Skip to contentSkip to conversation

Remove references to undefined symbols and fix two potential bugs

PATCH Applied Ulf Magnusson 22 participants 76 messages Feb 5, 2018 – Feb 19, 2018

Subsystems (29)
lore
Share & tools
lore
#0 lore raw

cover Remove references to undefined symbols and fix two potential bugs

Hello,

This patchset fixes most references to globally undefined symbols in Kconfig files, as reported by the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script. I will send emails in reply to the patchset for some cases where I'm not sure what the proper fix is.

All of the fixes are independent. I just created a patchset to track them more easily.

Summary of changes:

- Commits 1 and 2 fix what looks like potential bugs: A broken (intended) select of PL310_ERRATA_753970 for ARM/mvebu, and a misspelled dependency on PPC_PSERIES in KVM/PPC.

- Commits 3-5 remove leftover AVR32 stuff.

- Commits 6 and 7 remove some selects of removed symbols from riscv.

- Commits 8-17 remove various references to removed symbols.

- Commits 18-20 fix some (harmless) malformed defaults.

Ulf Magnusson (20):
ARM: mvebu: Fix broken PL310_ERRATA_753970 selects
KVM: PPC: Book3S PR: Fix broken select due to misspelling
kconfig: Remove leftover references to AVR32 symbol
misc: atmel: Remove CPU_AT32AP700X (AVR32) reference
rtc: at32ap700x: Remove PLATFORM_AT32AP dependency
riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select
riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select s390/kconfig: Remove ARCH_WANTS_PROT_NUMA_PROT_NONE select
clk: sunxi-ng: Remove SUNXI_CCU_* selects lib/Kconfig: Remove leftover select of GENERIC_IO
usb: gadget: udc: Remove USB_GADGET_DUALSPEED select
ARM: debug: Remove ARCH_MSM dep. from UNCOMPRESS_INCLUDE
ore: Remove PNFS_OBJLAYOUT dependency
iio: gyro: mpu3050: Remove INPUT_MPU3050 dep. from I2C
phy: Remove SOC_EXYNOS4212 dep. from PHY_EXYNOS4X12_USB
auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.
media: sec: Remove PLAT_S5P dependency
iwlwifi: fix malformed CONFIG_IWLWIFI_PCIE_RTPM default
ARC: Fix malformed ARC_EMUL_UNALIGNED default x86/PCI: VMD: Fix malformed default

 arch/arc/Kconfig                           |  1 -
 arch/arm/Kconfig.debug                     |  2 +-
 arch/arm/mach-mvebu/Kconfig                |  4 ++--
 arch/powerpc/kvm/Kconfig                   |  2 +-
 arch/riscv/Kconfig                         |  2 --
 arch/s390/Kconfig                          |  1 -
 drivers/auxdisplay/Kconfig                 |  2 +-
 drivers/clk/sunxi-ng/Kconfig               |  7 -------
 drivers/cpufreq/Kconfig                    | 10 ----------
 drivers/iio/gyro/Kconfig                   |  1 -
 drivers/media/platform/Kconfig             |  2 +-
 drivers/misc/Kconfig                       |  1 -
 drivers/net/wireless/intel/iwlwifi/Kconfig |  1 -
 drivers/pci/host/Kconfig                   |  1 -
 drivers/phy/samsung/Kconfig                |  2 +-
 drivers/pwm/Kconfig                        |  2 +-
 drivers/rtc/Kconfig                        |  2 +-
 drivers/spi/Kconfig                        |  2 +-
 drivers/usb/gadget/udc/Kconfig             |  1 -
 drivers/video/console/Kconfig              |  2 +-
 fs/exofs/Kconfig.ore                       |  2 +-
 lib/Kconfig                                |  1 -
 22 files changed, 12 insertions(+), 39 deletions(-)
Signature · 1 line
2.14.1
#1 lore raw

PATCH 1/20 ARM: mvebu: Fix broken PL310_ERRATA_753970 selects

The MACH_ARMADA_375 and MACH_ARMADA_38X boards select ARM_ERRATA_753970, but it was renamed to PL310_ERRATA_753970 by commit fa0ce4035d48 ("ARM:
7162/1: errata: tidy up Kconfig options for PL310 errata workarounds").

Fix the selects to use the new name.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/arm/mach-mvebu/Kconfig | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/arm/mach-mvebu/Kconfig +2 -2
index 6b32dc527edc..2c20599cc350 100644
@@ -41,7 +41,7 @@ config MACH_ARMADA_375
 	depends on ARCH_MULTI_V7
 	select ARMADA_370_XP_IRQ
 	select ARM_ERRATA_720789
-	select ARM_ERRATA_753970
+	select PL310_ERRATA_753970
 	select ARM_GIC
 	select ARMADA_375_CLK
 	select HAVE_ARM_SCU
@@ -57,7 +57,7 @@ config MACH_ARMADA_38X
 	bool "Marvell Armada 380/385 boards"
 	depends on ARCH_MULTI_V7
 	select ARM_ERRATA_720789
-	select ARM_ERRATA_753970
+	select PL310_ERRATA_753970
 	select ARM_GIC
 	select ARM_GLOBAL_TIMER
 	select CLKSRC_ARM_GLOBAL_TIMER_SCHED_CLOCK
Signature · 1 line
2.14.1
maintainer
#2 lore raw

Re: [PATCH 01/20] ARM: mvebu: Fix broken PL310_ERRATA_753970 selects

Hi Ulf,

Quoted 9 lines · Ulf Magnusson

The MACH_ARMADA_375 and MACH_ARMADA_38X boards select ARM_ERRATA_753970, but it was renamed to PL310_ERRATA_753970 by commit fa0ce4035d48 ("ARM:
7162/1: errata: tidy up Kconfig options for PL310 errata workarounds").

Fix the selects to use the new name.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Applied on mvebu/fixes with the "Fixes": and "CC:stable" flags

Thanks,

Gregory

Quoted 36 lines
---
 arch/arm/mach-mvebu/Kconfig | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
arch/arm/mach-mvebu/Kconfig +2 -2
index 6b32dc527edc..2c20599cc350 100644
@@ -41,7 +41,7 @@ config MACH_ARMADA_375
 	depends on ARCH_MULTI_V7
 	select ARMADA_370_XP_IRQ
 	select ARM_ERRATA_720789
-	select ARM_ERRATA_753970
+	select PL310_ERRATA_753970
 	select ARM_GIC
 	select ARMADA_375_CLK
 	select HAVE_ARM_SCU
@@ -57,7 +57,7 @@ config MACH_ARMADA_38X
 	bool "Marvell Armada 380/385 boards"
 	depends on ARCH_MULTI_V7
 	select ARM_ERRATA_720789
-	select ARM_ERRATA_753970
+	select PL310_ERRATA_753970
 	select ARM_GIC
 	select ARM_GLOBAL_TIMER
 	select CLKSRC_ARM_GLOBAL_TIMER_SCHED_CLOCK
Signature · 7 lines
2.14.1


_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel at lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Signature · 3 lines
Gregory Clement, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com
#3 lore raw

PATCH 2/20 KVM: PPC: Book3S PR: Fix broken select due to misspelling

In reply to Ulf Magnusson

Commit 76d837a4c0f9 ("KVM: PPC: Book3S PR: Don't include SPAPR TCE code on non-pseries platforms") added a reference to the globally undefined symbol PPC_SERIES. Looking at the rest of the commit, PPC_PSERIES was probably intended.

Change PPC_SERIES to PPC_PSERIES.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/powerpc/kvm/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/powerpc/kvm/Kconfig +1 -1
index f884a0529dfe..68a0e9d5b440 100644
@@ -69,7 +69,7 @@ config KVM_BOOK3S_64
 	select KVM_BOOK3S_64_HANDLER
 	select KVM
 	select KVM_BOOK3S_PR_POSSIBLE if !KVM_BOOK3S_HV_POSSIBLE
-	select SPAPR_TCE_IOMMU if IOMMU_SUPPORT && (PPC_SERIES || PPC_POWERNV)
+	select SPAPR_TCE_IOMMU if IOMMU_SUPPORT && (PPC_PSERIES || PPC_POWERNV)
 	---help---
 	  Support running unmodified book3s_64 and book3s_32 guest kernels
 	  in virtual machines on book3s_64 host processors.
Signature · 1 line
2.14.1
Acked-by on v1 2/20
#4 lore raw

Re: [PATCH 02/20] KVM: PPC: Book3S PR: Fix broken select due to misspelling

Quoted 5 lines · Ulf Magnusson

Commit 76d837a4c0f9 ("KVM: PPC: Book3S PR: Don't include SPAPR TCE code on non-pseries platforms") added a reference to the globally undefined symbol PPC_SERIES. Looking at the rest of the commit, PPC_PSERIES was probably intended.

Quoted 7 lines

Change PPC_SERIES to PPC_PSERIES.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Which tree is this series going into?

Paul.

#5 lore raw

Re: [PATCH 02/20] KVM: PPC: Book3S PR: Fix broken select due to misspelling

Quoted 12 lines · Paul Mackerras
Quoted 11 lines · Ulf Magnusson

Commit 76d837a4c0f9 ("KVM: PPC: Book3S PR: Don't include SPAPR TCE code on non-pseries platforms") added a reference to the globally undefined symbol PPC_SERIES. Looking at the rest of the commit, PPC_PSERIES was probably intended.

Change PPC_SERIES to PPC_PSERIES.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Quoted 7 lines
Quoted 1 line

Which tree is this series going into?

Paul.

I didn't have a particular one in mind. I imagined people would pick up individual patches into the trees that make the most sense. It was easiest to do the undefined symbol removal as a kernel-wide batch job.

All patches are against linux-next.

Cheers,
Ulf

#6 lore raw

Re: [PATCH 02/20] KVM: PPC: Book3S PR: Fix broken select due to misspelling

Quoted 17 lines · Ulf Magnusson
Quoted 16 lines · Paul Mackerras
Quoted 5 lines · Ulf Magnusson

Commit 76d837a4c0f9 ("KVM: PPC: Book3S PR: Don't include SPAPR TCE code on non-pseries platforms") added a reference to the globally undefined symbol PPC_SERIES. Looking at the rest of the commit, PPC_PSERIES was probably intended.

Quoted 7 lines

Change PPC_SERIES to PPC_PSERIES.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Quoted 7 lines
Quoted 3 lines

Which tree is this series going into?

Paul.

I didn't have a particular one in mind. I imagined people would pick up individual patches into the trees that make the most sense. It was easiest to do the undefined symbol removal as a kernel-wide batch job.

OK, I'll take this one then.

Paul.

#7 lore raw

PATCH 3/20 kconfig: Remove leftover references to AVR32 symbol

In reply to Ulf Magnusson

The AVR32 symbol was removed in commit 26202873bb51 ("avr32: remove support for AVR32 architecture"). Remove the remaining references to it from the Kconfig files.

