NetBSD/sys/dev/ic/sunscpal.c

2293 lines
57 KiB
C

/* $NetBSD: sunscpal.c,v 1.21 2006/11/24 19:46:59 christos Exp $ */
/*
* Copyright (c) 2001 Matthew Fredette
* Copyright (c) 1995 David Jones, Gordon W. Ross
* Copyright (c) 1994 Jarle Greipsland
* 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.
* 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 authors may not be used to endorse or promote products
* derived from this software without specific prior written permission.
* 4. All advertising materials mentioning features or use of this software
* must display the following acknowledgement:
* This product includes software developed by
* David Jones and Gordon Ross
*
* THIS SOFTWARE IS PROVIDED BY THE AUTHORS ``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 AUTHORS BE LIABLE FOR ANY DIRECT, INDIRECT,
* INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT
* NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
* DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
* THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
* (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF
* THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
*/
/*
* This is a machine-independent driver for the Sun "sc"
* SCSI Bus Controller (SBC).
*
* This code should work with any memory-mapped card,
* and can be shared by multiple adapters that address
* the card with different register offset spacings.
* (This can happen on the atari, for example.)
*
* Credits, history:
*
* Matthew Fredette completely copied revision 1.38 of
* ncr5380sbc.c, and then heavily modified it to match
* the Sun sc PAL. The remaining credits are for
* ncr5380sbc.c:
*
* David Jones is the author of most of the code that now
* appears in this file, and was the architect of the
* current overall structure (MI/MD code separation, etc.)
*
* Gordon Ross integrated the message phase code, added lots of
* comments about what happens when and why (re. SCSI spec.),
* debugged some reentrance problems, and added several new
* "hooks" needed for the Sun3 "si" adapters.
*
* The message in/out code was taken nearly verbatim from
* the aic6360 driver by Jarle Greipsland.
*
* Several other NCR5380 drivers were used for reference
* while developing this driver, including work by:
* The Alice Group (mac68k port) namely:
* Allen K. Briggs, Chris P. Caputo, Michael L. Finch,
* Bradley A. Grantham, and Lawrence A. Kesteloot
* Michael L. Hitch (amiga drivers: sci.c)
* Leo Weppelman (atari driver: ncr5380.c)
* There are others too. Thanks, everyone.
*
* Transliteration to bus_space() performed 9/17/98 by
* John Ruschmeyer (jruschme@exit109.com) for i386 'nca' driver.
* Thank you all.
*/
#include <sys/cdefs.h>
__KERNEL_RCSID(0, "$NetBSD: sunscpal.c,v 1.21 2006/11/24 19:46:59 christos Exp $");
#include "opt_ddb.h"
#include <sys/param.h>
#include <sys/systm.h>
#include <sys/kernel.h>
#include <sys/errno.h>
#include <sys/malloc.h>
#include <sys/device.h>
#include <sys/buf.h>
#include <sys/proc.h>
#include <sys/user.h>
#include <dev/scsipi/scsi_all.h>
#include <dev/scsipi/scsipi_all.h>
#include <dev/scsipi/scsipi_debug.h>
#include <dev/scsipi/scsi_message.h>
#include <dev/scsipi/scsiconf.h>
#ifdef DDB
#include <ddb/db_output.h>
#endif
#include <dev/ic/sunscpalreg.h>
#include <dev/ic/sunscpalvar.h>
static void sunscpal_reset_scsibus(struct sunscpal_softc *);
static void sunscpal_sched(struct sunscpal_softc *);
static void sunscpal_done(struct sunscpal_softc *);
static int sunscpal_select
(struct sunscpal_softc *, struct sunscpal_req *);
static void sunscpal_reselect(struct sunscpal_softc *);
static int sunscpal_msg_in(struct sunscpal_softc *);
static int sunscpal_msg_out(struct sunscpal_softc *);
static int sunscpal_data_xfer(struct sunscpal_softc *, int);
static int sunscpal_command(struct sunscpal_softc *);
static int sunscpal_status(struct sunscpal_softc *);
static void sunscpal_machine(struct sunscpal_softc *);
void sunscpal_abort(struct sunscpal_softc *);
void sunscpal_cmd_timeout(void *);
/*
* Action flags returned by the info_transfer functions:
* (These determine what happens next.)
*/
#define ACT_CONTINUE 0x00 /* No flags: expect another phase */
#define ACT_DISCONNECT 0x01 /* Target is disconnecting */
#define ACT_CMD_DONE 0x02 /* Need to call scsipi_done() */
#define ACT_RESET_BUS 0x04 /* Need bus reset (cmd timeout) */
#define ACT_WAIT_DMA 0x10 /* Wait for DMA to complete */
/*****************************************************************
* Debugging stuff
*****************************************************************/
#ifndef DDB
/* This is used only in recoverable places. */
#ifndef Debugger
#define Debugger() printf("Debug: sunscpal.c:%d\n", __LINE__)
#endif
#endif
#ifdef SUNSCPAL_DEBUG
#define SUNSCPAL_DBG_BREAK 1
#define SUNSCPAL_DBG_CMDS 2
#define SUNSCPAL_DBG_DMA 4
int sunscpal_debug = 0;
#define SUNSCPAL_BREAK() \
do { if (sunscpal_debug & SUNSCPAL_DBG_BREAK) Debugger(); } while (0)
static void sunscpal_show_scsi_cmd(struct scsipi_xfer *);
#ifdef DDB
void sunscpal_clear_trace(void);
void sunscpal_show_trace(void);
void sunscpal_show_req(struct sunscpal_req *);
void sunscpal_show_state(void);
#endif /* DDB */
#else /* SUNSCPAL_DEBUG */
#define SUNSCPAL_BREAK() /* nada */
#define sunscpal_show_scsi_cmd(xs) /* nada */
#endif /* SUNSCPAL_DEBUG */
static const char *
phase_names[8] = {
"DATA_OUT",
"DATA_IN",
"COMMAND",
"STATUS",
"UNSPEC1",
"UNSPEC2",
"MSG_OUT",
"MSG_IN",
};
#ifdef SUNSCPAL_USE_BUS_DMA
static void sunscpal_dma_alloc(struct sunscpal_softc *);
static void sunscpal_dma_free(struct sunscpal_softc *);
static void sunscpal_dma_setup(struct sunscpal_softc *);
#else
#define sunscpal_dma_alloc(sc) (*sc->sc_dma_alloc)(sc)
#define sunscpal_dma_free(sc) (*sc->sc_dma_free)(sc)
#define sunscpal_dma_setup(sc) (*sc->sc_dma_setup)(sc)
#endif
static void sunscpal_minphys(struct buf *);
/*****************************************************************
* Actual chip control
*****************************************************************/
/*
* XXX: These timeouts might need to be tuned...
*/
/* This one is used when waiting for a phase change. (X100uS.) */
int sunscpal_wait_phase_timo = 1000 * 10 * 300; /* 5 min. */
/* These are used in the following inline functions. */
int sunscpal_wait_req_timo = 1000 * 50; /* X2 = 100 mS. */
int sunscpal_wait_nrq_timo = 1000 * 25; /* X2 = 50 mS. */
static inline int sunscpal_wait_req(struct sunscpal_softc *);
static inline int sunscpal_wait_not_req(struct sunscpal_softc *);
static inline void sunscpal_sched_msgout(struct sunscpal_softc *, int);
/* Return zero on success. */
static inline int sunscpal_wait_req(sc)
struct sunscpal_softc *sc;
{
int timo = sunscpal_wait_req_timo;
for (;;) {
if (SUNSCPAL_READ_2(sc, sunscpal_icr) & SUNSCPAL_ICR_REQUEST) {
timo = 0; /* return 0 */
break;
}
if (--timo < 0)
break; /* return -1 */
delay(2);
}
return (timo);
}
/* Return zero on success. */
static inline int sunscpal_wait_not_req(sc)
struct sunscpal_softc *sc;
{
int timo = sunscpal_wait_nrq_timo;
for (;;) {
if ((SUNSCPAL_READ_2(sc, sunscpal_icr) & SUNSCPAL_ICR_REQUEST) == 0) {
timo = 0; /* return 0 */
break;
}
if (--timo < 0)
break; /* return -1 */
delay(2);
}
return (timo);
}
/*
* These functions control DMA functions in the chipset independent of
* the host DMA implementation.
*/
static void sunscpal_dma_start(struct sunscpal_softc *);
static void sunscpal_dma_poll(struct sunscpal_softc *);
static void sunscpal_dma_stop(struct sunscpal_softc *);
static void
sunscpal_dma_start(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr = sc->sc_current;
int xlen;
u_int16_t icr;
xlen = sc->sc_reqlen;
/* Let'er rip! */
icr = SUNSCPAL_READ_2(sc, sunscpal_icr);
icr |= SUNSCPAL_ICR_DMA_ENABLE |
((xlen & 1) ? 0 : SUNSCPAL_ICR_WORD_MODE) |
((sr->sr_flags & SR_IMMED) ? 0 : SUNSCPAL_ICR_INTERRUPT_ENABLE);
SUNSCPAL_WRITE_2(sc, sunscpal_icr, icr);
sc->sc_state |= SUNSCPAL_DOINGDMA;
#ifdef SUNSCPAL_DEBUG
if (sunscpal_debug & SUNSCPAL_DBG_DMA) {
printf("sunscpal_dma_start: started, flags=0x%x\n",
sc->sc_state);
}
#endif
}
#define ICR_MASK (SUNSCPAL_ICR_PARITY_ERROR | SUNSCPAL_ICR_BUS_ERROR | SUNSCPAL_ICR_INTERRUPT_REQUEST)
#define POLL_TIMO 50000 /* X100 = 5 sec. */
/*
* Poll (spin-wait) for DMA completion.
