2021-10-28 14:21:12

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 00/13] Kconfig symbol clean-up on ./arch/arm{64}

Dear arm maintainers,

The script ./scripts/checkkconfigsymbols.py warns on invalid references to
Kconfig symbols (often: minor typos, name confusions or outdated references).

This is a patch series addressing the issues reported by
./scripts/checkkconfigsymbols.py in the ./arch/arm{64}/ directories, quickly
filtered down with:

./scripts/checkkconfigsymbols.py | grep "arch/arm" -B 1 -A 2

without considering kernel configs, i.e., after removing arch/arm{64}/configs
in the working tree. It addresses some issues I considered to be "true positives";
so, issues that should be addressed and cleaned up.

The patches on updating the comments might arguably be considered of lower value
by itself, but as the other patches show, checkkconfigsymbols does detect some
relevant issues in the kernel tree otherwise being unnoticed.
So, it might be worth to clean up the kernel tree to have checkkconfigsymbols
produce a rather short list of issues and then continuously follow and check
what checkkconfigsymbols reports.

Some work from checkkconfigsymbols on arch/arm/ is deferred; the other remaining
issues that checkkconfigsymbols reports are (as far as I understand it now)
"false positives", i.e., the tool misinterprets the actual content) or
"intentional", e.g., they refer to historic configs for the purpose of
documentation (e.g., in changelogs) or are in various kernel configs that are
not continuously updated to match the current set of kernel configs. You can
see below a grouped list of remaining reports.

Each patch in this series can be discussed and applied individually if needed.
They are sent in one patch series, as they all orginate from the investigation
guided by the same tool and hence share similar topics and resolutions.

Please pick this series of minor clean-up patches on ./arch/arm{64}/.
It applies cleanly on next-20211027.

Best regards,

Lukas


Remaining reports from ./scripts/checkkconfigsymbols.py pointing to files in
./arch/arm{64}/ (excluding configs), grouped by the rationale why it is not
addressed in this patch series:


- Reference in historic comment:

ARM_NR_BANKS
Referencing files: arch/arm/mach-exynos/exynos.c

CPU
Referencing files: arch/arm/lib/backtrace.S, arch/arm/lib/csumpartialcopyuser.S,
arch/arm/mach-rpc/ecard-loader.S, arch/arm/mach-rpc/io-acorn.S, arch/arm/nwfpe/fpmodule.h

CPU_ARM92{0,2,5,6}_CPU_IDLE
Referencing files: arch/arm/mm/proc-arm92{0,2,5,6}.S

KVM_INDIRECT_VECTORS
Referencing files: arch/arm64/kvm/hyp/nvhe/host.S


- Use of CONFIG_* environment variable beyond the definition in Kconfig scripts

AS_DMB_ISHLD
Referencing files: arch/arm64/include/asm/vdso/compat_barrier.h, arch/arm64/kernel/vdso32/Makefile

CC_HAS_K_CONSTRAINT
Referencing files: arch/arm64/Makefile, arch/arm64/include/asm/atomic_ll_sc.h

SHELL
Ignore CONFIG_SHELL. Default variable in Kbuild build system.


- Parsing Mistake (Incomplete parsing heuristics) by checkkconfigsymbols.py:

ASC1, ASC2, SBC, UART
Referencing files: arch/arm/Kconfig.debug


- Future work to send clean-up patches:

IRDA{_MODULE}
Referencing files: arch/arm/mach-pxa/balloon3.c, arch/arm/mach-pxa/palm27x.c,
arch/arm/mach-pxa/palm27x.h, arch/arm/mach-pxa/palmtc.c, include/linux/atalk.h,
include/linux/netdevice.h

PXA_FICP{_MODULE}
Referencing files: arch/arm/mach-pxa/spitz.c

DEBUG_LL_SER3
Referencing files: arch/arm/boot/compressed/head.S

already discussed in 2014: https://lore.kernel.org/all/1400055127.31197.1.camel@x220/


Lukas Bulwahn (13):
arm: debug: remove obsolete debug code for DEBUG_ZTE_ZX
arm: debug: reuse the config DEBUG_OMAP2UART{1,2} for OMAP{3,4,5}
arm: Kconfig.debug: drop reference to removed ARCH_MSM
arm: drop an obsolete ifdef with the removed config PCI_HOST_ITE8152
arm: davinci: remove reference to obsolete BLK_DEV_PALMCHIP_BK3710
arm: ixp4xx: remove dead configs CPU_IXP43X and CPU_IXP46X
arm: imx: remove dead left-over from i.MX{27,31,35} removal
arm: imx: rename DEBUG_IMX21_IMX27_UART to DEBUG_IMX27_UART
arm: milbeaut: remove select of non-existing PINCTRL_MILBEAUT
arm: nomadik: drop selecting obsolete CLKSRC_NOMADIK_MTU_SCHED_CLOCK
arm: npcm: drop selecting non-existing ARM_ERRATA_794072
arm: socfpga: always select PL310_ERRATA_753970
arm: pgtable: refer to intended CONFIG_ARM_LPAE in comment

