This patch series cleans up a handful of selects that were redundant and
deletes presumably dead code with the goal of making it possible to add
kunit tests for the CCF in the future. To do that, we introduce a
"legacy" clk Kconfig option to mark code that hasn't migrated to the
common clk framework and then make the COMMON_CLK config option visible
in the menuconfig as long as that legacy option isn't enabled. I've also
included a couple patches at the end that may be more controversial but
helped me consolidate all this logic/code.
I haven't done more than compile test a few configs for arm, arm64,
h8300, and mips. More testing is welcome.
The plan is that I'll just merge the whole pile through the clk tree. If
the first five patches or the last three patches are better going
through another tree like arm-soc or architecture trees that's fine too,
but there are potential conflicts between trees so maybe it's better to
just leave it all in one tree.
Stephen Boyd (9):
ARM: Remove redundant COMMON_CLK selects
ARM: Remove redundant CLKDEV_LOOKUP selects
arm64: tegra: Remove redundant CLKDEV_LOOKUP selects
h8300: Remove redundant CLKDEV_LOOKUP selects
MIPS: Remove redundant CLKDEV_LOOKUP selects
clk: Allow the common clk framework to be selectable
ARM: mmp: Remove legacy clk code
MIPS: Loongson64: Drop asm/clock.h include
clk: Move HAVE_CLK config out of architecture layer
Cc: Alexander Shiyan <[email protected]>
Cc: "Andreas Färber" <[email protected]>
Cc: Arnd Bergmann <[email protected]>
Cc: Aurelien Jacquiot <[email protected]>
Cc: Catalin Marinas <[email protected]>
Cc: <[email protected]>
Cc: Geert Uytterhoeven <[email protected]>
Cc: Guan Xuetao <[email protected]>
Cc: Jiaxun Yang <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Cc: Lubomir Rintel <[email protected]>
Cc: Manivannan Sadhasivam <[email protected]>
Cc: Mark Brown <[email protected]>
Cc: Mark Salter <[email protected]>
Cc: Paul Burton <[email protected]>
Cc: Paul Walmsley <[email protected]>
Cc: Rich Felker <[email protected]>
Cc: Russell King <[email protected]>
Cc: Thierry Reding <[email protected]>
Cc: Thomas Bogendoerfer <[email protected]>
Cc: Tony Prisk <[email protected]>
Cc: [email protected]
Cc: Will Deacon <[email protected]>
Cc: Yoshinori Sato <[email protected]>
arch/Kconfig | 6 --
arch/arm/Kconfig | 5 +-
arch/arm/mach-actions/Kconfig | 1 -
arch/arm/mach-clps711x/Kconfig | 1 -
arch/arm/mach-mmp/Kconfig | 1 -
arch/arm/mach-mmp/Makefile | 6 --
arch/arm/mach-mmp/clock-mmp2.c | 114 -------------------------------
arch/arm/mach-mmp/clock-pxa168.c | 94 -------------------------
arch/arm/mach-mmp/clock-pxa910.c | 70 -------------------
arch/arm/mach-mmp/clock.c | 105 ----------------------------
arch/arm/mach-mmp/clock.h | 65 ------------------
arch/arm/mach-vt8500/Kconfig | 1 -
arch/arm64/Kconfig.platforms | 1 -
arch/c6x/Kconfig | 1 +
arch/h8300/Kconfig | 1 -
arch/m68k/Kconfig.cpu | 2 +-
arch/mips/Kconfig | 8 +--
arch/mips/loongson2ef/Kconfig | 2 +-
arch/mips/loongson64/smp.c | 1 -
arch/sh/boards/Kconfig | 5 ++
arch/unicore32/Kconfig | 2 +-
drivers/clk/Kconfig | 23 +++++--
22 files changed, 34 insertions(+), 481 deletions(-)
delete mode 100644 arch/arm/mach-mmp/clock-mmp2.c
delete mode 100644 arch/arm/mach-mmp/clock-pxa168.c
delete mode 100644 arch/arm/mach-mmp/clock-pxa910.c
delete mode 100644 arch/arm/mach-mmp/clock.c
delete mode 100644 arch/arm/mach-mmp/clock.h
base-commit: 7111951b8d4973bda27ff663f2cf18b663d15b48
--
Sent by a computer, using git, on the internet
This include isn't used by this file, so just remove it.
Cc: Jiaxun Yang <[email protected]>
Cc: Paul Burton <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Signed-off-by: Stephen Boyd <[email protected]>
---
arch/mips/loongson64/smp.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/arch/mips/loongson64/smp.c b/arch/mips/loongson64/smp.c
index de8e0741ce2d..b48d9c566c5a 100644
--- a/arch/mips/loongson64/smp.c
+++ b/arch/mips/loongson64/smp.c
@@ -14,7 +14,6 @@
#include <linux/kexec.h>
#include <asm/processor.h>
#include <asm/time.h>
-#include <asm/clock.h>
#include <asm/tlbflush.h>
#include <asm/cacheflush.h>
#include <loongson.h>
--
Sent by a computer, using git, on the internet
The h8300 architecture selects COMMON_CLK already, and the COMMON_CLK
config option already selects CLKDEV_LOOKUP so it's redundant to have
this selected again.
