1
0
mirror of https://git.FreeBSD.org/src.git synced 2024-12-25 11:37:56 +00:00
freebsd/sys/cam/cam_periph.c

1560 lines
38 KiB
C
Raw Normal View History

/*
* Common functions for CAM "type" (peripheral) drivers.
*
* Copyright (c) 1997, 1998 Justin T. Gibbs.
* Copyright (c) 1997, 1998 Kenneth D. Merry.
* All rights reserved.
*
* 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,
* without modification, immediately at the beginning of the file.
* 2. The name of the author 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.
*
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
* $Id: cam_periph.c,v 1.6 1998/10/22 22:16:48 ken Exp $
*/
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/types.h>
#include <sys/malloc.h>
#include <sys/kernel.h>
#include <sys/buf.h>
#include <sys/proc.h>
#include <sys/devicestat.h>
#include <vm/vm.h>
#include <vm/vm_extern.h>
#include <cam/cam.h>
#include <cam/cam_conf.h>
#include <cam/cam_ccb.h>
#include <cam/cam_xpt_periph.h>
#include <cam/cam_periph.h>
#include <cam/cam_debug.h>
#include <cam/scsi/scsi_all.h>
#include <cam/scsi/scsi_message.h>
#include <cam/scsi/scsi_da.h>
#include <cam/scsi/scsi_pass.h>
static u_int camperiphnextunit(struct periph_driver *p_drv,
u_int newunit, int wired);
static u_int camperiphunit(struct periph_driver *p_drv,
path_id_t path_id_t,
target_id_t target, lun_id_t lun);
static void camperiphdone(struct cam_periph *periph,
union ccb *done_ccb);
static void camperiphfree(struct cam_periph *periph);
cam_status
Fix a problem with the way we handled device invalidation when attaching to a device failed. In theory, the same steps that happen when we get an AC_LOST_DEVICE async notification should have been taken when a driver fails to attach. In practice, that wasn't the case. This only affected the da, cd and ch drivers, but the fix affects all peripheral drivers. There were several possible problems: - In the da driver, we didn't remove the peripheral's softc from the da driver's linked list of softcs. Once the peripheral and softc got removed, we'd get a kernel panic the next time the timeout routine called dasendorderedtag(). - In the da, cd and possibly ch drivers, we didn't remove the peripheral's devstat structure from the devstat queue. Once the peripheral and softc were removed, this could cause a panic if anyone tried to access device statistics. (one component of the linked list wouldn't exist anymore) - In the cd driver, we didn't take the peripheral off the changer run queue if it was scheduled to run. In practice, it's highly unlikely, and maybe impossible that the peripheral would have been on the changer run queue at that stage of the probe process. The fix is: - Add a new peripheral callback function (the "oninvalidate" function) that is called the first time cam_periph_invalidate() is called for a peripheral. - Create new foooninvalidate() routines for each peripheral driver. This routine is always called at splsoftcam(), and contains all the stuff that used to be in the AC_LOST_DEVICE case of the async callback handler. - Move the devstat cleanup call to the destructor/cleanup routines, since some of the drivers do I/O in their close routines. - Make sure that when we're flushing the buffer queue, we traverse it at splbio(). - Add a check for the invalid flag in the pt driver's open routine. Reviewed by: gibbs
1998-10-22 22:16:56 +00:00
cam_periph_alloc(periph_ctor_t *periph_ctor,
periph_oninv_t *periph_oninvalidate,
periph_dtor_t *periph_dtor, periph_start_t *periph_start,
char *name, cam_periph_type type, struct cam_path *path,
ac_callback_t *ac_callback, ac_code code, void *arg)
{
struct periph_driver **p_drv;
struct cam_periph *periph;
struct cam_periph *cur_periph;
path_id_t path_id;
target_id_t target_id;
lun_id_t lun_id;
cam_status status;
u_int init_level;
int s;
init_level = 0;
/*
* Handle Hot-Plug scenarios. If there is already a peripheral
* of our type assigned to this path, we are likely waiting for
* final close on an old, invalidated, peripheral. If this is
* the case, queue up a deferred call to the peripheral's async
* handler. If it looks like a mistaken re-alloation, complain.
*/
if ((periph = cam_periph_find(path, name)) != NULL) {
if ((periph->flags & CAM_PERIPH_INVALID) != 0
&& (periph->flags & CAM_PERIPH_NEW_DEV_FOUND) == 0) {
periph->flags |= CAM_PERIPH_NEW_DEV_FOUND;
periph->deferred_callback = ac_callback;
periph->deferred_ac = code;
return (CAM_REQ_INPROG);
} else {
printf("cam_periph_alloc: attempt to re-allocate "
"valid device %s%d rejected\n",
periph->periph_name, periph->unit_number);
}
return (CAM_REQ_INVALID);
}
periph = (struct cam_periph *)malloc(sizeof(*periph), M_DEVBUF,
M_NOWAIT);
if (periph == NULL)
return (CAM_RESRC_UNAVAIL);
init_level++;
for (p_drv = (struct periph_driver **)periphdriver_set.ls_items;
*p_drv != NULL; p_drv++) {
if (strcmp((*p_drv)->driver_name, name) == 0)
break;
}
path_id = xpt_path_path_id(path);
target_id = xpt_path_target_id(path);
lun_id = xpt_path_lun_id(path);
bzero(periph, sizeof(*periph));
cam_init_pinfo(&periph->pinfo);
periph->periph_start = periph_start;
periph->periph_dtor = periph_dtor;