No references remain to the removed AVR32_AT32AP_CPUFREQ symbol either, so remove that one too.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/cpufreq/Kconfig       | 10 ----------
 drivers/pwm/Kconfig           |  2 +-
 drivers/spi/Kconfig           |  2 +-
 drivers/video/console/Kconfig |  2 +-
 4 files changed, 3 insertions(+), 13 deletions(-)
drivers/cpufreq/Kconfig +0 -10
index d8addbce40bc..608af20a3494 100644
@@ -239,16 +239,6 @@ if PPC32 || PPC64
 source "drivers/cpufreq/Kconfig.powerpc"
 endif
 
-if AVR32
-config AVR32_AT32AP_CPUFREQ
-	bool "CPU frequency driver for AT32AP"
-	depends on PLATFORM_AT32AP
-	default n
-	help
-	  This enables the CPU frequency driver for AT32AP processors.
-	  If in doubt, say N.
-endif
-
 if IA64
 config IA64_ACPI_CPUFREQ
 	tristate "ACPI Processor P-States driver"
drivers/pwm/Kconfig +1 -1
index 763ee50ea57d..f16aad3bf5d6 100644
@@ -43,7 +43,7 @@ config PWM_AB8500
 
 config PWM_ATMEL
 	tristate "Atmel PWM support"
-	depends on ARCH_AT91 || AVR32
+	depends on ARCH_AT91
 	help
 	  Generic PWM framework driver for Atmel SoC.
 
drivers/spi/Kconfig +1 -1
index 603783976b81..c1e455d46c7f 100644
@@ -72,7 +72,7 @@ config SPI_ARMADA_3700
 config SPI_ATMEL
 	tristate "Atmel SPI Controller"
 	depends on HAS_DMA
-	depends on (ARCH_AT91 || AVR32 || COMPILE_TEST)
+	depends on (ARCH_AT91 || COMPILE_TEST)
 	help
 	  This selects a driver for the Atmel SPI Controller, present on
 	  many AT32 (AVR32) and AT91 (ARM) chips.
drivers/video/console/Kconfig +1 -1
index 7f1f1fbcef9e..be63759b6027 100644
@@ -7,7 +7,7 @@ menu "Console display driver support"
 config VGA_CONSOLE
 	bool "VGA text console" if EXPERT || !X86
 	depends on !4xx && !PPC_8xx && !SPARC && !M68K && !PARISC && !FRV && \
-		!SUPERH && !BLACKFIN && !AVR32 && !MN10300 && !CRIS && \
+		!SUPERH && !BLACKFIN && !MN10300 && !CRIS && \
 		(!ARM || ARCH_FOOTBRIDGE || ARCH_INTEGRATOR || ARCH_NETWINDER) && \
 		!ARM64 && !ARC && !MICROBLAZE && !OPENRISC
 	default y
Signature · 1 line
2.14.1
maintainer
#8 lore raw

Re: [PATCH 03/20] kconfig: Remove leftover references to AVR32 symbol

Quoted 27 lines · Ulf Magnusson

The AVR32 symbol was removed in commit 26202873bb51 ("avr32: remove support for AVR32 architecture"). Remove the remaining references to it from the Kconfig files.

No references remain to the removed AVR32_AT32AP_CPUFREQ symbol either, so remove that one too.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/cpufreq/Kconfig       | 10 ----------
 drivers/pwm/Kconfig           |  2 +-
 drivers/spi/Kconfig           |  2 +-
 drivers/video/console/Kconfig |  2 +-
 4 files changed, 3 insertions(+), 13 deletions(-)
drivers/cpufreq/Kconfig +0 -0
index d8addbce40bc..608af20a3494 100644
@@ -239,16 +239,6 @@ if PPC32 || PPC64
 source "drivers/cpufreq/Kconfig.powerpc"
 endif
 

drivers/cpufreq/Kconfig if PPC32 || PPC64 · line 251

Quoted 10 lines
drivers/cpufreq/Kconfig +0 -10
-if AVR32
-config AVR32_AT32AP_CPUFREQ
-	bool "CPU frequency driver for AT32AP"
-	depends on PLATFORM_AT32AP
-	default n
-	help
-	  This enables the CPU frequency driver for AT32AP processors.
-	  If in doubt, say N.
-endif
-

https://lkml.kernel.org/r/[email protected]

Signature · 1 line
viresh
#9 lore raw

PATCH v2 3/20 kconfig: Remove leftover references to AVR32 symbol

The AVR32 symbol was removed in commit 26202873bb51 ("avr32: remove support for AVR32 architecture"). Remove the remaining references to it from the Kconfig files.

---
Changes in v2:
Removal of AVR32_AT32AP_CPUFREQ is already pending in
https://marc.info/?i=20180118200202.11883-1-clabbe.montjoie%40gmail.com (which also does it properly -- I forgot to search for related files here). Just remove the other references.

 drivers/pwm/Kconfig           | 2 +-
 drivers/spi/Kconfig           | 2 +-
 drivers/video/console/Kconfig | 2 +-
 3 files changed, 3 insertions(+), 3 deletions(-)
drivers/pwm/Kconfig +1 -1
index 763ee50ea57d..f16aad3bf5d6 100644
@@ -43,7 +43,7 @@ config PWM_AB8500
 
 config PWM_ATMEL
 	tristate "Atmel PWM support"
-	depends on ARCH_AT91 || AVR32
+	depends on ARCH_AT91
 	help
 	  Generic PWM framework driver for Atmel SoC.
 
drivers/spi/Kconfig +1 -1
index 603783976b81..c1e455d46c7f 100644
@@ -72,7 +72,7 @@ config SPI_ARMADA_3700
 config SPI_ATMEL
 	tristate "Atmel SPI Controller"
 	depends on HAS_DMA
-	depends on (ARCH_AT91 || AVR32 || COMPILE_TEST)
+	depends on (ARCH_AT91 || COMPILE_TEST)
 	help
 	  This selects a driver for the Atmel SPI Controller, present on
 	  many AT32 (AVR32) and AT91 (ARM) chips.
drivers/video/console/Kconfig +1 -1
index 7f1f1fbcef9e..be63759b6027 100644
@@ -7,7 +7,7 @@ menu "Console display driver support"
 config VGA_CONSOLE
 	bool "VGA text console" if EXPERT || !X86
 	depends on !4xx && !PPC_8xx && !SPARC && !M68K && !PARISC && !FRV && \
-		!SUPERH && !BLACKFIN && !AVR32 && !MN10300 && !CRIS && \
+		!SUPERH && !BLACKFIN && !MN10300 && !CRIS && \
 		(!ARM || ARCH_FOOTBRIDGE || ARCH_INTEGRATOR || ARCH_NETWINDER) && \
 		!ARM64 && !ARC && !MICROBLAZE && !OPENRISC
 	default y
Signature · 1 line
2.14.1
#10 lore raw

PATCH 4/20 misc: atmel: Remove CPU_AT32AP700X (AVR32) reference

In reply to Ulf Magnusson

The CPU_AT32AP700X symbol symbol went away when when AVR32 was removed in commit 26202873bb51 ("avr32: remove support for AVR32 architecture")

Remove the prompt from ATMEL_TCB_CLKSRC_BLOCK. The prompt condition could never be satisfied now.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/misc/Kconfig | 1 -
 1 file changed, 1 deletion(-)
drivers/misc/Kconfig +0 -1
index 03605f8fc0dc..5d713008749b 100644
@@ -75,7 +75,6 @@ config ATMEL_TCB_CLKSRC
 config ATMEL_TCB_CLKSRC_BLOCK
 	int
 	depends on ATMEL_TCB_CLKSRC
-	prompt "TC Block" if CPU_AT32AP700X
 	default 0
 	range 0 1
 	help
Signature · 1 line
2.14.1
#11 lore raw

PATCH 5/20 rtc: at32ap700x: Remove PLATFORM_AT32AP dependency

In reply to Ulf Magnusson

The PLATFORM_AT32AP symbol went away when AVR32 was removed in commit 26202873bb51 ("avr32: remove support for AVR32 architecture").

Remove the RTC_DRV_AT32AP700X dependency from RTC_DRV_AT32AP700X.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/rtc/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rtc/Kconfig +1 -1
index b71cc7b5fbc9..1f5f3bd82ab6 100644
@@ -1404,7 +1404,7 @@ config RTC_DRV_PL031
 
 config RTC_DRV_AT32AP700X
 	tristate "AT32AP700X series RTC"
-	depends on PLATFORM_AT32AP || COMPILE_TEST
+	depends on COMPILE_TEST
 	help
 	  Driver for the internal RTC (Realtime Clock) on Atmel AVR32
 	  AT32AP700x family processors.
Signature · 1 line
2.14.1
#12 lore raw

Re: [PATCH 05/20] rtc: at32ap700x: Remove PLATFORM_AT32AP dependency

Quoted 28 lines · Ulf Magnusson

The PLATFORM_AT32AP symbol went away when AVR32 was removed in commit 26202873bb51 ("avr32: remove support for AVR32 architecture").

Remove the RTC_DRV_AT32AP700X dependency from RTC_DRV_AT32AP700X.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/rtc/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rtc/Kconfig +1 -2
index b71cc7b5fbc9..1f5f3bd82ab6 100644
@@ -1404,7 +1404,7 @@ config RTC_DRV_PL031
 
 config RTC_DRV_AT32AP700X
        tristate "AT32AP700X series RTC"
-       depends on PLATFORM_AT32AP || COMPILE_TEST
+       depends on COMPILE_TEST
        help
          Driver for the internal RTC (Realtime Clock) on Atmel AVR32
          AT32AP700x family processors.
--
Quoted 2 lines

2.14.1

RTC_DRV_AT32AP700X could probably be removed along with ./drivers/rtc/rtc-at32ap700x.c as well.

Cheers,
Ulf

#13 lore raw

Re: [PATCH 05/20] rtc: at32ap700x: Remove PLATFORM_AT32AP dependency

Hi,

Quoted 21 lines · Ulf Magnusson
Quoted 20 lines · Ulf Magnusson

The PLATFORM_AT32AP symbol went away when AVR32 was removed in commit 26202873bb51 ("avr32: remove support for AVR32 architecture").

Remove the RTC_DRV_AT32AP700X dependency from RTC_DRV_AT32AP700X.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/rtc/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rtc/Kconfig +0 -0
index b71cc7b5fbc9..1f5f3bd82ab6 100644
@@ -1404,7 +1404,7 @@ config RTC_DRV_PL031
 
Quoted 14 lines
Quoted 10 lines
drivers/rtc/Kconfig +1 -2
 config RTC_DRV_AT32AP700X
        tristate "AT32AP700X series RTC"
-       depends on PLATFORM_AT32AP || COMPILE_TEST
+       depends on COMPILE_TEST
        help
          Driver for the internal RTC (Realtime Clock) on Atmel AVR32
          AT32AP700x family processors.
--

2.14.1

RTC_DRV_AT32AP700X could probably be removed along with ./drivers/rtc/rtc-at32ap700x.c as well.

Note that the driver is already removed in Linus' tree.

Signature · 3 lines
Alexandre Belloni, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com
#14 lore raw

Re: [PATCH 05/20] rtc: at32ap700x: Remove PLATFORM_AT32AP dependency

Quoted 39 lines · Alexandre Belloni

Hi,

Quoted 21 lines · Ulf Magnusson
Quoted 20 lines · Ulf Magnusson

The PLATFORM_AT32AP symbol went away when AVR32 was removed in commit 26202873bb51 ("avr32: remove support for AVR32 architecture").

