mirror/freebsd
1
0
Fork 0
mirror of https://git.FreeBSD.org/src.git synced 2024-12-26 11:47:31 +00:00
Code Issues Releases Activity
87cfaf0e1f
freebsd/sys/cam/scsi/scsi_da.c

1624 lines
39 KiB
C
Raw Normal View History

SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/*
* Implementation of SCSI Direct Access Peripheral driver for CAM.
*
* Copyright (c) 1997 Justin T. Gibbs.
* 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.
*
Add a facility in the CAM error handling code to retry selection timeouts. If the client requests that the error recovery code retry a selection timeout, it will be retried after half a second. The delay is to give the device time to recover. For most of these drivers, I only added selection timeout retries where they were also retrying unit attention type errors. The sa(4) driver calls saerror() in a number of places, but most of them don't request retrying unit attentions. Also, bump the default minimum CD changer timeout from 2 to 5 seconds and the maximum timeout from 10 to 15 seconds. Some Pioneer changers seem to have trouble with the shorter timeout. Reviewed by: gibbs
1999-05-09 01:25:34 +00:00
* $Id: scsi_da.c,v 1.23 1999/05/07 07:02:59 phk Exp $
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
*/
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
#include "opt_hw_wdog.h"
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/buf.h>
#include <sys/devicestat.h>
#include <sys/dkbad.h>
#include <sys/disklabel.h>
#include <sys/diskslice.h>
#include <sys/malloc.h>
#include <sys/conf.h>
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
#include <machine/cons.h>
#include <machine/md_var.h>
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
#include <vm/vm.h>
#include <vm/vm_prot.h>
#include <vm/pmap.h>
#include <cam/cam.h>
#include <cam/cam_ccb.h>
#include <cam/cam_extend.h>
#include <cam/cam_periph.h>
#include <cam/cam_xpt_periph.h>
#include <cam/scsi/scsi_message.h>
typedef enum {
DA_STATE_PROBE,
DA_STATE_NORMAL
} da_state;
typedef enum {
DA_FLAG_PACK_INVALID = 0x001,
DA_FLAG_NEW_PACK = 0x002,
DA_FLAG_PACK_LOCKED = 0x004,
DA_FLAG_PACK_REMOVABLE = 0x008,
DA_FLAG_TAGGED_QUEUING = 0x010,
DA_FLAG_NEED_OTAG = 0x020,
DA_FLAG_WENT_IDLE = 0x040,
DA_FLAG_RETRY_UA = 0x080,
DA_FLAG_OPEN = 0x100
} da_flags;
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
typedef enum {
DA_Q_NONE = 0x00,
"Fix" a problem with the Quantum Viking. It appears that this drive does not like the 6-byte read and write commands! It returns illegal request, with the field pointer pointing to byte 9 of a 6 byte CDB. In any case, the work around is to put in a quirk mechanism that makes sure that we don't send 6-byte reads or writes to this device. It's rather sad that this is necessary. You'd think that they would be able to get something that basic to work right in their firmware... Reviewed by: gibbs Reported by: Adam McDougall <bsdx@spawnet.com>
1998-12-02 17:35:28 +00:00
DA_Q_NO_SYNC_CACHE = 0x01,
DA_Q_NO_6_BYTE = 0x02
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
} da_quirks;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
typedef enum {
DA_CCB_PROBE = 0x01,
DA_CCB_BUFFER_IO = 0x02,
DA_CCB_WAITING = 0x03,
DA_CCB_DUMP = 0x04,
DA_CCB_TYPE_MASK = 0x0F,
DA_CCB_RETRY_UA = 0x10
} da_ccb_state;
/* Offsets into our private area for storing information */
#define ccb_state ppriv_field0
#define ccb_bp ppriv_ptr1
struct disk_params {
u_int8_t heads;
u_int16_t cylinders;
u_int8_t secs_per_track;
u_int32_t secsize; /* Number of bytes/sector */
u_int32_t sectors; /* total number sectors */
};
struct da_softc {
struct buf_queue_head buf_queue;
struct devstat device_stats;
SLIST_ENTRY(da_softc) links;
LIST_HEAD(, ccb_hdr) pending_ccbs;
da_state state;
da_flags flags;
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
da_quirks quirks;
"Fix" a problem with the Quantum Viking. It appears that this drive does not like the 6-byte read and write commands! It returns illegal request, with the field pointer pointing to byte 9 of a 6 byte CDB. In any case, the work around is to put in a quirk mechanism that makes sure that we don't send 6-byte reads or writes to this device. It's rather sad that this is necessary. You'd think that they would be able to get something that basic to work right in their firmware... Reviewed by: gibbs Reported by: Adam McDougall <bsdx@spawnet.com>
1998-12-02 17:35:28 +00:00
int minimum_cmd_size;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
int ordered_tag_count;
struct disk_params params;
struct diskslices *dk_slices; /* virtual drives */
union ccb saved_ccb;
};
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
struct da_quirk_entry {
struct scsi_inquiry_pattern inq_pat;
da_quirks quirks;
};
static struct da_quirk_entry da_quirk_table[] =
{
{
Add quirk entries to disable the synchronize cache command for Micropolis 2217's (reported by Matthew Jacob in NetBSD PR kern/6027) and Fujitsu M2954's (reported by Tom Jackson). Some of the Fujitsus at least hang when they get a cache sync command. (Others just return illegal request.) Also, make error printing in dashutdown() a little more selective. Don't print any error when the sense key is illegal request. Drives that don't support the synchronize cache command usually return illegal request. Also, make sure the scsi status is check condition before going into scsi_sense_print(). Reviewed by: gibbs
1998-10-12 17:16:47 +00:00
/*
* This particular Fujitsu drive doesn't like the
* synchronize cache command.
* Reported by: Tom Jackson <toj@gorilla.net>
*/
{T_DIRECT, SIP_MEDIA_FIXED, "FUJITSU", "M2954*", "*"},
/*quirks*/ DA_Q_NO_SYNC_CACHE
},
{
/*
* This drive doesn't like the synchronize cache command
* either. Reported by: Matthew Jacob <mjacob@feral.com>
* in NetBSD PR kern/6027, August 24, 1998.
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
*/
Add quirk entries to disable the synchronize cache command for Micropolis 2217's (reported by Matthew Jacob in NetBSD PR kern/6027) and Fujitsu M2954's (reported by Tom Jackson). Some of the Fujitsus at least hang when they get a cache sync command. (Others just return illegal request.) Also, make error printing in dashutdown() a little more selective. Don't print any error when the sense key is illegal request. Drives that don't support the synchronize cache command usually return illegal request. Also, make sure the scsi status is check condition before going into scsi_sense_print(). Reviewed by: gibbs
1998-10-12 17:16:47 +00:00
{T_DIRECT, SIP_MEDIA_FIXED, "MICROP", "2217*", "*"},
/*quirks*/ DA_Q_NO_SYNC_CACHE
Disable cache syncs for a broken NEC drive. Reviewed by: gibbs Submitted by: Blaz Zupan <blaz@gold.amis.net>
1998-10-13 23:34:54 +00:00
},
Add a quirk that disables SYNCHRONIZE CACHE PR: 8882 Obtained from: Hellmuth Michaelis hm@kts.org
1999-01-05 20:43:41 +00:00
{
/*
* This drive doesn't like the synchronize cache command
* either. Reported by: Hellmuth Michaelis (hm@kts.org)
* (PR 8882).
*/
{T_DIRECT, SIP_MEDIA_FIXED, "MICROP", "2112*", "*"},
/*quirks*/ DA_Q_NO_SYNC_CACHE
},
Disable cache syncs for a broken NEC drive. Reviewed by: gibbs Submitted by: Blaz Zupan <blaz@gold.amis.net>
1998-10-13 23:34:54 +00:00
{
/*
* Doesn't like the synchronize cache command.
* Reported by: Blaz Zupan <blaz@gold.amis.net>
*/
{T_DIRECT, SIP_MEDIA_FIXED, "NEC", "D3847*", "*"},
/*quirks*/ DA_Q_NO_SYNC_CACHE
"Fix" a problem with the Quantum Viking. It appears that this drive does not like the 6-byte read and write commands! It returns illegal request, with the field pointer pointing to byte 9 of a 6 byte CDB. In any case, the work around is to put in a quirk mechanism that makes sure that we don't send 6-byte reads or writes to this device. It's rather sad that this is necessary. You'd think that they would be able to get something that basic to work right in their firmware... Reviewed by: gibbs Reported by: Adam McDougall <bsdx@spawnet.com>
1998-12-02 17:35:28 +00:00
},
{
/*
* Doesn't work correctly with 6 byte reads/writes.