* Called right after xx_dma_start(), and
* xx_dma_stop() will be called next.
*/
static void
sunscpal_dma_poll(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr = sc->sc_current;
int tmo;
/* Make sure DMA started successfully. */
if (sc->sc_state & SUNSCPAL_ABORTING)
return;
/* Wait for any "DMA complete" or error bits. */
tmo = POLL_TIMO;
for (;;) {
if (SUNSCPAL_READ_2(sc, sunscpal_icr) & ICR_MASK)
break;
if (--tmo <= 0) {
printf("sc: DMA timeout (while polling)\n");
/* Indicate timeout as MI code would. */
sr->sr_flags |= SR_OVERDUE;
break;
}
delay(100);
}
SUNSCPAL_TRACE("sunscpal_dma_poll: waited %d\n",
POLL_TIMO - tmo);
#ifdef SUNSCPAL_DEBUG
if (sunscpal_debug & SUNSCPAL_DBG_DMA) {
char buffer[64];
bitmask_snprintf(SUNSCPAL_READ_2(sc, sunscpal_icr), SUNSCPAL_ICR_BITS, buffer, sizeof(buffer));
printf("sunscpal_dma_poll: done, icr=%s\n", buffer);
}
#endif
}
static void
sunscpal_dma_stop(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr = sc->sc_current;
struct scsipi_xfer *xs = sr->sr_xs;
int resid, ntrans;
u_int16_t icr;
if ((sc->sc_state & SUNSCPAL_DOINGDMA) == 0) {
#ifdef DEBUG
printf("sunscpal_dma_stop: DMA not running\n");
#endif
return;
}
sc->sc_state &= ~SUNSCPAL_DOINGDMA;
/* First, halt the DMA engine. */
icr = SUNSCPAL_READ_2(sc, sunscpal_icr);
icr &= ~(SUNSCPAL_ICR_DMA_ENABLE | SUNSCPAL_ICR_WORD_MODE | SUNSCPAL_ICR_INTERRUPT_ENABLE);
SUNSCPAL_WRITE_2(sc, sunscpal_icr, icr);
#ifdef SUNSCPAL_USE_BUS_DMA
/*
* XXX - this function is supposed to be independent of
* the host's DMA implementation.
*/
{
sunscpal_dma_handle_t dh = sr->sr_dma_hand;
/* sync the DMA map: */
bus_dmamap_sync(sc->sunscpal_dmat, dh->dh_dmamap, 0, dh->dh_maplen,
((xs->xs_control & XS_CTL_DATA_OUT) == 0 ? BUS_DMASYNC_POSTREAD : BUS_DMASYNC_POSTWRITE));
}
#endif /* SUNSCPAL_USE_BUS_DMA */
if (icr & (SUNSCPAL_ICR_BUS_ERROR)) {
char buffer[64];
bitmask_snprintf(icr, SUNSCPAL_ICR_BITS, buffer, sizeof(buffer));
printf("sc: DMA error, icr=%s, reset\n", buffer);
sr->sr_xs->error = XS_DRIVER_STUFFUP;
sc->sc_state |= SUNSCPAL_ABORTING;
goto out;
}
/* Note that timeout may have set the error flag. */
if (sc->sc_state & SUNSCPAL_ABORTING)
goto out;
/* XXX: Wait for DMA to actually finish? */
/*
* Now try to figure out how much actually transferred
*/
resid = SUNSCPAL_DMA_COUNT_FLIP(SUNSCPAL_READ_2(sc, sunscpal_dma_count));
ntrans = sc->sc_reqlen - resid;
#ifdef SUNSCPAL_DEBUG
if (sunscpal_debug & SUNSCPAL_DBG_DMA) {
printf("sunscpal_dma_stop: resid=0x%x ntrans=0x%x\n",
resid, ntrans);
}
#endif
if (ntrans < sc->sc_min_dma_len) {
printf("sc: DMA count: 0x%x\n", resid);
sc->sc_state |= SUNSCPAL_ABORTING;
goto out;
}
if (ntrans > sc->sc_datalen)
panic("sunscpal_dma_stop: excess transfer");
/* Adjust data pointer */
sc->sc_dataptr += ntrans;
sc->sc_datalen -= ntrans;
/*
* After a read, we may need to clean-up
* "Left-over bytes" (yuck!)
*/
if (((xs->xs_control & XS_CTL_DATA_OUT) == 0) &&
((icr & SUNSCPAL_ICR_ODD_LENGTH) != 0))
{
#ifdef DEBUG
printf("sc: Got Left-over bytes!\n");
#endif
*(sc->sc_dataptr++) = SUNSCPAL_READ_1(sc, sunscpal_data);
sc->sc_datalen--;
}
out:
SUNSCPAL_WRITE_2(sc, sunscpal_dma_count, SUNSCPAL_DMA_COUNT_FLIP(0));
}
/* Ask the target for a MSG_OUT phase. */
static inline void
sunscpal_sched_msgout(sc, msg_code)
struct sunscpal_softc *sc;
int msg_code;
{
/*
* This controller does not allow you to assert ATN, which
* will eventually leave us with no option other than to reset
* the bus. We keep this function as a placeholder, though,
* and this printf will eventually go away or get #ifdef'ed:
*/
printf("sunscpal_sched_msgout: trying to schedule 0x%0x\n", msg_code);
sc->sc_msgpriq |= msg_code;
}
int
sunscpal_pio_out(sc, phase, count, data)
struct sunscpal_softc *sc;
int phase, count;
unsigned char *data;
{
int resid;
resid = count;
while (resid > 0) {
if (!SUNSCPAL_BUSY(sc)) {
SUNSCPAL_TRACE("pio_out: lost BSY, resid=%d\n", resid);
break;
}
if (sunscpal_wait_req(sc)) {
SUNSCPAL_TRACE("pio_out: no REQ, resid=%d\n", resid);
break;
}
if (SUNSCPAL_BUS_PHASE(SUNSCPAL_READ_2(sc, sunscpal_icr)) != phase)
break;
/* Put the data on the bus. */
if (data) {
SUNSCPAL_BYTE_WRITE(sc, phase, *data++);
} else {
SUNSCPAL_BYTE_WRITE(sc, phase, 0);
}
--resid;
}
return (count - resid);
}
int
sunscpal_pio_in(sc, phase, count, data)
struct sunscpal_softc *sc;
int phase, count;
unsigned char *data;
{
int resid;
resid = count;
while (resid > 0) {
if (!SUNSCPAL_BUSY(sc)) {
SUNSCPAL_TRACE("pio_in: lost BSY, resid=%d\n", resid);
break;
}
if (sunscpal_wait_req(sc)) {
SUNSCPAL_TRACE("pio_in: no REQ, resid=%d\n", resid);
break;
}
/* A phase change is not valid until AFTER REQ rises! */
if (SUNSCPAL_BUS_PHASE(SUNSCPAL_READ_2(sc, sunscpal_icr)) != phase)
break;
/* Read the data bus. */
if (data)
*data++ = SUNSCPAL_BYTE_READ(sc, phase);
else
(void) SUNSCPAL_BYTE_READ(sc, phase);
--resid;
}
return (count - resid);
}
void
sunscpal_init(sc)
struct sunscpal_softc *sc;
{
int i, j;
#ifdef SUNSCPAL_DEBUG
sunscpal_debug_sc = sc;
#endif
for (i = 0; i < SUNSCPAL_OPENINGS; i++)
sc->sc_ring[i].sr_xs = NULL;
for (i = 0; i < 8; i++)
for (j = 0; j < 8; j++)
sc->sc_matrix[i][j] = NULL;
sc->sc_prevphase = SUNSCPAL_PHASE_INVALID;
sc->sc_state = SUNSCPAL_IDLE;
SUNSCPAL_WRITE_2(sc, sunscpal_icr, 0);
SUNSCPAL_WRITE_2(sc, sunscpal_dma_addr_h, 0);
SUNSCPAL_WRITE_2(sc, sunscpal_dma_addr_l, 0);
SUNSCPAL_WRITE_2(sc, sunscpal_dma_count, SUNSCPAL_DMA_COUNT_FLIP(0));
SUNSCPAL_CLR_INTR(sc);
/* Another hack (Er.. hook!) for anything that needs it: */
if (sc->sc_intr_on) {
SUNSCPAL_TRACE("init: intr ON\n", 0);
sc->sc_intr_on(sc);
}
}
static void
sunscpal_reset_scsibus(sc)
struct sunscpal_softc *sc;
{
SUNSCPAL_TRACE("reset_scsibus, cur=0x%x\n",
(long) sc->sc_current);
SUNSCPAL_WRITE_2(sc, sunscpal_icr, SUNSCPAL_ICR_RESET);
delay(500);
SUNSCPAL_WRITE_2(sc, sunscpal_icr, 0);
SUNSCPAL_CLR_INTR(sc);
/* XXX - Need long delay here! */
delay(100000);
/* XXX - Need to cancel disconnected requests. */
}
/*
* Interrupt handler for the SCSI Bus Controller (SBC)
* This may also called for a DMA timeout (at splbio).
