Skip to contentSkip to conversation

Build errors in v3.18-stable-queue

DISCUSSION Guenter Roeck 4 participants 11 messages Feb 28, 2016 – Mar 6, 2016

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

Current build results:

	total: 137 pass: 121 fail: 16
Failed builds:
	arm:allmodconfig
	arm:omap2plus_defconfig
	arm:exynos_defconfig
	arm:s5pv210_defconfig
	powerpc:defconfig
	powerpc:allmodconfig
	powerpc:allnoconfig
	powerpc:ppc64e_defconfig
	powerpc:cell_defconfig
	powerpc:maple_defconfig
	powerpc:ppc6xx_defconfig
	powerpc:mpc83xx_defconfig
	powerpc:tqm8xx_defconfig
	powerpc:85xx/sbc8548_defconfig
	powerpc:83xx/mpc834x_mds_defconfig
	powerpc:86xx/sbc8641d_defconfig

Culprits are listed below. Copying the patch authors for advice.

Guenter

---
powerpc:

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list

Culprit: "powerpc/eeh: Do probe on pci_dn"

---
arm:

drivers/pinctrl/samsung/pinctrl-samsung.c: In function 'samsung_gpio_set':
drivers/pinctrl/samsung/pinctrl-samsung.c:551:9: error: implicit declaration of function 'gpiochip_get_data'

Culprit: "pinctrl: samsung: fix SMP race condition"

drivers/tty/serial/omap-serial.c: In function 'serial_omap_stop_tx':
drivers/tty/serial/omap-serial.c:318:10: error: 'struct uart_port' has no member named 'rs485'

drivers/tty/serial/omap-serial.c: In function 'serial_omap_probe':
drivers/tty/serial/omap-serial.c:1683:10: error: 'struct uart_port' has no member named 'rs485_config'

Culprit: "serial/omap: Use the rs485 functions on serial_core"

#1 lore raw
Quoted 40 lines · Guenter Roeck

Current build results:

    total: 137 pass: 121 fail: 16
Failed builds:
    arm:allmodconfig
    arm:omap2plus_defconfig
    arm:exynos_defconfig
    arm:s5pv210_defconfig
    powerpc:defconfig
    powerpc:allmodconfig
    powerpc:allnoconfig
    powerpc:ppc64e_defconfig
    powerpc:cell_defconfig
    powerpc:maple_defconfig
    powerpc:ppc6xx_defconfig
    powerpc:mpc83xx_defconfig
    powerpc:tqm8xx_defconfig
    powerpc:85xx/sbc8548_defconfig
    powerpc:83xx/mpc834x_mds_defconfig
    powerpc:86xx/sbc8641d_defconfig

Culprits are listed below. Copying the patch authors for advice.

Guenter

---
powerpc:

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list

Culprit: "powerpc/eeh: Do probe on pci_dn"

---
arm:

drivers/pinctrl/samsung/pinctrl-samsung.c: In function 'samsung_gpio_set':
drivers/pinctrl/samsung/pinctrl-samsung.c:551:9: error: implicit declaration of function 'gpiochip_get_data'

Culprit: "pinctrl: samsung: fix SMP race condition"

Quoted 7 lines

drivers/tty/serial/omap-serial.c: In function 'serial_omap_stop_tx':
drivers/tty/serial/omap-serial.c:318:10: error: 'struct uart_port' has no member named 'rs485'

drivers/tty/serial/omap-serial.c: In function 'serial_omap_probe':
drivers/tty/serial/omap-serial.c:1683:10: error: 'struct uart_port' has no member named 'rs485_config'

Culprit: "serial/omap: Use the rs485 functions on serial_core"

Thanks Guenter. I'll pull both of these out of 3.18 and 4.1 until we can get a proper backport since it doesn't seem trivial.