Cc: Yoshinori Sato <[email protected]>
Cc: [email protected]
Cc: Arnd Bergmann <[email protected]>
Signed-off-by: Stephen Boyd <[email protected]>
---
arch/h8300/Kconfig | 1 -
1 file changed, 1 deletion(-)
diff --git a/arch/h8300/Kconfig b/arch/h8300/Kconfig
index ec800e9d5aad..d11666d538fe 100644
--- a/arch/h8300/Kconfig
+++ b/arch/h8300/Kconfig
@@ -13,7 +13,6 @@ config H8300
select GENERIC_CPU_DEVICES
select MODULES_USE_ELF_RELA
select GENERIC_CLOCKEVENTS
- select CLKDEV_LOOKUP
select COMMON_CLK
select ARCH_WANT_FRAME_POINTERS
select OF
--
Sent by a computer, using git, on the internet
The implementation of 'struct clk' is not really an architectual detail
anymore now that most architectures have migrated to the common clk
framework. To sway new architecture ports away from trying to implement
their own 'struct clk', move the config next to the common clk framework
config.
Cc: Russell King <[email protected]>
Cc: Arnd Bergmann <[email protected]>
Signed-off-by: Stephen Boyd <[email protected]>
---
arch/Kconfig | 6 ------
drivers/clk/Kconfig | 6 ++++++
2 files changed, 6 insertions(+), 6 deletions(-)
diff --git a/arch/Kconfig b/arch/Kconfig
index 17fe351cdde0..903ac5547ee8 100644
--- a/arch/Kconfig
+++ b/arch/Kconfig
@@ -321,12 +321,6 @@ config HAVE_FUNCTION_ARG_ACCESS_API
the API needed to access function arguments from pt_regs,
declared in asm/ptrace.h
-config HAVE_CLK
- bool
- help
- The <linux/clk.h> calls support software clock gating and
- thus are a key power management tool on many systems.
-
config HAVE_HW_BREAKPOINT
bool
depends on PERF_EVENTS
diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig
index 890bed62196d..6ea0631e3956 100644
--- a/drivers/clk/Kconfig
+++ b/drivers/clk/Kconfig
@@ -1,5 +1,11 @@
# SPDX-License-Identifier: GPL-2.0
+config HAVE_CLK
+ bool
+ help
+ The <linux/clk.h> calls support software clock gating and
+ thus are a key power management tool on many systems.
+
config CLKDEV_LOOKUP
bool
select HAVE_CLK
--
Sent by a computer, using git, on the internet
The arm64 architecture selects COMMON_CLK at the toplevel ARM64 config.
The COMMON_CLK config option already selects CLKDEV_LOOKUP so it's
redundant to have this selected again for the Tegra specific config.
Cc: Paul Walmsley <[email protected]>
Cc: Thierry Reding <[email protected]>
Cc: Catalin Marinas <[email protected]>
Cc: Will Deacon <[email protected]>
Cc: <[email protected]>
Cc: Arnd Bergmann <[email protected]>
Signed-off-by: Stephen Boyd <[email protected]>
---
arch/arm64/Kconfig.platforms | 1 -
1 file changed, 1 deletion(-)
diff --git a/arch/arm64/Kconfig.platforms b/arch/arm64/Kconfig.platforms
index b2b504ea6fd6..5d5d91feca44 100644
--- a/arch/arm64/Kconfig.platforms
+++ b/arch/arm64/Kconfig.platforms
@@ -236,7 +236,6 @@ config ARCH_TEGRA
bool "NVIDIA Tegra SoC Family"
select ARCH_HAS_RESET_CONTROLLER
select ARM_GIC_PM
- select CLKDEV_LOOKUP
select CLKSRC_MMIO
select TIMER_OF
select GENERIC_CLOCKEVENTS
--
Sent by a computer, using git, on the internet
The ATH79 config selects COMMON_CLK already, and the COMMON_CLK config
option already selects CLKDEV_LOOKUP, and CLKDEV_LOOKUP already selects
HAVE_CLK so it's redundant to have these selected again.
Cc: Thomas Bogendoerfer <[email protected]>
Cc: <[email protected]>
Signed-off-by: Stephen Boyd <[email protected]>
---
arch/mips/Kconfig | 2 --
1 file changed, 2 deletions(-)
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index 797d7f1ad5fe..e53a8dd5c19b 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -209,9 +209,7 @@ config ATH79
select DMA_NONCOHERENT
select GPIOLIB
select PINCTRL
- select HAVE_CLK
select COMMON_CLK
- select CLKDEV_LOOKUP
select IRQ_MIPS_CPU
select SYS_HAS_CPU_MIPS32_R2
select SYS_HAS_EARLY_PRINTK
--
Sent by a computer, using git, on the internet
Enable build testing and configuration control of the common clk
framework so that more code coverage and testing can be done on the
common clk framework across various architectures. This also nicely
removes the requirement that architectures must select the framework
when they don't use it in architecture code.