Fix a problem with the way we handled device invalidation when attaching to a device failed. In theory, the same steps that happen when we get an AC_LOST_DEVICE async notification should have been taken when a driver fails to attach. In practice, that wasn't the case. This only affected the da, cd and ch drivers, but the fix affects all peripheral drivers. There were several possible problems: - In the da driver, we didn't remove the peripheral's softc from the da driver's linked list of softcs. Once the peripheral and softc got removed, we'd get a kernel panic the next time the timeout routine called dasendorderedtag(). - In the da, cd and possibly ch drivers, we didn't remove the peripheral's devstat structure from the devstat queue. Once the peripheral and softc were removed, this could cause a panic if anyone tried to access device statistics. (one component of the linked list wouldn't exist anymore) - In the cd driver, we didn't take the peripheral off the changer run queue if it was scheduled to run. In practice, it's highly unlikely, and maybe impossible that the peripheral would have been on the changer run queue at that stage of the probe process. The fix is: - Add a new peripheral callback function (the "oninvalidate" function) that is called the first time cam_periph_invalidate() is called for a peripheral. - Create new foooninvalidate() routines for each peripheral driver. This routine is always called at splsoftcam(), and contains all the stuff that used to be in the AC_LOST_DEVICE case of the async callback handler. - Move the devstat cleanup call to the destructor/cleanup routines, since some of the drivers do I/O in their close routines. - Make sure that when we're flushing the buffer queue, we traverse it at splbio(). - Add a check for the invalid flag in the pt driver's open routine. Reviewed by: gibbs
1998-10-22 22:16:56 +00:00
periph->periph_oninval = periph_oninvalidate;
periph->type = type;
periph->periph_name = name;
periph->unit_number = camperiphunit(*p_drv, path_id, target_id, lun_id);
periph->immediate_priority = CAM_PRIORITY_NONE;
periph->refcount = 0;
SLIST_INIT(&periph->ccb_list);
status = xpt_create_path(&path, periph, path_id, target_id, lun_id);
if (status != CAM_REQ_CMP)
goto failure;
periph->path = path;
init_level++;
status = xpt_add_periph(periph);
if (status != CAM_REQ_CMP)
goto failure;
s = splsoftcam();
cur_periph = TAILQ_FIRST(&(*p_drv)->units);
while (cur_periph != NULL
&& cur_periph->unit_number < periph->unit_number)
cur_periph = TAILQ_NEXT(cur_periph, unit_links);
if (cur_periph != NULL)
TAILQ_INSERT_BEFORE(cur_periph, periph, unit_links);
else {
TAILQ_INSERT_TAIL(&(*p_drv)->units, periph, unit_links);
(*p_drv)->generation++;
}
splx(s);
init_level++;
status = periph_ctor(periph, arg);
if (status == CAM_REQ_CMP)
init_level++;
failure:
switch (init_level) {
case 4:
/* Initialized successfully */
break;
case 3:
s = splsoftcam();
TAILQ_REMOVE(&(*p_drv)->units, periph, unit_links);
splx(s);
xpt_remove_periph(periph);
case 2:
xpt_free_path(periph->path);
case 1:
free(periph, M_DEVBUF);
case 0:
/* No cleanup to perform. */
break;
default:
panic("cam_periph_alloc: Unkown init level");
}
return(status);
}
/*
* Find a peripheral structure with the specified path, target, lun,
* and (optionally) type. If the name is NULL, this function will return
* the first peripheral driver that matches the specified path.
*/
struct cam_periph *
cam_periph_find(struct cam_path *path, char *name)
{
struct periph_driver **p_drv;
struct cam_periph *periph;
int s;
for (p_drv = (struct periph_driver **)periphdriver_set.ls_items;
*p_drv != NULL; p_drv++) {
if (name != NULL && (strcmp((*p_drv)->driver_name, name) != 0))
continue;
s = splsoftcam();
for (periph = TAILQ_FIRST(&(*p_drv)->units); periph != NULL;
periph = TAILQ_NEXT(periph, unit_links)) {
if (xpt_path_comp(periph->path, path) == 0) {
splx(s);
return(periph);
}
}
splx(s);
if (name != NULL)
return(NULL);
}
return(NULL);
}
cam_status
cam_periph_acquire(struct cam_periph *periph)
{
int s;
if (periph == NULL)
return(CAM_REQ_CMP_ERR);
s = splsoftcam();
periph->refcount++;
splx(s);
return(CAM_REQ_CMP);
}
void
cam_periph_release(struct cam_periph *periph)
{
int s;
if (periph == NULL)
return;
s = splsoftcam();
if ((--periph->refcount == 0)
&& (periph->flags & CAM_PERIPH_INVALID)) {
camperiphfree(periph);
}
splx(s);
}
/*
* Look for the next unit number that is not currently in use for this
* peripheral type starting at "newunit". Also exclude unit numbers that
* are reserved by for future "hardwiring" unless we already know that this
* is a potential wired device. Only assume that the device is "wired" the
* first time through the loop since after that we'll be looking at unit
* numbers that did not match a wiring entry.
*/
static u_int
camperiphnextunit(struct periph_driver *p_drv, u_int newunit, int wired)
{
struct cam_periph *periph;
struct cam_periph_config *periph_conf;
char *periph_name;
int s;
s = splsoftcam();
periph_name = p_drv->driver_name;
for (;;newunit++) {
for (periph = TAILQ_FIRST(&p_drv->units);
periph != NULL && periph->unit_number != newunit;
periph = TAILQ_NEXT(periph, unit_links))
;
if (periph != NULL && periph->unit_number == newunit) {
if (wired != 0) {
xpt_print_path(periph->path);
printf("Duplicate Wired Device entry!\n");
xpt_print_path(periph->path);
printf("Second device will not be wired\n");
wired = 0;
}
continue;
}
for (periph_conf = cam_pinit;
wired == 0 && periph_conf->periph_name != NULL;
periph_conf++) {
/*
* Don't match entries like "da 4" as a wired down
* device, but do match entries like "da 4 target 5"
* or even "da 4 scbus 1".
*/
if (IS_SPECIFIED(periph_conf->periph_unit)
&& (!strcmp(periph_name, periph_conf->periph_name))
&& (IS_SPECIFIED(periph_conf->target)
|| IS_SPECIFIED(periph_conf->pathid))
&& (newunit == periph_conf->periph_unit))
break;
}
if (wired != 0 || periph_conf->periph_name == NULL)
break;
}
splx(s);
return (newunit);
}
static u_int
camperiphunit(struct periph_driver *p_drv, path_id_t pathid,
target_id_t target, lun_id_t lun)
{
struct cam_periph_config *periph_conf;
u_int unit;
int hit;
unit = 0;
hit = 0;
for (periph_conf = cam_pinit;
periph_conf->periph_name != NULL;
periph_conf++, hit = 0) {
if (!strcmp(p_drv->driver_name, periph_conf->periph_name)
&& IS_SPECIFIED(periph_conf->periph_unit)) {
if (IS_SPECIFIED(periph_conf->pathid)) {
if (pathid != periph_conf->pathid)
continue;
hit++;
}
if (IS_SPECIFIED(periph_conf->target)) {
if (target != periph_conf->target)
continue;
hit++;
}
if (IS_SPECIFIED(periph_conf->lun)) {
if (lun != periph_conf->lun)
continue;
hit++;
}
if (hit != 0) {
unit = periph_conf->periph_unit;
break;
}
}
}
/*
* Either start from 0 looking for the next unit or from
* the unit number given in the periph_conf. This way,
* if we have wildcard matches, we don't return the same
* unit number twice.