Thanks,
Sasha

#2 lore raw
Quoted 21 lines · Sasha Levin
Quoted 20 lines · Guenter Roeck

Current build results:

    total: 137 pass: 121 fail: 16
Failed builds:
    arm:allmodconfig
    arm:omap2plus_defconfig
    arm:exynos_defconfig
    arm:s5pv210_defconfig
    powerpc:defconfig
    powerpc:allmodconfig
    powerpc:allnoconfig
    powerpc:ppc64e_defconfig
    powerpc:cell_defconfig
    powerpc:maple_defconfig
    powerpc:ppc6xx_defconfig
    powerpc:mpc83xx_defconfig
    powerpc:tqm8xx_defconfig
    powerpc:85xx/sbc8548_defconfig
    powerpc:83xx/mpc834x_mds_defconfig
    powerpc:86xx/sbc8641d_defconfig
Quoted 2 lines
Quoted 2 lines

Culprits are listed below. Copying the patch authors for advice.

Thanks for reporting it, Guenter. some questions as below:

- The subject says you were building 3.18 stable kernel, but the PowerPC related commit was merged to 4.1 kernel. So I assume you were building 4.1 stable kernel instead.
- All above PowerPC configuration failed to build? At least, those configuration that has (CONFIG_PPC64=y) shouldn't fail because of the error you pointed.
- Those configuration without (CONFIG_PPC64=y) needs below fix. Could you please have a try? If it works, I'll post a patch as soon as possible.

arch/powerpc/kernel/of_platform.c +2 -0
index b60a67d..25dec0b 100644
@@ -71,8 +71,10 @@ static int of_pci_phb_probe(struct platform_device *dev)
        eeh_dev_phb_init_dynamic(phb);
 
        /* Register devices with EEH */
+#ifdef CONFIG_EEH
        if (dev->dev.of_node->child)
                eeh_add_device_tree_early(PCI_DN(dev->dev.of_node));
+#endif
 

Thanks,
Gavin

Quoted 34 lines
Quoted 18 lines

Guenter

---
powerpc:

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list

Culprit: "powerpc/eeh: Do probe on pci_dn"

---
arm:

drivers/pinctrl/samsung/pinctrl-samsung.c: In function 'samsung_gpio_set':
drivers/pinctrl/samsung/pinctrl-samsung.c:551:9: error: implicit declaration of function 'gpiochip_get_data'

Culprit: "pinctrl: samsung: fix SMP race condition"

Quoted 7 lines

drivers/tty/serial/omap-serial.c: In function 'serial_omap_stop_tx':
drivers/tty/serial/omap-serial.c:318:10: error: 'struct uart_port' has no member named 'rs485'

drivers/tty/serial/omap-serial.c: In function 'serial_omap_probe':
drivers/tty/serial/omap-serial.c:1683:10: error: 'struct uart_port' has no member named 'rs485_config'

Culprit: "serial/omap: Use the rs485 functions on serial_core"

Thanks Guenter. I'll pull both of these out of 3.18 and 4.1 until we can get a proper backport since it doesn't seem trivial.

Thanks,
Sasha

#3 lore raw
Quoted 22 lines · Gavin Shan
Quoted 21 lines · Sasha Levin
Quoted 20 lines · Guenter Roeck

Current build results:

     total: 137 pass: 121 fail: 16
Failed builds:
     arm:allmodconfig
     arm:omap2plus_defconfig
     arm:exynos_defconfig
     arm:s5pv210_defconfig
     powerpc:defconfig
     powerpc:allmodconfig
     powerpc:allnoconfig
     powerpc:ppc64e_defconfig
     powerpc:cell_defconfig
     powerpc:maple_defconfig
     powerpc:ppc6xx_defconfig
     powerpc:mpc83xx_defconfig
     powerpc:tqm8xx_defconfig
     powerpc:85xx/sbc8548_defconfig
     powerpc:83xx/mpc834x_mds_defconfig
     powerpc:86xx/sbc8641d_defconfig
Quoted 8 lines
Quoted 2 lines
Quoted 2 lines

Culprits are listed below. Copying the patch authors for advice.

Thanks for reporting it, Guenter. some questions as below:

- The subject says you were building 3.18 stable kernel, but the PowerPC related commit was merged to 4.1 kernel. So I assume you were building 4.1 stable kernel instead.