There's one snag with doing this, and that's making sure that randconfig
builds don't select this option when some architecture or platform
implements 'struct clk' outside of the common clk framework. Introduce a
new config option 'HAVE_LEGACY_CLK' to indicate those platforms that
haven't migrated to the common clk framework and therefore shouldn't be
allowed to select this new config option. Also add a note that we hope
one day to remove this config entirely.
Based on a patch by Mark Brown <[email protected]>.
Cc: Mark Brown <[email protected]>
Cc: Geert Uytterhoeven <[email protected]>
Cc: Mark Salter <[email protected]>
Cc: Aurelien Jacquiot <[email protected]>
Cc: Jiaxun Yang <[email protected]>
Cc: Guan Xuetao <[email protected]>
Cc: Russell King <[email protected]>
Cc: Arnd Bergmann <[email protected]>
Cc: Yoshinori Sato <[email protected]>
Cc: Rich Felker <[email protected]>
Cc: Thomas Bogendoerfer <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Cc: <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Stephen Boyd <[email protected]>
---
arch/arm/Kconfig | 2 ++
arch/c6x/Kconfig | 1 +
arch/m68k/Kconfig.cpu | 2 +-
arch/mips/Kconfig | 6 ++++--
arch/mips/loongson2ef/Kconfig | 2 +-
arch/sh/boards/Kconfig | 5 +++++
arch/unicore32/Kconfig | 2 +-
drivers/clk/Kconfig | 17 +++++++++++++----
8 files changed, 28 insertions(+), 9 deletions(-)
diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
index eeabdc5a3dd2..0606e1bbd7b6 100644
--- a/arch/arm/Kconfig
+++ b/arch/arm/Kconfig
@@ -364,6 +364,7 @@ config ARCH_EP93XX
select CPU_ARM920T
select GENERIC_CLOCKEVENTS
select GPIOLIB
+ select HAVE_LEGACY_CLK
help
This enables support for the Cirrus EP93xx series of CPUs.
@@ -522,6 +523,7 @@ config ARCH_OMAP1
select GENERIC_IRQ_MULTI_HANDLER
select GPIOLIB
select HAVE_IDE
+ select HAVE_LEGACY_CLK
select IRQ_DOMAIN
select NEED_MACH_IO_H if PCCARD
select NEED_MACH_MEMORY_H
diff --git a/arch/c6x/Kconfig b/arch/c6x/Kconfig
index e65e8d82442a..6444ebfd06a6 100644
--- a/arch/c6x/Kconfig
+++ b/arch/c6x/Kconfig
@@ -11,6 +11,7 @@ config C6X
select ARCH_HAS_SYNC_DMA_FOR_CPU
select ARCH_HAS_SYNC_DMA_FOR_DEVICE
select CLKDEV_LOOKUP
+ select HAVE_LEGACY_CLK
select GENERIC_ATOMIC64
select GENERIC_IRQ_SHOW
select HAVE_ARCH_TRACEHOOK
diff --git a/arch/m68k/Kconfig.cpu b/arch/m68k/Kconfig.cpu
index 60ac1cd8b96f..bd2d29c22a10 100644
--- a/arch/m68k/Kconfig.cpu
+++ b/arch/m68k/Kconfig.cpu
@@ -28,7 +28,7 @@ config COLDFIRE
select CPU_HAS_NO_MULDIV64
select GENERIC_CSUM
select GPIOLIB
- select HAVE_CLK
+ select HAVE_LEGACY_CLK
endchoice
diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
index e53a8dd5c19b..fa02c715ab25 100644
--- a/arch/mips/Kconfig
+++ b/arch/mips/Kconfig
@@ -181,7 +181,7 @@ config AR7
select SYS_SUPPORTS_ZBOOT_UART16550
select GPIOLIB
select VLYNQ
- select HAVE_CLK
+ select HAVE_LEGACY_CLK
help
Support for the Texas Instruments AR7 System-on-a-Chip
family: TNETD7100, 7200 and 7300.