arch/arm/Kconfig.debug | 28 ++++++++++-------------
arch/arm/include/debug/imx-uart.h | 18 +++++++--------
arch/arm/include/debug/pl01x.S | 7 ------
arch/arm/kernel/bios32.c | 2 --
arch/arm/mach-davinci/board-dm644x-evm.c | 3 +--
arch/arm/mach-davinci/board-dm646x-evm.c | 3 +--
arch/arm/mach-davinci/board-neuros-osd2.c | 3 +--
arch/arm/mach-imx/Kconfig | 12 ----------
arch/arm/mach-imx/Makefile | 2 --
arch/arm/mach-ixp4xx/Kconfig | 13 -----------
arch/arm/mach-milbeaut/Kconfig | 1 -
arch/arm/mach-nomadik/Kconfig | 1 -
arch/arm/mach-npcm/Kconfig | 1 -
arch/arm/mach-socfpga/Kconfig | 2 +-
arch/arm/mm/pgd.c | 2 +-
15 files changed, 26 insertions(+), 72 deletions(-)

--
2.26.2


2021-10-28 14:21:13

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 01/13] arm: debug: remove obsolete debug code for DEBUG_ZTE_ZX

Commit 89d4f98ae90d ("ARM: remove zte zx platform") removes the config
DEBUG_ZTE_ZX. Hence, since then, the "ifdef CONFIG_DEBUG_ZTE_ZX" in
./arch/arm/include/debug/pl01x.S is dead code.

Fortunately, ./scripts/checkkconfigsymbols.py detects this and warns:

DEBUG_ZTE_ZX
Referencing files: arch/arm/include/debug/pl01x.S

So, remove the obsolete ifdef CONFIG_DEBUG_ZTE_ZX.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/include/debug/pl01x.S | 7 -------
1 file changed, 7 deletions(-)

diff --git a/arch/arm/include/debug/pl01x.S b/arch/arm/include/debug/pl01x.S
index 0c7bfa4c10db..c7e02d0628bf 100644
--- a/arch/arm/include/debug/pl01x.S
+++ b/arch/arm/include/debug/pl01x.S
@@ -8,13 +8,6 @@
*/
#include <linux/amba/serial.h>

-#ifdef CONFIG_DEBUG_ZTE_ZX
-#undef UART01x_DR
-#undef UART01x_FR
-#define UART01x_DR 0x04
-#define UART01x_FR 0x14
-#endif
-
#ifdef CONFIG_DEBUG_UART_PHYS
.macro addruart, rp, rv, tmp
ldr \rp, =CONFIG_DEBUG_UART_PHYS
--
2.26.2

2021-10-28 14:21:15

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 02/13] arm: debug: reuse the config DEBUG_OMAP2UART{1,2} for OMAP{3,4,5}

Commit d2b310b0234c ("ARM: debug: Use generic 8250 debug_ll for omap2 and
omap3/4/5 common uarts") adds address definitions of DEBUG_UART_PHYS for
OMAP2, OMAP3, OMAP4 and OMAP5 in ./arch/arm/Kconfig.debug.

These definitions depend on DEBUG_OMAP{2,3,4,5}UART{1,2}; however, only
DEBUG_OMAP2UART{1,2} are defined in ./arch/arm/Kconfig.debug, and
DEBUG_OMAP{3,4,5}UART{1,2} are not defined. Hence, the script
./scripts/checkkconfigsymbols.py warns here on non-existing symbols.
Simply reuse the config DEBUG_OMAP2UART{1,2}; there is no need to define
separate config symbols for OMAP{3,4,5}. So, just delete the dead
references to DEBUG_OMAP{3,4,5}UART{1,2}.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/Kconfig.debug | 12 ++++--------
1 file changed, 4 insertions(+), 8 deletions(-)

diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug
index 98436702e0c7..83484564b1d9 100644
--- a/arch/arm/Kconfig.debug
+++ b/arch/arm/Kconfig.debug
@@ -1643,10 +1643,8 @@ config DEBUG_UART_PHYS
default 0x48020000 if DEBUG_OMAP4UART3 || DEBUG_TI81XXUART1
default 0x48022000 if DEBUG_TI81XXUART2
default 0x48024000 if DEBUG_TI81XXUART3
- default 0x4806a000 if DEBUG_OMAP2UART1 || DEBUG_OMAP3UART1 || \
- DEBUG_OMAP4UART1 || DEBUG_OMAP5UART1
- default 0x4806c000 if DEBUG_OMAP2UART2 || DEBUG_OMAP3UART2 || \
- DEBUG_OMAP4UART2 || DEBUG_OMAP5UART2
+ default 0x4806a000 if DEBUG_OMAP2UART1
+ default 0x4806c000 if DEBUG_OMAP2UART2
default 0x4806e000 if DEBUG_OMAP2UART3 || DEBUG_OMAP4UART4
default 0x49020000 if DEBUG_OMAP3UART3
default 0x49042000 if DEBUG_OMAP3UART4
@@ -1782,10 +1780,8 @@ config DEBUG_UART_VIRT
default 0xfa020000 if DEBUG_OMAP4UART3 || DEBUG_TI81XXUART1
default 0xfa022000 if DEBUG_TI81XXUART2
default 0xfa024000 if DEBUG_TI81XXUART3
- default 0xfa06a000 if DEBUG_OMAP2UART1 || DEBUG_OMAP3UART1 || \
- DEBUG_OMAP4UART1 || DEBUG_OMAP5UART1
- default 0xfa06c000 if DEBUG_OMAP2UART2 || DEBUG_OMAP3UART2 || \
- DEBUG_OMAP4UART2 || DEBUG_OMAP5UART2
+ default 0xfa06a000 if DEBUG_OMAP2UART1
+ default 0xfa06c000 if DEBUG_OMAP2UART2
default 0xfa06e000 if DEBUG_OMAP2UART3 || DEBUG_OMAP4UART4
default 0xfa71e000 if DEBUG_QCOM_UARTDM
default 0xfb002000 if DEBUG_CNS3XXX
--
2.26.2

2021-10-28 14:21:23

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 03/13] arm: Kconfig.debug: drop reference to removed ARCH_MSM

Commit c0c89fafa289 ("ARM: Remove mach-msm and associated ARM architecture
code") removes the definition of the config ARCH_MSM. Since then, the
reference to ARCH_MSM in the dependencies of UNCOMPRESS_INCLUDE in
Kconfig.debug is dead.

Fortunately, ./scripts/checkkconfigsymbols.py warns:

ARCH_MSM
Referencing files: arch/arm/Kconfig.debug

Drop the dependency on this removed config.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/Kconfig.debug | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug
index 83484564b1d9..87aa6e92ee6e 100644
--- a/arch/arm/Kconfig.debug
+++ b/arch/arm/Kconfig.debug
@@ -1886,7 +1886,7 @@ 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"

--
2.26.2

2021-10-28 14:21:32

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 05/13] arm: davinci: remove reference to obsolete BLK_DEV_PALMCHIP_BK3710

Commit b7fb14d3ac63 ("ide: remove the legacy ide driver") removes the
definition of the config BLK_DEV_PALMCHIP_BK3710.

So, remove the obsolete references in the mach-davinci architecture.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/mach-davinci/board-dm644x-evm.c | 3 +--
arch/arm/mach-davinci/board-dm646x-evm.c | 3 +--
arch/arm/mach-davinci/board-neuros-osd2.c | 3 +--
3 files changed, 3 insertions(+), 6 deletions(-)

diff --git a/arch/arm/mach-davinci/board-dm644x-evm.c b/arch/arm/mach-davinci/board-dm644x-evm.c
index cce3a621eb20..df6dbb3f0faf 100644
--- a/arch/arm/mach-davinci/board-dm644x-evm.c
+++ b/arch/arm/mach-davinci/board-dm644x-evm.c
@@ -834,8 +834,7 @@ static int davinci_phy_fixup(struct phy_device *phydev)
return 0;
}

-#define HAS_ATA (IS_ENABLED(CONFIG_BLK_DEV_PALMCHIP_BK3710) || \
- IS_ENABLED(CONFIG_PATA_BK3710))
+#define HAS_ATA IS_ENABLED(CONFIG_PATA_BK3710)

#define HAS_NOR IS_ENABLED(CONFIG_MTD_PHYSMAP)

diff --git a/arch/arm/mach-davinci/board-dm646x-evm.c b/arch/arm/mach-davinci/board-dm646x-evm.c
index ee91d81ebbfd..744ece6b6b9a 100644
--- a/arch/arm/mach-davinci/board-dm646x-evm.c
+++ b/arch/arm/mach-davinci/board-dm646x-evm.c
@@ -151,8 +151,7 @@ static struct platform_device davinci_aemif_device = {
.num_resources = ARRAY_SIZE(davinci_aemif_resources),
};

-#define HAS_ATA (IS_ENABLED(CONFIG_BLK_DEV_PALMCHIP_BK3710) || \
- IS_ENABLED(CONFIG_PATA_BK3710))
+#define HAS_ATA IS_ENABLED(CONFIG_PATA_BK3710)