*/
unit = camperiphnextunit(p_drv, unit, /*wired*/hit);
return (unit);
}
void
cam_periph_invalidate(struct cam_periph *periph)
{
int s;
Fix a problem with the way we handled device invalidation when attaching to a device failed. In theory, the same steps that happen when we get an AC_LOST_DEVICE async notification should have been taken when a driver fails to attach. In practice, that wasn't the case. This only affected the da, cd and ch drivers, but the fix affects all peripheral drivers. There were several possible problems: - In the da driver, we didn't remove the peripheral's softc from the da driver's linked list of softcs. Once the peripheral and softc got removed, we'd get a kernel panic the next time the timeout routine called dasendorderedtag(). - In the da, cd and possibly ch drivers, we didn't remove the peripheral's devstat structure from the devstat queue. Once the peripheral and softc were removed, this could cause a panic if anyone tried to access device statistics. (one component of the linked list wouldn't exist anymore) - In the cd driver, we didn't take the peripheral off the changer run queue if it was scheduled to run. In practice, it's highly unlikely, and maybe impossible that the peripheral would have been on the changer run queue at that stage of the probe process. The fix is: - Add a new peripheral callback function (the "oninvalidate" function) that is called the first time cam_periph_invalidate() is called for a peripheral. - Create new foooninvalidate() routines for each peripheral driver. This routine is always called at splsoftcam(), and contains all the stuff that used to be in the AC_LOST_DEVICE case of the async callback handler. - Move the devstat cleanup call to the destructor/cleanup routines, since some of the drivers do I/O in their close routines. - Make sure that when we're flushing the buffer queue, we traverse it at splbio(). - Add a check for the invalid flag in the pt driver's open routine. Reviewed by: gibbs
1998-10-22 22:16:56 +00:00
s = splsoftcam();
/*
* We only call this routine the first time a peripheral is
* invalidated. The oninvalidate() routine is always called at
* splsoftcam().
*/
if (((periph->flags & CAM_PERIPH_INVALID) == 0)
&& (periph->periph_oninval != NULL))
periph->periph_oninval(periph);
periph->flags |= CAM_PERIPH_INVALID;
periph->flags &= ~CAM_PERIPH_NEW_DEV_FOUND;
if (periph->refcount == 0)
camperiphfree(periph);
else if (periph->refcount < 0)
printf("cam_invalidate_periph: refcount < 0!!\n");
splx(s);
}
static void
camperiphfree(struct cam_periph *periph)
{
int s;
struct periph_driver **p_drv;
for (p_drv = (struct periph_driver **)periphdriver_set.ls_items;
*p_drv != NULL; p_drv++) {
if (strcmp((*p_drv)->driver_name, periph->periph_name) == 0)
break;
}
if (periph->periph_dtor != NULL)
periph->periph_dtor(periph);
s = splsoftcam();
TAILQ_REMOVE(&(*p_drv)->units, periph, unit_links);
(*p_drv)->generation++;
splx(s);
xpt_remove_periph(periph);
if (periph->flags & CAM_PERIPH_NEW_DEV_FOUND) {
union ccb ccb;
void *arg;
switch (periph->deferred_ac) {
case AC_FOUND_DEVICE:
ccb.ccb_h.func_code = XPT_GDEV_TYPE;
xpt_setup_ccb(&ccb.ccb_h, periph->path, /*priority*/ 1);
xpt_action(&ccb);
arg = &ccb;
break;
case AC_PATH_REGISTERED:
ccb.ccb_h.func_code = XPT_PATH_INQ;
xpt_setup_ccb(&ccb.ccb_h, periph->path, /*priority*/ 1);
xpt_action(&ccb);
arg = &ccb;
break;
default:
arg = NULL;
break;
}
periph->deferred_callback(NULL, periph->deferred_ac,
periph->path, arg);
}
xpt_free_path(periph->path);
free(periph, M_DEVBUF);
}
/*
* Wait interruptibly for an exclusive lock.
*/
int
cam_periph_lock(struct cam_periph *periph, int priority)
{
int error;
while ((periph->flags & CAM_PERIPH_LOCKED) != 0) {
periph->flags |= CAM_PERIPH_LOCK_WANTED;
if ((error = tsleep(periph, priority, "caplck", 0)) != 0)
return error;
}
if (cam_periph_acquire(periph) != CAM_REQ_CMP)
return(ENXIO);
periph->flags |= CAM_PERIPH_LOCKED;
return 0;
}
/*
* Unlock and wake up any waiters.
*/
void
cam_periph_unlock(struct cam_periph *periph)
{
periph->flags &= ~CAM_PERIPH_LOCKED;
if ((periph->flags & CAM_PERIPH_LOCK_WANTED) != 0) {
periph->flags &= ~CAM_PERIPH_LOCK_WANTED;
wakeup(periph);
}
cam_periph_release(periph);
}
/*
* Map user virtual pointers into kernel virtual address space, so we can
* access the memory. This won't work on physical pointers, for now it's
* up to the caller to check for that. (XXX KDM -- should we do that here
* instead?) This also only works for up to MAXPHYS memory. Since we use
* buffers to map stuff in and out, we're limited to the buffer size.