@@ -296,9 +296,9 @@ config BCM63XX
select SYS_HAS_EARLY_PRINTK
select SWAP_IO_SPACE
select GPIOLIB
- select HAVE_CLK
select MIPS_L1_CACHE_SHIFT_4
select CLKDEV_LOOKUP
+ select HAVE_LEGACY_CLK
help
Support for BCM63XX based boards
@@ -419,6 +419,7 @@ config LANTIQ
select SWAP_IO_SPACE
select BOOT_RAW
select CLKDEV_LOOKUP
+ select HAVE_LEGACY_CLK
select USE_OF
select PINCTRL
select PINCTRL_LANTIQ
@@ -656,6 +657,7 @@ config RALINK
select SYS_SUPPORTS_MIPS16
select SYS_HAS_EARLY_PRINTK
select CLKDEV_LOOKUP
+ select HAVE_LEGACY_CLK
select ARCH_HAS_RESET_CONTROLLER
select RESET_CONTROLLER
diff --git a/arch/mips/loongson2ef/Kconfig b/arch/mips/loongson2ef/Kconfig
index 595dd48e1e4d..c9ec43afde73 100644
--- a/arch/mips/loongson2ef/Kconfig
+++ b/arch/mips/loongson2ef/Kconfig
@@ -46,7 +46,7 @@ config LEMOTE_MACH2F
select CSRC_R4K if ! MIPS_EXTERNAL_TIMER
select DMA_NONCOHERENT
select GENERIC_ISA_DMA_SUPPORT_BROKEN
- select HAVE_CLK
+ select HAVE_LEGACY_CLK
select FORCE_PCI
select I8259
select IRQ_MIPS_CPU
diff --git a/arch/sh/boards/Kconfig b/arch/sh/boards/Kconfig
index cee24c308337..fb0ca0c1efe1 100644
--- a/arch/sh/boards/Kconfig
+++ b/arch/sh/boards/Kconfig
@@ -7,6 +7,11 @@ config SOLUTION_ENGINE
config SH_ALPHA_BOARD
bool
+config SH_CUSTOM_CLK
+ def_bool y
+ depends on !SH_DEVICE_TREE
+ select HAVE_LEGACY_CLK
+
config SH_DEVICE_TREE
bool
select OF
diff --git a/arch/unicore32/Kconfig b/arch/unicore32/Kconfig
index 41fe944005f8..11ba1839d198 100644
--- a/arch/unicore32/Kconfig
+++ b/arch/unicore32/Kconfig
@@ -70,7 +70,7 @@ config ARCH_PUV3
def_bool y
select CPU_UCV2
select GENERIC_CLOCKEVENTS
- select HAVE_CLK
+ select HAVE_LEGACY_CLK
select GPIOLIB
# CONFIGs for ARCH_PUV3
diff --git a/drivers/clk/Kconfig b/drivers/clk/Kconfig
index bcb257baed06..890bed62196d 100644
--- a/drivers/clk/Kconfig
+++ b/drivers/clk/Kconfig
@@ -7,8 +7,18 @@ config CLKDEV_LOOKUP
config HAVE_CLK_PREPARE
bool
-config COMMON_CLK
+config HAVE_LEGACY_CLK # TODO: Remove once all legacy users are migrated
bool
+ select HAVE_CLK
+ help
+ Select this option when the clock API in <linux/clk.h> is implemented
+ by platform/architecture code. This method is deprecated. Modern
+ code should select COMMON_CLK instead and not define a custom
+ 'struct clk'.
+
+menuconfig COMMON_CLK
+ bool "Common Clock Framework"
+ depends on !HAVE_LEGACY_CLK
select HAVE_CLK_PREPARE
select CLKDEV_LOOKUP
select SRCU
@@ -20,8 +30,7 @@ config COMMON_CLK
Architectures utilizing the common struct clk should select
this option.
-menu "Common Clock Framework"
- depends on COMMON_CLK
+if COMMON_CLK
config COMMON_CLK_WM831X
tristate "Clock driver for WM831x/2x PMICs"
@@ -362,4 +371,4 @@ source "drivers/clk/ti/Kconfig"
source "drivers/clk/uniphier/Kconfig"
source "drivers/clk/zynqmp/Kconfig"
-endmenu
+endif
--
Sent by a computer, using git, on the internet
On Sun, Apr 5, 2020 at 4:51 AM Stephen Boyd <[email protected]> wrote:
> There's one snag with doing this, and that's making sure that randconfig
> builds don't select this option when some architecture or platform
> implements 'struct clk' outside of the common clk framework. Introduce a
> new config option 'HAVE_LEGACY_CLK' to indicate those platforms that
> haven't migrated to the common clk framework and therefore shouldn't be
> allowed to select this new config option. Also add a note that we hope
> one day to remove this config entirely.
Good idea!
I've looked through the individual ones and commented a bit on
what I think may or may not happen with them.
ralink SOC_MT7621 is the only one that I think you got wrong,
as it already has common-clk support.
