fsi: Patches for 5.5
Subsystems (4)
Joel Stanley
maintainerv2 cover fsi: Patches for 5.5
This series forms the FSI pull request for 5.5.
1-3 adds a FSI class type and updates the documentation for the sysfs files 4-7 makes some cleanups and fixes to the fsi core
8-10 adds the support for the FSI master in the ASPEED AST2600 BMC
The driver has been tested on hardware for most operations. Future enhancements include robust error recovery, DMA support and interrupt support.
The fix for byte order registers is not squashed in as Andrew's commit message is a piece of art that deserves to be in the kernel history. It does not need to go to stable as it fixes a patch earlier in this series.
Andrew Jeffery (3):
trace: fsi: Print transfer size unsigned
fsi: core: Fix small accesses and unaligned offsets via sysfs
fsi: aspeed: Fix OPB0 byte order register values
Jeremy Kerr (2):
fsi: Add fsi-master class
fsi: Move master attributes to fsi-master class
Joel Stanley (5):
ABI: Update FSI path documentation
fsi: Move defines to common header
dt-bindings: fsi: Add description of FSI master
fsi: Add ast2600 master driver
fsi: aspeed: Add trace points
kbuild test robot (1):
fsi: fsi_master_class can be static
Documentation/ABI/testing/sysfs-bus-fsi | 16 +- .../bindings/fsi/fsi-master-aspeed.txt | 24 + drivers/fsi/Kconfig | 8 + drivers/fsi/Makefile | 1 + drivers/fsi/fsi-core.c | 67 ++- drivers/fsi/fsi-master-aspeed.c | 544 ++++++++++++++++++ drivers/fsi/fsi-master-hub.c | 46 -- drivers/fsi/fsi-master.h | 71 +++ include/trace/events/fsi.h | 6 +- include/trace/events/fsi_master_aspeed.h | 77 +++ 10 files changed, 785 insertions(+), 75 deletions(-) create mode 100644 Documentation/devicetree/bindings/fsi/fsi-master-aspeed.txt create mode 100644 drivers/fsi/fsi-master-aspeed.c create mode 100644 include/trace/events/fsi_master_aspeed.h
Signature · 1 line
2.24.0.rc1
Joel Stanley
maintainerPATCH v2 1/11 fsi: Add fsi-master class
From: Jeremy Kerr <[email protected]>
This change adds a device class for FSI masters, allowing access under /sys/class/fsi-master/, and easier udev rules.
--- drivers/fsi/fsi-core.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)
drivers/fsi/fsi-core.c +13 -0
@@ -1241,6 +1241,10 @@ static ssize_t master_break_store(struct device *dev, static DEVICE_ATTR(break, 0200, NULL, master_break_store); +struct class fsi_master_class = { + .name ="fsi-master" , +}; +int fsi_master_register (struct fsi_master *master) {int rc;
@@ -1249,6 +1253,7 @@ int fsi_master_register(struct fsi_master *master)mutex_init (&master->scan_lock); master->idx =ida_simple_get (&master_ida,0 ,INT_MAX ,GFP_KERNEL );dev_set_name (&master->dev,"fsi%d" , master->idx); + master->dev.class = &fsi_master_class; rc =device_register (&master->dev);if (rc) {
@@ -1350,8 +1355,15 @@ static int __init fsi_init(void) rc =bus_register (&fsi_bus_type);if (rc) goto fail_bus; + + rc =class_register (&fsi_master_class); +if (rc) + goto fail_class; +return 0 ; + fail_class: +bus_unregister (&fsi_bus_type); fail_bus:unregister_chrdev_region (fsi_base_dev,FSI_CHAR_MAX_DEVICES );return rc;
@@ -1360,6 +1372,7 @@ postcore_initcall(fsi_init);static void fsi_exit (void ) { +class_unregister (&fsi_master_class);bus_unregister (&fsi_bus_type);unregister_chrdev_region (fsi_base_dev,FSI_CHAR_MAX_DEVICES );ida_destroy (&fsi_minor_ida);
Signature · 1 line
2.24.0.rc1
Re: [PATCH v2 01/11] fsi: Add fsi-master class
Support for the new paths has been added to our user space tools as well so it shouldn't change anything there.
Quoted 59 lines · AEDT Joel Stanley
From: Jeremy Kerr <[email protected]>
This change adds a device class for FSI masters, allowing access under
/sys/class/fsi-master/, and easier udev rules.Signed-off-by Jeremy Kerr <[email protected]>Signed-off-by Joel Stanley <[email protected]>--- drivers/fsi/fsi-core.c | 13 +++++++++++++ 1 file changed, 13 insertions(+)drivers/fsi/fsi-core.c +13
-0@@ -1241,6 +1241,10 @@ static ssize_t master_break_store(struct device *dev, static DEVICE_ATTR(break, 0200, NULL, master_break_store); +struct class fsi_master_class = { + .name ="fsi-master" , +}; +int fsi_master_register (struct fsi_master *master) {int rc;@@ -1249,6 +1253,7 @@ int fsi_master_register(struct fsi_master *master)mutex_init (&master->scan_lock); master->idx =ida_simple_get (&master_ida,0 ,INT_MAX ,GFP_KERNEL );dev_set_name (&master->dev,"fsi%d" , master->idx); + master->dev.class = &fsi_master_class; rc =device_register (&master->dev);if (rc) {@@ -1350,8 +1355,15 @@ static int __init fsi_init(void) rc =bus_register (&fsi_bus_type);if (rc) goto fail_bus; + + rc =class_register (&fsi_master_class); +if (rc) + goto fail_class; +return 0 ; + fail_class: +bus_unregister (&fsi_bus_type); fail_bus:unregister_chrdev_region (fsi_base_dev,FSI_CHAR_MAX_DEVICES );return rc;@@ -1360,6 +1372,7 @@ postcore_initcall(fsi_init);static void fsi_exit (void ) { +class_unregister (&fsi_master_class);bus_unregister (&fsi_bus_type);unregister_chrdev_region (fsi_base_dev,FSI_CHAR_MAX_DEVICES );ida_destroy (&fsi_minor_ida);
Joel Stanley
maintainerPATCH v2 2/11 fsi: Move master attributes to fsi-master class
In reply to Joel StanleyFrom: Jeremy Kerr <[email protected]>
Populate fsi_master_class->dev_attrs with the existing attribute definitions, so we don't need to explicitly register.