*/
int
sunscpal_intr(arg)
void *arg;
{
struct sunscpal_softc *sc = arg;
int claimed = 0;
/*
* Do not touch SBC regs here unless sc_current == NULL
* or it will complain about "register conflict" errors.
* Instead, just let sunscpal_machine() deal with it.
*/
SUNSCPAL_TRACE("intr: top, state=%d\n", sc->sc_state);
if (sc->sc_state == SUNSCPAL_IDLE) {
/*
* Might be reselect. sunscpal_reselect() will check,
* and set up the connection if so. This will verify
* that sc_current == NULL at the beginning...
*/
/* Another hack (Er.. hook!) for anything that needs it: */
if (sc->sc_intr_off) {
SUNSCPAL_TRACE("intr: for reselect, intr off\n", 0);
sc->sc_intr_off(sc);
}
sunscpal_reselect(sc);
}
/*
* The remaining documented interrupt causes are a DMA complete
* condition.
*
* The procedure is to let sunscpal_machine() figure out what
* to do next.
*/
if (sc->sc_state & SUNSCPAL_WORKING) {
SUNSCPAL_TRACE("intr: call machine, cur=0x%x\n",
(long) sc->sc_current);
/* This will usually free-up the nexus. */
sunscpal_machine(sc);
SUNSCPAL_TRACE("intr: machine done, cur=0x%x\n",
(long) sc->sc_current);
claimed = 1;
}
/* Maybe we can run some commands now... */
if (sc->sc_state == SUNSCPAL_IDLE) {
SUNSCPAL_TRACE("intr: call sched, cur=0x%x\n",
(long) sc->sc_current);
sunscpal_sched(sc);
SUNSCPAL_TRACE("intr: sched done, cur=0x%x\n",
(long) sc->sc_current);
}
return claimed;
}
/*
* Abort the current command (i.e. due to timeout)
*/
void
sunscpal_abort(sc)
struct sunscpal_softc *sc;
{
/*
* Finish it now. If DMA is in progress, we
* can not call sunscpal_sched_msgout() because
* that hits the SBC (avoid DMA conflict).
*/
/* Another hack (Er.. hook!) for anything that needs it: */
if (sc->sc_intr_off) {
SUNSCPAL_TRACE("abort: intr off\n", 0);
sc->sc_intr_off(sc);
}
sc->sc_state |= SUNSCPAL_ABORTING;
if ((sc->sc_state & SUNSCPAL_DOINGDMA) == 0) {
sunscpal_sched_msgout(sc, SEND_ABORT);
}
SUNSCPAL_TRACE("abort: call machine, cur=0x%x\n",
(long) sc->sc_current);
sunscpal_machine(sc);
SUNSCPAL_TRACE("abort: machine done, cur=0x%x\n",
(long) sc->sc_current);
/* Another hack (Er.. hook!) for anything that needs it: */
if (sc->sc_intr_on) {
SUNSCPAL_TRACE("abort: intr ON\n", 0);
sc->sc_intr_on(sc);
}
}
/*
* Timeout handler, scheduled for each SCSI command.
*/
void
sunscpal_cmd_timeout(arg)
void *arg;
{
struct sunscpal_req *sr = arg;
struct scsipi_xfer *xs;
struct scsipi_periph *periph;
struct sunscpal_softc *sc;
int s;
s = splbio();
/* Get all our variables... */
xs = sr->sr_xs;
if (xs == NULL) {
printf("sunscpal_cmd_timeout: no scsipi_xfer\n");
goto out;
}
periph = xs->xs_periph;
sc = (void *)periph->periph_channel->chan_adapter->adapt_dev;
printf("%s: cmd timeout, targ=%d, lun=%d\n",
sc->sc_dev.dv_xname,
sr->sr_target, sr->sr_lun);
/*
* Mark the overdue job as failed, and arrange for
* sunscpal_machine to terminate it. If the victim
* is the current job, call sunscpal_machine() now.
* Otherwise arrange for sunscpal_sched() to do it.
*/
sr->sr_flags |= SR_OVERDUE;
if (sc->sc_current == sr) {
SUNSCPAL_TRACE("cmd_tmo: call abort, sr=0x%x\n", (long) sr);
sunscpal_abort(sc);
} else {
/*
* The driver may be idle, or busy with another job.
* Arrange for sunscpal_sched() to do the deed.
*/
SUNSCPAL_TRACE("cmd_tmo: clear matrix, t/l=0x%02x\n",
(sr->sr_target << 4) | sr->sr_lun);
sc->sc_matrix[sr->sr_target][sr->sr_lun] = NULL;
}
/*
* We may have aborted the current job, or may have
* already been idle. In either case, we should now
* be idle, so try to start another job.
*/
if (sc->sc_state == SUNSCPAL_IDLE) {
SUNSCPAL_TRACE("cmd_tmo: call sched, cur=0x%x\n",
(long) sc->sc_current);
sunscpal_sched(sc);
SUNSCPAL_TRACE("cmd_tmo: sched done, cur=0x%x\n",
(long) sc->sc_current);
}
out:
splx(s);
}
/*****************************************************************
* Interface to higher level
*****************************************************************/
/*
* Enter a new SCSI command into the "issue" queue, and
* if there is work to do, start it going.
*
* WARNING: This can be called recursively!
* (see comment in sunscpal_done)
*/
void
sunscpal_scsipi_request(chan, req, arg)
struct scsipi_channel *chan;
scsipi_adapter_req_t req;
void *arg;
{
struct scsipi_xfer *xs;
struct scsipi_periph *periph;
struct sunscpal_softc *sc = (void *)chan->chan_adapter->adapt_dev;
struct sunscpal_req *sr;
int s, i, flags;
switch (req) {
case ADAPTER_REQ_RUN_XFER:
xs = arg;
periph = xs->xs_periph;
flags = xs->xs_control;
if (flags & XS_CTL_DATA_UIO)
panic("sunscpal: scsi data uio requested");
s = splbio();
if (flags & XS_CTL_POLL) {
/* Terminate any current command. */
sr = sc->sc_current;
if (sr) {
printf("%s: polled request aborting %d/%d\n",
sc->sc_dev.dv_xname,
sr->sr_target, sr->sr_lun);
sunscpal_abort(sc);
}
if (sc->sc_state != SUNSCPAL_IDLE) {
panic("sunscpal_scsi_cmd: polled request, abort failed");
}
}
/*
* Find lowest empty slot in ring buffer.
* XXX: What about "fairness" and cmd order?
*/
for (i = 0; i < SUNSCPAL_OPENINGS; i++)
if (sc->sc_ring[i].sr_xs == NULL)
goto new;
xs->error = XS_RESOURCE_SHORTAGE;
SUNSCPAL_TRACE("scsipi_cmd: no openings, rv=%d\n", rv);
goto out;
new:
/* Create queue entry */
sr = &sc->sc_ring[i];
sr->sr_xs = xs;
sr->sr_target = xs->xs_periph->periph_target;
sr->sr_lun = xs->xs_periph->periph_lun;
sr->sr_dma_hand = NULL;
sr->sr_dataptr = xs->data;
sr->sr_datalen = xs->datalen;
sr->sr_flags = (flags & XS_CTL_POLL) ? SR_IMMED : 0;
sr->sr_status = -1; /* no value */
sc->sc_ncmds++;
SUNSCPAL_TRACE("scsipi_cmd: new sr=0x%x\n", (long)sr);
if (flags & XS_CTL_POLL) {
/* Force this new command to be next. */
sc->sc_rr = i;
}
/*
* If we were idle, run some commands...
*/
if (sc->sc_state == SUNSCPAL_IDLE) {
SUNSCPAL_TRACE("scsipi_cmd: call sched, cur=0x%x\n",
(long) sc->sc_current);
sunscpal_sched(sc);
SUNSCPAL_TRACE("scsipi_cmd: sched done, cur=0x%x\n",
(long) sc->sc_current);
}
if (flags & XS_CTL_POLL) {
/* Make sure sunscpal_sched() finished it. */
if ((xs->xs_status & XS_STS_DONE) == 0)
panic("sunscpal_scsi_cmd: poll didn't finish");
}
out:
splx(s);
return;
case ADAPTER_REQ_GROW_RESOURCES:
/* XXX Not supported. */
return;
case ADAPTER_REQ_SET_XFER_MODE:
{
/*
* We don't support Sync, Wide, or Tagged Queueing.
* Just callback now, to report this.
*/
struct scsipi_xfer_mode *xm = arg;
xm->xm_mode = 0;
xm->xm_period = 0;
xm->xm_offset = 0;
scsipi_async_event(chan, ASYNC_EVENT_XFER_MODE, xm);
return;
}
}
}
/*
* POST PROCESSING OF SCSI_CMD (usually current)
* Called by sunscpal_sched(), sunscpal_machine()
*/
static void
sunscpal_done(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr;
struct scsipi_xfer *xs;
#ifdef DIAGNOSTIC
if (sc->sc_state == SUNSCPAL_IDLE)
panic("sunscpal_done: state=idle");
if (sc->sc_current == NULL)
panic("sunscpal_done: current=0");
#endif
sr = sc->sc_current;
xs = sr->sr_xs;
SUNSCPAL_TRACE("done: top, cur=0x%x\n", (long) sc->sc_current);
/*
* Clean up DMA resources for this command.