*/
int
cam_periph_mapmem(union ccb *ccb, struct cam_periph_map_info *mapinfo)
{
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
int numbufs, i;
int flags[CAM_PERIPH_MAXMAPS];
u_int8_t **data_ptrs[CAM_PERIPH_MAXMAPS];
u_int32_t lengths[CAM_PERIPH_MAXMAPS];
u_int32_t dirs[CAM_PERIPH_MAXMAPS];
switch(ccb->ccb_h.func_code) {
case XPT_DEV_MATCH:
if (ccb->cdm.match_buf_len == 0) {
printf("cam_periph_mapmem: invalid match buffer "
"length 0\n");
return(EINVAL);
}
if (ccb->cdm.pattern_buf_len > 0) {
data_ptrs[0] = (u_int8_t **)&ccb->cdm.patterns;
lengths[0] = ccb->cdm.pattern_buf_len;
dirs[0] = CAM_DIR_OUT;
data_ptrs[1] = (u_int8_t **)&ccb->cdm.matches;
lengths[1] = ccb->cdm.match_buf_len;
dirs[1] = CAM_DIR_IN;
numbufs = 2;
} else {
data_ptrs[0] = (u_int8_t **)&ccb->cdm.matches;
lengths[0] = ccb->cdm.match_buf_len;
dirs[0] = CAM_DIR_IN;
numbufs = 1;
}
break;
case XPT_SCSI_IO:
if ((ccb->ccb_h.flags & CAM_DIR_MASK) == CAM_DIR_NONE)
return(0);
data_ptrs[0] = &ccb->csio.data_ptr;
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
lengths[0] = ccb->csio.dxfer_len;
dirs[0] = ccb->ccb_h.flags & CAM_DIR_MASK;
numbufs = 1;
break;
default:
return(EINVAL);
break; /* NOTREACHED */
}
/*
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
* Check the transfer length and permissions first, so we don't
* have to unmap any previously mapped buffers.
*/
for (i = 0; i < numbufs; i++) {
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
flags[i] = 0;
/*
* The userland data pointer passed in may not be page
* aligned. vmapbuf() truncates the address to a page
* boundary, so if the address isn't page aligned, we'll
* need enough space for the given transfer length, plus
* whatever extra space is necessary to make it to the page
* boundary.
*/
if ((lengths[i] +
(((vm_offset_t)(*data_ptrs[i])) & PAGE_MASK)) > DFLTPHYS){
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
printf("cam_periph_mapmem: attempt to map %u bytes, "
"which is greater than DFLTPHYS(%d)\n",
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
lengths[i] +
(((vm_offset_t)(*data_ptrs[i])) & PAGE_MASK),
DFLTPHYS);
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
return(E2BIG);
}
if (dirs[i] & CAM_DIR_IN) {
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
flags[i] = B_READ;
if (useracc(*data_ptrs[i], lengths[i], B_READ) == 0){
printf("cam_periph_mapmem: error, "
"address %p, length %lu isn't "
"user accessible for READ\n",
(void *)*data_ptrs[i],
(u_long)lengths[i]);
return(EACCES);
}
}
/*
* XXX this check is really bogus, since B_WRITE currently
* is all 0's, and so it is "set" all the time.
*/
if (dirs[i] & CAM_DIR_OUT) {
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
flags[i] |= B_WRITE;
if (useracc(*data_ptrs[i], lengths[i], B_WRITE) == 0){
printf("cam_periph_mapmem: error, "
"address %p, length %lu isn't "
"user accessible for WRITE\n",
(void *)*data_ptrs[i],
(u_long)lengths[i]);
return(EACCES);
}
}
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
}
/* this keeps the current process from getting swapped */
/*
* XXX KDM should I use P_NOSWAP instead?
*/
curproc->p_flag |= P_PHYSIO;
for (i = 0; i < numbufs; i++) {
/*
* Get the buffer.
*/
mapinfo->bp[i] = getpbuf();
/* save the buffer's data address */
mapinfo->bp[i]->b_saveaddr = mapinfo->bp[i]->b_data;
/* put our pointer in the data slot */
mapinfo->bp[i]->b_data = *data_ptrs[i];
/* set the transfer length, we know it's < DFLTPHYS */
mapinfo->bp[i]->b_bufsize = lengths[i];
/* set the flags */
Probable fix for the "cdda2wav" panics that various people have been reporting since this past summer. (I think Daniel O'Conner was the first.) The problem appears to have been something like this: - cdda2wav by default passes in a buffer that is close to the 128K MAXPHYS limit. - many times, the buffer is not page aligned - vmapbuf() truncates the address, so that it is page aligned - that causes the total size of the buffer to be greater than MAXPHYS, which of course is a bad thing. Here's a quote from the PR (kern/9067): ================== In particular, note bp->b_bufsize = 0x0001f950 and bp->b_data = 0xf2219960 (which does not start on a page boundary). vunmapbuf() loops through all the pages without any difficulty until addr reaches 0xf2239000, and then the panic occurs. This seems to indicate that we are exceeding MAXPHYS since we actually started from the middle of a page (the data is being transfered to a non page aligned location). To complete the description, note that the system call originates from ReadCddaMMC12() (in scsi_cmds.c of cdda2wav) with a request to read 55 audio sectors of 2352 bytes (which is calculated to fall under MAXPHYS). This in turn ends up calling scsi_send() (in scsi-bsd.c) which calls cam_fill_csio() and cam_send_ccb(). This results in a CAMIOCOMMAND ioctl with a ccb function code of XPT_SCSI_IO. ================== The fix is to change the size check in cam_periph_mapmem() so that it is like the one in minphys(). In particular, it is something like: if ((buffer_length + (buf_ptr & PAGE_MASK)) > MAXPHYS) buffer is too big My fix is based on the one in the PR, but I cleaned up a fair number of things in cam_periph_mapmem(). The checks for each buffer to be mapped are now in a separate loop from the actual mapping operation. With the new arrangement, we don't have to bother with unmapping any previously mapped buffers if one of the checks fails. Many thanks to James Liu for tracking this down. I'd appreciate it if some vm-savvy folks would look this over. I believe this fix is correct, but I could be wrong. PR: kern/9067 (also, kern/8112) Reviewed by: gibbs Submitted by: "James T. Liu" <jtliu@phlebas.rockefeller.edu>
1998-12-16 18:00:39 +00:00
mapinfo->bp[i]->b_flags = flags[i] | B_PHYS | B_BUSY;
/* map the buffer into kernel memory */
vmapbuf(mapinfo->bp[i]);
/* set our pointer to the new mapped area */
*data_ptrs[i] = mapinfo->bp[i]->b_data;
mapinfo->num_bufs_used++;
}
return(0);
}
/*
* Unmap memory segments mapped into kernel virtual address space by
* cam_periph_mapmem().
