UbiquitousOS
/
OpenCloudOS-Kernel
mirror of https://gitee.com/OpenCloudOS/OpenCloudOS-Kernel.git
11
0
Fork 0
Code Issues Projects Releases Wiki Activity
OpenCloudOS-Kernel/drivers/scsi/qlogicpti.c

1483 lines
40 KiB
C
Raw Normal View History

Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* qlogicpti.c: Performance Technologies QlogicISP sbus card driver.
*
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
* Copyright (C) 1996, 2006, 2008 David S. Miller (davem@davemloft.net)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
*
* A lot of this driver was directly stolen from Erik H. Moe's PCI
* Qlogic ISP driver. Mucho kudos to him for this code.
*
* An even bigger kudos to John Grana at Performance Technologies
* for providing me with the hardware to write this driver, you rule
* John you really do.
*
* May, 2, 1997: Added support for QLGC,isp --jj
*/
#include <linux/kernel.h>
#include <linux/delay.h>
#include <linux/types.h>
#include <linux/string.h>
include cleanup: Update gfp.h and slab.h includes to prepare for breaking implicit slab.h inclusion from percpu.h percpu.h is included by sched.h and module.h and thus ends up being included when building most .c files. percpu.h includes slab.h which in turn includes gfp.h making everything defined by the two files universally available and complicating inclusion dependencies. percpu.h -> slab.h dependency is about to be removed. Prepare for this change by updating users of gfp and slab facilities include those headers directly instead of assuming availability. As this conversion needs to touch large number of source files, the following script is used as the basis of conversion. http://userweb.kernel.org/~tj/misc/slabh-sweep.py The script does the followings. * Scan files for gfp and slab usages and update includes such that only the necessary includes are there. ie. if only gfp is used, gfp.h, if slab is used, slab.h. * When the script inserts a new include, it looks at the include blocks and try to put the new include such that its order conforms to its surrounding. It's put in the include block which contains core kernel includes, in the same order that the rest are ordered - alphabetical, Christmas tree, rev-Xmas-tree or at the end if there doesn't seem to be any matching order. * If the script can't find a place to put a new include (mostly because the file doesn't have fitting include block), it prints out an error message indicating which .h file needs to be added to the file. The conversion was done in the following steps. 1. The initial automatic conversion of all .c files updated slightly over 4000 files, deleting around 700 includes and adding ~480 gfp.h and ~3000 slab.h inclusions. The script emitted errors for ~400 files. 2. Each error was manually checked. Some didn't need the inclusion, some needed manual addition while adding it to implementation .h or embedding .c file was more appropriate for others. This step added inclusions to around 150 files. 3. The script was run again and the output was compared to the edits from #2 to make sure no file was left behind. 4. Several build tests were done and a couple of problems were fixed. e.g. lib/decompress_*.c used malloc/free() wrappers around slab APIs requiring slab.h to be added manually. 5. The script was run on all .h files but without automatically editing them as sprinkling gfp.h and slab.h inclusions around .h files could easily lead to inclusion dependency hell. Most gfp.h inclusion directives were ignored as stuff from gfp.h was usually wildly available and often used in preprocessor macros. Each slab.h inclusion directive was examined and added manually as necessary. 6. percpu.h was updated not to include slab.h. 7. Build test were done on the following configurations and failures were fixed. CONFIG_GCOV_KERNEL was turned off for all tests (as my distributed build env didn't work with gcov compiles) and a few more options had to be turned off depending on archs to make things build (like ipr on powerpc/64 which failed due to missing writeq). * x86 and x86_64 UP and SMP allmodconfig and a custom test config. * powerpc and powerpc64 SMP allmodconfig * sparc and sparc64 SMP allmodconfig * ia64 SMP allmodconfig * s390 SMP allmodconfig * alpha SMP allmodconfig * um on x86_64 SMP allmodconfig 8. percpu.h modifications were reverted so that it could be applied as a separate patch and serve as bisection point. Given the fact that I had only a couple of failures from tests on step 6, I'm fairly confident about the coverage of this conversion patch. If there is a breakage, it's likely to be something in one of the arch headers which should be easily discoverable easily on most builds of the specific arch. Signed-off-by: Tejun Heo <tj@kernel.org> Guess-its-ok-by: Christoph Lameter <cl@linux-foundation.org> Cc: Ingo Molnar <mingo@redhat.com> Cc: Lee Schermerhorn <Lee.Schermerhorn@hp.com>
2010-03-24 16:04:11 +08:00
#include <linux/gfp.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#include <linux/blkdev.h>
#include <linux/proc_fs.h>
#include <linux/stat.h>
#include <linux/init.h>
#include <linux/spinlock.h>
#include <linux/interrupt.h>
#include <linux/module.h>
[PATCH] drivers/scsi/*: use time_after() and friends They deal with wrapping correctly and are nicer to read. Signed-off-by: Marcelo Feitoza Parisi <marcelo@feitoza.com.br> Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Cc: James Bottomley <James.Bottomley@steeleye.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-28 17:56:47 +08:00
#include <linux/jiffies.h>
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
#include <linux/dma-mapping.h>
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
#include <linux/of.h>
#include <linux/of_device.h>
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
#include <linux/firmware.h>
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#include <asm/byteorder.h>
#include "qlogicpti.h"
#include <asm/dma.h>
#include <asm/ptrace.h>
#include <asm/pgtable.h>
#include <asm/oplib.h>
#include <asm/io.h>
#include <asm/irq.h>
#include <scsi/scsi.h>
#include <scsi/scsi_cmnd.h>
#include <scsi/scsi_device.h>
#include <scsi/scsi_eh.h>
#include <scsi/scsi_tcq.h>
#include <scsi/scsi_host.h>
#define MAX_TARGETS 16
#define MAX_LUNS 8 /* 32 for 1.31 F/W */
#define DEFAULT_LOOP_COUNT 10000
static struct qlogicpti *qptichain = NULL;
static DEFINE_SPINLOCK(qptichain_lock);
#define PACKB(a, b) (((a)<<4)|(b))
static const u_char mbox_param[] = {
PACKB(1, 1), /* MBOX_NO_OP */
PACKB(5, 5), /* MBOX_LOAD_RAM */
PACKB(2, 0), /* MBOX_EXEC_FIRMWARE */
PACKB(5, 5), /* MBOX_DUMP_RAM */
PACKB(3, 3), /* MBOX_WRITE_RAM_WORD */
PACKB(2, 3), /* MBOX_READ_RAM_WORD */
PACKB(6, 6), /* MBOX_MAILBOX_REG_TEST */
PACKB(2, 3), /* MBOX_VERIFY_CHECKSUM */
PACKB(1, 3), /* MBOX_ABOUT_FIRMWARE */
PACKB(0, 0), /* 0x0009 */
PACKB(0, 0), /* 0x000a */
PACKB(0, 0), /* 0x000b */
PACKB(0, 0), /* 0x000c */
PACKB(0, 0), /* 0x000d */
PACKB(1, 2), /* MBOX_CHECK_FIRMWARE */
PACKB(0, 0), /* 0x000f */
PACKB(5, 5), /* MBOX_INIT_REQ_QUEUE */
PACKB(6, 6), /* MBOX_INIT_RES_QUEUE */
PACKB(4, 4), /* MBOX_EXECUTE_IOCB */
PACKB(2, 2), /* MBOX_WAKE_UP */
PACKB(1, 6), /* MBOX_STOP_FIRMWARE */
PACKB(4, 4), /* MBOX_ABORT */
PACKB(2, 2), /* MBOX_ABORT_DEVICE */
PACKB(3, 3), /* MBOX_ABORT_TARGET */
PACKB(2, 2), /* MBOX_BUS_RESET */
PACKB(2, 3), /* MBOX_STOP_QUEUE */
PACKB(2, 3), /* MBOX_START_QUEUE */
PACKB(2, 3), /* MBOX_SINGLE_STEP_QUEUE */
PACKB(2, 3), /* MBOX_ABORT_QUEUE */