* Returns illegal request, and points to byte 9 of the
* 6-byte CDB.
* Reported by: Adam McDougall <bsdx@spawnet.com>
*/
{T_DIRECT, SIP_MEDIA_FIXED, "QUANTUM", "VIKING 4*", "*"},
/*quirks*/ DA_Q_NO_6_BYTE
},
{
/*
* See above.
*/
{T_DIRECT, SIP_MEDIA_FIXED, "QUANTUM", "VIKING 2*", "*"},
/*quirks*/ DA_Q_NO_6_BYTE
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
}
};
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
static d_open_t daopen;
static d_close_t daclose;
static d_strategy_t dastrategy;
static d_ioctl_t daioctl;
static d_dump_t dadump;
static d_psize_t dasize;
static periph_init_t dainit;
static void daasync(void *callback_arg, u_int32_t code,
struct cam_path *path, void *arg);
static periph_ctor_t daregister;
static periph_dtor_t dacleanup;
static periph_start_t dastart;
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
static periph_oninv_t daoninvalidate;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
static void dadone(struct cam_periph *periph,
union ccb *done_ccb);
static int daerror(union ccb *ccb, u_int32_t cam_flags,
u_int32_t sense_flags);
static void daprevent(struct cam_periph *periph, int action);
static void dasetgeom(struct cam_periph *periph,
struct scsi_read_capacity_data * rdcap);
static timeout_t dasendorderedtag;
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
static void dashutdown(int howto, void *arg);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
#ifndef DA_DEFAULT_TIMEOUT
#define DA_DEFAULT_TIMEOUT 60 /* Timeout in seconds */
#endif
/*
* DA_ORDEREDTAG_INTERVAL determines how often, relative
* to the default timeout, we check to see whether an ordered
* tagged transaction is appropriate to prevent simple tag
* starvation. Since we'd like to ensure that there is at least
* 1/2 of the timeout length left for a starved transaction to
* complete after we've sent an ordered tag, we must poll at least
* four times in every timeout period. This takes care of the worst
* case where a starved transaction starts during an interval that
* meets the requirement "don't send an ordered tag" test so it takes
* us two intervals to determine that a tag must be sent.
*/
#ifndef DA_ORDEREDTAG_INTERVAL
#define DA_ORDEREDTAG_INTERVAL 4
#endif
static struct periph_driver dadriver =
{
dainit, "da",
TAILQ_HEAD_INITIALIZER(dadriver.units), /* generation */ 0
};
DATA_SET(periphdriver_set, dadriver);
#define DA_CDEV_MAJOR 13
#define DA_BDEV_MAJOR 4
/* For 2.2-stable support */
#ifndef D_DISK
#define D_DISK 0
#endif
static struct cdevsw da_cdevsw =
{
/*d_open*/ daopen,
/*d_close*/ daclose,
Introduce two functions: physread() and physwrite() and use these directly in *devsw[] rather than the 46 local copies of the same functions. (grog will do the same for vinum when he has time)
1999-05-07 07:03:47 +00:00
/*d_read*/ physread,
/*d_write*/ physwrite,
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/*d_ioctl*/ daioctl,
/*d_stop*/ nostop,
/*d_reset*/ noreset,
/*d_devtotty*/ nodevtotty,
/*d_poll*/ seltrue,
/*d_mmap*/ nommap,
/*d_strategy*/ dastrategy,
/*d_name*/ "da",
/*d_spare*/ NULL,
/*d_maj*/ -1,
/*d_dump*/ dadump,
/*d_psize*/ dasize,
/*d_flags*/ D_DISK,
/*d_maxio*/ 0,
/*b_maj*/ -1
};
static SLIST_HEAD(,da_softc) softc_list;
static struct extend_array *daperiphs;
static int
daopen(dev_t dev, int flags, int fmt, struct proc *p)
{
struct cam_periph *periph;
struct da_softc *softc;
struct disklabel label;
int unit;
int part;
int error;
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
int s;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
unit = dkunit(dev);
part = dkpart(dev);
periph = cam_extend_get(daperiphs, unit);
if (periph == NULL)
return (ENXIO);
softc = (struct da_softc *)periph->softc;
CAM_DEBUG(periph->path, CAM_DEBUG_TRACE,
("daopen: dev=0x%x (unit %d , partition %d)\n", dev,
unit, part));
if ((error = cam_periph_lock(periph, PRIBIO|PCATCH)) != 0) {
return (error); /* error code from tsleep */
}
if ((softc->flags & DA_FLAG_OPEN) == 0) {
if (cam_periph_acquire(periph) != CAM_REQ_CMP)
return(ENXIO);
softc->flags |= DA_FLAG_OPEN;
}
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();
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
if ((softc->flags & DA_FLAG_PACK_INVALID) != 0) {
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
/*
* If any partition is open, although the disk has
* been invalidated, disallow further opens.
*/
if (dsisopen(softc->dk_slices)) {
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
splx(s);
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
cam_periph_unlock(periph);
return (ENXIO);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
}
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
/* Invalidate our pack information. */
dsgone(&softc->dk_slices);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
softc->flags &= ~DA_FLAG_PACK_INVALID;
}
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
splx(s);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/* Do a read capacity */
{
struct scsi_read_capacity_data *rcap;
union ccb *ccb;
rcap = (struct scsi_read_capacity_data *)malloc(sizeof(*rcap),
M_TEMP,
M_WAITOK);
ccb = cam_periph_getccb(periph, /*priority*/1);
scsi_read_capacity(&ccb->csio,
/*retries*/1,
/*cbfncp*/dadone,
MSG_SIMPLE_Q_TAG,
rcap,
SSD_FULL_SIZE,
Up the read capacity timeout from 20 seconds to 60 seconds to keep my JAZ drive happy. This shouldn't impact fast drives, and will keep cam from failing on very slow ones (that are spinning up, say). 20 seconds was almost long enough, but not in all cases. Suggested by: gibbs
1998-10-07 03:09:19 +00:00
/*timeout*/60000);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
ccb->ccb_h.ccb_bp = NULL;
error = cam_periph_runccb(ccb, daerror, /*cam_flags*/0,
Add a facility in the CAM error handling code to retry selection timeouts. If the client requests that the error recovery code retry a selection timeout, it will be retried after half a second. The delay is to give the device time to recover. For most of these drivers, I only added selection timeout retries where they were also retrying unit attention type errors. The sa(4) driver calls saerror() in a number of places, but most of them don't request retrying unit attentions. Also, bump the default minimum CD changer timeout from 2 to 5 seconds and the maximum timeout from 10 to 15 seconds. Some Pioneer changers seem to have trouble with the shorter timeout. Reviewed by: gibbs
1999-05-09 01:25:34 +00:00
/*sense_flags*/SF_RETRY_UA |
SF_RETRY_SELTO,
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
&softc->device_stats);
xpt_release_ccb(ccb);
if (error == 0) {
dasetgeom(periph, rcap);
}
free(rcap, M_TEMP);
}
if (error == 0) {
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
struct ccb_getdev cgd;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/* Build label for whole disk. */
bzero(&label, sizeof(label));
label.d_type = DTYPE_SCSI;
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
/*
* Grab the inquiry data to get the vendor and product names.
* Put them in the typename and packname for the label.
*/
xpt_setup_ccb(&cgd.ccb_h, periph->path, /*priority*/ 1);
cgd.ccb_h.func_code = XPT_GDEV_TYPE;
xpt_action((union ccb *)&cgd);
strncpy(label.d_typename, cgd.inq_data.vendor,
min(SID_VENDOR_SIZE, sizeof(label.d_typename)));
strncpy(label.d_packname, cgd.inq_data.product,
min(SID_PRODUCT_SIZE, sizeof(label.d_packname)));
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
label.d_secsize = softc->params.secsize;
label.d_nsectors = softc->params.secs_per_track;
label.d_ntracks = softc->params.heads;
label.d_ncylinders = softc->params.cylinders;
label.d_secpercyl = softc->params.heads
* softc->params.secs_per_track;
label.d_secperunit = softc->params.sectors;
if ((dsisopen(softc->dk_slices) == 0)
&& ((softc->flags & DA_FLAG_PACK_REMOVABLE) != 0)) {
daprevent(periph, PR_PREVENT);
}
/* Initialize slice tables. */
error = dsopen("da", dev, fmt, 0, &softc->dk_slices, &label,
dastrategy, (ds_setgeom_t *)NULL,
&da_cdevsw);
/*
* Check to see whether or not the blocksize is set yet.