> arch/arm/Kconfig | 2 ++
> arch/c6x/Kconfig | 1 +
> arch/m68k/Kconfig.cpu | 2 +-
> arch/mips/Kconfig | 6 ++++--
> arch/mips/loongson2ef/Kconfig | 2 +-
> arch/sh/boards/Kconfig | 5 +++++
> arch/unicore32/Kconfig | 2 +-
> drivers/clk/Kconfig | 17 +++++++++++++----
> 8 files changed, 28 insertions(+), 9 deletions(-)
>
> diff --git a/arch/arm/Kconfig b/arch/arm/Kconfig
> index eeabdc5a3dd2..0606e1bbd7b6 100644
> --- a/arch/arm/Kconfig
> +++ b/arch/arm/Kconfig
> @@ -364,6 +364,7 @@ config ARCH_EP93XX
> select CPU_ARM920T
> select GENERIC_CLOCKEVENTS
> select GPIOLIB
> + select HAVE_LEGACY_CLK
> help
> This enables support for the Cirrus EP93xx series of CPUs.
>
> @@ -522,6 +523,7 @@ config ARCH_OMAP1
> select GENERIC_IRQ_MULTI_HANDLER
> select GPIOLIB
> select HAVE_IDE
> + select HAVE_LEGACY_CLK
> select IRQ_DOMAIN
> select NEED_MACH_IO_H if PCCARD
> select NEED_MACH_MEMORY_H
Ah nice, I thought there were more than two implementations
left. I need to refresh the series for OMAP1 that I did last year
and debug the regression that caused. I think Linus Walleij
has looked into the ep93xx one at some point.
> diff --git a/arch/c6x/Kconfig b/arch/c6x/Kconfig
> index e65e8d82442a..6444ebfd06a6 100644
> --- a/arch/c6x/Kconfig
> +++ b/arch/c6x/Kconfig
> @@ -11,6 +11,7 @@ config C6X
> select ARCH_HAS_SYNC_DMA_FOR_CPU
> select ARCH_HAS_SYNC_DMA_FOR_DEVICE
> select CLKDEV_LOOKUP
> + select HAVE_LEGACY_CLK
> select GENERIC_ATOMIC64
> select GENERIC_IRQ_SHOW
> select HAVE_ARCH_TRACEHOOK
This was last touched in 2012 when TMS320C6678 support
was added. It seems to be based on the original same code
as omap1 and omap2 that eventually turned into the common
clock subsystem, so it would clearly be possible to do a
step-by-step conversion like I did for omap1, but I suspect
we can wait for all of c6x-linux to become obsolete enough.
> diff --git a/arch/m68k/Kconfig.cpu b/arch/m68k/Kconfig.cpu
> index 60ac1cd8b96f..bd2d29c22a10 100644
> --- a/arch/m68k/Kconfig.cpu
> +++ b/arch/m68k/Kconfig.cpu
> @@ -28,7 +28,7 @@ config COLDFIRE
> select CPU_HAS_NO_MULDIV64
> select GENERIC_CSUM
> select GPIOLIB
> - select HAVE_CLK
> + select HAVE_LEGACY_CLK
>
> endchoice
The coldfire clock implementation looks rather simple compared
to chips from the 2010s: most chips have only fixed clocks,
and three of them have one of two registers of clock gates.
It shouldn't be hard to convert, but enabling common-clk will
cause a noticeable kernel size increase on the fairly limited
hardware.
Simply enabling COMMON_CLK in m5475evb_defconfig
results in a 1.7% or 40KB growth in kernel size, plus there
would be additional dynamic memory usage:
text data bss dec hex filename
1934726 263616 83284 2281626 22d09a obj/vmlinux-before
1971989 266192 83308 2321489 236c51 obj/vmlinux-after
> diff --git a/arch/mips/Kconfig b/arch/mips/Kconfig
> index e53a8dd5c19b..fa02c715ab25 100644
> --- a/arch/mips/Kconfig
> +++ b/arch/mips/Kconfig
> @@ -181,7 +181,7 @@ config AR7
> select SYS_SUPPORTS_ZBOOT_UART16550
> select GPIOLIB
> select VLYNQ
> - select HAVE_CLK
> + select HAVE_LEGACY_CLK
> help
> Support for the Texas Instruments AR7 System-on-a-Chip
> family: TNETD7100, 7200 and 7300.
I suspect this platform is fairly dead, OpenWRT dropped support for it,
and the known machines have at most 16MB of RAM, which limits
the number of use cases that also want to build recent kernels.
> @@ -296,9 +296,9 @@ config BCM63XX
> select SYS_HAS_EARLY_PRINTK
> select SWAP_IO_SPACE
> select GPIOLIB
> - select HAVE_CLK
> select MIPS_L1_CACHE_SHIFT_4
> select CLKDEV_LOOKUP
> + select HAVE_LEGACY_CLK
> help
> Support for BCM63XX based boards
>
> @@ -419,6 +419,7 @@ config LANTIQ
> select SWAP_IO_SPACE
> select BOOT_RAW
> select CLKDEV_LOOKUP
> + select HAVE_LEGACY_CLK
> select USE_OF
> select PINCTRL
> select PINCTRL_LANTIQ
> @@ -656,6 +657,7 @@ config RALINK
> select SYS_SUPPORTS_MIPS16
> select SYS_HAS_EARLY_PRINTK
> select CLKDEV_LOOKUP
> + select HAVE_LEGACY_CLK
> select ARCH_HAS_RESET_CONTROLLER
> select RESET_CONTROLLER
All three of these are older platforms that still have a number of machines
supported in OpenWRT, so there is hope that someone would do the work
to add common-clk support. Both Lantiq and Ralink have been converted
to device tree not too long ago.