PACKB(2, 4), /* MBOX_GET_DEV_QUEUE_STATUS */
PACKB(0, 0), /* 0x001e */
PACKB(1, 3), /* MBOX_GET_FIRMWARE_STATUS */
PACKB(1, 2), /* MBOX_GET_INIT_SCSI_ID */
PACKB(1, 2), /* MBOX_GET_SELECT_TIMEOUT */
PACKB(1, 3), /* MBOX_GET_RETRY_COUNT */
PACKB(1, 2), /* MBOX_GET_TAG_AGE_LIMIT */
PACKB(1, 2), /* MBOX_GET_CLOCK_RATE */
PACKB(1, 2), /* MBOX_GET_ACT_NEG_STATE */
PACKB(1, 2), /* MBOX_GET_ASYNC_DATA_SETUP_TIME */
PACKB(1, 3), /* MBOX_GET_SBUS_PARAMS */
PACKB(2, 4), /* MBOX_GET_TARGET_PARAMS */
PACKB(2, 4), /* MBOX_GET_DEV_QUEUE_PARAMS */
PACKB(0, 0), /* 0x002a */
PACKB(0, 0), /* 0x002b */
PACKB(0, 0), /* 0x002c */
PACKB(0, 0), /* 0x002d */
PACKB(0, 0), /* 0x002e */
PACKB(0, 0), /* 0x002f */
PACKB(2, 2), /* MBOX_SET_INIT_SCSI_ID */
PACKB(2, 2), /* MBOX_SET_SELECT_TIMEOUT */
PACKB(3, 3), /* MBOX_SET_RETRY_COUNT */
PACKB(2, 2), /* MBOX_SET_TAG_AGE_LIMIT */
PACKB(2, 2), /* MBOX_SET_CLOCK_RATE */
PACKB(2, 2), /* MBOX_SET_ACTIVE_NEG_STATE */
PACKB(2, 2), /* MBOX_SET_ASYNC_DATA_SETUP_TIME */
PACKB(3, 3), /* MBOX_SET_SBUS_CONTROL_PARAMS */
PACKB(4, 4), /* MBOX_SET_TARGET_PARAMS */
PACKB(4, 4), /* MBOX_SET_DEV_QUEUE_PARAMS */
PACKB(0, 0), /* 0x003a */
PACKB(0, 0), /* 0x003b */
PACKB(0, 0), /* 0x003c */
PACKB(0, 0), /* 0x003d */
PACKB(0, 0), /* 0x003e */
PACKB(0, 0), /* 0x003f */
PACKB(0, 0), /* 0x0040 */
PACKB(0, 0), /* 0x0041 */
PACKB(0, 0) /* 0x0042 */
};
[SCSI] drivers/scsi: Use ARRAY_SIZE macro Use ARRAY_SIZE macro instead of sizeof(x)/sizeof(x[0]) and remove duplicates of the macro. Signed-off-by: Tobias Klauser <tklauser@nuerscht.ch> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2006-06-09 13:23:48 +08:00
#define MAX_MBOX_COMMAND ARRAY_SIZE(mbox_param)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* queue length's _must_ be power of two: */
#define QUEUE_DEPTH(in, out, ql) ((in - out) & (ql))
#define REQ_QUEUE_DEPTH(in, out) QUEUE_DEPTH(in, out, \
QLOGICPTI_REQ_QUEUE_LEN)
#define RES_QUEUE_DEPTH(in, out) QUEUE_DEPTH(in, out, RES_QUEUE_LEN)
static inline void qlogicpti_enable_irqs(struct qlogicpti *qpti)
{
sbus_writew(SBUS_CTRL_ERIRQ | SBUS_CTRL_GENAB,
qpti->qregs + SBUS_CTRL);
}
static inline void qlogicpti_disable_irqs(struct qlogicpti *qpti)
{
sbus_writew(0, qpti->qregs + SBUS_CTRL);
}
static inline void set_sbus_cfg1(struct qlogicpti *qpti)
{
u16 val;
u8 bursts = qpti->bursts;
#if 0 /* It appears that at least PTI cards do not support
* 64-byte bursts and that setting the B64 bit actually
* is a nop and the chip ends up using the smallest burst
* size. -DaveM
*/
sparc: Move SBUS DMA attribute interfaces out of asm/sbus.h This is in preparation for the subsequent asm/sbus.h removal. Also, make these routines take a "struct device" or no arguments, as appropriate. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 14:33:42 +08:00
if (sbus_can_burst64() && (bursts & DMA_BURST64)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
val = (SBUS_CFG1_BENAB | SBUS_CFG1_B64);
} else
#endif
if (bursts & DMA_BURST32) {
val = (SBUS_CFG1_BENAB | SBUS_CFG1_B32);
} else if (bursts & DMA_BURST16) {
val = (SBUS_CFG1_BENAB | SBUS_CFG1_B16);
} else if (bursts & DMA_BURST8) {
val = (SBUS_CFG1_BENAB | SBUS_CFG1_B8);
} else {
val = 0; /* No sbus bursts for you... */
}
sbus_writew(val, qpti->qregs + SBUS_CFG1);
}
static int qlogicpti_mbox_command(struct qlogicpti *qpti, u_short param[], int force)
{
int loop_count;
u16 tmp;
if (mbox_param[param[0]] == 0)
return 1;
/* Set SBUS semaphore. */
tmp = sbus_readw(qpti->qregs + SBUS_SEMAPHORE);
tmp |= SBUS_SEMAPHORE_LCK;
sbus_writew(tmp, qpti->qregs + SBUS_SEMAPHORE);
/* Wait for host IRQ bit to clear. */
loop_count = DEFAULT_LOOP_COUNT;
while (--loop_count && (sbus_readw(qpti->qregs + HCCTRL) & HCCTRL_HIRQ)) {
barrier();
cpu_relax();
}
if (!loop_count)
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_EMERG "qlogicpti%d: mbox_command loop timeout #1\n",
qpti->qpti_id);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Write mailbox command registers. */
switch (mbox_param[param[0]] >> 4) {
case 6: sbus_writew(param[5], qpti->qregs + MBOX5);
case 5: sbus_writew(param[4], qpti->qregs + MBOX4);
case 4: sbus_writew(param[3], qpti->qregs + MBOX3);
case 3: sbus_writew(param[2], qpti->qregs + MBOX2);
case 2: sbus_writew(param[1], qpti->qregs + MBOX1);
case 1: sbus_writew(param[0], qpti->qregs + MBOX0);
}
/* Clear RISC interrupt. */
tmp = sbus_readw(qpti->qregs + HCCTRL);
tmp |= HCCTRL_CRIRQ;
sbus_writew(tmp, qpti->qregs + HCCTRL);
/* Clear SBUS semaphore. */
sbus_writew(0, qpti->qregs + SBUS_SEMAPHORE);
/* Set HOST interrupt. */
tmp = sbus_readw(qpti->qregs + HCCTRL);
tmp |= HCCTRL_SHIRQ;
sbus_writew(tmp, qpti->qregs + HCCTRL);
/* Wait for HOST interrupt clears. */
loop_count = DEFAULT_LOOP_COUNT;
while (--loop_count &&
(sbus_readw(qpti->qregs + HCCTRL) & HCCTRL_CRIRQ))
udelay(20);
if (!loop_count)
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_EMERG "qlogicpti%d: mbox_command[%04x] loop timeout #2\n",
qpti->qpti_id, param[0]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Wait for SBUS semaphore to get set. */
loop_count = DEFAULT_LOOP_COUNT;
while (--loop_count &&
!(sbus_readw(qpti->qregs + SBUS_SEMAPHORE) & SBUS_SEMAPHORE_LCK)) {
udelay(20);
/* Workaround for some buggy chips. */
if (sbus_readw(qpti->qregs + MBOX0) & 0x4000)
break;
}
if (!loop_count)
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_EMERG "qlogicpti%d: mbox_command[%04x] loop timeout #3\n",
qpti->qpti_id, param[0]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Wait for MBOX busy condition to go away. */
loop_count = DEFAULT_LOOP_COUNT;
while (--loop_count && (sbus_readw(qpti->qregs + MBOX0) == 0x04))
udelay(20);
if (!loop_count)
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_EMERG "qlogicpti%d: mbox_command[%04x] loop timeout #4\n",
qpti->qpti_id, param[0]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* Read back output parameters. */
switch (mbox_param[param[0]] & 0xf) {
case 6: param[5] = sbus_readw(qpti->qregs + MBOX5);
case 5: param[4] = sbus_readw(qpti->qregs + MBOX4);
case 4: param[3] = sbus_readw(qpti->qregs + MBOX3);
case 3: param[2] = sbus_readw(qpti->qregs + MBOX2);
case 2: param[1] = sbus_readw(qpti->qregs + MBOX1);
case 1: param[0] = sbus_readw(qpti->qregs + MBOX0);
}
/* Clear RISC interrupt. */
tmp = sbus_readw(qpti->qregs + HCCTRL);
tmp |= HCCTRL_CRIRQ;
sbus_writew(tmp, qpti->qregs + HCCTRL);
/* Release SBUS semaphore. */
tmp = sbus_readw(qpti->qregs + SBUS_SEMAPHORE);
tmp &= ~(SBUS_SEMAPHORE_LCK);
sbus_writew(tmp, qpti->qregs + SBUS_SEMAPHORE);
/* We're done. */
return 0;
}
static inline void qlogicpti_set_hostdev_defaults(struct qlogicpti *qpti)
{
int i;
qpti->host_param.initiator_scsi_id = qpti->scsi_id;
qpti->host_param.bus_reset_delay = 3;
qpti->host_param.retry_count = 0;
qpti->host_param.retry_delay = 5;
qpti->host_param.async_data_setup_time = 3;
qpti->host_param.req_ack_active_negation = 1;
qpti->host_param.data_line_active_negation = 1;
qpti->host_param.data_dma_burst_enable = 1;
qpti->host_param.command_dma_burst_enable = 1;
qpti->host_param.tag_aging = 8;
qpti->host_param.selection_timeout = 250;
qpti->host_param.max_queue_depth = 256;
for(i = 0; i < MAX_TARGETS; i++) {
/*
* disconnect, parity, arq, reneg on reset, and, oddly enough
* tags...the midlayer's notion of tagged support has to match
* our device settings, and since we base whether we enable a
* tag on a per-cmnd basis upon what the midlayer sez, we
* actually enable the capability here.