Remove the RTC_DRV_AT32AP700X dependency from RTC_DRV_AT32AP700X.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/rtc/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/rtc/Kconfig +0 -0
index b71cc7b5fbc9..1f5f3bd82ab6 100644
@@ -1404,7 +1404,7 @@ config RTC_DRV_PL031
 
Quoted 14 lines
Quoted 10 lines
drivers/rtc/Kconfig +1 -2
 config RTC_DRV_AT32AP700X
        tristate "AT32AP700X series RTC"
-       depends on PLATFORM_AT32AP || COMPILE_TEST
+       depends on COMPILE_TEST
        help
          Driver for the internal RTC (Realtime Clock) on Atmel AVR32
          AT32AP700x family processors.
--

2.14.1

RTC_DRV_AT32AP700X could probably be removed along with ./drivers/rtc/rtc-at32ap700x.c as well.

Quoted 7 lines

Note that the driver is already removed in Linus' tree.

--
Alexandre Belloni, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering
http://bootlin.com

That got rid of the config symbol too. This patch is moot then.

Cheers,
Ulf

#15 lore raw

PATCH 6/20 riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select

In reply to Ulf Magnusson

The ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE symbol was removed in commit 51a021244b9d ("atomic64: no need for CONFIG_ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE").

Remove the ARCH_HAS_ATOMIC64_DEC_IS_POSITIVE select from RISCV.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -1
index b6722c246d9c..ff69c77b9e78 100644
@@ -8,7 +8,6 @@ config RISCV
 	select OF
 	select OF_EARLY_FLATTREE
 	select OF_IRQ
-	select ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE
 	select ARCH_WANT_FRAME_POINTERS
 	select CLONE_BACKWARDS
 	select COMMON_CLK
Signature · 1 line
2.14.1
Reviewed-by on v1 6/20
#16 lore raw

Re: [PATCH 06/20] riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select

Quoted 6 lines · Ulf Magnusson

The ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE symbol was removed in commit 51a021244b9d ("atomic64: no need for CONFIG_ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE").

Remove the ARCH_HAS_ATOMIC64_DEC_IS_POSITIVE select from RISCV.

Quoted 8 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)

Looks good.

arch/riscv/Kconfig config RISCV · line 13

Quoted 13 lines
arch/riscv/Kconfig +0 -1
index b6722c246d9c..ff69c77b9e78 100644
@@ -8,7 +8,6 @@ config RISCV
 	select OF
 	select OF_EARLY_FLATTREE
 	select OF_IRQ
-	select ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE
 	select ARCH_WANT_FRAME_POINTERS
 	select CLONE_BACKWARDS
 	select COMMON_CLK

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 819 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-riscv/attachments/20180205/3c543c9a/attachment.sig>

Reviewed-by on v1 6/20
#17 lore raw

Re: [PATCH 06/20] riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select

In reply to Ulf Magnusson
Quoted 20 lines · ulfalizer at gmail.com

The ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE symbol was removed in commit 51a021244b9d ("atomic64: no need for CONFIG_ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE").

Remove the ARCH_HAS_ATOMIC64_DEC_IS_POSITIVE select from RISCV.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -0
index b6722c246d9c..ff69c77b9e78 100644

arch/riscv/Kconfig config RISCV · line 13

Quoted 7 lines
arch/riscv/Kconfig +0 -1
@@ -8,7 +8,6 @@ config RISCV
 	select OF
 	select OF_EARLY_FLATTREE
 	select OF_IRQ
-	select ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE
 	select ARCH_WANT_FRAME_POINTERS
 	select CLONE_BACKWARDS
 	select COMMON_CLK

Thanks!

#18 lore raw

Re: [PATCH 06/20] riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select

Quoted 34 lines · Palmer Dabbelt
Quoted 22 lines · ulfalizer at gmail.com

The ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE symbol was removed in commit 51a021244b9d ("atomic64: no need for CONFIG_ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE").

Remove the ARCH_HAS_ATOMIC64_DEC_IS_POSITIVE select from RISCV.

Discovered with the

https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -0
index b6722c246d9c..ff69c77b9e78 100644

arch/riscv/Kconfig config RISCV · line 13

Quoted 7 lines
arch/riscv/Kconfig +0 -1
@@ -8,7 +8,6 @@ config RISCV
        select OF
        select OF_EARLY_FLATTREE
        select OF_IRQ
-       select ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE
        select ARCH_WANT_FRAME_POINTERS
        select CLONE_BACKWARDS
        select COMMON_CLK

Thanks!

Quoted 1 line

I didn't have a particular tree in mind for these patches by the way, so feel free to pull it in yourselves.

Cheers,
Ulf

#19 lore raw

PATCH 7/20 riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

In reply to Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -1
index ff69c77b9e78..716e90e60e5c 100644
@@ -19,7 +19,6 @@ config RISCV
 	select GENERIC_STRNLEN_USER
 	select GENERIC_SMP_IDLE_THREAD
 	select GENERIC_ATOMIC64 if !64BIT || !RISCV_ISA_A
-	select ARCH_WANT_OPTIONAL_GPIOLIB
 	select HAVE_MEMBLOCK
 	select HAVE_MEMBLOCK_NODE_MAP
 	select HAVE_DMA_API_DEBUG
Signature · 1 line
2.14.1
#20 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

Quoted 32 lines · Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -2
index ff69c77b9e78..716e90e60e5c 100644
@@ -19,7 +19,6 @@ config RISCV
        select GENERIC_STRNLEN_USER
        select GENERIC_SMP_IDLE_THREAD
        select GENERIC_ATOMIC64 if !64BIT || !RISCV_ISA_A
-       select ARCH_WANT_OPTIONAL_GPIOLIB
        select HAVE_MEMBLOCK
        select HAVE_MEMBLOCK_NODE_MAP
        select HAVE_DMA_API_DEBUG
--
Quoted 2 lines

2.14.1

There's also a select of a globally undefined symbol RISCV_IRQ_INTC here. I couldn't dig up any historical or unmerged definition for it either. Googling just gets you the select.

Cheers,
Ulf

Reviewed-by
#21 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

Quoted 31 lines · ulfalizer at gmail.com
Quoted 30 lines · Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -1
index ff69c77b9e78..716e90e60e5c 100644
@@ -19,7 +19,6 @@ config RISCV
        select GENERIC_STRNLEN_USER
        select GENERIC_SMP_IDLE_THREAD
        select GENERIC_ATOMIC64 if !64BIT || !RISCV_ISA_A
-       select ARCH_WANT_OPTIONAL_GPIOLIB
        select HAVE_MEMBLOCK
        select HAVE_MEMBLOCK_NODE_MAP
Quoted 8 lines
Quoted 4 lines
arch/riscv/Kconfig +0 -1
        select HAVE_DMA_API_DEBUG
--

2.14.1

There's also a select of a globally undefined symbol RISCV_IRQ_INTC here. I couldn't dig up any historical or unmerged definition for it either. Googling just gets you the select.

It looks like this slipped through, the symbol has been renamed RISCV_INTC. A patch to remove it is

(and like the other ones, if you want it through my tree just say something)

#22 lore raw

PATCH 1/1 riscv: kconfig: Remove RISCV_IRQ_INTC select

The RISCV_IRQ_INTC configuration symbol is undefined, but RISCV selects it. Quoting Palmer Dabbelt:

It looks like this slipped through, the symbol has been renamed RISCV_INTC.

No RISCV_INTC configuration symbol has been merged either. Just remove the RISCV_IRQ_INTC select for now.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -1
index b6722c246d9c..2be44b6651c5 100644
@@ -34,7 +34,6 @@ config RISCV
 	select HAVE_ARCH_TRACEHOOK
 	select MODULES_USE_ELF_RELA if MODULES
 	select THREAD_INFO_IN_TASK
-	select RISCV_IRQ_INTC
 	select RISCV_TIMER
 
 config MMU
Signature · 1 line
2.14.1
#23 lore raw

Re: [PATCH] riscv: kconfig: Remove RISCV_IRQ_INTC select

Quoted 19 lines · ulfalizer at gmail.com

The RISCV_IRQ_INTC configuration symbol is undefined, but RISCV selects it. Quoting Palmer Dabbelt:

It looks like this slipped through, the symbol has been renamed RISCV_INTC.

No RISCV_INTC configuration symbol has been merged either. Just remove the RISCV_IRQ_INTC select for now.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -0
index b6722c246d9c..2be44b6651c5 100644

arch/riscv/Kconfig config RISCV · line 39

Quoted 7 lines
arch/riscv/Kconfig +0 -1
@@ -34,7 +34,6 @@ config RISCV
 	select HAVE_ARCH_TRACEHOOK
 	select MODULES_USE_ELF_RELA if MODULES
 	select THREAD_INFO_IN_TASK
-	select RISCV_IRQ_INTC
 	select RISCV_TIMER
 
 config MMU

Thanks! I'll take this (and the other relevant ones) into the RISC-V tree.

#24 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

In reply to Ulf Magnusson
Quoted 10 lines · Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Quoted 5 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

This should probaly be replaced with
select GPIOLIB

Yours,
Linus Walleij

#25 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

Quoted 16 lines · Linus Walleij
Quoted 14 lines · Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Quoted 7 lines
Quoted 1 line

This should probaly be replaced with
select GPIOLIB

Yours,
Linus Walleij

That'd make more sense if GPIOLIB is actually used. The select has been a no-op since mid-2016 at least, but maybe something else selects GPIOLIB.

Cheers,
Ulf

#26 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

Quoted 25 lines · Ulf Magnusson
Quoted 16 lines · Linus Walleij
Quoted 14 lines · Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Quoted 7 lines
Quoted 1 line

This should probaly be replaced with
select GPIOLIB

Yours,
Linus Walleij

Quoted 6 lines

That'd make more sense if GPIOLIB is actually used. The select has been a no-op since mid-2016 at least, but maybe something else selects GPIOLIB.

Cheers,
Ulf

Just replace it with your own patch once you've figured out the proper fix. More efficient than having a middle man making (sometimes poor) guesses.

Cheers,
Ulf

#27 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

Quoted 30 lines · Ulf Magnusson
Quoted 29 lines · Ulf Magnusson
Quoted 16 lines · Linus Walleij
Quoted 14 lines · Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Quoted 7 lines
Quoted 1 line

This should probaly be replaced with
select GPIOLIB

Yours,
Linus Walleij

That'd make more sense if GPIOLIB is actually used. The select has been a no-op since mid-2016 at least, but maybe something else selects GPIOLIB.

Quoted 6 lines
Quoted 2 lines

Cheers,
Ulf

Just replace it with your own patch once you've figured out the proper fix. More efficient than having a middle man making (sometimes poor) guesses.

I think it's just a side effect of the new RISCV arch being developed outside of the mainline kernel. And they may have had a gpio driver outside of the mainline as well.

Normally we don't encourage out-of-tree development, but new archs are a bit of an exception since they have to be merged as a big piece of upfront code at some point...

Anyways, if they don't have any GPIO in their machine yet, let's let the RISCV maintainers apply this.

Yours,
Linus Walleij

Reviewed-by on v1 7/20
#28 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

In reply to Ulf Magnusson
Quoted 10 lines · Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Quoted 5 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

RISCV maintainers: please apply and carry this patch.

Yours,
Linus Walleij

#29 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

Quoted 16 lines · linus.walleij at linaro.org
Quoted 14 lines · Ulf Magnusson

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Quoted 5 lines
Quoted 1 line

RISCV maintainers: please apply and carry this patch.

OK, I'll atke it into my tree. Thanks!