In this case it was 3.18. There were failures in 4.1 as well, but they are different. All builds in 3.18 fail, but only two out of 12 builds in 4.1. For 3.18, this includes both 32 bit and 64 bit builds.

Quoted 3 lines

- All above PowerPC configuration failed to build? At least, those configuration that has (CONFIG_PPC64=y) shouldn't fail because of the error you pointed.

Possibly, but they still fail. Sorry if my report missed the 64 bit error(s).

Quoted 8 lines

- Those configuration without (CONFIG_PPC64=y) needs below fix. Could you please have a try? If it works, I'll post a patch as soon as possible.

arch/powerpc/kernel/of_platform.c +0 -0
index b60a67d..25dec0b 100644

arch/powerpc/kernel/of_platform.c static int of_pci_phb_probe(struct platform_device *dev) · line 78

Quoted 8 lines
arch/powerpc/kernel/of_platform.c +2 -0
@@ -71,8 +71,10 @@ static int of_pci_phb_probe(struct platform_device *dev)
         eeh_dev_phb_init_dynamic(phb);
 
         /* Register devices with EEH */
+#ifdef CONFIG_EEH
         if (dev->dev.of_node->child)
                 eeh_add_device_tree_early(PCI_DN(dev->dev.of_node));
+#endif
 

Looking into a ppc 64 bit log (from a qemu test attempt), I see

In file included from arch/powerpc/mm/init_64.c:59:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list [-Werror]
static inline void eeh_add_device_early(struct pci_dn *pdn) { }

which won't be fixed by the above patch.

The build logs are public at http://kerneltests.org/builders.
They are typically available for at least 30 days, so feel free to have a look yourself.

Thanks,
Guenter

Quoted 39 lines
arch/powerpc/kernel/of_platform.c +0 -0
 Thanks,

Gavin

Quoted 34 lines
Quoted 18 lines

Guenter

---
powerpc:

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list

Culprit: "powerpc/eeh: Do probe on pci_dn"

---
arm:

drivers/pinctrl/samsung/pinctrl-samsung.c: In function 'samsung_gpio_set':
drivers/pinctrl/samsung/pinctrl-samsung.c:551:9: error: implicit declaration of function 'gpiochip_get_data'

Culprit: "pinctrl: samsung: fix SMP race condition"

Quoted 7 lines

drivers/tty/serial/omap-serial.c: In function 'serial_omap_stop_tx':
drivers/tty/serial/omap-serial.c:318:10: error: 'struct uart_port' has no member named 'rs485'

drivers/tty/serial/omap-serial.c: In function 'serial_omap_probe':
drivers/tty/serial/omap-serial.c:1683:10: error: 'struct uart_port' has no member named 'rs485_config'

Culprit: "serial/omap: Use the rs485 functions on serial_core"

Thanks Guenter. I'll pull both of these out of 3.18 and 4.1 until we can get a proper backport since it doesn't seem trivial.

Thanks,
Sasha

#4 lore raw
Quoted 65 lines · Guenter Roeck
Quoted 22 lines · Gavin Shan
Quoted 21 lines · Sasha Levin
Quoted 20 lines · Guenter Roeck

Current build results:

    total: 137 pass: 121 fail: 16
Failed builds:
    arm:allmodconfig
    arm:omap2plus_defconfig
    arm:exynos_defconfig
    arm:s5pv210_defconfig
    powerpc:defconfig
    powerpc:allmodconfig
    powerpc:allnoconfig
    powerpc:ppc64e_defconfig
    powerpc:cell_defconfig
    powerpc:maple_defconfig
    powerpc:ppc6xx_defconfig
    powerpc:mpc83xx_defconfig
    powerpc:tqm8xx_defconfig
    powerpc:85xx/sbc8548_defconfig
    powerpc:83xx/mpc834x_mds_defconfig
    powerpc:86xx/sbc8641d_defconfig
Quoted 8 lines
Quoted 2 lines
Quoted 2 lines

Culprits are listed below. Copying the patch authors for advice.