*/
qpti->dev_param[i].device_flags = 0xcd;
qpti->dev_param[i].execution_throttle = 16;
if (qpti->ultra) {
qpti->dev_param[i].synchronous_period = 12;
qpti->dev_param[i].synchronous_offset = 8;
} else {
qpti->dev_param[i].synchronous_period = 25;
qpti->dev_param[i].synchronous_offset = 12;
}
qpti->dev_param[i].device_enable = 1;
}
}
static int qlogicpti_reset_hardware(struct Scsi_Host *host)
{
struct qlogicpti *qpti = (struct qlogicpti *) host->hostdata;
u_short param[6];
unsigned short risc_code_addr;
int loop_count, i;
unsigned long flags;
risc_code_addr = 0x1000; /* all load addresses are at 0x1000 */
spin_lock_irqsave(host->host_lock, flags);
sbus_writew(HCCTRL_PAUSE, qpti->qregs + HCCTRL);
/* Only reset the scsi bus if it is not free. */
if (sbus_readw(qpti->qregs + CPU_PCTRL) & CPU_PCTRL_BSY) {
sbus_writew(CPU_ORIDE_RMOD, qpti->qregs + CPU_ORIDE);
sbus_writew(CPU_CMD_BRESET, qpti->qregs + CPU_CMD);
udelay(400);
}
sbus_writew(SBUS_CTRL_RESET, qpti->qregs + SBUS_CTRL);
sbus_writew((DMA_CTRL_CCLEAR | DMA_CTRL_CIRQ), qpti->qregs + CMD_DMA_CTRL);
sbus_writew((DMA_CTRL_CCLEAR | DMA_CTRL_CIRQ), qpti->qregs + DATA_DMA_CTRL);
loop_count = DEFAULT_LOOP_COUNT;
while (--loop_count && ((sbus_readw(qpti->qregs + MBOX0) & 0xff) == 0x04))
udelay(20);
if (!loop_count)
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_EMERG "qlogicpti%d: reset_hardware loop timeout\n",
qpti->qpti_id);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sbus_writew(HCCTRL_PAUSE, qpti->qregs + HCCTRL);
set_sbus_cfg1(qpti);
qlogicpti_enable_irqs(qpti);
if (sbus_readw(qpti->qregs + RISC_PSR) & RISC_PSR_ULTRA) {
qpti->ultra = 1;
sbus_writew((RISC_MTREG_P0ULTRA | RISC_MTREG_P1ULTRA),
qpti->qregs + RISC_MTREG);
} else {
qpti->ultra = 0;
sbus_writew((RISC_MTREG_P0DFLT | RISC_MTREG_P1DFLT),
qpti->qregs + RISC_MTREG);
}
/* reset adapter and per-device default values. */
/* do it after finding out whether we're ultra mode capable */
qlogicpti_set_hostdev_defaults(qpti);
/* Release the RISC processor. */
sbus_writew(HCCTRL_REL, qpti->qregs + HCCTRL);
/* Get RISC to start executing the firmware code. */
param[0] = MBOX_EXEC_FIRMWARE;
param[1] = risc_code_addr;
if (qlogicpti_mbox_command(qpti, param, 1)) {
printk(KERN_EMERG "qlogicpti%d: Cannot execute ISP firmware.\n",
qpti->qpti_id);
spin_unlock_irqrestore(host->host_lock, flags);
return 1;
}
/* Set initiator scsi ID. */
param[0] = MBOX_SET_INIT_SCSI_ID;
param[1] = qpti->host_param.initiator_scsi_id;
if (qlogicpti_mbox_command(qpti, param, 1) ||
(param[0] != MBOX_COMMAND_COMPLETE)) {
printk(KERN_EMERG "qlogicpti%d: Cannot set initiator SCSI ID.\n",
qpti->qpti_id);
spin_unlock_irqrestore(host->host_lock, flags);
return 1;
}
/* Initialize state of the queues, both hw and sw. */
qpti->req_in_ptr = qpti->res_out_ptr = 0;
param[0] = MBOX_INIT_RES_QUEUE;
param[1] = RES_QUEUE_LEN + 1;
param[2] = (u_short) (qpti->res_dvma >> 16);
param[3] = (u_short) (qpti->res_dvma & 0xffff);
param[4] = param[5] = 0;
if (qlogicpti_mbox_command(qpti, param, 1)) {
printk(KERN_EMERG "qlogicpti%d: Cannot init response queue.\n",
qpti->qpti_id);
spin_unlock_irqrestore(host->host_lock, flags);
return 1;
}
param[0] = MBOX_INIT_REQ_QUEUE;
param[1] = QLOGICPTI_REQ_QUEUE_LEN + 1;
param[2] = (u_short) (qpti->req_dvma >> 16);
param[3] = (u_short) (qpti->req_dvma & 0xffff);
param[4] = param[5] = 0;
if (qlogicpti_mbox_command(qpti, param, 1)) {
printk(KERN_EMERG "qlogicpti%d: Cannot init request queue.\n",
qpti->qpti_id);
spin_unlock_irqrestore(host->host_lock, flags);
return 1;
}
param[0] = MBOX_SET_RETRY_COUNT;
param[1] = qpti->host_param.retry_count;
param[2] = qpti->host_param.retry_delay;
qlogicpti_mbox_command(qpti, param, 0);
param[0] = MBOX_SET_TAG_AGE_LIMIT;
param[1] = qpti->host_param.tag_aging;
qlogicpti_mbox_command(qpti, param, 0);
for (i = 0; i < MAX_TARGETS; i++) {
param[0] = MBOX_GET_DEV_QUEUE_PARAMS;
param[1] = (i << 8);
qlogicpti_mbox_command(qpti, param, 0);
}
param[0] = MBOX_GET_FIRMWARE_STATUS;
qlogicpti_mbox_command(qpti, param, 0);
param[0] = MBOX_SET_SELECT_TIMEOUT;
param[1] = qpti->host_param.selection_timeout;
qlogicpti_mbox_command(qpti, param, 0);
for (i = 0; i < MAX_TARGETS; i++) {
param[0] = MBOX_SET_TARGET_PARAMS;
param[1] = (i << 8);
param[2] = (qpti->dev_param[i].device_flags << 8);
/*
* Since we're now loading 1.31 f/w, force narrow/async.
*/
param[2] |= 0xc0;
param[3] = 0; /* no offset, we do not have sync mode yet */
qlogicpti_mbox_command(qpti, param, 0);
}
/*
* Always (sigh) do an initial bus reset (kicks f/w).
*/
param[0] = MBOX_BUS_RESET;
param[1] = qpti->host_param.bus_reset_delay;
qlogicpti_mbox_command(qpti, param, 0);
qpti->send_marker = 1;
spin_unlock_irqrestore(host->host_lock, flags);
return 0;
}
#define PTI_RESET_LIMIT 400
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static int qlogicpti_load_firmware(struct qlogicpti *qpti)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
const struct firmware *fw;
const char fwname[] = "qlogic/isp1000.bin";
const __le16 *fw_data;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
struct Scsi_Host *host = qpti->qhost;
unsigned short csum = 0;
unsigned short param[6];
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
unsigned short risc_code_addr, risc_code_length;
int err;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
unsigned long flags;
int i, timeout;
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
err = request_firmware(&fw, fwname, &qpti->op->dev);
if (err) {
printk(KERN_ERR "Failed to load image \"%s\" err %d\n",
fwname, err);
return err;
}
if (fw->size % 2) {
printk(KERN_ERR "Bogus length %zu in image \"%s\"\n",
fw->size, fwname);
err = -EINVAL;
goto outfirm;
}
fw_data = (const __le16 *)&fw->data[0];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
risc_code_addr = 0x1000; /* all f/w modules load at 0x1000 */
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
risc_code_length = fw->size / 2;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
spin_lock_irqsave(host->host_lock, flags);
/* Verify the checksum twice, one before loading it, and once
* afterwards via the mailbox commands.
*/
for (i = 0; i < risc_code_length; i++)
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
csum += __le16_to_cpu(fw_data[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (csum) {
printk(KERN_EMERG "qlogicpti%d: Aieee, firmware checksum failed!",
qpti->qpti_id);
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
err = 1;
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
sbus_writew(SBUS_CTRL_RESET, qpti->qregs + SBUS_CTRL);
sbus_writew((DMA_CTRL_CCLEAR | DMA_CTRL_CIRQ), qpti->qregs + CMD_DMA_CTRL);
sbus_writew((DMA_CTRL_CCLEAR | DMA_CTRL_CIRQ), qpti->qregs + DATA_DMA_CTRL);
timeout = PTI_RESET_LIMIT;
while (--timeout && (sbus_readw(qpti->qregs + SBUS_CTRL) & SBUS_CTRL_RESET))
udelay(20);
if (!timeout) {
printk(KERN_EMERG "qlogicpti%d: Cannot reset the ISP.", qpti->qpti_id);
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
err = 1;
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
sbus_writew(HCCTRL_RESET, qpti->qregs + HCCTRL);
mdelay(1);
sbus_writew((SBUS_CTRL_GENAB | SBUS_CTRL_ERIRQ), qpti->qregs + SBUS_CTRL);
set_sbus_cfg1(qpti);
sbus_writew(0, qpti->qregs + SBUS_SEMAPHORE);
if (sbus_readw(qpti->qregs + RISC_PSR) & RISC_PSR_ULTRA) {
qpti->ultra = 1;
sbus_writew((RISC_MTREG_P0ULTRA | RISC_MTREG_P1ULTRA),
qpti->qregs + RISC_MTREG);
} else {
qpti->ultra = 0;
sbus_writew((RISC_MTREG_P0DFLT | RISC_MTREG_P1DFLT),
qpti->qregs + RISC_MTREG);
}
sbus_writew(HCCTRL_REL, qpti->qregs + HCCTRL);
/* Pin lines are only stable while RISC is paused. */
sbus_writew(HCCTRL_PAUSE, qpti->qregs + HCCTRL);
if (sbus_readw(qpti->qregs + CPU_PDIFF) & CPU_PDIFF_MODE)
qpti->differential = 1;
else
qpti->differential = 0;
sbus_writew(HCCTRL_REL, qpti->qregs + HCCTRL);
/* This shouldn't be necessary- we've reset things so we should be
running from the ROM now.. */
param[0] = MBOX_STOP_FIRMWARE;
param[1] = param[2] = param[3] = param[4] = param[5] = 0;
if (qlogicpti_mbox_command(qpti, param, 1)) {
printk(KERN_EMERG "qlogicpti%d: Cannot stop firmware for reload.\n",
qpti->qpti_id);
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
err = 1;
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
/* Load it up.. */
for (i = 0; i < risc_code_length; i++) {
param[0] = MBOX_WRITE_RAM_WORD;
param[1] = risc_code_addr + i;
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
param[2] = __le16_to_cpu(fw_data[i]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (qlogicpti_mbox_command(qpti, param, 1) ||
param[0] != MBOX_COMMAND_COMPLETE) {
printk("qlogicpti%d: Firmware dload failed, I'm bolixed!\n",
qpti->qpti_id);
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
err = 1;
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
}
/* Reset the ISP again. */
sbus_writew(HCCTRL_RESET, qpti->qregs + HCCTRL);
mdelay(1);
qlogicpti_enable_irqs(qpti);
sbus_writew(0, qpti->qregs + SBUS_SEMAPHORE);
sbus_writew(HCCTRL_REL, qpti->qregs + HCCTRL);
/* Ask ISP to verify the checksum of the new code. */
param[0] = MBOX_VERIFY_CHECKSUM;
param[1] = risc_code_addr;
if (qlogicpti_mbox_command(qpti, param, 1) ||
(param[0] != MBOX_COMMAND_COMPLETE)) {
printk(KERN_EMERG "qlogicpti%d: New firmware csum failure!\n",
qpti->qpti_id);
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
err = 1;
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
/* Start using newly downloaded firmware. */
param[0] = MBOX_EXEC_FIRMWARE;
param[1] = risc_code_addr;
qlogicpti_mbox_command(qpti, param, 1);
param[0] = MBOX_ABOUT_FIRMWARE;
if (qlogicpti_mbox_command(qpti, param, 1) ||
(param[0] != MBOX_COMMAND_COMPLETE)) {
printk(KERN_EMERG "qlogicpti%d: AboutFirmware cmd fails.\n",
qpti->qpti_id);
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
err = 1;
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
/* Snag the major and minor revisions from the result. */
qpti->fware_majrev = param[1];
qpti->fware_minrev = param[2];
qpti->fware_micrev = param[3];
/* Set the clock rate */
param[0] = MBOX_SET_CLOCK_RATE;
param[1] = qpti->clock;
if (qlogicpti_mbox_command(qpti, param, 1) ||
(param[0] != MBOX_COMMAND_COMPLETE)) {
printk(KERN_EMERG "qlogicpti%d: could not set clock rate.\n",
qpti->qpti_id);