Reviewed-by on v1 7/20
#30 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

In reply to Ulf Magnusson
Quoted 24 lines · ulfalizer at gmail.com

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -0
index ff69c77b9e78..716e90e60e5c 100644

arch/riscv/Kconfig config RISCV · line 24

Quoted 7 lines
arch/riscv/Kconfig +0 -1
@@ -19,7 +19,6 @@ config RISCV
 	select GENERIC_STRNLEN_USER
 	select GENERIC_SMP_IDLE_THREAD
 	select GENERIC_ATOMIC64 if !64BIT || !RISCV_ISA_A
-	select ARCH_WANT_OPTIONAL_GPIOLIB
 	select HAVE_MEMBLOCK
 	select HAVE_MEMBLOCK_NODE_MAP
 	select HAVE_DMA_API_DEBUG

Thanks!

(If you want these through my tree, just say something.)

#31 lore raw

Re: [PATCH 07/20] riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select

Quoted 35 lines · Palmer Dabbelt
Quoted 24 lines · ulfalizer at gmail.com

The ARCH_WANT_OPTIONAL_GPIOLIB symbol was removed in commit 65053e1a7743 ("gpio: delete ARCH_[WANTS_OPTIONAL|REQUIRE]_GPIOLIB"). GPIOLIB should just be selected explicitly if needed.

Remove the ARCH_WANT_OPTIONAL_GPIOLIB select from RISCV.

See commit 0145071b3314 ("x86: Do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") and commit da9a1c6767 ("arm64: do away with ARCH_[WANT_OPTIONAL|REQUIRE]_GPIOLIB") as well.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/riscv/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/riscv/Kconfig +0 -0
index ff69c77b9e78..716e90e60e5c 100644

arch/riscv/Kconfig config RISCV · line 24

Quoted 7 lines
arch/riscv/Kconfig +0 -1
@@ -19,7 +19,6 @@ config RISCV
 	select GENERIC_STRNLEN_USER
 	select GENERIC_SMP_IDLE_THREAD
 	select GENERIC_ATOMIC64 if !64BIT || !RISCV_ISA_A
-	select ARCH_WANT_OPTIONAL_GPIOLIB
 	select HAVE_MEMBLOCK
 	select HAVE_MEMBLOCK_NODE_MAP
 	select HAVE_DMA_API_DEBUG

Thanks!

Quoted 1 line

(If you want these through my tree, just say something.)

I didn't have a particular tree in mind for these patches, so feel free to take it.

Quoted 2 lines

Cheers,
Ulf

#32 lore raw

PATCH 8/20 s390/kconfig: Remove ARCH_WANTS_PROT_NUMA_PROT_NONE select

In reply to Ulf Magnusson

The ARCH_WANTS_PROT_NUMA_PROT_NONE symbol was removed by commit 6a33979d5bd7 ("mm: remove misleading ARCH_USES_NUMA_PROT_NONE"), but S390 still selects it.

Remove the ARCH_WANTS_PROT_NUMA_PROT_NONE select from the S390 symbol.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/s390/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/s390/Kconfig +0 -1
index 0105ce28e246..bb090b7992e4 100644
@@ -112,7 +112,6 @@ config S390
 	select ARCH_USE_BUILTIN_BSWAP
 	select ARCH_USE_CMPXCHG_LOCKREF
 	select ARCH_WANTS_DYNAMIC_TASK_STRUCT
-	select ARCH_WANTS_PROT_NUMA_PROT_NONE
 	select ARCH_WANTS_UBSAN_NO_NULL
 	select ARCH_WANT_IPC_PARSE_VERSION
 	select BUILDTIME_EXTABLE_SORT
Signature · 1 line
2.14.1
#33 lore raw

Re: [PATCH 08/20] s390/kconfig: Remove ARCH_WANTS_PROT_NUMA_PROT_NONE select

Quoted 6 lines · Ulf Magnusson

The ARCH_WANTS_PROT_NUMA_PROT_NONE symbol was removed by commit 6a33979d5bd7 ("mm: remove misleading ARCH_USES_NUMA_PROT_NONE"), but S390 still selects it.

Remove the ARCH_WANTS_PROT_NUMA_PROT_NONE select from the S390 symbol.

Quoted 8 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/s390/Kconfig | 1 -
 1 file changed, 1 deletion(-)

Applied, thanks!

#34 lore raw

PATCH 9/20 clk: sunxi-ng: Remove SUNXI_CCU_* selects

In reply to Ulf Magnusson

The selected symbols were removed in commit 06e226c7fb23 ("clk:
sunxi-ng: Move all clock types to a library").

Remove the selects from SUN4I_A10_CCU.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/clk/sunxi-ng/Kconfig | 7 -------
 1 file changed, 7 deletions(-)
drivers/clk/sunxi-ng/Kconfig +0 -7
index 6427d0ebe2de..33168f94ee39 100644
@@ -13,13 +13,6 @@ config SUN50I_A64_CCU
 
 config SUN4I_A10_CCU
 	bool "Support for the Allwinner A10/A20 CCU"
-	select SUNXI_CCU_DIV
-	select SUNXI_CCU_MULT
-	select SUNXI_CCU_NK
-	select SUNXI_CCU_NKM
-	select SUNXI_CCU_NM
-	select SUNXI_CCU_MP
-	select SUNXI_CCU_PHASE
 	default MACH_SUN4I
 	default MACH_SUN7I
 	depends on MACH_SUN4I || MACH_SUN7I || COMPILE_TEST
Signature · 1 line
2.14.1
#35 lore raw

Re: [PATCH 09/20] clk: sunxi-ng: Remove SUNXI_CCU_* selects

Hi Ulf,

Quoted 10 lines · Ulf Magnusson

The selected symbols were removed in commit 06e226c7fb23 ("clk:
sunxi-ng: Move all clock types to a library").

Remove the selects from SUN4I_A10_CCU.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

This has already been submitted by Corentin:
https://patchwork.kernel.org/patch/10175435/

Thanks!
Maxime

Signature · 10 lines
Maxime Ripard, Bootlin (formerly Free Electrons)
Embedded Linux and Kernel engineering
http://bootlin.com
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20180205/5dadb51f/attachment.sig>
#36 lore raw

PATCH 10/20 lib/Kconfig: Remove leftover select of GENERIC_IO

In reply to Ulf Magnusson

GENERIC_IO was removed by commit 9de8da47742b ("kconfig: kill off GENERIC_IO option"), but HAS_IOMEM still selects it.

Remove the select of GENERIC_IO from HAS_IOMEM.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 lib/Kconfig | 1 -
 1 file changed, 1 deletion(-)
lib/Kconfig +0 -1
index e96089499371..8ce8bcad6eff 100644
@@ -396,7 +396,6 @@ config ASSOCIATIVE_ARRAY
 config HAS_IOMEM
 	bool
 	depends on !NO_IOMEM
-	select GENERIC_IO
 	default y
 
 config HAS_IOPORT_MAP
Signature · 1 line
2.14.1
Acked-by on v1 10/20
#37 lore raw

Re: [PATCH 10/20] lib/Kconfig: Remove leftover select of GENERIC_IO

Quoted 5 lines · Ulf Magnusson

GENERIC_IO was removed by commit 9de8da47742b ("kconfig: kill off GENERIC_IO option"), but HAS_IOMEM still selects it.

Remove the select of GENERIC_IO from HAS_IOMEM.

Quoted 8 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 lib/Kconfig | 1 -
 1 file changed, 1 deletion(-)

There was another fix[1] for this but it seems to have gotten lost.

Rob

[1] https://lkml.org/lkml/2018/1/17/924

Acked-by on v1 10/20
#38 lore raw

Re: [PATCH 10/20] lib/Kconfig: Remove leftover select of GENERIC_IO

In reply to Ulf Magnusson
Quoted 8 lines · Ulf Magnusson

GENERIC_IO was removed by commit 9de8da47742b ("kconfig: kill off GENERIC_IO option"), but HAS_IOMEM still selects it.

Remove the select of GENERIC_IO from HAS_IOMEM.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Since I also stumbled over this and tried to fix it,

Quoted 18 lines
---
  lib/Kconfig | 1 -
  1 file changed, 1 deletion(-)
lib/Kconfig +0 -1
index e96089499371..8ce8bcad6eff 100644
@@ -396,7 +396,6 @@ config ASSOCIATIVE_ARRAY
  config HAS_IOMEM
  	bool
  	depends on !NO_IOMEM
-	select GENERIC_IO
  	default y
 
  config HAS_IOPORT_MAP
#39 lore raw

Re: [PATCH 10/20] lib/Kconfig: Remove leftover select of GENERIC_IO

In reply to Ulf Magnusson
Quoted 5 lines · Ulf Magnusson

GENERIC_IO was removed by commit 9de8da47742b ("kconfig: kill off GENERIC_IO option"), but HAS_IOMEM still selects it.

Remove the select of GENERIC_IO from HAS_IOMEM.

Quoted 8 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 lib/Kconfig | 1 -
 1 file changed, 1 deletion(-)

Who do you intend to apply this because people keep sending the same fix?

Rob

#40 lore raw

Re: [PATCH 10/20] lib/Kconfig: Remove leftover select of GENERIC_IO

Quoted 10 lines · Rob Herring
Quoted 9 lines · Ulf Magnusson

GENERIC_IO was removed by commit 9de8da47742b ("kconfig: kill off GENERIC_IO option"), but HAS_IOMEM still selects it.

Remove the select of GENERIC_IO from HAS_IOMEM.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Quoted 8 lines
Quoted 4 lines
---
 lib/Kconfig | 1 -
 1 file changed, 1 deletion(-)

Who do you intend to apply this because people keep sending the same fix?

Rob

I didn't have a particular tree in mind. I imagined people would pick up patches into the trees where they make the most sense.

I just turned it into a patchset to keep track of it easier myself. Looks like that might have caused some confusion in retrospect.

Cheers,
Ulf

#41 lore raw

PATCH 11/20 usb: gadget: udc: Remove USB_GADGET_DUALSPEED select

In reply to Ulf Magnusson

USB_GADGET_DUALSPEED was removed by commit 85b8614d7223 ("usb: gadget:
get rid of USB_GADGET_{DUAL,SUPER}SPEED"), but the USB_SNP_UDC_PLAT symbol still selects it.

Remove the USB_GADGET_DUALSPEED select from USB_SNP_UDC_PLAT.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/usb/gadget/udc/Kconfig | 1 -
 1 file changed, 1 deletion(-)
drivers/usb/gadget/udc/Kconfig +0 -1
index 1e9567091d86..0875d38476ee 100644
@@ -274,7 +274,6 @@ config USB_SNP_UDC_PLAT
 	tristate "Synopsys USB 2.0 Device controller"
 	depends on USB_GADGET && OF && HAS_DMA
 	depends on EXTCON || EXTCON=n
-	select USB_GADGET_DUALSPEED
 	select USB_SNP_CORE
 	default ARCH_BCM_IPROC
 	help
#42 lore raw

PATCH 12/20 ARM: debug: Remove ARCH_MSM dep. from UNCOMPRESS_INCLUDE

In reply to Ulf Magnusson

The ARCH_MSM symbol was removed in commit c0c89fafa289 ("ARM: Remove mach-msm and associated ARM architecture code"), but a default in UNCOMPRESS_INCLUDE still depends on it.