*/
if (sr->sr_dma_hand) {
SUNSCPAL_TRACE("done: dma_free, dh=0x%x\n",
(long) sr->sr_dma_hand);
sunscpal_dma_free(sc);
}
#ifdef DIAGNOSTIC
if (sr->sr_dma_hand)
panic("sunscpal_done: DMA free did not");
#endif
if (sc->sc_state & SUNSCPAL_ABORTING) {
SUNSCPAL_TRACE("done: aborting, error=%d\n", xs->error);
if (xs->error == XS_NOERROR)
xs->error = XS_TIMEOUT;
}
SUNSCPAL_TRACE("done: check error=%d\n", (long) xs->error);
/* If error is already set, ignore sr_status value. */
if (xs->error != XS_NOERROR)
goto finish;
SUNSCPAL_TRACE("done: check status=%d\n", sr->sr_status);
xs->status = sr->sr_status;
switch (sr->sr_status) {
case SCSI_OK: /* 0 */
break;
case SCSI_CHECK:
case SCSI_BUSY:
xs->error = XS_BUSY;
break;
case -1:
/* This is our "impossible" initial value. */
/* fallthrough */
default:
printf("%s: target %d, bad status=%d\n",
sc->sc_dev.dv_xname, sr->sr_target, sr->sr_status);
xs->error = XS_DRIVER_STUFFUP;
break;
}
finish:
SUNSCPAL_TRACE("done: finish, error=%d\n", xs->error);
/*
* Dequeue the finished command, but don't clear sc_state until
* after the call to scsipi_done(), because that may call back to
* sunscpal_scsi_cmd() - unwanted recursion!
*
* Keeping sc->sc_state != idle terminates the recursion.
*/
#ifdef DIAGNOSTIC
if ((sc->sc_state & SUNSCPAL_WORKING) == 0)
panic("sunscpal_done: bad state");
#endif
/* Clear our pointers to the request. */
sc->sc_current = NULL;
sc->sc_matrix[sr->sr_target][sr->sr_lun] = NULL;
callout_stop(&sr->sr_xs->xs_callout);
/* Make the request free. */
sr->sr_xs = NULL;
sc->sc_ncmds--;
/* Tell common SCSI code it is done. */
scsipi_done(xs);
sc->sc_state = SUNSCPAL_IDLE;
/* Now sunscpal_sched() may be called again. */
}
/*
* Schedule a SCSI operation. This routine should return
* only after it achieves one of the following conditions:
* Busy (sc->sc_state != SUNSCPAL_IDLE)
* No more work can be started.
*/
static void
sunscpal_sched(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr;
struct scsipi_xfer *xs;
int target = 0, lun = 0;
int error, i;
/* Another hack (Er.. hook!) for anything that needs it: */
if (sc->sc_intr_off) {
SUNSCPAL_TRACE("sched: top, intr off\n", 0);
sc->sc_intr_off(sc);
}
next_job:
/*
* Grab the next job from queue. Must be idle.
*/
#ifdef DIAGNOSTIC
if (sc->sc_state != SUNSCPAL_IDLE)
panic("sunscpal_sched: not idle");
if (sc->sc_current)
panic("sunscpal_sched: current set");
#endif
/*
* Always start the search where we last looked.
*/
i = sc->sc_rr;
sr = NULL;
do {
if (sc->sc_ring[i].sr_xs) {
target = sc->sc_ring[i].sr_target;
lun = sc->sc_ring[i].sr_lun;
if (sc->sc_matrix[target][lun] == NULL) {
/*
* Do not mark the target/LUN busy yet,
* because reselect may cause some other
* job to become the current one, so we
* might not actually start this job.
* Instead, set sc_matrix later on.
*/
sc->sc_rr = i;
sr = &sc->sc_ring[i];
break;
}
}
i++;
if (i == SUNSCPAL_OPENINGS)
i = 0;
} while (i != sc->sc_rr);
if (sr == NULL) {
SUNSCPAL_TRACE("sched: no work, cur=0x%x\n",
(long) sc->sc_current);
/* Another hack (Er.. hook!) for anything that needs it: */
if (sc->sc_intr_on) {
SUNSCPAL_TRACE("sched: ret, intr ON\n", 0);
sc->sc_intr_on(sc);
}
return; /* No more work to do. */
}
SUNSCPAL_TRACE("sched: select for t/l=0x%02x\n",
(sr->sr_target << 4) | sr->sr_lun);
sc->sc_state = SUNSCPAL_WORKING;
error = sunscpal_select(sc, sr);
if (sc->sc_current) {
/* Lost the race! reselected out from under us! */
/* Work with the reselected job. */
if (sr->sr_flags & SR_IMMED) {
printf("%s: reselected while polling (abort)\n",
sc->sc_dev.dv_xname);
/* Abort the reselected job. */
sc->sc_state |= SUNSCPAL_ABORTING;
sc->sc_msgpriq |= SEND_ABORT;
}
sr = sc->sc_current;
xs = sr->sr_xs;
SUNSCPAL_TRACE("sched: reselect, new sr=0x%x\n", (long)sr);
goto have_nexus;
}
/* Normal selection result. Target/LUN is now busy. */
sc->sc_matrix[target][lun] = sr;
sc->sc_current = sr; /* connected */
xs = sr->sr_xs;
/*
* Initialize pointers, etc. for this job
*/
sc->sc_dataptr = sr->sr_dataptr;
sc->sc_datalen = sr->sr_datalen;
sc->sc_prevphase = SUNSCPAL_PHASE_INVALID;
sc->sc_msgpriq = SEND_IDENTIFY;
sc->sc_msgoutq = 0;
sc->sc_msgout = 0;
SUNSCPAL_TRACE("sched: select rv=%d\n", error);
switch (error) {
case XS_NOERROR:
break;
case XS_BUSY:
/* XXX - Reset and try again. */
printf("%s: select found SCSI bus busy, resetting...\n",
sc->sc_dev.dv_xname);
sunscpal_reset_scsibus(sc);
/* fallthrough */
case XS_SELTIMEOUT:
default:
xs->error = error; /* from select */
SUNSCPAL_TRACE("sched: call done, sr=0x%x\n", (long)sr);
sunscpal_done(sc);
/* Paranoia: clear everything. */
sc->sc_dataptr = NULL;
sc->sc_datalen = 0;
sc->sc_prevphase = SUNSCPAL_PHASE_INVALID;
sc->sc_msgpriq = 0;
sc->sc_msgoutq = 0;
sc->sc_msgout = 0;
goto next_job;
}
/*
* Selection was successful. Normally, this means
* we are starting a new command. However, this
* might be the termination of an overdue job.
*/
if (sr->sr_flags & SR_OVERDUE) {
SUNSCPAL_TRACE("sched: overdue, sr=0x%x\n", (long)sr);
sc->sc_state |= SUNSCPAL_ABORTING;
sc->sc_msgpriq |= SEND_ABORT;
goto have_nexus;
}
/*
* OK, we are starting a new command.
* Initialize and allocate resources for the new command.
* Device reset is special (only uses MSG_OUT phase).
* Normal commands start in MSG_OUT phase where we will
* send and IDENDIFY message, and then expect CMD phase.
*/
#ifdef SUNSCPAL_DEBUG
if (sunscpal_debug & SUNSCPAL_DBG_CMDS) {
printf("sunscpal_sched: begin, target=%d, LUN=%d\n",
xs->xs_periph->periph_target, xs->xs_periph->periph_lun);
sunscpal_show_scsi_cmd(xs);
}
#endif
if (xs->xs_control & XS_CTL_RESET) {
SUNSCPAL_TRACE("sched: cmd=reset, sr=0x%x\n", (long)sr);
/* Not an error, so do not set SUNSCPAL_ABORTING */
sc->sc_msgpriq |= SEND_DEV_RESET;
goto have_nexus;
}
#ifdef DIAGNOSTIC
if ((xs->xs_control & (XS_CTL_DATA_IN | XS_CTL_DATA_OUT)) == 0) {
if (sc->sc_dataptr) {
printf("%s: ptr but no data in/out flags?\n",
sc->sc_dev.dv_xname);
SUNSCPAL_BREAK();
sc->sc_dataptr = NULL;
}
}
#endif
/* Allocate DMA space (maybe) */
if (sc->sc_dataptr && (sc->sc_flags & SUNSCPAL_DISABLE_DMA) == 0 &&
(sc->sc_datalen >= sc->sc_min_dma_len))
{
SUNSCPAL_TRACE("sched: dma_alloc, len=%d\n", sc->sc_datalen);
sunscpal_dma_alloc(sc);
}
/*
* Initialization hook called just after select,
* at the beginning of COMMAND phase.
* (but AFTER the DMA allocation is done)
*
* We need to set up the DMA engine BEFORE the target puts
* the SCSI bus into any DATA phase.
