.Nm
dæmon. This command is temporary and will be replaced. Currently, the bit mask
may contain the bits 1 (log every action to syslog) and 4 (don't update
configuration). Option bit 4 can be useful for error recovery.
.It Nm start
.Op volume | plex | subdisk
.Pp
@ -1019,14 +1219,62 @@ commands, though especially
.Nm resetconfig
outputs all sort of dire warnings. Don't use these commands unless you have a
good reason to do so.
.It
Some state transitions are not very intuitive. In fact, it's not clear whether
this is a bug or a feature. If you find that you can't start an object in some
strange state, such as a
.Ar reborn
subdisk, try first to get it into
.Ar stopped
state, with the
.Nm stop
or
.Nm stop Ar -f
commands. If that works, you should then be able to start it. If you find
that this is the only way to get out of a position where easier methods fail,
please report the situation.
.It
If you build the kernel module with the
.Ar -DVINUMDEBUG
option, you must also build
.Nm vinum(8)
with the
.Ar -DVINUMDEBUG
option, since the size of some data objects used by both components depends on
this option. If you don't do so, commands will fail with the message
.Ar Invalid argument ,
and a console message will be logged such as
.Pp
.Bd -literal
vinumioctl: invalid ioctl from process 247 (vinum): c0e44642
.Ed
.Pp
This error may also occur if you use old versions of kld or userland program.
.El
.Sh BUGS
.Bl -enum
.It
.Nm
is currently in beta test. Many bugs can be expected. The configuration
mechanism is not yet fully functional. If you have difficulties, please look at
http://www.lemis.com/vinum_beta.html and
http://www.lemis.com/vinum_debugging.html before reporting problems.
.Pp
.It
It is possible to unload
.Nm
with the
.Nm kldunload
command. This is buggy, and the only reason it is present at all is to make it
easier for people testing the system: the alternative is a reboot. It works
about 80% of the time: expect about one panic every five unloads.
.It
It is possible to configure
.Nm
statically, but it has never been tested in this form. Don't even bother to
report the problem if you have trouble with a static
.Nm
pseudo-device.
.It
This man page tickles a bug in the
.Ar doc
man page macros: depending on your screen size, you may get the error message
@ -1037,6 +1285,7 @@ warning: page 6: table text block will not fit on one page
This will be fixed in a later rewrite of the page, when the
.Nm tbl
macros will be removed.
.El
.Sh FILES
.Ar /dev/vinum
- directory with device nodes for