Remove the ARCH_MSM dependency from the default in UNCOMPRESS_INCLUDE.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/arm/Kconfig.debug | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/Kconfig.debug +1 -1
index 78a647080ebc..20c601a045d2 100644
@@ -1825,7 +1825,7 @@ config DEBUG_UNCOMPRESS
 
 config UNCOMPRESS_INCLUDE
 	string
-	default "debug/uncompress.h" if ARCH_MULTIPLATFORM || ARCH_MSM || \
+	default "debug/uncompress.h" if ARCH_MULTIPLATFORM || \
 					PLAT_SAMSUNG || ARM_SINGLE_ARMV7M
 	default "mach/uncompress.h"
 
Signature · 1 line
2.14.1
#43 lore raw

Re: [PATCH 12/20] ARM: debug: Remove ARCH_MSM dep. from UNCOMPRESS_INCLUDE

Quoted 29 lines · Ulf Magnusson

The ARCH_MSM symbol was removed in commit c0c89fafa289 ("ARM: Remove mach-msm and associated ARM architecture code"), but a default in UNCOMPRESS_INCLUDE still depends on it.

Remove the ARCH_MSM dependency from the default in UNCOMPRESS_INCLUDE.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 arch/arm/Kconfig.debug | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/Kconfig.debug +1 -2
index 78a647080ebc..20c601a045d2 100644
@@ -1825,7 +1825,7 @@ config DEBUG_UNCOMPRESS
 
 config UNCOMPRESS_INCLUDE
        string
-       default "debug/uncompress.h" if ARCH_MULTIPLATFORM || ARCH_MSM || \
+       default "debug/uncompress.h" if ARCH_MULTIPLATFORM || \
                                        PLAT_SAMSUNG || ARM_SINGLE_ARMV7M
        default "mach/uncompress.h"
 
--
Quoted 2 lines

2.14.1

These globally undefined symbols (with no historical definitions that I could find) are also referenced in this file:

DEBUG_OMAP4UART2
DEBUG_OMAP4UART1
DEBUG_OMAP5UART1
DEBUG_OMAP5UART2
DEBUG_OMAP3UART2
DEBUG_OMAP3UART1

Cheers,
Ulf

#44 lore raw

PATCH 13/20 ore: Remove PNFS_OBJLAYOUT dependency

In reply to Ulf Magnusson

The PNFS_OBJLAYOUT symbol was removed in commit 6d22323b2e9f ("nfs:
remove the objlayout driver"), but the ORE symbol still depends on it.

Remove the PNFS_OBJLAYOUT dependency from ORE.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 fs/exofs/Kconfig.ore | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
fs/exofs/Kconfig.ore +1 -1
index 2daf2329c28d..73cab8d92dfd 100644
@@ -7,7 +7,7 @@
 # selected by any of the users.
 config ORE
 	tristate
-	depends on EXOFS_FS || PNFS_OBJLAYOUT
+	depends on EXOFS_FS
 	select ASYNC_XOR
 	select RAID6_PQ
 	select ASYNC_PQ
Signature · 1 line
2.14.1
#45 lore raw

PATCH 14/20 iio: gyro: mpu3050: Remove INPUT_MPU3050 dep. from I2C

In reply to Ulf Magnusson

The INPUT_MPU3050 symbol was removed in commit b1fe0cf06f92 ("Input:
delete MPU3050 driver").

Remove the negative dependency on it from the MPU3050_I2C symbol. It is always satisfied now.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/iio/gyro/Kconfig | 1 -
 1 file changed, 1 deletion(-)
drivers/iio/gyro/Kconfig +0 -1
index 3126cf05e6b9..b702b9176006 100644
@@ -92,7 +92,6 @@ config MPU3050
 
 config MPU3050_I2C
 	tristate "Invensense MPU3050 devices on I2C"
-	depends on !(INPUT_MPU3050=y || INPUT_MPU3050=m)
 	depends on I2C
 	select MPU3050
 	select REGMAP_I2C
Signature · 1 line
2.14.1
Reviewed-by on v1 14/20
#46 lore raw

Re: [PATCH 14/20] iio: gyro: mpu3050: Remove INPUT_MPU3050 dep. from I2C

Quoted 11 lines · Ulf Magnusson

The INPUT_MPU3050 symbol was removed in commit b1fe0cf06f92 ("Input:
delete MPU3050 driver").

Remove the negative dependency on it from the MPU3050_I2C symbol. It is always satisfied now.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Jonathan, please pick this up to IIO!

Yours,
Linus Walleij

#47 lore raw

PATCH 15/20 phy: Remove SOC_EXYNOS4212 dep. from PHY_EXYNOS4X12_USB

In reply to Ulf Magnusson

Exynos4212 support was removed by commit bca9085e0ae9 ("ARM: dts:
exynos: remove Exynos4212 support (dead code)").

Remove the SOC_EXYNOS4212 dependency from PHY_EXYNOS4X12_USB.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/phy/samsung/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/phy/samsung/Kconfig +1 -1
index b7e0645a7bd9..2a5d33cb0e7e 100644
@@ -49,7 +49,7 @@ config PHY_EXYNOS4210_USB2
 config PHY_EXYNOS4X12_USB2
 	bool
 	depends on PHY_SAMSUNG_USB2
-	default SOC_EXYNOS3250 || SOC_EXYNOS4212 || SOC_EXYNOS4412
+	default SOC_EXYNOS3250 || SOC_EXYNOS4412
 
 config PHY_EXYNOS5250_USB2
 	bool
Signature · 1 line
2.14.1
Acked-by on v1 15/20
#48 lore raw

Re: [PATCH 15/20] phy: Remove SOC_EXYNOS4212 dep. from PHY_EXYNOS4X12_USB

Quoted 5 lines · Ulf Magnusson

Exynos4212 support was removed by commit bca9085e0ae9 ("ARM: dts:
exynos: remove Exynos4212 support (dead code)").

Remove the SOC_EXYNOS4212 dependency from PHY_EXYNOS4X12_USB.

Quoted 8 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/phy/samsung/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Best regards,
Krzysztof

#49 lore raw

Re: [PATCH 15/20] phy: Remove SOC_EXYNOS4212 dep. from PHY_EXYNOS4X12_USB

Quoted 10 lines · Krzysztof Kozlowski
Quoted 9 lines · Ulf Magnusson

Exynos4212 support was removed by commit bca9085e0ae9 ("ARM: dts:
exynos: remove Exynos4212 support (dead code)").

Remove the SOC_EXYNOS4212 dependency from PHY_EXYNOS4X12_USB.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Quoted 6 lines
Quoted 4 lines
---
 drivers/phy/samsung/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

merged, thanks!

-Kishon

Quoted 4 lines

Best regards,
Krzysztof

#50 lore raw

PATCH 16/20 auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.

In reply to Ulf Magnusson

The MIPS_SEAD3 symbol was removed in commit 64601cb1343f ("leds: Remove SEAD-3 driver").

Remove the MIPS_SEAD3 dependency from IMG_ASCII_LCD.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/Kconfig +1 -1
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
 	tristate "Imagination Technologies ASCII LCD Display"
 	depends on HAS_IOMEM
-	default y if MIPS_MALTA || MIPS_SEAD3
+	default y if MIPS_MALTA
 	select SYSCON
 	help
 	  Enable this to support the simple ASCII LCD displays found on
Signature · 1 line
2.14.1
#51 lore raw

Re: [PATCH 16/20] auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.

Quoted 28 lines · Ulf Magnusson

The MIPS_SEAD3 symbol was removed in commit 64601cb1343f ("leds: Remove SEAD-3 driver").

Remove the MIPS_SEAD3 dependency from IMG_ASCII_LCD.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/Kconfig +1 -2
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
        tristate "Imagination Technologies ASCII LCD Display"
        depends on HAS_IOMEM
-       default y if MIPS_MALTA || MIPS_SEAD3
+       default y if MIPS_MALTA
        select SYSCON
        help
          Enable this to support the simple ASCII LCD displays found on
--
Quoted 2 lines

2.14.1

The SYSCON symbol selected here is also globally undefined. Maybe MFD_SYSCON was intended.

Cheers,
Ulf

#52 lore raw

Re: [PATCH 16/20] auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.

Quoted 26 lines · Ulf Magnusson
Quoted 25 lines · Ulf Magnusson

The MIPS_SEAD3 symbol was removed in commit 64601cb1343f ("leds: Remove SEAD-3 driver").

Remove the MIPS_SEAD3 dependency from IMG_ASCII_LCD.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/Kconfig +1 -1
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
        tristate "Imagination Technologies ASCII LCD Display"
        depends on HAS_IOMEM
-       default y if MIPS_MALTA || MIPS_SEAD3
+       default y if MIPS_MALTA
        select SYSCON
Quoted 8 lines
Quoted 5 lines
drivers/auxdisplay/Kconfig +0 -1
        help
          Enable this to support the simple ASCII LCD displays found on
--

2.14.1

The SYSCON symbol selected here is also globally undefined. Maybe MFD_SYSCON was intended.

Yes.

from 2018-JAN-18, Corentin Labbe:
https://marc.info/?l=linux-kernel&m=151630656607526&w=2

Signature · 1 line
~Randy
#53 lore raw

Re: [PATCH 16/20] auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.

In reply to Ulf Magnusson
Quoted 2 lines · Ulf Magnusson

The MIPS_SEAD3 symbol was removed in commit 64601cb1343f ("leds: Remove SEAD-3 driver").

Commit 64601cb1343f did not remove the MIPS_SEAD3 symbol, it seemed to a few months before, in 3f5f0a4475e1 ("MIPS: generic: Convert SEAD-3 to a generic board"). Can you please amend the description?

Thanks!
Miguel

Quoted 28 lines

Remove the MIPS_SEAD3 dependency from IMG_ASCII_LCD.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/Kconfig +1 -2
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
        tristate "Imagination Technologies ASCII LCD Display"
        depends on HAS_IOMEM
-       default y if MIPS_MALTA || MIPS_SEAD3
+       default y if MIPS_MALTA
        select SYSCON
        help
          Enable this to support the simple ASCII LCD displays found on
--

2.14.1

#54 lore raw

PATCH v2 1/1 auxdisplay: img-ascii-lcd: kconfig: Remove MIPS_SEAD3 reference

Commit 3f5f0a4475e1 ("MIPS: generic: Convert SEAD-3 to a generic board") removed the MIPS_SEAD3 symbol and moved the setting of IMG_ASCII_LCD to the board-sead-3.config defconfig file, but IMG_ASCII_LCD still references the removed MIPS_SEAD3 symbol.

Remove the reference.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
Changes in v2:

- The wrong commit was referenced in the commit message.

- Mention that IMG_ASCII_LCD is now set via the board-sead-3.config defconfig file.

- Clarify that Kconfig is involved. The previous commit title was "auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 reference".

 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/Kconfig +1 -1
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
 	tristate "Imagination Technologies ASCII LCD Display"
 	depends on HAS_IOMEM
-	default y if MIPS_MALTA || MIPS_SEAD3
+	default y if MIPS_MALTA
 	select SYSCON
 	help
 	  Enable this to support the simple ASCII LCD displays found on
Signature · 1 line
2.14.1
#55 lore raw

Re: [PATCH v2] auxdisplay: img-ascii-lcd: kconfig: Remove MIPS_SEAD3 reference

Quoted 15 lines · Ulf Magnusson

Commit 3f5f0a4475e1 ("MIPS: generic: Convert SEAD-3 to a generic board") removed the MIPS_SEAD3 symbol and moved the setting of IMG_ASCII_LCD to the board-sead-3.config defconfig file, but IMG_ASCII_LCD still references the removed MIPS_SEAD3 symbol.