I think the RALINK entry needs to be changed, as this includes
SOC_MT7621, which already uses COMMON_CLK.
> diff --git a/arch/mips/loongson2ef/Kconfig b/arch/mips/loongson2ef/Kconfig
> index 595dd48e1e4d..c9ec43afde73 100644
> --- a/arch/mips/loongson2ef/Kconfig
> +++ b/arch/mips/loongson2ef/Kconfig
> @@ -46,7 +46,7 @@ config LEMOTE_MACH2F
> select CSRC_R4K if ! MIPS_EXTERNAL_TIMER
> select DMA_NONCOHERENT
> select GENERIC_ISA_DMA_SUPPORT_BROKEN
> - select HAVE_CLK
> + select HAVE_LEGACY_CLK
> select FORCE_PCI
> select I8259
> select IRQ_MIPS_CPU
For all I can tell, this driver only uses the clock interface for
its cpufreq driver, and the actual API definition looks bogus.
Half of it could be folded into the cpufreq driver, and the rest
of it removed, along with arch/mips/include/asm/clock.h
> diff --git a/arch/sh/boards/Kconfig b/arch/sh/boards/Kconfig
> index cee24c308337..fb0ca0c1efe1 100644
> --- a/arch/sh/boards/Kconfig
> +++ b/arch/sh/boards/Kconfig
> @@ -7,6 +7,11 @@ config SOLUTION_ENGINE
> config SH_ALPHA_BOARD
> bool
>
> +config SH_CUSTOM_CLK
> + def_bool y
> + depends on !SH_DEVICE_TREE
> + select HAVE_LEGACY_CLK
> +
> config SH_DEVICE_TREE
> bool
> select OF
It's going to be hard to find anyone willing to touch this, as the code
is very large but has few users these days.
drivers/sh/clk/ is basically a generic clk subsystem on its own, and
there are at least 28 different arch/sh SoCs with drivers for it.
Arnd
Hi Stephen,
I love your patch! Yet something to improve:
[auto build test ERROR on arm-soc/for-next]
[also build test ERROR on clk/clk-next arm/for-next linus/master v5.6 next-20200405]
[if your patch is applied to the wrong git tree, please drop us a note to help
improve the system. BTW, we also suggest to use '--base' option to specify the
base tree in git format-patch, please see https://stackoverflow.com/a/37406982]
url: https://github.com/0day-ci/linux/commits/Stephen-Boyd/Allow-COMMON_CLK-to-be-selectable/20200405-105253
base: https://git.kernel.org/pub/scm/linux/kernel/git/arm/arm-soc.git for-next
config: sparc-allyesconfig (attached as .config)
compiler: sparc64-linux-gcc (GCC) 9.3.0
reproduce:
wget https://raw.githubusercontent.com/intel/lkp-tests/master/sbin/make.cross -O ~/bin/make.cross
chmod +x ~/bin/make.cross
# save the attached .config to linux build tree
GCC_VERSION=9.3.0 make.cross ARCH=sparc
If you fix the issue, kindly add following tag as appropriate
Reported-by: kbuild test robot <[email protected]>
All errors (new ones prefixed by >>):
sparc64-linux-ld: drivers/mmc/host/meson-mx-sdio.o: in function `meson_mx_mmc_remove':
>> meson-mx-sdio.c:(.text+0x70): undefined reference to `of_platform_device_destroy'
sparc64-linux-ld: drivers/mmc/host/meson-mx-sdio.o: in function `meson_mx_mmc_probe':
>> meson-mx-sdio.c:(.text+0x9e4): undefined reference to `of_platform_device_create'
>> sparc64-linux-ld: meson-mx-sdio.c:(.text+0xdd4): undefined reference to `of_platform_device_destroy'
---
0-DAY CI Kernel Test Service, Intel Corporation
https://lists.01.org/hyperkitty/list/[email protected]
On Mon, Apr 6, 2020 at 5:01 AM Stephen Boyd <[email protected]> wrote:
>
> Quoting Arnd Bergmann (2020-04-05 05:45:20)
> > On Sun, Apr 5, 2020 at 4:51 AM Stephen Boyd <[email protected]> wrote:
> > > There's one snag with doing this, and that's making sure that randconfig
> > > builds don't select this option when some architecture or platform
> > > implements 'struct clk' outside of the common clk framework. Introduce a
> > > new config option 'HAVE_LEGACY_CLK' to indicate those platforms that
> > > haven't migrated to the common clk framework and therefore shouldn't be
> > > allowed to select this new config option. Also add a note that we hope
> > > one day to remove this config entirely.