Thanks for reporting it, Guenter. some questions as below:

- The subject says you were building 3.18 stable kernel, but the PowerPC related commit was merged to 4.1 kernel. So I assume you were building 4.1 stable kernel instead.

In this case it was 3.18. There were failures in 4.1 as well, but they are different. All builds in 3.18 fail, but only two out of 12 builds in 4.1. For 3.18, this includes both 32 bit and 64 bit builds.

Quoted 3 lines

- All above PowerPC configuration failed to build? At least, those configuration that has (CONFIG_PPC64=y) shouldn't fail because of the error you pointed.

Possibly, but they still fail. Sorry if my report missed the 64 bit error(s).

Quoted 8 lines

- Those configuration without (CONFIG_PPC64=y) needs below fix. Could you please have a try? If it works, I'll post a patch as soon as possible.

arch/powerpc/kernel/of_platform.c +0 -0
index b60a67d..25dec0b 100644

arch/powerpc/kernel/of_platform.c static int of_pci_phb_probe(struct platform_device *dev) · line 78

Quoted 8 lines
arch/powerpc/kernel/of_platform.c +2 -0
@@ -71,8 +71,10 @@ static int of_pci_phb_probe(struct platform_device *dev)
        eeh_dev_phb_init_dynamic(phb);
 
        /* Register devices with EEH */
+#ifdef CONFIG_EEH
        if (dev->dev.of_node->child)
                eeh_add_device_tree_early(PCI_DN(dev->dev.of_node));
+#endif
 

Looking into a ppc 64 bit log (from a qemu test attempt), I see

In file included from arch/powerpc/mm/init_64.c:59:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list [-Werror] static inline void eeh_add_device_early(struct pci_dn *pdn) { }

Quoted 6 lines

which won't be fixed by the above patch.

The build logs are public at http://kerneltests.org/builders.
They are typically available for at least 30 days, so feel free to have a look yourself.

Yeah, there're more info included in above link. The attached patch would fix the build error.

Thanks,
Gavin

Quoted 37 lines

Thanks,
Guenter

Quoted 34 lines
arch/powerpc/kernel/of_platform.c +0 -0
 Thanks,

Gavin

Quoted 31 lines
Quoted 18 lines

Guenter

---
powerpc:

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list

Culprit: "powerpc/eeh: Do probe on pci_dn"

---
arm:

drivers/pinctrl/samsung/pinctrl-samsung.c: In function 'samsung_gpio_set':
drivers/pinctrl/samsung/pinctrl-samsung.c:551:9: error: implicit declaration of function 'gpiochip_get_data'

Culprit: "pinctrl: samsung: fix SMP race condition"

Quoted 7 lines

drivers/tty/serial/omap-serial.c: In function 'serial_omap_stop_tx':
drivers/tty/serial/omap-serial.c:318:10: error: 'struct uart_port' has no member named 'rs485'

drivers/tty/serial/omap-serial.c: In function 'serial_omap_probe':
drivers/tty/serial/omap-serial.c:1683:10: error: 'struct uart_port' has no member named 'rs485_config'

Culprit: "serial/omap: Use the rs485 functions on serial_core"

Thanks Guenter. I'll pull both of these out of 3.18 and 4.1 until we can get a proper backport since it doesn't seem trivial.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

Thanks,
Sasha

From 0c43881f78dbdf62ba02364e6306ea271903fc06 Mon Sep 17 00:00:00 2001

From: Gavin Shan <[email protected]>
Date: Tue, 1 Mar 2016 16:02:43 +1100
Subject: [PATCH] powerpc/eeh: Fix build error caused by pci_dn
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

eeh.h could be included when we have following condition. Then we run into build error as below: (CONFIG_PPC64 && !CONFIG_EEH) || (!CONFIG_PPC64 && !CONFIG_EEH)

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: ‘struct pci_dn’ \
declared inside parameter list [-Werror]

    :

In file included from arch/powerpc/mm/hash_utils_64.c:49:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: ‘struct pci_dn’ \
declared inside parameter list [-Werror]

This fixes the issue by replacing those empty inline functions with macro so that we don't rely on @pci_dn when CONFIG_EEH is disabled.