* If it isn't, set it and then clear the blocksize
* unavailable flag for the device statistics.
*/
if ((softc->device_stats.flags & DEVSTAT_BS_UNAVAILABLE) != 0){
softc->device_stats.block_size = softc->params.secsize;
softc->device_stats.flags &= ~DEVSTAT_BS_UNAVAILABLE;
}
}
if (error != 0) {
if ((dsisopen(softc->dk_slices) == 0)
&& ((softc->flags & DA_FLAG_PACK_REMOVABLE) != 0)) {
daprevent(periph, PR_ALLOW);
}
}
cam_periph_unlock(periph);
return (error);
}
static int
daclose(dev_t dev, int flag, int fmt, struct proc *p)
{
struct cam_periph *periph;
struct da_softc *softc;
int unit;
int error;
unit = dkunit(dev);
periph = cam_extend_get(daperiphs, unit);
if (periph == NULL)
return (ENXIO);
softc = (struct da_softc *)periph->softc;
if ((error = cam_periph_lock(periph, PRIBIO)) != 0) {
return (error); /* error code from tsleep */
}
dsclose(dev, fmt, softc->dk_slices);
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
if (dsisopen(softc->dk_slices)) {
cam_periph_unlock(periph);
return (0);
}
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
if ((softc->quirks & DA_Q_NO_SYNC_CACHE) == 0) {
union ccb *ccb;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
ccb = cam_periph_getccb(periph, /*priority*/1);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
scsi_synchronize_cache(&ccb->csio,
/*retries*/1,
/*cbfcnp*/dadone,
MSG_SIMPLE_Q_TAG,
/*begin_lba*/0,/* Cover the whole disk */
/*lb_count*/0,
SSD_FULL_SIZE,
5 * 60 * 1000);
cam_periph_runccb(ccb, /*error_routine*/NULL, /*cam_flags*/0,
you can retry SYNC CACHE on UA errors
1998-12-23 16:48:17 +00:00
/*sense_flags*/SF_RETRY_UA,
&softc->device_stats);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
Silence complaints about synchronize cache requests that fail with illegal request.
1999-03-05 23:20:20 +00:00
if ((ccb->ccb_h.status & CAM_STATUS_MASK) != CAM_REQ_CMP) {
if ((ccb->ccb_h.status & CAM_STATUS_MASK) ==
CAM_SCSI_STATUS_ERROR) {
int asc, ascq;
int sense_key, error_code;
scsi_extract_sense(&ccb->csio.sense_data,
&error_code,
&sense_key,
&asc, &ascq);
if (sense_key != SSD_KEY_ILLEGAL_REQUEST)
scsi_sense_print(&ccb->csio);
} else {
xpt_print_path(periph->path);
printf("Synchronize cache failed, status "
"== 0x%x, scsi status == 0x%x\n",
ccb->csio.ccb_h.status,
ccb->csio.scsi_status);
}
}
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
if ((ccb->ccb_h.status & CAM_DEV_QFRZN) != 0)
cam_release_devq(ccb->ccb_h.path,
/*relsim_flags*/0,
/*reduction*/0,
/*timeout*/0,
/*getcount_only*/0);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
xpt_release_ccb(ccb);
}
Don't leave the device queue in a frozen state if the Synchronize Cache command on close fails.
1998-09-19 04:59:35 +00:00
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
if ((softc->flags & DA_FLAG_PACK_REMOVABLE) != 0) {
daprevent(periph, PR_ALLOW);
/*
* If we've got removeable media, mark the blocksize as
* unavailable, since it could change when new media is
* inserted.
*/
softc->device_stats.flags |= DEVSTAT_BS_UNAVAILABLE;
}
softc->flags &= ~DA_FLAG_OPEN;
cam_periph_unlock(periph);
cam_periph_release(periph);
return (0);
}
/*
* Actually translate the requested transfer into one the physical driver
* can understand. The transfer is described by a buf and will include
* only one physical transfer.
*/
static void
dastrategy(struct buf *bp)
{
struct cam_periph *periph;
struct da_softc *softc;
u_int unit;
u_int part;
int s;
unit = dkunit(bp->b_dev);
part = dkpart(bp->b_dev);
periph = cam_extend_get(daperiphs, unit);
if (periph == NULL) {
bp->b_error = ENXIO;
goto bad;
}
softc = (struct da_softc *)periph->softc;
#if 0
/*
* check it's not too big a transfer for our adapter
*/
scsi_minphys(bp,&sd_switch);
#endif
/*
* Do bounds checking, adjust transfer, set b_cylin and b_pbklno.
*/
if (dscheck(bp, softc->dk_slices) <= 0)
goto done;
/*
* Mask interrupts so that the pack cannot be invalidated until
* after we are in the queue. Otherwise, we might not properly
* clean up one of the buffers.
*/
s = splbio();
/*
* If the device has been made invalid, error out
*/
if ((softc->flags & DA_FLAG_PACK_INVALID)) {
splx(s);
bp->b_error = ENXIO;
goto bad;
}
/*
* Place it in the queue of disk activities for this disk
*/
bufqdisksort(&softc->buf_queue, bp);
splx(s);
/*
* Schedule ourselves for performing the work.
*/
xpt_schedule(periph, /* XXX priority */1);
return;
bad:
bp->b_flags |= B_ERROR;
done:
/*
* Correctly set the buf to indicate a completed xfer
*/
bp->b_resid = bp->b_bcount;
biodone(bp);
return;
}
/* For 2.2-stable support */
#ifndef ENOIOCTL
#define ENOIOCTL -1
#endif
static int
daioctl(dev_t dev, u_long cmd, caddr_t addr, int flag, struct proc *p)
{
struct cam_periph *periph;
struct da_softc *softc;
int unit;
int error;
unit = dkunit(dev);
periph = cam_extend_get(daperiphs, unit);
if (periph == NULL)
return (ENXIO);
softc = (struct da_softc *)periph->softc;
CAM_DEBUG(periph->path, CAM_DEBUG_TRACE, ("daioctl\n"));
if (cmd == DIOCSBAD)
return (EINVAL); /* XXX */
if ((error = cam_periph_lock(periph, PRIBIO|PCATCH)) != 0) {
return (error); /* error code from tsleep */
}
error = dsioctl("da", dev, cmd, addr, flag, &softc->dk_slices,
dastrategy, (ds_setgeom_t *)NULL);
if (error == ENOIOCTL)
error = cam_periph_ioctl(periph, cmd, addr, daerror);
cam_periph_unlock(periph);
return (error);
}
static int
dadump(dev_t dev)
{
Convert dadump to use reasonable data types so that some casting is unecessary.
1998-12-11 03:54:43 +00:00
struct cam_periph *periph;
struct da_softc *softc;
struct disklabel *lp;
u_int unit;
u_int part;
long num; /* number of sectors to write */
long blkoff;
long blknum;
long blkcnt;
vm_offset_t addr;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
static int dadoingadump = 0;
Convert dadump to use reasonable data types so that some casting is unecessary.
1998-12-11 03:54:43 +00:00
struct ccb_scsiio csio;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/* toss any characters present prior to dump */
while (cncheckc() != -1)
;
unit = dkunit(dev);
part = dkpart(dev);
periph = cam_extend_get(daperiphs, unit);
if (periph == NULL) {
return (ENXIO);
}
softc = (struct da_softc *)periph->softc;
if ((softc->flags & DA_FLAG_PACK_INVALID) != 0
|| (softc->dk_slices == NULL)
|| (lp = dsgetlabel(dev, softc->dk_slices)) == NULL)
return (ENXIO);
/* Size of memory to dump, in disk sectors. */
/* XXX Fix up for non DEV_BSIZE sectors!!! */
num = (u_long)Maxmem * PAGE_SIZE / softc->params.secsize;
blkoff = lp->d_partitions[part].p_offset;
blkoff += softc->dk_slices->dss_slices[dkslice(dev)].ds_offset;
/* check transfer bounds against partition size */
if ((dumplo < 0) || ((dumplo + num) > lp->d_partitions[part].p_size))
return (EINVAL);
if (dadoingadump != 0)
return (EFAULT);
dadoingadump = 1;
blknum = dumplo + blkoff;
blkcnt = PAGE_SIZE / softc->params.secsize;
Convert dadump to use reasonable data types so that some casting is unecessary.
1998-12-11 03:54:43 +00:00
addr = 0; /* starting address */
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
while (num > 0) {
Convert dadump to use reasonable data types so that some casting is unecessary.
1998-12-11 03:54:43 +00:00
if (is_physical_memory(addr)) {
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
pmap_enter(kernel_pmap, (vm_offset_t)CADDR1,
Convert dadump to use reasonable data types so that some casting is unecessary.
1998-12-11 03:54:43 +00:00
trunc_page(addr), VM_PROT_READ, TRUE);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
} else {
pmap_enter(kernel_pmap, (vm_offset_t)CADDR1,
trunc_page(0), VM_PROT_READ, TRUE);
}
xpt_setup_ccb(&csio.ccb_h, periph->path, /*priority*/1);
csio.ccb_h.ccb_state = DA_CCB_DUMP;
scsi_read_write(&csio,
/*retries*/1,
dadone,
MSG_ORDERED_Q_TAG,
/*read*/FALSE,
/*byte2*/0,
"Fix" a problem with the Quantum Viking. It appears that this drive does not like the 6-byte read and write commands! It returns illegal request, with the field pointer pointing to byte 9 of a 6 byte CDB. In any case, the work around is to put in a quirk mechanism that makes sure that we don't send 6-byte reads or writes to this device. It's rather sad that this is necessary. You'd think that they would be able to get something that basic to work right in their firmware... Reviewed by: gibbs Reported by: Adam McDougall <bsdx@spawnet.com>
1998-12-02 17:35:28 +00:00
/*minimum_cmd_size*/ softc->minimum_cmd_size,
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
blknum,
blkcnt,
/*data_ptr*/CADDR1,
/*dxfer_len*/blkcnt * softc->params.secsize,
/*sense_len*/SSD_FULL_SIZE,
DA_DEFAULT_TIMEOUT * 1000);
xpt_polled_action((union ccb *)&csio);
if ((csio.ccb_h.status & CAM_STATUS_MASK) != CAM_REQ_CMP) {
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
printf("Aborting dump due to I/O error.\n");
if ((csio.ccb_h.status & CAM_STATUS_MASK) ==
CAM_SCSI_STATUS_ERROR)
scsi_sense_print(&csio);
else
printf("status == 0x%x, scsi status == 0x%x\n",
csio.ccb_h.status, csio.scsi_status);
return(EIO);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
}
Convert dadump to use reasonable data types so that some casting is unecessary.
1998-12-11 03:54:43 +00:00
if (addr % (1024 * 1024) == 0) {
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
#ifdef HW_WDOG
if (wdog_tickler)
(*wdog_tickler)();
#endif /* HW_WDOG */
/* Count in MB of data left to write */
printf("%ld ", (num * softc->params.secsize)
/ (1024 * 1024));
}
/* update block count */
num -= blkcnt;
blknum += blkcnt;
Convert dadump to use reasonable data types so that some casting is unecessary.
1998-12-11 03:54:43 +00:00
addr += blkcnt * softc->params.secsize;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/* operator aborting dump? */
if (cncheckc() != -1)
return (EINTR);
}
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
/*
* Sync the disk cache contents to the physical media.
*/
if ((softc->quirks & DA_Q_NO_SYNC_CACHE) == 0) {
xpt_setup_ccb(&csio.ccb_h, periph->path, /*priority*/1);
csio.ccb_h.ccb_state = DA_CCB_DUMP;
scsi_synchronize_cache(&csio,
/*retries*/1,
/*cbfcnp*/dadone,
MSG_SIMPLE_Q_TAG,
/*begin_lba*/0,/* Cover the whole disk */
/*lb_count*/0,
SSD_FULL_SIZE,
5 * 60 * 1000);
xpt_polled_action((union ccb *)&csio);
if ((csio.ccb_h.status & CAM_STATUS_MASK) != CAM_REQ_CMP) {
if ((csio.ccb_h.status & CAM_STATUS_MASK) ==
Silence complaints about synchronize cache requests that fail with illegal request.
1999-03-05 23:20:20 +00:00
CAM_SCSI_STATUS_ERROR) {
int asc, ascq;
int sense_key, error_code;
scsi_extract_sense(&csio.sense_data,
&error_code,
&sense_key,
&asc, &ascq);
if (sense_key != SSD_KEY_ILLEGAL_REQUEST)
scsi_sense_print(&csio);
} else {
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
xpt_print_path(periph->path);
printf("Synchronize cache failed, status "
"== 0x%x, scsi status == 0x%x\n",
csio.ccb_h.status, csio.scsi_status);
}
}
}
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
return (0);
}
static int
dasize(dev_t dev)
{
struct cam_periph *periph;
struct da_softc *softc;
periph = cam_extend_get(daperiphs, dkunit(dev));
if (periph == NULL)
return (ENXIO);
softc = (struct da_softc *)periph->softc;
return (dssize(dev, &softc->dk_slices, daopen, daclose));
}
static void
dainit(void)
{
cam_status status;
struct cam_path *path;
/*
* Create our extend array for storing the devices we attach to.
*/
daperiphs = cam_extend_new();
SLIST_INIT(&softc_list);
if (daperiphs == NULL) {
printf("da: Failed to alloc extend array!\n");
return;
}
/*
* Install a global async callback. This callback will
* receive async callbacks like "new device found".
*/
status = xpt_create_path(&path, /*periph*/NULL, CAM_XPT_PATH_ID,
CAM_TARGET_WILDCARD, CAM_LUN_WILDCARD);
if (status == CAM_REQ_CMP) {
struct ccb_setasync csa;
xpt_setup_ccb(&csa.ccb_h, path, /*priority*/5);
csa.ccb_h.func_code = XPT_SASYNC_CB;
csa.event_enable = AC_FOUND_DEVICE;
csa.callback = daasync;
csa.callback_arg = NULL;
xpt_action((union ccb *)&csa);
status = csa.ccb_h.status;
xpt_free_path(path);
}
if (status != CAM_REQ_CMP) {
printf("da: Failed to attach master async callback "
"due to status 0x%x!\n", status);
} else {
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
int err;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/* If we were successfull, register our devsw */
cdevsw_add_generic(DA_BDEV_MAJOR, DA_CDEV_MAJOR, &da_cdevsw);
/*
* Schedule a periodic event to occasioanly send an
* ordered tag to a device.
*/
timeout(dasendorderedtag, NULL,
(DA_DEFAULT_TIMEOUT * hz) / DA_ORDEREDTAG_INTERVAL);
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
if ((err = at_shutdown(dashutdown, NULL,
SHUTDOWN_POST_SYNC)) != 0)
printf("dainit: at_shutdown returned %d!\n", err);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
}
}
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
static void
daoninvalidate(struct cam_periph *periph)
{
int s;
struct da_softc *softc;
struct buf *q_bp;
struct ccb_setasync csa;
softc = (struct da_softc *)periph->softc;
/*
* De-register any async callbacks.
*/
xpt_setup_ccb(&csa.ccb_h, periph->path,
/* priority */ 5);
csa.ccb_h.func_code = XPT_SASYNC_CB;
csa.event_enable = 0;
csa.callback = daasync;
csa.callback_arg = periph;
xpt_action((union ccb *)&csa);
softc->flags |= DA_FLAG_PACK_INVALID;
/*
* Although the oninvalidate() routines are always called at
* splsoftcam, we need to be at splbio() here to keep the buffer
* queue from being modified while we traverse it.
*/
s = splbio();
/*
* Return all queued I/O with ENXIO.
* XXX Handle any transactions queued to the card
* with XPT_ABORT_CCB.
*/
while ((q_bp = bufq_first(&softc->buf_queue)) != NULL){
bufq_remove(&softc->buf_queue, q_bp);
q_bp->b_resid = q_bp->b_bcount;
q_bp->b_error = ENXIO;
q_bp->b_flags |= B_ERROR;
biodone(q_bp);
}
splx(s);
SLIST_REMOVE(&softc_list, softc, da_softc, links);
xpt_print_path(periph->path);
printf("lost device\n");
}
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
static void
dacleanup(struct cam_periph *periph)
{
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
struct da_softc *softc;
softc = (struct da_softc *)periph->softc;
devstat_remove_entry(&softc->device_stats);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
cam_extend_release(daperiphs, periph->unit_number);
xpt_print_path(periph->path);
printf("removing device entry\n");
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
free(softc, M_DEVBUF);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
}
static void
daasync(void *callback_arg, u_int32_t code,
struct cam_path *path, void *arg)
{
struct cam_periph *periph;
periph = (struct cam_periph *)callback_arg;
switch (code) {
case AC_FOUND_DEVICE:
{
struct ccb_getdev *cgd;
cam_status status;
cgd = (struct ccb_getdev *)arg;
if ((cgd->pd_type != T_DIRECT) && (cgd->pd_type != T_OPTICAL))
break;
/*
* Allocate a peripheral instance for
* this device and start the probe
* process.
*/
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
status = cam_periph_alloc(daregister, daoninvalidate,
dacleanup, dastart,
"da", CAM_PERIPH_BIO,
cgd->ccb_h.path, daasync,
AC_FOUND_DEVICE, cgd);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
if (status != CAM_REQ_CMP
&& status != CAM_REQ_INPROG)
printf("daasync: Unable to attach to new device "
"due to status 0x%x\n", status);
break;
}
case AC_LOST_DEVICE:
cam_periph_invalidate(periph);
break;
case AC_SENT_BDR:
case AC_BUS_RESET:
{
struct da_softc *softc;
struct ccb_hdr *ccbh;
int s;
softc = (struct da_softc *)periph->softc;
s = splsoftcam();
/*
* Don't fail on the expected unit attention
* that will occur.
*/
softc->flags |= DA_FLAG_RETRY_UA;
for (ccbh = LIST_FIRST(&softc->pending_ccbs);
ccbh != NULL; ccbh = LIST_NEXT(ccbh, periph_links.le))
ccbh->ccb_state |= DA_CCB_RETRY_UA;
splx(s);
break;
}
case AC_TRANSFER_NEG:
case AC_SCSI_AEN:
case AC_UNSOL_RESEL:
default:
break;
}
}
static cam_status
daregister(struct cam_periph *periph, void *arg)
{
int s;
struct da_softc *softc;
struct ccb_setasync csa;
struct ccb_getdev *cgd;
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
caddr_t match;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
cgd = (struct ccb_getdev *)arg;
if (periph == NULL) {
printf("daregister: periph was NULL!!\n");
return(CAM_REQ_CMP_ERR);
}
if (cgd == NULL) {
printf("daregister: no getdev CCB, can't register device\n");
return(CAM_REQ_CMP_ERR);
}
softc = (struct da_softc *)malloc(sizeof(*softc),M_DEVBUF,M_NOWAIT);
if (softc == NULL) {
printf("daregister: Unable to probe new device. "
"Unable to allocate softc\n");
return(CAM_REQ_CMP_ERR);
}
bzero(softc, sizeof(*softc));
LIST_INIT(&softc->pending_ccbs);
softc->state = DA_STATE_PROBE;
bufq_init(&softc->buf_queue);
if (SID_IS_REMOVABLE(&cgd->inq_data))
softc->flags |= DA_FLAG_PACK_REMOVABLE;
if ((cgd->inq_data.flags & SID_CmdQue) != 0)
softc->flags |= DA_FLAG_TAGGED_QUEUING;
periph->softc = softc;
cam_extend_set(daperiphs, periph->unit_number, periph);
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
/*
* See if this device has any quirks.
*/
match = cam_quirkmatch((caddr_t)&cgd->inq_data,
(caddr_t)da_quirk_table,
sizeof(da_quirk_table)/sizeof(*da_quirk_table),
sizeof(*da_quirk_table), scsi_inquiry_match);
if (match != NULL)
softc->quirks = ((struct da_quirk_entry *)match)->quirks;
else
softc->quirks = DA_Q_NONE;
"Fix" a problem with the Quantum Viking. It appears that this drive does not like the 6-byte read and write commands! It returns illegal request, with the field pointer pointing to byte 9 of a 6 byte CDB. In any case, the work around is to put in a quirk mechanism that makes sure that we don't send 6-byte reads or writes to this device. It's rather sad that this is necessary. You'd think that they would be able to get something that basic to work right in their firmware... Reviewed by: gibbs Reported by: Adam McDougall <bsdx@spawnet.com>
1998-12-02 17:35:28 +00:00
if (softc->quirks & DA_Q_NO_6_BYTE)
softc->minimum_cmd_size = 10;
else
softc->minimum_cmd_size = 6;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/*
* Block our timeout handler while we
* add this softc to the dev list.
*/
s = splsoftclock();
SLIST_INSERT_HEAD(&softc_list, softc, links);
splx(s);
/*
* The DA driver supports a blocksize, but
* we don't know the blocksize until we do
* a read capacity. So, set a flag to
* indicate that the blocksize is
* unavailable right now. We'll clear the
* flag as soon as we've done a read capacity.
*/
devstat_add_entry(&softc->device_stats, "da",
periph->unit_number, 0,
DEVSTAT_BS_UNAVAILABLE,
Add a prioritization field to the devstat_add_entry() call so that peripheral drivers can determine where in the devstat(9) list they are inserted. This requires recompilation of libdevstat, systat, vmstat, rpc.rstatd, and any ports that depend on the devstat code, since the size of the devstat structure has changed. The devstat version number has been incremented as well to reflect the change. This sorts devices in the devstat list in "more interesting" to "less interesting" order. So, for instance, da devices are now more important than floppy drives, and so will appear before floppy drives in the default output from systat, iostat, vmstat, etc. The order of devices is, for now, kept in a central table in devicestat.h. If individual drivers were able to make a meaningful decision on what priority they should be at attach time, we could consider splitting the priority information out into the various drivers. For now, though, they have no way of knowing that, so it's easier to put them in an easy to find table. Also, move the checkversion() call in vmstat(8) to a more logical place. Thanks to Bruce and David O'Brien for suggestions, for reviewing this, and for putting up with the long time it has taken me to commit it. Bruce did object somewhat to the central priority table (he would rather the priorities be distributed in each driver), so his objection is duly noted here. Reviewed by: bde, obrien
1999-02-10 00:04:13 +00:00
cgd->pd_type | DEVSTAT_TYPE_IF_SCSI,
DEVSTAT_PRIORITY_DA);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/*
* Add async callbacks for bus reset and
* bus device reset calls. I don't bother
* checking if this fails as, in most cases,
* the system will function just fine without
* them and the only alternative would be to
* not attach the device on failure.
*/
xpt_setup_ccb(&csa.ccb_h, periph->path, /*priority*/5);
csa.ccb_h.func_code = XPT_SASYNC_CB;
csa.event_enable = AC_SENT_BDR | AC_BUS_RESET | AC_LOST_DEVICE;
csa.callback = daasync;
csa.callback_arg = periph;
xpt_action((union ccb *)&csa);
/*
* Lock this peripheral until we are setup.
* This first call can't block
*/
(void)cam_periph_lock(periph, PRIBIO);
xpt_schedule(periph, /*priority*/5);
return(CAM_REQ_CMP);
}
static void
dastart(struct cam_periph *periph, union ccb *start_ccb)
{
struct da_softc *softc;
softc = (struct da_softc *)periph->softc;
switch (softc->state) {
case DA_STATE_NORMAL:
{
/* Pull a buffer from the queue and get going on it */
struct buf *bp;
int s;
/*
* See if there is a buf with work for us to do..
*/
s = splbio();
bp = bufq_first(&softc->buf_queue);
if (periph->immediate_priority <= periph->pinfo.priority) {
CAM_DEBUG_PRINT(CAM_DEBUG_SUBTRACE,
("queuing for immediate ccb\n"));
start_ccb->ccb_h.ccb_state = DA_CCB_WAITING;
SLIST_INSERT_HEAD(&periph->ccb_list, &start_ccb->ccb_h,
periph_links.sle);
periph->immediate_priority = CAM_PRIORITY_NONE;
splx(s);
wakeup(&periph->ccb_list);
} else if (bp == NULL) {
splx(s);
xpt_release_ccb(start_ccb);
} else {
int oldspl;
u_int8_t tag_code;
bufq_remove(&softc->buf_queue, bp);
devstat_start_transaction(&softc->device_stats);
if ((bp->b_flags & B_ORDERED) != 0
|| (softc->flags & DA_FLAG_NEED_OTAG) != 0) {
softc->flags &= ~DA_FLAG_NEED_OTAG;
softc->ordered_tag_count++;
tag_code = MSG_ORDERED_Q_TAG;
} else {
tag_code = MSG_SIMPLE_Q_TAG;
}
scsi_read_write(&start_ccb->csio,
/*retries*/4,
dadone,
tag_code,
bp->b_flags & B_READ,
/*byte2*/0,
"Fix" a problem with the Quantum Viking. It appears that this drive does not like the 6-byte read and write commands! It returns illegal request, with the field pointer pointing to byte 9 of a 6 byte CDB. In any case, the work around is to put in a quirk mechanism that makes sure that we don't send 6-byte reads or writes to this device. It's rather sad that this is necessary. You'd think that they would be able to get something that basic to work right in their firmware... Reviewed by: gibbs Reported by: Adam McDougall <bsdx@spawnet.com>
1998-12-02 17:35:28 +00:00
softc->minimum_cmd_size,
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
bp->b_pblkno,
bp->b_bcount / softc->params.secsize,
bp->b_data,
bp->b_bcount,
/*sense_len*/SSD_FULL_SIZE,
DA_DEFAULT_TIMEOUT * 1000);
start_ccb->ccb_h.ccb_state = DA_CCB_BUFFER_IO;
/*
* Block out any asyncronous callbacks
* while we touch the pending ccb list.
*/
oldspl = splcam();
LIST_INSERT_HEAD(&softc->pending_ccbs,
&start_ccb->ccb_h, periph_links.le);
splx(oldspl);
/* We expect a unit attention from this device */
if ((softc->flags & DA_FLAG_RETRY_UA) != 0) {
start_ccb->ccb_h.ccb_state |= DA_CCB_RETRY_UA;
softc->flags &= ~DA_FLAG_RETRY_UA;
}
start_ccb->ccb_h.ccb_bp = bp;
bp = bufq_first(&softc->buf_queue);
splx(s);
xpt_action(start_ccb);
}
if (bp != NULL) {
/* Have more work to do, so ensure we stay scheduled */
xpt_schedule(periph, /* XXX priority */1);
}
break;
}
case DA_STATE_PROBE:
{
struct ccb_scsiio *csio;
struct scsi_read_capacity_data *rcap;
rcap = (struct scsi_read_capacity_data *)malloc(sizeof(*rcap),
M_TEMP,
M_NOWAIT);
if (rcap == NULL) {
printf("dastart: Couldn't malloc read_capacity data\n");
/* da_free_periph??? */
break;
}
csio = &start_ccb->csio;
scsi_read_capacity(csio,
/*retries*/4,
dadone,
MSG_SIMPLE_Q_TAG,
rcap,
SSD_FULL_SIZE,
/*timeout*/5000);
start_ccb->ccb_h.ccb_bp = NULL;
start_ccb->ccb_h.ccb_state = DA_CCB_PROBE;
xpt_action(start_ccb);
break;
}
}
}
static void
dadone(struct cam_periph *periph, union ccb *done_ccb)
{
struct da_softc *softc;
struct ccb_scsiio *csio;
softc = (struct da_softc *)periph->softc;
csio = &done_ccb->csio;
switch (csio->ccb_h.ccb_state & DA_CCB_TYPE_MASK) {
case DA_CCB_BUFFER_IO:
{
struct buf *bp;
int oldspl;
bp = (struct buf *)done_ccb->ccb_h.ccb_bp;
if ((done_ccb->ccb_h.status & CAM_STATUS_MASK) != CAM_REQ_CMP) {
int error;
int s;
int sf;
if ((csio->ccb_h.ccb_state & DA_CCB_RETRY_UA) != 0)
sf = SF_RETRY_UA;
else
sf = 0;
Add a facility in the CAM error handling code to retry selection timeouts. If the client requests that the error recovery code retry a selection timeout, it will be retried after half a second. The delay is to give the device time to recover. For most of these drivers, I only added selection timeout retries where they were also retrying unit attention type errors. The sa(4) driver calls saerror() in a number of places, but most of them don't request retrying unit attentions. Also, bump the default minimum CD changer timeout from 2 to 5 seconds and the maximum timeout from 10 to 15 seconds. Some Pioneer changers seem to have trouble with the shorter timeout. Reviewed by: gibbs
1999-05-09 01:25:34 +00:00
/* Retry selection timeouts */
sf |= SF_RETRY_SELTO;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
if ((error = daerror(done_ccb, 0, sf)) == ERESTART) {
/*
* A retry was scheuled, so
* just return.
*/
return;
}
if (error != 0) {
struct buf *q_bp;
s = splbio();
if (error == ENXIO) {
/*
* Catastrophic error. Mark our pack as
* invalid.
*/
/* XXX See if this is really a media
* change first.
*/
xpt_print_path(periph->path);
printf("Invalidating pack\n");
softc->flags |= DA_FLAG_PACK_INVALID;
}
/*
* return all queued I/O with EIO, so that
* the client can retry these I/Os in the
* proper order should it attempt to recover.
*/
while ((q_bp = bufq_first(&softc->buf_queue))
!= NULL) {
bufq_remove(&softc->buf_queue, q_bp);
q_bp->b_resid = q_bp->b_bcount;
q_bp->b_error = EIO;
q_bp->b_flags |= B_ERROR;
biodone(q_bp);
}
splx(s);
bp->b_error = error;
bp->b_resid = bp->b_bcount;
bp->b_flags |= B_ERROR;
} else {
bp->b_resid = csio->resid;
bp->b_error = 0;
if (bp->b_resid != 0) {
/* Short transfer ??? */
bp->b_flags |= B_ERROR;
}
}
if ((done_ccb->ccb_h.status & CAM_DEV_QFRZN) != 0)
cam_release_devq(done_ccb->ccb_h.path,
/*relsim_flags*/0,
/*reduction*/0,
/*timeout*/0,
/*getcount_only*/0);
} else {
bp->b_resid = csio->resid;
if (csio->resid > 0)
bp->b_flags |= B_ERROR;
}
/*
* Block out any asyncronous callbacks
* while we touch the pending ccb list.
*/
oldspl = splcam();
LIST_REMOVE(&done_ccb->ccb_h, periph_links.le);
splx(oldspl);
devstat_end_transaction(&softc->device_stats,
bp->b_bcount - bp->b_resid,
done_ccb->csio.tag_action & 0xf,
(bp->b_flags & B_READ) ? DEVSTAT_READ
: DEVSTAT_WRITE);
if (softc->device_stats.busy_count == 0)
softc->flags |= DA_FLAG_WENT_IDLE;
biodone(bp);
break;
}
case DA_CCB_PROBE:
{
struct scsi_read_capacity_data *rdcap;
char announce_buf[80];
rdcap = (struct scsi_read_capacity_data *)csio->data_ptr;
if ((csio->ccb_h.status & CAM_STATUS_MASK) == CAM_REQ_CMP) {
struct disk_params *dp;
dasetgeom(periph, rdcap);
dp = &softc->params;
Examine all occurrences of sprintf(), strcat(), and str[n]cpy() for possible buffer overflow problems. Replaced most sprintf()'s with snprintf(); for others cases, added terminating NUL bytes where appropriate, replaced constants like "16" with sizeof(), etc. These changes include several bug fixes, but most changes are for maintainability's sake. Any instance where it wasn't "immediately obvious" that a buffer overflow could not occur was made safer. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Matthew Dillon <dillon@apollo.backplane.com> Reviewed by: Mike Spengler <mks@networkcs.com>
1998-12-04 22:54:57 +00:00
snprintf(announce_buf, sizeof(announce_buf),
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
"%ldMB (%d %d byte sectors: %dH %dS/T %dC)",
dp->sectors / ((1024L * 1024L) / dp->secsize),
dp->sectors, dp->secsize, dp->heads,
dp->secs_per_track, dp->cylinders);
} else {
int error;
/*
* Retry any UNIT ATTENTION type errors. They
* are expected at boot.
*/
Add a facility in the CAM error handling code to retry selection timeouts. If the client requests that the error recovery code retry a selection timeout, it will be retried after half a second. The delay is to give the device time to recover. For most of these drivers, I only added selection timeout retries where they were also retrying unit attention type errors. The sa(4) driver calls saerror() in a number of places, but most of them don't request retrying unit attentions. Also, bump the default minimum CD changer timeout from 2 to 5 seconds and the maximum timeout from 10 to 15 seconds. Some Pioneer changers seem to have trouble with the shorter timeout. Reviewed by: gibbs
1999-05-09 01:25:34 +00:00
error = daerror(done_ccb, 0, SF_RETRY_UA |
SF_RETRY_SELTO | SF_NO_PRINT);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
if (error == ERESTART) {
/*
* A retry was scheuled, so
* just return.
