freebsd-dev/sys/dev/dpt/dpt_scsi.c
Konstantin Belousov dd0b4fb6d5 Reform the busdma API so that new types may be added without modifying
every architecture's busdma_machdep.c.  It is done by unifying the
bus_dmamap_load_buffer() routines so that they may be called from MI
code.  The MD busdma is then given a chance to do any final processing
in the complete() callback.

The cam changes unify the bus_dmamap_load* handling in cam drivers.

The arm and mips implementations are updated to track virtual
addresses for sync().  Previously this was done in a type specific
way.  Now it is done in a generic way by recording the list of
virtuals in the map.

Submitted by:	jeff (sponsored by EMC/Isilon)
Reviewed by:	kan (previous version), scottl,
	mjacob (isp(4), no objections for target mode changes)
Discussed with:	     ian (arm changes)
Tested by:	marius (sparc64), mips (jmallet), isci(4) on x86 (jharris),
	amd64 (Fabian Keil <freebsd-listen@fabiankeil.de>)
2013-02-12 16:57:20 +00:00

2678 lines
70 KiB
C

/*-
* Copyright (c) 1997 by Simon Shapiro
* 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. Redistributions in binary form must reproduce the above copyright
* notice, this list of conditions and the following disclaimer in the
* documentation and/or other materials provided with the distribution.
* 3. The 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.
*/
#include <sys/cdefs.h>
__FBSDID("$FreeBSD$");
/*
* dpt_scsi.c: SCSI dependant code for the DPT driver
*
* credits: Assisted by Mike Neuffer in the early low level DPT code
* Thanx to Mark Salyzyn of DPT for his assistance.
* Special thanx to Justin Gibbs for invaluable help in
* making this driver look and work like a FreeBSD component.
* Last but not least, many thanx to UCB and the FreeBSD
* team for creating and maintaining such a wonderful O/S.
*
* TODO: * Add ISA probe code.
* * Add driver-level RAID-0. This will allow interoperability with
* NiceTry, M$-Doze, Win-Dog, Slowlaris, etc., in recognizing RAID
* arrays that span controllers (Wow!).
*/
#define _DPT_C_
#include "opt_dpt.h"
#include "opt_eisa.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/conf.h>
#include <sys/eventhandler.h>
#include <sys/malloc.h>
#include <sys/kernel.h>
#include <sys/bus.h>
#include <machine/bus.h>
#include <machine/resource.h>
#include <sys/rman.h>
#include <cam/cam.h>
#include <cam/cam_ccb.h>
#include <cam/cam_sim.h>
#include <cam/cam_xpt_sim.h>
#include <cam/cam_debug.h>
#include <cam/scsi/scsi_all.h>
#include <cam/scsi/scsi_message.h>
#include <vm/vm.h>
#include <vm/pmap.h>
#include <dev/dpt/dpt.h>
/* dpt_isa.c, dpt_eisa.c, and dpt_pci.c need this in a central place */
devclass_t dpt_devclass;
#define microtime_now dpt_time_now()
#define dpt_inl(dpt, port) \
bus_read_4((dpt)->io_res, (dpt)->io_offset + port)
#define dpt_inb(dpt, port) \
bus_read_1((dpt)->io_res, (dpt)->io_offset + port)
#define dpt_outl(dpt, port, value) \
bus_write_4((dpt)->io_res, (dpt)->io_offset + port, value)
#define dpt_outb(dpt, port, value) \
bus_write_1((dpt)->io_res, (dpt)->io_offset + port, value)
/*
* These will have to be setup by parameters passed at boot/load time. For
* perfromance reasons, we make them constants for the time being.
*/
#define dpt_min_segs DPT_MAX_SEGS
#define dpt_max_segs DPT_MAX_SEGS
/* Definitions for our use of the SIM private CCB area */
#define ccb_dccb_ptr spriv_ptr0
#define ccb_dpt_ptr spriv_ptr1
/* ================= Private Inline Function declarations ===================*/
static __inline int dpt_just_reset(dpt_softc_t * dpt);
static __inline int dpt_raid_busy(dpt_softc_t * dpt);
#ifdef DEV_EISA
static __inline int dpt_pio_wait (u_int32_t, u_int, u_int, u_int);
#endif
static __inline int dpt_wait(dpt_softc_t *dpt, u_int bits,
u_int state);
static __inline struct dpt_ccb* dptgetccb(struct dpt_softc *dpt);
static __inline void dptfreeccb(struct dpt_softc *dpt,
struct dpt_ccb *dccb);
static __inline bus_addr_t dptccbvtop(struct dpt_softc *dpt,
struct dpt_ccb *dccb);
static __inline int dpt_send_immediate(dpt_softc_t *dpt,
eata_ccb_t *cmd_block,
u_int32_t cmd_busaddr,
u_int retries,
u_int ifc, u_int code,
u_int code2);
/* ==================== Private Function declarations =======================*/
static void dptmapmem(void *arg, bus_dma_segment_t *segs,
int nseg, int error);
static struct sg_map_node*
dptallocsgmap(struct dpt_softc *dpt);
static int dptallocccbs(dpt_softc_t *dpt);
static int dpt_get_conf(dpt_softc_t *dpt, dpt_ccb_t *dccb,
u_int32_t dccb_busaddr, u_int size,
u_int page, u_int target, int extent);
static void dpt_detect_cache(dpt_softc_t *dpt, dpt_ccb_t *dccb,
u_int32_t dccb_busaddr,
u_int8_t *buff);
static void dpt_poll(struct cam_sim *sim);
static void dpt_intr_locked(dpt_softc_t *dpt);
static void dptexecuteccb(void *arg, bus_dma_segment_t *dm_segs,
int nseg, int error);
static void dpt_action(struct cam_sim *sim, union ccb *ccb);
static int dpt_send_eata_command(dpt_softc_t *dpt, eata_ccb_t *cmd,
u_int32_t cmd_busaddr,
u_int command, u_int retries,
u_int ifc, u_int code,
u_int code2);
static void dptprocesserror(dpt_softc_t *dpt, dpt_ccb_t *dccb,
union ccb *ccb, u_int hba_stat,
u_int scsi_stat, u_int32_t resid);
static void dpttimeout(void *arg);
static void dptshutdown(void *arg, int howto);
/* ================= Private Inline Function definitions ====================*/
static __inline int
dpt_just_reset(dpt_softc_t * dpt)
{
if ((dpt_inb(dpt, 2) == 'D')
&& (dpt_inb(dpt, 3) == 'P')
&& (dpt_inb(dpt, 4) == 'T')
&& (dpt_inb(dpt, 5) == 'H'))
return (1);
else
return (0);
}
static __inline int
dpt_raid_busy(dpt_softc_t * dpt)
{
if ((dpt_inb(dpt, 0) == 'D')
&& (dpt_inb(dpt, 1) == 'P')
&& (dpt_inb(dpt, 2) == 'T'))
return (1);
else
return (0);
}
#ifdef DEV_EISA
static __inline int
dpt_pio_wait (u_int32_t base, u_int reg, u_int bits, u_int state)
{
int i;
u_int c;
for (i = 0; i < 20000; i++) { /* wait 20ms for not busy */
c = inb(base + reg) & bits;
if (!(c == state))
return (0);
else
DELAY(50);
}
return (-1);
}
#endif
static __inline int
dpt_wait(dpt_softc_t *dpt, u_int bits, u_int state)
{
int i;
u_int c;
for (i = 0; i < 20000; i++) { /* wait 20ms for not busy */
c = dpt_inb(dpt, HA_RSTATUS) & bits;
if (c == state)
return (0);
else
DELAY(50);
}
return (-1);
}
static __inline struct dpt_ccb*
dptgetccb(struct dpt_softc *dpt)
{
struct dpt_ccb* dccb;
if (!dumping)
mtx_assert(&dpt->lock, MA_OWNED);
if ((dccb = SLIST_FIRST(&dpt->free_dccb_list)) != NULL) {
SLIST_REMOVE_HEAD(&dpt->free_dccb_list, links);
dpt->free_dccbs--;
} else if (dpt->total_dccbs < dpt->max_dccbs) {
dptallocccbs(dpt);
dccb = SLIST_FIRST(&dpt->free_dccb_list);
if (dccb == NULL)
device_printf(dpt->dev, "Can't malloc DCCB\n");
else {
SLIST_REMOVE_HEAD(&dpt->free_dccb_list, links);
dpt->free_dccbs--;
}
}
return (dccb);
}
static __inline void
dptfreeccb(struct dpt_softc *dpt, struct dpt_ccb *dccb)
{
if (!dumping)
mtx_assert(&dpt->lock, MA_OWNED);
if ((dccb->state & DCCB_ACTIVE) != 0)
LIST_REMOVE(&dccb->ccb->ccb_h, sim_links.le);
if ((dccb->state & DCCB_RELEASE_SIMQ) != 0)
dccb->ccb->ccb_h.status |= CAM_RELEASE_SIMQ;
else if (dpt->resource_shortage != 0
&& (dccb->ccb->ccb_h.status & CAM_RELEASE_SIMQ) == 0) {
dccb->ccb->ccb_h.status |= CAM_RELEASE_SIMQ;
dpt->resource_shortage = FALSE;
}
dccb->state = DCCB_FREE;
SLIST_INSERT_HEAD(&dpt->free_dccb_list, dccb, links);
++dpt->free_dccbs;
}
static __inline bus_addr_t
dptccbvtop(struct dpt_softc *dpt, struct dpt_ccb *dccb)
{
return (dpt->dpt_ccb_busbase
+ (u_int32_t)((caddr_t)dccb - (caddr_t)dpt->dpt_dccbs));
}
static __inline struct dpt_ccb *
dptccbptov(struct dpt_softc *dpt, bus_addr_t busaddr)
{
return (dpt->dpt_dccbs
+ ((struct dpt_ccb *)busaddr
- (struct dpt_ccb *)dpt->dpt_ccb_busbase));
}
/*
* Send a command for immediate execution by the DPT
* See above function for IMPORTANT notes.
*/
static __inline int
dpt_send_immediate(dpt_softc_t *dpt, eata_ccb_t *cmd_block,
u_int32_t cmd_busaddr, u_int retries,
u_int ifc, u_int code, u_int code2)
{
return (dpt_send_eata_command(dpt, cmd_block, cmd_busaddr,
EATA_CMD_IMMEDIATE, retries, ifc,
code, code2));
}
/* ===================== Private Function definitions =======================*/
static void
dptmapmem(void *arg, bus_dma_segment_t *segs, int nseg, int error)
{
bus_addr_t *busaddrp;
busaddrp = (bus_addr_t *)arg;
*busaddrp = segs->ds_addr;
}
static struct sg_map_node *
dptallocsgmap(struct dpt_softc *dpt)
{
struct sg_map_node *sg_map;
sg_map = malloc(sizeof(*sg_map), M_DEVBUF, M_NOWAIT);
if (sg_map == NULL)
return (NULL);
/* Allocate S/G space for the next batch of CCBS */
if (bus_dmamem_alloc(dpt->sg_dmat, (void **)&sg_map->sg_vaddr,
BUS_DMA_NOWAIT, &sg_map->sg_dmamap) != 0) {
free(sg_map, M_DEVBUF);
return (NULL);
}
(void)bus_dmamap_load(dpt->sg_dmat, sg_map->sg_dmamap, sg_map->sg_vaddr,
PAGE_SIZE, dptmapmem, &sg_map->sg_physaddr,
/*flags*/0);
SLIST_INSERT_HEAD(&dpt->sg_maps, sg_map, links);
return (sg_map);
}
/*
* Allocate another chunk of CCB's. Return count of entries added.
