[PATCH v1 0/3]
cover [PATCH v1 0/3]
Here is a third [1] version of the SFI GPIO helpers. This series contains SFI GPIO helpers along with the update of GPIO library to lookup for SFI resources.
It has been tested on Medfield device on top of v3.13-rc2 + Alexandre's GPIO descriptor API update [2], Mika's ACPI gpio patches [3] and my GPIO library clean up [4].
[1] https://lkml.org/lkml/2013/6/5/316
[2] http://www.spinics.net/lists/kernel/msg1645907.html
[3] http://www.spinics.net/lists/linux-acpi/msg47572.html
[4] https://www.mail-archive.com/[email protected]/msg01515.html
Andy Shevchenko (3):
SFI: fix compiler warnings
SFI: store GPIO table and export lookup function
gpiolib: append SFI helpers for GPIO API
drivers/gpio/Kconfig | 4 ++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 +++++++++++ drivers/gpio/gpiolib.c | 3 ++ drivers/gpio/gpiolib.h | 13 +++++ drivers/sfi/Makefile | 2 +- drivers/sfi/sfi_acpi.c | 9 ++-- drivers/sfi/sfi_core.c | 10 +++- drivers/sfi/sfi_core.h | 3 ++ drivers/sfi/sfi_gpio.c | 123 +++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi.h | 8 +++ 11 files changed, 197 insertions(+), 7 deletions(-) create mode 100644 drivers/gpio/gpiolib-sfi.c create mode 100644 drivers/sfi/sfi_gpio.c
Signature · 1 line
1.8.4.4
PATCH 1/3 SFI: fix compiler warnings
This fixes the following compiler warnings when build is done by make W=1.
drivers/sfi/sfi_acpi.c:154:5: warning: no previous prototype for ‘sfi_acpi_table_parse’ [-Wmissing-prototypes]
drivers/sfi/sfi_core.c:164:26: warning: no previous prototype for ‘sfi_map_table’ [-Wmissing-prototypes] drivers/sfi/sfi_core.c:192:6: warning: no previous prototype for ‘sfi_unmap_table’ [-Wmissing-prototypes]
Indentation fixes are also included.
--- drivers/sfi/sfi_acpi.c | 9 +++++---- drivers/sfi/sfi_core.c | 4 ++-- 2 files changed, 7 insertions(+), 6 deletions(-)
drivers/sfi/sfi_acpi.c +5 -4
@@ -63,6 +63,8 @@#include <acpi/acpi.h> #include <linux/sfi.h> +#include <linux/sfi_acpi.h> +#include "sfi_core.h" /*
@@ -152,7 +154,7 @@ static void sfi_acpi_put_table(struct acpi_table_header *table)* Find specified table in XSDT, run handler on it and return its return value */ int sfi_acpi_table_parse (char *signature,char *oem_id,char *oem_table_id, -int (*handler)(struct acpi_table_header *)) +int (*handler)(struct acpi_table_header *)) {struct acpi_table_header *table =NULL ;struct sfi_table_key key;
@@ -175,8 +177,8 @@ int sfi_acpi_table_parse(char *signature, char *oem_id, char *oem_table_id, }static ssize_t sfi_acpi_table_show (struct file *filp,struct kobject *kobj, -struct bin_attribute *bin_attr,char *buf, -loff_t offset,size_t count) +struct bin_attribute *bin_attr,char *buf, +loff_t offset,size_t count) {struct sfi_table_attr *tbl_attr = container_of(bin_attr, struct sfi_table_attr, attr);
@@ -199,7 +201,6 @@ static ssize_t sfi_acpi_table_show(struct file *filp, struct kobject *kobj,return cnt; } -void __init sfi_acpi_sysfs_init (void ) {u32 tbl_cnt, i;
drivers/sfi/sfi_core.c +2 -2
@@ -161,7 +161,7 @@ static int sfi_verify_table(struct sfi_table_header *table)* Check for common case that we can re-use mapping to SYST, * which requires syst_pa, syst_va to be initialized. */ -struct sfi_table_header *sfi_map_table (u64 pa) +static struct sfi_table_header *sfi_map_table (u64 pa) {struct sfi_table_header *th;u32 length;
@@ -189,7 +189,7 @@ struct sfi_table_header *sfi_map_table(u64 pa)* Undoes effect of sfi_map_table() by unmapping table * if it did not completely fit on same page as SYST. */ -void sfi_unmap_table (struct sfi_table_header *th) +static void sfi_unmap_table (struct sfi_table_header *th) {if (!TABLE_ON_PAGE (syst_va, th, th->len)) sfi_unmap_memory(th, TABLE_ON_PAGE(th, th, th->len) ?
Signature · 6 lines
1.8.4.4 -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH v1 1/3] SFI: fix compiler warnings
Quoted 10 lines · Andy Shevchenko
This fixes the following compiler warnings when build is done by make W=1.
drivers/sfi/sfi_acpi.c:154:5: warning: no previous prototype for ‘sfi_acpi_table_parse’ [-Wmissing-prototypes]
drivers/sfi/sfi_core.c:164:26: warning: no previous prototype for ‘sfi_map_table’ [-Wmissing-prototypes]drivers/sfi/sfi_core.c:192:6: warning: no previous prototype for ‘sfi_unmap_table’ [-Wmissing-prototypes]Indentation fixes are also included.
Signed-off-by Andy Shevchenko <[email protected]>
IMO this patch could be accepted regardless of this series feedback. It's a simple fix without much dependence with the other patches.
Br, David
--
To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Re: [PATCH v1 1/3] SFI: fix compiler warnings
Quoted 8 lines · David Cohen
Quoted 7 lines · Andy Shevchenko
This fixes the following compiler warnings when build is done by make W=1.
drivers/sfi/sfi_acpi.c:154:5: warning: no previous prototype for ‘sfi_acpi_table_parse’ [-Wmissing-prototypes]
drivers/sfi/sfi_core.c:164:26: warning: no previous prototype for ‘sfi_map_table’ [-Wmissing-prototypes]drivers/sfi/sfi_core.c:192:6: warning: no previous prototype for ‘sfi_unmap_table’ [-Wmissing-prototypes]
Quoted 6 lines
Quoted 3 lines
Indentation fixes are also included.
Signed-off-by Andy Shevchenko <[email protected]>IMO this patch could be accepted regardless of this series feedback. It's a simple fix without much dependence with the other patches.
Yes, but it was ignored all times. Do you know a maintainer who can upstream this one?
Signature · 7 lines
Andy Shevchenko <[email protected]> Intel Finland Oy -- To unsubscribe from this list: send the line "unsubscribe linux-gpio" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
PATCH 2/3 SFI: store GPIO table and export lookup function
In reply to Andy ShevchenkoWe have to provide a mechanism to retrive GPIO information from SFI. For this we store SFI GPIO table and provide the lookup function sfi_gpio_get_entry_by_name() that will be used later in GPIO framework.
