Skip to contentSkip to conversation

x86/Kconfig: Select PCI_LOCKLESS_CONFIG if PCI is enabled

PATCH In review Sinan Kaya 3 participants 10 messages Jan 17, 2019 – Jan 21, 2019

Subsystems (3)
lore
Share & tools
lore

Sinan Kaya

author
#0 lore raw

PATCH 1/1 x86/Kconfig: Select PCI_LOCKLESS_CONFIG if PCI is enabled

After 'commit 5d32a66541c4 ("PCI/ACPI: Allow ACPI to be built without CONFIG_PCI set")' dependencies on CONFIG_PCI that previously were satisfied implicitly through dependencies on CONFIG_ACPI have to be specified directly.

PCI_LOCKLESS_CONFIG depends on PCI but this dependency has not been mentioned in the Kconfig.

Add an explicit dependency here.

Fixes 5d32a66541c46 ("PCI/ACPI: Allow ACPI to be built without CONFIG_PCI set")
---
 arch/x86/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/Kconfig +1 -1
index 568f339595ed..0519da6f8ee4 100644
@@ -199,7 +199,7 @@ config X86
 	select IRQ_FORCED_THREADING
 	select NEED_SG_DMA_LENGTH
 	select PCI_DOMAINS			if PCI
-	select PCI_LOCKLESS_CONFIG
+	select PCI_LOCKLESS_CONFIG		if PCI
 	select PERF_EVENTS
 	select RTC_LIB
 	select RTC_MC146818_LIB
Signature · 1 line
2.19.0
maintainer
#1 lore raw
Quoted 22 lines · Sinan Kaya

After 'commit 5d32a66541c4 ("PCI/ACPI: Allow ACPI to be built without CONFIG_PCI set")' dependencies on CONFIG_PCI that previously were satisfied implicitly through dependencies on CONFIG_ACPI have to be specified directly.

PCI_LOCKLESS_CONFIG depends on PCI but this dependency has not been mentioned in the Kconfig.

Add an explicit dependency here.

Fixes 5d32a66541c46 ("PCI/ACPI: Allow ACPI to be built without CONFIG_PCI set")
---
 arch/x86/Kconfig | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/Kconfig +0 -0
index 568f339595ed..0519da6f8ee4 100644

arch/x86/Kconfig config X86 · line 205

Quoted 8 lines
arch/x86/Kconfig +1 -1
@@ -199,7 +199,7 @@ config X86
 	select IRQ_FORCED_THREADING
 	select NEED_SG_DMA_LENGTH
 	select PCI_DOMAINS			if PCI
-	select PCI_LOCKLESS_CONFIG
+	select PCI_LOCKLESS_CONFIG		if PCI
 	select PERF_EVENTS
 	select RTC_LIB
 	select RTC_MC146818_LIB

Are there any more arch/x86/ fixes for the 5d32a66541c4 fallout floating around?

If so, pls merge them all together into a single patch so that we're done with this once and for all.

Also, I see a lot of build failures when doing randconfig builds for the stuff in drivers/platform/x86/Kconfig. Is someone picking those up too?

If not, I'd take 'em too in a single patch, if Darren and Andy are fine with it.

Thx.

Signature · 4 lines
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.
author
#2 lore raw
Quoted 35 lines · Borislav Petkov
Quoted 22 lines · Sinan Kaya

After 'commit 5d32a66541c4 ("PCI/ACPI: Allow ACPI to be built without CONFIG_PCI set")' dependencies on CONFIG_PCI that previously were satisfied implicitly through dependencies on CONFIG_ACPI have to be specified directly.

PCI_LOCKLESS_CONFIG depends on PCI but this dependency has not been mentioned in the Kconfig.

Add an explicit dependency here.

Fixes 5d32a66541c46 ("PCI/ACPI: Allow ACPI to be built without CONFIG_PCI set")
---
  arch/x86/Kconfig | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/Kconfig +0 -0
index 568f339595ed..0519da6f8ee4 100644

arch/x86/Kconfig config X86 · line 205

Quoted 8 lines
arch/x86/Kconfig +1 -1
@@ -199,7 +199,7 @@ config X86
  	select IRQ_FORCED_THREADING
  	select NEED_SG_DMA_LENGTH
  	select PCI_DOMAINS			if PCI
-	select PCI_LOCKLESS_CONFIG
+	select PCI_LOCKLESS_CONFIG		if PCI
  	select PERF_EVENTS
  	select RTC_LIB
  	select RTC_MC146818_LIB

Are there any more arch/x86/ fixes for the 5d32a66541c4 fallout floating around?

Quoted 8 lines

If so, pls merge them all together into a single patch so that we're done with this once and for all.

Also, I see a lot of build failures when doing randconfig builds for the stuff in drivers/platform/x86/Kconfig. Is someone picking those up too?

If not, I'd take 'em too in a single patch, if Darren and Andy are fine with it.

I'm only looking at PCI and ACPI related changes to be honest. The rest is out of my expertise area.