Remove the reference.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
Changes in v2:

Quoted 8 lines

- The wrong commit was referenced in the commit message.

- Mention that IMG_ASCII_LCD is now set via the board-sead-3.config defconfig file.

- Clarify that Kconfig is involved. The previous commit title was "auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 reference".

Thanks for the new version!

Quoted 20 lines
 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/Kconfig +1 -2
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
        tristate "Imagination Technologies ASCII LCD Display"
        depends on HAS_IOMEM
-       default y if MIPS_MALTA || MIPS_SEAD3
+       default y if MIPS_MALTA
        select SYSCON
        help
          Enable this to support the simple ASCII LCD displays found on
--

2.14.1

#56 lore raw

Re: [PATCH v2] auxdisplay: img-ascii-lcd: kconfig: Remove MIPS_SEAD3 reference

Quoted 21 lines · Miguel Ojeda
Quoted 20 lines · Ulf Magnusson

Commit 3f5f0a4475e1 ("MIPS: generic: Convert SEAD-3 to a generic board") removed the MIPS_SEAD3 symbol and moved the setting of IMG_ASCII_LCD to the board-sead-3.config defconfig file, but IMG_ASCII_LCD still references the removed MIPS_SEAD3 symbol.

Remove the reference.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
Changes in v2:

- The wrong commit was referenced in the commit message.

- Mention that IMG_ASCII_LCD is now set via the board-sead-3.config defconfig file.

Quoted 7 lines
Quoted 3 lines

- Clarify that Kconfig is involved. The previous commit title was "auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 reference".

Thanks for the new version!

To clarify, I am picking this patch up for my queue (sorry for any confusion!).

Quoted 21 lines
Quoted 21 lines

 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/Kconfig +1 -2
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
        tristate "Imagination Technologies ASCII LCD Display"
        depends on HAS_IOMEM
-       default y if MIPS_MALTA || MIPS_SEAD3
+       default y if MIPS_MALTA
        select SYSCON
        help
          Enable this to support the simple ASCII LCD displays found on
--

2.14.1

#57 lore raw

Re: [PATCH 16/20] auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.

In reply to Ulf Magnusson
Quoted 4 lines · Ulf Magnusson

The MIPS_SEAD3 symbol was removed in commit 64601cb1343f ("leds: Remove SEAD-3 driver").

Remove the MIPS_SEAD3 dependency from IMG_ASCII_LCD.

Its not a dependency, just a default (presumably so that existing kernel configs get updated automatically).

Note that sead3 now uses MIPS_GENERIC (or more specifically CONFIG_LEGACY_BOARD_SEAD3), but this driver is enabled in the config snippets, e.g.:
arch/mips/configs/generic/board-sead-3.config arch/mips/configs/generic/board-boston.config

so indeed there is no real need to replace MIPS_SEAD3 with anything here.

Cheers
James

Quoted 26 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/auxdisplay/Kconfig +1 -2
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
        tristate "Imagination Technologies ASCII LCD Display"
        depends on HAS_IOMEM
-       default y if MIPS_MALTA || MIPS_SEAD3
+       default y if MIPS_MALTA
        select SYSCON
        help
          Enable this to support the simple ASCII LCD displays found on
--

2.14.1

Signature · 1 line
James Hogan
#58 lore raw

Re: [PATCH 16/20] auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.

Quoted 8 lines · James Hogan
Quoted 4 lines · Ulf Magnusson

The MIPS_SEAD3 symbol was removed in commit 64601cb1343f ("leds: Remove SEAD-3 driver").

Remove the MIPS_SEAD3 dependency from IMG_ASCII_LCD.

Its not a dependency, just a default (presumably so that existing kernel configs get updated automatically).

Note that it appears in the default's condition rather than as part of the default value.

Undefined Kconfig symbols always evaluate to n in a tristate state, regardless of whether they're given a user value in a .config file or not, so this is a no-op as far as Kconfig in concerned.

The motivation in the original commit message was a bit lackluster though. I've submitted a v2.

Quoted 22 lines

Note that sead3 now uses MIPS_GENERIC (or more specifically CONFIG_LEGACY_BOARD_SEAD3), but this driver is enabled in the config snippets, e.g.:
arch/mips/configs/generic/board-sead-3.config arch/mips/configs/generic/board-boston.config

so indeed there is no real need to replace MIPS_SEAD3 with anything here.

Cheers
James

Quoted 10 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/auxdisplay/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
Quoted 21 lines
Quoted 16 lines
drivers/auxdisplay/Kconfig +1 -2
index 2c2ed9cf8796..3cba78d36eec 100644
@@ -137,7 +137,7 @@ config CFAG12864B_RATE
 config IMG_ASCII_LCD
        tristate "Imagination Technologies ASCII LCD Display"
        depends on HAS_IOMEM
-       default y if MIPS_MALTA || MIPS_SEAD3
+       default y if MIPS_MALTA
        select SYSCON
        help
          Enable this to support the simple ASCII LCD displays found on
--

2.14.1

Signature · 1 line
James Hogan

Cheers,
Ulf

#59 lore raw

PATCH 17/20 media: sec: Remove PLAT_S5P dependency

In reply to Ulf Magnusson

The PLAT_S5P symbol was removed in commit d78c16ccde96 ("ARM: SAMSUNG:
Remove remaining legacy code").

Remove the PLAT_S5P dependency from VIDEO_SAMSUNG_S5P_CEC.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/media/platform/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/media/platform/Kconfig +1 -1
index 614fbef08ddc..8b4cd02e2938 100644
@@ -582,7 +582,7 @@ config CEC_GPIO
 
 config VIDEO_SAMSUNG_S5P_CEC
        tristate "Samsung S5P CEC driver"
-       depends on PLAT_S5P || ARCH_EXYNOS || COMPILE_TEST
+       depends on ARCH_EXYNOS || COMPILE_TEST
        select CEC_CORE
        select CEC_NOTIFIER
        ---help---
Signature · 1 line
2.14.1
#60 lore raw

PATCH 18/20 iwlwifi: fix malformed CONFIG_IWLWIFI_PCIE_RTPM default

In reply to Ulf Magnusson

'default false' should be 'default n', though they happen to have the same effect here, due to undefined symbols ('false' in this case) evaluating to n in a tristate sense.

Remove the default instead of changing it. bool and tristate symbols implicitly default to n.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/net/wireless/intel/iwlwifi/Kconfig | 1 -
 1 file changed, 1 deletion(-)
drivers/net/wireless/intel/iwlwifi/Kconfig +0 -1
index c5f2ddf9b0fe..e5a2fc738ac3 100644
@@ -91,7 +91,6 @@ config IWLWIFI_BCAST_FILTERING
 config IWLWIFI_PCIE_RTPM
        bool "Enable runtime power management mode for PCIe devices"
        depends on IWLMVM && PM && EXPERT
-       default false
        help
          Say Y here to enable runtime power management for PCIe
          devices.  If enabled, the device will go into low power mode
Signature · 1 line
2.14.1
#61 lore raw

Re: [PATCH 18/20] iwlwifi: fix malformed CONFIG_IWLWIFI_PCIE_RTPM default

Quoted 7 lines · Ulf Magnusson

'default false' should be 'default n', though they happen to have the same effect here, due to undefined symbols ('false' in this case) evaluating to n in a tristate sense.

Remove the default instead of changing it. bool and tristate symbols implicitly default to n.

Quoted 7 lines

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_und
efined.py
script.

---

Thanks, Ulf!

This is applied in our internal tree and will eventually reach the mainline, following our usual upstreaming process.

--
Cheers,
Luca.

#62 lore raw

PATCH 19/20 ARC: Fix malformed ARC_EMUL_UNALIGNED default

In reply to Ulf Magnusson

'default N' should be 'default n', though they happen to have the same effect here, due to undefined symbols (N in this case) evaluating to n in a tristate sense.

Remove the default from ARC_EMUL_UNALIGNED instead of changing it. bool and tristate symbols implicitly default to n.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Signed-off-by Ulf Magnusson <ulfalizer at gmail.com>
---
 arch/arc/Kconfig | 1 -
 1 file changed, 1 deletion(-)
arch/arc/Kconfig +0 -1
index f3a80cf164cc..d76bf4a83740 100644
@@ -484,7 +484,6 @@ config ARC_CURR_IN_REG
 
 config ARC_EMUL_UNALIGNED
 	bool "Emulate unaligned memory access (userspace only)"
-	default N
 	select SYSCTL_ARCH_UNALIGN_NO_WARN
 	select SYSCTL_ARCH_UNALIGN_ALLOW
 	depends on ISA_ARCOMPACT
Signature · 1 line
2.14.1
#63 lore raw

Re: [PATCH 19/20] ARC: Fix malformed ARC_EMUL_UNALIGNED default

Quoted 4 lines · Ulf Magnusson

'default N' should be 'default n', though they happen to have the same effect here, due to undefined symbols (N in this case) evaluating to n in a tristate sense.

Quoted 8 lines

Remove the default from ARC_EMUL_UNALIGNED instead of changing it. bool and tristate symbols implicitly default to n.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

Signed-off-by Ulf Magnusson <ulfalizer at gmail.com>

Added to ARC for-curr.

Thx,
-Vineet

Quoted 17 lines
---
  arch/arc/Kconfig | 1 -
  1 file changed, 1 deletion(-)
arch/arc/Kconfig +0 -1
index f3a80cf164cc..d76bf4a83740 100644
@@ -484,7 +484,6 @@ config ARC_CURR_IN_REG
 
  config ARC_EMUL_UNALIGNED
  	bool "Emulate unaligned memory access (userspace only)"
-	default N
  	select SYSCTL_ARCH_UNALIGN_NO_WARN
  	select SYSCTL_ARCH_UNALIGN_ALLOW
  	depends on ISA_ARCOMPACT
#64 lore raw

PATCH 20/20 x86/PCI: VMD: Fix malformed default

In reply to Ulf Magnusson

'default N' should be 'default n', though they happen to have the same effect here, due to undefined symbols (N in this case) evaluating to n in a tristate sense.

Remove the default from VMD instead of changing it. bool and tristate symbols implicitly default to n.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/pci/host/Kconfig | 1 -
 1 file changed, 1 deletion(-)
drivers/pci/host/Kconfig +0 -1
index a4ed7484d127..dc8a2a175f19 100644
@@ -215,7 +215,6 @@ config PCIE_TANGO_SMP8759
 config VMD
 	depends on PCI_MSI && X86_64 && SRCU
 	tristate "Intel Volume Management Device Driver"
-	default N
 	---help---
 	  Adds support for the Intel Volume Management Device (VMD). VMD is a
 	  secondary PCI host bridge that allows PCI Express root ports,
Signature · 1 line
2.14.1
#65 lore raw

Re: [PATCH 20/20] x86/PCI: VMD: Fix malformed default

Quoted 29 lines · Ulf Magnusson

'default N' should be 'default n', though they happen to have the same effect here, due to undefined symbols (N in this case) evaluating to n in a tristate sense.