--- drivers/sfi/Makefile | 2 +- drivers/sfi/sfi_core.c | 6 +++ drivers/sfi/sfi_core.h | 3 ++ drivers/sfi/sfi_gpio.c | 123 +++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi.h | 8 ++++ 5 files changed, 141 insertions(+), 1 deletion(-) create mode 100644 drivers/sfi/sfi_gpio.c
drivers/sfi/Makefile +1 -1
@@ -1,3 +1,3 @@ obj-y += sfi_acpi.o obj-y += sfi_core.o - +obj-y += sfi_gpio.o
drivers/sfi/sfi_core.c +6 -0
@@ -512,6 +512,12 @@ void __init sfi_init_late(void) syst_va =sfi_map_memory (syst_pa, length);sfi_acpi_init (); + +/* +* Parsing GPIO table first, since the DEVS table will need this table +* to map the pin name to the actual pin. +*/ +sfi_gpio_init (); }/*
drivers/sfi/sfi_core.h +3 -0
@@ -79,3 +79,6 @@ struct sfi_table_header *sfi_get_table(struct sfi_table_key *key);extern void sfi_put_table (struct sfi_table_header *table);extern struct sfi_table_attr __init *sfi_sysfs_install_table (u64 pa);extern void __init sfi_acpi_sysfs_init (void ); + +/* sfi_gpio.c */ +int sfi_gpio_init (void );
drivers/sfi/sfi_gpio.c new file +123 -0
123 lines not on this page · raw message
include/linux/sfi.h +8 -0
@@ -185,6 +185,8 @@ static inline void disable_sfi(void) sfi_disabled = 1; } +struct sfi_gpio_table_entry *sfi_gpio_get_entry_by_name (const char *name); +#else /* !CONFIG_SFI */ static inline void sfi_init (void )
@@ -204,6 +206,12 @@ static inline int sfi_table_parse(char *signature, char *oem_id,return -1 ; } +static inline +struct sfi_gpio_table_entry *sfi_gpio_get_entry_by_name (const char *name) +{ +return NULL ; +} +#endif /* !CONFIG_SFI */ #endif /*_LINUX_SFI_H*/
Signature · 1 line
1.8.4.4
Re: [PATCH v1 2/3] SFI: store GPIO table and export lookup function
Hi Andy,
Quoted 127 lines · Andy Shevchenko
We have to provide a mechanism to retrive GPIO information from SFI. For this we store SFI GPIO table and provide the lookup function
sfi_gpio_get_entry_by_name()that will be used later in GPIO framework.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/sfi/Makefile | 2 +- drivers/sfi/sfi_core.c | 6 +++ drivers/sfi/sfi_core.h | 3 ++ drivers/sfi/sfi_gpio.c | 123 +++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi.h | 8 ++++ 5 files changed, 141 insertions(+), 1 deletion(-) create mode 100644 drivers/sfi/sfi_gpio.cdrivers/sfi/Makefile +1
-1@@ -1,3 +1,3 @@ obj-y += sfi_acpi.o obj-y += sfi_core.o - +obj-y += sfi_gpio.odrivers/sfi/sfi_core.c +6
-0@@ -512,6 +512,12 @@ void __init sfi_init_late(void) syst_va =sfi_map_memory (syst_pa, length);sfi_acpi_init (); + +/* +* Parsing GPIO table first, since the DEVS table will need this table +* to map the pin name to the actual pin. +*/ +sfi_gpio_init (); }/* drivers/sfi/sfi_core.h +3
-0@@ -79,3 +79,6 @@ struct sfi_table_header *sfi_get_table(struct sfi_table_key *key);extern void sfi_put_table (struct sfi_table_header *table);extern struct sfi_table_attr __init *sfi_sysfs_install_table (u64 pa);extern void __init sfi_acpi_sysfs_init (void ); + +/* sfi_gpio.c */ +int sfi_gpio_init (void );drivers/sfi/sfi_gpio.c new file +70
-0@@ -0,0 +1,123 @@ +/* sfi_gpio.c Simple Firmware Interface - GPIO extensions */ + +/* + +This file is provided under a dual BSD/GPLv2 license. When using or +redistributing this file, you may do so under either license. + +GPL LICENSE SUMMARY + +Copyright(c) 2013 Intel Corporation. All rights reserved. + +This program is free software; you can redistribute it and/or modify +it under the terms of version 2 of the GNU General Public License as +published by the Free Software Foundation. + +This program is distributed in the hope that it will be useful, but +WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; if not, write to the Free Software +Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. +The full GNU General Public License is included in this distribution +in the file called LICENSE.GPL. + +BSD LICENSE + +Copyright(c) 2013 Intel Corporation. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions +are met: + +* Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. +* Redistributions in binary form must reproduce the above copyright +notice, this list of conditions and the following disclaimer in +the documentation and/or other materials provided with the +distribution. +* Neither the name of Intel Corporation nor the names of its +contributors may be used to endorse or promote products derived +from this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +*/ + +#define KMSG_COMPONENT "SFI" +#define pr_fmt (fmt) KMSG_COMPONENT ": " fmt + +#include <linux/errno.h> +#include <linux/kernel.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/export.h> +#include <linux/sfi.h> + +#include "sfi_core.h" +
drivers/sfi/sfi_gpio.c line 80
Quoted 10 lines
drivers/sfi/sfi_gpio.c new file +10
-0+static struct sfi_gpio_table_entry *sfi_gpio_table; +static int sfi_gpio_num_entry; + +struct sfi_gpio_table_entry *sfi_gpio_get_entry_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry = sfi_gpio_table; +int i; + +if (!pentry) +return NULL ;
If this function is called prior to sfi_gpio_table initialization, it will have
same result as if the gpio name doesn't exist.
For development sanity, how about return ERR_PTR(-EBUSY) or ERR_PTR(-EAGAIN) if it's too early?
Quoted 8 lines
drivers/sfi/sfi_gpio.c new file +8
-0+ +for (i =0 ; i < sfi_gpio_num_entry; i++, pentry++) { +if (!strncmp (name, pentry->pin_name,SFI_NAME_LEN )) +return pentry; + } + +return NULL ; +}
drivers/sfi/sfi_gpio.c line 98
Quoted 10 lines
drivers/sfi/sfi_gpio.c new file +10
-0+EXPORT_SYMBOL_GPL (sfi_gpio_get_entry_by_name); + +static int __init sfi_gpio_parse (struct sfi_table_header *table) +{ +struct sfi_table_simple *sb; +struct sfi_gpio_table_entry *pentry; +int num, i; + +if (sfi_gpio_table) +return -EBUSY ;
It can't ever happen. This is a static function called once in this file. IMO kernel doesn't need such kind of tests :)
Quoted 39 lines
drivers/sfi/sfi_gpio.c new file +25
-0+ + sb = container_of(table, struct sfi_table_simple, header); + + num = SFI_GET_NUM_ENTRIES(sb, struct sfi_gpio_table_entry); + pentry = (struct sfi_gpio_table_entry *)sb->pentry; + + sfi_gpio_table =kmemdup (pentry, num *sizeof (*pentry),GFP_KERNEL ); +if (!sfi_gpio_table) +return -ENOMEM ; + + sfi_gpio_num_entry = num; + +pr_debug ("GPIO pin info:\n" ); +for (i =0 ; i < num; i++, pentry++) +pr_debug ("GPIO [%2d] chip = %16.16s, name = %16.16s, pin=%d\n" , + i, pentry->controller_name, pentry->pin_name, + pentry->pin_no); + +return 0 ; +} + +int __init sfi_gpio_init (void ) +{ +return sfi_table_parse (SFI_SIG_GPIO ,NULL ,NULL , sfi_gpio_parse); +}include/linux/sfi.h +2
-0@@ -185,6 +185,8 @@ static inline void disable_sfi(void) sfi_disabled = 1; } +struct sfi_gpio_table_entry *sfi_gpio_get_entry_by_name (const char *name); +#else /* !CONFIG_SFI */ static inline void sfi_init (void )
include/linux/sfi.h static inline int sfi_table_parse(char *signature, char *oem_id, · line 212
Quoted 7 lines
include/linux/sfi.h +4
-0@@ -204,6 +206,12 @@ static inline int sfi_table_parse(char *signature, char *oem_id,return -1 ; } +static inline +struct sfi_gpio_table_entry *sfi_gpio_get_entry_by_name (const char *name) +{ +return NULL ;
I'd suggest something more meaningful like ERR_PTR(-ENODEV). Again, for development sanity.