> >
> > Good idea!
> >
> > I've looked through the individual ones and commented a bit on
> > what I think may or may not happen with them.
> >
> > ralink SOC_MT7621 is the only one that I think you got wrong,
> > as it already has common-clk support.
>
> Ah I missed that it was inside a big if RALINK. Thanks. I suppose I
> should just remove the select then for that config and not worry about
> the duplication of clkdev and common clk configs.
Won't that cause build failures in those configurations that have
both implementations?
According to the Makefile, the clk.c file is built whenever CONFIG_MIPS_GIC
is unset, so I think we need
select HAVE_LEGACY_CLK if !MIPS_GIC
or maybe move the select into the per-chip configs that need it:
RT288X, RT305X, RT3883, and MT7620.
> > > diff --git a/arch/m68k/Kconfig.cpu b/arch/m68k/Kconfig.cpu
> > > index 60ac1cd8b96f..bd2d29c22a10 100644
> > > --- a/arch/m68k/Kconfig.cpu
> > > +++ b/arch/m68k/Kconfig.cpu
> >
> > text data bss dec hex filename
> > 1934726 263616 83284 2281626 22d09a obj/vmlinux-before
> > 1971989 266192 83308 2321489 236c51 obj/vmlinux-after
> >
> > The coldfire clock implementation looks rather simple compared
> > to chips from the 2010s: most chips have only fixed clocks,
> > and three of them have one of two registers of clock gates.
> >
> > It shouldn't be hard to convert, but enabling common-clk will
> > cause a noticeable kernel size increase on the fairly limited
> > hardware.
> >
> > Simply enabling COMMON_CLK in m5475evb_defconfig
> > results in a 1.7% or 40KB growth in kernel size, plus there
> > would be additional dynamic memory usage:
> There could certainly be some work done to reduce the code size of the
> CCF. I haven't looked but perhaps we could save some memory by making
> the basic types selectable too and then push a bunch of kconfig updates
> through for that.
Right, that might help. Another possibility would be to support both
the common clk layer and the custom clk implementation on coldfire
until we remove the other custom implementations, by which point
even fewer people will care about coldfire.
Let's see what Geert and Greg think would be the best path for coldfire,
maybe the added 40KB is less of a problem after all.
Arnd
On Sat, Apr 04, 2020 at 07:51:17PM -0700, Stephen Boyd wrote:
> The arm64 architecture selects COMMON_CLK at the toplevel ARM64 config.
> The COMMON_CLK config option already selects CLKDEV_LOOKUP so it's
> redundant to have this selected again for the Tegra specific config.
>
> Cc: Paul Walmsley <[email protected]>
> Cc: Thierry Reding <[email protected]>
> Cc: Catalin Marinas <[email protected]>
> Cc: Will Deacon <[email protected]>
> Cc: <[email protected]>
> Cc: Arnd Bergmann <[email protected]>
> Signed-off-by: Stephen Boyd <[email protected]>
> ---
> arch/arm64/Kconfig.platforms | 1 -
> 1 file changed, 1 deletion(-)
We rarely touch this, so I think it's fine for this to go through the
clock tree. I'll make a mental note just in case there's going to be
something touching this area in the next release, but I'm sure it'll be
all minor so we could easily work that out:
Acked-by: Thierry Reding <[email protected]>
Hi Arnd, Stephen
On 6/4/20 5:35 pm, Arnd Bergmann wrote:
> On Mon, Apr 6, 2020 at 5:01 AM Stephen Boyd <[email protected]> wrote:
>> Quoting Arnd Bergmann (2020-04-05 05:45:20)
>>> On Sun, Apr 5, 2020 at 4:51 AM Stephen Boyd <[email protected]> wrote:
>>>> There's one snag with doing this, and that's making sure that randconfig
>>>> builds don't select this option when some architecture or platform
>>>> implements 'struct clk' outside of the common clk framework. Introduce a
>>>> new config option 'HAVE_LEGACY_CLK' to indicate those platforms that
>>>> haven't migrated to the common clk framework and therefore shouldn't be
>>>> allowed to select this new config option. Also add a note that we hope
>>>> one day to remove this config entirely.
>>>
>>> Good idea!
>>>
>>> I've looked through the individual ones and commented a bit on
>>> what I think may or may not happen with them.
>>>
>>> ralink SOC_MT7621 is the only one that I think you got wrong,
>>> as it already has common-clk support.
>>
>> Ah I missed that it was inside a big if RALINK. Thanks. I suppose I
>> should just remove the select then for that config and not worry about
>> the duplication of clkdev and common clk configs.
>
> Won't that cause build failures in those configurations that have
> both implementations?