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
err = 1;
goto out;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
if (qpti->is_pti != 0) {
/* Load scsi initiator ID and interrupt level into sbus static ram. */
param[0] = MBOX_WRITE_RAM_WORD;
param[1] = 0xff80;
param[2] = (unsigned short) qpti->scsi_id;
qlogicpti_mbox_command(qpti, param, 1);
param[0] = MBOX_WRITE_RAM_WORD;
param[1] = 0xff00;
param[2] = (unsigned short) 3;
qlogicpti_mbox_command(qpti, param, 1);
}
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
out:
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
spin_unlock_irqrestore(host->host_lock, flags);
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
outfirm:
release_firmware(fw);
return err;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
static int qlogicpti_verify_tmon(struct qlogicpti *qpti)
{
int curstat = sbus_readb(qpti->sreg);
curstat &= 0xf0;
if (!(curstat & SREG_FUSE) && (qpti->swsreg & SREG_FUSE))
printk("qlogicpti%d: Fuse returned to normal state.\n", qpti->qpti_id);
if (!(curstat & SREG_TPOWER) && (qpti->swsreg & SREG_TPOWER))
printk("qlogicpti%d: termpwr back to normal state.\n", qpti->qpti_id);
if (curstat != qpti->swsreg) {
int error = 0;
if (curstat & SREG_FUSE) {
error++;
printk("qlogicpti%d: Fuse is open!\n", qpti->qpti_id);
}
if (curstat & SREG_TPOWER) {
error++;
printk("qlogicpti%d: termpwr failure\n", qpti->qpti_id);
}
if (qpti->differential &&
(curstat & SREG_DSENSE) != SREG_DSENSE) {
error++;
printk("qlogicpti%d: You have a single ended device on a "
"differential bus! Please fix!\n", qpti->qpti_id);
}
qpti->swsreg = curstat;
return error;
}
return 0;
}
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t qpti_intr(int irq, void *dev_id);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static void qpti_chain_add(struct qlogicpti *qpti)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
spin_lock_irq(&qptichain_lock);
if (qptichain != NULL) {
struct qlogicpti *qlink = qptichain;
while(qlink->next)
qlink = qlink->next;
qlink->next = qpti;
} else {
qptichain = qpti;
}
qpti->next = NULL;
spin_unlock_irq(&qptichain_lock);
}
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static void qpti_chain_del(struct qlogicpti *qpti)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
spin_lock_irq(&qptichain_lock);
if (qptichain == qpti) {
qptichain = qpti->next;
} else {
struct qlogicpti *qlink = qptichain;
while(qlink->next != qpti)
qlink = qlink->next;
qlink->next = qpti->next;
}
qpti->next = NULL;
spin_unlock_irq(&qptichain_lock);
}
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static int qpti_map_regs(struct qlogicpti *qpti)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
of/device: Replace struct of_device with struct platform_device of_device is just an alias for platform_device, so remove it entirely. Also replace to_of_device() with to_platform_device() and update comment blocks. This patch was initially generated from the following semantic patch, and then edited by hand to pick up the bits that coccinelle didn't catch. @@ @@ -struct of_device +struct platform_device Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Reviewed-by: David S. Miller <davem@davemloft.net>
2010-08-06 23:25:50 +08:00
struct platform_device *op = qpti->op;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
qpti->qregs = of_ioremap(&op->resource[0], 0,
resource_size(&op->resource[0]),
"PTI Qlogic/ISP");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!qpti->qregs) {
printk("PTI: Qlogic/ISP registers are unmappable\n");
return -1;
}
if (qpti->is_pti) {
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
qpti->sreg = of_ioremap(&op->resource[0], (16 * 4096),
sizeof(unsigned char),
"PTI Qlogic/ISP statreg");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (!qpti->sreg) {
printk("PTI: Qlogic/ISP status register is unmappable\n");
return -1;
}
}
return 0;
}
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static int qpti_register_irq(struct qlogicpti *qpti)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
of/device: Replace struct of_device with struct platform_device of_device is just an alias for platform_device, so remove it entirely. Also replace to_of_device() with to_platform_device() and update comment blocks. This patch was initially generated from the following semantic patch, and then edited by hand to pick up the bits that coccinelle didn't catch. @@ @@ -struct of_device +struct platform_device Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Reviewed-by: David S. Miller <davem@davemloft.net>
2010-08-06 23:25:50 +08:00
struct platform_device *op = qpti->op;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
sparc/of: Move of_device fields into struct pdev_archdata This patch moves SPARC architecture specific data members out of struct of_device and into the pdev_archdata structure. The reason for this change is to unify the struct of_device definition amongst all the architectures. It also remvoes the .sysdata, .slot, .portid and .clock_freq properties because they aren't actually used by anything. A subsequent patch will replace struct of_device entirely with struct platform_device and the of_platform support code will share common routines with the platform bus (but the bus instances themselves can remain separate). This patch also adds 'struct resources *resource' and num_resources to match the fields defined in struct platform_device. After this change, 'struct platform_device' can be used as a drop-in replacement for 'struct of_platform'. This change is in preparation for merging the of_platform_bus_type with the platform_bus_type. Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: David S. Miller <davem@davemloft.net> Cc: Stephen Rothwell <sfr@canb.auug.org.au>
2010-06-19 01:09:58 +08:00
qpti->qhost->irq = qpti->irq = op->archdata.irqs[0];
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
/* We used to try various overly-clever things to
* reduce the interrupt processing overhead on
* sun4c/sun4m when multiple PTI's shared the
* same IRQ. It was too complex and messy to
* sanely maintain.
*/
if (request_irq(qpti->irq, qpti_intr,
qlogicpti: Remove slash in QlogicPTI irq name qlogicpti driver registers its irq with a name containing slash. This results in [ 71.049735] WARNING: at fs/proc/generic.c:316 __xlate_proc_name+0xa8/0xb8() [ 71.132815] name 'Qlogic/PTI' because proc_mkdir with the name of the irq fails. Fix it by just removing the slash from irq name. Discovered and tested on real hardware (Sun Ultra 1). Signed-off-by: Meelis Roos <mroos@linux.ee> Signed-off-by: David S. Miller <davem@davemloft.net>
2010-03-09 02:53:08 +08:00
IRQF_SHARED, "QlogicPTI", qpti))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
goto fail;
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk("qlogicpti%d: IRQ %d ", qpti->qpti_id, qpti->irq);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return 0;
fail:
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk("qlogicpti%d: Cannot acquire irq line\n", qpti->qpti_id);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return -1;
}
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static void qpti_get_scsi_id(struct qlogicpti *qpti)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
of/device: Replace struct of_device with struct platform_device of_device is just an alias for platform_device, so remove it entirely. Also replace to_of_device() with to_platform_device() and update comment blocks. This patch was initially generated from the following semantic patch, and then edited by hand to pick up the bits that coccinelle didn't catch. @@ @@ -struct of_device +struct platform_device Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Reviewed-by: David S. Miller <davem@davemloft.net>
2010-08-06 23:25:50 +08:00
struct platform_device *op = qpti->op;
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
struct device_node *dp;
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-14 07:12:29 +08:00
dp = op->dev.of_node;
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
qpti->scsi_id = of_getintprop_default(dp, "initiator-id", -1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (qpti->scsi_id == -1)
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
qpti->scsi_id = of_getintprop_default(dp, "scsi-initiator-id",
-1);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (qpti->scsi_id == -1)
qpti->scsi_id =
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
of_getintprop_default(dp->parent,
"scsi-initiator-id", 7);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
qpti->qhost->this_id = qpti->scsi_id;
qpti->qhost->max_sectors = 64;
printk("SCSI ID %d ", qpti->scsi_id);
}
static void qpti_get_bursts(struct qlogicpti *qpti)
{
of/device: Replace struct of_device with struct platform_device of_device is just an alias for platform_device, so remove it entirely. Also replace to_of_device() with to_platform_device() and update comment blocks. This patch was initially generated from the following semantic patch, and then edited by hand to pick up the bits that coccinelle didn't catch. @@ @@ -struct of_device +struct platform_device Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Reviewed-by: David S. Miller <davem@davemloft.net>
2010-08-06 23:25:50 +08:00
struct platform_device *op = qpti->op;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
u8 bursts, bmask;
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-14 07:12:29 +08:00
bursts = of_getintprop_default(op->dev.of_node, "burst-sizes", 0xff);
bmask = of_getintprop_default(op->dev.of_node->parent, "burst-sizes", 0xff);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (bmask != 0xff)
bursts &= bmask;
if (bursts == 0xff ||
(bursts & DMA_BURST16) == 0 ||
(bursts & DMA_BURST32) == 0)
bursts = (DMA_BURST32 - 1);
qpti->bursts = bursts;
}
static void qpti_get_clock(struct qlogicpti *qpti)
{
unsigned int cfreq;
/* Check for what the clock input to this card is.
* Default to 40Mhz.
*/
cfreq = prom_getintdefault(qpti->prom_node,"clock-frequency",40000000);
qpti->clock = (cfreq + 500000)/1000000;
if (qpti->clock == 0) /* bullshit */
qpti->clock = 40;
}
/* The request and response queues must each be aligned
* on a page boundary.