--- drivers/fsi/fsi-core.c | 23 +++++++++-------------- 1 file changed, 9 insertions(+), 14 deletions(-)
drivers/fsi/fsi-core.c +9 -14
@@ -1241,8 +1241,17 @@ static ssize_t master_break_store(struct device *dev, static DEVICE_ATTR(break, 0200, NULL, master_break_store); +static struct attribute *master_attrs[] = { + &dev_attr_break.attr, + &dev_attr_rescan.attr, +NULL +}; + +ATTRIBUTE_GROUPS (master); +struct class fsi_master_class = { .name ="fsi-master" , + .dev_groups = master_groups, };int fsi_master_register (struct fsi_master *master)
@@ -1261,20 +1270,6 @@ int fsi_master_register(struct fsi_master *master)return rc; } - rc =device_create_file (&master->dev, &dev_attr_rescan); -if (rc) { -device_del (&master->dev); -ida_simple_remove (&master_ida, master->idx); -return rc; - } - - rc =device_create_file (&master->dev, &dev_attr_break); -if (rc) { -device_del (&master->dev); -ida_simple_remove (&master_ida, master->idx); -return rc; - } - np =dev_of_node (&master->dev);if (!of_property_read_bool (np,"no-scan-on-init" )) {mutex_lock (&master->scan_lock);
Signature · 1 line
2.24.0.rc1
Re: [PATCH v2 02/11] fsi: Move master attributes to fsi-master class
Quoted 55 lines · AEDT Joel Stanley
From: Jeremy Kerr <[email protected]>
Populate
fsi_master_class->dev_attrswith the existing attribute definitions, so we don't need to explicitly register.Signed-off-by Jeremy Kerr <[email protected]>Signed-off-by Joel Stanley <[email protected]>--- drivers/fsi/fsi-core.c | 23 +++++++++-------------- 1 file changed, 9 insertions(+), 14 deletions(-)drivers/fsi/fsi-core.c +9
-14@@ -1241,8 +1241,17 @@ static ssize_t master_break_store(struct device *dev, static DEVICE_ATTR(break, 0200, NULL, master_break_store); +static struct attribute *master_attrs[] = { + &dev_attr_break.attr, + &dev_attr_rescan.attr, +NULL +}; + +ATTRIBUTE_GROUPS (master); +struct class fsi_master_class = { .name ="fsi-master" , + .dev_groups = master_groups, };int fsi_master_register (struct fsi_master *master)@@ -1261,20 +1270,6 @@ int fsi_master_register(struct fsi_master *master)return rc; } - rc =device_create_file (&master->dev, &dev_attr_rescan); -if (rc) { -device_del (&master->dev); -ida_simple_remove (&master_ida, master->idx); -return rc; - } - - rc =device_create_file (&master->dev, &dev_attr_break); -if (rc) { -device_del (&master->dev); -ida_simple_remove (&master_ida, master->idx); -return rc; - } - np =dev_of_node (&master->dev);if (!of_property_read_bool (np,"no-scan-on-init" )) {mutex_lock (&master->scan_lock);
Joel Stanley
maintainerPATCH v2 3/11 ABI: Update FSI path documentation
In reply to Joel StanleyThe paths added back in 4.13 weren't quite correct. The in reality the files documented lived under
/sys/devices/../fsi0/rescan
/sys/devices/../fsi0/break
/sys/devices/../fsi0/slave@00:00/term
/sys/devices/../fsi0/slave@00:00/raw
In 5.5 with the addition of the FSI class they move to
/sys/devices/../fsi-master/fsi0/rescan /sys/devices/../fsi-master/fsi0/break /sys/devices/../fsi-master/fsi0/slave@00:00/term /sys/devices/../fsi-master/fsi0/slave@00:00/raw
This is closer to how the (incorrect) documentation described them.
--- Documentation/ABI/testing/sysfs-bus-fsi | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-)
Documentation/ABI/testing/sysfs-bus-fsi +8 -8
@@ -1,25 +1,25 @@ -What: /sys/bus/platform/devices/fsi-master/rescan +What: /sys/bus/platform/devices/../fsi-master/fsi0/rescan Date: May 2017 KernelVersion: 4.12 -Contact: [email protected] +Contact: [email protected] Description: Initiates a FSI master scan for all connected slave devices on its links. -What: /sys/bus/platform/devices/fsi-master/break +What: /sys/bus/platform/devices/../fsi-master/fsi0/break Date: May 2017 KernelVersion: 4.12 -Contact: [email protected] +Contact: [email protected] Description: Sends an FSI BREAK command on a master's communication link to any connnected slaves. A BREAK resets connected device's logic and preps it to receive further commands from the master. -What: /sys/bus/platform/devices/fsi-master/slave@00:00/term +What: /sys/bus/platform/devices/../fsi-master/fsi0/slave@00:00/term Date: May 2017 KernelVersion: 4.12 -Contact: [email protected] +Contact: [email protected] Description: Sends an FSI terminate command from the master to its connected slave. A terminate resets the slave's state machines
@@ -29,10 +29,10 @@ Description: ongoing operation in case of an expired 'Master Time Out' timer. -What: /sys/bus/platform/devices/fsi-master/slave@00:00/raw +What: /sys/bus/platform/devices/../fsi-master/fsi0/slave@00:00/raw Date: May 2017 KernelVersion: 4.12 -Contact: [email protected] +Contact: [email protected] Description: Provides a means of reading/writing a 32 bit value from/to a specified FSI bus address.
Signature · 1 line
2.24.0.rc1
Re: [PATCH v2 03/11] ABI: Update FSI path documentation
Quoted 1 line · AEDT Joel Stanley
The paths added back in 4.13 weren't quite correct. The in reality the
Minor nit-pick, it should read "In reality the...". Otherwise looks reasonable.
Quoted 71 lines
files documented lived under
/sys/devices/../fsi0/rescan
/sys/devices/../fsi0/break
/sys/devices/../fsi0/slave@00:00/term
/sys/devices/../fsi0/slave@00:00/rawIn 5.5 with the addition of the FSI class they move to
/sys/devices/../fsi-master/fsi0/rescan/sys/devices/../fsi-master/fsi0/break/sys/devices/../fsi-master/fsi0/slave@00:00/term /sys/devices/../fsi-master/fsi0/slave@00:00/rawThis is closer to how the (incorrect) documentation described them.
Signed-off-by Joel Stanley <[email protected]>--- Documentation/ABI/testing/sysfs-bus-fsi | 16 ++++++++-------- 1 file changed, 8 insertions(+), 8 deletions(-)Documentation/ABI/testing/sysfs-bus-fsi +8
-8@@ -1,25 +1,25 @@ -What: /sys/bus/platform/devices/fsi-master/rescan +What: /sys/bus/platform/devices/../fsi-master/fsi0/rescan Date: May 2017 KernelVersion: 4.12 -Contact: [email protected] +Contact: [email protected] Description: Initiates a FSI master scan for all connected slave devices on its links. -What: /sys/bus/platform/devices/fsi-master/break +What: /sys/bus/platform/devices/../fsi-master/fsi0/break Date: May 2017 KernelVersion: 4.12 -Contact: [email protected] +Contact: [email protected] Description: Sends an FSI BREAK command on a master's communication link to any connnected slaves. A BREAK resets connected device's logic and preps it to receive further commands from the master. -What: /sys/bus/platform/devices/fsi-master/slave@00:00/term +What: /sys/bus/platform/devices/../fsi-master/fsi0/slave@00:00/term Date: May 2017 KernelVersion: 4.12 -Contact: [email protected] +Contact: [email protected] Description: Sends an FSI terminate command from the master to its connected slave. A terminate resets the slave's state machines@@ -29,10 +29,10 @@ Description: ongoing operation in case of an expired 'Master Time Out' timer. -What: /sys/bus/platform/devices/fsi-master/slave@00:00/raw +What: /sys/bus/platform/devices/../fsi-master/fsi0/slave@00:00/raw Date: May 2017 KernelVersion: 4.12 -Contact: [email protected] +Contact: [email protected] Description: Provides a means of reading/writing a 32 bit value from/to a specified FSI bus address.
Joel Stanley
maintainerPATCH v2 4/11 trace: fsi: Print transfer size unsigned
In reply to Joel StanleyFrom: Andrew Jeffery <[email protected]>
Due to other bugs I observed a spurious -1 transfer size.
--- include/trace/events/fsi.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
include/trace/events/fsi.h +3 -3
@@ -26,7 +26,7 @@ TRACE_EVENT(fsi_master_read, __entry->addr = addr; __entry->size = size; ), - TP_printk("fsi%d:%02d:%02d %08x[%zd]", + TP_printk("fsi%d:%02d:%02d %08x[%zu]", __entry->master_idx, __entry->link, __entry->id,
@@ -56,7 +56,7 @@ TRACE_EVENT(fsi_master_write, __entry->data =0 ;memcpy (&__entry->data, data, size); ), - TP_printk("fsi%d:%02d:%02d %08x[%zd] <= {%*ph}", + TP_printk("fsi%d:%02d:%02d %08x[%zu] <= {%*ph}", __entry->master_idx, __entry->link, __entry->id,
@@ -93,7 +93,7 @@ TRACE_EVENT(fsi_master_rw_result,if (__entry->write || !__entry->ret)memcpy (&__entry->data, data, size); ), - TP_printk("fsi%d:%02d:%02d %08x[%zd] %s {%*ph} ret %d", + TP_printk("fsi%d:%02d:%02d %08x[%zu] %s {%*ph} ret %d", __entry->master_idx, __entry->link, __entry->id,
Signature · 1 line
2.24.0.rc1
Re: [PATCH v2 04/11] trace: fsi: Print transfer size unsigned
Quoted 42 lines · AEDT Joel Stanley
From: Andrew Jeffery <[email protected]>
Due to other bugs I observed a spurious -1 transfer size.