*/
if (sr->sr_dma_hand) {
SUNSCPAL_TRACE("sched: dma_setup, dh=0x%x\n",
(long) sr->sr_dma_hand);
sunscpal_dma_setup(sc);
}
/*
* Schedule a timeout for the job we are starting.
*/
if ((sr->sr_flags & SR_IMMED) == 0) {
i = mstohz(xs->timeout);
SUNSCPAL_TRACE("sched: set timeout=%d\n", i);
callout_reset(&sr->sr_xs->xs_callout, i,
sunscpal_cmd_timeout, sr);
}
have_nexus:
SUNSCPAL_TRACE("sched: call machine, cur=0x%x\n",
(long) sc->sc_current);
sunscpal_machine(sc);
SUNSCPAL_TRACE("sched: machine done, cur=0x%x\n",
(long) sc->sc_current);
/*
* What state did sunscpal_machine() leave us in?
* Hopefully it sometimes completes a job...
*/
if (sc->sc_state == SUNSCPAL_IDLE)
goto next_job;
return; /* Have work in progress. */
}
/*
* Reselect handler: checks for reselection, and if we are being
* reselected, it sets up sc->sc_current.
*
* We are reselected when:
* SEL is TRUE
* IO is TRUE
* BSY is FALSE
*/
void
sunscpal_reselect(sc)
struct sunscpal_softc *sc;
{
/*
* This controller does not implement disconnect/reselect, so
* we really don't have anything to do here. We keep this
* function as a placeholder, though.
*/
}
/*
* Select target: xs is the transfer that we are selecting for.
* sc->sc_current should be NULL.
*
* Returns:
* sc->sc_current != NULL ==> we were reselected (race!)
* XS_NOERROR ==> selection worked
* XS_BUSY ==> lost arbitration
* XS_SELTIMEOUT ==> no response to selection
*/
static int
sunscpal_select(sc, sr)
struct sunscpal_softc *sc;
struct sunscpal_req *sr;
{
int timo, target_mask;
u_short mode;
/* Check for reselect */
sunscpal_reselect(sc);
if (sc->sc_current) {
SUNSCPAL_TRACE("select: reselect, cur=0x%x\n",
(long) sc->sc_current);
return XS_BUSY; /* reselected */
}
/*
* Select the target.
*/
target_mask = (1 << sr->sr_target);
SUNSCPAL_WRITE_1(sc, sunscpal_data, target_mask);
SUNSCPAL_WRITE_2(sc, sunscpal_icr, SUNSCPAL_ICR_SELECT);
/*
* Wait for the target to assert BSY.
* SCSI spec. says wait for 250 mS.
*/
for (timo = 25000;;) {
if (SUNSCPAL_READ_2(sc, sunscpal_icr) & SUNSCPAL_ICR_BUSY)
goto success;
if (--timo <= 0)
break;
delay(10);
}
SUNSCPAL_WRITE_1(sc, sunscpal_data, 0);
SUNSCPAL_WRITE_2(sc, sunscpal_icr, 0);
SUNSCPAL_TRACE("select: device down, rc=%d\n", XS_SELTIMEOUT);
return XS_SELTIMEOUT;
success:
/*
* The target is now driving BSY, so we can stop
* driving SEL and the data bus. We do set up
* whether or not this target needs parity.
*/
mode = 0;
if ((sc->sc_parity_disable & target_mask) == 0)
mode |= SUNSCPAL_ICR_PARITY_ENABLE;
SUNSCPAL_WRITE_2(sc, sunscpal_icr, mode);
return XS_NOERROR;
}
/*****************************************************************
* Functions to handle each info. transfer phase:
*****************************************************************/
/*
* The message system:
*
* This is a revamped message system that now should easier accommodate
* new messages, if necessary.
*
* Currently we accept these messages:
* IDENTIFY (when reselecting)
* COMMAND COMPLETE # (expect bus free after messages marked #)
* NOOP
* MESSAGE REJECT
* SYNCHRONOUS DATA TRANSFER REQUEST
* SAVE DATA POINTER
* RESTORE POINTERS
* DISCONNECT #
*
* We may send these messages in prioritized order:
* BUS DEVICE RESET # if XS_CTL_RESET & xs->xs_control (or in
* weird sits.)
* MESSAGE PARITY ERROR par. err. during MSGI
* MESSAGE REJECT If we get a message we don't know how to handle
* ABORT # send on errors
* INITIATOR DETECTED ERROR also on errors (SCSI2) (during info xfer)
* IDENTIFY At the start of each transfer
* SYNCHRONOUS DATA TRANSFER REQUEST if appropriate
* NOOP if nothing else fits the bill ...
*/
/*
* Precondition:
* The SCSI bus is already in the MSGI phase and there is a message byte
* on the bus, along with an asserted REQ signal.
*
* Our return value determines whether our caller, sunscpal_machine()
* will expect to see another REQ (and possibly phase change).
*/
static int
sunscpal_msg_in(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr = sc->sc_current;
struct scsipi_xfer *xs = sr->sr_xs;
int n, phase;
int act_flags;
act_flags = ACT_CONTINUE;
if (sc->sc_prevphase == SUNSCPAL_PHASE_MSG_IN) {
/* This is a continuation of the previous message. */
n = sc->sc_imp - sc->sc_imess;
SUNSCPAL_TRACE("msg_in: continuation, n=%d\n", n);
goto nextbyte;
}
/* This is a new MESSAGE IN phase. Clean up our state. */
sc->sc_state &= ~SUNSCPAL_DROP_MSGIN;
nextmsg:
n = 0;
sc->sc_imp = &sc->sc_imess[n];
nextbyte:
/*
* Read a whole message, but don't ack the last byte. If we reject the
* message, we have to assert ATN during the message transfer phase
* itself.
*/
for (;;) {
/*
* Read a message byte.
* First, check BSY, REQ, phase...
*/
if (!SUNSCPAL_BUSY(sc)) {
SUNSCPAL_TRACE("msg_in: lost BSY, n=%d\n", n);
/* XXX - Assume the command completed? */
act_flags |= (ACT_DISCONNECT | ACT_CMD_DONE);
return (act_flags);
}
if (sunscpal_wait_req(sc)) {
SUNSCPAL_TRACE("msg_in: BSY but no REQ, n=%d\n", n);
/* Just let sunscpal_machine() handle it... */
return (act_flags);
}
phase = SUNSCPAL_BUS_PHASE(SUNSCPAL_READ_2(sc, sunscpal_icr));
if (phase != SUNSCPAL_PHASE_MSG_IN) {
/*
* Target left MESSAGE IN, probably because it
* a) noticed our ATN signal, or
* b) ran out of messages.
*/
return (act_flags);
}
/* Still in MESSAGE IN phase, and REQ is asserted. */
if (SUNSCPAL_READ_2(sc, sunscpal_icr) & SUNSCPAL_ICR_PARITY_ERROR) {
sunscpal_sched_msgout(sc, SEND_PARITY_ERROR);
sc->sc_state |= SUNSCPAL_DROP_MSGIN;
}
/* Gather incoming message bytes if needed. */
if ((sc->sc_state & SUNSCPAL_DROP_MSGIN) == 0) {
if (n >= SUNSCPAL_MAX_MSG_LEN) {
sunscpal_sched_msgout(sc, SEND_REJECT);
sc->sc_state |= SUNSCPAL_DROP_MSGIN;
} else {
*sc->sc_imp++ = SUNSCPAL_READ_1(sc, sunscpal_cmd_stat);
n++;
/*
* This testing is suboptimal, but most
* messages will be of the one byte variety, so
* it should not affect performance
* significantly.
*/
if (n == 1 && MSG_IS1BYTE(sc->sc_imess[0]))
goto have_msg;
if (n == 2 && MSG_IS2BYTE(sc->sc_imess[0]))
goto have_msg;
if (n >= 3 && MSG_ISEXTENDED(sc->sc_imess[0]) &&
n == sc->sc_imess[1] + 2)
goto have_msg;
}
}
/*
* If we reach this spot we're either:
* a) in the middle of a multi-byte message, or
* b) dropping bytes.