#ifdef CONFIG_I2C
/* CPLD Register 0 bits to control ATA */
diff --git a/arch/arm/mach-davinci/board-neuros-osd2.c b/arch/arm/mach-davinci/board-neuros-osd2.c
index b4843f68bb57..81c21fe5748a 100644
--- a/arch/arm/mach-davinci/board-neuros-osd2.c
+++ b/arch/arm/mach-davinci/board-neuros-osd2.c
@@ -164,8 +164,7 @@ static struct davinci_mmc_config davinci_ntosd2_mmc_config = {
.wires = 4,
};

-#define HAS_ATA (IS_ENABLED(CONFIG_BLK_DEV_PALMCHIP_BK3710) || \
- IS_ENABLED(CONFIG_PATA_BK3710))
+#define HAS_ATA IS_ENABLED(CONFIG_PATA_BK3710)

#define HAS_NAND IS_ENABLED(CONFIG_MTD_NAND_DAVINCI)

--
2.26.2

2021-10-28 14:21:40

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 06/13] arm: ixp4xx: remove dead configs CPU_IXP43X and CPU_IXP46X

Commit 73d04ca5f4ac ("ARM: ixp4xx: Delete Intel reference design
boardfiles") removes the definition of the configs MACH_IXDP465 and
MACH_KIXRP435, but misses to remove the configs CPU_IXP43X and CPU_IXP46X
that depend on those removed configs, and hence are dead now.

Fortunately, ./scripts/checkkconfigsymbols.py warns:

MACH_IXDP465
Referencing files: arch/arm/mach-ixp4xx/Kconfig

MACH_KIXRP435
Referencing files: arch/arm/mach-ixp4xx/Kconfig

Remove the dead configs CPU_IXP43X and CPU_IXP46X.

A further quick grep for the name of those two symbols did not show any
use of the two config symbols; so, there are no further clean-up activities
beyond this config removal needed.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/mach-ixp4xx/Kconfig | 13 -------------
1 file changed, 13 deletions(-)

diff --git a/arch/arm/mach-ixp4xx/Kconfig b/arch/arm/mach-ixp4xx/Kconfig
index 365a5853d310..4c787b4be62b 100644
--- a/arch/arm/mach-ixp4xx/Kconfig
+++ b/arch/arm/mach-ixp4xx/Kconfig
@@ -39,19 +39,6 @@ config ARCH_PRPMC1100
PrPCM1100 Processor Mezanine Module. For more information on
this platform, see <file:Documentation/arm/ixp4xx.rst>.

-#
-# Certain registers and IRQs are only enabled if supporting IXP465 CPUs
-#
-config CPU_IXP46X
- bool
- depends on MACH_IXDP465
- default y
-
-config CPU_IXP43X
- bool
- depends on MACH_KIXRP435
- default y
-
comment "IXP4xx Options"

config IXP4XX_PCI_LEGACY
--
2.26.2

2021-10-28 14:21:58

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 08/13] arm: imx: rename DEBUG_IMX21_IMX27_UART to DEBUG_IMX27_UART

Since commit 4b563a066611 ("ARM: imx: Remove imx21 support"), the config
DEBUG_IMX21_IMX27_UART is really only debug support for IMX27.

So, rename this option to DEBUG_IMX27_UART and adjust dependencies in
Kconfig and rename the definitions to IMX27 as further clean-up.

This issue was discovered with ./scripts/checkkconfigsymbols.py, which
reported that DEBUG_IMX21_IMX27_UART depends on the non-existing config
SOC_IMX21.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/Kconfig.debug | 14 +++++++-------
arch/arm/include/debug/imx-uart.h | 18 +++++++++---------
2 files changed, 16 insertions(+), 16 deletions(-)

diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug
index 87aa6e92ee6e..7dad01729683 100644
--- a/arch/arm/Kconfig.debug
+++ b/arch/arm/Kconfig.debug
@@ -410,12 +410,12 @@ choice
Say Y here if you want kernel low-level debugging support
on i.MX25.

- config DEBUG_IMX21_IMX27_UART
- bool "i.MX21 and i.MX27 Debug UART"
- depends on SOC_IMX21 || SOC_IMX27
+ config DEBUG_IMX27_UART
+ bool "i.MX27 Debug UART"
+ depends on SOC_IMX27
help
Say Y here if you want kernel low-level debugging support
- on i.MX21 or i.MX27.
+ on i.MX27.