Br, David Cohen
Quoted 5 lines
include/linux/sfi.h +2
-0+} +#endif /* !CONFIG_SFI */ #endif /*_LINUX_SFI_H*/
Re: [PATCH v1 2/3] SFI: store GPIO table and export lookup function
Quoted 138 lines · David Cohen
Hi Andy,
Quoted 135 lines · Andy Shevchenko
We have to provide a mechanism to retrive GPIO information from SFI. For this we store SFI GPIO table and provide the lookup function
sfi_gpio_get_entry_by_name()that will be used later in GPIO framework.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/sfi/Makefile | 2 +- drivers/sfi/sfi_core.c | 6 +++ drivers/sfi/sfi_core.h | 3 ++ drivers/sfi/sfi_gpio.c | 123 +++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi.h | 8 ++++ 5 files changed, 141 insertions(+), 1 deletion(-) create mode 100644 drivers/sfi/sfi_gpio.cdrivers/sfi/Makefile +1
-1@@ -1,3 +1,3 @@ obj-y += sfi_acpi.o obj-y += sfi_core.o - +obj-y += sfi_gpio.odrivers/sfi/sfi_core.c +6
-0@@ -512,6 +512,12 @@ void __init sfi_init_late(void) syst_va =sfi_map_memory (syst_pa, length);sfi_acpi_init (); + +/* +* Parsing GPIO table first, since the DEVS table will need this table +* to map the pin name to the actual pin. +*/ +sfi_gpio_init (); }/* drivers/sfi/sfi_core.h +3
-0@@ -79,3 +79,6 @@ struct sfi_table_header *sfi_get_table(struct sfi_table_key *key);extern void sfi_put_table (struct sfi_table_header *table);extern struct sfi_table_attr __init *sfi_sysfs_install_table (u64 pa);extern void __init sfi_acpi_sysfs_init (void ); + +/* sfi_gpio.c */ +int sfi_gpio_init (void );drivers/sfi/sfi_gpio.c new file +78
-0@@ -0,0 +1,123 @@ +/* sfi_gpio.c Simple Firmware Interface - GPIO extensions */ + +/* + +This file is provided under a dual BSD/GPLv2 license. When using or +redistributing this file, you may do so under either license. + +GPL LICENSE SUMMARY + +Copyright(c) 2013 Intel Corporation. All rights reserved. + +This program is free software; you can redistribute it and/or modify +it under the terms of version 2 of the GNU General Public License as +published by the Free Software Foundation. + +This program is distributed in the hope that it will be useful, but +WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU +General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; if not, write to the Free Software +Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. +The full GNU General Public License is included in this distribution +in the file called LICENSE.GPL. + +BSD LICENSE + +Copyright(c) 2013 Intel Corporation. All rights reserved. + +Redistribution and use in source and binary forms, with or without +modification, are permitted provided that the following conditions +are met: + +* Redistributions of source code must retain the above copyright +notice, this list of conditions and the following disclaimer. +* Redistributions in binary form must reproduce the above copyright +notice, this list of conditions and the following disclaimer in +the documentation and/or other materials provided with the +distribution. +* Neither the name of Intel Corporation nor the names of its +contributors may be used to endorse or promote products derived +from this software without specific prior written permission. + +THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS +"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT +LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR +A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT +OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, +SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT +LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, +DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY +THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT +(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE +OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. + +*/ + +#define KMSG_COMPONENT "SFI" +#define pr_fmt (fmt) KMSG_COMPONENT ": " fmt + +#include <linux/errno.h> +#include <linux/kernel.h> +#include <linux/string.h> +#include <linux/slab.h> +#include <linux/export.h> +#include <linux/sfi.h> + +#include "sfi_core.h" + +static struct sfi_gpio_table_entry *sfi_gpio_table; +static int sfi_gpio_num_entry; + +struct sfi_gpio_table_entry *sfi_gpio_get_entry_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry = sfi_gpio_table; +int i; +
Quoted 8 lines
drivers/sfi/sfi_gpio.c line 80
Quoted 2 lines
drivers/sfi/sfi_gpio.c new file +2
-0+if (!pentry) +return NULL ;If this function is called prior to
sfi_gpio_tableinitialization, it will have
same result as if the gpio name doesn't exist.
For development sanity, how about returnERR_PTR(-EBUSY)orERR_PTR(-EAGAIN)if it's too early?
Okay, I switch to error pointer.
Quoted 14 lines
Quoted 14 lines
drivers/sfi/sfi_gpio.c new file +13
-0+ +for (i =0 ; i < sfi_gpio_num_entry; i++, pentry++) { +if (!strncmp (name, pentry->pin_name,SFI_NAME_LEN )) +return pentry; + } + +return NULL ; +} +EXPORT_SYMBOL_GPL (sfi_gpio_get_entry_by_name); + +static int __init sfi_gpio_parse (struct sfi_table_header *table) +{ +struct sfi_table_simple *sb;
Quoted 8 lines
drivers/sfi/sfi_gpio.c line 98
Quoted 5 lines
drivers/sfi/sfi_gpio.c new file +5
-0+struct sfi_gpio_table_entry *pentry; +int num, i; + +if (sfi_gpio_table) +return -EBUSY ;It can't ever happen. This is a static function called once in this file. IMO kernel doesn't need such kind of tests :)
Will fix it.
Quoted 43 lines
Quoted 43 lines
drivers/sfi/sfi_gpio.c new file +25
-0+ + sb = container_of(table, struct sfi_table_simple, header); + + num = SFI_GET_NUM_ENTRIES(sb, struct sfi_gpio_table_entry); + pentry = (struct sfi_gpio_table_entry *)sb->pentry; + + sfi_gpio_table =kmemdup (pentry, num *sizeof (*pentry),GFP_KERNEL ); +if (!sfi_gpio_table) +return -ENOMEM ; + + sfi_gpio_num_entry = num; + +pr_debug ("GPIO pin info:\n" ); +for (i =0 ; i < num; i++, pentry++) +pr_debug ("GPIO [%2d] chip = %16.16s, name = %16.16s, pin=%d\n" , + i, pentry->controller_name, pentry->pin_name, + pentry->pin_no); + +return 0 ; +} + +int __init sfi_gpio_init (void ) +{ +return sfi_table_parse (SFI_SIG_GPIO ,NULL ,NULL , sfi_gpio_parse); +}include/linux/sfi.h +2
-0@@ -185,6 +185,8 @@ static inline void disable_sfi(void) sfi_disabled = 1; } +struct sfi_gpio_table_entry *sfi_gpio_get_entry_by_name (const char *name); +#else /* !CONFIG_SFI */ static inline void sfi_init (void )@@ -204,6 +206,12 @@ static inline int sfi_table_parse(char *signature, char *oem_id,return -1 ; }
Quoted 7 lines
include/linux/sfi.h static inline int sfi_table_parse(char *signature, char *oem_id, · line 212
Quoted 4 lines
include/linux/sfi.h +4
-0+static inline +struct sfi_gpio_table_entry *sfi_gpio_get_entry_by_name (const char *name) +{ +return NULL ;I'd suggest something more meaningful like
ERR_PTR(-ENODEV). Again, for development sanity.
Okay.
Quoted 9 lines
Br, David Cohen
Quoted 5 lines
include/linux/sfi.h +2
-0+} +#endif /* !CONFIG_SFI */ #endif /*_LINUX_SFI_H*/
Signature · 2 lines
Andy Shevchenko <[email protected]> Intel Finland Oy
Re: [PATCH v1 2/3] SFI: store GPIO table and export lookup function
In reply to Andy ShevchenkoQuoted 72 lines · Andy Shevchenko
We have to provide a mechanism to retrive GPIO information from SFI. For this we store SFI GPIO table and provide the lookup function
sfi_gpio_get_entry_by_name()that will be used later in GPIO framework.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/sfi/Makefile | 2 +- drivers/sfi/sfi_core.c | 6 +++ drivers/sfi/sfi_core.h | 3 ++ drivers/sfi/sfi_gpio.c | 123 +++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi.h | 8 ++++ 5 files changed, 141 insertions(+), 1 deletion(-)create mode 100644
drivers/sfi/sfi_gpio.cdrivers/sfi/Makefile +1
-1@@ -1,3 +1,3 @@ obj-y += sfi_acpi.o obj-y += sfi_core.o - +obj-y += sfi_gpio.odrivers/sfi/sfi_core.c +6
-0@@ -512,6 +512,12 @@ void __init sfi_init_late(void) syst_va =sfi_map_memory (syst_pa, length);sfi_acpi_init (); + +/* +* Parsing GPIO table first, since the DEVS table will need this table +* to map the pin name to the actual pin. +*/ +sfi_gpio_init (); }/* drivers/sfi/sfi_core.h +3
-0@@ -79,3 +79,6 @@ struct sfi_table_header *sfi_get_table(struct sfi_table_key *key);extern void sfi_put_table (struct sfi_table_header *table);extern struct sfi_table_attr __init *sfi_sysfs_install_table (u64 pa);extern void __init sfi_acpi_sysfs_init (void ); + +/* sfi_gpio.c */ +int sfi_gpio_init (void );drivers/sfi/sfi_gpio.c new file +15
-0@@ -0,0 +1,123 @@ +/* sfi_gpio.c Simple Firmware Interface - GPIO extensions */ + +/* + +This file is provided under a dual BSD/GPLv2 license. When using or +redistributing this file, you may do so under either license. + +GPL LICENSE SUMMARY + +Copyright(c) 2013 Intel Corporation. All rights reserved. + +This program is free software; you can redistribute it and/or modify +it under the terms of version 2 of the GNU General Public License as +published by the Free Software Foundation. +
drivers/sfi/sfi_gpio.c line 25
Quoted 10 lines
drivers/sfi/sfi_gpio.c new file +10
-0+ This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + The full GNU General Public License is included in this distribution + in the file called LICENSE.GPL.
As I've been told many times in the past, this last paragraph should not be included unless you are willing to update all your patches that include it should the FSF move to another address. ;)
Otherwise I'm not familiar with SFI, but this looks ok to me.
Re: [PATCH v1 2/3] SFI: store GPIO table and export lookup function
Quoted 88 lines · Alex Courbot
Quoted 72 lines · Andy Shevchenko
We have to provide a mechanism to retrive GPIO information from SFI. For this we store SFI GPIO table and provide the lookup function
sfi_gpio_get_entry_by_name()that will be used later in GPIO framework.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/sfi/Makefile | 2 +- drivers/sfi/sfi_core.c | 6 +++ drivers/sfi/sfi_core.h | 3 ++ drivers/sfi/sfi_gpio.c | 123 +++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi.h | 8 ++++ 5 files changed, 141 insertions(+), 1 deletion(-)create mode 100644
drivers/sfi/sfi_gpio.cdrivers/sfi/Makefile +1
-1@@ -1,3 +1,3 @@ obj-y += sfi_acpi.o obj-y += sfi_core.o - +obj-y += sfi_gpio.odrivers/sfi/sfi_core.c +6
-0@@ -512,6 +512,12 @@ void __init sfi_init_late(void) syst_va =sfi_map_memory (syst_pa, length);sfi_acpi_init (); + +/* +* Parsing GPIO table first, since the DEVS table will need this table +* to map the pin name to the actual pin. +*/ +sfi_gpio_init (); }/* drivers/sfi/sfi_core.h +3
-0@@ -79,3 +79,6 @@ struct sfi_table_header *sfi_get_table(struct sfi_table_key *key);extern void sfi_put_table (struct sfi_table_header *table);extern struct sfi_table_attr __init *sfi_sysfs_install_table (u64 pa);extern void __init sfi_acpi_sysfs_init (void ); + +/* sfi_gpio.c */ +int sfi_gpio_init (void );drivers/sfi/sfi_gpio.c new file +15
-0@@ -0,0 +1,123 @@ +/* sfi_gpio.c Simple Firmware Interface - GPIO extensions */ + +/* + +This file is provided under a dual BSD/GPLv2 license. When using or +redistributing this file, you may do so under either license. + +GPL LICENSE SUMMARY + +Copyright(c) 2013 Intel Corporation. All rights reserved. + +This program is free software; you can redistribute it and/or modify +it under the terms of version 2 of the GNU General Public License as +published by the Free Software Foundation. +drivers/sfi/sfi_gpio.c line 25
Quoted 10 lines
drivers/sfi/sfi_gpio.c new file +10
-0+ This program is distributed in the hope that it will be useful, but + WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 51 Franklin St - Fifth Floor, Boston, MA 02110-1301 USA. + The full GNU General Public License is included in this distribution + in the file called LICENSE.GPL.As I've been told many times in the past, this last paragraph should not be included unless you are willing to update all your patches that include it should the FSF move to another address. ;)
Quoted 1 line
Otherwise I'm not familiar with SFI, but this looks ok to me.
I have just copied the top from other file in that folder (to keep licensing the same). So, it probably requires to be updated in all related files at once.
Meanwhile I'll prepare that patch and reduce this one.
Signature · 2 lines
Andy Shevchenko <[email protected]> Intel Finland Oy
PATCH 3/3 gpiolib: append SFI helpers for GPIO API
In reply to Andy ShevchenkoTo support some (legacy) firmwares and platforms let's make life easier for their customers.
This patch provides a function which converts sfi_gpio_table_entry to gpio_desc. The use of it is integrated into GPIO library to enable generic access to the SFI GPIO resources.
--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+) create mode 100644 drivers/gpio/gpiolib-sfi.c
drivers/gpio/Kconfig +4 -0
@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPI
drivers/gpio/Makefile +1 -0
@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabetically
drivers/gpio/gpiolib-sfi.c new file +28 -0
@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +*/ + +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> + +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV ); + +return gpio_to_desc (pentry->pin_no); +}
drivers/gpio/gpiolib.c +3 -0
@@ -2451,6 +2451,9 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, }else if (IS_ENABLED (CONFIG_ACPI ) && dev &&ACPI_HANDLE (dev)) {dev_dbg (dev,"using ACPI for GPIO lookup\n" ); desc =acpi_find_gpio (dev, con_id, idx, &flags); + }else if (IS_ENABLED (CONFIG_SFI )) { +dev_dbg (dev,"using SFI for GPIO lookup\n" ); + desc =sfi_get_gpiod_by_name (con_id); }/*
drivers/gpio/gpiolib.h +13 -0
@@ -43,4 +43,17 @@ acpi_get_gpiod_by_index(struct device *dev, int index, }#endif +#ifdef CONFIG_GPIO_SFI + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name); + +#else /* !CONFIG_GPIO_SFI */ + +static inline struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +return ERR_PTR (-ENODEV ); +} + +#endif /* !CONFIG_GPIO_SFI */ +#endif /* GPIOLIB_H */
Signature · 1 line
1.8.4.4
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Quoted 65 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for their customers.
This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+) create mode 100644 drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +15
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +*/ + +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h>
drivers/gpio/gpiolib-sfi.c line 25
Quoted 10 lines
drivers/gpio/gpiolib-sfi.c new file +10
-0+ +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV );
With my comments on patch 2/3, I'd suggest here:
if (!pentry)
return ERR_PTR(-EINVAL);
if (IS_ERR(pentry))
return pentry;!pentry == could not findIS_ERR(pentry) == something else went wrong
Br, David
Quoted 4 lines
drivers/gpio/gpiolib-sfi.c new file +3
-0+ +return gpio_to_desc (pentry->pin_no); +}
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Quoted 79 lines · David Cohen
Quoted 65 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for their customers.
This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+) create mode 100644 drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +15
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +*/ + +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> drivers/gpio/gpiolib-sfi.c line 25
Quoted 10 lines
drivers/gpio/gpiolib-sfi.c new file +10
-0+ +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV );With my comments on patch 2/3, I'd suggest here:
if (!pentry)
Quoted 1 line
return ERR_PTR(-EINVAL);
It should never be here. We have to avoid tristate return pointers.
Quoted 2 lines
if (
IS_ERR(pentry))return pentry;
This one correct, I agree.
Quoted 1 line
!pentry == could not find
Rather -ENOENT.
Quoted 10 lines
IS_ERR(pentry)== something else went wrongBr, David
Quoted 4 lines
drivers/gpio/gpiolib-sfi.c new file +3
-0+ +return gpio_to_desc (pentry->pin_no); +}
Signature · 2 lines
Andy Shevchenko <[email protected]> Intel Finland Oy
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
In reply to Andy ShevchenkoQuoted 83 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for their customers.
This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+)create mode 100644
drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +28
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +*/ + +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> + +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV ); + +return gpio_to_desc (pentry->pin_no); +}drivers/gpio/gpiolib.c +0
-0
drivers/gpio/gpiolib.c struct gpio_desc *__must_check gpiod_get_index(struct device *dev, · line 2456
Quoted 6 lines
drivers/gpio/gpiolib.c +3
-0@@ -2451,6 +2451,9 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, }else if (IS_ENABLED (CONFIG_ACPI ) && dev &&ACPI_HANDLE (dev)) {dev_dbg (dev,"using ACPI for GPIO lookup\n" ); desc =acpi_find_gpio (dev, con_id, idx, &flags); + }else if (IS_ENABLED (CONFIG_SFI )) { +dev_dbg (dev,"using SFI for GPIO lookup\n" ); + desc =sfi_get_gpiod_by_name (con_id);
Your lookup function is ignoring the dev argument. Are SFI GPIOs always supposed to be system-global? In this case, your if condition should likely be
} else if (IS_ENABLED(CONFIG_SFI) && !dev) {So that a global SFI GPIO does not get mistakenly assigned to a device that has, say, a more suited platform mapping on the same con_id.
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Quoted 85 lines · Alex Courbot
Quoted 84 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for their customers.
This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+)create mode 100644
drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +28
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +*/ + +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> + +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV ); + +return gpio_to_desc (pentry->pin_no); +}drivers/gpio/gpiolib.c +0
-0@@ -2451,6 +2451,9 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, }else if (IS_ENABLED (CONFIG_ACPI ) && dev &&ACPI_HANDLE (dev)) {
Quoted 8 lines
drivers/gpio/gpiolib.c struct gpio_desc *__must_check gpiod_get_index(struct device *dev, · line 2456
Quoted 5 lines
drivers/gpio/gpiolib.c +3
-0dev_dbg (dev,"using ACPI for GPIO lookup\n" ); desc =acpi_find_gpio (dev, con_id, idx, &flags); + }else if (IS_ENABLED (CONFIG_SFI )) { +dev_dbg (dev,"using SFI for GPIO lookup\n" ); + desc =sfi_get_gpiod_by_name (con_id);Your lookup function is ignoring the dev argument. Are SFI GPIOs always supposed to be system-global?
It's not clear. It could be device related, though SFI itself has probably wrong design. I rather prefer to avoid a dev parameter check at all.
Quoted 7 lines
In this case, your if condition should
likely be} else if (IS_ENABLED(CONFIG_SFI) && !dev) {So that a global SFI GPIO does not get mistakenly assigned to a device that has, say, a more suited platform mapping on the same
con_id.
So, for example in the driver that could be enumerated from SFI, DT, and via platform data you suggest to have something like
desc = gpiod_get(dev, "con_id_device_tree");
...
if (IS_ERR(desc))
desc = gpiod_get(NULL, "con_id_sfi");
if (IS_ERR(desc))
desc = gpiod_get(???, "con_id_from_platdata");
Correct?
Quoted 1 line
From my point of view, the SFI, platform data, and DT cases should have
the same con_id name. And I don't see a problem here, if we use dev parameter != NULL for SFI case. Since you proposed to distinguish no entity case vs. others, it would be easy to return it from SFI gpio helper.
Signature · 2 lines
Andy Shevchenko <[email protected]> Intel Finland Oy
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Quoted 114 lines · Andy Shevchenko
Quoted 85 lines · Alex Courbot
Quoted 84 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for their customers.
This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+)create mode 100644
drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +28
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +*/ + +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> + +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV ); + +return gpio_to_desc (pentry->pin_no); +}drivers/gpio/gpiolib.c +0
-0@@ -2451,6 +2451,9 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, }else if (IS_ENABLED (CONFIG_ACPI ) && dev &&ACPI_HANDLE (dev)) {Quoted 8 lines
drivers/gpio/gpiolib.c struct gpio_desc *__must_check gpiod_get_index(struct device *dev, · line 2456
Quoted 5 lines
drivers/gpio/gpiolib.c +3
-0dev_dbg (dev,"using ACPI for GPIO lookup\n" ); desc =acpi_find_gpio (dev, con_id, idx, &flags); + }else if (IS_ENABLED (CONFIG_SFI )) { +dev_dbg (dev,"using SFI for GPIO lookup\n" ); + desc =sfi_get_gpiod_by_name (con_id);Your lookup function is ignoring the dev argument. Are SFI GPIOs always supposed to be system-global?
It's not clear. It could be device related, though SFI itself has probably wrong design. I rather prefer to avoid a dev parameter check at all.
Quoted 7 lines
In this case, your if condition should
likely be} else if (IS_ENABLED(CONFIG_SFI) && !dev) {So that a global SFI GPIO does not get mistakenly assigned to a device that has, say, a more suited platform mapping on the same
con_id.So, for example in the driver that could be enumerated from SFI, DT, and via platform data you suggest to have something like
desc =
gpiod_get(dev, "con_id_device_tree");
...
Quoted 7 lines
if (
IS_ERR(desc))
desc =gpiod_get(NULL, "con_id_sfi");if (
IS_ERR(desc))
desc =gpiod_get(???, "con_id_from_platdata");Correct?
No, this is not what I'm suggesting. Device drivers should not care who provides the GPIO, they should just ask for it, and obtain it (or not).
The scope of the problem actually depends on what SFI GPIOs are used for. For instance, let's say you have two devices each using an enabling GPIO which happens to be provided by SFI. Both drivers for these devices obtain the enable GPIO as follows:
enable_gpio = gpiod_get(dev, "enable");
Here you actually have two problems:
1) Since you only look for con_id, how to you discriminate the enable GPIOs for these devices?