I started a run of 1000 randconfig runs in the meantime. I'll collect the results soon and follow up if there is something I can help.

Quoted 3 lines

Thx.

author
#3 lore raw
In reply to Borislav Petkov
Quoted 2 lines · Borislav Petkov

Also, I see a lot of build failures when doing randconfig builds for the stuff in drivers/platform/x86/Kconfig. Is someone picking those up too?

Can you share the build failures you are seeing?

maintainer
#4 lore raw
Quoted 5 lines · Sinan Kaya
Quoted 2 lines · Borislav Petkov

Also, I see a lot of build failures when doing randconfig builds for the stuff in drivers/platform/x86/Kconfig. Is someone picking those up too?

Can you share the build failures you are seeing?

Here are three configs which fail.

They have in common:

# CONFIG_PCI is not set

As a tip for the future - always build the default def/allno/yes/modconfig targets and as many randconfigs as possible when changing Kconfig dependencies. I've learned that the hard way.

Here are the build failures:

arch/x86/kernel/ftrace.c:668:23: warning: ‘ftrace_jmp_replace’ defined but not used [-Wunused-function] drivers/platform/x86/intel_ips.c: In function ‘ips_probe’:
drivers/platform/x86/intel_ips.c:1463:2: error: implicit declaration of function ‘pcim_enable_device’ [-Werror=implicit-function-declaration] drivers/platform/x86/intel_ips.c:1525:2: error: implicit declaration of function ‘pci_alloc_irq_vectors’ [-Werror=implicit-function-declaration] drivers/platform/x86/intel_ips.c:1525:41: error: ‘PCI_IRQ_LEGACY’ undeclared (first use in this function) drivers/platform/x86/intel_ips.c:1525:41: note: each undeclared identifier is reported only once for each function it appears in drivers/platform/x86/intel_ips.c:1529:2: error: implicit declaration of function ‘pci_irq_vector’ [-Werror=implicit-function-declaration] drivers/platform/x86/intel_ips.c:1592:2: error: implicit declaration of function ‘pci_free_irq_vectors’ [-Werror=implicit-function-declaration] drivers/platform/x86/intel_ips.c: At top level:
drivers/platform/x86/intel_ips.c:1636:1: warning: data definition has no type or storage class [enabled by default] drivers/platform/x86/intel_ips.c:1636:1: error: type defaults to ‘int’ in declaration of ‘module_pci_driver’ [-Werror=implicit-int] drivers/platform/x86/intel_ips.c:1636:1: warning: parameter names (without types) in function declaration [enabled by default] drivers/platform/x86/intel_ips.c:1629:26: warning: ‘ips_pci_driver’ defined but not used [-Wunused-variable]
make[3]: *** [drivers/platform/x86/intel_ips.o] Error 1
make[2]: *** [drivers/platform/x86] Error 2

drivers/platform/x86/apple-gmux.c: In function ‘gmux_probe’:
drivers/platform/x86/apple-gmux.c:746:22: error: ‘pci_bus_type’ undeclared (first use in this function) drivers/platform/x86/apple-gmux.c:746:22: note: each undeclared identifier is reported only once for each function it appears in make[3]: *** [drivers/platform/x86/apple-gmux.o] Error 1

drivers/platform/x86/intel_pmc_ipc.c: In function ‘ipc_pci_probe’:
drivers/platform/x86/intel_pmc_ipc.c:520:2: error: implicit declaration of function ‘pcim_enable_device’ [-Werror=implicit-function-declaration] make[3]: *** [drivers/platform/x86/intel_pmc_ipc.o] Error 1

You get the idea...

Signature · 4 lines
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.
#5 lore raw
Quoted 41 lines · Borislav Petkov
Quoted 5 lines · Sinan Kaya
Quoted 2 lines · Borislav Petkov

Also, I see a lot of build failures when doing randconfig builds for the stuff in drivers/platform/x86/Kconfig. Is someone picking those up too?

Can you share the build failures you are seeing?

Here are three configs which fail.

They have in common:

# CONFIG_PCI is not set

As a tip for the future - always build the default def/allno/yes/modconfig targets and as many randconfigs as possible when changing Kconfig dependencies. I've learned that the hard way.