config DEBUG_IMX28_UART
bool "i.MX28 Debug UART"
@@ -1481,7 +1481,7 @@ config DEBUG_IMX_UART_PORT
int "i.MX Debug UART Port Selection"
depends on DEBUG_IMX1_UART || \
DEBUG_IMX25_UART || \
- DEBUG_IMX21_IMX27_UART || \
+ DEBUG_IMX27_UART || \
DEBUG_IMX31_UART || \
DEBUG_IMX35_UART || \
DEBUG_IMX50_UART || \
@@ -1540,12 +1540,12 @@ config DEBUG_LL_INCLUDE
default "debug/icedcc.S" if DEBUG_ICEDCC
default "debug/imx.S" if DEBUG_IMX1_UART || \
DEBUG_IMX25_UART || \
- DEBUG_IMX21_IMX27_UART || \
+ DEBUG_IMX27_UART || \
DEBUG_IMX31_UART || \
DEBUG_IMX35_UART || \
DEBUG_IMX50_UART || \
DEBUG_IMX51_UART || \
- DEBUG_IMX53_UART ||\
+ DEBUG_IMX53_UART || \
DEBUG_IMX6Q_UART || \
DEBUG_IMX6SL_UART || \
DEBUG_IMX6SX_UART || \
diff --git a/arch/arm/include/debug/imx-uart.h b/arch/arm/include/debug/imx-uart.h
index c8eb83d4b896..3edbb3c5b42b 100644
--- a/arch/arm/include/debug/imx-uart.h
+++ b/arch/arm/include/debug/imx-uart.h
@@ -11,13 +11,6 @@
#define IMX1_UART_BASE_ADDR(n) IMX1_UART##n##_BASE_ADDR
#define IMX1_UART_BASE(n) IMX1_UART_BASE_ADDR(n)

-#define IMX21_UART1_BASE_ADDR 0x1000a000
-#define IMX21_UART2_BASE_ADDR 0x1000b000
-#define IMX21_UART3_BASE_ADDR 0x1000c000
-#define IMX21_UART4_BASE_ADDR 0x1000d000
-#define IMX21_UART_BASE_ADDR(n) IMX21_UART##n##_BASE_ADDR
-#define IMX21_UART_BASE(n) IMX21_UART_BASE_ADDR(n)
-
#define IMX25_UART1_BASE_ADDR 0x43f90000
#define IMX25_UART2_BASE_ADDR 0x43f94000
#define IMX25_UART3_BASE_ADDR 0x5000c000
@@ -26,6 +19,13 @@
#define IMX25_UART_BASE_ADDR(n) IMX25_UART##n##_BASE_ADDR
#define IMX25_UART_BASE(n) IMX25_UART_BASE_ADDR(n)

+#define IMX27_UART1_BASE_ADDR 0x1000a000
+#define IMX27_UART2_BASE_ADDR 0x1000b000
+#define IMX27_UART3_BASE_ADDR 0x1000c000
+#define IMX27_UART4_BASE_ADDR 0x1000d000
+#define IMX27_UART_BASE_ADDR(n) IMX27_UART##n##_BASE_ADDR
+#define IMX27_UART_BASE(n) IMX27_UART_BASE_ADDR(n)
+
#define IMX31_UART1_BASE_ADDR 0x43f90000
#define IMX31_UART2_BASE_ADDR 0x43f94000
#define IMX31_UART3_BASE_ADDR 0x5000c000
@@ -112,10 +112,10 @@

#ifdef CONFIG_DEBUG_IMX1_UART
#define UART_PADDR IMX_DEBUG_UART_BASE(IMX1)
-#elif defined(CONFIG_DEBUG_IMX21_IMX27_UART)
-#define UART_PADDR IMX_DEBUG_UART_BASE(IMX21)
#elif defined(CONFIG_DEBUG_IMX25_UART)
#define UART_PADDR IMX_DEBUG_UART_BASE(IMX25)
+#elif defined(CONFIG_DEBUG_IMX27_UART)
+#define UART_PADDR IMX_DEBUG_UART_BASE(IMX27)
#elif defined(CONFIG_DEBUG_IMX31_UART)
#define UART_PADDR IMX_DEBUG_UART_BASE(IMX31)
#elif defined(CONFIG_DEBUG_IMX35_UART)
--
2.26.2

2021-10-28 14:22:17

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 10/13] arm: nomadik: drop selecting obsolete CLKSRC_NOMADIK_MTU_SCHED_CLOCK

Commit 85b6fcadcf66 ("clocksource/drivers/ux500: Drop Ux500 custom
SCHED_CLOCK") removes a sched_clock workaround and its corresponding
config CLKSRC_NOMADIK_MTU_SCHED_CLOCK. Since then, selecting
CLKSRC_NOMADIK_MTU_SCHED_CLOCK in ./arch/arm/mach-nomadik/Kconfig has no
effect and ./scripts/checkkconfigsymbols.py warns:

CLKSRC_NOMADIK_MTU_SCHED_CLOCK
Referencing files: arch/arm/mach-nomadik/Kconfig

Simply drop selecting the obsolete CLKSRC_NOMADIK_MTU_SCHED_CLOCK.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/mach-nomadik/Kconfig | 1 -
1 file changed, 1 deletion(-)

diff --git a/arch/arm/mach-nomadik/Kconfig b/arch/arm/mach-nomadik/Kconfig
index e98429be2b18..ab52b1abc453 100644
--- a/arch/arm/mach-nomadik/Kconfig
+++ b/arch/arm/mach-nomadik/Kconfig
@@ -5,7 +5,6 @@ menuconfig ARCH_NOMADIK
select ARM_AMBA
select ARM_VIC
select CLKSRC_NOMADIK_MTU
- select CLKSRC_NOMADIK_MTU_SCHED_CLOCK
select CPU_ARM926T
select GPIOLIB
select MFD_SYSCON
--
2.26.2