*/
if (act_flags != ACT_CONTINUE)
return (act_flags);
/* back to nextbyte */
}
have_msg:
/* We now have a complete message. Parse it. */
switch (sc->sc_imess[0]) {
case MSG_CMDCOMPLETE:
SUNSCPAL_TRACE("msg_in: CMDCOMPLETE\n", 0);
/* Target is about to disconnect. */
act_flags |= (ACT_DISCONNECT | ACT_CMD_DONE);
break;
case MSG_PARITY_ERROR:
SUNSCPAL_TRACE("msg_in: PARITY_ERROR\n", 0);
/* Resend the last message. */
sunscpal_sched_msgout(sc, sc->sc_msgout);
break;
case MSG_MESSAGE_REJECT:
/* The target rejects the last message we sent. */
SUNSCPAL_TRACE("msg_in: got reject for 0x%x\n", sc->sc_msgout);
switch (sc->sc_msgout) {
case SEND_IDENTIFY:
/* Really old target controller? */
/* XXX ... */
break;
case SEND_INIT_DET_ERR:
goto abort;
}
break;
case MSG_NOOP:
SUNSCPAL_TRACE("msg_in: NOOP\n", 0);
break;
case MSG_DISCONNECT:
SUNSCPAL_TRACE("msg_in: DISCONNECT\n", 0);
/* Target is about to disconnect. */
act_flags |= ACT_DISCONNECT;
if ((xs->xs_periph->periph_quirks & PQUIRK_AUTOSAVE) == 0)
break;
/*FALLTHROUGH*/
case MSG_SAVEDATAPOINTER:
SUNSCPAL_TRACE("msg_in: SAVE_PTRS\n", 0);
sr->sr_dataptr = sc->sc_dataptr;
sr->sr_datalen = sc->sc_datalen;
break;
case MSG_RESTOREPOINTERS:
SUNSCPAL_TRACE("msg_in: RESTORE_PTRS\n", 0);
sc->sc_dataptr = sr->sr_dataptr;
sc->sc_datalen = sr->sr_datalen;
break;
case MSG_EXTENDED:
switch (sc->sc_imess[2]) {
case MSG_EXT_SDTR:
case MSG_EXT_WDTR:
/* The ncr5380 can not do synchronous mode. */
goto reject;
default:
printf("%s: unrecognized MESSAGE EXTENDED; sending REJECT\n",
sc->sc_dev.dv_xname);
SUNSCPAL_BREAK();
goto reject;
}
break;
default:
SUNSCPAL_TRACE("msg_in: eh? imsg=0x%x\n", sc->sc_imess[0]);
printf("%s: unrecognized MESSAGE; sending REJECT\n",
sc->sc_dev.dv_xname);
SUNSCPAL_BREAK();
/* fallthrough */
reject:
sunscpal_sched_msgout(sc, SEND_REJECT);
break;
abort:
sc->sc_state |= SUNSCPAL_ABORTING;
sunscpal_sched_msgout(sc, SEND_ABORT);
break;
}
/* Go get the next message, if any. */
if (act_flags == ACT_CONTINUE)
goto nextmsg;
return (act_flags);
}
/*
* The message out (and in) stuff is a bit complicated:
* If the target requests another message (sequence) without
* having changed phase in between it really asks for a
* retransmit, probably due to parity error(s).
* The following messages can be sent:
* IDENTIFY @ These 4 stem from SCSI command activity
* SDTR @
* WDTR @
* DEV_RESET @
* REJECT if MSGI doesn't make sense
* PARITY_ERROR if parity error while in MSGI
* INIT_DET_ERR if parity error while not in MSGI
* ABORT if INIT_DET_ERR rejected
* NOOP if asked for a message and there's nothing to send
*
* Note that we call this one with (sc_current == NULL)
* when sending ABORT for unwanted reselections.
*/
static int
sunscpal_msg_out(sc)
struct sunscpal_softc *sc;
{
/*
* This controller does not allow you to assert ATN, which
* means we will never get the opportunity to send messages to
* the target (the bus will never enter this MSG_OUT phase).
* This will eventually leave us with no option other than to
* reset the bus. We keep this function as a placeholder,
* though, and this printf will eventually go away or get
* #ifdef'ed:
*/
printf("sunscpal_msg_out: bus is in MSG_OUT phase?\n");
return (ACT_CONTINUE | ACT_RESET_BUS);
}
/*
* Handle command phase.
*/
static int
sunscpal_command(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr = sc->sc_current;
struct scsipi_xfer *xs = sr->sr_xs;
int len;
/* Assume command can be sent in one go. */
/* XXX: Do this using DMA, and get a phase change intr? */
len = sunscpal_pio_out(sc, SUNSCPAL_PHASE_COMMAND, xs->cmdlen,
(u_char *)xs->cmd);
if (len != xs->cmdlen) {
#ifdef SUNSCPAL_DEBUG
printf("sunscpal_command: short transfer: wanted %d got %d.\n",
xs->cmdlen, len);
sunscpal_show_scsi_cmd(xs);
SUNSCPAL_BREAK();
#endif
if (len < 6) {
xs->error = XS_DRIVER_STUFFUP;
sc->sc_state |= SUNSCPAL_ABORTING;
sunscpal_sched_msgout(sc, SEND_ABORT);
}
}
return ACT_CONTINUE;
}
/*
* Handle either data_in or data_out
*/
static int
sunscpal_data_xfer(sc, phase)
struct sunscpal_softc *sc;
int phase;
{
struct sunscpal_req *sr = sc->sc_current;
struct scsipi_xfer *xs = sr->sr_xs;
int expected_phase;
int len;
/*
* When aborting a command, disallow any data phase.
*/
if (sc->sc_state & SUNSCPAL_ABORTING) {
printf("%s: aborting, bus phase=%s (reset)\n",
sc->sc_dev.dv_xname, phase_names[(phase >> 8) & 7]);
return ACT_RESET_BUS; /* XXX */
}
/* Validate expected phase (data_in or data_out) */
expected_phase = (xs->xs_control & XS_CTL_DATA_OUT) ?
SUNSCPAL_PHASE_DATA_OUT : SUNSCPAL_PHASE_DATA_IN;
if (phase != expected_phase) {
printf("%s: data phase error\n", sc->sc_dev.dv_xname);
goto abort;
}
/* Make sure we have some data to move. */
if (sc->sc_datalen <= 0) {
/* Device needs padding. */
if (phase == SUNSCPAL_PHASE_DATA_IN)
sunscpal_pio_in(sc, phase, 4096, NULL);
else
sunscpal_pio_out(sc, phase, 4096, NULL);
/* Make sure that caused a phase change. */
if (SUNSCPAL_BUS_PHASE(SUNSCPAL_READ_2(sc, sunscpal_icr)) == phase) {
/* More than 4k is just too much! */
printf("%s: too much data padding\n",
sc->sc_dev.dv_xname);
goto abort;
}
return ACT_CONTINUE;
}
/*
* Attempt DMA only if dma_alloc gave us a DMA handle AND
* there is enough left to transfer so DMA is worth while.
*/
if (sr->sr_dma_hand &&
(sc->sc_datalen >= sc->sc_min_dma_len))
{
/*
* OK, really start DMA. Note, the MD start function
* is responsible for setting the TCMD register, etc.
* (Acknowledge the phase change there, not here.)
*/
SUNSCPAL_TRACE("data_xfer: dma_start, dh=0x%x\n",
(long) sr->sr_dma_hand);
sunscpal_dma_start(sc);
return ACT_WAIT_DMA;
}
/*
* Doing PIO for data transfer. (Possibly "Pseudo DMA")
* XXX: Do PDMA functions need to set tcmd later?
*/
SUNSCPAL_TRACE("data_xfer: doing PIO, len=%d\n", sc->sc_datalen);
if (phase == SUNSCPAL_PHASE_DATA_OUT) {
len = sunscpal_pio_out(sc, phase, sc->sc_datalen, sc->sc_dataptr);
} else {
len = sunscpal_pio_in(sc, phase, sc->sc_datalen, sc->sc_dataptr);
}
sc->sc_dataptr += len;
sc->sc_datalen -= len;
SUNSCPAL_TRACE("data_xfer: did PIO, resid=%d\n", sc->sc_datalen);
return (ACT_CONTINUE);
abort:
sc->sc_state |= SUNSCPAL_ABORTING;
sunscpal_sched_msgout(sc, SEND_ABORT);
return (ACT_CONTINUE);
}
static int
sunscpal_status(sc)
struct sunscpal_softc *sc;
{
int len;
u_char status;
struct sunscpal_req *sr = sc->sc_current;
len = sunscpal_pio_in(sc, SUNSCPAL_PHASE_STATUS, 1, &status);
if (len) {
sr->sr_status = status;
} else {
printf("sunscpal_status: none?\n");
}
return ACT_CONTINUE;
}
/*
* This is the big state machine that follows SCSI phase changes.
* This is somewhat like a co-routine. It will do a SCSI command,
* and exit if the command is complete, or if it must wait, i.e.
* for DMA to complete or for reselect to resume the job.
*
* The bus must be selected, and we need to know which command is
* being undertaken.
*/
static void
sunscpal_machine(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr;
struct scsipi_xfer *xs;
int act_flags, phase, timo;
#ifdef DIAGNOSTIC
if (sc->sc_state == SUNSCPAL_IDLE)
panic("sunscpal_machine: state=idle");
if (sc->sc_current == NULL)
panic("sunscpal_machine: no current cmd");
#endif
sr = sc->sc_current;
xs = sr->sr_xs;
act_flags = ACT_CONTINUE;
/*
* This will be called by sunscpal_intr() when DMA is
* complete. Must stop DMA before touching the PAL or
* there will be "register conflict" errors.
*/
if (sc->sc_state & SUNSCPAL_DOINGDMA) {
/* Pick-up where where we left off... */
goto dma_done;
}
next_phase:
if (!SUNSCPAL_BUSY(sc)) {
/* Unexpected disconnect */
printf("sunscpal_machine: unexpected disconnect.\n");
xs->error = XS_DRIVER_STUFFUP;
act_flags |= (ACT_DISCONNECT | ACT_CMD_DONE);
goto do_actions;
}
/*
* Wait for REQ before reading the phase.
* Need to wait longer than usual here, because
* some devices are just plain slow...