*/
static int
dptallocccbs(dpt_softc_t *dpt)
{
struct dpt_ccb *next_ccb;
struct sg_map_node *sg_map;
bus_addr_t physaddr;
dpt_sg_t *segs;
int newcount;
int i;
if (!dumping)
mtx_assert(&dpt->lock, MA_OWNED);
next_ccb = &dpt->dpt_dccbs[dpt->total_dccbs];
if (next_ccb == dpt->dpt_dccbs) {
/*
* First time through. Re-use the S/G
* space we allocated for initialization
* CCBS.
*/
sg_map = SLIST_FIRST(&dpt->sg_maps);
} else {
sg_map = dptallocsgmap(dpt);
}
if (sg_map == NULL)
return (0);
segs = sg_map->sg_vaddr;
physaddr = sg_map->sg_physaddr;
newcount = (PAGE_SIZE / (dpt->sgsize * sizeof(dpt_sg_t)));
for (i = 0; dpt->total_dccbs < dpt->max_dccbs && i < newcount; i++) {
int error;
error = bus_dmamap_create(dpt->buffer_dmat, /*flags*/0,
&next_ccb->dmamap);
if (error != 0)
break;
callout_init_mtx(&next_ccb->timer, &dpt->lock, 0);
next_ccb->sg_list = segs;
next_ccb->sg_busaddr = htonl(physaddr);
next_ccb->eata_ccb.cp_dataDMA = htonl(physaddr);
next_ccb->eata_ccb.cp_statDMA = htonl(dpt->sp_physaddr);
next_ccb->eata_ccb.cp_reqDMA =
htonl(dptccbvtop(dpt, next_ccb)
+ offsetof(struct dpt_ccb, sense_data));
next_ccb->eata_ccb.cp_busaddr = dpt->dpt_ccb_busend;
next_ccb->state = DCCB_FREE;
next_ccb->tag = dpt->total_dccbs;
SLIST_INSERT_HEAD(&dpt->free_dccb_list, next_ccb, links);
segs += dpt->sgsize;
physaddr += (dpt->sgsize * sizeof(dpt_sg_t));
dpt->dpt_ccb_busend += sizeof(*next_ccb);
next_ccb++;
dpt->total_dccbs++;
}
return (i);
}
#ifdef DEV_EISA
dpt_conf_t *
dpt_pio_get_conf (u_int32_t base)
{
static dpt_conf_t * conf;
u_int16_t * p;
int i;
/*
* Allocate a dpt_conf_t
*/
if (!conf) {
conf = (dpt_conf_t *)malloc(sizeof(dpt_conf_t),
M_DEVBUF, M_NOWAIT | M_ZERO);
}
/*
* If we didn't get one then we probably won't ever get one.
*/
if (!conf) {
printf("dpt: unable to allocate dpt_conf_t\n");
return (NULL);
}
/*
* Reset the controller.
*/
outb((base + HA_WCOMMAND), EATA_CMD_RESET);
/*
* Wait for the controller to become ready.
* For some reason there can be -no- delays after calling reset
* before we wait on ready status.
*/
if (dpt_pio_wait(base, HA_RSTATUS, HA_SBUSY, 0)) {
printf("dpt: timeout waiting for controller to become ready\n");
return (NULL);
}
if (dpt_pio_wait(base, HA_RAUXSTAT, HA_ABUSY, 0)) {
printf("dpt: timetout waiting for adapter ready.\n");
return (NULL);
}
/*
* Send the PIO_READ_CONFIG command.
*/
outb((base + HA_WCOMMAND), EATA_CMD_PIO_READ_CONFIG);
/*
* Read the data into the struct.
*/
p = (u_int16_t *)conf;
for (i = 0; i < (sizeof(dpt_conf_t) / 2); i++) {
if (dpt_pio_wait(base, HA_RSTATUS, HA_SDRQ, 0)) {
if (bootverbose)
printf("dpt: timeout in data read.\n");
return (NULL);
}
(*p) = inw(base + HA_RDATA);
p++;
}
if (inb(base + HA_RSTATUS) & HA_SERROR) {
if (bootverbose)
printf("dpt: error reading configuration data.\n");
return (NULL);
}
#define BE_EATA_SIGNATURE 0x45415441
#define LE_EATA_SIGNATURE 0x41544145
/*
* Test to see if we have a valid card.
*/
if ((conf->signature == BE_EATA_SIGNATURE) ||
(conf->signature == LE_EATA_SIGNATURE)) {
while (inb(base + HA_RSTATUS) & HA_SDRQ) {
inw(base + HA_RDATA);
}
return (conf);
}
return (NULL);
}
#endif
/*
* Read a configuration page into the supplied dpt_cont_t buffer.
*/
static int
dpt_get_conf(dpt_softc_t *dpt, dpt_ccb_t *dccb, u_int32_t dccb_busaddr,
u_int size, u_int page, u_int target, int extent)
{
eata_ccb_t *cp;
u_int8_t status;
int ndx;
int result;
mtx_assert(&dpt->lock, MA_OWNED);
cp = &dccb->eata_ccb;
bzero((void *)(uintptr_t)(volatile void *)dpt->sp, sizeof(*dpt->sp));
cp->Interpret = 1;
cp->DataIn = 1;
cp->Auto_Req_Sen = 1;
cp->reqlen = sizeof(struct scsi_sense_data);
cp->cp_id = target;
cp->cp_LUN = 0; /* In the EATA packet */
cp->cp_lun = 0; /* In the SCSI command */
cp->cp_scsi_cmd = INQUIRY;
cp->cp_len = size;
cp->cp_extent = extent;
cp->cp_page = page;
cp->cp_channel = 0; /* DNC, Interpret mode is set */
cp->cp_identify = 1;
cp->cp_datalen = htonl(size);
/*
* This could be a simple for loop, but we suspected the compiler To
* have optimized it a bit too much. Wait for the controller to
* become ready
*/
while (((status = dpt_inb(dpt, HA_RSTATUS)) != (HA_SREADY | HA_SSC)
&& (status != (HA_SREADY | HA_SSC | HA_SERROR))
&& (status != (HA_SDRDY | HA_SERROR | HA_SDRQ)))
|| (dpt_wait(dpt, HA_SBUSY, 0))) {
/*
* RAID Drives still Spinning up? (This should only occur if
* the DPT controller is in a NON PC (PCI?) platform).
*/
if (dpt_raid_busy(dpt)) {
device_printf(dpt->dev,
"WARNING: Get_conf() RSUS failed.\n");
return (0);
}
}
DptStat_Reset_BUSY(dpt->sp);
/*
* XXXX We might want to do something more clever than aborting at
* this point, like resetting (rebooting) the controller and trying
* again.
*/
if ((result = dpt_send_eata_command(dpt, cp, dccb_busaddr,
EATA_CMD_DMA_SEND_CP,
10000, 0, 0, 0)) != 0) {
device_printf(dpt->dev,
"WARNING: Get_conf() failed (%d) to send "
"EATA_CMD_DMA_READ_CONFIG\n",
result);
return (0);
}
/* Wait for two seconds for a response. This can be slow */
for (ndx = 0;
(ndx < 20000)
&& !((status = dpt_inb(dpt, HA_RAUXSTAT)) & HA_AIRQ);
ndx++) {
DELAY(50);
}
/* Grab the status and clear interrupts */
status = dpt_inb(dpt, HA_RSTATUS);
/*
* Check the status carefully. Return only if the
* command was successful.
*/
if (((status & HA_SERROR) == 0)
&& (dpt->sp->hba_stat == 0)
&& (dpt->sp->scsi_stat == 0)
&& (dpt->sp->residue_len == 0))
return (0);
if (dpt->sp->scsi_stat == SCSI_STATUS_CHECK_COND)
return (0);
return (1);
}
/* Detect Cache parameters and size */
static void
dpt_detect_cache(dpt_softc_t *dpt, dpt_ccb_t *dccb, u_int32_t dccb_busaddr,
u_int8_t *buff)
{
eata_ccb_t *cp;
u_int8_t *param;
int bytes;
int result;
int ndx;
u_int8_t status;
mtx_assert(&dpt->lock, MA_OWNED);
/*
* Default setting, for best perfromance..
* This is what virtually all cards default to..
*/
dpt->cache_type = DPT_CACHE_WRITEBACK;
dpt->cache_size = 0;
cp = &dccb->eata_ccb;
bzero((void *)(uintptr_t)(volatile void *)dpt->sp, sizeof(dpt->sp));
bzero(buff, 512);
/* Setup the command structure */
cp->Interpret = 1;
cp->DataIn = 1;
cp->Auto_Req_Sen = 1;
cp->reqlen = sizeof(struct scsi_sense_data);
cp->cp_id = 0; /* who cares? The HBA will interpret.. */
cp->cp_LUN = 0; /* In the EATA packet */
cp->cp_lun = 0; /* In the SCSI command */
cp->cp_channel = 0;
cp->cp_scsi_cmd = EATA_CMD_DMA_SEND_CP;
cp->cp_len = 56;
cp->cp_extent = 0;
cp->cp_page = 0;
cp->cp_identify = 1;
cp->cp_dispri = 1;
/*
* Build the EATA Command Packet structure
* for a Log Sense Command.
*/
cp->cp_cdb[0] = 0x4d;
cp->cp_cdb[1] = 0x0;
cp->cp_cdb[2] = 0x40 | 0x33;
cp->cp_cdb[7] = 1;
cp->cp_datalen = htonl(512);
result = dpt_send_eata_command(dpt, cp, dccb_busaddr,
EATA_CMD_DMA_SEND_CP,
10000, 0, 0, 0);
if (result != 0) {
device_printf(dpt->dev,
"WARNING: detect_cache() failed (%d) to send "
"EATA_CMD_DMA_SEND_CP\n", result);
return;
}
/* Wait for two seconds for a response. This can be slow... */
for (ndx = 0;
(ndx < 20000) &&
!((status = dpt_inb(dpt, HA_RAUXSTAT)) & HA_AIRQ);
ndx++) {
DELAY(50);
}
/* Grab the status and clear interrupts */
status = dpt_inb(dpt, HA_RSTATUS);
/*
* Sanity check
*/
if (buff[0] != 0x33) {
return;
}
bytes = DPT_HCP_LENGTH(buff);
param = DPT_HCP_FIRST(buff);
if (DPT_HCP_CODE(param) != 1) {
/*
* DPT Log Page layout error
*/
device_printf(dpt->dev, "NOTICE: Log Page (1) layout error\n");
return;
}
if (!(param[4] & 0x4)) {
dpt->cache_type = DPT_NO_CACHE;
return;
}
while (DPT_HCP_CODE(param) != 6) {
param = DPT_HCP_NEXT(param);
if ((param < buff)
|| (param >= &buff[bytes])) {
return;
}
}
if (param[4] & 0x2) {
/*
* Cache disabled
*/
dpt->cache_type = DPT_NO_CACHE;
return;
}
if (param[4] & 0x4) {
dpt->cache_type = DPT_CACHE_WRITETHROUGH;
}
/* XXX This isn't correct. This log parameter only has two bytes.... */
#if 0
dpt->cache_size = param[5]
| (param[6] << 8)
| (param[7] << 16)
| (param[8] << 24);
#endif
}
static void
dpt_poll(struct cam_sim *sim)
{
dpt_intr_locked(cam_sim_softc(sim));
}
static void
dptexecuteccb(void *arg, bus_dma_segment_t *dm_segs, int nseg, int error)
{
struct dpt_ccb *dccb;
union ccb *ccb;
struct dpt_softc *dpt;
dccb = (struct dpt_ccb *)arg;
ccb = dccb->ccb;
dpt = (struct dpt_softc *)ccb->ccb_h.ccb_dpt_ptr;
if (!dumping)
mtx_assert(&dpt->lock, MA_OWNED);
if (error != 0) {
if (error != EFBIG)
device_printf(dpt->dev,
"Unexepected error 0x%x returned from "
"bus_dmamap_load\n", error);
if (ccb->ccb_h.status == CAM_REQ_INPROG) {
xpt_freeze_devq(ccb->ccb_h.path, /*count*/1);
ccb->ccb_h.status = CAM_REQ_TOO_BIG|CAM_DEV_QFRZN;
}
dptfreeccb(dpt, dccb);
xpt_done(ccb);
return;
}
if (nseg != 0) {
dpt_sg_t *sg;
bus_dma_segment_t *end_seg;
bus_dmasync_op_t op;
end_seg = dm_segs + nseg;
/* Copy the segments into our SG list */
sg = dccb->sg_list;
while (dm_segs < end_seg) {
sg->seg_len = htonl(dm_segs->ds_len);
sg->seg_addr = htonl(dm_segs->ds_addr);
sg++;
dm_segs++;
}
if (nseg > 1) {
dccb->eata_ccb.scatter = 1;
dccb->eata_ccb.cp_dataDMA = dccb->sg_busaddr;
dccb->eata_ccb.cp_datalen =
htonl(nseg * sizeof(dpt_sg_t));
} else {
dccb->eata_ccb.cp_dataDMA = dccb->sg_list[0].seg_addr;
dccb->eata_ccb.cp_datalen = dccb->sg_list[0].seg_len;
}
if ((ccb->ccb_h.flags & CAM_DIR_MASK) == CAM_DIR_IN)
op = BUS_DMASYNC_PREREAD;
else
op = BUS_DMASYNC_PREWRITE;
bus_dmamap_sync(dpt->buffer_dmat, dccb->dmamap, op);
} else {
dccb->eata_ccb.cp_dataDMA = 0;
dccb->eata_ccb.cp_datalen = 0;
}
/*
* Last time we need to check if this CCB needs to
* be aborted.