>
> According to the Makefile, the clk.c file is built whenever CONFIG_MIPS_GIC
> is unset, so I think we need
>
> select HAVE_LEGACY_CLK if !MIPS_GIC
>
> or maybe move the select into the per-chip configs that need it:
> RT288X, RT305X, RT3883, and MT7620.
>
>>>> diff --git a/arch/m68k/Kconfig.cpu b/arch/m68k/Kconfig.cpu
>>>> index 60ac1cd8b96f..bd2d29c22a10 100644
>>>> --- a/arch/m68k/Kconfig.cpu
>>>> +++ b/arch/m68k/Kconfig.cpu
>>>
>>> text data bss dec hex filename
>>> 1934726 263616 83284 2281626 22d09a obj/vmlinux-before
>>> 1971989 266192 83308 2321489 236c51 obj/vmlinux-after
>>>
>>> The coldfire clock implementation looks rather simple compared
>>> to chips from the 2010s: most chips have only fixed clocks,
>>> and three of them have one of two registers of clock gates.
>>>
>>> It shouldn't be hard to convert, but enabling common-clk will
>>> cause a noticeable kernel size increase on the fairly limited
>>> hardware.
>>>
>>> Simply enabling COMMON_CLK in m5475evb_defconfig
>>> results in a 1.7% or 40KB growth in kernel size, plus there
>>> would be additional dynamic memory usage:
>> There could certainly be some work done to reduce the code size of the
>> CCF. I haven't looked but perhaps we could save some memory by making
>> the basic types selectable too and then push a bunch of kconfig updates
>> through for that.
>
> Right, that might help. Another possibility would be to support both
> the common clk layer and the custom clk implementation on coldfire
> until we remove the other custom implementations, by which point
> even fewer people will care about coldfire.
>
> Let's see what Geert and Greg think would be the best path for coldfire,
> maybe the added 40KB is less of a problem after all.
Losing another 40k is not ideal, but not the end of the world.
It would not stop me running it on any platforms I regularly
run on. For sure some of the really old hardware just doesn't
have the RAM to spare.
Any way, I say we have to move forward and and move to using
the common clock framework for ColdFire sooner than later.
Regards
Greg
Hi Greg,
On Tue, Apr 7, 2020 at 6:57 AM Greg Ungerer <[email protected]> wrote:
> On 6/4/20 5:35 pm, Arnd Bergmann wrote:
> > On Mon, Apr 6, 2020 at 5:01 AM Stephen Boyd <[email protected]> wrote:
> >> Quoting Arnd Bergmann (2020-04-05 05:45:20)
> >>> On Sun, Apr 5, 2020 at 4:51 AM Stephen Boyd <[email protected]> wrote:
> >>>> There's one snag with doing this, and that's making sure that randconfig
> >>>> builds don't select this option when some architecture or platform
> >>>> implements 'struct clk' outside of the common clk framework. Introduce a
> >>>> new config option 'HAVE_LEGACY_CLK' to indicate those platforms that
> >>>> haven't migrated to the common clk framework and therefore shouldn't be
> >>>> allowed to select this new config option. Also add a note that we hope
> >>>> one day to remove this config entirely.
> >>>> --- a/arch/m68k/Kconfig.cpu
> >>>> +++ b/arch/m68k/Kconfig.cpu
> >>>
> >>> text data bss dec hex filename
> >>> 1934726 263616 83284 2281626 22d09a obj/vmlinux-before
> >>> 1971989 266192 83308 2321489 236c51 obj/vmlinux-after
> >>>
> >>> The coldfire clock implementation looks rather simple compared
> >>> to chips from the 2010s: most chips have only fixed clocks,
> >>> and three of them have one of two registers of clock gates.
> >>>
> >>> It shouldn't be hard to convert, but enabling common-clk will
> >>> cause a noticeable kernel size increase on the fairly limited
> >>> hardware.
> >>>
> >>> Simply enabling COMMON_CLK in m5475evb_defconfig
> >>> results in a 1.7% or 40KB growth in kernel size, plus there
> >>> would be additional dynamic memory usage:
> >> There could certainly be some work done to reduce the code size of the
> >> CCF. I haven't looked but perhaps we could save some memory by making
> >> the basic types selectable too and then push a bunch of kconfig updates
> >> through for that.
> >
> > Right, that might help. Another possibility would be to support both
> > the common clk layer and the custom clk implementation on coldfire
> > until we remove the other custom implementations, by which point
> > even fewer people will care about coldfire.
> >
> > Let's see what Geert and Greg think would be the best path for coldfire,
> > maybe the added 40KB is less of a problem after all.
>
> Losing another 40k is not ideal, but not the end of the world.
> It would not stop me running it on any platforms I regularly
> run on. For sure some of the really old hardware just doesn't
> have the RAM to spare.
>
> Any way, I say we have to move forward and and move to using
> the common clock framework for ColdFire sooner than later.
Fine for me.
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds