Skip to contentSkip to conversation

[PATCH v1 0/3]

PATCH In review Andy Shevchenko 4 participants 23 messages Dec 5, 2013 – Dec 12, 2013 GPIO

lore
Share & tools
lore
#0 lore raw

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
#1 lore raw

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
index f5b4ca5..1332546 100644
@@ -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
index 1e824fb..296db7a 100644
@@ -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
#2 lore raw

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.

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

#3 lore raw

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.

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
#4 lore raw

PATCH 2/3 SFI: store GPIO table and export lookup function

In reply to 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.

---
 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
index 2343732..dc011db 100644
@@ -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
index 296db7a..e9ff6f0 100644
@@ -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
index 1d5cfe8..18c663d 100644
@@ -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
index d9b436f..510e74d 100644
@@ -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
#5 lore raw

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.

---
 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
index 2343732..dc011db 100644
@@ -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
index 296db7a..e9ff6f0 100644
@@ -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
index 1d5cfe8..18c663d 100644
@@ -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
new file mode 100644
index 0000000..677368d
@@ -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
index d9b436f..510e74d 100644
@@ -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*/
#6 lore raw

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.

---
 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
index 2343732..dc011db 100644
@@ -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
index 296db7a..e9ff6f0 100644
@@ -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
index 1d5cfe8..18c663d 100644
@@ -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
new file mode 100644
index 0000000..677368d
@@ -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_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?

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
index d9b436f..510e74d 100644
@@ -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
#7 lore raw

Re: [PATCH v1 2/3] SFI: store GPIO table and export lookup function

In reply to Andy Shevchenko
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.

---
  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
index 2343732..dc011db 100644
@@ -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
index 296db7a..e9ff6f0 100644
@@ -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
index 1d5cfe8..18c663d 100644
@@ -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
new file mode 100644
index 0000000..677368d
@@ -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.

#8 lore raw

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.

---
  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
index 2343732..dc011db 100644
@@ -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
index 296db7a..e9ff6f0 100644
@@ -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
index 1d5cfe8..18c663d 100644
@@ -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
new file mode 100644
index 0000000..677368d
@@ -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
#9 lore raw

PATCH 3/3 gpiolib: append SFI helpers for GPIO API

In reply to 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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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
new file mode 100644
index 0000000..c804314
@@ -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
index bad400c..789ae1c 100644
@@ -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
index 82be586..ca0615a 100644
@@ -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
#10 lore raw

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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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 +15 -0
new file mode 100644
index 0000000..c804314
@@ -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 find
IS_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);
+}
#11 lore raw

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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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 +15 -0
new file mode 100644
index 0000000..c804314
@@ -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 wrong

Br, 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
#12 lore raw

Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API

In reply to Andy Shevchenko
Quoted 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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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
new file mode 100644
index 0000000..c804314
@@ -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
index bad400c..789ae1c 100644

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.

#13 lore raw

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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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
new file mode 100644
index 0000000..c804314
@@ -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
index bad400c..789ae1c 100644
@@ -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 -0
  		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?

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
#14 lore raw

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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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
new file mode 100644
index 0000000..c804314
@@ -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
index bad400c..789ae1c 100644
@@ -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 -0
   		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?

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.

#15 lore raw

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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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
new file mode 100644
index 0000000..c804314
@@ -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
index bad400c..789ae1c 100644
@@ -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 -0
   		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?

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
#16 lore raw

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

#17 lore raw

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
#18 lore raw

Re: [PATCH v1 3/3] gpiolib: append SFI helpers for GPIO API

In reply to Andy Shevchenko
Quoted 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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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
new file mode 100644
index 0000000..c804314
@@ -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
index bad400c..789ae1c 100644
@@ -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 -0
    		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?

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.

#19 lore raw

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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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 +7 -0
new file mode 100644
index 0000000..c804314
@@ -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
index bad400c..789ae1c 100644
@@ -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 -0
            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 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 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).

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.

#20 lore raw

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_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
index ae3682d..a12752a 100644
@@ -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
index ee95154..5373e3a 100644
@@ -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 +7 -0
new file mode 100644
index 0000000..c804314
@@ -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
index bad400c..789ae1c 100644
@@ -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 -0
             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 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 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.

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.

#21 lore raw
In reply to Andy Shevchenko

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.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
#22 lore raw
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.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 · 2 lines
Andy Shevchenko <[email protected]>
Intel Finland Oy