*/
if (ccb->ccb_h.status != CAM_REQ_INPROG) {
if (nseg != 0)
bus_dmamap_unload(dpt->buffer_dmat, dccb->dmamap);
dptfreeccb(dpt, dccb);
xpt_done(ccb);
return;
}
dccb->state |= DCCB_ACTIVE;
ccb->ccb_h.status |= CAM_SIM_QUEUED;
LIST_INSERT_HEAD(&dpt->pending_ccb_list, &ccb->ccb_h, sim_links.le);
callout_reset(&dccb->timer, (ccb->ccb_h.timeout * hz) / 1000,
dpttimeout, dccb);
if (dpt_send_eata_command(dpt, &dccb->eata_ccb,
dccb->eata_ccb.cp_busaddr,
EATA_CMD_DMA_SEND_CP, 0, 0, 0, 0) != 0) {
ccb->ccb_h.status = CAM_NO_HBA; /* HBA dead or just busy?? */
if (nseg != 0)
bus_dmamap_unload(dpt->buffer_dmat, dccb->dmamap);
dptfreeccb(dpt, dccb);
xpt_done(ccb);
}
}
static void
dpt_action(struct cam_sim *sim, union ccb *ccb)
{
struct dpt_softc *dpt;
CAM_DEBUG(ccb->ccb_h.path, CAM_DEBUG_TRACE, ("dpt_action\n"));
dpt = (struct dpt_softc *)cam_sim_softc(sim);
mtx_assert(&dpt->lock, MA_OWNED);
if ((dpt->state & DPT_HA_SHUTDOWN_ACTIVE) != 0) {
xpt_print_path(ccb->ccb_h.path);
printf("controller is shutdown. Aborting CCB.\n");
ccb->ccb_h.status = CAM_NO_HBA;
xpt_done(ccb);
return;
}
switch (ccb->ccb_h.func_code) {
/* Common cases first */
case XPT_SCSI_IO: /* Execute the requested I/O operation */
{
struct ccb_scsiio *csio;
struct ccb_hdr *ccbh;
struct dpt_ccb *dccb;
struct eata_ccb *eccb;
csio = &ccb->csio;
ccbh = &ccb->ccb_h;
/* Max CDB length is 12 bytes */
if (csio->cdb_len > 12) {
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
return;
}
if ((dccb = dptgetccb(dpt)) == NULL) {
dpt->resource_shortage = 1;
xpt_freeze_simq(sim, /*count*/1);
ccb->ccb_h.status = CAM_REQUEUE_REQ;
xpt_done(ccb);
return;
}
eccb = &dccb->eata_ccb;
/* Link dccb and ccb so we can find one from the other */
dccb->ccb = ccb;
ccb->ccb_h.ccb_dccb_ptr = dccb;
ccb->ccb_h.ccb_dpt_ptr = dpt;
/*
* Explicitly set all flags so that the compiler can
* be smart about setting them.
*/
eccb->SCSI_Reset = 0;
eccb->HBA_Init = 0;
eccb->Auto_Req_Sen = (ccb->ccb_h.flags & CAM_DIS_AUTOSENSE)
? 0 : 1;
eccb->scatter = 0;
eccb->Quick = 0;
eccb->Interpret =
ccb->ccb_h.target_id == dpt->hostid[cam_sim_bus(sim)]
? 1 : 0;
eccb->DataOut = (ccb->ccb_h.flags & CAM_DIR_OUT) ? 1 : 0;
eccb->DataIn = (ccb->ccb_h.flags & CAM_DIR_IN) ? 1 : 0;
eccb->reqlen = csio->sense_len;
eccb->cp_id = ccb->ccb_h.target_id;
eccb->cp_channel = cam_sim_bus(sim);
eccb->cp_LUN = ccb->ccb_h.target_lun;
eccb->cp_luntar = 0;
eccb->cp_dispri = (ccb->ccb_h.flags & CAM_DIS_DISCONNECT)
? 0 : 1;
eccb->cp_identify = 1;
if ((ccb->ccb_h.flags & CAM_TAG_ACTION_VALID) != 0
&& csio->tag_action != CAM_TAG_ACTION_NONE) {
eccb->cp_msg[0] = csio->tag_action;
eccb->cp_msg[1] = dccb->tag;
} else {
eccb->cp_msg[0] = 0;
eccb->cp_msg[1] = 0;
}
eccb->cp_msg[2] = 0;
if ((ccb->ccb_h.flags & CAM_CDB_POINTER) != 0) {
if ((ccb->ccb_h.flags & CAM_CDB_PHYS) == 0) {
bcopy(csio->cdb_io.cdb_ptr,
eccb->cp_cdb, csio->cdb_len);
} else {
/* I guess I could map it in... */
ccb->ccb_h.status = CAM_REQ_INVALID;
dptfreeccb(dpt, dccb);
xpt_done(ccb);
return;
}
} else {
bcopy(csio->cdb_io.cdb_bytes,
eccb->cp_cdb, csio->cdb_len);
}
/*
* If we have any data to send with this command,
* map it into bus space.
*/
/* Only use S/G if there is a transfer */
if ((ccbh->flags & CAM_DIR_MASK) != CAM_DIR_NONE) {
int error;
error = bus_dmamap_load_ccb(dpt->buffer_dmat,
dccb->dmamap,
ccb,
dptexecuteccb,
dccb, /*flags*/0);
if (error == EINPROGRESS) {
/*
* So as to maintain ordering,
* freeze the controller queue
* until our mapping is
* returned.
*/
xpt_freeze_simq(sim, 1);
dccb->state |= CAM_RELEASE_SIMQ;
}
} else {
/*
* XXX JGibbs.
* Does it want them both on or both off?
* CAM_DIR_NONE is both on, so this code can
* be removed if this is also what the DPT
* exptects.
*/
eccb->DataOut = 0;
eccb->DataIn = 0;
dptexecuteccb(dccb, NULL, 0, 0);
}
break;
}
case XPT_RESET_DEV: /* Bus Device Reset the specified SCSI device */
case XPT_ABORT: /* Abort the specified CCB */
/* XXX Implement */
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
break;
case XPT_SET_TRAN_SETTINGS:
{
ccb->ccb_h.status = CAM_FUNC_NOTAVAIL;
xpt_done(ccb);
break;
}
case XPT_GET_TRAN_SETTINGS:
/* Get default/user set transfer settings for the target */
{
struct ccb_trans_settings *cts = &ccb->cts;
struct ccb_trans_settings_scsi *scsi =
&cts->proto_specific.scsi;
struct ccb_trans_settings_spi *spi =
&cts->xport_specific.spi;
cts->protocol = PROTO_SCSI;
cts->protocol_version = SCSI_REV_2;
cts->transport = XPORT_SPI;
cts->transport_version = 2;
if (cts->type == CTS_TYPE_USER_SETTINGS) {
spi->flags = CTS_SPI_FLAGS_DISC_ENB;
spi->bus_width = (dpt->max_id > 7)
? MSG_EXT_WDTR_BUS_8_BIT
: MSG_EXT_WDTR_BUS_16_BIT;
spi->sync_period = 25; /* 10MHz */
if (spi->sync_period != 0)
spi->sync_offset = 15;
scsi->flags = CTS_SCSI_FLAGS_TAG_ENB;
spi->valid = CTS_SPI_VALID_SYNC_RATE
| CTS_SPI_VALID_SYNC_OFFSET
| CTS_SPI_VALID_BUS_WIDTH
| CTS_SPI_VALID_DISC;
scsi->valid = CTS_SCSI_VALID_TQ;
ccb->ccb_h.status = CAM_REQ_CMP;
} else {
ccb->ccb_h.status = CAM_FUNC_NOTAVAIL;
}
xpt_done(ccb);
break;
}
case XPT_CALC_GEOMETRY:
{
/*
* XXX Use Adaptec translation until I find out how to
* get this information from the card.
*/
cam_calc_geometry(&ccb->ccg, /*extended*/1);
xpt_done(ccb);
break;
}
case XPT_RESET_BUS: /* Reset the specified SCSI bus */
{
/* XXX Implement */
ccb->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
break;
}
case XPT_TERM_IO: /* Terminate the I/O process */
/* XXX Implement */
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
break;
case XPT_PATH_INQ: /* Path routing inquiry */
{
struct ccb_pathinq *cpi = &ccb->cpi;
cpi->version_num = 1;
cpi->hba_inquiry = PI_SDTR_ABLE|PI_TAG_ABLE;
if (dpt->max_id > 7)
cpi->hba_inquiry |= PI_WIDE_16;
cpi->target_sprt = 0;
cpi->hba_misc = 0;
cpi->hba_eng_cnt = 0;
cpi->max_target = dpt->max_id;
cpi->max_lun = dpt->max_lun;
cpi->initiator_id = dpt->hostid[cam_sim_bus(sim)];
cpi->bus_id = cam_sim_bus(sim);
cpi->base_transfer_speed = 3300;
strncpy(cpi->sim_vid, "FreeBSD", SIM_IDLEN);
strncpy(cpi->hba_vid, "DPT", HBA_IDLEN);
strncpy(cpi->dev_name, cam_sim_name(sim), DEV_IDLEN);
cpi->unit_number = cam_sim_unit(sim);
cpi->transport = XPORT_SPI;
cpi->transport_version = 2;
cpi->protocol = PROTO_SCSI;
cpi->protocol_version = SCSI_REV_2;
cpi->ccb_h.status = CAM_REQ_CMP;
xpt_done(ccb);
break;
}
default:
ccb->ccb_h.status = CAM_REQ_INVALID;
xpt_done(ccb);
break;
}
}
/*
* This routine will try to send an EATA command to the DPT HBA.