2021-10-28 14:22:21

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 11/13] arm: npcm: drop selecting non-existing ARM_ERRATA_794072

There is no and never was a Kconfig for ARM_ERRATA_794072 in the kernel
tree. So, there is no need to select ARM_ERRATA_794072 in
./arch/arm/mach-npcm/Kconfig.

Simply drop selecting the non-existing ARM_ERRATA_794072.

This issue was discovered with ./scripts/checkkconfigsymbols.py.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/mach-npcm/Kconfig | 1 -
1 file changed, 1 deletion(-)

diff --git a/arch/arm/mach-npcm/Kconfig b/arch/arm/mach-npcm/Kconfig
index a71cf1d189ae..9c137462fbe4 100644
--- a/arch/arm/mach-npcm/Kconfig
+++ b/arch/arm/mach-npcm/Kconfig
@@ -30,7 +30,6 @@ config ARCH_NPCM7XX
select ARM_ERRATA_764369 if SMP
select ARM_ERRATA_720789
select ARM_ERRATA_754322
- select ARM_ERRATA_794072
select PL310_ERRATA_588369
select PL310_ERRATA_727915
select MFD_SYSCON
--
2.26.2

2021-10-28 14:22:35

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 13/13] arm: pgtable: refer to intended CONFIG_ARM_LPAE in comment

Commit 5615f69bc209 ("ARM: 9016/2: Initialize the mapping of KASan shadow
memory") adds some nested ifdef's in ./arch/arm/mm/pgd.c, and follows the
good practice to annotate the endif's with a comment to indicate the
corresponding ifdef condition.

One comment annotation refers to CONFIG_LPAE, whereas the config is
actually called CONFIG_ARM_LPAE. That imprecision in a comment is probably
tolerable for all human readers.

However, the script ./scripts/checkkconfigsymbols.py, which checks the
kernel tree for references to non-existing Kconfig symbols, identifies and
reports that the reference to CONFIG_LPAE is invalid.

The script ./scripts/checkkconfigsymbols.py has been quite useful to
identify a number of bugs with Kconfig symbols and deserves to be executed
and checked regularly.

So, repair the comment to reduce the reports from this script and simplify
to use this script, as new issues are easier to spot when the list of
reports is shorter.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/mm/pgd.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/mm/pgd.c b/arch/arm/mm/pgd.c
index f8e9bc58a84f..bf2935f21c12 100644
--- a/arch/arm/mm/pgd.c
+++ b/arch/arm/mm/pgd.c
@@ -80,7 +80,7 @@ pgd_t *pgd_alloc(struct mm_struct *mm)
* sizeof(pmd_t));
clean_dcache_area(new_pmd, PTRS_PER_PMD * sizeof(pmd_t));
#endif /* CONFIG_KASAN */
-#endif /* CONFIG_LPAE */
+#endif /* CONFIG_ARM_LPAE */

if (!vectors_high()) {
/*
--
2.26.2

2021-10-28 14:23:03

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 04/13] arm: drop an obsolete ifdef with the removed config PCI_HOST_ITE8152

Commit 6da5238fa384 ("ARM: 8993/1: remove it8152 PCI controller driver")
removes the config PCI_HOST_ITE8152, but left a dangling obsolete ifndef
in ./arch/arm/kernel/bios32.c.

Hence, ./scripts/checkkconfigsymbols.py warns:

PCI_HOST_ITE8152
Referencing files: arch/arm/kernel/bios32.c

Remove this obsolete ifndef.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/kernel/bios32.c | 2 --
1 file changed, 2 deletions(-)

diff --git a/arch/arm/kernel/bios32.c b/arch/arm/kernel/bios32.c
index e7ef2b5bea9c..26b5d8361980 100644
--- a/arch/arm/kernel/bios32.c
+++ b/arch/arm/kernel/bios32.c
@@ -529,12 +529,10 @@ void pci_common_init_dev(struct device *parent, struct hw_pci *hw)
}
}

-#ifndef CONFIG_PCI_HOST_ITE8152
void pcibios_set_master(struct pci_dev *dev)
{
/* No special bus mastering setup handling */
}
-#endif

char * __init pcibios_setup(char *str)
{
--
2.26.2

2021-10-28 14:23:09

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 07/13] arm: imx: remove dead left-over from i.MX{27,31,35} removal

The commits:

commit 879c0e5e0ac7 ("ARM: imx: Remove i.MX27 board files")
commit c93197b0041d ("ARM: imx: Remove i.MX31 board files")
commit e1324ece2af4 ("ARM: imx: Remove i.MX35 board files")

remove the config MACH_MX27_3DS, MACH_MX31_3DS and MACH_MX35_3DS.
Commit a542fc18168c ("ARM: imx31: Remove remaining i.MX31 board code")
further removes arch/arm/mach-imx/3ds_debugboard.{c,h}. So, only some
dead left-over in Kconfig and Makefile remains.