Signed-off-by Andrew Jeffery <[email protected]>Signed-off-by Joel Stanley <[email protected]>--- include/trace/events/fsi.h | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)include/trace/events/fsi.h +3
-3@@ -26,7 +26,7 @@ TRACE_EVENT(fsi_master_read, __entry->addr = addr; __entry->size = size; ), - TP_printk("fsi%d:%02d:%02d %08x[%zd]", + TP_printk("fsi%d:%02d:%02d %08x[%zu]", __entry->master_idx, __entry->link, __entry->id,@@ -56,7 +56,7 @@ TRACE_EVENT(fsi_master_write, __entry->data =0 ;memcpy (&__entry->data, data, size); ), - TP_printk("fsi%d:%02d:%02d %08x[%zd] <= {%*ph}", + TP_printk("fsi%d:%02d:%02d %08x[%zu] <= {%*ph}", __entry->master_idx, __entry->link, __entry->id,@@ -93,7 +93,7 @@ TRACE_EVENT(fsi_master_rw_result,if (__entry->write || !__entry->ret)memcpy (&__entry->data, data, size); ), - TP_printk("fsi%d:%02d:%02d %08x[%zd] %s {%*ph} ret %d", + TP_printk("fsi%d:%02d:%02d %08x[%zu] %s {%*ph} ret %d", __entry->master_idx, __entry->link, __entry->id,
Joel Stanley
maintainerPATCH v2 5/11 fsi: core: Fix small accesses and unaligned offsets via sysfs
In reply to Joel StanleyFrom: Andrew Jeffery <[email protected]>
Subtracting the offset delta from four-byte alignment lead to wrapping of the requested length where count is less than off. Generalise the length handling to enable and optimise aligned access sizes for all offset and size combinations. The new formula produces the following results for given offset and count values:
offset count | length
--------------+-------
0 1 | 1
0 2 | 2
0 3 | 2
0 4 | 4
0 5 | 4
1 1 | 1
1 2 | 1
1 3 | 1
1 4 | 1
1 5 | 1
2 1 | 1
2 2 | 2
2 3 | 2
2 4 | 2
2 5 | 2
3 1 | 1
3 2 | 1
3 3 | 1
3 4 | 1
3 5 | 1
We might need something like this for the cfam chardevs as well, for example we don't currently implement any alignment restrictions / handling in the hardware master driver.
--- drivers/fsi/fsi-core.c | 31 +++++++++++++++++++++++++++---- 1 file changed, 27 insertions(+), 4 deletions(-)
drivers/fsi/fsi-core.c +27 -4
@@ -544,6 +544,31 @@ static int fsi_slave_scan(struct fsi_slave *slave)return 0 ; } +static unsigned long aligned_access_size (size_t offset,size_t count) +{ +unsigned long offset_unit, count_unit; + +/* Criteria: +* +* 1. Access size must be less than or equal to the maximum access +* width or the highest power-of-two factor of offset +* 2. Access size must be less than or equal to the amount specified by +* count +* +* The access width is optimal if we can calculate 1 to be strictly +* equal while still satisfying 2. +*/ + +/* Find 1 by the bottom bit of offset (with a 4 byte access cap) */ + offset_unit =BIT (__builtin_ctzl (offset |4 )); + +/* Find 2 by the top bit of count */ + count_unit =BIT (8 *sizeof (unsigned long ) -1 -__builtin_clzl (count)); + +/* Constrain the maximum access width to the minimum of both criteria */ +return BIT (__builtin_ctzl (offset_unit | count_unit)); +} +static ssize_t fsi_slave_sysfs_raw_read (struct file *file,struct kobject *kobj,struct bin_attribute *attr,char *buf,loff_t off,size_t count)
@@ -559,8 +584,7 @@ static ssize_t fsi_slave_sysfs_raw_read(struct file *file,return -EINVAL ;for (total_len =0 ; total_len < count; total_len += read_len) { - read_len =min_t (size_t, count,4 ); - read_len -= off &0x3 ; + read_len =aligned_access_size (off, count - total_len); rc =fsi_slave_read (slave, off, buf + total_len, read_len);if (rc)
@@ -587,8 +611,7 @@ static ssize_t fsi_slave_sysfs_raw_write(struct file *file,return -EINVAL ;for (total_len =0 ; total_len < count; total_len += write_len) { - write_len =min_t (size_t, count,4 ); - write_len -= off &0x3 ; + write_len =aligned_access_size (off, count - total_len); rc =fsi_slave_write (slave, off, buf + total_len, write_len);if (rc)
Signature · 1 line
2.24.0.rc1
Joel Stanley
maintainerPATCH v2 6/11 fsi: fsi_master_class can be static
In reply to Joel StanleyFrom: kbuild test robot <[email protected]>
There are no users outside of this file.
--- drivers/fsi/fsi-core.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/fsi/fsi-core.c +1 -1
@@ -1272,7 +1272,7 @@ static struct attribute *master_attrs[] = {ATTRIBUTE_GROUPS (master); -struct class fsi_master_class = { +static struct class fsi_master_class = { .name ="fsi-master" , .dev_groups = master_groups, };
Signature · 1 line
2.24.0.rc1
Joel Stanley
maintainerPATCH v2 7/11 fsi: Move defines to common header
In reply to Joel StanleyThe FSI master registers are common to the hub and AST2600 master (and the FSP2, if someone was to upstream a driver for that).
Add defines to the fsi-master.h header, and introduce headings to delineate the existing low level details.
--- drivers/fsi/fsi-master-hub.c | 46 ----------------------- drivers/fsi/fsi-master.h | 71 ++++++++++++++++++++++++++++++++++++ 2 files changed, 71 insertions(+), 46 deletions(-)
drivers/fsi/fsi-master-hub.c +0 -46
@@ -13,53 +13,7 @@#include "fsi-master.h" -/* Control Registers */ -#define FSI_MMODE 0x0/* R/W: mode */ -#define FSI_MDLYR 0x4/* R/W: delay */ -#define FSI_MCRSP 0x8/* R/W: clock rate */ -#define FSI_MENP0 0x10/* R/W: enable */ -#define FSI_MLEVP0 0x18/* R: plug detect */ -#define FSI_MSENP0 0x18/* S: Set enable */ -#define FSI_MCENP0 0x20/* C: Clear enable */ -#define FSI_MAEB 0x70/* R: Error address */ -#define FSI_MVER 0x74/* R: master version/type */ -#define FSI_MRESP0 0xd0/* W: Port reset */ -#define FSI_MESRB0 0x1d0/* R: Master error status */ -#define FSI_MRESB0 0x1d0/* W: Reset bridge */ -#define FSI_MECTRL 0x2e0/* W: Error control */ - -/* MMODE: Mode control */ -#define FSI_MMODE_EIP 0x80000000/* Enable interrupt polling */ -#define FSI_MMODE_ECRC 0x40000000/* Enable error recovery */ -#define FSI_MMODE_EPC 0x10000000/* Enable parity checking */ -#define FSI_MMODE_P8_TO_LSB 0x00000010/* Timeout value LSB */ -/* MSB=1, LSB=0 is 0.8 ms */ -/* MSB=0, LSB=1 is 0.9 ms */ -#define FSI_MMODE_CRS0SHFT 18/* Clk rate selection 0 shift */ -#define FSI_MMODE_CRS0MASK 0x3ff/* Clk rate selection 0 mask */ -#define FSI_MMODE_CRS1SHFT 8/* Clk rate selection 1 shift */ -#define FSI_MMODE_CRS1MASK 0x3ff/* Clk rate selection 1 mask */ - -/* MRESB: Reset brindge */ -#define FSI_MRESB_RST_GEN 0x80000000/* General reset */ -#define FSI_MRESB_RST_ERR 0x40000000/* Error Reset */ - -/* MRESB: Reset port */ -#define FSI_MRESP_RST_ALL_MASTER 0x20000000/* Reset all FSI masters */ -#define FSI_MRESP_RST_ALL_LINK 0x10000000/* Reset all FSI port contr. */ -#define FSI_MRESP_RST_MCR 0x08000000/* Reset FSI master reg. */ -#define FSI_MRESP_RST_PYE 0x04000000/* Reset FSI parity error */ -#define FSI_MRESP_RST_ALL 0xfc000000/* Reset any error */ - -/* MECTRL: Error control */ -#define FSI_MECTRL_EOAE 0x8000/* Enable machine check when */ -/* master 0 in error */ -#define FSI_MECTRL_P8_AUTO_TERM 0x4000/* Auto terminate */ -#define FSI_ENGID_HUB_MASTER 0x1c -#define FSI_HUB_LINK_OFFSET 0x80000 -#define FSI_HUB_LINK_SIZE 0x80000 -#define FSI_HUB_MASTER_MAX_LINKS 8#define FSI_LINK_ENABLE_SETUP_TIME 10/* in mS */
drivers/fsi/fsi-master.h +71 -0
83 lines not on this page · raw message
Signature · 1 line
2.24.0.rc1
Joel Stanley
maintainerPATCH v2 8/11 dt-bindings: fsi: Add description of FSI master
In reply to Joel StanleyThis describes the FSI master present in the AST2600.