* It will, by default, try 20,000 times, waiting 50us between tries.
* It returns 0 on success and 1 on failure.
*/
static int
dpt_send_eata_command(dpt_softc_t *dpt, eata_ccb_t *cmd_block,
u_int32_t cmd_busaddr, u_int command, u_int retries,
u_int ifc, u_int code, u_int code2)
{
u_int loop;
if (!retries)
retries = 20000;
/*
* I hate this polling nonsense. Wish there was a way to tell the DPT
* to go get commands at its own pace, or to interrupt when ready.
* In the mean time we will measure how many itterations it really
* takes.
*/
for (loop = 0; loop < retries; loop++) {
if ((dpt_inb(dpt, HA_RAUXSTAT) & HA_ABUSY) == 0)
break;
else
DELAY(50);
}
if (loop < retries) {
#ifdef DPT_MEASURE_PERFORMANCE
if (loop > dpt->performance.max_eata_tries)
dpt->performance.max_eata_tries = loop;
if (loop < dpt->performance.min_eata_tries)
dpt->performance.min_eata_tries = loop;
#endif
} else {
#ifdef DPT_MEASURE_PERFORMANCE
++dpt->performance.command_too_busy;
#endif
return (1);
}
/* The controller is alive, advance the wedge timer */
#ifdef DPT_RESET_HBA
dpt->last_contact = microtime_now;
#endif
if (cmd_block == NULL)
cmd_busaddr = 0;
#if (BYTE_ORDER == BIG_ENDIAN)
else {
cmd_busaddr = ((cmd_busaddr >> 24) & 0xFF)
| ((cmd_busaddr >> 16) & 0xFF)
| ((cmd_busaddr >> 8) & 0xFF)
| (cmd_busaddr & 0xFF);
}
#endif
/* And now the address */
dpt_outl(dpt, HA_WDMAADDR, cmd_busaddr);
if (command == EATA_CMD_IMMEDIATE) {
if (cmd_block == NULL) {
dpt_outb(dpt, HA_WCODE2, code2);
dpt_outb(dpt, HA_WCODE, code);
}
dpt_outb(dpt, HA_WIFC, ifc);
}
dpt_outb(dpt, HA_WCOMMAND, command);
return (0);
}
/* ==================== Exported Function definitions =======================*/
void
dpt_alloc(device_t dev)
{
dpt_softc_t *dpt = device_get_softc(dev);
int i;
mtx_init(&dpt->lock, "dpt", NULL, MTX_DEF);
SLIST_INIT(&dpt->free_dccb_list);
LIST_INIT(&dpt->pending_ccb_list);
for (i = 0; i < MAX_CHANNELS; i++)
dpt->resetlevel[i] = DPT_HA_OK;
#ifdef DPT_MEASURE_PERFORMANCE
dpt_reset_performance(dpt);
#endif /* DPT_MEASURE_PERFORMANCE */
return;
}
void
dpt_free(struct dpt_softc *dpt)
{
switch (dpt->init_level) {
default:
case 5:
bus_dmamap_unload(dpt->dccb_dmat, dpt->dccb_dmamap);
case 4:
bus_dmamem_free(dpt->dccb_dmat, dpt->dpt_dccbs,
dpt->dccb_dmamap);
bus_dmamap_destroy(dpt->dccb_dmat, dpt->dccb_dmamap);
case 3:
bus_dma_tag_destroy(dpt->dccb_dmat);
case 2:
bus_dma_tag_destroy(dpt->buffer_dmat);
case 1:
{
struct sg_map_node *sg_map;
while ((sg_map = SLIST_FIRST(&dpt->sg_maps)) != NULL) {
SLIST_REMOVE_HEAD(&dpt->sg_maps, links);
bus_dmamap_unload(dpt->sg_dmat,
sg_map->sg_dmamap);
bus_dmamem_free(dpt->sg_dmat, sg_map->sg_vaddr,
sg_map->sg_dmamap);
free(sg_map, M_DEVBUF);
}
bus_dma_tag_destroy(dpt->sg_dmat);
}
case 0:
break;
}
mtx_destroy(&dpt->lock);
}
int
dpt_alloc_resources (device_t dev)
{
dpt_softc_t * dpt;
int error;
dpt = device_get_softc(dev);
dpt->io_res = bus_alloc_resource_any(dev, dpt->io_type, &dpt->io_rid,
RF_ACTIVE);
if (dpt->io_res == NULL) {
device_printf(dev, "No I/O space?!\n");
error = ENOMEM;
goto bad;
}
dpt->irq_res = bus_alloc_resource_any(dev, SYS_RES_IRQ, &dpt->irq_rid,
RF_ACTIVE);
if (dpt->irq_res == NULL) {
device_printf(dev, "No IRQ!\n");
error = ENOMEM;
goto bad;
}
return (0);
bad:
return(error);
}
void
dpt_release_resources (device_t dev)
{
struct dpt_softc * dpt;
dpt = device_get_softc(dev);
if (dpt->ih)
bus_teardown_intr(dev, dpt->irq_res, dpt->ih);
if (dpt->io_res)
bus_release_resource(dev, dpt->io_type, dpt->io_rid, dpt->io_res);
if (dpt->irq_res)
bus_release_resource(dev, SYS_RES_IRQ, dpt->irq_rid, dpt->irq_res);
if (dpt->drq_res)
bus_release_resource(dev, SYS_RES_DRQ, dpt->drq_rid, dpt->drq_res);
return;
}
static u_int8_t string_sizes[] =
{
sizeof(((dpt_inq_t*)NULL)->vendor),
sizeof(((dpt_inq_t*)NULL)->modelNum),
sizeof(((dpt_inq_t*)NULL)->firmware),
sizeof(((dpt_inq_t*)NULL)->protocol),
};
int
dpt_init(struct dpt_softc *dpt)
{
dpt_conf_t conf;
struct sg_map_node *sg_map;
dpt_ccb_t *dccb;
u_int8_t *strp;
int index;
int i;
int retval;
dpt->init_level = 0;
SLIST_INIT(&dpt->sg_maps);
mtx_lock(&dpt->lock);
#ifdef DPT_RESET_BOARD
device_printf(dpt->dev, "resetting HBA\n");
dpt_outb(dpt, HA_WCOMMAND, EATA_CMD_RESET);
DELAY(750000);
/* XXX Shouldn't we poll a status register or something??? */
#endif
/* DMA tag for our S/G structures. We allocate in page sized chunks */
if (bus_dma_tag_create( /* parent */ dpt->parent_dmat,
/* alignment */ 1,
/* boundary */ 0,
/* lowaddr */ BUS_SPACE_MAXADDR,
/* highaddr */ BUS_SPACE_MAXADDR,
/* filter */ NULL,
/* filterarg */ NULL,
/* maxsize */ PAGE_SIZE,
/* nsegments */ 1,
/* maxsegsz */ BUS_SPACE_MAXSIZE_32BIT,
/* flags */ 0,
/* lockfunc */ NULL,
/* lockarg */ NULL,
&dpt->sg_dmat) != 0) {
goto error_exit;
}
dpt->init_level++;
/*
* We allocate our DPT ccbs as a contiguous array of bus dma'able
* memory. To get the allocation size, we need to know how many
* ccbs the card supports. This requires a ccb. We solve this
* chicken and egg problem by allocating some re-usable S/G space
* up front, and treating it as our status packet, CCB, and target
* memory space for these commands.
*/
sg_map = dptallocsgmap(dpt);
if (sg_map == NULL)
goto error_exit;
dpt->sp = (volatile dpt_sp_t *)sg_map->sg_vaddr;
dccb = (struct dpt_ccb *)(uintptr_t)(volatile void *)&dpt->sp[1];
bzero(dccb, sizeof(*dccb));
dpt->sp_physaddr = sg_map->sg_physaddr;
dccb->eata_ccb.cp_dataDMA =
htonl(sg_map->sg_physaddr + sizeof(dpt_sp_t) + sizeof(*dccb));
dccb->eata_ccb.cp_busaddr = ~0;
dccb->eata_ccb.cp_statDMA = htonl(dpt->sp_physaddr);
dccb->eata_ccb.cp_reqDMA = htonl(dpt->sp_physaddr + sizeof(*dccb)
+ offsetof(struct dpt_ccb, sense_data));
/* Okay. Fetch our config */
bzero(&dccb[1], sizeof(conf)); /* data area */
retval = dpt_get_conf(dpt, dccb, sg_map->sg_physaddr + sizeof(dpt_sp_t),
sizeof(conf), 0xc1, 7, 1);
if (retval != 0) {
device_printf(dpt->dev, "Failed to get board configuration\n");
goto error_exit;
}
bcopy(&dccb[1], &conf, sizeof(conf));
bzero(&dccb[1], sizeof(dpt->board_data));
retval = dpt_get_conf(dpt, dccb, sg_map->sg_physaddr + sizeof(dpt_sp_t),
sizeof(dpt->board_data), 0, conf.scsi_id0, 0);
if (retval != 0) {
device_printf(dpt->dev, "Failed to get inquiry information\n");
goto error_exit;
}
bcopy(&dccb[1], &dpt->board_data, sizeof(dpt->board_data));
dpt_detect_cache(dpt, dccb, sg_map->sg_physaddr + sizeof(dpt_sp_t),
(u_int8_t *)&dccb[1]);
switch (ntohl(conf.splen)) {
case DPT_EATA_REVA:
dpt->EATA_revision = 'a';
break;
case DPT_EATA_REVB:
dpt->EATA_revision = 'b';
break;
case DPT_EATA_REVC:
dpt->EATA_revision = 'c';
break;
case DPT_EATA_REVZ:
dpt->EATA_revision = 'z';
break;
default:
dpt->EATA_revision = '?';
}
dpt->max_id = conf.MAX_ID;
dpt->max_lun = conf.MAX_LUN;
dpt->irq = conf.IRQ;
dpt->dma_channel = (8 - conf.DMA_channel) & 7;
dpt->channels = conf.MAX_CHAN + 1;
dpt->state |= DPT_HA_OK;
if (conf.SECOND)
dpt->primary = FALSE;
else