Remove this remaining left-over.

This issue was identified with ./scripts/checkkconfigsymbols.py,
which warns on references to the non-existing configs
MACH_MX{27,31,35}_3DS in ./arch/arm/mach-imx/Kconfig.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/mach-imx/Kconfig | 12 ------------
arch/arm/mach-imx/Makefile | 2 --
2 files changed, 14 deletions(-)

diff --git a/arch/arm/mach-imx/Kconfig b/arch/arm/mach-imx/Kconfig
index b407b024dde3..f296bac467c8 100644
--- a/arch/arm/mach-imx/Kconfig
+++ b/arch/arm/mach-imx/Kconfig
@@ -21,18 +21,6 @@ config MXC_TZIC
config MXC_AVIC
bool

-config MXC_DEBUG_BOARD
- bool "Enable MXC debug board(for 3-stack)"
- depends on MACH_MX27_3DS || MACH_MX31_3DS || MACH_MX35_3DS
- help
- The debug board is an integral part of the MXC 3-stack(PDK)
- platforms, it can be attached or removed from the peripheral
- board. On debug board, several debug devices(ethernet, UART,
- buttons, LEDs and JTAG) are implemented. Between the MCU and
- these devices, a CPLD is added as a bridge which performs
- data/address de-multiplexing and decode, signal level shift,
- interrupt control and various board functions.
-
config HAVE_IMX_ANATOP
bool

diff --git a/arch/arm/mach-imx/Makefile b/arch/arm/mach-imx/Makefile
index d1506ef7a537..d5291ed9186a 100644
--- a/arch/arm/mach-imx/Makefile
+++ b/arch/arm/mach-imx/Makefile
@@ -14,8 +14,6 @@ obj-$(CONFIG_SOC_IMX5) += cpu-imx5.o $(imx5-pm-y)
obj-$(CONFIG_MXC_TZIC) += tzic.o
obj-$(CONFIG_MXC_AVIC) += avic.o

-obj-$(CONFIG_MXC_DEBUG_BOARD) += 3ds_debugboard.o
-
ifeq ($(CONFIG_CPU_IDLE),y)
obj-$(CONFIG_SOC_IMX5) += cpuidle-imx5.o
obj-$(CONFIG_SOC_IMX6Q) += cpuidle-imx6q.o
--
2.26.2

2021-10-28 14:23:15

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 09/13] arm: milbeaut: remove select of non-existing PINCTRL_MILBEAUT

The patch series "Add basic support for Socionext Milbeaut M10V SoC" (see
Link) introduced the config ARCH_MILBEAUT_M10V "Milbeaut SC2000/M10V
platform" in ./arch/arm/mach-milbeaut/ and intended to introduce timer,
clock, pinctrl and serial controller drivers.

However, during patch submission in March 2019, the introduction of the
milbeaut pinctrl driver was dropped from v2 to v3 of the patch series.
Since then, there was no further patch series to add this pinctrl driver
later on.

Hence, selecting PINCTRL_MILBEAUT in config is simply dangling and
referring to a non-existing config symbols.
Fortunately, ./scripts/checkkconfigsymbols.py warns:

PINCTRL_MILBEAUT
Referencing files: arch/arm/mach-milbeaut/Kconfig

Remove this select of the non-existing PINCTRL_MILBEAUT for now.

Link: https://lore.kernel.org/linux-arm-kernel/[email protected]/

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/mach-milbeaut/Kconfig | 1 -
1 file changed, 1 deletion(-)

diff --git a/arch/arm/mach-milbeaut/Kconfig b/arch/arm/mach-milbeaut/Kconfig
index 6a576fd8521e..f9d1006f9442 100644
--- a/arch/arm/mach-milbeaut/Kconfig
+++ b/arch/arm/mach-milbeaut/Kconfig
@@ -13,7 +13,6 @@ config ARCH_MILBEAUT_M10V
select ARM_ARCH_TIMER
select MILBEAUT_TIMER
select PINCTRL
- select PINCTRL_MILBEAUT
help
Support for Socionext's MILBEAUT M10V based systems

--
2.26.2

2021-10-28 14:23:49

by Lukas Bulwahn

[permalink] [raw]
Subject: [PATCH 12/13] arm: socfpga: always select PL310_ERRATA_753970

Commit fbc125afdc50 ("ARM: socfpga: Turn on ARM errata for L2 cache") adds
some selects in ./arch/arm/mach-socfpga/Kconfig, with one select being
conditionally selected on the non-existing Kconfig symbol PL310.