Fixes ff57b45 ("powerpc/eeh: Do probe on pci_dn")
---
 arch/powerpc/include/asm/eeh.h | 20 +++++++-------------
 1 file changed, 7 insertions(+), 13 deletions(-)
arch/powerpc/include/asm/eeh.h +7 -13
index c3a2d14..5f7ec12 100644
@@ -339,19 +339,13 @@ static inline int eeh_check_failure(const volatile void __iomem *token)
 
 #define eeh_dev_check_failure(x) (0)
 
-static inline void eeh_addr_cache_build(void) { }
-
-static inline void eeh_add_device_early(struct pci_dn *pdn) { }
-
-static inline void eeh_add_device_tree_early(struct pci_dn *pdn) { }
-
-static inline void eeh_add_device_late(struct pci_dev *dev) { }
-
-static inline void eeh_add_device_tree_late(struct pci_bus *bus) { }
-
-static inline void eeh_add_sysfs_files(struct pci_bus *bus) { }
-
-static inline void eeh_remove_device(struct pci_dev *dev) { }
+#define eeh_addr_cache_build()
+#define eeh_add_device_early(pdn)
+#define eeh_add_device_tree_early(pdn)
+#define eeh_add_device_late(pdev)
+#define eeh_add_device_tree_late(pbus)
+#define eeh_add_sysfs_files(pbus)
+#define eeh_remove_device(pdev)
 
 #define EEH_POSSIBLE_ERROR(val, type) (0)
 #define EEH_IO_ERROR_VALUE(size) (-1UL)
Signature · 1 line
2.1.0
#6 lore raw
Quoted 5 lines · Sasha Levin
Quoted 1 line · Gavin Shan

What about the 3.18 breakages? Is there a different fix for those?

This build failure is introduced by commit ff57b45 ("powerpc/eeh:
Do probe on pci_dn") which was merged to 4.1 kernel. I don't know why 3.18 stable kernel needs it. If 3.18 stable kernel needs the commit, yes, the attached fix should be put into 3.18 stable kernel as well.

Let me know if there're anything else I need to handle.

gwshan@gwshan:~/sandbox/linux.main$ git tag --contains ff57b45
v4.1
v4.1-rc1
v4.1-rc2
v4.1-rc3
v4.1-rc4
v4.1-rc5
v4.1-rc6
v4.1-rc7
v4.1-rc8
v4.2
 :
 :

Thanks,
Gavin

Quoted 4 lines

Thanks,
Sasha

#7 lore raw
Quoted 7 lines · Gavin Shan
Quoted 5 lines · Sasha Levin
Quoted 1 line · Gavin Shan

What about the 3.18 breakages? Is there a different fix for those?

Quoted 8 lines

This build failure is introduced by commit ff57b45 ("powerpc/eeh:
Do probe on pci_dn") which was merged to 4.1 kernel. I don't know why 3.18 stable kernel needs it. If 3.18 stable kernel needs the commit, yes, the attached fix should be put into 3.18 stable kernel as well.

Let me know if there're anything else I need to handle.

FYI, the patch sent for review/merge: https://patchwork.ozlabs.org/patch/591691/

Thanks,
Gavin

Quoted 21 lines
gwshan@gwshan:~/sandbox/linux.main$ git tag --contains ff57b45
v4.1
v4.1-rc1
v4.1-rc2
v4.1-rc3
v4.1-rc4
v4.1-rc5
v4.1-rc6
v4.1-rc7
v4.1-rc8
v4.2
:
:

Thanks,
Gavin

Quoted 4 lines

Thanks,
Sasha

#8 lore raw
In reply to Sasha Levin
Quoted 44 lines · Sasha Levin
Quoted 43 lines · Guenter Roeck