*/
void
cam_periph_unmapmem(union ccb *ccb, struct cam_periph_map_info *mapinfo)
{
int numbufs, i;
u_int8_t **data_ptrs[CAM_PERIPH_MAXMAPS];
if (mapinfo->num_bufs_used <= 0) {
/* allow ourselves to be swapped once again */
curproc->p_flag &= ~P_PHYSIO;
return;
}
switch (ccb->ccb_h.func_code) {
case XPT_DEV_MATCH:
numbufs = min(mapinfo->num_bufs_used, 2);
if (numbufs == 1) {
data_ptrs[0] = (u_int8_t **)&ccb->cdm.matches;
} else {
data_ptrs[0] = (u_int8_t **)&ccb->cdm.patterns;
data_ptrs[1] = (u_int8_t **)&ccb->cdm.matches;
}
break;
case XPT_SCSI_IO:
data_ptrs[0] = &ccb->csio.data_ptr;
numbufs = min(mapinfo->num_bufs_used, 1);
break;
default:
/* allow ourselves to be swapped once again */
curproc->p_flag &= ~P_PHYSIO;
return;
break; /* NOTREACHED */
}
for (i = 0; i < numbufs; i++) {
/* Set the user's pointer back to the original value */
*data_ptrs[i] = mapinfo->bp[i]->b_saveaddr;
/* unmap the buffer */
vunmapbuf(mapinfo->bp[i]);
/* clear the flags we set above */
mapinfo->bp[i]->b_flags &= ~(B_PHYS|B_BUSY);
/* release the buffer */
relpbuf(mapinfo->bp[i]);
}
/* allow ourselves to be swapped once again */
curproc->p_flag &= ~P_PHYSIO;
}
union ccb *
cam_periph_getccb(struct cam_periph *periph, u_int32_t priority)
{
struct ccb_hdr *ccb_h;
int s;
CAM_DEBUG(periph->path, CAM_DEBUG_TRACE, ("entering cdgetccb\n"));
s = splsoftcam();
while (periph->ccb_list.slh_first == NULL) {
if (periph->immediate_priority > priority)
periph->immediate_priority = priority;
xpt_schedule(periph, priority);
if ((periph->ccb_list.slh_first != NULL)
&& (periph->ccb_list.slh_first->pinfo.priority == priority))
break;
tsleep(&periph->ccb_list, PRIBIO, "cgticb", 0);
}
ccb_h = periph->ccb_list.slh_first;
SLIST_REMOVE_HEAD(&periph->ccb_list, periph_links.sle);
splx(s);
return ((union ccb *)ccb_h);
}
void
cam_periph_ccbwait(union ccb *ccb)
{
int s;
s = splsoftcam();
if ((ccb->ccb_h.pinfo.index != CAM_UNQUEUED_INDEX)
|| ((ccb->ccb_h.status & CAM_STATUS_MASK) == CAM_REQ_INPROG))
tsleep(&ccb->ccb_h.cbfcnp, PRIBIO, "cbwait", 0);
splx(s);
}
int
cam_periph_ioctl(struct cam_periph *periph, int cmd, caddr_t addr,
int (*error_routine)(union ccb *ccb,
cam_flags camflags,
u_int32_t sense_flags))
{
union ccb *ccb;
int error;
int found;
error = found = 0;
switch(cmd){
case CAMGETPASSTHRU:
ccb = cam_periph_getccb(periph, /* priority */ 1);
xpt_setup_ccb(&ccb->ccb_h,
ccb->ccb_h.path,
/*priority*/1);
ccb->ccb_h.func_code = XPT_GDEVLIST;
/*
* Basically, the point of this is that we go through
* getting the list of devices, until we find a passthrough
* device. In the current version of the CAM code, the
* only way to determine what type of device we're dealing
* with is by its name.
*/
while (found == 0) {
ccb->cgdl.index = 0;
ccb->cgdl.status = CAM_GDEVLIST_MORE_DEVS;
while (ccb->cgdl.status == CAM_GDEVLIST_MORE_DEVS) {
/* we want the next device in the list */
xpt_action(ccb);
if (strncmp(ccb->cgdl.periph_name,
"pass", 4) == 0){
found = 1;
break;
}
}
if ((ccb->cgdl.status == CAM_GDEVLIST_LAST_DEVICE) &&
(found == 0)) {
ccb->cgdl.periph_name[0] = '\0';
ccb->cgdl.unit_number = 0;
break;
}
}
/* copy the result back out */
bcopy(ccb, addr, sizeof(union ccb));
/* and release the ccb */
xpt_release_ccb(ccb);
break;
default:
error = ENOTTY;
break;
}
return(error);
}
int
cam_periph_runccb(union ccb *ccb,
int (*error_routine)(union ccb *ccb,
cam_flags camflags,
u_int32_t sense_flags),
cam_flags camflags, u_int32_t sense_flags,
struct devstat *ds)
{
int error;
error = 0;
/*
* If the user has supplied a stats structure, and if we understand
* this particular type of ccb, record the transaction start.
*/
if ((ds != NULL) && (ccb->ccb_h.func_code == XPT_SCSI_IO))
devstat_start_transaction(ds);
xpt_action(ccb);
do {
cam_periph_ccbwait(ccb);
if ((ccb->ccb_h.status & CAM_STATUS_MASK) == CAM_REQ_CMP)
error = 0;
else if (error_routine != NULL)
error = (*error_routine)(ccb, camflags, sense_flags);
else
error = 0;
} while (error == ERESTART);
if ((ccb->ccb_h.status & CAM_DEV_QFRZN) != 0)
cam_release_devq(ccb->ccb_h.path,
/* relsim_flags */0,
/* openings */0,
/* timeout */0,
/* getcount_only */ FALSE);
if ((ds != NULL) && (ccb->ccb_h.func_code == XPT_SCSI_IO))
devstat_end_transaction(ds,
ccb->csio.dxfer_len,
ccb->csio.tag_action & 0xf,
((ccb->ccb_h.flags & CAM_DIR_MASK) ==
CAM_DIR_NONE) ? DEVSTAT_NO_DATA :
(ccb->ccb_h.flags & CAM_DIR_OUT) ?