Here are the build failures:

arch/x86/kernel/ftrace.c:668:23: warning: ‘ftrace_jmp_replace’ defined but not used [-Wunused-function] drivers/platform/x86/intel_ips.c: In function ‘ips_probe’:
drivers/platform/x86/intel_ips.c:1463:2: error: implicit declaration of function ‘pcim_enable_device’ [-Werror=implicit-function-declaration] drivers/platform/x86/intel_ips.c:1525:2: error: implicit declaration of function ‘pci_alloc_irq_vectors’ [-Werror=implicit-function-declaration] drivers/platform/x86/intel_ips.c:1525:41: error: ‘PCI_IRQ_LEGACY’ undeclared (first use in this function) drivers/platform/x86/intel_ips.c:1525:41: note: each undeclared identifier is reported only once for each function it appears in drivers/platform/x86/intel_ips.c:1529:2: error: implicit declaration of function ‘pci_irq_vector’ [-Werror=implicit-function-declaration] drivers/platform/x86/intel_ips.c:1592:2: error: implicit declaration of function ‘pci_free_irq_vectors’ [-Werror=implicit-function-declaration] drivers/platform/x86/intel_ips.c: At top level:
drivers/platform/x86/intel_ips.c:1636:1: warning: data definition has no type or storage class [enabled by default] drivers/platform/x86/intel_ips.c:1636:1: error: type defaults to ‘int’ in declaration of ‘module_pci_driver’ [-Werror=implicit-int] drivers/platform/x86/intel_ips.c:1636:1: warning: parameter names (without types) in function declaration [enabled by default] drivers/platform/x86/intel_ips.c:1629:26: warning: ‘ips_pci_driver’ defined but not used [-Wunused-variable]
make[3]: *** [drivers/platform/x86/intel_ips.o] Error 1
make[2]: *** [drivers/platform/x86] Error 2

drivers/platform/x86/apple-gmux.c: In function ‘gmux_probe’:
drivers/platform/x86/apple-gmux.c:746:22: error: ‘pci_bus_type’ undeclared (first use in this function) drivers/platform/x86/apple-gmux.c:746:22: note: each undeclared identifier is reported only once for each function it appears in make[3]: *** [drivers/platform/x86/apple-gmux.o] Error 1

Quoted 5 lines

drivers/platform/x86/intel_pmc_ipc.c: In function ‘ipc_pci_probe’:
drivers/platform/x86/intel_pmc_ipc.c:520:2: error: implicit declaration of function ‘pcim_enable_device’ [-Werror=implicit-function-declaration] make[3]: *** [drivers/platform/x86/intel_pmc_ipc.o] Error 1

You get the idea...

I have patches for intel_ips and intel_pmc_ipc queued up (will be pushed for -rc3), plus some others. apple-gmux is a new one.

maintainer
#6 lore raw
Quoted 2 lines · Rafael J. Wysocki

I have patches for intel_ips and intel_pmc_ipc queued up (will be pushed for -rc3), plus some others.

Yeah, I saw the patchset and applied some of them locally so that I be able to do randconfig builds. Do you have a branch somewhere which I can merge locally for testing?

Signature · 4 lines
Regards/Gruss,
    Boris.

Good mailing practices for 400: avoid top-posting and trim the reply.
author
#7 lore raw
Quoted 8 lines · Borislav Petkov
Quoted 2 lines · Rafael J. Wysocki

I have patches for intel_ips and intel_pmc_ipc queued up (will be pushed for -rc3), plus some others.

Yeah, I saw the patchset and applied some of them locally so that I be able to do randconfig builds. Do you have a branch somewhere which I can merge locally for testing?

You can use this tag (next-20190116) for local test.

#8 lore raw
In reply to Borislav Petkov
Quoted 8 lines · Borislav Petkov
Quoted 2 lines · Rafael J. Wysocki

I have patches for intel_ips and intel_pmc_ipc queued up (will be pushed for -rc3), plus some others.

Yeah, I saw the patchset and applied some of them locally so that I be able to do randconfig builds. Do you have a branch somewhere which I can merge locally for testing?

You can pull the tag I've just pushed to Linus:
https://lore.kernel.org/lkml/CAJZ5v0jiK=zLP4cUsw=y9ea7PLHHgy=xshAsHFhTgyDJH+kOiQ@mail.gmail.com/T/#u

author
#9 lore raw
Quoted 6 lines · Rafael J. Wysocki
Quoted 5 lines · Borislav Petkov
Quoted 2 lines · Rafael J. Wysocki

I have patches for intel_ips and intel_pmc_ipc queued up (will be pushed for -rc3), plus some others.

Quoted 7 lines
Quoted 3 lines

Yeah, I saw the patchset and applied some of them locally so that I be able to do randconfig builds. Do you have a branch somewhere which I can merge locally for testing?

You can pull the tag I've just pushed to Linus:
https://lore.kernel.org/lkml/CAJZ5v0jiK=zLP4cUsw=y9ea7PLHHgy=xshAsHFhTgyDJH+kOiQ@mail.gmail.com/T/#u

Here is a result of my randconfig analysis using next-20190116.

There are two issues that are related to my change:
1. x86/Kconfig: Select PCI_LOCKLESS_CONFIG if PCI is enabled
2. mfd: Fix unmet dependency warning for MFD_TPS68470

I'll post v2 for this patch to pick up the above commit #2.

and three unrelated changes as follows:
1. platform/x86: Fix unmet dependency warning for SAMSUNG_Q10
2. platform/x86: Fix unmet dependency warning for ACPI_CMPC
3. iwlwifi: Fix unmet dependency warning for MAC80211_LEDS

I'm posting these unrelated changes independently.