2) The device drivers are the one to decide which GPIO name they request. How are SFI GPIO names decided? If you don't have any kind of flexibility for their naming, and want to use them with devices drivers, you will very likely need another naming layer that associates a (device_name, con_id) pair to the right SFI GPIO, similarly to what is done with platform GPIOs. If the only consumer of SFI GPIOs is platform code, and SFI GPIOs are all uniquely named, then you may as well request the device to be NULL for their lookup so that they don't interfere with more precisely-mapped GPIOs.
I don't know anything about SFI GPIOs, how they are defined, where their name comes from, and how they are used so my vision may be incomplete. But AFAICT it all comes down to one of these two scenarios:
1) SFI GPIOs are only used in platform code -> using their pin name is ok, device argument should be assumed to be NULL for their matching
2) SFI GPIOs are also consumed by device drivers -> you need a way to match a (device, con_id) pair to your SFI GPIOs so they can be matched exactly and through the names drivers will request.
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Quoted 126 lines · Alex Courbot
Quoted 114 lines · Andy Shevchenko
Quoted 85 lines · Alex Courbot
Quoted 84 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for their customers.
This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+)create mode 100644
drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +28
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +*/ + +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> + +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV ); + +return gpio_to_desc (pentry->pin_no); +}drivers/gpio/gpiolib.c +0
-0@@ -2451,6 +2451,9 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, }else if (IS_ENABLED (CONFIG_ACPI ) && dev &&ACPI_HANDLE (dev)) {Quoted 8 lines
drivers/gpio/gpiolib.c struct gpio_desc *__must_check gpiod_get_index(struct device *dev, · line 2456
Quoted 5 lines
drivers/gpio/gpiolib.c +3
-0dev_dbg (dev,"using ACPI for GPIO lookup\n" ); desc =acpi_find_gpio (dev, con_id, idx, &flags); + }else if (IS_ENABLED (CONFIG_SFI )) { +dev_dbg (dev,"using SFI for GPIO lookup\n" ); + desc =sfi_get_gpiod_by_name (con_id);Your lookup function is ignoring the dev argument. Are SFI GPIOs always supposed to be system-global?
It's not clear. It could be device related, though SFI itself has probably wrong design. I rather prefer to avoid a dev parameter check at all.
Quoted 7 lines
In this case, your if condition should
likely be} else if (IS_ENABLED(CONFIG_SFI) && !dev) {So that a global SFI GPIO does not get mistakenly assigned to a device that has, say, a more suited platform mapping on the same
con_id.So, for example in the driver that could be enumerated from SFI, DT, and via platform data you suggest to have something like
desc =
gpiod_get(dev, "con_id_device_tree");
...Quoted 7 lines
if (
IS_ERR(desc))
desc =gpiod_get(NULL, "con_id_sfi");if (
IS_ERR(desc))
desc =gpiod_get(???, "con_id_from_platdata");Correct?
No, this is not what I'm suggesting. Device drivers should not care who provides the GPIO, they should just ask for it, and obtain it (or not).
Quoted 6 lines
The scope of the problem actually depends on what SFI GPIOs are used for. For instance, let's say you have two devices each using an enabling GPIO which happens to be provided by SFI. Both drivers for these devices obtain the enable GPIO as follows:
enable_gpio = gpiod_get(dev, "enable");
You rather can't do that in sfi case. Yes, the structure has (stringy) reference to gpio chip that provides a line, but in practice the gpio line name should be unique. Consider this is misdesign of SFI as I mentioned earlier.
Quoted 6 lines
Here you actually have two problems:
1) Since you only look for
con_id, how to you discriminate the enable GPIOs for these devices?
Quoted 9 lines
2) The device drivers are the one to decide which GPIO name they request. How are SFI GPIO names decided? If you don't have any kind of flexibility for their naming, and want to use them with devices drivers, you will very likely need another naming layer that associates a (
device_name,con_id) pair to the right SFI GPIO, similarly to what is done with platform GPIOs. If the only consumer of SFI GPIOs is platform code, and SFI GPIOs are all uniquely named, then you may as well request the device to be NULL for their lookup so that they don't interfere with more precisely-mapped GPIOs.
So, you insist to have !dev there?
Quoted 7 lines
I don't know anything about SFI GPIOs, how they are defined, where their name comes from, and how they are used so my vision may be incomplete. But AFAICT it all comes down to one of these two scenarios:
1) SFI GPIOs are only used in platform code -> using their pin name is ok, device argument should be assumed to be NULL for their matching
Let's stick to this.
Quoted 3 lines
2) SFI GPIOs are also consumed by device drivers -> you need a way to match a (device,
con_id) pair to your SFI GPIOs so they can be matched exactly and through the names drivers will request.
Mostly unlikely we go this way. It would mean we don't need SFI at all.
Signature · 2 lines
Andy Shevchenko <[email protected]> Intel Finland Oy
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Hi,
[snip]
Quoted 4 lines
Quoted 4 lines
I don't know anything about SFI GPIOs, how they are defined, where their name comes from, and how they are used so my vision may be incomplete. But AFAICT it all comes down to one of these two scenarios:
Quoted 8 lines
Quoted 2 lines
1) SFI GPIOs are only used in platform code -> using their pin name is ok, device argument should be assumed to be NULL for their matching
Let's stick to this.
Quoted 3 lines
2) SFI GPIOs are also consumed by device drivers -> you need a way to match a (device,
con_id) pair to your SFI GPIOs so they can be matched exactly and through the names drivers will request.Mostly unlikely we go this way. It would mean we don't need SFI at all.
I'm currently upstreaming Intel-MID's SFI users. This is a forbidden situation which I'd change before submit the patch.
Br, David Cohen
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Quoted 12 lines · David Cohen
Hi,
[snip]
Quoted 8 lines
Quoted 6 lines
I don't know anything about SFI GPIOs, how they are defined, where their name comes from, and how they are used so my vision may be incomplete. But AFAICT it all comes down to one of these two scenarios:
1) SFI GPIOs are only used in platform code -> using their pin name is ok, device argument should be assumed to be NULL for their matching
Let's stick to this.
Quoted 7 lines
Quoted 4 lines
Quoted 3 lines
2) SFI GPIOs are also consumed by device drivers -> you need a way to match a (device,
con_id) pair to your SFI GPIOs so they can be matched exactly and through the names drivers will request.Mostly unlikely we go this way. It would mean we don't need SFI at all.
I'm currently upstreaming Intel-MID's SFI users. This is a forbidden situation which I'd change before submit the patch.
I'm sorry I didn't get this. You mean we have to have another mapping, or you agree with system-global GPIOs in SFI case?
Quoted 2 lines
Br, David Cohen
Signature · 2 lines
Andy Shevchenko <[email protected]> Intel Finland Oy
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
In reply to Andy ShevchenkoQuoted 132 lines · Andy Shevchenko
Quoted 131 lines · Alex Courbot
Quoted 114 lines · Andy Shevchenko
Quoted 85 lines · Alex Courbot
Quoted 84 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for their customers.
This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+)create mode 100644
drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +28
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify +* it under the terms of the GNU General Public License version 2 as +* published by the Free Software Foundation. +*/ + +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> + +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV ); + +return gpio_to_desc (pentry->pin_no); +}drivers/gpio/gpiolib.c +0
-0@@ -2451,6 +2451,9 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, }else if (IS_ENABLED (CONFIG_ACPI ) && dev &&ACPI_HANDLE (dev)) {Quoted 8 lines
drivers/gpio/gpiolib.c struct gpio_desc *__must_check gpiod_get_index(struct device *dev, · line 2456
Quoted 5 lines
drivers/gpio/gpiolib.c +3
-0dev_dbg (dev,"using ACPI for GPIO lookup\n" ); desc =acpi_find_gpio (dev, con_id, idx, &flags); + }else if (IS_ENABLED (CONFIG_SFI )) { +dev_dbg (dev,"using SFI for GPIO lookup\n" ); + desc =sfi_get_gpiod_by_name (con_id);Your lookup function is ignoring the dev argument. Are SFI GPIOs always supposed to be system-global?