Remove the default from VMD instead of changing it. bool and tristate symbols implicitly default to n.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/pci/host/Kconfig | 1 -
 1 file changed, 1 deletion(-)
drivers/pci/host/Kconfig +0 -2
index a4ed7484d127..dc8a2a175f19 100644
@@ -215,7 +215,6 @@ config PCIE_TANGO_SMP8759
 config VMD
        depends on PCI_MSI && X86_64 && SRCU
        tristate "Intel Volume Management Device Driver"
-       default N
        ---help---
          Adds support for the Intel Volume Management Device (VMD). VMD is a
          secondary PCI host bridge that allows PCI Express root ports,
--
Quoted 2 lines

2.14.1

The PCIE_ALTERA symbol in this file also depends on the globally undefined symbol OF_PCI, which was removed in 4670d610d592 ("PCI: Move OF-related PCI functions into PCI core").

Cheers,
Ulf

#66 lore raw

Re: [PATCH 20/20] x86/PCI: VMD: Fix malformed default

In reply to Ulf Magnusson
Quoted 1 line

x86/PCI: VMD: Fix malformed default

In the title, please include a clue about what default you're talking about, e.g., mention Kconfig somehow.

Quoted 31 lines · Ulf Magnusson

'default N' should be 'default n', though they happen to have the same effect here, due to undefined symbols (N in this case) evaluating to n in a tristate sense.

Remove the default from VMD instead of changing it. bool and tristate symbols implicitly default to n.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/pci/host/Kconfig | 1 -
 1 file changed, 1 deletion(-)
drivers/pci/host/Kconfig +0 -1
index a4ed7484d127..dc8a2a175f19 100644
@@ -215,7 +215,6 @@ config PCIE_TANGO_SMP8759
 config VMD
 	depends on PCI_MSI && X86_64 && SRCU
 	tristate "Intel Volume Management Device Driver"
-	default N
 	---help---
 	  Adds support for the Intel Volume Management Device (VMD). VMD is a
 	  secondary PCI host bridge that allows PCI Express root ports,
Signature · 1 line
2.14.1
#67 lore raw

Re: [PATCH 20/20] x86/PCI: VMD: Fix malformed default

Quoted 22 lines · Bjorn Helgaas
Quoted 1 line

x86/PCI: VMD: Fix malformed default

In the title, please include a clue about what default you're talking about, e.g., mention Kconfig somehow.

Quoted 16 lines · Ulf Magnusson

'default N' should be 'default n', though they happen to have the same effect here, due to undefined symbols (N in this case) evaluating to n in a tristate sense.

Remove the default from VMD instead of changing it. bool and tristate symbols implicitly default to n.

Discovered with the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script.

---
 drivers/pci/host/Kconfig | 1 -
 1 file changed, 1 deletion(-)
Quoted 15 lines
Quoted 13 lines
drivers/pci/host/Kconfig +0 -2
index a4ed7484d127..dc8a2a175f19 100644
@@ -215,7 +215,6 @@ config PCIE_TANGO_SMP8759
 config VMD
      depends on PCI_MSI && X86_64 && SRCU
      tristate "Intel Volume Management Device Driver"
-     default N
      ---help---
        Adds support for the Intel Volume Management Device (VMD). VMD is a
        secondary PCI host bridge that allows PCI Express root ports,
--
Quoted 2 lines

2.14.1

Typoed the --in-reply-to. v2 is at https://lkml.org/lkml/2018/2/6/770

Cheers,
Ulf

#68 lore raw
In reply to Ulf Magnusson
Quoted 71 lines · Ulf Magnusson

Hello,

This patchset fixes most references to globally undefined symbols in Kconfig files, as reported by the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script. I will send emails in reply to the patchset for some cases where I'm not sure what the proper fix is.

All of the fixes are independent. I just created a patchset to track them more easily.

Summary of changes:

- Commits 1 and 2 fix what looks like potential bugs: A broken (intended) select of PL310_ERRATA_753970 for ARM/mvebu, and a misspelled dependency on PPC_PSERIES in KVM/PPC.

- Commits 3-5 remove leftover AVR32 stuff.

- Commits 6 and 7 remove some selects of removed symbols from riscv.

- Commits 8-17 remove various references to removed symbols.

- Commits 18-20 fix some (harmless) malformed defaults.

Ulf Magnusson (20):
ARM: mvebu: Fix broken PL310_ERRATA_753970 selects
KVM: PPC: Book3S PR: Fix broken select due to misspelling
kconfig: Remove leftover references to AVR32 symbol
misc: atmel: Remove CPU_AT32AP700X (AVR32) reference
rtc: at32ap700x: Remove PLATFORM_AT32AP dependency
riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select
riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select s390/kconfig: Remove ARCH_WANTS_PROT_NUMA_PROT_NONE select
clk: sunxi-ng: Remove SUNXI_CCU_* selects lib/Kconfig: Remove leftover select of GENERIC_IO
usb: gadget: udc: Remove USB_GADGET_DUALSPEED select
ARM: debug: Remove ARCH_MSM dep. from UNCOMPRESS_INCLUDE
ore: Remove PNFS_OBJLAYOUT dependency
iio: gyro: mpu3050: Remove INPUT_MPU3050 dep. from I2C
phy: Remove SOC_EXYNOS4212 dep. from PHY_EXYNOS4X12_USB
auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.
media: sec: Remove PLAT_S5P dependency
iwlwifi: fix malformed CONFIG_IWLWIFI_PCIE_RTPM default
ARC: Fix malformed ARC_EMUL_UNALIGNED default x86/PCI: VMD: Fix malformed default

 arch/arc/Kconfig                           |  1 -
 arch/arm/Kconfig.debug                     |  2 +-
 arch/arm/mach-mvebu/Kconfig                |  4 ++--
 arch/powerpc/kvm/Kconfig                   |  2 +-
 arch/riscv/Kconfig                         |  2 --
 arch/s390/Kconfig                          |  1 -
 drivers/auxdisplay/Kconfig                 |  2 +-
 drivers/clk/sunxi-ng/Kconfig               |  7 -------
 drivers/cpufreq/Kconfig                    | 10 ----------
 drivers/iio/gyro/Kconfig                   |  1 -
 drivers/media/platform/Kconfig             |  2 +-
 drivers/misc/Kconfig                       |  1 -
 drivers/net/wireless/intel/iwlwifi/Kconfig |  1 -
 drivers/pci/host/Kconfig                   |  1 -
 drivers/phy/samsung/Kconfig                |  2 +-
 drivers/pwm/Kconfig                        |  2 +-
 drivers/rtc/Kconfig                        |  2 +-
 drivers/spi/Kconfig                        |  2 +-
 drivers/usb/gadget/udc/Kconfig             |  1 -
 drivers/video/console/Kconfig              |  2 +-
 fs/exofs/Kconfig.ore                       |  2 +-
 lib/Kconfig                                |  1 -
 22 files changed, 12 insertions(+), 39 deletions(-)
Quoted 3 lines

--
2.14.1

Since 05a4a9527931 ("kernel/watchdog: split up config options"), lib/Kconfig.debug references a globally undefined symbol HARDLOCK_DETECTOR_ARCH. My guess is that it was supposed to be defined alongside HARDLOCK_DETECTOR_PERF in the same file.

Cheers,
Ulf

#69 lore raw
In reply to Ulf Magnusson
Quoted 71 lines · Ulf Magnusson

Hello,

This patchset fixes most references to globally undefined symbols in Kconfig files, as reported by the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script. I will send emails in reply to the patchset for some cases where I'm not sure what the proper fix is.

All of the fixes are independent. I just created a patchset to track them more easily.

Summary of changes:

- Commits 1 and 2 fix what looks like potential bugs: A broken (intended) select of PL310_ERRATA_753970 for ARM/mvebu, and a misspelled dependency on PPC_PSERIES in KVM/PPC.

- Commits 3-5 remove leftover AVR32 stuff.

- Commits 6 and 7 remove some selects of removed symbols from riscv.

- Commits 8-17 remove various references to removed symbols.

- Commits 18-20 fix some (harmless) malformed defaults.

Ulf Magnusson (20):
ARM: mvebu: Fix broken PL310_ERRATA_753970 selects
KVM: PPC: Book3S PR: Fix broken select due to misspelling
kconfig: Remove leftover references to AVR32 symbol
misc: atmel: Remove CPU_AT32AP700X (AVR32) reference
rtc: at32ap700x: Remove PLATFORM_AT32AP dependency
riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select
riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select s390/kconfig: Remove ARCH_WANTS_PROT_NUMA_PROT_NONE select
clk: sunxi-ng: Remove SUNXI_CCU_* selects lib/Kconfig: Remove leftover select of GENERIC_IO
usb: gadget: udc: Remove USB_GADGET_DUALSPEED select
ARM: debug: Remove ARCH_MSM dep. from UNCOMPRESS_INCLUDE
ore: Remove PNFS_OBJLAYOUT dependency
iio: gyro: mpu3050: Remove INPUT_MPU3050 dep. from I2C
phy: Remove SOC_EXYNOS4212 dep. from PHY_EXYNOS4X12_USB
auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.
media: sec: Remove PLAT_S5P dependency
iwlwifi: fix malformed CONFIG_IWLWIFI_PCIE_RTPM default
ARC: Fix malformed ARC_EMUL_UNALIGNED default x86/PCI: VMD: Fix malformed default

 arch/arc/Kconfig                           |  1 -
 arch/arm/Kconfig.debug                     |  2 +-
 arch/arm/mach-mvebu/Kconfig                |  4 ++--
 arch/powerpc/kvm/Kconfig                   |  2 +-
 arch/riscv/Kconfig                         |  2 --
 arch/s390/Kconfig                          |  1 -
 drivers/auxdisplay/Kconfig                 |  2 +-
 drivers/clk/sunxi-ng/Kconfig               |  7 -------
 drivers/cpufreq/Kconfig                    | 10 ----------
 drivers/iio/gyro/Kconfig                   |  1 -
 drivers/media/platform/Kconfig             |  2 +-
 drivers/misc/Kconfig                       |  1 -
 drivers/net/wireless/intel/iwlwifi/Kconfig |  1 -
 drivers/pci/host/Kconfig                   |  1 -
 drivers/phy/samsung/Kconfig                |  2 +-
 drivers/pwm/Kconfig                        |  2 +-
 drivers/rtc/Kconfig                        |  2 +-
 drivers/spi/Kconfig                        |  2 +-
 drivers/usb/gadget/udc/Kconfig             |  1 -
 drivers/video/console/Kconfig              |  2 +-
 fs/exofs/Kconfig.ore                       |  2 +-
 lib/Kconfig                                |  1 -
 22 files changed, 12 insertions(+), 39 deletions(-)
Quoted 3 lines

--
2.14.1

GDBSTUB in arch/mn10300/Kconfig.debug depends on the globally undefined symbol DEPRECATED, which seems to have been removed a long time ago. Here's an old reference I found:
https://lwn.net/Articles/113867/

Cheers,
Ulf

#70 lore raw
In reply to Ulf Magnusson
Quoted 71 lines · Ulf Magnusson

Hello,

This patchset fixes most references to globally undefined symbols in Kconfig files, as reported by the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script. I will send emails in reply to the patchset for some cases where I'm not sure what the proper fix is.

All of the fixes are independent. I just created a patchset to track them more easily.

Summary of changes:

- Commits 1 and 2 fix what looks like potential bugs: A broken (intended) select of PL310_ERRATA_753970 for ARM/mvebu, and a misspelled dependency on PPC_PSERIES in KVM/PPC.