DEVSTAT_WRITE :
DEVSTAT_READ);
return(error);
}
u_int32_t
cam_release_devq(struct cam_path *path, u_int32_t relsim_flags,
u_int32_t openings, u_int32_t timeout,
int getcount_only)
{
struct ccb_relsim crs;
xpt_setup_ccb(&crs.ccb_h, path,
/*priority*/1);
crs.ccb_h.func_code = XPT_REL_SIMQ;
crs.ccb_h.flags = getcount_only ? CAM_DEV_QFREEZE : 0;
crs.release_flags = relsim_flags;
crs.openings = openings;
crs.release_timeout = timeout;
xpt_action((union ccb *)&crs);
return (crs.qfrozen_cnt);
}
#define saved_ccb_ptr ppriv_ptr0
static void
camperiphdone(struct cam_periph *periph, union ccb *done_ccb)
{
cam_status status;
int frozen;
int sense;
struct scsi_start_stop_unit *scsi_cmd;
u_int32_t relsim_flags, timeout;
u_int32_t qfrozen_cnt;
status = done_ccb->ccb_h.status;
frozen = (status & CAM_DEV_QFRZN) != 0;
sense = (status & CAM_AUTOSNS_VALID) != 0;
status &= CAM_STATUS_MASK;
timeout = 0;
relsim_flags = 0;
/*
* Unfreeze the queue once if it is already frozen..
*/
if (frozen != 0) {
qfrozen_cnt = cam_release_devq(done_ccb->ccb_h.path,
/*relsim_flags*/0,
/*openings*/0,
/*timeout*/0,
/*getcount_only*/0);
}
switch (status) {
case CAM_REQ_CMP:
/*
* If we have successfully taken a device from the not
* ready to ready state, re-scan the device and re-get the
* inquiry information. Many devices (mostly disks) don't
* properly report their inquiry information unless they
* are spun up.
*/
if (done_ccb->ccb_h.func_code == XPT_SCSI_IO) {
scsi_cmd = (struct scsi_start_stop_unit *)
&done_ccb->csio.cdb_io.cdb_bytes;
if (scsi_cmd->opcode == START_STOP_UNIT)
xpt_async(AC_INQ_CHANGED,
done_ccb->ccb_h.path, NULL);
}
bcopy(done_ccb->ccb_h.saved_ccb_ptr, done_ccb,
sizeof(union ccb));
periph->flags &= ~CAM_PERIPH_RECOVERY_INPROG;
xpt_action(done_ccb);
break;
case CAM_SCSI_STATUS_ERROR:
scsi_cmd = (struct scsi_start_stop_unit *)
&done_ccb->csio.cdb_io.cdb_bytes;
if (sense != 0) {
struct scsi_sense_data *sense;
int error_code, sense_key, asc, ascq;
sense = &done_ccb->csio.sense_data;
scsi_extract_sense(sense, &error_code,
&sense_key, &asc, &ascq);
/*
* If the error is "invalid field in CDB",
* and the load/eject flag is set, turn the
* flag off and try again. This is just in
* case the drive in question barfs on the
* load eject flag. The CAM code should set
* the load/eject flag by default for
* removable media.
*/
/* XXX KDM
* Should we check to see what the specific
* scsi status is?? Or does it not matter
* since we already know that there was an
* error, and we know what the specific
* error code was, and we know what the
* opcode is..
*/
if ((scsi_cmd->opcode == START_STOP_UNIT) &&
((scsi_cmd->how & SSS_LOEJ) != 0) &&
(asc == 0x24) && (ascq == 0x00) &&
(done_ccb->ccb_h.retry_count > 0)) {
scsi_cmd->how &= ~SSS_LOEJ;
xpt_action(done_ccb);
} else if (done_ccb->ccb_h.retry_count > 0) {
/*
* In this case, the error recovery
* command failed, but we've got
* some retries left on it. Give
* it another try.
*/
/* set the timeout to .5 sec */
relsim_flags =
RELSIM_RELEASE_AFTER_TIMEOUT;
timeout = 500;
xpt_action(done_ccb);
break;
} else {
/*
* Copy the original CCB back and
* send it back to the caller.
*/
bcopy(done_ccb->ccb_h.saved_ccb_ptr,
done_ccb, sizeof(union ccb));
periph->flags &= ~CAM_PERIPH_RECOVERY_INPROG;
xpt_action(done_ccb);
}
} else {
/*
* Eh?? The command failed, but we don't
* have any sense. What's up with that?
* Fire the CCB again to return it to the
* caller.
*/
bcopy(done_ccb->ccb_h.saved_ccb_ptr,
done_ccb, sizeof(union ccb));
periph->flags &= ~CAM_PERIPH_RECOVERY_INPROG;
xpt_action(done_ccb);
}
break;
default:
bcopy(done_ccb->ccb_h.saved_ccb_ptr, done_ccb,
sizeof(union ccb));
periph->flags &= ~CAM_PERIPH_RECOVERY_INPROG;
xpt_action(done_ccb);
break;
}
/* decrement the retry count */
if (done_ccb->ccb_h.retry_count > 0)
done_ccb->ccb_h.retry_count--;
qfrozen_cnt = cam_release_devq(done_ccb->ccb_h.path,
/*relsim_flags*/relsim_flags,
/*openings*/0,
/*timeout*/timeout,
/*getcount_only*/0);
}
/*
* Generic error handler. Peripheral drivers usually filter
* out the errors that they handle in a unique mannor, then
* call this function.
*/
int
cam_periph_error(union ccb *ccb, cam_flags camflags,
u_int32_t sense_flags, union ccb *save_ccb)
{
cam_status status;
int frozen;
int sense;
int error;
int openings;
int retry;
u_int32_t relsim_flags;
u_int32_t timeout;
status = ccb->ccb_h.status;
frozen = (status & CAM_DEV_QFRZN) != 0;
sense = (status & CAM_AUTOSNS_VALID) != 0;
status &= CAM_STATUS_MASK;
relsim_flags = 0;
switch (status) {
case CAM_REQ_CMP:
/* decrement the number of retries */
retry = ccb->ccb_h.retry_count > 0;
if (retry)
ccb->ccb_h.retry_count--;
error = 0;
break;
case CAM_SCSI_STATUS_ERROR:
switch (ccb->csio.scsi_status) {
case SCSI_STATUS_OK:
case SCSI_STATUS_COND_MET:
case SCSI_STATUS_INTERMED:
case SCSI_STATUS_INTERMED_COND_MET:
error = 0;
break;
case SCSI_STATUS_CMD_TERMINATED:
case SCSI_STATUS_CHECK_COND:
if (sense != 0) {
struct scsi_sense_data *sense;
int error_code, sense_key, asc, ascq;
struct cam_periph *periph;
scsi_sense_action err_action;
struct ccb_getdev cgd;
sense = &ccb->csio.sense_data;
scsi_extract_sense(sense, &error_code,
&sense_key, &asc, &ascq);
periph = xpt_path_periph(ccb->ccb_h.path);
/*
* Grab the inquiry data for this device.