dpt->primary = TRUE;
dpt->more_support = conf.MORE_support;
if (strncmp(dpt->board_data.firmware, "07G0", 4) >= 0)
dpt->immediate_support = 1;
else
dpt->immediate_support = 0;
dpt->broken_INQUIRY = FALSE;
dpt->cplen = ntohl(conf.cplen);
dpt->cppadlen = ntohs(conf.cppadlen);
dpt->max_dccbs = ntohs(conf.queuesiz);
if (dpt->max_dccbs > 256) {
device_printf(dpt->dev, "Max CCBs reduced from %d to "
"256 due to tag algorithm\n", dpt->max_dccbs);
dpt->max_dccbs = 256;
}
dpt->hostid[0] = conf.scsi_id0;
dpt->hostid[1] = conf.scsi_id1;
dpt->hostid[2] = conf.scsi_id2;
if (conf.SG_64K)
dpt->sgsize = 8192;
else
dpt->sgsize = ntohs(conf.SGsiz);
/* We can only get 64k buffers, so don't bother to waste space. */
if (dpt->sgsize < 17 || dpt->sgsize > 32)
dpt->sgsize = 32;
if (dpt->sgsize > dpt_max_segs)
dpt->sgsize = dpt_max_segs;
/* DMA tag for mapping buffers into device visible space. */
if (bus_dma_tag_create( /* parent */ dpt->parent_dmat,
/* alignment */ 1,
/* boundary */ 0,
/* lowaddr */ BUS_SPACE_MAXADDR,
/* highaddr */ BUS_SPACE_MAXADDR,
/* filter */ NULL,
/* filterarg */ NULL,
/* maxsize */ MAXBSIZE,
/* nsegments */ dpt->sgsize,
/* maxsegsz */ BUS_SPACE_MAXSIZE_32BIT,
/* flags */ BUS_DMA_ALLOCNOW,
/* lockfunc */ busdma_lock_mutex,
/* lockarg */ &dpt->lock,
&dpt->buffer_dmat) != 0) {
device_printf(dpt->dev,
"bus_dma_tag_create(...,dpt->buffer_dmat) failed\n");
goto error_exit;
}
dpt->init_level++;
/* DMA tag for our ccb structures and interrupt status packet */
if (bus_dma_tag_create( /* parent */ dpt->parent_dmat,
/* alignment */ 1,
/* boundary */ 0,
/* lowaddr */ BUS_SPACE_MAXADDR,
/* highaddr */ BUS_SPACE_MAXADDR,
/* filter */ NULL,
/* filterarg */ NULL,
/* maxsize */ (dpt->max_dccbs *
sizeof(struct dpt_ccb)) +
sizeof(dpt_sp_t),
/* nsegments */ 1,
/* maxsegsz */ BUS_SPACE_MAXSIZE_32BIT,
/* flags */ 0,
/* lockfunc */ NULL,
/* lockarg */ NULL,
&dpt->dccb_dmat) != 0) {
device_printf(dpt->dev,
"bus_dma_tag_create(...,dpt->dccb_dmat) failed\n");
goto error_exit;
}
dpt->init_level++;
/* Allocation for our ccbs and interrupt status packet */
if (bus_dmamem_alloc(dpt->dccb_dmat, (void **)&dpt->dpt_dccbs,
BUS_DMA_NOWAIT, &dpt->dccb_dmamap) != 0) {
device_printf(dpt->dev,
"bus_dmamem_alloc(dpt->dccb_dmat,...) failed\n");
goto error_exit;
}
dpt->init_level++;
/* And permanently map them */
bus_dmamap_load(dpt->dccb_dmat, dpt->dccb_dmamap,
dpt->dpt_dccbs,
(dpt->max_dccbs * sizeof(struct dpt_ccb))
+ sizeof(dpt_sp_t),
dptmapmem, &dpt->dpt_ccb_busbase, /*flags*/0);
/* Clear them out. */
bzero(dpt->dpt_dccbs,
(dpt->max_dccbs * sizeof(struct dpt_ccb)) + sizeof(dpt_sp_t));
dpt->dpt_ccb_busend = dpt->dpt_ccb_busbase;
dpt->sp = (dpt_sp_t*)&dpt->dpt_dccbs[dpt->max_dccbs];
dpt->sp_physaddr = dpt->dpt_ccb_busbase
+ (dpt->max_dccbs * sizeof(dpt_ccb_t));
dpt->init_level++;
/* Allocate our first batch of ccbs */
if (dptallocccbs(dpt) == 0) {
device_printf(dpt->dev, "dptallocccbs(dpt) == 0\n");
mtx_unlock(&dpt->lock);
return (2);
}
/* Prepare for Target Mode */
dpt->target_mode_enabled = 1;
/* Nuke excess spaces from inquiry information */
strp = dpt->board_data.vendor;
for (i = 0; i < sizeof(string_sizes); i++) {
index = string_sizes[i] - 1;
while (index && (strp[index] == ' '))
strp[index--] = '\0';
strp += string_sizes[i];
}
device_printf(dpt->dev, "%.8s %.16s FW Rev. %.4s, ",
dpt->board_data.vendor,
dpt->board_data.modelNum, dpt->board_data.firmware);
printf("%d channel%s, ", dpt->channels, dpt->channels > 1 ? "s" : "");
if (dpt->cache_type != DPT_NO_CACHE
&& dpt->cache_size != 0) {
printf("%s Cache, ",
dpt->cache_type == DPT_CACHE_WRITETHROUGH
? "Write-Through" : "Write-Back");
}
printf("%d CCBs\n", dpt->max_dccbs);
mtx_unlock(&dpt->lock);
return (0);
error_exit:
mtx_unlock(&dpt->lock);
return (1);
}
int
dpt_attach(dpt_softc_t *dpt)
{
struct cam_devq *devq;
int i;
/*
* Create the device queue for our SIM.
*/
devq = cam_simq_alloc(dpt->max_dccbs);
if (devq == NULL)
return (0);
mtx_lock(&dpt->lock);
for (i = 0; i < dpt->channels; i++) {
/*
* Construct our SIM entry
*/
dpt->sims[i] = cam_sim_alloc(dpt_action, dpt_poll, "dpt",
dpt, device_get_unit(dpt->dev), &dpt->lock,
/*untagged*/2,
/*tagged*/dpt->max_dccbs, devq);
if (dpt->sims[i] == NULL) {
if (i == 0)
cam_simq_free(devq);
else
printf( "%s(): Unable to attach bus %d "
"due to resource shortage\n",
__func__, i);
break;
}
if (xpt_bus_register(dpt->sims[i], dpt->dev, i) != CAM_SUCCESS){
cam_sim_free(dpt->sims[i], /*free_devq*/i == 0);
dpt->sims[i] = NULL;
break;
}
if (xpt_create_path(&dpt->paths[i], /*periph*/NULL,
cam_sim_path(dpt->sims[i]),
CAM_TARGET_WILDCARD,
CAM_LUN_WILDCARD) != CAM_REQ_CMP) {
xpt_bus_deregister(cam_sim_path(dpt->sims[i]));
cam_sim_free(dpt->sims[i], /*free_devq*/i == 0);
dpt->sims[i] = NULL;
break;
}
}
mtx_unlock(&dpt->lock);
if (i > 0)
EVENTHANDLER_REGISTER(shutdown_final, dptshutdown,
dpt, SHUTDOWN_PRI_DEFAULT);
return (i);
}
int
dpt_detach (device_t dev)
{
struct dpt_softc * dpt;
int i;
dpt = device_get_softc(dev);
mtx_lock(&dpt->lock);
for (i = 0; i < dpt->channels; i++) {
#if 0
xpt_async(AC_LOST_DEVICE, dpt->paths[i], NULL);
#endif
xpt_free_path(dpt->paths[i]);
xpt_bus_deregister(cam_sim_path(dpt->sims[i]));
cam_sim_free(dpt->sims[i], /*free_devq*/TRUE);
}
mtx_unlock(&dpt->lock);
dptshutdown((void *)dpt, SHUTDOWN_PRI_DEFAULT);
dpt_release_resources(dev);
dpt_free(dpt);
return (0);
}
/*
* This is the interrupt handler for the DPT driver.
*/
void
dpt_intr(void *arg)
{
dpt_softc_t *dpt;
dpt = arg;
mtx_lock(&dpt->lock);
dpt_intr_locked(dpt);
mtx_unlock(&dpt->lock);
}
void
dpt_intr_locked(dpt_softc_t *dpt)
{
dpt_ccb_t *dccb;
union ccb *ccb;
u_int status;
u_int aux_status;
u_int hba_stat;
u_int scsi_stat;
u_int32_t residue_len; /* Number of bytes not transferred */
/* First order of business is to check if this interrupt is for us */
while (((aux_status = dpt_inb(dpt, HA_RAUXSTAT)) & HA_AIRQ) != 0) {
/*
* What we want to do now, is to capture the status, all of it,
* move it where it belongs, wake up whoever sleeps waiting to
* process this result, and get out of here.
*/
if (dpt->sp->ccb_busaddr < dpt->dpt_ccb_busbase
|| dpt->sp->ccb_busaddr >= dpt->dpt_ccb_busend) {
device_printf(dpt->dev,
"Encountered bogus status packet\n");
status = dpt_inb(dpt, HA_RSTATUS);
return;
}
dccb = dptccbptov(dpt, dpt->sp->ccb_busaddr);
dpt->sp->ccb_busaddr = ~0;
/* Ignore status packets with EOC not set */
if (dpt->sp->EOC == 0) {
device_printf(dpt->dev,
"ERROR: Request %d received with "
"clear EOC.\n Marking as LOST.\n",
dccb->transaction_id);
#ifdef DPT_HANDLE_TIMEOUTS
dccb->state |= DPT_CCB_STATE_MARKED_LOST;
#endif
/* This CLEARS the interrupt! */
status = dpt_inb(dpt, HA_RSTATUS);
continue;
}
dpt->sp->EOC = 0;
/*
* Double buffer the status information so the hardware can
* work on updating the status packet while we decifer the
* one we were just interrupted for.
* According to Mark Salyzyn, we only need few pieces of it.