- Commits 3-5 remove leftover AVR32 stuff.

- Commits 6 and 7 remove some selects of removed symbols from riscv.

- Commits 8-17 remove various references to removed symbols.

- Commits 18-20 fix some (harmless) malformed defaults.

Ulf Magnusson (20):
ARM: mvebu: Fix broken PL310_ERRATA_753970 selects
KVM: PPC: Book3S PR: Fix broken select due to misspelling
kconfig: Remove leftover references to AVR32 symbol
misc: atmel: Remove CPU_AT32AP700X (AVR32) reference
rtc: at32ap700x: Remove PLATFORM_AT32AP dependency
riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select
riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select s390/kconfig: Remove ARCH_WANTS_PROT_NUMA_PROT_NONE select
clk: sunxi-ng: Remove SUNXI_CCU_* selects lib/Kconfig: Remove leftover select of GENERIC_IO
usb: gadget: udc: Remove USB_GADGET_DUALSPEED select
ARM: debug: Remove ARCH_MSM dep. from UNCOMPRESS_INCLUDE
ore: Remove PNFS_OBJLAYOUT dependency
iio: gyro: mpu3050: Remove INPUT_MPU3050 dep. from I2C
phy: Remove SOC_EXYNOS4212 dep. from PHY_EXYNOS4X12_USB
auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.
media: sec: Remove PLAT_S5P dependency
iwlwifi: fix malformed CONFIG_IWLWIFI_PCIE_RTPM default
ARC: Fix malformed ARC_EMUL_UNALIGNED default x86/PCI: VMD: Fix malformed default

 arch/arc/Kconfig                           |  1 -
 arch/arm/Kconfig.debug                     |  2 +-
 arch/arm/mach-mvebu/Kconfig                |  4 ++--
 arch/powerpc/kvm/Kconfig                   |  2 +-
 arch/riscv/Kconfig                         |  2 --
 arch/s390/Kconfig                          |  1 -
 drivers/auxdisplay/Kconfig                 |  2 +-
 drivers/clk/sunxi-ng/Kconfig               |  7 -------
 drivers/cpufreq/Kconfig                    | 10 ----------
 drivers/iio/gyro/Kconfig                   |  1 -
 drivers/media/platform/Kconfig             |  2 +-
 drivers/misc/Kconfig                       |  1 -
 drivers/net/wireless/intel/iwlwifi/Kconfig |  1 -
 drivers/pci/host/Kconfig                   |  1 -
 drivers/phy/samsung/Kconfig                |  2 +-
 drivers/pwm/Kconfig                        |  2 +-
 drivers/rtc/Kconfig                        |  2 +-
 drivers/spi/Kconfig                        |  2 +-
 drivers/usb/gadget/udc/Kconfig             |  1 -
 drivers/video/console/Kconfig              |  2 +-
 fs/exofs/Kconfig.ore                       |  2 +-
 lib/Kconfig                                |  1 -
 22 files changed, 12 insertions(+), 39 deletions(-)
Quoted 3 lines

--
2.14.1

The symbol AC97_BUS_NEW in sound/ac97/Kconfig selects the globally undefined symbol AC97.

It doesn't seem to simply be a misspelling of AC97_BUS, because then the dependencies of AC97_BUS_COMPAT would be impossible to satisfy.

Cheers,
Ulf

#71 lore raw
Quoted 5 lines · Ulf Magnusson

The symbol AC97_BUS_NEW in sound/ac97/Kconfig selects the globally undefined symbol AC97.

It doesn't seem to simply be a misspelling of AC97_BUS, because then the dependencies of AC97_BUS_COMPAT would be impossible to satisfy.

You're correct.

I don't know what happened to me when coding this, but you can safely remove this useless select statement.

Cheers.

Signature · 1 line
Robert
#72 lore raw

PATCH 1/1 ALSA: ac97: kconfig: Remove select of undefined symbol AC97

The AC97_BUS_NEW Kconfig symbol selects the globally undefined symbol AC97.

Robert Jarzmik confirmed in https://lkml.org/lkml/2018/2/7/96 that the select was put in by mistake and can be safely removed, with no other changes required. Remove it.

---
 sound/ac97/Kconfig | 1 -
 1 file changed, 1 deletion(-)
sound/ac97/Kconfig +0 -1
index f8a64e15e5bf..baa5f8ef89d2 100644
@@ -5,7 +5,6 @@
 
 config AC97_BUS_NEW
 	tristate
-	select AC97
 	help
 	  This is the new AC97 bus type, successor of AC97_BUS. The ported
 	  drivers which benefit from the AC97 automatic probing should "select"
Signature · 1 line
2.14.1
#73 lore raw

Re: [PATCH] ALSA: ac97: kconfig: Remove select of undefined symbol AC97

Quoted 27 lines · Ulf Magnusson

The AC97_BUS_NEW Kconfig symbol selects the globally undefined symbol AC97.

Robert Jarzmik confirmed in https://lkml.org/lkml/2018/2/7/96 that the select was put in by mistake and can be safely removed, with no other changes required. Remove it.

---
 sound/ac97/Kconfig | 1 -
 1 file changed, 1 deletion(-)
sound/ac97/Kconfig +0 -1
index f8a64e15e5bf..baa5f8ef89d2 100644
@@ -5,7 +5,6 @@
 
 config AC97_BUS_NEW
 	tristate
-	select AC97
 	help
 	  This is the new AC97 bus type, successor of AC97_BUS. The ported
 	  drivers which benefit from the AC97 automatic probing should "select"
Signature · 1 line
2.14.1

I didn't have a particular tree in mind for this patchset by the way. Feel free to take the patch in wherever it makes sense.

Cheers,
Ulf

#74 lore raw

Re: [PATCH] ALSA: ac97: kconfig: Remove select of undefined symbol AC97

Quoted 20 lines · Ulf Magnusson
Quoted 18 lines · Ulf Magnusson

The AC97_BUS_NEW Kconfig symbol selects the globally undefined symbol AC97.

Robert Jarzmik confirmed in https://lkml.org/lkml/2018/2/7/96 that the select was put in by mistake and can be safely removed, with no other changes required. Remove it.

---
 sound/ac97/Kconfig | 1 -
 1 file changed, 1 deletion(-)
sound/ac97/Kconfig +0 -0
index f8a64e15e5bf..baa5f8ef89d2 100644
@@ -5,7 +5,6 @@
 
Quoted 12 lines
Quoted 9 lines
sound/ac97/Kconfig +0 -1
 config AC97_BUS_NEW
 	tristate
-	select AC97
 	help
 	  This is the new AC97 bus type, successor of AC97_BUS. The ported
 	  drivers which benefit from the AC97 automatic probing should "select"
Signature · 1 line
2.14.1

I didn't have a particular tree in mind for this patchset by the way. Feel free to take the patch in wherever it makes sense.

Applied now, thanks.

Takashi

#75 lore raw
In reply to Ulf Magnusson
Quoted 71 lines · Ulf Magnusson

Hello,

This patchset fixes most references to globally undefined symbols in Kconfig files, as reported by the
https://github.com/ulfalizer/Kconfiglib/blob/master/examples/list_undefined.py
script. I will send emails in reply to the patchset for some cases where I'm not sure what the proper fix is.

All of the fixes are independent. I just created a patchset to track them more easily.

Summary of changes:

- Commits 1 and 2 fix what looks like potential bugs: A broken (intended) select of PL310_ERRATA_753970 for ARM/mvebu, and a misspelled dependency on PPC_PSERIES in KVM/PPC.

- Commits 3-5 remove leftover AVR32 stuff.

- Commits 6 and 7 remove some selects of removed symbols from riscv.

- Commits 8-17 remove various references to removed symbols.

- Commits 18-20 fix some (harmless) malformed defaults.

Ulf Magnusson (20):
ARM: mvebu: Fix broken PL310_ERRATA_753970 selects
KVM: PPC: Book3S PR: Fix broken select due to misspelling
kconfig: Remove leftover references to AVR32 symbol
misc: atmel: Remove CPU_AT32AP700X (AVR32) reference
rtc: at32ap700x: Remove PLATFORM_AT32AP dependency
riscv: Remove ARCH_HAS_ATOMIC64_DEC_IF_POSITIVE select
riscv: Remove ARCH_WANT_OPTIONAL_GPIOLIB select s390/kconfig: Remove ARCH_WANTS_PROT_NUMA_PROT_NONE select
clk: sunxi-ng: Remove SUNXI_CCU_* selects lib/Kconfig: Remove leftover select of GENERIC_IO
usb: gadget: udc: Remove USB_GADGET_DUALSPEED select
ARM: debug: Remove ARCH_MSM dep. from UNCOMPRESS_INCLUDE
ore: Remove PNFS_OBJLAYOUT dependency
iio: gyro: mpu3050: Remove INPUT_MPU3050 dep. from I2C
phy: Remove SOC_EXYNOS4212 dep. from PHY_EXYNOS4X12_USB
auxdisplay: img-ascii-lcd: Remove MIPS_SEAD3 dep.
media: sec: Remove PLAT_S5P dependency
iwlwifi: fix malformed CONFIG_IWLWIFI_PCIE_RTPM default
ARC: Fix malformed ARC_EMUL_UNALIGNED default x86/PCI: VMD: Fix malformed default

 arch/arc/Kconfig                           |  1 -
 arch/arm/Kconfig.debug                     |  2 +-
 arch/arm/mach-mvebu/Kconfig                |  4 ++--
 arch/powerpc/kvm/Kconfig                   |  2 +-
 arch/riscv/Kconfig                         |  2 --
 arch/s390/Kconfig                          |  1 -
 drivers/auxdisplay/Kconfig                 |  2 +-
 drivers/clk/sunxi-ng/Kconfig               |  7 -------
 drivers/cpufreq/Kconfig                    | 10 ----------
 drivers/iio/gyro/Kconfig                   |  1 -
 drivers/media/platform/Kconfig             |  2 +-
 drivers/misc/Kconfig                       |  1 -
 drivers/net/wireless/intel/iwlwifi/Kconfig |  1 -
 drivers/pci/host/Kconfig                   |  1 -
 drivers/phy/samsung/Kconfig                |  2 +-
 drivers/pwm/Kconfig                        |  2 +-
 drivers/rtc/Kconfig                        |  2 +-
 drivers/spi/Kconfig                        |  2 +-
 drivers/usb/gadget/udc/Kconfig             |  1 -
 drivers/video/console/Kconfig              |  2 +-
 fs/exofs/Kconfig.ore                       |  2 +-
 lib/Kconfig                                |  1 -
 22 files changed, 12 insertions(+), 39 deletions(-)
Quoted 3 lines

--
2.14.1

Another one that looks like a potential bug:

b49efd762482 ("dma-mapping: move dma_mark_clean to dma-direct.h") selects ARCH_HAS_DMA_MARK_CLEAN for ia64 but never defines it. It's used in include/linux/dma-direct.h:

  #ifdef CONFIG_ARCH_HAS_DMA_MARK_CLEAN
  void dma_mark_clean(void *addr, size_t size);
  #else
  static inline void dma_mark_clean(void *addr, size_t size)
  {
  }
  #endif /* CONFIG_ARCH_HAS_DMA_MARK_CLEAN */

Cheers,
Ulf