*/
xpt_setup_ccb(&cgd.ccb_h, ccb->ccb_h.path,
/*priority*/ 1);
cgd.ccb_h.func_code = XPT_GDEV_TYPE;
xpt_action((union ccb *)&cgd);
err_action = scsi_error_action(asc, ascq,
&cgd.inq_data);
/*
* Send a Test Unit Ready to the device.
* If the 'many' flag is set, we send 120
* test unit ready commands, one every half
* second. Otherwise, we just send one TUR.
* We only want to do this if the retry
* count has not been exhausted.
*/
if (((err_action & SS_MASK) == SS_TUR)
&& save_ccb != NULL
&& ccb->ccb_h.retry_count > 0) {
/*
* Since error recovery is already
* in progress, don't attempt to
* process this error. It is probably
* related to the error that caused
* the currently active error recovery
* action. Also, we only have
* space for one saved CCB, so if we
* had two concurrent error recovery
* actions, we would end up
* over-writing one error recovery
* CCB with another one.
*/
if (periph->flags &
CAM_PERIPH_RECOVERY_INPROG) {
error = ERESTART;
break;
}
periph->flags |=
CAM_PERIPH_RECOVERY_INPROG;
/* decrement the number of retries */
if ((err_action &
SSQ_DECREMENT_COUNT) != 0) {
retry = 1;
ccb->ccb_h.retry_count--;
}
bcopy(ccb, save_ccb, sizeof(*save_ccb));
/*
* We retry this one every half
* second for a minute. If the
* device hasn't become ready in a
* minute's time, it's unlikely to
* ever become ready. If the table
* doesn't specify SSQ_MANY, we can
* only try this once. Oh well.
*/
if ((err_action & SSQ_MANY) != 0)
scsi_test_unit_ready(&ccb->csio,
/*retries*/120,
camperiphdone,
MSG_SIMPLE_Q_TAG,
SSD_FULL_SIZE,
/*timeout*/5000);
else
scsi_test_unit_ready(&ccb->csio,
/*retries*/1,
camperiphdone,
MSG_SIMPLE_Q_TAG,
SSD_FULL_SIZE,
/*timeout*/5000);
/* release the queue after .5 sec. */
relsim_flags =
RELSIM_RELEASE_AFTER_TIMEOUT;
timeout = 500;
/*
* Drop the priority to 0 so that
* we are the first to execute. Also
* freeze the queue after this command
* is sent so that we can restore the
* old csio and have it queued in the
* proper order before we let normal
* transactions go to the drive.
*/
ccb->ccb_h.pinfo.priority = 0;
ccb->ccb_h.flags |= CAM_DEV_QFREEZE;
/*
* Save a pointer to the original
* CCB in the new CCB.
*/
ccb->ccb_h.saved_ccb_ptr = save_ccb;
error = ERESTART;
}
/*
* Send a start unit command to the device,
* and then retry the command. We only
* want to do this if the retry count has
* not been exhausted. If the user
* specified 0 retries, then we follow
* their request and do not retry.
*/
else if (((err_action & SS_MASK) == SS_START)
&& save_ccb != NULL
&& ccb->ccb_h.retry_count > 0) {
int le;
/*
* Only one error recovery action
* at a time. See above.
*/
if (periph->flags &
CAM_PERIPH_RECOVERY_INPROG) {
error = ERESTART;
break;
}
periph->flags |=
CAM_PERIPH_RECOVERY_INPROG;
/* decrement the number of retries */
retry = 1;
ccb->ccb_h.retry_count--;
/*
* Check for removable media and
* set load/eject flag
* appropriately.
*/
if (SID_IS_REMOVABLE(&cgd.inq_data))
le = TRUE;
else
le = FALSE;
/*
* Attempt to start the drive up.
*
* Save the current ccb so it can
* be restored and retried once the
* drive is started up.
*/
bcopy(ccb, save_ccb, sizeof(*save_ccb));
scsi_start_stop(&ccb->csio,
/*retries*/1,
camperiphdone,
MSG_SIMPLE_Q_TAG,
/*start*/TRUE,
/*load/eject*/le,
/*immediate*/FALSE,
SSD_FULL_SIZE,
/*timeout*/50000);
/*
* Drop the priority to 0 so that
* we are the first to execute. Also
* freeze the queue after this command
* is sent so that we can restore the
* old csio and have it queued in the
* proper order before we let normal
* transactions go to the drive.
*/
ccb->ccb_h.pinfo.priority = 0;
ccb->ccb_h.flags |= CAM_DEV_QFREEZE;
/*
* Save a pointer to the original
* CCB in the new CCB.
*/
ccb->ccb_h.saved_ccb_ptr = save_ccb;
error = ERESTART;
} else if ((sense_flags & SF_RETRY_UA) != 0) {
/*
* XXX KDM this is a *horrible*
* hack.
*/
error = scsi_interpret_sense(ccb,
sense_flags,
&relsim_flags,
&openings,
&timeout,
err_action);
}
/*
* Theoretically, this code should send a
* test unit ready to the given device, and
* if it returns and error, send a start
* unit command. Since we don't yet have
* the capability to do two-command error
* recovery, just send a start unit.
* XXX KDM fix this!
*/
else if (((err_action & SS_MASK) == SS_TURSTART)
&& save_ccb != NULL
&& ccb->ccb_h.retry_count > 0) {
int le;
/*
* Only one error recovery action
* at a time. See above.
*/
if (periph->flags &
CAM_PERIPH_RECOVERY_INPROG) {
error = ERESTART;
break;
}
periph->flags |=
CAM_PERIPH_RECOVERY_INPROG;
/* decrement the number of retries */
retry = 1;
ccb->ccb_h.retry_count--;
/*
* Check for removable media and
* set load/eject flag
* appropriately.
*/
if (SID_IS_REMOVABLE(&cgd.inq_data))
le = TRUE;
else
le = FALSE;
/*
* Attempt to start the drive up.
*
* Save the current ccb so it can
* be restored and retried once the
* drive is started up.