*/
return;
} else if (error != 0) {
struct scsi_sense_data *sense;
int asc, ascq;
int sense_key, error_code;
int have_sense;
cam_status status;
struct ccb_getdev cgd;
/* Don't wedge this device's queue */
cam_release_devq(done_ccb->ccb_h.path,
/*relsim_flags*/0,
/*reduction*/0,
/*timeout*/0,
/*getcount_only*/0);
status = done_ccb->ccb_h.status;
xpt_setup_ccb(&cgd.ccb_h,
done_ccb->ccb_h.path,
/* priority */ 1);
cgd.ccb_h.func_code = XPT_GDEV_TYPE;
xpt_action((union ccb *)&cgd);
if (((csio->ccb_h.flags & CAM_SENSE_PHYS) != 0)
|| ((csio->ccb_h.flags & CAM_SENSE_PTR) != 0)
|| ((status & CAM_AUTOSNS_VALID) == 0))
have_sense = FALSE;
else
have_sense = TRUE;
if (have_sense) {
sense = &csio->sense_data;
scsi_extract_sense(sense, &error_code,
&sense_key,
&asc, &ascq);
}
/*
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
* Attach to anything that claims to be a
* direct access or optical disk device,
* as long as it doesn't return a "Logical
* unit not supported" (0x25) error.
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
*/
Add a number of interrelated CAM feature enhancements and bug fixes. NOTE: These changes will require recompilation of any userland applications, like cdrecord, xmcd, etc., that use the CAM passthrough interface. A make world is recommended. camcontrol.[c8]: - We now support two new commands, "tags" and "negotiate". - The tags commands allows users to view the number of tagged openings for a device as well as a number of other related parameters, and it allows users to set tagged openings for a device. - The negotiate command allows users to enable and disable disconnection and tagged queueing, set sync rates, offsets and bus width. Note that not all of those features are available for all controllers. Only the adv, ahc, and ncr drivers fully support all of the features at this point. Some cards do not allow the setting of sync rates, offsets and the like, and some of the drivers don't have any facilities to do so. Some drivers, like the adw driver, only support enabling or disabling sync negotiation, but do not support setting sync rates. - new description in the camcontrol man page of how to format a disk - cleanup of the camcontrol inquiry command - add support in the 'devlist' command for skipping unconfigured devices if -v was not specified on the command line. - make use of the new base_transfer_speed in the path inquiry CCB. - fix CCB bzero cases cam_xpt.c, cam_sim.[ch], cam_ccb.h: - new flags on many CCB function codes to designate whether they're non-immediate, use a user-supplied CCB, and can only be passed from userland programs via the xpt device. Use these flags in the transport layer and pass driver to categorize CCBs. - new flag in the transport layer device matching code for device nodes that indicates whether a device is unconfigured - bump the CAM version from 0x10 to 0x11 - Change the CAM ioctls to use the version as their group code, so we can force users to recompile code even when the CCB size doesn't change. - add + fill in a new value in the path inquiry CCB, base_transfer_speed. Remove a corresponding field from the cam_sim structure, and add code to every SIM to set this field to the proper value. - Fix the set transfer settings code in the transport layer. scsi_cd.c: - make some variables volatile instead of just casting them in various places - fix a race condition in the changer code - attach unless we get a "logical unit not supported" error. This should fix all of the cases where people have devices that return weird errors when they don't have media in the drive. scsi_da.c: - attach unless we get a "logical unit not supported" error scsi_pass.c: - for immediate CCBs, just malloc a CCB to send the user request in. This gets rid of the 'held' count problem in camcontrol tags. scsi_pass.h: - change the CAM ioctls to use the CAM version as their group code. adv driver: - Allow changing the sync rate and offset separately. adw driver - Allow changing the sync rate and offset separately. aha driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. ahc driver: - Allow setting offset and sync rate separately bt driver: - Don't return CAM_REQ_CMP for SET_TRAN_SETTINGS CCBs. NCR driver: - Fix the ultra/ultra 2 negotiation bug - allow setting both the sync rate and offset separately Other HBA drivers: - Put code in to set the base_transfer_speed field for XPT_GET_TRAN_SETTINGS CCBs. Reviewed by: gibbs, mjacob (isp), imp (aha)
1999-05-06 20:16:39 +00:00
if ((have_sense) && (asc != 0x25)
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
&& (error_code == SSD_CURRENT_ERROR))
Examine all occurrences of sprintf(), strcat(), and str[n]cpy() for possible buffer overflow problems. Replaced most sprintf()'s with snprintf(); for others cases, added terminating NUL bytes where appropriate, replaced constants like "16" with sizeof(), etc. These changes include several bug fixes, but most changes are for maintainability's sake. Any instance where it wasn't "immediately obvious" that a buffer overflow could not occur was made safer. Reviewed by: Bruce Evans <bde@zeta.org.au> Reviewed by: Matthew Dillon <dillon@apollo.backplane.com> Reviewed by: Mike Spengler <mks@networkcs.com>
1998-12-04 22:54:57 +00:00
snprintf(announce_buf,
sizeof(announce_buf),
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
"Attempt to query device "
"size failed: %s, %s",
scsi_sense_key_text[sense_key],
scsi_sense_desc(asc,ascq,
&cgd.inq_data));
else {
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
if (have_sense)
scsi_sense_print(
&done_ccb->csio);
else {
xpt_print_path(periph->path);
printf("got CAM status %#x\n",
done_ccb->ccb_h.status);
}
A better fix to avoid race conditions between failed probes and peripheral removal. Obtained from:gibbs@freebsd.org
1999-01-07 20:19:09 +00:00
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
xpt_print_path(periph->path);
printf("fatal error, failed"
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
" to attach to device\n");
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
/*
* Free up resources.
*/
cam_periph_invalidate(periph);
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
announce_buf[0] = '\0';
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
}
}
}
free(rdcap, M_TEMP);
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
if (announce_buf[0] != '\0')
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
xpt_announce_periph(periph, announce_buf);
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
softc->state = DA_STATE_NORMAL;
A better fix to avoid race conditions between failed probes and peripheral removal. Obtained from:gibbs@freebsd.org
1999-01-07 20:19:09 +00:00
/*
* Since our peripheral may be invalidated by an error
* above or an external event, we must release our CCB
* before releasing the probe lock on the peripheral.
* The peripheral will only go away once the last lock
* is removed, and we need it around for the CCB release
* operation.
*/
xpt_release_ccb(done_ccb);
Some fixes for the CD and DA drivers from bde. (and some of my own as well) Among them: [ cd driver ] 1. Old labeling code was still there. 2. Error handling for dsopen() was broken (no test for the `error' returned by dsopen(); bogus test of an `error' that is known to be 0). 3. cdopen() closed the physical device after certain errors although there may still be open partitions on it. 4. cdclose() closed the physical device although there may still be open partitions on it. 5. Some printf format fixes was incomplete or missing. 6. cdioctl() truncated unit numbers mod 256. 7. cdioctl() was missing locking. [ da driver ] 1. daclose() closed the physical device although there may still be open partitions on it. This was fixed many years ago in sd.c rev.1.57. 2. A minor optimization (the dk_slices != NULL test) in sdopen() became uglier in daopen(). It is not worth doing. da only regressed compared with od and my version of sd, since I never committed the change to sd. daopen() should probably do less if some partition is already open. This is not addressed by the diffs. [ ... ] 5. "opt_hw_wdog.h" was not included, so the HW_WDOG code was unreachable. - Added a getdev CCB call in the cdopen() and daopen() calls so that the vendor name and device name are available for the disklabel. (suggested by bde) - Removed vestigal devfs support in both drivers, since we can't properly work with devfs yet. (ask bde for details on this) - Cleaned up the probe code in both drivers in the failure cases. There were a number of things wrong here. The peripheral driver instances weren't getting properly cleaned up. Sometimes the wrong probe message would get printed out (with the failure message appended), so it wasn't very clear that we failed to attach. SCSI sense information was printed, even when the error in question wasn't a SCSI error. I put similar fixes into the changer driver in revision 1.2 of scsi_ch.c. Reviewed by: gibbs Submitted by: bde (partially)
1998-10-07 02:57:57 +00:00
cam_periph_unlock(periph);
A better fix to avoid race conditions between failed probes and peripheral removal. Obtained from:gibbs@freebsd.org
1999-01-07 20:19:09 +00:00
return;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
}
case DA_CCB_WAITING:
{
/* Caller will release the CCB */
wakeup(&done_ccb->ccb_h.cbfcnp);
return;
}
case DA_CCB_DUMP:
/* No-op. We're polling */
return;
A better fix to avoid race conditions between failed probes and peripheral removal. Obtained from:gibbs@freebsd.org
1999-01-07 20:19:09 +00:00
default:
break;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
}
A better fix to avoid race conditions between failed probes and peripheral removal. Obtained from:gibbs@freebsd.org
1999-01-07 20:19:09 +00:00
xpt_release_ccb(done_ccb);
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
}
static int
daerror(union ccb *ccb, u_int32_t cam_flags, u_int32_t sense_flags)
{
struct da_softc *softc;
struct cam_periph *periph;
periph = xpt_path_periph(ccb->ccb_h.path);
softc = (struct da_softc *)periph->softc;
Don't invalidate devices due to unexpected unit attention errors. In a perfect world, we'd notice the UA and do some device validation to ensure that the device hasn't changed. We may get this before the year ends, but not before 3.0R. This change gives the adminstrator ample ammunition to take off a foot or two, but hey this *is* UN*X.