It's not clear. It could be device related, though SFI itself has probably wrong design. I rather prefer to avoid a dev parameter check at all.
Quoted 7 lines
In this case, your if condition should
likely be} else if (IS_ENABLED(CONFIG_SFI) && !dev) {So that a global SFI GPIO does not get mistakenly assigned to a device that has, say, a more suited platform mapping on the same
con_id.So, for example in the driver that could be enumerated from SFI, DT, and via platform data you suggest to have something like
desc =
gpiod_get(dev, "con_id_device_tree");
...Quoted 7 lines
if (
IS_ERR(desc))
desc =gpiod_get(NULL, "con_id_sfi");if (
IS_ERR(desc))
desc =gpiod_get(???, "con_id_from_platdata");Correct?
No, this is not what I'm suggesting. Device drivers should not care who provides the GPIO, they should just ask for it, and obtain it (or not).
The scope of the problem actually depends on what SFI GPIOs are used for. For instance, let's say you have two devices each using an enabling GPIO which happens to be provided by SFI. Both drivers for these devices obtain the enable GPIO as follows:
Quoted 6 lines
Quoted 1 line
enable_gpio = gpiod_get(dev, "enable");You rather can't do that in sfi case. Yes, the structure has (stringy) reference to gpio chip that provides a line, but in practice the gpio line name should be unique. Consider this is misdesign of SFI as I mentioned earlier.
Ok. Note that this is not necessarily a bad thing, it just means that SFI GPIOs should only be used by platform-specific code that is aware of their existence (and particular naming).
Device drivers on the other hand cannot make any assumption on who will provide the GPIO, so if you want them to be able to use SFI GPIOs, you will need an actual way to map then to (device, function) pairs. If this use-case is out of the equation, then I'm glad with the v6 of your patch.
But David raised a point that kind of sounds like he might want to use them that way, so I'm holding off my ack until this is solved.
Alex.
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Quoted 157 lines · Alex Courbot
Quoted 145 lines · Andy Shevchenko
Quoted 144 lines · Alex Courbot
Quoted 125 lines · Andy Shevchenko
Quoted 93 lines · Alex Courbot
Quoted 92 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for
their customers.This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic
access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+)create mode 100644
drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +7
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */drivers/gpio/gpiolib-sfi.c new file +18
-0+ +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> + +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV ); + +return gpio_to_desc (pentry->pin_no); +}drivers/gpio/gpiolib.c +0
-0@@ -2451,6 +2451,9 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, }else if (IS_ENABLED (CONFIG_ACPI ) && dev &&ACPI_HANDLE (dev)) {dev_dbg (dev,"using ACPI for GPIO lookup\n" );Quoted 8 lines
drivers/gpio/gpiolib.c struct gpio_desc *__must_check gpiod_get_index(struct device *dev, · line 2456
Quoted 4 lines
drivers/gpio/gpiolib.c +3
-0desc =acpi_find_gpio (dev, con_id, idx, &flags); + }else if (IS_ENABLED (CONFIG_SFI )) { +dev_dbg (dev,"using SFI for GPIO lookup\n" ); + desc =sfi_get_gpiod_by_name (con_id);Your lookup function is ignoring the dev argument. Are SFI GPIOs always
supposed to be system-global?It's not clear. It could be device related, though SFI itself has probably wrong design. I rather prefer to avoid a dev parameter check at
all.Quoted 8 lines
In this case, your if condition should
likely be} else if (IS_ENABLED(CONFIG_SFI) && !dev) {So that a global SFI GPIO does not get mistakenly assigned to a device
that has, say, a more suited platform mapping on the samecon_id.So, for example in the driver that could be enumerated from SFI, DT, and
via platform data you suggest to have something likedesc =
gpiod_get(dev, "con_id_device_tree");
...Quoted 7 lines
if (
IS_ERR(desc))
desc =gpiod_get(NULL, "con_id_sfi");if (
IS_ERR(desc))
desc =gpiod_get(???, "con_id_from_platdata");Correct?
No, this is not what I'm suggesting. Device drivers should not care who provides the GPIO, they should just ask for it, and obtain it (or not).
The scope of the problem actually depends on what SFI GPIOs are used for. For instance, let's say you have two devices each using an enabling
GPIO which happens to be provided by SFI. Both drivers for these devices
obtain the enable GPIO as follows:Quoted 6 lines
Quoted 1 line
enable_gpio = gpiod_get(dev, "enable");You rather can't do that in sfi case. Yes, the structure has (stringy) reference to gpio chip that provides a line, but in practice the gpio line name should be unique. Consider this is misdesign of SFI as I mentioned earlier.
Ok. Note that this is not necessarily a bad thing, it just means that SFI GPIOs should only be used by platform-specific code that is aware of their existence (and particular naming).
Quoted 8 lines
Device drivers on the other hand cannot make any assumption on who will provide the GPIO, so if you want them to be able to use SFI GPIOs, you will need an actual way to map then to (device, function) pairs. If this use-case is out of the equation, then I'm glad with the v6 of your patch.
But David raised a point that kind of sounds like he might want to use them that way, so I'm holding off my ack until this is solved.
Intel-MID needs this approach on platform code inside arch/x86/platform/intel-mid/device_libs/ directory. But I agree it should not
be found on device drivers in general.
Br, David
Quoted 3 lines
Alex.
Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API
Quoted 164 lines · David Cohen
Quoted 163 lines · Alex Courbot
Quoted 145 lines · Andy Shevchenko
Quoted 144 lines · Alex Courbot
Quoted 125 lines · Andy Shevchenko
Quoted 93 lines · Alex Courbot
Quoted 92 lines · Andy Shevchenko
To support some (legacy) firmwares and platforms let's make life easier for
their customers.This patch provides a function which converts
sfi_gpio_table_entrytogpio_desc. The use of it is integrated into GPIO library to enable generic
access to the SFI GPIO resources.Signed-off-by Andy Shevchenko <[email protected]>--- drivers/gpio/Kconfig | 4 ++++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 ++++++++++++++++++++++++++++ drivers/gpio/gpiolib.c | 3 +++ drivers/gpio/gpiolib.h | 13 +++++++++++++ 5 files changed, 49 insertions(+)create mode 100644
drivers/gpio/gpiolib-sfi.cdrivers/gpio/Kconfig +4
-0@@ -51,6 +51,10 @@ config OF_GPIO def_bool y depends on OF +config GPIO_SFI + def_bool y + depends on SFI + config GPIO_ACPI def_bool y depends on ACPIdrivers/gpio/Makefile +1
-0@@ -5,6 +5,7 @@ ccflags-$(CONFIG_DEBUG_GPIO) += -DDEBUG obj-$(CONFIG_GPIO_DEVRES) += devres.o obj-$(CONFIG_GPIOLIB) += gpiolib.o obj-$(CONFIG_OF_GPIO) += gpiolib-of.o +obj-$(CONFIG_GPIO_SFI) += gpiolib-sfi.o obj-$(CONFIG_GPIO_ACPI) += gpiolib-acpi.o # Device drivers. Generally keep list sorted alphabeticallydrivers/gpio/gpiolib-sfi.c new file +7
-0@@ -0,0 +1,28 @@ +/* +* Simple Firmware Interface (SFI) helpers for GPIO API +* +* Copyright (C) 2013, Intel Corporation +* Author: Andy Shevchenko <[email protected]> +* +* This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License version 2 as
+ * published by the Free Software Foundation.