--- .../bindings/fsi/fsi-master-aspeed.txt | 24 +++++++++++++++++++ 1 file changed, 24 insertions(+) create mode 100644 Documentation/devicetree/bindings/fsi/fsi-master-aspeed.txt
Documentation/devicetree/bindings/fsi/fsi-master-aspeed.txt new file +24 -0
@@ -0,0 +1,24 @@ +Device-tree bindings for AST2600 FSI master +------------------------------------------- + +The AST2600 contains two identical FSI masters. They share a clock and have a +separate interrupt line and output pins. + +Required properties: + - compatible: "aspeed,ast2600-fsi-master" + - reg: base address and length + - clocks: phandle and clock number + - interrupts: platform dependent interrupt description + - pinctrl-0: phandle to pinctrl node + - pinctrl-names: pinctrl state + +Examples: + + fsi-master { + compatible = "aspeed,ast2600-fsi-master", "fsi-master"; + reg = <0x1e79b000 0x94>; + interrupts = <GIC_SPI 100 IRQ_TYPE_LEVEL_HIGH>; + pinctrl-names = "default"; + pinctrl-0 = <&pinctrl_fsi1_default>; + clocks = <&syscon ASPEED_CLK_GATE_FSICLK>; + };
Signature · 1 line
2.24.0.rc1
Re: [PATCH v2 08/11] dt-bindings: fsi: Add description of FSI master
Quoted 39 lines · AEDT Joel Stanley
This describes the FSI master present in the AST2600.
Signed-off-by Joel Stanley <[email protected]>--- .../bindings/fsi/fsi-master-aspeed.txt | 24 +++++++++++++++++++ 1 file changed, 24 insertions(+) create mode 100644 Documentation/devicetree/bindings/fsi/fsi-master-aspeed.txtDocumentation/devicetree/bindings/fsi/fsi-master-aspeed.txt new file +24
-0@@ -0,0 +1,24 @@ +Device-tree bindings for AST2600 FSI master +------------------------------------------- + +The AST2600 contains two identical FSI masters. They share a clock and have a +separate interrupt line and output pins. + +Required properties: + - compatible: "aspeed,ast2600-fsi-master" + - reg: base address and length + - clocks: phandle and clock number + - interrupts: platform dependent interrupt description + - pinctrl-0: phandle to pinctrl node + - pinctrl-names: pinctrl state + +Examples: + + fsi-master { + compatible = "aspeed,ast2600-fsi-master", "fsi-master"; + reg = <0x1e79b000 0x94>; + interrupts = <GIC_SPI 100 IRQ_TYPE_LEVEL_HIGH>; + pinctrl-names = "default"; + pinctrl-0 = <&pinctrl_fsi1_default>; + clocks = <&syscon ASPEED_CLK_GATE_FSICLK>; + };
Re: [PATCH v2 08/11] dt-bindings: fsi: Add description of FSI master
In reply to Joel StanleyQuoted 42 lines · Joel Stanley
This describes the FSI master present in the AST2600.
Signed-off-by Joel Stanley <[email protected]>Acked-by Alistair Popple <[email protected]>--- .../bindings/fsi/fsi-master-aspeed.txt | 24 +++++++++++++++++++ 1 file changed, 24 insertions(+) create mode 100644 Documentation/devicetree/bindings/fsi/fsi-master-aspeed.txtDocumentation/devicetree/bindings/fsi/fsi-master-aspeed.txt new file +24
-0@@ -0,0 +1,24 @@ +Device-tree bindings for AST2600 FSI master +------------------------------------------- + +The AST2600 contains two identical FSI masters. They share a clock and have a +separate interrupt line and output pins. + +Required properties: + - compatible: "aspeed,ast2600-fsi-master" + - reg: base address and length + - clocks: phandle and clock number + - interrupts: platform dependent interrupt description + - pinctrl-0: phandle to pinctrl node + - pinctrl-names: pinctrl state + +Examples: + + fsi-master { + compatible = "aspeed,ast2600-fsi-master", "fsi-master"; + reg = <0x1e79b000 0x94>; + interrupts = <GIC_SPI 100 IRQ_TYPE_LEVEL_HIGH>; + pinctrl-names = "default"; + pinctrl-0 = <&pinctrl_fsi1_default>; + clocks = <&syscon ASPEED_CLK_GATE_FSICLK>; + };Signature · 1 line
2.24.0.rc1
As these all seem like bog-standard properties, I'll take this now, thanks.
greg k-h
Joel Stanley
maintainerPATCH v2 9/11 fsi: Add ast2600 master driver
In reply to Joel StanleyThe ast2600 BMC has a pair of FSI masters in it, behind an AHB to OPB bridge.
The master driver supports reads and writes of full words, half word and byte accesses to remote CFAMs. It can perform very basic error recovery through resetting of the FSI port when an error is detected, and the issuing of breaks and terms.
--
v2:
- remove debugging
- squash in fixes
--- drivers/fsi/Kconfig | 8 + drivers/fsi/Makefile | 1 + drivers/fsi/fsi-master-aspeed.c | 522 ++++++++++++++++++++++++++++++++ 3 files changed, 531 insertions(+) create mode 100644 drivers/fsi/fsi-master-aspeed.c
drivers/fsi/Kconfig +8 -0
@@ -53,6 +53,14 @@ config FSI_MASTER_AST_CF lines driven by the internal ColdFire coprocessor. This requires the corresponding machine specific ColdFire firmware to be available. +config FSI_MASTER_ASPEED + tristate "FSI ASPEED master" + help + This option enables a FSI master that is present behind an OPB bridge + in the AST2600. + + Enable it for your BMC kernel in an OpenPower or IBM Power system. + config FSI_SCOM tristate "SCOM FSI client device driver" ---help---
drivers/fsi/Makefile +1 -0
@@ -2,6 +2,7 @@ obj-$(CONFIG_FSI) += fsi-core.o obj-$(CONFIG_FSI_MASTER_HUB) += fsi-master-hub.o +obj-$(CONFIG_FSI_MASTER_ASPEED) += fsi-master-aspeed.o obj-$(CONFIG_FSI_MASTER_GPIO) += fsi-master-gpio.o obj-$(CONFIG_FSI_MASTER_AST_CF) += fsi-master-ast-cf.o obj-$(CONFIG_FSI_SCOM) += fsi-scom.o
drivers/fsi/fsi-master-aspeed.c new file +522 -0
522 lines not on this page · raw message
Signature · 1 line
2.24.0.rc1
Re: [PATCH v2 09/11] fsi: Add ast2600 master driver
Having a hardware master should make things quicker.