Current build results:

     total: 137 pass: 121 fail: 16
Failed builds:
     arm:allmodconfig
     arm:omap2plus_defconfig
     arm:exynos_defconfig
     arm:s5pv210_defconfig
     powerpc:defconfig
     powerpc:allmodconfig
     powerpc:allnoconfig
     powerpc:ppc64e_defconfig
     powerpc:cell_defconfig
     powerpc:maple_defconfig
     powerpc:ppc6xx_defconfig
     powerpc:mpc83xx_defconfig
     powerpc:tqm8xx_defconfig
     powerpc:85xx/sbc8548_defconfig
     powerpc:83xx/mpc834x_mds_defconfig
     powerpc:86xx/sbc8641d_defconfig

Culprits are listed below. Copying the patch authors for advice.

Guenter

---
powerpc:

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list

Culprit: "powerpc/eeh: Do probe on pci_dn"

---
arm:

drivers/pinctrl/samsung/pinctrl-samsung.c: In function 'samsung_gpio_set':
drivers/pinctrl/samsung/pinctrl-samsung.c:551:9: error: implicit declaration of function 'gpiochip_get_data'

Culprit: "pinctrl: samsung: fix SMP race condition"

drivers/tty/serial/omap-serial.c: In function 'serial_omap_stop_tx':
drivers/tty/serial/omap-serial.c:318:10: error: 'struct uart_port' has no member named 'rs485'

Quoted 9 lines
Quoted 4 lines

drivers/tty/serial/omap-serial.c: In function 'serial_omap_probe':
drivers/tty/serial/omap-serial.c:1683:10: error: 'struct uart_port' has no member named 'rs485_config'

Culprit: "serial/omap: Use the rs485 functions on serial_core"

Thanks Guenter. I'll pull both of these out of 3.18 and 4.1 until we can get a proper backport since it doesn't seem trivial.

Hi Sasha,

the ppc error is now gone, as is the gpio error. 3.18 builds still fail due to the rs485 error, though.

Guenter

#9 lore raw
In reply to Guenter Roeck
Quoted 40 lines · Guenter Roeck

Current build results:

    total: 137 pass: 121 fail: 16
Failed builds:
    arm:allmodconfig
    arm:omap2plus_defconfig
    arm:exynos_defconfig
    arm:s5pv210_defconfig
    powerpc:defconfig
    powerpc:allmodconfig
    powerpc:allnoconfig
    powerpc:ppc64e_defconfig
    powerpc:cell_defconfig
    powerpc:maple_defconfig
    powerpc:ppc6xx_defconfig
    powerpc:mpc83xx_defconfig
    powerpc:tqm8xx_defconfig
    powerpc:85xx/sbc8548_defconfig
    powerpc:83xx/mpc834x_mds_defconfig
    powerpc:86xx/sbc8641d_defconfig

Culprits are listed below. Copying the patch authors for advice.

Guenter

---
powerpc:

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list

Culprit: "powerpc/eeh: Do probe on pci_dn"

---
arm:

drivers/pinctrl/samsung/pinctrl-samsung.c: In function 'samsung_gpio_set':
drivers/pinctrl/samsung/pinctrl-samsung.c:551:9: error: implicit declaration of function 'gpiochip_get_data'

Culprit: "pinctrl: samsung: fix SMP race condition"

Quoted 7 lines

drivers/tty/serial/omap-serial.c: In function 'serial_omap_stop_tx':
drivers/tty/serial/omap-serial.c:318:10: error: 'struct uart_port' has no member named 'rs485'

drivers/tty/serial/omap-serial.c: In function 'serial_omap_probe':
drivers/tty/serial/omap-serial.c:1683:10: error: 'struct uart_port' has no member named 'rs485_config'

Culprit: "serial/omap: Use the rs485 functions on serial_core"

Thanks Guenter. I think below Linus' patch will also need to fix the build error.

commit b08ea35a3296ee25c4cb53a977b752266dafa2c2
Author: Linus Walleij <[email protected]>
Date: Thu Dec 3 15:14:13 2015 +0100

gpio: add a data pointer to gpio_chip

This adds a void * pointer to gpio_chip so that driver can assign and retrieve some states. This is done to get rid of container_of() calls for gpio_chips embedded inside state containers, so we can remove the need to have the gpio_chip or later (planned) struct gpio_device be dynamically allocated at registration time, so that its struct device can be properly reference counted and not bound to its parent device (e.g. a platform_device) but instead live on after unregistration if it is opened by e.g. a char device or sysfs.