*/
timo = sunscpal_wait_phase_timo;
for (;;) {
if (SUNSCPAL_READ_2(sc, sunscpal_icr) & SUNSCPAL_ICR_REQUEST)
break;
if (--timo <= 0) {
if (sc->sc_state & SUNSCPAL_ABORTING) {
printf("%s: no REQ while aborting, reset\n",
sc->sc_dev.dv_xname);
act_flags |= ACT_RESET_BUS;
goto do_actions;
}
printf("%s: no REQ for next phase, abort\n",
sc->sc_dev.dv_xname);
sc->sc_state |= SUNSCPAL_ABORTING;
sunscpal_sched_msgout(sc, SEND_ABORT);
goto next_phase;
}
delay(100);
}
phase = SUNSCPAL_BUS_PHASE(SUNSCPAL_READ_2(sc, sunscpal_icr));
SUNSCPAL_TRACE("machine: phase=%s\n",
(long) phase_names[(phase >> 8) & 7]);
/*
* We assume that the device knows what it's doing,
* so any phase is good.
*/
switch (phase) {
case SUNSCPAL_PHASE_DATA_OUT:
case SUNSCPAL_PHASE_DATA_IN:
act_flags = sunscpal_data_xfer(sc, phase);
break;
case SUNSCPAL_PHASE_COMMAND:
act_flags = sunscpal_command(sc);
break;
case SUNSCPAL_PHASE_STATUS:
act_flags = sunscpal_status(sc);
break;
case SUNSCPAL_PHASE_MSG_OUT:
act_flags = sunscpal_msg_out(sc);
break;
case SUNSCPAL_PHASE_MSG_IN:
act_flags = sunscpal_msg_in(sc);
break;
default:
printf("sunscpal_machine: Unexpected phase 0x%x\n", phase);
sc->sc_state |= SUNSCPAL_ABORTING;
sunscpal_sched_msgout(sc, SEND_ABORT);
goto next_phase;
} /* switch */
sc->sc_prevphase = phase;
do_actions:
if (act_flags & ACT_WAIT_DMA) {
act_flags &= ~ACT_WAIT_DMA;
/* Wait for DMA to complete (polling, or interrupt). */
if ((sr->sr_flags & SR_IMMED) == 0) {
SUNSCPAL_TRACE("machine: wait for DMA intr.\n", 0);
return; /* will resume at dma_done */
}
/* Busy-wait for it to finish. */
SUNSCPAL_TRACE("machine: dma_poll, dh=0x%x\n",
(long) sr->sr_dma_hand);
sunscpal_dma_poll(sc);
dma_done:
/* Return here after interrupt. */
if (sr->sr_flags & SR_OVERDUE)
sc->sc_state |= SUNSCPAL_ABORTING;
SUNSCPAL_TRACE("machine: dma_stop, dh=0x%x\n",
(long) sr->sr_dma_hand);
sunscpal_dma_stop(sc);
SUNSCPAL_CLR_INTR(sc); /* XXX */
/*
* While DMA is running we can not touch the SBC,
* so various places just set SUNSCPAL_ABORTING and
* expect us the "kick it" when DMA is done.
*/
if (sc->sc_state & SUNSCPAL_ABORTING) {
sunscpal_sched_msgout(sc, SEND_ABORT);
}
}
/*
* Check for parity error.
* XXX - better place to check?
*/
if (SUNSCPAL_READ_2(sc, sunscpal_icr) & SUNSCPAL_ICR_PARITY_ERROR) {
printf("%s: parity error!\n", sc->sc_dev.dv_xname);
/* XXX: sc->sc_state |= SUNSCPAL_ABORTING; */
sunscpal_sched_msgout(sc, SEND_PARITY_ERROR);
}
if (act_flags == ACT_CONTINUE)
goto next_phase;
/* All other actions "break" from the loop. */
SUNSCPAL_TRACE("machine: act_flags=0x%x\n", act_flags);
if (act_flags & ACT_RESET_BUS) {
act_flags |= ACT_CMD_DONE;
/*
* Reset the SCSI bus, usually due to a timeout.
* The error code XS_TIMEOUT allows retries.
*/
sc->sc_state |= SUNSCPAL_ABORTING;
printf("%s: reset SCSI bus for TID=%d LUN=%d\n",
sc->sc_dev.dv_xname, sr->sr_target, sr->sr_lun);
sunscpal_reset_scsibus(sc);
}
if (act_flags & ACT_CMD_DONE) {
act_flags |= ACT_DISCONNECT;
/* Need to call scsipi_done() */
/* XXX: from the aic6360 driver, but why? */
if (sc->sc_datalen < 0) {
printf("%s: %d extra bytes from %d:%d\n",
sc->sc_dev.dv_xname, -sc->sc_datalen,
sr->sr_target, sr->sr_lun);
sc->sc_datalen = 0;
}
xs->resid = sc->sc_datalen;
/* Note: this will clear sc_current */
SUNSCPAL_TRACE("machine: call done, cur=0x%x\n", (long)sr);
sunscpal_done(sc);
}
if (act_flags & ACT_DISCONNECT) {
/*
* The device has dropped BSY (or will soon).
* We have to wait here for BSY to drop, otherwise
* the next command may decide we need a bus reset.
*/
timo = sunscpal_wait_req_timo; /* XXX */
for (;;) {
if (!SUNSCPAL_BUSY(sc))
goto busfree;
if (--timo <= 0)
break;
delay(2);
}
/* Device is sitting on the bus! */
printf("%s: Target %d LUN %d stuck busy, resetting...\n",
sc->sc_dev.dv_xname, sr->sr_target, sr->sr_lun);
sunscpal_reset_scsibus(sc);
busfree:
SUNSCPAL_TRACE("machine: discon, waited %d\n",
sunscpal_wait_req_timo - timo);
SUNSCPAL_WRITE_2(sc, sunscpal_icr, 0);
if ((act_flags & ACT_CMD_DONE) == 0) {
SUNSCPAL_TRACE("machine: discon, cur=0x%x\n", (long)sr);
}
/*
* We may be here due to a disconnect message,
* in which case we did NOT call sunscpal_done,
* and we need to clear sc_current.
*/
sc->sc_state = SUNSCPAL_IDLE;
sc->sc_current = NULL;
/* Paranoia: clear everything. */
sc->sc_dataptr = NULL;
sc->sc_datalen = 0;
sc->sc_prevphase = SUNSCPAL_PHASE_INVALID;
sc->sc_msgpriq = 0;
sc->sc_msgoutq = 0;
sc->sc_msgout = 0;
/* Our caller will re-enable interrupts. */
}
}
#ifdef SUNSCPAL_DEBUG
static void
sunscpal_show_scsi_cmd(xs)
struct scsipi_xfer *xs;
{
u_char *b = (u_char *) xs->cmd;
int i = 0;
scsipi_printaddr(xs->xs_periph);
if ( ! ( xs->xs_control & XS_CTL_RESET ) ) {
printf("-");
while (i < xs->cmdlen) {
if (i) printf(",");
printf("%x",b[i++]);
}
printf("-\n");
} else {
printf("-RESET-\n");
}
}
int sunscpal_traceidx = 0;
#define TRACE_MAX 1024
struct trace_ent {
char *msg;
long val;
} sunscpal_tracebuf[TRACE_MAX];
void
sunscpal_trace(msg, val)
char *msg;
long val;
{
struct trace_ent *tr;
int s;
s = splbio();
tr = &sunscpal_tracebuf[sunscpal_traceidx];
sunscpal_traceidx++;
if (sunscpal_traceidx >= TRACE_MAX)
sunscpal_traceidx = 0;
tr->msg = msg;
tr->val = val;
splx(s);
}
#ifdef DDB
void
sunscpal_clear_trace()
{
sunscpal_traceidx = 0;
memset((char*) sunscpal_tracebuf, 0, sizeof(sunscpal_tracebuf));
}
void
sunscpal_show_trace()
{
struct trace_ent *tr;
int idx;
idx = sunscpal_traceidx;
do {
tr = &sunscpal_tracebuf[idx];
idx++;
if (idx >= TRACE_MAX)
idx = 0;
if (tr->msg)
db_printf(tr->msg, tr->val);
} while (idx != sunscpal_traceidx);
}
void
sunscpal_show_req(sr)
struct sunscpal_req *sr;
{
struct scsipi_xfer *xs = sr->sr_xs;
db_printf("TID=%d ", sr->sr_target);
db_printf("LUN=%d ", sr->sr_lun);
db_printf("dh=%p ", sr->sr_dma_hand);
db_printf("dptr=%p ", sr->sr_dataptr);
db_printf("dlen=0x%x ", sr->sr_datalen);
db_printf("flags=%d ", sr->sr_flags);
db_printf("stat=%d ", sr->sr_status);
if (xs == NULL) {
db_printf("(xs=NULL)\n");
return;
}
db_printf("\n");
#ifdef SCSIDEBUG
show_scsipi_xs(xs);
#else
db_printf("xs=%p\n", xs);
#endif
}
void
sunscpal_show_state()
{
struct sunscpal_softc *sc;
struct sunscpal_req *sr;
int i, j, k;
sc = sunscpal_debug_sc;
if (sc == NULL) {
db_printf("sunscpal_debug_sc == NULL\n");
return;
}
db_printf("sc_ncmds=%d\n", sc->sc_ncmds);
k = -1; /* which is current? */
for (i = 0; i < SUNSCPAL_OPENINGS; i++) {
sr = &sc->sc_ring[i];
if (sr->sr_xs) {
if (sr == sc->sc_current)
k = i;
db_printf("req %d: (sr=%p)", i, sr);
sunscpal_show_req(sr);
}
}
db_printf("sc_rr=%d, current=%d\n", sc->sc_rr, k);
db_printf("Active request matrix:\n");
for(i = 0; i < 8; i++) { /* targets */
for (j = 0; j < 8; j++) { /* LUN */
sr = sc->sc_matrix[i][j];
if (sr) {
db_printf("TID=%d LUN=%d sr=%p\n", i, j, sr);
}
}
}
db_printf("sc_state=0x%x\n", sc->sc_state);
db_printf("sc_current=%p\n", sc->sc_current);
db_printf("sc_dataptr=%p\n", sc->sc_dataptr);
db_printf("sc_datalen=0x%x\n", sc->sc_datalen);
db_printf("sc_prevphase=%d\n", sc->sc_prevphase);
db_printf("sc_msgpriq=0x%x\n", sc->sc_msgpriq);
}
#endif /* DDB */
#endif /* SUNSCPAL_DEBUG */
void
sunscpal_attach(sc, options)
struct sunscpal_softc *sc;
int options;
{
/*
* Handle our options.