Quoted 570 lines · AEDT Joel Stanley
The ast2600 BMC has a pair of FSI masters in it, behind an AHB to OPB bridge.
The master driver supports reads and writes of full words, half word and byte accesses to remote CFAMs. It can perform very basic error recovery through resetting of the FSI port when an error is detected, and the issuing of breaks and terms.
Signed-off-by Joel Stanley <[email protected]>--
v2:
- remove debugging
- squash in fixes--- drivers/fsi/Kconfig | 8 + drivers/fsi/Makefile | 1 + drivers/fsi/fsi-master-aspeed.c | 522 ++++++++++++++++++++++++++++++++ 3 files changed, 531 insertions(+) create mode 100644 drivers/fsi/fsi-master-aspeed.cdrivers/fsi/Kconfig +8
-0@@ -53,6 +53,14 @@ config FSI_MASTER_AST_CF lines driven by the internal ColdFire coprocessor. This requires the corresponding machine specific ColdFire firmware to be available. +config FSI_MASTER_ASPEED + tristate "FSI ASPEED master" + help + This option enables a FSI master that is present behind an OPB bridge + in the AST2600. + + Enable it for your BMC kernel in an OpenPower or IBM Power system. + config FSI_SCOM tristate "SCOM FSI client device driver" ---help---drivers/fsi/Makefile +1
-0@@ -2,6 +2,7 @@ obj-$(CONFIG_FSI) += fsi-core.o obj-$(CONFIG_FSI_MASTER_HUB) += fsi-master-hub.o +obj-$(CONFIG_FSI_MASTER_ASPEED) += fsi-master-aspeed.o obj-$(CONFIG_FSI_MASTER_GPIO) += fsi-master-gpio.o obj-$(CONFIG_FSI_MASTER_AST_CF) += fsi-master-ast-cf.o obj-$(CONFIG_FSI_SCOM) += fsi-scom.odrivers/fsi/fsi-master-aspeed.c new file +512
-0@@ -0,0 +1,522 @@ +// SPDX-License-Identifier: GPL-2.0-or-later +// Copyright (C) IBM Corporation 2018 +// FSI master driver for AST2600 + +#include <linux/clk.h> +#include <linux/delay.h> +#include <linux/fsi.h> +#include <linux/io.h> +#include <linux/mfd/syscon.h> +#include <linux/module.h> +#include <linux/of.h> +#include <linux/platform_device.h> +#include <linux/regmap.h> +#include <linux/slab.h> +#include <linux/iopoll.h> + +#include "fsi-master.h" + +struct fsi_master_aspeed { +struct fsi_master master; +struct device *dev; +void __iomem *base; +struct clk *clk; +}; + +#define to_fsi_master_aspeed (m) \ + container_of(m, struct fsi_master_aspeed, master) + +/* Control register (size 0x400) */ +static const u32 ctrl_base =0x80000000 ; + +static const u32 fsi_base =0xa0000000 ; + +#define OPB_FSI_VER 0x00 +#define OPB_TRIGGER 0x04 +#define OPB_CTRL_BASE 0x08 +#define OPB_FSI_BASE 0x0c +#define OPB_CLK_SYNC 0x3c +#define OPB_IRQ_CLEAR 0x40 +#define OPB_IRQ_MASK 0x44 +#define OPB_IRQ_STATUS 0x48 + +#define OPB0_SELECT 0x10 +#define OPB0_RW 0x14 +#define OPB0_XFER_SIZE 0x18 +#define OPB0_FSI_ADDR 0x1c +#define OPB0_FSI_DATA_W 0x20 +#define OPB0_STATUS 0x80 +#define OPB0_FSI_DATA_R 0x84 + +#define OPB0_WRITE_ORDER1 0x4c +#define OPB0_WRITE_ORDER2 0x50 +#define OPB1_WRITE_ORDER1 0x54 +#define OPB1_WRITE_ORDER2 0x58 +#define OPB0_READ_ORDER1 0x5c +#define OPB1_READ_ORDER2 0x60 + +#define OPB_RETRY_COUNTER 0x64 + +/* OPBn_STATUS */ +#define STATUS_HALFWORD_ACK BIT(0) +#define STATUS_FULLWORD_ACK BIT(1) +#define STATUS_ERR_ACK BIT(2) +#define STATUS_RETRY BIT(3) +#define STATUS_TIMEOUT BIT(4) + +/* OPB_IRQ_MASK */ +#define OPB1_XFER_ACK_EN BIT(17) +#define OPB0_XFER_ACK_EN BIT(16) + +/* OPB_RW */ +#define CMD_READ BIT(0) +#define CMD_WRITE 0 + +/* OPBx_XFER_SIZE */ +#define XFER_FULLWORD (BIT(1) | BIT(0)) +#define XFER_HALFWORD (BIT(0)) +#define XFER_BYTE (0) + +#define FSI_LINK_ENABLE_SETUP_TIME 10/* in mS */ + +#define DEFAULT_DIVISOR 14 +#define OPB_POLL_TIMEOUT 10000 + +static int __opb_write (struct fsi_master_aspeed *aspeed,u32 addr, +u32 val,u32 transfer_size) +{ +void __iomem *base = aspeed->base; +u32 reg, status; +int ret; + +writel (CMD_WRITE , base +OPB0_RW ); +writel (transfer_size, base +OPB0_XFER_SIZE ); +writel (addr, base +OPB0_FSI_ADDR ); +writel (val, base +OPB0_FSI_DATA_W ); +writel (0x1 , base +OPB_IRQ_CLEAR ); +writel (0x1 , base +OPB_TRIGGER ); + + ret =readl_poll_timeout (base +OPB_IRQ_STATUS , reg, + (reg &OPB0_XFER_ACK_EN ) !=0 , +0 ,OPB_POLL_TIMEOUT ); + + status =readl (base +OPB0_STATUS ); + +/* Return error when poll timed out */ +if (ret) +return ret; + +/* Command failed, master will reset */ +if (status &STATUS_ERR_ACK ) +return -EIO ; + +return 0 ; +} + +static int opb_writeb (struct fsi_master_aspeed *aspeed,u32 addr,u8 val) +{ +return __opb_write (aspeed, addr, val,XFER_BYTE ); +} + +static int opb_writew (struct fsi_master_aspeed *aspeed,u32 addr,__be16 val) +{ +return __opb_write (aspeed, addr, (__force u16 )val,XFER_HALFWORD ); +} + +static int opb_writel (struct fsi_master_aspeed *aspeed,u32 addr,__be32 val) +{ +return __opb_write (aspeed, addr, (__force u32 )val,XFER_FULLWORD ); +} + +static int __opb_read (struct fsi_master_aspeed *aspeed,uint32_t addr, +u32 transfer_size,void *out) +{ +void __iomem *base = aspeed->base; +u32 result, reg; +int status, ret; + +writel (CMD_READ , base +OPB0_RW ); +writel (transfer_size, base +OPB0_XFER_SIZE ); +writel (addr, base +OPB0_FSI_ADDR ); +writel (0x1 , base +OPB_IRQ_CLEAR ); +writel (0x1 , base +OPB_TRIGGER ); + + ret =readl_poll_timeout (base +OPB_IRQ_STATUS , reg, + (reg &OPB0_XFER_ACK_EN ) !