The data is added with the new function gpiochip_add_data() and for compatibility we add static inline wrapper function gpiochip_add() that will call gpiochip_add_data() with NULL as argument. The latter will be removed once we have exorcised gpiochip_add() from the kernel.

gpiochip_get_data() is added as a static inline accessor for drivers to quickly get their data out.

#10 lore raw
Quoted 73 lines · Youngmin Nam
Quoted 40 lines · Guenter Roeck

Current build results:

    total: 137 pass: 121 fail: 16
Failed builds:
    arm:allmodconfig
    arm:omap2plus_defconfig
    arm:exynos_defconfig
    arm:s5pv210_defconfig
    powerpc:defconfig
    powerpc:allmodconfig
    powerpc:allnoconfig
    powerpc:ppc64e_defconfig
    powerpc:cell_defconfig
    powerpc:maple_defconfig
    powerpc:ppc6xx_defconfig
    powerpc:mpc83xx_defconfig
    powerpc:tqm8xx_defconfig
    powerpc:85xx/sbc8548_defconfig
    powerpc:83xx/mpc834x_mds_defconfig
    powerpc:86xx/sbc8641d_defconfig

Culprits are listed below. Copying the patch authors for advice.

Guenter

---
powerpc:

In file included from arch/powerpc/kernel/of_platform.c:30:0:
./arch/powerpc/include/asm/eeh.h:344:48: error: 'struct pci_dn' declared inside parameter list

Culprit: "powerpc/eeh: Do probe on pci_dn"

---
arm:

drivers/pinctrl/samsung/pinctrl-samsung.c: In function 'samsung_gpio_set':
drivers/pinctrl/samsung/pinctrl-samsung.c:551:9: error: implicit declaration of function 'gpiochip_get_data'

Culprit: "pinctrl: samsung: fix SMP race condition"

Quoted 7 lines

drivers/tty/serial/omap-serial.c: In function 'serial_omap_stop_tx':
drivers/tty/serial/omap-serial.c:318:10: error: 'struct uart_port' has no member named 'rs485'

drivers/tty/serial/omap-serial.c: In function 'serial_omap_probe':
drivers/tty/serial/omap-serial.c:1683:10: error: 'struct uart_port' has no member named 'rs485_config'

Culprit: "serial/omap: Use the rs485 functions on serial_core"

Thanks Guenter. I think below Linus' patch will also need to fix the build error.

commit b08ea35a3296ee25c4cb53a977b752266dafa2c2
Author: Linus Walleij <[email protected]>
Date: Thu Dec 3 15:14:13 2015 +0100

gpio: add a data pointer to gpio_chip

This adds a void * pointer to gpio_chip so that driver can assign and retrieve some states. This is done to get rid of container_of() calls for gpio_chips embedded inside state containers, so we can remove the need to have the gpio_chip or later (planned) struct gpio_device be dynamically allocated at registration time, so that its struct device can be properly reference counted and not bound to its parent device (e.g. a platform_device) but instead live on after unregistration if it is opened by e.g. a char device or sysfs.

The data is added with the new function gpiochip_add_data() and for compatibility we add static inline wrapper function gpiochip_add() that will call gpiochip_add_data() with NULL as argument. The latter will be removed once we have exorcised gpiochip_add() from the kernel.

Quoted 5 lines

gpiochip_get_data() is added as a static inline accessor for drivers to quickly get their data out.

Hi Youngmin,

This won't be enough as that commit merely adds a field but doesn't add any usages of that field, so adding that patch into -stable would "fix" the build error but would break the device itself.

Thanks,
Sasha