*/
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static int qpti_map_queues(struct qlogicpti *qpti)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
of/device: Replace struct of_device with struct platform_device of_device is just an alias for platform_device, so remove it entirely. Also replace to_of_device() with to_platform_device() and update comment blocks. This patch was initially generated from the following semantic patch, and then edited by hand to pick up the bits that coccinelle didn't catch. @@ @@ -struct of_device +struct platform_device Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Reviewed-by: David S. Miller <davem@davemloft.net>
2010-08-06 23:25:50 +08:00
struct platform_device *op = qpti->op;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
#define QSIZE(entries) (((entries) + 1) * QUEUE_ENTRY_LEN)
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
qpti->res_cpu = dma_alloc_coherent(&op->dev,
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
QSIZE(RES_QUEUE_LEN),
&qpti->res_dvma, GFP_ATOMIC);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (qpti->res_cpu == NULL ||
qpti->res_dvma == 0) {
printk("QPTI: Cannot map response queue.\n");
return -1;
}
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
qpti->req_cpu = dma_alloc_coherent(&op->dev,
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
QSIZE(QLOGICPTI_REQ_QUEUE_LEN),
&qpti->req_dvma, GFP_ATOMIC);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (qpti->req_cpu == NULL ||
qpti->req_dvma == 0) {
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
dma_free_coherent(&op->dev, QSIZE(RES_QUEUE_LEN),
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
qpti->res_cpu, qpti->res_dvma);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
printk("QPTI: Cannot map request queue.\n");
return -1;
}
memset(qpti->res_cpu, 0, QSIZE(RES_QUEUE_LEN));
memset(qpti->req_cpu, 0, QSIZE(QLOGICPTI_REQ_QUEUE_LEN));
return 0;
}
const char *qlogicpti_info(struct Scsi_Host *host)
{
static char buf[80];
struct qlogicpti *qpti = (struct qlogicpti *) host->hostdata;
[SPARC]: Kill __irq_itoa(). This ugly hack was long overdue to die. It was a way to print out Sparc interrupts in a more freindly format, since IRQ numbers were arbitrary opaque 32-bit integers which vectored into PIL levels. These 32-bit integers were not necessarily in the 0-->NR_IRQS range, but the PILs they vectored to were. The idea now is that we will increase NR_IRQS a little bit and use a virtual<-->real IRQ number mapping scheme similar to PowerPC. That makes this IRQ printing hack irrelevant, and furthermore only a handful of drivers actually used __irq_itoa() making it even less useful. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-20 16:21:29 +08:00
sprintf(buf, "PTI Qlogic,ISP SBUS SCSI irq %d regs at %p",
qpti->qhost->irq, qpti->qregs);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return buf;
}
/* I am a certified frobtronicist. */
static inline void marker_frob(struct Command_Entry *cmd)
{
struct Marker_Entry *marker = (struct Marker_Entry *) cmd;
memset(marker, 0, sizeof(struct Marker_Entry));
marker->hdr.entry_cnt = 1;
marker->hdr.entry_type = ENTRY_MARKER;
marker->modifier = SYNC_ALL;
marker->rsvd = 0;
}
static inline void cmd_frob(struct Command_Entry *cmd, struct scsi_cmnd *Cmnd,
struct qlogicpti *qpti)
{
memset(cmd, 0, sizeof(struct Command_Entry));
cmd->hdr.entry_cnt = 1;
cmd->hdr.entry_type = ENTRY_COMMAND;
cmd->target_id = Cmnd->device->id;
cmd->target_lun = Cmnd->device->lun;
cmd->cdb_length = Cmnd->cmd_len;
cmd->control_flags = 0;
if (Cmnd->device->tagged_supported) {
if (qpti->cmd_count[Cmnd->device->id] == 0)
qpti->tag_ages[Cmnd->device->id] = jiffies;
[PATCH] drivers/scsi/*: use time_after() and friends They deal with wrapping correctly and are nicer to read. Signed-off-by: Marcelo Feitoza Parisi <marcelo@feitoza.com.br> Signed-off-by: Alexey Dobriyan <adobriyan@gmail.com> Cc: James Bottomley <James.Bottomley@steeleye.com> Signed-off-by: Andrew Morton <akpm@osdl.org> Signed-off-by: Linus Torvalds <torvalds@osdl.org>
2006-03-28 17:56:47 +08:00
if (time_after(jiffies, qpti->tag_ages[Cmnd->device->id] + (5*HZ))) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
cmd->control_flags = CFLAG_ORDERED_TAG;
qpti->tag_ages[Cmnd->device->id] = jiffies;
} else
cmd->control_flags = CFLAG_SIMPLE_TAG;
}
if ((Cmnd->cmnd[0] == WRITE_6) ||
(Cmnd->cmnd[0] == WRITE_10) ||
(Cmnd->cmnd[0] == WRITE_12))
cmd->control_flags |= CFLAG_WRITE;
else
cmd->control_flags |= CFLAG_READ;
[SCSI] qlogicpti: fix timeout qlogicpti times out for some tape library operations (like mtx inventory). It seems SCSI command timeout is hardcoded into the driver. Fix it by propagating the timeout from scsi request to the controller as suggested by James Bottomley. Tested on Sun Ultra 1 with Sun StorEdge L8 Autoloader. Signed-off-by: Meelis Roos <mroos@linux.ee> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <JBottomley@Parallels.com>
2011-10-07 23:05:08 +08:00
cmd->time_out = Cmnd->request->timeout/HZ;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
memcpy(cmd->cdb, Cmnd->cmnd, Cmnd->cmd_len);
}
/* Do it to it baby. */
static inline int load_cmd(struct scsi_cmnd *Cmnd, struct Command_Entry *cmd,
struct qlogicpti *qpti, u_int in_ptr, u_int out_ptr)
{
struct dataseg *ds;
qlogicpti: sg chaining support Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2007-05-09 19:43:12 +08:00
struct scatterlist *sg, *s;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
int i, n;
[SCSI] qlogicpti: convert to accessors and !use_sg cleanup - convert to accessors and !use_sg cleanup Signed-off-by: Boaz Harrosh <bharrosh@panasas.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2007-09-10 02:21:35 +08:00
if (scsi_bufflen(Cmnd)) {
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
int sg_count;
[SCSI] qlogicpti: convert to accessors and !use_sg cleanup - convert to accessors and !use_sg cleanup Signed-off-by: Boaz Harrosh <bharrosh@panasas.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2007-09-10 02:21:35 +08:00
sg = scsi_sglist(Cmnd);
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
sg_count = dma_map_sg(&qpti->op->dev, sg,
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
scsi_sg_count(Cmnd),
Cmnd->sc_data_direction);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
ds = cmd->dataseg;
cmd->segment_cnt = sg_count;
/* Fill in first four sg entries: */
n = sg_count;
if (n > 4)
n = 4;
qlogicpti: sg chaining support Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2007-05-09 19:43:12 +08:00
for_each_sg(sg, s, n, i) {
ds[i].d_base = sg_dma_address(s);
ds[i].d_count = sg_dma_len(s);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
sg_count -= 4;
qlogicpti: sg chaining support Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2007-05-09 19:43:12 +08:00
sg = s;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
while (sg_count > 0) {
struct Continuation_Entry *cont;
++cmd->hdr.entry_cnt;
cont = (struct Continuation_Entry *) &qpti->req_cpu[in_ptr];
in_ptr = NEXT_REQ_PTR(in_ptr);
if (in_ptr == out_ptr)
return -1;
cont->hdr.entry_type = ENTRY_CONTINUATION;
cont->hdr.entry_cnt = 0;
cont->hdr.sys_def_1 = 0;
cont->hdr.flags = 0;
cont->reserved = 0;
ds = cont->dataseg;
n = sg_count;
if (n > 7)
n = 7;
qlogicpti: sg chaining support Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2007-05-09 19:43:12 +08:00
for_each_sg(sg, s, n, i) {
ds[i].d_base = sg_dma_address(s);
ds[i].d_count = sg_dma_len(s);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
sg_count -= n;
scsi: fix fall out of sg-chaining patch in qlogicpti Boaz writes: "I've reviewed all patches since Matthew's, and I find one small problem. In the load_cmd() there is a compound loop where the first 4 sg's are set then the rest are set into a memory structure in group of 7 sg's. Well the second 7-group and on is a bug because sg pointer does not advance. This is a fall out from Jens's patch." The reporter, Meelis Roos <mroos@ut.ee>, verified that this patch does indeed fix his problem with qlogicpti. Signed-off-by: Jens Axboe <jens.axboe@oracle.com>
2008-09-29 15:38:55 +08:00
sg = s;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
} else {
cmd->dataseg[0].d_base = 0;
cmd->dataseg[0].d_count = 0;
cmd->segment_cnt = 1; /* Shouldn't this be 0? */
}
/* Committed, record Scsi_Cmd so we can find it later. */
cmd->handle = in_ptr;
qpti->cmd_slots[in_ptr] = Cmnd;
qpti->cmd_count[Cmnd->device->id]++;
sbus_writew(in_ptr, qpti->qregs + MBOX4);
qpti->req_in_ptr = in_ptr;
return in_ptr;
}
static inline void update_can_queue(struct Scsi_Host *host, u_int in_ptr, u_int out_ptr)
{
/* Temporary workaround until bug is found and fixed (one bug has been found
already, but fixing it makes things even worse) -jj */
int num_free = QLOGICPTI_REQ_QUEUE_LEN - REQ_QUEUE_DEPTH(in_ptr, out_ptr) - 64;
host->can_queue = host->host_busy + num_free;
host->sg_tablesize = QLOGICPTI_MAX_SG(num_free);
}
[SCSI] qlogicpti: Add a slave_configure method By configuring targets in slave_configure, we can eliminate a shadow queuecommand, a shadow scsi_done, a write to the host template, abuse of SCp->Message and SCp->Status, a use of kmap_atomic() and sniffing the results of INQUIRY. Signed-off-by: Matthew Wilcox <willy@linux.intel.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-10-01 22:07:52 +08:00
static int qlogicpti_slave_configure(struct scsi_device *sdev)
[QLOGICPTI]: Handle INQUIRY response sniffing correctly. These days, in 2.6.x, even INQUIRY commands are sent using scatter gather lists. Bug reported by Tom 'spot' Callaway. Signed-off-by: David S. Miller <davem@davemloft.net>
2005-10-15 04:44:32 +08:00
{
[SCSI] qlogicpti: Add a slave_configure method By configuring targets in slave_configure, we can eliminate a shadow queuecommand, a shadow scsi_done, a write to the host template, abuse of SCp->Message and SCp->Status, a use of kmap_atomic() and sniffing the results of INQUIRY. Signed-off-by: Matthew Wilcox <willy@linux.intel.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-10-01 22:07:52 +08:00
struct qlogicpti *qpti = shost_priv(sdev->host);
int tgt = sdev->id;
u_short param[6];
[QLOGICPTI]: Handle INQUIRY response sniffing correctly. These days, in 2.6.x, even INQUIRY commands are sent using scatter gather lists. Bug reported by Tom 'spot' Callaway. Signed-off-by: David S. Miller <davem@davemloft.net>
2005-10-15 04:44:32 +08:00
[SCSI] qlogicpti: Add a slave_configure method By configuring targets in slave_configure, we can eliminate a shadow queuecommand, a shadow scsi_done, a write to the host template, abuse of SCp->Message and SCp->Status, a use of kmap_atomic() and sniffing the results of INQUIRY. Signed-off-by: Matthew Wilcox <willy@linux.intel.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-10-01 22:07:52 +08:00
/* tags handled in midlayer */
/* enable sync mode? */
if (sdev->sdtr) {
qpti->dev_param[tgt].device_flags |= 0x10;
[QLOGICPTI]: Handle INQUIRY response sniffing correctly. These days, in 2.6.x, even INQUIRY commands are sent using scatter gather lists. Bug reported by Tom 'spot' Callaway. Signed-off-by: David S. Miller <davem@davemloft.net>
2005-10-15 04:44:32 +08:00
} else {
[SCSI] qlogicpti: Add a slave_configure method By configuring targets in slave_configure, we can eliminate a shadow queuecommand, a shadow scsi_done, a write to the host template, abuse of SCp->Message and SCp->Status, a use of kmap_atomic() and sniffing the results of INQUIRY. Signed-off-by: Matthew Wilcox <willy@linux.intel.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-10-01 22:07:52 +08:00
qpti->dev_param[tgt].synchronous_offset = 0;
qpti->dev_param[tgt].synchronous_period = 0;
[QLOGICPTI]: Handle INQUIRY response sniffing correctly. These days, in 2.6.x, even INQUIRY commands are sent using scatter gather lists. Bug reported by Tom 'spot' Callaway. Signed-off-by: David S. Miller <davem@davemloft.net>
2005-10-15 04:44:32 +08:00
}
[SCSI] qlogicpti: Add a slave_configure method By configuring targets in slave_configure, we can eliminate a shadow queuecommand, a shadow scsi_done, a write to the host template, abuse of SCp->Message and SCp->Status, a use of kmap_atomic() and sniffing the results of INQUIRY. Signed-off-by: Matthew Wilcox <willy@linux.intel.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-10-01 22:07:52 +08:00
/* are we wide capable? */
if (sdev->wdtr)
qpti->dev_param[tgt].device_flags |= 0x20;
param[0] = MBOX_SET_TARGET_PARAMS;
param[1] = (tgt << 8);
param[2] = (qpti->dev_param[tgt].device_flags << 8);
if (qpti->dev_param[tgt].device_flags & 0x10) {
param[3] = (qpti->dev_param[tgt].synchronous_offset << 8) |
qpti->dev_param[tgt].synchronous_period;
} else {
param[3] = 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
[SCSI] qlogicpti: Add a slave_configure method By configuring targets in slave_configure, we can eliminate a shadow queuecommand, a shadow scsi_done, a write to the host template, abuse of SCp->Message and SCp->Status, a use of kmap_atomic() and sniffing the results of INQUIRY. Signed-off-by: Matthew Wilcox <willy@linux.intel.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-10-01 22:07:52 +08:00
qlogicpti_mbox_command(qpti, param, 0);
return 0;
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
}
/*
* The middle SCSI layer ensures that queuecommand never gets invoked
* concurrently with itself or the interrupt handler (though the
* interrupt handler may call this routine as part of
* request-completion handling).