1998-09-20 07:17:11 +00:00
/*
* XXX
* Until we have a better way of doing pack validation,
* don't treat UAs as errors.
*/
sense_flags |= SF_RETRY_UA;
SCSI Peripheral drivers for CAM: da - Direct Access Devices (disks, optical devices, SS disks) cd - CDROM (or devices that can act like them, WORM, CD-RW, etc) ch - Medium Changer devices. sa - Sequential Access Devices (tape drives) pass - Application pass-thru driver targ - Target Mode "Processor Target" Emulator pt - Processor Target Devices (scanners, cpus, etc.) Submitted by: The CAM Team
1998-09-15 06:36:34 +00:00
return(cam_periph_error(ccb, cam_flags, sense_flags,
&softc->saved_ccb));
}
static void
daprevent(struct cam_periph *periph, int action)
{
struct da_softc *softc;
union ccb *ccb;
int error;
softc = (struct da_softc *)periph->softc;
if (((action == PR_ALLOW)
&& (softc->flags & DA_FLAG_PACK_LOCKED) == 0)
|| ((action == PR_PREVENT)
&& (softc->flags & DA_FLAG_PACK_LOCKED) != 0)) {
return;
}
ccb = cam_periph_getccb(periph, /*priority*/1);
scsi_prevent(&ccb->csio,
/*retries*/1,
/*cbcfp*/dadone,
MSG_SIMPLE_Q_TAG,
action,
SSD_FULL_SIZE,
5000);
error = cam_periph_runccb(ccb, /*error_routine*/NULL, /*cam_flags*/0,
/*sense_flags*/0, &softc->device_stats);
if (error == 0) {
if (action == PR_ALLOW)
softc->flags &= ~DA_FLAG_PACK_LOCKED;
else
softc->flags |= DA_FLAG_PACK_LOCKED;
}
xpt_release_ccb(ccb);
}
static void
dasetgeom(struct cam_periph *periph, struct scsi_read_capacity_data * rdcap)
{
struct ccb_calc_geometry ccg;
struct da_softc *softc;
struct disk_params *dp;
softc = (struct da_softc *)periph->softc;
dp = &softc->params;
dp->secsize = scsi_4btoul(rdcap->length);
dp->sectors = scsi_4btoul(rdcap->addr) + 1;
/*
* Have the controller provide us with a geometry
* for this disk. The only time the geometry
* matters is when we boot and the controller
* is the only one knowledgeable enough to come
* up with something that will make this a bootable
* device.
*/
xpt_setup_ccb(&ccg.ccb_h, periph->path, /*priority*/1);
ccg.ccb_h.func_code = XPT_CALC_GEOMETRY;
ccg.block_size = dp->secsize;
ccg.volume_size = dp->sectors;
ccg.heads = 0;
ccg.secs_per_track = 0;
ccg.cylinders = 0;
xpt_action((union ccb*)&ccg);
dp->heads = ccg.heads;
dp->secs_per_track = ccg.secs_per_track;
dp->cylinders = ccg.cylinders;
}
static void
dasendorderedtag(void *arg)
{
struct da_softc *softc;
int s;
for (softc = SLIST_FIRST(&softc_list);
softc != NULL;
softc = SLIST_NEXT(softc, links)) {
s = splsoftcam();
if ((softc->ordered_tag_count == 0)
&& ((softc->flags & DA_FLAG_WENT_IDLE) == 0)) {
softc->flags |= DA_FLAG_NEED_OTAG;
}
if (softc->device_stats.busy_count > 0)
softc->flags &= ~DA_FLAG_WENT_IDLE;
softc->ordered_tag_count = 0;
splx(s);
}
/* Queue us up again */
timeout(dasendorderedtag, NULL,
(DA_DEFAULT_TIMEOUT * hz) / DA_ORDEREDTAG_INTERVAL);
}
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
/*
* Step through all DA peripheral drivers, and if the device is still open,
* sync the disk cache to physical media.
*/
static void
dashutdown(int howto, void *arg)
{
struct cam_periph *periph;
struct da_softc *softc;
for (periph = TAILQ_FIRST(&dadriver.units); periph != NULL;
periph = TAILQ_NEXT(periph, unit_links)) {
union ccb ccb;
softc = (struct da_softc *)periph->softc;
/*
* We only sync the cache if the drive is still open, and
* if the drive is capable of it..
*/
if (((softc->flags & DA_FLAG_OPEN) == 0)
|| (softc->quirks & DA_Q_NO_SYNC_CACHE))
continue;
xpt_setup_ccb(&ccb.ccb_h, periph->path, /*priority*/1);
ccb.ccb_h.ccb_state = DA_CCB_DUMP;
scsi_synchronize_cache(&ccb.csio,
/*retries*/1,
/*cbfcnp*/dadone,
MSG_SIMPLE_Q_TAG,
/*begin_lba*/0, /* whole disk */
/*lb_count*/0,
SSD_FULL_SIZE,
5 * 60 * 1000);
xpt_polled_action(&ccb);
if ((ccb.ccb_h.status & CAM_STATUS_MASK) != CAM_REQ_CMP) {
Add quirk entries to disable the synchronize cache command for Micropolis 2217's (reported by Matthew Jacob in NetBSD PR kern/6027) and Fujitsu M2954's (reported by Tom Jackson). Some of the Fujitsus at least hang when they get a cache sync command. (Others just return illegal request.) Also, make error printing in dashutdown() a little more selective. Don't print any error when the sense key is illegal request. Drives that don't support the synchronize cache command usually return illegal request. Also, make sure the scsi status is check condition before going into scsi_sense_print(). Reviewed by: gibbs
1998-10-12 17:16:47 +00:00
if (((ccb.ccb_h.status & CAM_STATUS_MASK) ==
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
CAM_SCSI_STATUS_ERROR)
Add quirk entries to disable the synchronize cache command for Micropolis 2217's (reported by Matthew Jacob in NetBSD PR kern/6027) and Fujitsu M2954's (reported by Tom Jackson). Some of the Fujitsus at least hang when they get a cache sync command. (Others just return illegal request.) Also, make error printing in dashutdown() a little more selective. Don't print any error when the sense key is illegal request. Drives that don't support the synchronize cache command usually return illegal request. Also, make sure the scsi status is check condition before going into scsi_sense_print(). Reviewed by: gibbs
1998-10-12 17:16:47 +00:00
&& (ccb.csio.scsi_status == SCSI_STATUS_CHECK_COND)){
int error_code, sense_key, asc, ascq;
scsi_extract_sense(&ccb.csio.sense_data,
&error_code, &sense_key,
&asc, &ascq);
if (sense_key != SSD_KEY_ILLEGAL_REQUEST)
scsi_sense_print(&ccb.csio);
} else {
Add the quirk entry framework to handle disabling the synchronize cache command on drives that don't like it. Right now, there's just a bogus quirk entry in the table that doesn't do anything, but that should be changed once we get actual inquiry data for drives that don't like the synchronize cache command. Also, add a shutdown hook that runs through all direct access peripherals and runs a synchronize cache on them if they're still open, and if synchronize cache isn't disabled via a quirk entry. Add a synchronize cache call at the end of dadump() (again, conditionalized on the quirk entry), so we can insure that the disk cache contents get flushed to physical media after a dump. Check the new quirk entry in daclose() to decide whether or not to synchronize the cache for a disk at final close. Reviewed by: gibbs
1998-10-08 05:46:38 +00:00
xpt_print_path(periph->path);
printf("Synchronize cache failed, status "
"== 0x%x, scsi status == 0x%x\n",
ccb.ccb_h.status, ccb.csio.scsi_status);
}
}
if ((ccb.ccb_h.status & CAM_DEV_QFRZN) != 0)
cam_release_devq(ccb.ccb_h.path,
/*relsim_flags*/0,
/*reduction*/0,
/*timeout*/0,
/*getcount_only*/0);
}
}
Reference in New Issue Copy Permalink