+ */drivers/gpio/gpiolib-sfi.c new file +18
-0+ +#include <linux/gpio/consumer.h> +#include <linux/sfi.h> +#include <linux/errno.h> +#include <linux/err.h> + +#include "gpiolib.h" + +struct gpio_desc *sfi_get_gpiod_by_name (const char *name) +{ +struct sfi_gpio_table_entry *pentry; + + pentry =sfi_gpio_get_entry_by_name (name); +if (!pentry) +return ERR_PTR (-ENODEV ); + +return gpio_to_desc (pentry->pin_no); +}drivers/gpio/gpiolib.c +0
-0@@ -2451,6 +2451,9 @@ struct gpio_desc *__must_check gpiod_get_index(struct device *dev, }else if (IS_ENABLED (CONFIG_ACPI ) && dev &&ACPI_HANDLE (dev)) {dev_dbg (dev,"using ACPI for GPIO lookup\n" );Quoted 8 lines
drivers/gpio/gpiolib.c struct gpio_desc *__must_check · line 2458
Quoted 4 lines
drivers/gpio/gpiolib.c +3
-0desc =acpi_find_gpio (dev, con_id, idx, &flags); + }else if (IS_ENABLED (CONFIG_SFI )) { +dev_dbg (dev,"using SFI for GPIO lookup\n" ); + desc =sfi_get_gpiod_by_name (con_id);Your lookup function is ignoring the dev argument. Are SFI GPIOs always
supposed to be system-global?It's not clear. It could be device related, though SFI itself has probably wrong design. I rather prefer to avoid a dev parameter check at
all.Quoted 8 lines
In this case, your if condition should
likely be} else if (IS_ENABLED(CONFIG_SFI) && !dev) {So that a global SFI GPIO does not get mistakenly assigned to a device
that has, say, a more suited platform mapping on the samecon_id.So, for example in the driver that could be enumerated from SFI, DT, and
via platform data you suggest to have something likedesc =
gpiod_get(dev, "con_id_device_tree");
...Quoted 7 lines
if (
IS_ERR(desc))desc = gpiod_get(NULL, "con_id_sfi");if (
IS_ERR(desc))desc = gpiod_get(???, "con_id_from_platdata");Correct?
No, this is not what I'm suggesting. Device drivers should not care who provides the GPIO, they should just ask for it, and obtain it (or not).
The scope of the problem actually depends on what SFI GPIOs are used for. For instance, let's say you have two devices each using an enabling
GPIO which happens to be provided by SFI. Both drivers for these devices
obtain the enable GPIO as follows:Quoted 6 lines
Quoted 1 line
enable_gpio = gpiod_get(dev, "enable");You rather can't do that in sfi case. Yes, the structure has (stringy) reference to gpio chip that provides a line, but in practice the gpio line name should be unique. Consider this is misdesign of SFI as I mentioned earlier.
Ok. Note that this is not necessarily a bad thing, it just means that SFI GPIOs should only be used by platform-specific code that is aware of their existence (and particular naming).
Device drivers on the other hand cannot make any assumption on who will provide the GPIO, so if you want them to be able to use SFI GPIOs, you will need an actual way to map then to (device, function) pairs. If this use-case is out of the equation, then I'm glad with the v6 of your patch.
Quoted 7 lines
Quoted 2 lines
But David raised a point that kind of sounds like he might want to use them that way, so I'm holding off my ack until this is solved.
Intel-MID needs this approach on platform code inside
arch/x86/platform/intel-mid/device_libs/directory. But I agree it should not
be found on device drivers in general.
That was my only concern. If these GPIOs are only used by code that expects them to come from SFI, then I'm fine with Andy's latest version.
Hi Andy,
Quoted 1 line · Andy Shevchenko
Here is a third [1] version of the SFI GPIO helpers. This series contains SFI
Your patch set subject says it's the first version :)
Quoted 5 lines
GPIO helpers along with the update of GPIO library to lookup for SFI resources.
It has been tested on Medfield device on top of v3.13-rc2 + Alexandre's GPIO descriptor API update [2], Mika's ACPI gpio patches [3] and my GPIO library clean up [4].
Good test case. I'm able to extend it to Merrifield.
Br, David
Quoted 26 lines
[1] https://lkml.org/lkml/2013/6/5/316
[2] http://www.spinics.net/lists/kernel/msg1645907.html
[3] http://www.spinics.net/lists/linux-acpi/msg47572.html
[4] https://www.mail-archive.com/[email protected]/msg01515.htmlAndy Shevchenko (3):
SFI: fix compiler warnings
SFI: store GPIO table and export lookup function
gpiolib: append SFI helpers for GPIO APIdrivers/gpio/Kconfig | 4 ++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 +++++++++++ drivers/gpio/gpiolib.c | 3 ++ drivers/gpio/gpiolib.h | 13 +++++ drivers/sfi/Makefile | 2 +- drivers/sfi/sfi_acpi.c | 9 ++-- drivers/sfi/sfi_core.c | 10 +++- drivers/sfi/sfi_core.h | 3 ++ drivers/sfi/sfi_gpio.c | 123 +++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi.h | 8 +++ 11 files changed, 197 insertions(+), 7 deletions(-) create mode 100644 drivers/gpio/gpiolib-sfi.c create mode 100644 drivers/sfi/sfi_gpio.c
Quoted 6 lines · David Cohen
Hi Andy,
Quoted 1 line · Andy Shevchenko
Here is a third [1] version of the SFI GPIO helpers. This series contains SFI
Your patch set subject says it's the first version :)
Will fix in next iteration.
Quoted 8 lines
Quoted 6 lines
GPIO helpers along with the update of GPIO library to lookup for SFI resources.
It has been tested on Medfield device on top of v3.13-rc2 + Alexandre's GPIO descriptor API update [2], Mika's ACPI gpio patches [3] and my GPIO library clean up [4].
Good test case. I'm able to extend it to Merrifield.
Be informed, that no driver is modified by this patch series. To test you have to update one or more drivers. I've tested against gpio_keys, but I'm going to make that solution upstreamable (it requires to update gpio_keys driver itself first).
Quoted 30 lines
Br, David
Quoted 26 lines
[1] https://lkml.org/lkml/2013/6/5/316
[2] http://www.spinics.net/lists/kernel/msg1645907.html
[3] http://www.spinics.net/lists/linux-acpi/msg47572.html
[4] https://www.mail-archive.com/[email protected]/msg01515.htmlAndy Shevchenko (3):
SFI: fix compiler warnings
SFI: store GPIO table and export lookup function
gpiolib: append SFI helpers for GPIO APIdrivers/gpio/Kconfig | 4 ++ drivers/gpio/Makefile | 1 + drivers/gpio/gpiolib-sfi.c | 28 +++++++++++ drivers/gpio/gpiolib.c | 3 ++ drivers/gpio/gpiolib.h | 13 +++++ drivers/sfi/Makefile | 2 +- drivers/sfi/sfi_acpi.c | 9 ++-- drivers/sfi/sfi_core.c | 10 +++- drivers/sfi/sfi_core.h | 3 ++ drivers/sfi/sfi_gpio.c | 123 +++++++++++++++++++++++++++++++++++++++++++++ include/linux/sfi.h | 8 +++ 11 files changed, 197 insertions(+), 7 deletions(-) create mode 100644 drivers/gpio/gpiolib-sfi.c create mode 100644 drivers/sfi/sfi_gpio.c
Signature · 2 lines
Andy Shevchenko <[email protected]> Intel Finland Oy