*/
bcopy(ccb, save_ccb, sizeof(*save_ccb));
scsi_start_stop(&ccb->csio,
/*retries*/1,
camperiphdone,
MSG_SIMPLE_Q_TAG,
/*start*/TRUE,
/*load/eject*/le,
/*immediate*/FALSE,
SSD_FULL_SIZE,
/*timeout*/50000);
/* release the queue after .5 sec. */
relsim_flags =
RELSIM_RELEASE_AFTER_TIMEOUT;
timeout = 500;
/*
* Drop the priority to 0 so that
* we are the first to execute. Also
* freeze the queue after this command
* is sent so that we can restore the
* old csio and have it queued in the
* proper order before we let normal
* transactions go to the drive.
*/
ccb->ccb_h.pinfo.priority = 0;
ccb->ccb_h.flags |= CAM_DEV_QFREEZE;
/*
* Save a pointer to the original
* CCB in the new CCB.
*/
ccb->ccb_h.saved_ccb_ptr = save_ccb;
error = ERESTART;
} else {
error = scsi_interpret_sense(ccb,
sense_flags,
&relsim_flags,
&openings,
&timeout,
err_action);
}
} else if (ccb->csio.scsi_status ==
SCSI_STATUS_CHECK_COND) {
/* no point in decrementing the retry count */
panic("cam_periph_error: scsi status of "
"CHECK COND returned but no sense "
"information is availible. "
"Controller should have returned "
"CAM_AUTOSENSE_FAILED");
/* NOTREACHED */
error = EIO;
} else if (ccb->ccb_h.retry_count > 0) {
/*
* XXX KDM shouldn't there be a better
* argument to return??
*/
error = EIO;
} else {
/* decrement the number of retries */
retry = ccb->ccb_h.retry_count > 0;
if (retry)
ccb->ccb_h.retry_count--;
/*
* If it was aborted with no
* clue as to the reason, just
* retry it again.
*/
error = ERESTART;
}
break;
case SCSI_STATUS_QUEUE_FULL:
{
/* no decrement */
struct ccb_getdev cgd;
/*
* First off, find out what the current
* transaction counts are.
*/
xpt_setup_ccb(&cgd.ccb_h,
ccb->ccb_h.path,
/*priority*/1);
cgd.ccb_h.func_code = XPT_GDEV_TYPE;
xpt_action((union ccb *)&cgd);
/*
* If we were the only transaction active, treat
* the QUEUE FULL as if it were a BUSY condition.
*/
if (cgd.dev_active != 0) {
/*
* Reduce the number of openings to
* be 1 less than the amount it took
* to get a queue full bounded by the
* minimum allowed tag count for this
* device.
*/
openings = cgd.dev_active;
if (openings < cgd.mintags)
openings = cgd.mintags;
if (openings < cgd.dev_active+cgd.dev_openings)
relsim_flags = RELSIM_ADJUST_OPENINGS;
else {
/*
* Some devices report queue full for
* temporary resource shortages. For
* this reason, we allow a minimum
* tag count to be entered via a
* quirk entry to prevent the queue
* count on these devices from falling
* to a pessimisticly low value. We
* still wait for the next successful
* completion, however, before queueing
* more transactions to the device.
*/
relsim_flags =
RELSIM_RELEASE_AFTER_CMDCMPLT;
}
timeout = 0;
error = ERESTART;
break;
}
/* FALLTHROUGH */
}
case SCSI_STATUS_BUSY:
/*
* Restart the queue after either another
* command completes or a 1 second timeout.
*/
/*
* XXX KDM ask JTG about this again, do we need to
* be looking at the retry count here?
*/
error = ERESTART;
relsim_flags = RELSIM_RELEASE_AFTER_TIMEOUT
| RELSIM_RELEASE_AFTER_CMDCMPLT;
timeout = 1000;
break;
case SCSI_STATUS_RESERV_CONFLICT:
error = EIO;
break;
default:
error = EIO;
break;
}
break;
case CAM_REQ_CMP_ERR:
case CAM_AUTOSENSE_FAIL:
case CAM_CMD_TIMEOUT:
case CAM_UNEXP_BUSFREE:
case CAM_UNCOR_PARITY:
case CAM_DATA_RUN_ERR:
/* decrement the number of retries */
retry = ccb->ccb_h.retry_count > 0;
if (retry) {
ccb->ccb_h.retry_count--;
error = ERESTART;
} else {
error = EIO;
}
break;
case CAM_UA_ABORT:
case CAM_UA_TERMIO:
case CAM_MSG_REJECT_REC:
/* XXX Don't know that these are correct */
error = EIO;
break;
case CAM_SEL_TIMEOUT:
{
/*
* XXX
* A single selection timeout should not be enough
* to invalidate a device. We should retry for multiple
* seconds assuming this isn't a probe. We'll probably
* need a special flag for that.
*/
#if 0
struct cam_path *newpath;
/* Should we do more if we can't create the path?? */
if (xpt_create_path(&newpath, xpt_path_periph(ccb->ccb_h.path),
xpt_path_path_id(ccb->ccb_h.path),
xpt_path_target_id(ccb->ccb_h.path),
CAM_LUN_WILDCARD) != CAM_REQ_CMP)
break;
/*
* Let peripheral drivers know that this device has gone
* away.
*/
xpt_async(AC_LOST_DEVICE, newpath, NULL);
xpt_free_path(newpath);
#endif
error = ENXIO;
break;
}
case CAM_REQ_INVALID:
case CAM_PATH_INVALID:
case CAM_DEV_NOT_THERE:
case CAM_NO_HBA:
case CAM_PROVIDE_FAIL:
case CAM_REQ_TOO_BIG:
error = EINVAL;
break;
case CAM_SCSI_BUS_RESET:
case CAM_BDR_SENT:
case CAM_REQUEUE_REQ:
/* Unconditional requeue, dammit */
error = ERESTART;
break;
case CAM_RESRC_UNAVAIL:
case CAM_BUSY:
/* timeout??? */
default:
/* decrement the number of retries */
retry = ccb->ccb_h.retry_count > 0;
if (retry) {
ccb->ccb_h.retry_count--;
error = ERESTART;
} else {
/* Check the sense codes */
error = EIO;
}
break;
}
/* Attempt a retry */
if (error == ERESTART || error == 0) {
if (frozen != 0)
ccb->ccb_h.status &= ~CAM_DEV_QFRZN;
if (error == ERESTART)
xpt_action(ccb);
if (frozen != 0) {
cam_release_devq(ccb->ccb_h.path,
relsim_flags,
openings,
timeout,
/*getcount_only*/0);
}
}
return (error);
}