*
* "This code must fly." -davem
*/
SCSI host lock push-down Move the mid-layer's ->queuecommand() invocation from being locked with the host lock to being unlocked to facilitate speeding up the critical path for drivers who don't need this lock taken anyway. The patch below presents a simple SCSI host lock push-down as an equivalent transformation. No locking or other behavior should change with this patch. All existing bugs and locking orders are preserved. Additionally, add one parameter to queuecommand, struct Scsi_Host * and remove one parameter from queuecommand, void (*done)(struct scsi_cmnd *) Scsi_Host* is a convenient pointer that most host drivers need anyway, and 'done' is redundant to struct scsi_cmnd->scsi_done. Minimal code disturbance was attempted with this change. Most drivers needed only two one-line modifications for their host lock push-down. Signed-off-by: Jeff Garzik <jgarzik@redhat.com> Acked-by: James Bottomley <James.Bottomley@suse.de> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2010-11-16 15:10:29 +08:00
static int qlogicpti_queuecommand_lck(struct scsi_cmnd *Cmnd, void (*done)(struct scsi_cmnd *))
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct Scsi_Host *host = Cmnd->device->host;
struct qlogicpti *qpti = (struct qlogicpti *) host->hostdata;
struct Command_Entry *cmd;
u_int out_ptr;
int in_ptr;
Cmnd->scsi_done = done;
in_ptr = qpti->req_in_ptr;
cmd = (struct Command_Entry *) &qpti->req_cpu[in_ptr];
out_ptr = sbus_readw(qpti->qregs + MBOX4);
in_ptr = NEXT_REQ_PTR(in_ptr);
if (in_ptr == out_ptr)
goto toss_command;
if (qpti->send_marker) {
marker_frob(cmd);
qpti->send_marker = 0;
if (NEXT_REQ_PTR(in_ptr) == out_ptr) {
sbus_writew(in_ptr, qpti->qregs + MBOX4);
qpti->req_in_ptr = in_ptr;
goto toss_command;
}
cmd = (struct Command_Entry *) &qpti->req_cpu[in_ptr];
in_ptr = NEXT_REQ_PTR(in_ptr);
}
cmd_frob(cmd, Cmnd, qpti);
if ((in_ptr = load_cmd(Cmnd, cmd, qpti, in_ptr, out_ptr)) == -1)
goto toss_command;
update_can_queue(host, in_ptr, out_ptr);
return 0;
toss_command:
printk(KERN_EMERG "qlogicpti%d: request queue overflow\n",
qpti->qpti_id);
/* Unfortunately, unless you use the new EH code, which
* we don't, the midlayer will ignore the return value,
* which is insane. We pick up the pieces like this.
*/
Cmnd->result = DID_BUS_BUSY;
done(Cmnd);
return 1;
}
SCSI host lock push-down Move the mid-layer's ->queuecommand() invocation from being locked with the host lock to being unlocked to facilitate speeding up the critical path for drivers who don't need this lock taken anyway. The patch below presents a simple SCSI host lock push-down as an equivalent transformation. No locking or other behavior should change with this patch. All existing bugs and locking orders are preserved. Additionally, add one parameter to queuecommand, struct Scsi_Host * and remove one parameter from queuecommand, void (*done)(struct scsi_cmnd *) Scsi_Host* is a convenient pointer that most host drivers need anyway, and 'done' is redundant to struct scsi_cmnd->scsi_done. Minimal code disturbance was attempted with this change. Most drivers needed only two one-line modifications for their host lock push-down. Signed-off-by: Jeff Garzik <jgarzik@redhat.com> Acked-by: James Bottomley <James.Bottomley@suse.de> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2010-11-16 15:10:29 +08:00
static DEF_SCSI_QCMD(qlogicpti_queuecommand)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
static int qlogicpti_return_status(struct Status_Entry *sts, int id)
{
int host_status = DID_ERROR;
switch (sts->completion_status) {
case CS_COMPLETE:
host_status = DID_OK;
break;
case CS_INCOMPLETE:
if (!(sts->state_flags & SF_GOT_BUS))
host_status = DID_NO_CONNECT;
else if (!(sts->state_flags & SF_GOT_TARGET))
host_status = DID_BAD_TARGET;
else if (!(sts->state_flags & SF_SENT_CDB))
host_status = DID_ERROR;
else if (!(sts->state_flags & SF_TRANSFERRED_DATA))
host_status = DID_ERROR;
else if (!(sts->state_flags & SF_GOT_STATUS))
host_status = DID_ERROR;
else if (!(sts->state_flags & SF_GOT_SENSE))
host_status = DID_ERROR;
break;
case CS_DMA_ERROR:
case CS_TRANSPORT_ERROR:
host_status = DID_ERROR;
break;
case CS_RESET_OCCURRED:
case CS_BUS_RESET:
host_status = DID_RESET;
break;
case CS_ABORTED:
host_status = DID_ABORT;
break;
case CS_TIMEOUT:
host_status = DID_TIME_OUT;
break;
case CS_DATA_OVERRUN:
case CS_COMMAND_OVERRUN:
case CS_STATUS_OVERRUN:
case CS_BAD_MESSAGE:
case CS_NO_MESSAGE_OUT:
case CS_EXT_ID_FAILED:
case CS_IDE_MSG_FAILED:
case CS_ABORT_MSG_FAILED:
case CS_NOP_MSG_FAILED:
case CS_PARITY_ERROR_MSG_FAILED:
case CS_DEVICE_RESET_MSG_FAILED:
case CS_ID_MSG_FAILED:
case CS_UNEXP_BUS_FREE:
host_status = DID_ERROR;
break;
case CS_DATA_UNDERRUN:
host_status = DID_OK;
break;
default:
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_EMERG "qlogicpti%d: unknown completion status 0x%04x\n",
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
id, sts->completion_status);
host_status = DID_ERROR;
break;
}
return (sts->scsi_status & STATUS_MASK) | (host_status << 16);
}
static struct scsi_cmnd *qlogicpti_intr_handler(struct qlogicpti *qpti)
{
struct scsi_cmnd *Cmnd, *done_queue = NULL;
struct Status_Entry *sts;
u_int in_ptr, out_ptr;
if (!(sbus_readw(qpti->qregs + SBUS_STAT) & SBUS_STAT_RINT))
return NULL;
in_ptr = sbus_readw(qpti->qregs + MBOX5);
sbus_writew(HCCTRL_CRIRQ, qpti->qregs + HCCTRL);
if (sbus_readw(qpti->qregs + SBUS_SEMAPHORE) & SBUS_SEMAPHORE_LCK) {
switch (sbus_readw(qpti->qregs + MBOX0)) {
case ASYNC_SCSI_BUS_RESET:
case EXECUTION_TIMEOUT_RESET:
qpti->send_marker = 1;
break;
case INVALID_COMMAND:
case HOST_INTERFACE_ERROR:
case COMMAND_ERROR:
case COMMAND_PARAM_ERROR:
break;
};
sbus_writew(0, qpti->qregs + SBUS_SEMAPHORE);
}
/* This looks like a network driver! */
out_ptr = qpti->res_out_ptr;
while (out_ptr != in_ptr) {
u_int cmd_slot;
sts = (struct Status_Entry *) &qpti->res_cpu[out_ptr];
out_ptr = NEXT_RES_PTR(out_ptr);
/* We store an index in the handle, not the pointer in
* some form. This avoids problems due to the fact
* that the handle provided is only 32-bits. -DaveM
*/
cmd_slot = sts->handle;
Cmnd = qpti->cmd_slots[cmd_slot];
qpti->cmd_slots[cmd_slot] = NULL;
if (sts->completion_status == CS_RESET_OCCURRED ||
sts->completion_status == CS_ABORTED ||
(sts->status_flags & STF_BUS_RESET))
qpti->send_marker = 1;
if (sts->state_flags & SF_GOT_SENSE)
memcpy(Cmnd->sense_buffer, sts->req_sense_data,
[SCSI] replace sizeof sense_buffer with SCSI_SENSE_BUFFERSIZE This replaces sizeof sense_buffer with SCSI_SENSE_BUFFERSIZE in several LLDs. It's a preparation for the future changes to remove sense_buffer array in scsi_cmnd structure. Signed-off-by: FUJITA Tomonori <fujita.tomonori@lab.ntt.co.jp> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2008-01-13 14:46:13 +08:00
SCSI_SENSE_BUFFERSIZE);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
if (sts->hdr.entry_type == ENTRY_STATUS)
Cmnd->result =
qlogicpti_return_status(sts, qpti->qpti_id);
else
Cmnd->result = DID_ERROR << 16;
[SCSI] qlogicpti: convert to accessors and !use_sg cleanup - convert to accessors and !use_sg cleanup Signed-off-by: Boaz Harrosh <bharrosh@panasas.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2007-09-10 02:21:35 +08:00
if (scsi_bufflen(Cmnd))
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
dma_unmap_sg(&qpti->op->dev,
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
scsi_sglist(Cmnd), scsi_sg_count(Cmnd),
Cmnd->sc_data_direction);
[SCSI] qlogicpti: convert to accessors and !use_sg cleanup - convert to accessors and !use_sg cleanup Signed-off-by: Boaz Harrosh <bharrosh@panasas.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2007-09-10 02:21:35 +08:00
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
qpti->cmd_count[Cmnd->device->id]--;
sbus_writew(out_ptr, qpti->qregs + MBOX5);
Cmnd->host_scribble = (unsigned char *) done_queue;
done_queue = Cmnd;
}
qpti->res_out_ptr = out_ptr;
return done_queue;
}
IRQ: Maintain regs pointer globally rather than passing to IRQ handlers Maintain a per-CPU global "struct pt_regs *" variable which can be used instead of passing regs around manually through all ~1800 interrupt handlers in the Linux kernel. The regs pointer is used in few places, but it potentially costs both stack space and code to pass it around. On the FRV arch, removing the regs parameter from all the genirq function results in a 20% speed up of the IRQ exit path (ie: from leaving timer_interrupt() to leaving do_IRQ()). Where appropriate, an arch may override the generic storage facility and do something different with the variable. On FRV, for instance, the address is maintained in GR28 at all times inside the kernel as part of general exception handling. Having looked over the code, it appears that the parameter may be handed down through up to twenty or so layers of functions. Consider a USB character device attached to a USB hub, attached to a USB controller that posts its interrupts through a cascaded auxiliary interrupt controller. A character device driver may want to pass regs to the sysrq handler through the input layer which adds another few layers of parameter passing. I've build this code with allyesconfig for x86_64 and i386. I've runtested the main part of the code on FRV and i386, though I can't test most of the drivers. I've also done partial conversion for powerpc and MIPS - these at least compile with minimal configurations. This will affect all archs. Mostly the changes should be relatively easy. Take do_IRQ(), store the regs pointer at the beginning, saving the old one: struct pt_regs *old_regs = set_irq_regs(regs); And put the old one back at the end: set_irq_regs(old_regs); Don't pass regs through to generic_handle_irq() or __do_IRQ(). In timer_interrupt(), this sort of change will be necessary: - update_process_times(user_mode(regs)); - profile_tick(CPU_PROFILING, regs); + update_process_times(user_mode(get_irq_regs())); + profile_tick(CPU_PROFILING); I'd like to move update_process_times()'s use of get_irq_regs() into itself, except that i386, alone of the archs, uses something other than user_mode(). Some notes on the interrupt handling in the drivers: (*) input_dev() is now gone entirely. The regs pointer is no longer stored in the input_dev struct. (*) finish_unlinks() in drivers/usb/host/ohci-q.c needs checking. It does something different depending on whether it's been supplied with a regs pointer or not. (*) Various IRQ handler function pointers have been moved to type irq_handler_t. Signed-Off-By: David Howells <dhowells@redhat.com> (cherry picked from 1b16e7ac850969f38b375e511e3fa2f474a33867 commit)
2006-10-05 21:55:46 +08:00
static irqreturn_t qpti_intr(int irq, void *dev_id)
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
{
struct qlogicpti *qpti = dev_id;
unsigned long flags;
struct scsi_cmnd *dq;
spin_lock_irqsave(qpti->qhost->host_lock, flags);
dq = qlogicpti_intr_handler(qpti);
if (dq != NULL) {
do {
struct scsi_cmnd *next;
next = (struct scsi_cmnd *) dq->host_scribble;
dq->scsi_done(dq);
dq = next;
} while (dq != NULL);
}
spin_unlock_irqrestore(qpti->qhost->host_lock, flags);
return IRQ_HANDLED;
}
static int qlogicpti_abort(struct scsi_cmnd *Cmnd)
{
u_short param[6];
struct Scsi_Host *host = Cmnd->device->host;
struct qlogicpti *qpti = (struct qlogicpti *) host->hostdata;
int return_status = SUCCESS;
u32 cmd_cookie;
int i;
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_WARNING "qlogicpti%d: Aborting cmd for tgt[%d] lun[%d]\n",
qpti->qpti_id, (int)Cmnd->device->id, (int)Cmnd->device->lun);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
qlogicpti_disable_irqs(qpti);
/* Find the 32-bit cookie we gave to the firmware for
* this command.
*/
for (i = 0; i < QLOGICPTI_REQ_QUEUE_LEN + 1; i++)
if (qpti->cmd_slots[i] == Cmnd)
break;
cmd_cookie = i;
param[0] = MBOX_ABORT;
param[1] = (((u_short) Cmnd->device->id) << 8) | Cmnd->device->lun;
param[2] = cmd_cookie >> 16;
param[3] = cmd_cookie & 0xffff;
if (qlogicpti_mbox_command(qpti, param, 0) ||
(param[0] != MBOX_COMMAND_COMPLETE)) {
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_EMERG "qlogicpti%d: scsi abort failure: %x\n",
qpti->qpti_id, param[0]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return_status = FAILED;
}
qlogicpti_enable_irqs(qpti);
return return_status;
}
static int qlogicpti_reset(struct scsi_cmnd *Cmnd)
{
u_short param[6];
struct Scsi_Host *host = Cmnd->device->host;
struct qlogicpti *qpti = (struct qlogicpti *) host->hostdata;
int return_status = SUCCESS;
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_WARNING "qlogicpti%d: Resetting SCSI bus!\n",
qpti->qpti_id);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
qlogicpti_disable_irqs(qpti);
param[0] = MBOX_BUS_RESET;
param[1] = qpti->host_param.bus_reset_delay;
if (qlogicpti_mbox_command(qpti, param, 0) ||
(param[0] != MBOX_COMMAND_COMPLETE)) {
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk(KERN_EMERG "qlogicisp%d: scsi bus reset failure: %x\n",
qpti->qpti_id, param[0]);
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
return_status = FAILED;
}
qlogicpti_enable_irqs(qpti);
return return_status;
}
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
static struct scsi_host_template qpti_template = {
.module = THIS_MODULE,
.name = "qlogicpti",
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
.info = qlogicpti_info,
[SCSI] qlogicpti: Add a slave_configure method By configuring targets in slave_configure, we can eliminate a shadow queuecommand, a shadow scsi_done, a write to the host template, abuse of SCp->Message and SCp->Status, a use of kmap_atomic() and sniffing the results of INQUIRY. Signed-off-by: Matthew Wilcox <willy@linux.intel.com> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-10-01 22:07:52 +08:00
.queuecommand = qlogicpti_queuecommand,
.slave_configure = qlogicpti_slave_configure,
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
.eh_abort_handler = qlogicpti_abort,
.eh_bus_reset_handler = qlogicpti_reset,
.can_queue = QLOGICPTI_REQ_QUEUE_LEN,
.this_id = 7,
.sg_tablesize = QLOGICPTI_MAX_SG(QLOGICPTI_REQ_QUEUE_LEN),
.cmd_per_lun = 1,
.use_clustering = ENABLE_CLUSTERING,
};
drivercore: revert addition of of_match to struct device Commit b826291c, "drivercore/dt: add a match table pointer to struct device" added an of_match pointer to struct device to cache the of_match_table entry discovered at driver match time. This was unsafe because matching is not an atomic operation with probing a driver. If two or more drivers are attempted to be matched to a driver at the same time, then the cached matching entry pointer could get overwritten. This patch reverts the of_match cache pointer and reworks all users to call of_match_device() directly instead. Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-05-19 01:19:24 +08:00
static const struct of_device_id qpti_match[];
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static int qpti_sbus_probe(struct platform_device *op)
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
{
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-14 07:12:29 +08:00
struct device_node *dp = op->dev.of_node;
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
struct Scsi_Host *host;
struct qlogicpti *qpti;
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
static int nqptis;
[SPARC/64] constify of_get_property return: drivers The only unfortunate bit here is that the name field of struct map_info is not const, so for now we put a cast on the assignment of it. Signed-off-by: Stephen Rothwell <sfr@canb.auug.org.au> Signed-off-by: David S. Miller <davem@davemloft.net>
2007-03-29 15:49:54 +08:00
const char *fcode;
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
/* Sometimes Antares cards come up not completely
* setup, and we get a report of a zero IRQ.
*/
sparc/of: Move of_device fields into struct pdev_archdata This patch moves SPARC architecture specific data members out of struct of_device and into the pdev_archdata structure. The reason for this change is to unify the struct of_device definition amongst all the architectures. It also remvoes the .sysdata, .slot, .portid and .clock_freq properties because they aren't actually used by anything. A subsequent patch will replace struct of_device entirely with struct platform_device and the of_platform support code will share common routines with the platform bus (but the bus instances themselves can remain separate). This patch also adds 'struct resources *resource' and num_resources to match the fields defined in struct platform_device. After this change, 'struct platform_device' can be used as a drop-in replacement for 'struct of_platform'. This change is in preparation for merging the of_platform_bus_type with the platform_bus_type. Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: David S. Miller <davem@davemloft.net> Cc: Stephen Rothwell <sfr@canb.auug.org.au>
2010-06-19 01:09:58 +08:00
if (op->archdata.irqs[0] == 0)
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
return -ENODEV;
qlogicpti: Fix build warning. The build warns: drivers/scsi/qlogicpti.c: In function 'qpti_sbus_probe': drivers/scsi/qlogicpti.c:1316:45: warning: passing argument 1 of 'scsi_host_alloc' discards 'const' qualifier from pointer target type [enabled by default] include/scsi/scsi_host.h:778:26: note: expected 'struct scsi_host_template *' but argument is of type 'const struct scsi_host_template *' The problem is that of_device_id->data is a const void pointer. This is pretty silly in this specific instance, because for all matched device IDs we set match->data to the same value, &qpti_template. So just use that directly instead of the unnecessary and improperly typed abstraction. Signed-off-by: David S. Miller <davem@davemloft.net>
2012-10-29 04:26:37 +08:00
host = scsi_host_alloc(&qpti_template, sizeof(struct qlogicpti));
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
if (!host)
return -ENOMEM;
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
qpti = shost_priv(host);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
host->max_id = MAX_TARGETS;
qpti->qhost = host;
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
qpti->op = op;
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
qpti->qpti_id = nqptis;
of: Always use 'struct device.of_node' to get device node pointer. The following structure elements duplicate the information in 'struct device.of_node' and so are being eliminated. This patch makes all readers of these elements use device.of_node instead. (struct of_device *)->node (struct dev_archdata *)->prom_node (sparc) (struct dev_archdata *)->of_node (powerpc & microblaze) Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2010-04-14 07:12:29 +08:00
strcpy(qpti->prom_name, op->dev.of_node->name);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
qpti->is_pti = strcmp(qpti->prom_name, "QLGC,isp");
if (qpti_map_regs(qpti) < 0)
goto fail_unlink;
if (qpti_register_irq(qpti) < 0)
goto fail_unmap_regs;
qpti_get_scsi_id(qpti);
qpti_get_bursts(qpti);
qpti_get_clock(qpti);
/* Clear out scsi_cmnd array. */
memset(qpti->cmd_slots, 0, sizeof(qpti->cmd_slots));
if (qpti_map_queues(qpti) < 0)
goto fail_free_irq;
/* Load the firmware. */
if (qlogicpti_load_firmware(qpti))
goto fail_unmap_queues;
if (qpti->is_pti) {
/* Check the PTI status reg. */
if (qlogicpti_verify_tmon(qpti))
goto fail_unmap_queues;
}
/* Reset the ISP and init res/req queues. */
if (qlogicpti_reset_hardware(host))
goto fail_unmap_queues;
printk("(Firmware v%d.%d.%d)", qpti->fware_majrev,
qpti->fware_minrev, qpti->fware_micrev);
fcode = of_get_property(dp, "isp-fcode", NULL);
if (fcode && fcode[0])
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk("(FCode %s)", fcode);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
if (of_find_property(dp, "differential", NULL) != NULL)
qpti->differential = 1;
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk("\nqlogicpti%d: [%s Wide, using %s interface]\n",
qpti->qpti_id,
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
(qpti->ultra ? "Ultra" : "Fast"),
(qpti->differential ? "differential" : "single ended"));
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
if (scsi_add_host(host, &op->dev)) {
[SCSI] qlogicpti: Some cosmetic changes Change a printk sequencing issue where <6> ... was coming up in the middle of a line when scsi_add_host was being called. Reduce the length of some printk messages and make the messages more consistant. All cosmetic but it makes it easier to read as it scrolles off the screen during boot. Signed-off-by: Mark Fortescue <mark@mtfhpc.demon.co.uk> Acked-by: David S. Miller <davem@davemloft.net> Signed-off-by: James Bottomley <James.Bottomley@SteelEye.com>
2007-08-01 03:10:45 +08:00
printk("qlogicpti%d: Failed scsi_add_host\n", qpti->qpti_id);
goto fail_unmap_queues;
}
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
dev_set_drvdata(&op->dev, qpti);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
qpti_chain_add(qpti);
scsi_scan_host(host);
nqptis++;
return 0;
fail_unmap_queues:
#define QSIZE(entries) (((entries) + 1) * QUEUE_ENTRY_LEN)
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
dma_free_coherent(&op->dev,
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
QSIZE(RES_QUEUE_LEN),
qpti->res_cpu, qpti->res_dvma);
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
dma_free_coherent(&op->dev,
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
QSIZE(QLOGICPTI_REQ_QUEUE_LEN),
qpti->req_cpu, qpti->req_dvma);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
#undef QSIZE
fail_unmap_regs:
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
of_iounmap(&op->resource[0], qpti->qregs,
resource_size(&op->resource[0]));
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
if (qpti->is_pti)
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
of_iounmap(&op->resource[0], qpti->sreg,
sizeof(unsigned char));
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
fail_free_irq:
free_irq(qpti->irq, qpti);
fail_unlink:
scsi_host_put(host);
return -ENODEV;
}
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
static int qpti_sbus_remove(struct platform_device *op)
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
{
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
struct qlogicpti *qpti = dev_get_drvdata(&op->dev);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
qpti_chain_del(qpti);
scsi_remove_host(qpti->qhost);
/* Shut up the card. */
sbus_writew(0, qpti->qregs + SBUS_CTRL);
/* Free IRQ handler and unmap Qlogic,ISP and PTI status regs. */
free_irq(qpti->irq, qpti);
#define QSIZE(entries) (((entries) + 1) * QUEUE_ENTRY_LEN)
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
dma_free_coherent(&op->dev,
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
QSIZE(RES_QUEUE_LEN),
qpti->res_cpu, qpti->res_dvma);
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
dma_free_coherent(&op->dev,
sparc: Convert all SBUS drivers to dma_*() interfaces. And all the SBUS dma interfaces are deleted. A private implementation remains inside of the 32-bit sparc port which exists only for the sake of the implementation of dma_*(). Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-28 09:09:11 +08:00
QSIZE(QLOGICPTI_REQ_QUEUE_LEN),
qpti->req_cpu, qpti->req_dvma);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
#undef QSIZE
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
of_iounmap(&op->resource[0], qpti->qregs,
resource_size(&op->resource[0]));
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
if (qpti->is_pti)
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
of_iounmap(&op->resource[0], qpti->sreg, sizeof(unsigned char));
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
scsi_host_put(qpti->qhost);
return 0;
}
sparc: Annotate of_device_id arrays with const or __initdata. As suggested by Stephen Rothwell. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-31 16:23:17 +08:00
static const struct of_device_id qpti_match[] = {
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
{
.name = "ptisp",
},
{
.name = "PTI,ptisp",
},
{
.name = "QLGC,isp",
},
{
.name = "SUNW,isp",
},
{},
};
MODULE_DEVICE_TABLE(of, qpti_match);
dt/sparc: Eliminate users of of_platform_{,un}register_driver Get rid of old users of of_platform_driver in arch/sparc. Most of_platform_driver users can be converted to use the platform_bus directly. Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-02-23 11:01:33 +08:00
static struct platform_driver qpti_sbus_driver = {
of: Remove duplicate fields from of_platform_driver .name, .match_table and .owner are duplicated in both of_platform_driver and device_driver. This patch is a removes the extra copies from struct of_platform_driver and converts all users to the device_driver members. This patch is a pretty mechanical change. The usage model doesn't change and if any drivers have been missed, or if anything has been fixed up incorrectly, then it will fail with a compile time error, and the fixup will be trivial. This patch looks big and scary because it touches so many files, but it should be pretty safe. Signed-off-by: Grant Likely <grant.likely@secretlab.ca> Acked-by: Sean MacLennan <smaclennan@pikatech.com>
2010-04-14 07:13:02 +08:00
.driver = {
.name = "qpti",
.owner = THIS_MODULE,
.of_match_table = qpti_match,
},
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
.probe = qpti_sbus_probe,
Drivers: scsi: remove __dev* attributes. CONFIG_HOTPLUG is going away as an option. As a result, the __dev* markings need to be removed. This change removes the use of __devinit, __devexit_p, __devinitdata, __devinitconst, and __devexit from these drivers. Based on patches originally written by Bill Pemberton, but redone by me in order to handle some of the coding style issues better, by hand. Cc: Bill Pemberton <wfp5p@virginia.edu> Cc: Adam Radford <linuxraid@lsi.com> Cc: "James E.J. Bottomley" <JBottomley@parallels.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2012-12-22 05:08:55 +08:00
.remove = qpti_sbus_remove,
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
};
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
static int __init qpti_init(void)
{
dt/sparc: Eliminate users of of_platform_{,un}register_driver Get rid of old users of of_platform_driver in arch/sparc. Most of_platform_driver users can be converted to use the platform_bus directly. Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-02-23 11:01:33 +08:00
return platform_driver_register(&qpti_sbus_driver);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
}
static void __exit qpti_exit(void)
{
dt/sparc: Eliminate users of of_platform_{,un}register_driver Get rid of old users of of_platform_driver in arch/sparc. Most of_platform_driver users can be converted to use the platform_bus directly. Signed-off-by: Grant Likely <grant.likely@secretlab.ca>
2011-02-23 11:01:33 +08:00
platform_driver_unregister(&qpti_sbus_driver);
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
}
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
MODULE_DESCRIPTION("QlogicISP SBUS driver");
MODULE_AUTHOR("David S. Miller (davem@davemloft.net)");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
MODULE_LICENSE("GPL");
qlogicpti: Convert to pure OF driver. Signed-off-by: David S. Miller <davem@davemloft.net>
2008-08-27 15:24:00 +08:00
MODULE_VERSION("2.1");
[SCSI] qlogicpti: use request_firmware Firmware blob is little endian Thanks to Stephen Rothwell for fixing typos Signed-off-by: Jaswinder Singh Rajput <jaswinderrajput@gmail.com> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com>
2009-04-03 04:12:50 +08:00
MODULE_FIRMWARE("qlogic/isp1000.bin");
Linux-2.6.12-rc2 Initial git repository build. I'm not bothering with the full history, even though we have it. We can create a separate "historical" git archive of that later if we want to, and in the meantime it's about 3.2GB when imported into git - space that would just make the early git days unnecessarily complicated, when we don't have a lot of good infrastructure for it. Let it rip!
2005-04-17 06:20:36 +08:00
[SCSI] qlogicpti: Convert to new SBUS device framework. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-06-24 13:00:34 +08:00
module_init(qpti_init);
module_exit(qpti_exit);