*/
printf(": options=0x%x\n", options);
sc->sc_parity_disable = (options & SUNSCPAL_OPT_NO_PARITY_CHK);
if (options & SUNSCPAL_OPT_DISABLE_DMA)
sc->sc_flags |= SUNSCPAL_DISABLE_DMA;
/*
* Fill in the adapter.
*/
memset(&sc->sc_adapter, 0, sizeof(sc->sc_adapter));
sc->sc_adapter.adapt_dev = &sc->sc_dev;
sc->sc_adapter.adapt_nchannels = 1;
sc->sc_adapter.adapt_openings = SUNSCPAL_OPENINGS;
sc->sc_adapter.adapt_max_periph = 1;
sc->sc_adapter.adapt_request = sunscpal_scsipi_request;
sc->sc_adapter.adapt_minphys = sunscpal_minphys;
if (options & SUNSCPAL_OPT_FORCE_POLLING)
sc->sc_adapter.adapt_flags |= SCSIPI_ADAPT_POLL_ONLY;
sc->sc_channel.chan_adapter = &sc->sc_adapter;
sc->sc_channel.chan_bustype = &scsi_bustype;
sc->sc_channel.chan_channel = 0;
sc->sc_channel.chan_ntargets = 8;
sc->sc_channel.chan_nluns = 8;
sc->sc_channel.chan_id = 7;
/*
* Add reference to adapter so that we drop the reference after
* config_found() to make sure the adatper is disabled.
*/
if (scsipi_adapter_addref(&sc->sc_adapter) != 0) {
printf("%s: unable to enable controller\n",
sc->sc_dev.dv_xname);
return;
}
sunscpal_init(sc); /* Init chip and driver */
sunscpal_reset_scsibus(sc);
/*
* Ask the adapter what subunits are present
*/
(void) config_found(&sc->sc_dev, &sc->sc_channel, scsiprint);
scsipi_adapter_delref(&sc->sc_adapter);
}
int
sunscpal_detach(sc, flags)
struct sunscpal_softc *sc;
int flags;
{
return (EOPNOTSUPP);
}
static void
sunscpal_minphys(struct buf *bp)
{
if (bp->b_bcount > SUNSCPAL_MAX_DMA_LEN) {
#ifdef SUNSCPAL_DEBUG
if (sunscpal_debug & SUNSCPAL_DBG_DMA) {
printf("sunscpal_minphys len = 0x%lx.\n", bp->b_bcount);
Debugger();
}
#endif
bp->b_bcount = SUNSCPAL_MAX_DMA_LEN;
}
return (minphys(bp));
}
#ifdef SUNSCPAL_USE_BUS_DMA
/*
* Allocate a DMA handle and put it in sr->sr_dma_hand. Prepare
* for DMA transfer.
*/
static void
sunscpal_dma_alloc(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr = sc->sc_current;
sunscpal_dma_handle_t dh;
int i, xlen;
u_long addr;
#ifdef DIAGNOSTIC
if (sr->sr_dma_hand != NULL)
panic("sunscpal_dma_alloc: already have DMA handle");
#endif
addr = (u_long) sc->sc_dataptr;
xlen = sc->sc_datalen;
/* If the DMA start addr is misaligned then do PIO */
if ((addr & 1) || (xlen & 1)) {
printf("sunscpal_dma_alloc: misaligned.\n");
return;
}
/* Make sure our caller checked sc_min_dma_len. */
if (xlen < sc->sc_min_dma_len)
panic("sunscpal_dma_alloc: xlen=0x%x", xlen);
/*
* Never attempt single transfers of more than 63k, because
* our count register is only 16 bits.
* This should never happen since already bounded by minphys().
* XXX - Should just segment these...
*/
if (xlen > SUNSCPAL_MAX_DMA_LEN) {
printf("sunscpal_dma_alloc: excessive xlen=0x%x\n", xlen);
Debugger();
sc->sc_datalen = xlen = SUNSCPAL_MAX_DMA_LEN;
}
/* Find free DMA handle. Guaranteed to find one since we have
as many DMA handles as the driver has processes. */
for (i = 0; i < SUNSCPAL_OPENINGS; i++) {
if ((sc->sc_dma_handles[i].dh_flags & SUNSCDH_BUSY) == 0)
goto found;
}
panic("sc: no free DMA handles.");
found:
dh = &sc->sc_dma_handles[i];
dh->dh_flags = SUNSCDH_BUSY;
dh->dh_mapaddr = (u_char*) addr;
dh->dh_maplen = xlen;
dh->dh_dvma = 0;
/* Load the DMA map. */
if (bus_dmamap_load(sc->sunscpal_dmat, dh->dh_dmamap, dh->dh_mapaddr, dh->dh_maplen, NULL, BUS_DMA_NOWAIT) != 0) {
/* Can't load map */
printf("sunscpal_dma_alloc: can't DMA %p/0x%x\n",
dh->dh_mapaddr, dh->dh_maplen);
dh->dh_flags = 0;
return;
}
/* success */
sr->sr_dma_hand = dh;
return;
}
static void
sunscpal_dma_free(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr = sc->sc_current;
sunscpal_dma_handle_t dh = sr->sr_dma_hand;
#ifdef DIAGNOSTIC
if (dh == NULL)
panic("sunscpal_dma_free: no DMA handle");
#endif
if (sc->sc_state & SUNSCPAL_DOINGDMA)
panic("sunscpal_dma_free: free while in progress");
if (dh->dh_flags & SUNSCDH_BUSY) {
/* XXX - Should separate allocation and mapping. */
/* Give back the DVMA space. */
bus_dmamap_unload(sc->sunscpal_dmat, dh->dh_dmamap);
dh->dh_flags = 0;
}
sr->sr_dma_hand = NULL;
}
/*
* This function is called during the SELECT phase that
* precedes a COMMAND phase, in case we need to setup the
* DMA engine before the bus enters a DATA phase.
*
* On the sc version, setup the start address and the count.
*/
static void
sunscpal_dma_setup(sc)
struct sunscpal_softc *sc;
{
struct sunscpal_req *sr = sc->sc_current;
struct scsipi_xfer *xs = sr->sr_xs;
sunscpal_dma_handle_t dh = sr->sr_dma_hand;
long data_pa;
int xlen;
/*
* Get the DVMA mapping for this segment.
* XXX - Should separate allocation and mapin.
*/
data_pa = dh->dh_dvma;
data_pa += (sc->sc_dataptr - dh->dh_mapaddr);
if (data_pa & 1)
panic("sunscpal_dma_setup: bad pa=0x%lx", data_pa);
xlen = sc->sc_datalen;
if (xlen & 1)
panic("sunscpal_dma_setup: bad xlen=0x%x", xlen);
sc->sc_reqlen = xlen; /* XXX: or less? */
#ifdef SUNSCPAL_DEBUG
if (sunscpal_debug & SUNSCPAL_DBG_DMA) {
printf("sunscpal_dma_setup: dh=%p, pa=0x%lx, xlen=0x%x\n",
dh, data_pa, xlen);
}
#endif
/* sync the DMA map: */
bus_dmamap_sync(sc->sunscpal_dmat, dh->dh_dmamap, 0, dh->dh_maplen,
((xs->xs_control & XS_CTL_DATA_OUT) == 0 ? BUS_DMASYNC_PREREAD : BUS_DMASYNC_PREWRITE));
/* Load the start address and the count. */
SUNSCPAL_WRITE_2(sc, sunscpal_dma_addr_h, (data_pa >> 16) & 0xFFFF);
SUNSCPAL_WRITE_2(sc, sunscpal_dma_addr_l, (data_pa >> 0) & 0xFFFF);
SUNSCPAL_WRITE_2(sc, sunscpal_dma_count, SUNSCPAL_DMA_COUNT_FLIP(xlen));
}
#endif /* SUNSCPAL_USE_BUS_DMA */