*/
hba_stat = dpt->sp->hba_stat;
scsi_stat = dpt->sp->scsi_stat;
residue_len = dpt->sp->residue_len;
/* Clear interrupts, check for error */
if ((status = dpt_inb(dpt, HA_RSTATUS)) & HA_SERROR) {
/*
* Error Condition. Check for magic cookie. Exit
* this test on earliest sign of non-reset condition
*/
/* Check that this is not a board reset interrupt */
if (dpt_just_reset(dpt)) {
device_printf(dpt->dev, "HBA rebooted.\n"
" All transactions should be "
"resubmitted\n");
device_printf(dpt->dev,
">>---->> This is incomplete, "
"fix me.... <<----<<");
panic("DPT Rebooted");
}
}
/* Process CCB */
ccb = dccb->ccb;
callout_stop(&dccb->timer);
if ((ccb->ccb_h.flags & CAM_DIR_MASK) != CAM_DIR_NONE) {
bus_dmasync_op_t op;
if ((ccb->ccb_h.flags & CAM_DIR_MASK) == CAM_DIR_IN)
op = BUS_DMASYNC_POSTREAD;
else
op = BUS_DMASYNC_POSTWRITE;
bus_dmamap_sync(dpt->buffer_dmat, dccb->dmamap, op);
bus_dmamap_unload(dpt->buffer_dmat, dccb->dmamap);
}
/* Common Case inline... */
if (hba_stat == HA_NO_ERROR) {
ccb->csio.scsi_status = scsi_stat;
ccb->ccb_h.status = 0;
switch (scsi_stat) {
case SCSI_STATUS_OK:
ccb->ccb_h.status |= CAM_REQ_CMP;
break;
case SCSI_STATUS_CHECK_COND:
case SCSI_STATUS_CMD_TERMINATED:
bcopy(&dccb->sense_data, &ccb->csio.sense_data,
ccb->csio.sense_len);
ccb->ccb_h.status |= CAM_AUTOSNS_VALID;
/* FALLTHROUGH */
default:
ccb->ccb_h.status |= CAM_SCSI_STATUS_ERROR;
/* XXX Freeze DevQ */
break;
}
ccb->csio.resid = residue_len;
dptfreeccb(dpt, dccb);
xpt_done(ccb);
} else {
dptprocesserror(dpt, dccb, ccb, hba_stat, scsi_stat,
residue_len);
}
}
}
static void
dptprocesserror(dpt_softc_t *dpt, dpt_ccb_t *dccb, union ccb *ccb,
u_int hba_stat, u_int scsi_stat, u_int32_t resid)
{
ccb->csio.resid = resid;
switch (hba_stat) {
case HA_ERR_SEL_TO:
ccb->ccb_h.status = CAM_SEL_TIMEOUT;
break;
case HA_ERR_CMD_TO:
ccb->ccb_h.status = CAM_CMD_TIMEOUT;
break;
case HA_SCSIBUS_RESET:
case HA_HBA_POWER_UP: /* Similar effect to a bus reset??? */
ccb->ccb_h.status = CAM_SCSI_BUS_RESET;
break;
case HA_CP_ABORTED:
case HA_CP_RESET: /* XXX ??? */
case HA_CP_ABORT_NA: /* XXX ??? */
case HA_CP_RESET_NA: /* XXX ??? */
if ((ccb->ccb_h.status & CAM_STATUS_MASK) == CAM_REQ_INPROG)
ccb->ccb_h.status = CAM_REQ_ABORTED;
break;
case HA_PCI_PARITY:
case HA_PCI_MABORT:
case HA_PCI_TABORT:
case HA_PCI_STABORT:
case HA_BUS_PARITY:
case HA_PARITY_ERR:
case HA_ECC_ERR:
ccb->ccb_h.status = CAM_UNCOR_PARITY;
break;
case HA_UNX_MSGRJCT:
ccb->ccb_h.status = CAM_MSG_REJECT_REC;
break;
case HA_UNX_BUSPHASE:
ccb->ccb_h.status = CAM_SEQUENCE_FAIL;
break;
case HA_UNX_BUS_FREE:
ccb->ccb_h.status = CAM_UNEXP_BUSFREE;
break;
case HA_SCSI_HUNG:
case HA_RESET_STUCK:
/*
* Dead??? Can the controller get unstuck
* from these conditions
*/
ccb->ccb_h.status = CAM_NO_HBA;
break;
case HA_RSENSE_FAIL:
ccb->ccb_h.status = CAM_AUTOSENSE_FAIL;
break;
default:
device_printf(dpt->dev, "Undocumented Error %x\n", hba_stat);
printf("Please mail this message to shimon@simon-shapiro.org\n");
ccb->ccb_h.status = CAM_REQ_CMP_ERR;
break;
}
dptfreeccb(dpt, dccb);
xpt_done(ccb);
}
static void
dpttimeout(void *arg)
{
struct dpt_ccb *dccb;
union ccb *ccb;
struct dpt_softc *dpt;
dccb = (struct dpt_ccb *)arg;
ccb = dccb->ccb;
dpt = (struct dpt_softc *)ccb->ccb_h.ccb_dpt_ptr;
mtx_assert(&dpt->lock, MA_OWNED);
xpt_print_path(ccb->ccb_h.path);
printf("CCB %p - timed out\n", (void *)dccb);
/*
* Try to clear any pending jobs. FreeBSD will lose interrupts,
* leaving the controller suspended, and commands timed-out.
* By calling the interrupt handler, any command thus stuck will be
* completed.
*/
dpt_intr_locked(dpt);
if ((dccb->state & DCCB_ACTIVE) == 0) {
xpt_print_path(ccb->ccb_h.path);
printf("CCB %p - timed out CCB already completed\n",
(void *)dccb);
return;
}
/* Abort this particular command. Leave all others running */
dpt_send_immediate(dpt, &dccb->eata_ccb, dccb->eata_ccb.cp_busaddr,
/*retries*/20000, EATA_SPECIFIC_ABORT, 0, 0);
ccb->ccb_h.status = CAM_CMD_TIMEOUT;
}
/*
* Shutdown the controller and ensure that the cache is completely flushed.
* Called from the shutdown_final event after all disk access has completed.
*/
static void
dptshutdown(void *arg, int howto)
{
dpt_softc_t *dpt;
dpt = (dpt_softc_t *)arg;
device_printf(dpt->dev,
"Shutting down (mode %x) HBA. Please wait...\n", howto);
/*
* What we do for a shutdown, is give the DPT early power loss warning
*/
mtx_lock(&dpt->lock);
dpt_send_immediate(dpt, NULL, 0, EATA_POWER_OFF_WARN, 0, 0, 0);
mtx_unlock(&dpt->lock);
DELAY(1000 * 1000 * 5);
device_printf(dpt->dev, "Controller was warned of shutdown and is now "
"disabled\n");
}
/*============================================================================*/
#if 0
#ifdef DPT_RESET_HBA
/*
** Function name : dpt_reset_hba
**
** Description : Reset the HBA and properly discard all pending work
** Input : Softc
** Output : Nothing
*/
static void
dpt_reset_hba(dpt_softc_t *dpt)
{
eata_ccb_t *ccb;
dpt_ccb_t dccb, *dccbp;
int result;
struct scsi_xfer *xs;
mtx_assert(&dpt->lock, MA_OWNED);
/* Prepare a control block. The SCSI command part is immaterial */
dccb.xs = NULL;
dccb.flags = 0;
dccb.state = DPT_CCB_STATE_NEW;
dccb.std_callback = NULL;
dccb.wrbuff_callback = NULL;
ccb = &dccb.eata_ccb;
ccb->CP_OpCode = EATA_CMD_RESET;
ccb->SCSI_Reset = 0;
ccb->HBA_Init = 1;
ccb->Auto_Req_Sen = 1;
ccb->cp_id = 0; /* Should be ignored */
ccb->DataIn = 1;
ccb->DataOut = 0;
ccb->Interpret = 1;
ccb->reqlen = htonl(sizeof(struct scsi_sense_data));
ccb->cp_statDMA = htonl(vtophys(&ccb->cp_statDMA));
ccb->cp_reqDMA = htonl(vtophys(&ccb->cp_reqDMA));
ccb->cp_viraddr = (u_int32_t) & ccb;
ccb->cp_msg[0] = HA_IDENTIFY_MSG | HA_DISCO_RECO;
ccb->cp_scsi_cmd = 0; /* Should be ignored */
/* Lock up the submitted queue. We are very persistant here */
while (dpt->queue_status & DPT_SUBMITTED_QUEUE_ACTIVE) {
DELAY(100);
}
dpt->queue_status |= DPT_SUBMITTED_QUEUE_ACTIVE;
/* Send the RESET message */
if ((result = dpt_send_eata_command(dpt, &dccb.eata_ccb,
EATA_CMD_RESET, 0, 0, 0, 0)) != 0) {
device_printf(dpt->dev, "Failed to send the RESET message.\n"
" Trying cold boot (ouch!)\n");
if ((result = dpt_send_eata_command(dpt, &dccb.eata_ccb,
EATA_COLD_BOOT, 0, 0,
0, 0)) != 0) {
panic("%s: Faild to cold boot the HBA\n",
device_get_nameunit(dpt->dev));
}
#ifdef DPT_MEASURE_PERFORMANCE
dpt->performance.cold_boots++;
#endif /* DPT_MEASURE_PERFORMANCE */
}
#ifdef DPT_MEASURE_PERFORMANCE
dpt->performance.warm_starts++;
#endif /* DPT_MEASURE_PERFORMANCE */
device_printf(dpt->dev,
"Aborting pending requests. O/S should re-submit\n");
while ((dccbp = TAILQ_FIRST(&dpt->completed_ccbs)) != NULL) {
struct scsi_xfer *xs = dccbp->xs;
/* Not all transactions have xs structs */
if (xs != NULL) {
/* Tell the kernel proper this did not complete well */
xs->error |= XS_SELTIMEOUT;
xs->flags |= SCSI_ITSDONE;
scsi_done(xs);
}
dpt_Qremove_submitted(dpt, dccbp);
/* Remember, Callbacks are NOT in the standard queue */
if (dccbp->std_callback != NULL) {
(dccbp->std_callback)(dpt, dccbp->eata_ccb.cp_channel,
dccbp);
} else {
dpt_Qpush_free(dpt, dccbp);
}
}
device_printf(dpt->dev, "reset done aborting all pending commands\n");
dpt->queue_status &= ~DPT_SUBMITTED_QUEUE_ACTIVE;
}
#endif /* DPT_RESET_HBA */
/*
* Build a Command Block for target mode READ/WRITE BUFFER,
* with the ``sync'' bit ON.
*
* Although the length and offset are 24 bit fields in the command, they cannot
* exceed 8192 bytes, so we take them as short integers andcheck their range.
* If they are sensless, we round them to zero offset, maximum length and
* complain.
*/
static void
dpt_target_ccb(dpt_softc_t * dpt, int bus, u_int8_t target, u_int8_t lun,
dpt_ccb_t * ccb, int mode, u_int8_t command,
u_int16_t length, u_int16_t offset)
{
eata_ccb_t *cp;
mtx_assert(&dpt->lock, MA_OWNED);
if ((length + offset) > DPT_MAX_TARGET_MODE_BUFFER_SIZE) {
device_printf(dpt->dev,
"Length of %d, and offset of %d are wrong\n",
length, offset);
length = DPT_MAX_TARGET_MODE_BUFFER_SIZE;
offset = 0;
}
ccb->xs = NULL;
ccb->flags = 0;
ccb->state = DPT_CCB_STATE_NEW;
ccb->std_callback = (ccb_callback) dpt_target_done;
ccb->wrbuff_callback = NULL;
cp = &ccb->eata_ccb;
cp->CP_OpCode = EATA_CMD_DMA_SEND_CP;
cp->SCSI_Reset = 0;
cp->HBA_Init = 0;
cp->Auto_Req_Sen = 1;
cp->cp_id = target;
cp->DataIn = 1;
cp->DataOut = 0;
cp->Interpret = 0;
cp->reqlen = htonl(sizeof(struct scsi_sense_data));
cp->cp_statDMA = htonl(vtophys(&cp->cp_statDMA));
cp->cp_reqDMA = htonl(vtophys(&cp->cp_reqDMA));
cp->cp_viraddr = (u_int32_t) & ccb;
cp->cp_msg[0] = HA_IDENTIFY_MSG | HA_DISCO_RECO;
cp->cp_scsi_cmd = command;
cp->cp_cdb[1] = (u_int8_t) (mode & SCSI_TM_MODE_MASK);
cp->cp_lun = lun; /* Order is important here! */
cp->cp_cdb[2] = 0x00; /* Buffer Id, only 1 :-( */
cp->cp_cdb[3] = (length >> 16) & 0xFF; /* Buffer offset MSB */
cp->cp_cdb[4] = (length >> 8) & 0xFF;
cp->cp_cdb[5] = length & 0xFF;
cp->cp_cdb[6] = (length >> 16) & 0xFF; /* Length MSB */
cp->cp_cdb[7] = (length >> 8) & 0xFF;
cp->cp_cdb[8] = length & 0xFF; /* Length LSB */
cp->cp_cdb[9] = 0; /* No sync, no match bits */
/*
* This could be optimized to live in dpt_register_buffer.
* We keep it here, just in case the kernel decides to reallocate pages
*/
if (dpt_scatter_gather(dpt, ccb, DPT_RW_BUFFER_SIZE,
dpt->rw_buffer[bus][target][lun])) {
device_printf(dpt->dev, "Failed to setup Scatter/Gather for "
"Target-Mode buffer\n");
}
}
/* Setup a target mode READ command */
static void
dpt_set_target(int redo, dpt_softc_t * dpt,
u_int8_t bus, u_int8_t target, u_int8_t lun, int mode,
u_int16_t length, u_int16_t offset, dpt_ccb_t * ccb)
{
mtx_assert(&dpt->lock, MA_OWNED);
if (dpt->target_mode_enabled) {
if (!redo)
dpt_target_ccb(dpt, bus, target, lun, ccb, mode,
SCSI_TM_READ_BUFFER, length, offset);
ccb->transaction_id = ++dpt->commands_processed;
#ifdef DPT_MEASURE_PERFORMANCE
dpt->performance.command_count[ccb->eata_ccb.cp_scsi_cmd]++;
ccb->command_started = microtime_now;
#endif
dpt_Qadd_waiting(dpt, ccb);
dpt_sched_queue(dpt);
} else {
device_printf(dpt->dev,
"Target Mode Request, but Target Mode is OFF\n");
}
}
/*
* Schedule a buffer to be sent to another target.