=0 , +0 ,OPB_POLL_TIMEOUT ); + + status =readl (base +OPB0_STATUS ); + + result =readl (base +OPB0_FSI_DATA_R ); + +/* Return error when poll timed out */ +if (ret) +return ret; + +/* Command failed, master will reset */ +if (status &STATUS_ERR_ACK ) +return -EIO ; + +if (out) { +switch (transfer_size) { +case XFER_BYTE : + *(u8 *)out = result; +break ; +case XFER_HALFWORD : + *(u16 *)out = result; +break ; +case XFER_FULLWORD : + *(u32 *)out = result; +break ; +default : +return -EINVAL ; + } + + } + +return 0 ; +} + +static int opb_readl (struct fsi_master_aspeed *aspeed,uint32_t addr,__be32 *out) +{ +return __opb_read (aspeed, addr,XFER_FULLWORD , out); +} + +static int opb_readw (struct fsi_master_aspeed *aspeed,uint32_t addr,__be16 *out) +{ +return __opb_read (aspeed, addr,XFER_HALFWORD , (void *)out); +} + +static int opb_readb (struct fsi_master_aspeed *aspeed,uint32_t addr,u8 *out) +{ +return __opb_read (aspeed, addr,XFER_BYTE , (void *)out); +} + +static int check_errors (struct fsi_master_aspeed *aspeed,int err) +{ +int ret; + +if (err == -EIO ) { +/* Check MAEB (0x70) ? */ + +/* Then clear errors in master */ + ret =opb_writel (aspeed, ctrl_base +FSI_MRESP0 , +cpu_to_be32 (FSI_MRESP_RST_ALL_MASTER )); +if (ret) { +/* TODO: log? return different code? */ +return ret; + } +/* TODO: confirm that 0x70 was okay */ + } + +/* This will pass through timeout errors */ +return err; +} + +static int aspeed_master_read (struct fsi_master *master,int link, +uint8_t id,uint32_t addr,void *val,size_t size) +{ +struct fsi_master_aspeed *aspeed =to_fsi_master_aspeed (master); +int ret; + +if (id !=0 ) +return -EINVAL ; + + addr += link *FSI_HUB_LINK_SIZE ; + +switch (size) { +case 1 : + ret =opb_readb (aspeed, fsi_base + addr, val); +break ; +case 2 : + ret =opb_readw (aspeed, fsi_base + addr, val); +break ; +case 4 : + ret =opb_readl (aspeed, fsi_base + addr, val); +break ; +default : +return -EINVAL ; + } + + ret =check_errors (aspeed, ret); +if (ret) +return ret; + +return 0 ; +} + +static int aspeed_master_write (struct fsi_master *master,int link, +uint8_t id,uint32_t addr,const void *val,size_t size) +{ +struct fsi_master_aspeed *aspeed =to_fsi_master_aspeed (master); +int ret; + +if (id !=0 ) +return -EINVAL ; + + addr += link *FSI_HUB_LINK_SIZE ; + +switch (size) { +case 1 : + ret =opb_writeb (aspeed, fsi_base + addr, *(u8 *)val); +break ; +case 2 : + ret =opb_writew (aspeed, fsi_base + addr, *(__be16 *)val); +break ; +case 4 : + ret =opb_writel (aspeed, fsi_base + addr, *(__be32 *)val); +break ; +default : +return -EINVAL ; + } + + ret =check_errors (aspeed, ret); +if (ret) +return ret; + +return 0 ; +} + +static int aspeed_master_link_enable (struct fsi_master *master,int link) +{ +struct fsi_master_aspeed *aspeed =to_fsi_master_aspeed (master); +int idx, bit, ret; +__be32 reg, result; + + idx = link /32 ; + bit = link %32 ; + + reg =cpu_to_be32 (0x80000000 >> bit); + + ret =opb_writel (aspeed, ctrl_base +FSI_MSENP0 + (4 * idx), reg); +if (ret) +return ret; + +mdelay (FSI_LINK_ENABLE_SETUP_TIME ); + + ret =opb_readl (aspeed, ctrl_base +FSI_MENP0 + (4 * idx), &result); +if (ret) +return ret; + +if (result != reg) { +dev_err (aspeed->dev,"%s failed: %08x\n" , __func__, result); +return -EIO ; + } + +return 0 ; +} + +static int aspeed_master_term (struct fsi_master *master,int link,uint8_t id) +{ +uint32_t addr; +__be32 cmd; + + addr =0x4 ; + cmd =cpu_to_be32 (0xecc00000 ); + +return aspeed_master_write (master, link, id, addr, &cmd,4 ); +} + +static int aspeed_master_break (struct fsi_master *master,int link) +{ +uint32_t addr; +__be32 cmd; + + addr =0x0 ; + cmd =cpu_to_be32 (0xc0de0000 ); + +return aspeed_master_write (master, link,0 , addr, &cmd,4 ); +} + +static void aspeed_master_release (struct device *dev) +{ +struct fsi_master_aspeed *aspeed = +to_fsi_master_aspeed (dev_to_fsi_master (dev)); + +kfree (aspeed); +} + +/* mmode encoders */ +static inline u32 fsi_mmode_crs0 (u32 x) +{ +return (x &FSI_MMODE_CRS0MASK ) <<FSI_MMODE_CRS0SHFT ; +} + +static inline u32 fsi_mmode_crs1 (u32 x) +{ +return (x &FSI_MMODE_CRS1MASK ) <<FSI_MMODE_CRS1SHFT ; +} + +static int aspeed_master_init (struct fsi_master_aspeed *aspeed) +{ +__be32 reg; + + reg =cpu_to_be32 (FSI_MRESP_RST_ALL_MASTER |FSI_MRESP_RST_ALL_LINK + |FSI_MRESP_RST_MCR |FSI_MRESP_RST_PYE ); +opb_writel (aspeed, ctrl_base +FSI_MRESP0 , reg); + +/* Initialize the MFSI (hub master) engine */ + reg =cpu_to_be32 (FSI_MRESP_RST_ALL_MASTER |FSI_MRESP_RST_ALL_LINK + |FSI_MRESP_RST_MCR |FSI_MRESP_RST_PYE ); +opb_writel (aspeed, ctrl_base +FSI_MRESP0 , reg); + + reg =cpu_to_be32 (FSI_MECTRL_EOAE |FSI_MECTRL_P8_AUTO_TERM ); +opb_writel (aspeed, ctrl_base +FSI_MECTRL , reg); + + reg =cpu_to_be32 (FSI_MMODE_ECRC |FSI_MMODE_EPC |FSI_MMODE_RELA + |fsi_mmode_crs0 (DEFAULT_DIVISOR ) + |fsi_mmode_crs1 (DEFAULT_DIVISOR ) + |FSI_MMODE_P8_TO_LSB ); +opb_writel (aspeed, ctrl_base +FSI_MMODE , reg); + + reg =cpu_to_be32 (0xffff0000 ); +opb_writel (aspeed, ctrl_base +FSI_MDLYR , reg); + + reg =cpu_to_be32 (~0 ); +opb_writel (aspeed, ctrl_base +FSI_MSENP0 , reg); + +/* Leave enabled long enough for master logic to set up */ +mdelay (FSI_LINK_ENABLE_SETUP_TIME ); + +opb_writel (aspeed, ctrl_base +FSI_MCENP0 , reg); + +opb_readl (aspeed, ctrl_base +FSI_MAEB ,NULL ); + + reg =cpu_to_be32 (FSI_MRESP_RST_ALL_MASTER |FSI_MRESP_RST_ALL_LINK ); +opb_writel (aspeed, ctrl_base +FSI_MRESP0 , reg); + +opb_readl (aspeed, ctrl_base +FSI_MLEVP0 ,NULL ); + +/* Reset the master bridge */ + reg =cpu_to_be32 (FSI_MRESB_RST_GEN ); +opb_writel (aspeed, ctrl_base +FSI_MRESB0 , reg); + + reg =cpu_to_be32 (FSI_MRESB_RST_ERR ); +opb_writel (aspeed, ctrl_base +FSI_MRESB0 , reg); + +return 0 ; +} + +static int fsi_master_aspeed_probe (struct platform_device *pdev) +{ +struct fsi_master_aspeed *aspeed; +struct resource *res; +int rc, links, reg; +__be32 raw; + + aspeed =devm_kzalloc (&pdev->dev,sizeof (*aspeed),GFP_KERNEL ); +if (!aspeed) +return -ENOMEM ; + + aspeed->dev = &pdev->dev; + + res =platform_get_resource (pdev,IORESOURCE_MEM ,0 ); + aspeed->base =devm_ioremap_resource (&pdev->dev, res); +if (IS_ERR (aspeed->base)) +return PTR_ERR (aspeed->base); + + aspeed->clk =devm_clk_get (aspeed->dev,NULL ); +if (IS_ERR (aspeed->clk)) { +dev_err (aspeed->dev,"couldn't get clock\n" ); +return PTR_ERR (aspeed->clk); + } + rc =clk_prepare_enable (aspeed->clk); +if (rc) { +dev_err (aspeed->dev,"couldn't enable clock\n" ); +return rc; + } + +writel (0x1 , aspeed->base +OPB_CLK_SYNC ); +writel (OPB1_XFER_ACK_EN |OPB0_XFER_ACK_EN , + aspeed->base +OPB_IRQ_MASK ); + +/* TODO: determine an appropriate value */ +writel (0x10 , aspeed->base +OPB_RETRY_COUNTER ); + +writel (ctrl_base, aspeed->base +OPB_CTRL_BASE ); +writel (fsi_base, aspeed->base +OPB_FSI_BASE ); + +/* Set read data order */ +writel (0x0011bb1b , aspeed->base +OPB0_READ_ORDER1 ); + +/* Set write data order */ +writel (0x0011bb1b , aspeed->base +OPB0_WRITE_ORDER1 ); +writel (0xffaa5500 , aspeed->base +OPB0_WRITE_ORDER2 ); + +/* +* Select OPB0 for all operations. +* Will need to be reworked when enabling DMA or anything that uses +* OPB1. +*/ +writel (0x1 , aspeed->base +OPB0_SELECT ); + + rc =opb_readl (aspeed, ctrl_base +FSI_MVER , &raw); +if (rc) { +dev_err (&pdev->dev,"failed to read hub version\n" ); +return rc; + } + + reg =be32_to_cpu (raw); + links = (reg >>8 ) &0xff ; +dev_info (&pdev->dev,"hub version %08x (%d links)\n" , reg, links); + + aspeed->master.dev.parent = &pdev->dev; + aspeed->master.dev.release = aspeed_master_release; + aspeed->master.dev.of_node =of_node_get (dev_of_node (&pdev->dev)); + + aspeed->master.n_links = links; + aspeed->master.read = aspeed_master_read; + aspeed->master.write = aspeed_master_write; + aspeed->master.send_break = aspeed_master_break; + aspeed->master.term = aspeed_master_term; + aspeed->master.link_enable = aspeed_master_link_enable; + +dev_set_drvdata (&pdev->dev, aspeed); + +aspeed_master_init (aspeed); + + rc =fsi_master_register (&aspeed->master); +if (rc) + goto err_release; + +/* At this point, fsi_master_register performs the device_initialize(), +* and holds the sole reference on master.dev. This means the device +* will be freed (via ->release) during any subsequent call to +* fsi_master_unregister. We add our own reference to it here, so we +* can perform cleanup (in _remove()) without it being freed before +* we're ready. +*/ +get_device (&aspeed->master.dev); +return 0 ; + +err_release: +clk_disable_unprepare (aspeed->clk); +return rc; +} + +static int fsi_master_aspeed_remove (struct platform_device *pdev) +{ +struct fsi_master_aspeed *aspeed =platform_get_drvdata (pdev); + +fsi_master_unregister (&aspeed->master); +clk_disable_unprepare (aspeed->clk); + +return 0 ; +} + +static const struct of_device_id fsi_master_aspeed_match[] = { + { .compatible ="aspeed,ast2600-fsi-master" }, + { }, +}; + +static struct platform_driver fsi_master_aspeed_driver = {
drivers/fsi/fsi-master-aspeed.c line 522
Quoted 10 lines
drivers/fsi/fsi-master-aspeed.c new file +10
-0+ .driver = { + .name = "fsi-master-aspeed", + .of_match_table = fsi_master_aspeed_match, + }, + .probe = fsi_master_aspeed_probe, + .remove = fsi_master_aspeed_remove, +}; + +module_platform_driver (fsi_master_aspeed_driver); +MODULE_LICENSE ("GPL" );
_______________________________________________
linux-arm-kernel mailing list
[email protected]
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Joel Stanley
maintainerPATCH v2 10/11 fsi: aspeed: Add trace points
In reply to Joel StanleyThese trace points help with debugging the FSI master. They show the low level reads, writes and error states of the master.
---
v2: remove unnecessary semicolon
--- drivers/fsi/fsi-master-aspeed.c | 22 +++++++ include/trace/events/fsi_master_aspeed.h | 77 ++++++++++++++++++++++++ 2 files changed, 99 insertions(+) create mode 100644 include/trace/events/fsi_master_aspeed.h
drivers/fsi/fsi-master-aspeed.c +22 -0
@@ -77,6 +77,9 @@ static const u32 fsi_base = 0xa0000000;#define XFER_HALFWORD (BIT(0))#define XFER_BYTE (0) +#define CREATE_TRACE_POINTS +#include <trace/events/fsi_master_aspeed.h> +#define FSI_LINK_ENABLE_SETUP_TIME 10/* in mS */ #define DEFAULT_DIVISOR 14
@@ -102,6 +105,8 @@ static int __opb_write(struct fsi_master_aspeed *aspeed, u32 addr, status =readl (base +OPB0_STATUS ); +trace_fsi_master_aspeed_opb_write (addr, val, transfer_size, status, reg); +/* Return error when poll timed out */ if (ret)return ret;
@@ -149,6 +154,10 @@ static int __opb_read(struct fsi_master_aspeed *aspeed, uint32_t addr, result =readl (base +OPB0_FSI_DATA_R ); +trace_fsi_master_aspeed_opb_read (addr, transfer_size, result, +readl (base +OPB0_STATUS ), + reg); +/* Return error when poll timed out */ if (ret)return ret;
@@ -196,6 +205,19 @@ static int check_errors(struct fsi_master_aspeed *aspeed, int err) {int ret; +if (trace_fsi_master_aspeed_opb_error_enabled ()) { +__be32 mresp0, mstap0, mesrb0; + +opb_readl (aspeed, ctrl_base +FSI_MRESP0 , &mresp0); +opb_readl (aspeed, ctrl_base +FSI_MSTAP0 , &mstap0); +opb_readl (aspeed, ctrl_base +FSI_MESRB0 , &mesrb0); + +trace_fsi_master_aspeed_opb_error ( +be32_to_cpu (mresp0), +be32_to_cpu (mstap0), +be32_to_cpu (mesrb0)); + } +if (err == -EIO ) {/* Check MAEB (0x70) ? */
include/trace/events/fsi_master_aspeed.h new file +77 -0
77 lines not on this page · raw message
Signature · 1 line
2.24.0.rc1
Joel Stanley
maintainerPATCH v2 11/11 fsi: aspeed: Fix OPB0 byte order register values
In reply to Joel StanleyFrom: Andrew Jeffery <[email protected]>
The data byte order selection registers in the APB2OPB primarily expose some internal plumbing necessary to get correct write accesses onto the OPB. OPB write cycles require "data mirroring" across the 32-bit data bus to support variable data width slaves that don't implement "byte enables". For slaves that do implement byte enables the master can signal which bytes on the data bus the slave should consider valid.