* The work will be scheduled and the callback provided will be called when
* the work is actually done.
*
* Please NOTE: ``Anyone'' can send a buffer, but only registered clients
* get notified of receipt of buffers.
*/
int
dpt_send_buffer(int unit, u_int8_t channel, u_int8_t target, u_int8_t lun,
u_int8_t mode, u_int16_t length, u_int16_t offset, void *data,
buff_wr_done callback)
{
dpt_softc_t *dpt;
dpt_ccb_t *ccb = NULL;
/* This is an external call. Be a bit paranoid */
dpt = devclass_get_device(dpt_devclass, unit);
if (dpt == NULL)
return (INVALID_UNIT);
mtx_lock(&dpt->lock);
if (dpt->target_mode_enabled) {
if ((channel >= dpt->channels) || (target > dpt->max_id) ||
(lun > dpt->max_lun)) {
mtx_unlock(&dpt->lock);
return (INVALID_SENDER);
}
if ((dpt->rw_buffer[channel][target][lun] == NULL) ||
(dpt->buffer_receiver[channel][target][lun] == NULL)) {
mtx_unlock(&dpt->lock);
return (NOT_REGISTERED);
}
/* Process the free list */
if ((TAILQ_EMPTY(&dpt->free_ccbs)) && dpt_alloc_freelist(dpt)) {
device_printf(dpt->dev,
"ERROR: Cannot allocate any more free CCB's.\n"
" Please try later\n");
mtx_unlock(&dpt->lock);
return (NO_RESOURCES);
}
/* Now grab the newest CCB */
if ((ccb = dpt_Qpop_free(dpt)) == NULL) {
mtx_unlock(&dpt->lock);
panic("%s: Got a NULL CCB from pop_free()\n",
device_get_nameunit(dpt->dev));
}
bcopy(dpt->rw_buffer[channel][target][lun] + offset, data, length);
dpt_target_ccb(dpt, channel, target, lun, ccb, mode,
SCSI_TM_WRITE_BUFFER,
length, offset);
ccb->std_callback = (ccb_callback) callback; /* Potential trouble */
ccb->transaction_id = ++dpt->commands_processed;
#ifdef DPT_MEASURE_PERFORMANCE
dpt->performance.command_count[ccb->eata_ccb.cp_scsi_cmd]++;
ccb->command_started = microtime_now;
#endif
dpt_Qadd_waiting(dpt, ccb);
dpt_sched_queue(dpt);
mtx_unlock(&dpt->lock);
return (0);
}
mtx_unlock(&dpt->lock);
return (DRIVER_DOWN);
}
static void
dpt_target_done(dpt_softc_t * dpt, int bus, dpt_ccb_t * ccb)
{
eata_ccb_t *cp;
cp = &ccb->eata_ccb;
/*
* Remove the CCB from the waiting queue.
* We do NOT put it back on the free, etc., queues as it is a special
* ccb, owned by the dpt_softc of this unit.
*/
dpt_Qremove_completed(dpt, ccb);
#define br_channel (ccb->eata_ccb.cp_channel)
#define br_target (ccb->eata_ccb.cp_id)
#define br_lun (ccb->eata_ccb.cp_LUN)
#define br_index [br_channel][br_target][br_lun]
#define read_buffer_callback (dpt->buffer_receiver br_index )
#define read_buffer (dpt->rw_buffer[br_channel][br_target][br_lun])
#define cb(offset) (ccb->eata_ccb.cp_cdb[offset])
#define br_offset ((cb(3) << 16) | (cb(4) << 8) | cb(5))
#define br_length ((cb(6) << 16) | (cb(7) << 8) | cb(8))
/* Different reasons for being here, you know... */
switch (ccb->eata_ccb.cp_scsi_cmd) {
case SCSI_TM_READ_BUFFER:
if (read_buffer_callback != NULL) {
/* This is a buffer generated by a kernel process */
read_buffer_callback(device_get_unit(dpt->dev),
br_channel, br_target, br_lun,
read_buffer,
br_offset, br_length);
} else {
/*
* This is a buffer waited for by a user (sleeping)
* command
*/
wakeup(ccb);
}
/* We ALWAYS re-issue the same command; args are don't-care */
dpt_set_target(1, 0, 0, 0, 0, 0, 0, 0, 0);
break;
case SCSI_TM_WRITE_BUFFER:
(ccb->wrbuff_callback) (device_get_unit(dpt->dev), br_channel,
br_target, br_offset, br_length,
br_lun, ccb->status_packet.hba_stat);
break;
default:
device_printf(dpt->dev,
"%s is an unsupported command for target mode\n",
scsi_cmd_name(ccb->eata_ccb.cp_scsi_cmd));
}
dpt->target_ccb[br_channel][br_target][br_lun] = NULL;
dpt_Qpush_free(dpt, ccb);
}
/*
* Use this function to register a client for a buffer read target operation.
* The function you register will be called every time a buffer is received
* by the target mode code.
*/
dpt_rb_t
dpt_register_buffer(int unit, u_int8_t channel, u_int8_t target, u_int8_t lun,
u_int8_t mode, u_int16_t length, u_int16_t offset,
dpt_rec_buff callback, dpt_rb_op_t op)
{
dpt_softc_t *dpt;
dpt_ccb_t *ccb = NULL;
int ospl;
dpt = devclass_get_device(dpt_devclass, unit);
if (dpt == NULL)
return (INVALID_UNIT);
mtx_lock(&dpt->lock);
if (dpt->state & DPT_HA_SHUTDOWN_ACTIVE) {
mtx_unlock(&dpt->lock);
return (DRIVER_DOWN);
}
if ((channel > (dpt->channels - 1)) || (target > (dpt->max_id - 1)) ||
(lun > (dpt->max_lun - 1))) {
mtx_unlock(&dpt->lock);
return (INVALID_SENDER);
}
if (dpt->buffer_receiver[channel][target][lun] == NULL) {
if (op == REGISTER_BUFFER) {
/* Assign the requested callback */
dpt->buffer_receiver[channel][target][lun] = callback;
/* Get a CCB */
/* Process the free list */
if ((TAILQ_EMPTY(&dpt->free_ccbs)) && dpt_alloc_freelist(dpt)) {
device_printf(dpt->dev,
"ERROR: Cannot allocate any more free CCB's.\n"
" Please try later\n");
mtx_unlock(&dpt->lock);
return (NO_RESOURCES);
}
/* Now grab the newest CCB */
if ((ccb = dpt_Qpop_free(dpt)) == NULL) {
mtx_unlock(&dpt->lock);
panic("%s: Got a NULL CCB from pop_free()\n",
device_get_nameunit(dpt->dev));
}
/* Clean up the leftover of the previous tenant */
ccb->status = DPT_CCB_STATE_NEW;
dpt->target_ccb[channel][target][lun] = ccb;
dpt->rw_buffer[channel][target][lun] =
malloc(DPT_RW_BUFFER_SIZE, M_DEVBUF, M_NOWAIT);
if (dpt->rw_buffer[channel][target][lun] == NULL) {
device_printf(dpt->dev, "Failed to allocate "
"Target-Mode buffer\n");
dpt_Qpush_free(dpt, ccb);
mtx_unlock(&dpt->lock);
return (NO_RESOURCES);
}
dpt_set_target(0, dpt, channel, target, lun, mode,
length, offset, ccb);
mtx_unlock(&dpt->lock);
return (SUCCESSFULLY_REGISTERED);
} else {
mtx_unlock(&dpt->lock);
return (NOT_REGISTERED);
}
} else {
if (op == REGISTER_BUFFER) {
if (dpt->buffer_receiver[channel][target][lun] == callback) {
mtx_unlock(&dpt->lock);
return (ALREADY_REGISTERED);
} else {
mtx_unlock(&dpt->lock);
return (REGISTERED_TO_ANOTHER);
}
} else {
if (dpt->buffer_receiver[channel][target][lun] == callback) {
dpt->buffer_receiver[channel][target][lun] = NULL;
dpt_Qpush_free(dpt, ccb);
free(dpt->rw_buffer[channel][target][lun], M_DEVBUF);
mtx_unlock(&dpt->lock);
return (SUCCESSFULLY_REGISTERED);
} else {
mtx_unlock(&dpt->lock);
return (INVALID_CALLBACK);
}
}
}
mtx_unlock(&dpt->lock);
}
/* Return the state of the blinking DPT LED's */
u_int8_t
dpt_blinking_led(dpt_softc_t * dpt)
{
int ndx;
u_int32_t state;
u_int32_t previous;
u_int8_t result;
mtx_assert(&dpt->lock, MA_OWNED);
result = 0;
for (ndx = 0, state = 0, previous = 0;
(ndx < 10) && (state != previous);
ndx++) {
previous = state;
state = dpt_inl(dpt, 1);
}
if ((state == previous) && (state == DPT_BLINK_INDICATOR))
result = dpt_inb(dpt, 5);
return (result);
}
/*
* Execute a command which did not come from the kernel's SCSI layer.
* The only way to map user commands to bus and target is to comply with the
* standard DPT wire-down scheme:
*/
int
dpt_user_cmd(dpt_softc_t * dpt, eata_pt_t * user_cmd,
caddr_t cmdarg, int minor_no)
{
dpt_ccb_t *ccb;
void *data;
int channel, target, lun;
int huh;
int result;
int submitted;
mtx_assert(&dpt->lock, MA_OWNED);
data = NULL;
channel = minor2hba(minor_no);
target = minor2target(minor_no);
lun = minor2lun(minor_no);
if ((channel > (dpt->channels - 1))
|| (target > dpt->max_id)
|| (lun > dpt->max_lun))
return (ENXIO);
if (target == dpt->sc_scsi_link[channel].adapter_targ) {
/* This one is for the controller itself */
if ((user_cmd->eataID[0] != 'E')
|| (user_cmd->eataID[1] != 'A')
|| (user_cmd->eataID[2] != 'T')
|| (user_cmd->eataID[3] != 'A')) {
return (ENXIO);
}
}
/* Get a DPT CCB, so we can prepare a command */
/* Process the free list */
if ((TAILQ_EMPTY(&dpt->free_ccbs)) && dpt_alloc_freelist(dpt)) {
device_printf(dpt->dev,
"ERROR: Cannot allocate any more free CCB's.\n"
" Please try later\n");
return (EFAULT);
}
/* Now grab the newest CCB */
if ((ccb = dpt_Qpop_free(dpt)) == NULL) {
panic("%s: Got a NULL CCB from pop_free()\n",
device_get_nameunit(dpt->dev));
} else {
/* Clean up the leftover of the previous tenant */
ccb->status = DPT_CCB_STATE_NEW;
}
bcopy((caddr_t) & user_cmd->command_packet, (caddr_t) & ccb->eata_ccb,
sizeof(eata_ccb_t));
/* We do not want to do user specified scatter/gather. Why?? */
if (ccb->eata_ccb.scatter == 1)
return (EINVAL);
ccb->eata_ccb.Auto_Req_Sen = 1;
ccb->eata_ccb.reqlen = htonl(sizeof(struct scsi_sense_data));
ccb->eata_ccb.cp_datalen = htonl(sizeof(ccb->eata_ccb.cp_datalen));
ccb->eata_ccb.cp_dataDMA = htonl(vtophys(ccb->eata_ccb.cp_dataDMA));
ccb->eata_ccb.cp_statDMA = htonl(vtophys(&ccb->eata_ccb.cp_statDMA));
ccb->eata_ccb.cp_reqDMA = htonl(vtophys(&ccb->eata_ccb.cp_reqDMA));
ccb->eata_ccb.cp_viraddr = (u_int32_t) & ccb;
if (ccb->eata_ccb.DataIn || ccb->eata_ccb.DataOut) {
/* Data I/O is involved in this command. Alocate buffer */
if (ccb->eata_ccb.cp_datalen > PAGE_SIZE) {
data = contigmalloc(ccb->eata_ccb.cp_datalen,
M_TEMP, M_WAITOK, 0, ~0,
ccb->eata_ccb.cp_datalen,
0x10000);
} else {
data = malloc(ccb->eata_ccb.cp_datalen, M_TEMP,
M_WAITOK);
}
if (data == NULL) {
device_printf(dpt->dev, "Cannot allocate %d bytes "
"for EATA command\n",
ccb->eata_ccb.cp_datalen);
return (EFAULT);
}
#define usr_cmd_DMA (caddr_t)user_cmd->command_packet.cp_dataDMA
if (ccb->eata_ccb.DataIn == 1) {
if (copyin(usr_cmd_DMA,
data, ccb->eata_ccb.cp_datalen) == -1)
return (EFAULT);
}
} else {
/* No data I/O involved here. Make sure the DPT knows that */
ccb->eata_ccb.cp_datalen = 0;
data = NULL;
}
if (ccb->eata_ccb.FWNEST == 1)
ccb->eata_ccb.FWNEST = 0;
if (ccb->eata_ccb.cp_datalen != 0) {
if (dpt_scatter_gather(dpt, ccb, ccb->eata_ccb.cp_datalen,
data) != 0) {
if (data != NULL)
free(data, M_TEMP);
return (EFAULT);
}
}
/**
* We are required to quiet a SCSI bus.
* since we do not queue comands on a bus basis,
* we wait for ALL commands on a controller to complete.
* In the mean time, sched_queue() will not schedule new commands.
*/
if ((ccb->eata_ccb.cp_cdb[0] == MULTIFUNCTION_CMD)
&& (ccb->eata_ccb.cp_cdb[2] == BUS_QUIET)) {
/* We wait for ALL traffic for this HBa to subside */
dpt->state |= DPT_HA_QUIET;
while ((submitted = dpt->submitted_ccbs_count) != 0) {
huh = mtx_sleep((void *) dpt, &dpt->lock,
PCATCH | PRIBIO, "dptqt", 100 * hz);
switch (huh) {
case 0:
/* Wakeup call received */
break;
case EWOULDBLOCK:
/* Timer Expired */
break;
default:
/* anything else */
break;
}
}
}
/* Resume normal operation */
if ((ccb->eata_ccb.cp_cdb[0] == MULTIFUNCTION_CMD)
&& (ccb->eata_ccb.cp_cdb[2] == BUS_UNQUIET)) {
dpt->state &= ~DPT_HA_QUIET;
}
/**
* Schedule the command and submit it.
* We bypass dpt_sched_queue, as it will block on DPT_HA_QUIET
*/
ccb->xs = NULL;
ccb->flags = 0;
ccb->eata_ccb.Auto_Req_Sen = 1; /* We always want this feature */
ccb->transaction_id = ++dpt->commands_processed;
ccb->std_callback = (ccb_callback) dpt_user_cmd_done;
ccb->result = (u_int32_t) & cmdarg;
ccb->data = data;
#ifdef DPT_MEASURE_PERFORMANCE
++dpt->performance.command_count[ccb->eata_ccb.cp_scsi_cmd];
ccb->command_started = microtime_now;
#endif
dpt_Qadd_waiting(dpt, ccb);
dpt_sched_queue(dpt);
/* Wait for the command to complete */
(void) mtx_sleep((void *) ccb, &dpt->lock, PCATCH | PRIBIO, "dptucw",
100 * hz);
/* Free allocated memory */
if (data != NULL)
free(data, M_TEMP);
return (0);
}
static void
dpt_user_cmd_done(dpt_softc_t * dpt, int bus, dpt_ccb_t * ccb)
{
u_int32_t result;
caddr_t cmd_arg;
mtx_unlock(&dpt->lock);
/**
* If Auto Request Sense is on, copyout the sense struct
*/
#define usr_pckt_DMA (caddr_t)(intptr_t)ntohl(ccb->eata_ccb.cp_reqDMA)
#define usr_pckt_len ntohl(ccb->eata_ccb.cp_datalen)
if (ccb->eata_ccb.Auto_Req_Sen == 1) {
if (copyout((caddr_t) & ccb->sense_data, usr_pckt_DMA,
sizeof(struct scsi_sense_data))) {
mtx_lock(&dpt->lock);
ccb->result = EFAULT;
dpt_Qpush_free(dpt, ccb);
wakeup(ccb);
return;
}
}
/* If DataIn is on, copyout the data */
if ((ccb->eata_ccb.DataIn == 1)
&& (ccb->status_packet.hba_stat == HA_NO_ERROR)) {
if (copyout(ccb->data, usr_pckt_DMA, usr_pckt_len)) {
mtx_lock(&dpt->lock);
dpt_Qpush_free(dpt, ccb);
ccb->result = EFAULT;
wakeup(ccb);
return;
}
}
/* Copyout the status */
result = ccb->status_packet.hba_stat;
cmd_arg = (caddr_t) ccb->result;
if (copyout((caddr_t) & result, cmd_arg, sizeof(result))) {
mtx_lock(&dpt->lock);
dpt_Qpush_free(dpt, ccb);
ccb->result = EFAULT;
wakeup(ccb);
return;
}
mtx_lock(&dpt->lock);
/* Put the CCB back in the freelist */
ccb->state |= DPT_CCB_STATE_COMPLETED;
dpt_Qpush_free(dpt, ccb);
/* Free allocated memory */
return;
}
#ifdef DPT_HANDLE_TIMEOUTS
/**
* This function walks down the SUBMITTED queue.
* Every request that is too old gets aborted and marked.
* Since the DPT will complete (interrupt) immediately (what does that mean?),
* We just walk the list, aborting old commands and marking them as such.
* The dpt_complete function will get rid of the that were interrupted in the
* normal manner.
*
* This function needs to run at splcam(), as it interacts with the submitted
* queue, as well as the completed and free queues. Just like dpt_intr() does.
* To run it at any ISPL other than that of dpt_intr(), will mean that dpt_intr
* willbe able to pre-empt it, grab a transaction in progress (towards
* destruction) and operate on it. The state of this transaction will be not
* very clear.
* The only other option, is to lock it only as long as necessary but have
* dpt_intr() spin-wait on it. In a UP environment this makes no sense and in
* a SMP environment, the advantage is dubvious for a function that runs once
* every ten seconds for few microseconds and, on systems with healthy
* hardware, does not do anything anyway.
*/
static void
dpt_handle_timeouts(dpt_softc_t * dpt)
{
dpt_ccb_t *ccb;
if (dpt->state & DPT_HA_TIMEOUTS_ACTIVE) {
device_printf(dpt->dev, "WARNING: Timeout Handling Collision\n");
return;
}
dpt->state |= DPT_HA_TIMEOUTS_ACTIVE;
/* Loop through the entire submitted queue, looking for lost souls */
TAILQ_FIRST(ccb, &&dpt->submitted_ccbs, links) {
struct scsi_xfer *xs;
u_int32_t age, max_age;
xs = ccb->xs;
age = dpt_time_delta(ccb->command_started, microtime_now);
#define TenSec 10000000
if (xs == NULL) { /* Local, non-kernel call */
max_age = TenSec;
} else {
max_age = (((xs->timeout * (dpt->submitted_ccbs_count
+ DPT_TIMEOUT_FACTOR))
> TenSec)
? (xs->timeout * (dpt->submitted_ccbs_count
+ DPT_TIMEOUT_FACTOR))
: TenSec);
}
/*
* If a transaction is marked lost and is TWICE as old as we
* care, then, and only then do we destroy it!
*/
if (ccb->state & DPT_CCB_STATE_MARKED_LOST) {
/* Remember who is next */
if (age > (max_age * 2)) {
dpt_Qremove_submitted(dpt, ccb);
ccb->state &= ~DPT_CCB_STATE_MARKED_LOST;
ccb->state |= DPT_CCB_STATE_ABORTED;
#define cmd_name scsi_cmd_name(ccb->eata_ccb.cp_scsi_cmd)
if (ccb->retries++ > DPT_RETRIES) {
device_printf(dpt->dev,
"ERROR: Destroying stale "
"%d (%s)\n"
" on "
"c%db%dt%du%d (%d/%d)\n",
ccb->transaction_id,
cmd_name,
device_get_unit(dpt->dev),
ccb->eata_ccb.cp_channel,
ccb->eata_ccb.cp_id,
ccb->eata_ccb.cp_LUN, age,
ccb->retries);
#define send_ccb &ccb->eata_ccb
#define ESA EATA_SPECIFIC_ABORT
(void) dpt_send_immediate(dpt,
send_ccb,
ESA,
0, 0);
dpt_Qpush_free(dpt, ccb);
/* The SCSI layer should re-try */
xs->error |= XS_TIMEOUT;
xs->flags |= SCSI_ITSDONE;
scsi_done(xs);
} else {
device_printf(dpt->dev,
"ERROR: Stale %d (%s) on "
"c%db%dt%du%d (%d)\n"
" gets another "
"chance(%d/%d)\n",
ccb->transaction_id,
cmd_name,
device_get_unit(dpt->dev),
ccb->eata_ccb.cp_channel,
ccb->eata_ccb.cp_id,
ccb->eata_ccb.cp_LUN,
age, ccb->retries, DPT_RETRIES);
dpt_Qpush_waiting(dpt, ccb);
dpt_sched_queue(dpt);
}
}
} else {
/*
* This is a transaction that is not to be destroyed
* (yet) But it is too old for our liking. We wait as
* long as the upper layer thinks. Not really, we
* multiply that by the number of commands in the
* submitted queue + 1.
*/
if (!(ccb->state & DPT_CCB_STATE_MARKED_LOST) &&
(age != ~0) && (age > max_age)) {
device_printf(dpt->dev,
"ERROR: Marking %d (%s) on "
"c%db%dt%du%d \n"
" as late after %dusec\n",
ccb->transaction_id,
cmd_name,
device_get_unit(dpt->dev),
ccb->eata_ccb.cp_channel,
ccb->eata_ccb.cp_id,
ccb->eata_ccb.cp_LUN, age);
ccb->state |= DPT_CCB_STATE_MARKED_LOST;
}
}
}
dpt->state &= ~DPT_HA_TIMEOUTS_ACTIVE;
}
static void
dpt_timeout(void *arg)
{
dpt_softc_t *dpt = (dpt_softc_t *) arg;
mtx_assert(&dpt->lock, MA_OWNED);
if (!(dpt->state & DPT_HA_TIMEOUTS_ACTIVE))
dpt_handle_timeouts(dpt);
callout_reset(&dpt->timer, hz * 10, dpt_timeout, dpt);
}
#endif /* DPT_HANDLE_TIMEOUTS */
#endif