The data mirroring behaviour is specified by the following table:
+-----------------+----------+-----------------------------------+
| | | 32-bit Data Bus |
+---------+-------+----------+---------+---------+-------+-------+
| | | | | | | |
| ABus | Mn_BE | Request | Dbus | Dbus | Dbus | Dbus |
| (30:31) | (0:3) | Transfer | 0:7 | 8:15 | 16:23 | 24:31 |
| | | Size | byte0 | byte1 | byte2 | byte3 |
+---------+-------+----------+---------+---------+-------+-------+
| 00 | 1111 | fullword | byte0 | byte1 | byte2 | byte3 |
+---------+-------+----------+---------+---------+-------+-------+
| 00 | 1110 | halfword | byte0 | byte1 | byte2 | |
+---------+-------+----------+---------+---------+-------+-------+
| 01 | 0111 | byte | _byte1_ | byte1 | byte2 | byte3 |
+---------+-------+----------+---------+---------+-------+-------+
| 00 | 1100 | halfword | byte0 | byte1 | | |
+---------+-------+----------+---------+---------+-------+-------+
| 01 | 0110 | byte | _byte1_ | byte1 | byte2 | |
+---------+-------+----------+---------+---------+-------+-------+
| 10 | 0011 | halfword | _byte2_ | _byte3_ | byte2 | byte3 |
+---------+-------+----------+---------+---------+-------+-------+
| 00 | 1000 | byte | byte0 | | | |
+---------+-------+----------+---------+---------+-------+-------+
| 01 | 0100 | byte | _byte1_ | byte1 | | |
+---------+-------+----------+---------+---------+-------+-------+
| 10 | 0010 | byte | _byte2_ | | byte2 | |
+---------+-------+----------+---------+---------+-------+-------+
| 11 | 0001 | byte | _byte3_ | _byte3_ | | byte3 |
+---------+-------+----------+---------+---------+-------+-------+Mirrored data values are highlighted by underscores in the Dbus columns. The values in the ABus and Request Transfer Size columns correspond to values in the field names listed in the write data order select register descriptions.
Similar configuration registers are exposed for reads which enables the secondary purpose of configuring hardware endian conversions. It appears the data bus byte order is switched around in hardware so set the registers such that we can access the correct values for all widths. The values were determined by experimentation on hardware against fixed CFAM register values to configure the read data order, then in combination with the table above and the register layout documentation in the AST2600 datasheet performing write/read cycles to configure the write data order registers.
--- drivers/fsi/fsi-master-aspeed.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)
drivers/fsi/fsi-master-aspeed.c +3 -3
@@ -459,11 +459,11 @@ static int fsi_master_aspeed_probe(struct platform_device *pdev)writel (fsi_base, aspeed->base +OPB_FSI_BASE );/* Set read data order */ -writel (0x0011bb1b , aspeed->base +OPB0_READ_ORDER1 ); +writel (0x00030b1b , aspeed->base +OPB0_READ_ORDER1 );/* Set write data order */ -writel (0x0011bb1b , aspeed->base +OPB0_WRITE_ORDER1 ); -writel (0xffaa5500 , aspeed->base +OPB0_WRITE_ORDER2 ); +writel (0x0011101b , aspeed->base +OPB0_WRITE_ORDER1 ); +writel (0x0c330f3f , aspeed->base +OPB0_WRITE_ORDER2 );/* * Select OPB0 for all operations.
Signature · 1 line
2.24.0.rc1
Re: [PATCH v2 11/11] fsi: aspeed: Fix OPB0 byte order register values
OPB data mirroring is pretty special, glad someone has figured it out and made some artwork in the process.
drivers/fsi/fsi-master-
Quoted 62 lines · AEDT Joel Stanley
From: Andrew Jeffery <[email protected]>
The data byte order selection registers in the APB2OPB primarily expose some internal plumbing necessary to get correct write accesses onto the OPB. OPB write cycles require "data mirroring" across the 32-bit data bus to support variable data width slaves that don't implement "byte enables". For slaves that do implement byte enables the master can signal which bytes on the data bus the slave should consider valid.
The data mirroring behaviour is specified by the following table:
+-----------------+----------+-----------------------------------+ | | | 32-bit Data Bus | +---------+-------+----------+---------+---------+-------+-------+ | | | | | | | | | ABus | Mn_BE | Request | Dbus | Dbus | Dbus | Dbus | | (30:31) | (0:3) | Transfer | 0:7 | 8:15 | 16:23 | 24:31 | | | | Size | byte0 | byte1 | byte2 | byte3 | +---------+-------+----------+---------+---------+-------+-------+ | 00 | 1111 | fullword | byte0 | byte1 | byte2 | byte3 | +---------+-------+----------+---------+---------+-------+-------+ | 00 | 1110 | halfword | byte0 | byte1 | byte2 | | +---------+-------+----------+---------+---------+-------+-------+ | 01 | 0111 | byte | _byte1_ | byte1 | byte2 | byte3 | +---------+-------+----------+---------+---------+-------+-------+ | 00 | 1100 | halfword | byte0 | byte1 | | | +---------+-------+----------+---------+---------+-------+-------+ | 01 | 0110 | byte | _byte1_ | byte1 | byte2 | | +---------+-------+----------+---------+---------+-------+-------+ | 10 | 0011 | halfword | _byte2_ | _byte3_ | byte2 | byte3 | +---------+-------+----------+---------+---------+-------+-------+ | 00 | 1000 | byte | byte0 | | | | +---------+-------+----------+---------+---------+-------+-------+ | 01 | 0100 | byte | _byte1_ | byte1 | | | +---------+-------+----------+---------+---------+-------+-------+ | 10 | 0010 | byte | _byte2_ | | byte2 | | +---------+-------+----------+---------+---------+-------+-------+ | 11 | 0001 | byte | _byte3_ | _byte3_ | | byte3 | +---------+-------+----------+---------+---------+-------+-------+Mirrored data values are highlighted by underscores in the Dbus columns. The values in the ABus and Request Transfer Size columns correspond to values in the field names listed in the write data order select register descriptions.
Similar configuration registers are exposed for reads which enables the secondary purpose of configuring hardware endian conversions. It appears the data bus byte order is switched around in hardware so set the registers such that we can access the correct values for all widths. The values were determined by experimentation on hardware against fixed CFAM register values to configure the read data order, then in combination with the table above and the register layout documentation in the AST2600 datasheet performing write/read cycles to configure the write data order registers.
Signed-off-by Andrew Jeffery <[email protected]>Signed-off-by Joel Stanley <[email protected]>--- drivers/fsi/fsi-master-aspeed.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)drivers/fsi/fsi-master- renamed from drivers/fsi/fsi-master-aspeed.c +0
-0
aspeed.c
drivers/fsi/fsi-master-aspeed.c static int fsi_master_aspeed_probe(struct
Quoted 4 lines
index
95e226ac78b9..f49742b310c2100644drivers/fsi/fsi-master-aspeed.c +0
-0@@ -459,11 +459,11 @@ static int fsi_master_aspeed_probe(struct
platform_device *pdev)
Quoted 15 lines
drivers/fsi/fsi-master-aspeed.c +3
-3writel (fsi_base, aspeed->base +OPB_FSI_BASE );/* Set read data order */ -writel (0x0011bb1b , aspeed->base +OPB0_READ_ORDER1 ); +writel (0x00030b1b , aspeed->base +OPB0_READ_ORDER1 );/* Set write data order */ -writel (0x0011bb1b , aspeed->base +OPB0_WRITE_ORDER1 ); -writel (0xffaa5500 , aspeed->base +OPB0_WRITE_ORDER2 ); +writel (0x0011101b , aspeed->base +OPB0_WRITE_ORDER1 ); +writel (0x0c330f3f , aspeed->base +OPB0_WRITE_ORDER2 );/* * Select OPB0 for all operations.