Hence, ./scripts/checkkconfigsymbols.py warns:

PL310
Referencing files: arch/arm/mach-socfpga/Kconfig

Assuming that this errata should actually be selected for
ARCH_INTEL_SOCFPGA, simply select this config unconditionally.

Signed-off-by: Lukas Bulwahn <[email protected]>
---
arch/arm/mach-socfpga/Kconfig | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/arch/arm/mach-socfpga/Kconfig b/arch/arm/mach-socfpga/Kconfig
index 43ddec677c0b..c21df17202a6 100644
--- a/arch/arm/mach-socfpga/Kconfig
+++ b/arch/arm/mach-socfpga/Kconfig
@@ -16,7 +16,7 @@ menuconfig ARCH_INTEL_SOCFPGA
select ARM_ERRATA_775420
select PL310_ERRATA_588369
select PL310_ERRATA_727915
- select PL310_ERRATA_753970 if PL310
+ select PL310_ERRATA_753970
select PL310_ERRATA_769419

if ARCH_INTEL_SOCFPGA
--
2.26.2

2021-10-28 14:40:53

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 01/13] arm: debug: remove obsolete debug code for DEBUG_ZTE_ZX

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit 89d4f98ae90d ("ARM: remove zte zx platform") removes the config
> DEBUG_ZTE_ZX. Hence, since then, the "ifdef CONFIG_DEBUG_ZTE_ZX" in
> ./arch/arm/include/debug/pl01x.S is dead code.
>
> Fortunately, ./scripts/checkkconfigsymbols.py detects this and warns:
>
> DEBUG_ZTE_ZX
> Referencing files: arch/arm/include/debug/pl01x.S
>
> So, remove the obsolete ifdef CONFIG_DEBUG_ZTE_ZX.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Arnd Bergmann <[email protected]>

I see another copy of these constants in drivers/tty/serial/amba-pl011.c,
which we should probably clean up as well.

Arnd

2021-10-28 14:41:50

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 02/13] arm: debug: reuse the config DEBUG_OMAP2UART{1,2} for OMAP{3,4,5}

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit d2b310b0234c ("ARM: debug: Use generic 8250 debug_ll for omap2 and
> omap3/4/5 common uarts") adds address definitions of DEBUG_UART_PHYS for
> OMAP2, OMAP3, OMAP4 and OMAP5 in ./arch/arm/Kconfig.debug.
>
> These definitions depend on DEBUG_OMAP{2,3,4,5}UART{1,2}; however, only
> DEBUG_OMAP2UART{1,2} are defined in ./arch/arm/Kconfig.debug, and
> DEBUG_OMAP{3,4,5}UART{1,2} are not defined. Hence, the script
> ./scripts/checkkconfigsymbols.py warns here on non-existing symbols.
> Simply reuse the config DEBUG_OMAP2UART{1,2}; there is no need to define
> separate config symbols for OMAP{3,4,5}. So, just delete the dead
> references to DEBUG_OMAP{3,4,5}UART{1,2}.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Arnd Bergmann <[email protected]>

2021-10-28 14:44:50

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 03/13] arm: Kconfig.debug: drop reference to removed ARCH_MSM

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit c0c89fafa289 ("ARM: Remove mach-msm and associated ARM architecture
> code") removes the definition of the config ARCH_MSM. Since then, the
> reference to ARCH_MSM in the dependencies of UNCOMPRESS_INCLUDE in
> Kconfig.debug is dead.
>
> Fortunately, ./scripts/checkkconfigsymbols.py warns:
>
> ARCH_MSM
> Referencing files: arch/arm/Kconfig.debug
>
> Drop the dependency on this removed config.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>
> ---
> arch/arm/Kconfig.debug | 2 +-
> 1 file changed, 1 insertion(+), 1 deletion(-)
>
> diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug
> index 83484564b1d9..87aa6e92ee6e 100644
> --- a/arch/arm/Kconfig.debug
> +++ b/arch/arm/Kconfig.debug
> @@ -1886,7 +1886,7 @@ 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

The PLAT_SAMSUNG reference is also misplaced here, I think you just want
ARCH_S3C24XX instead, since the other samsung ones already require
ARCH_MULTIPLATFORM.

Arnd

2021-10-28 14:49:10

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 05/13] arm: davinci: remove reference to obsolete BLK_DEV_PALMCHIP_BK3710

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit b7fb14d3ac63 ("ide: remove the legacy ide driver") removes the
> definition of the config BLK_DEV_PALMCHIP_BK3710.
>
> So, remove the obsolete references in the mach-davinci architecture.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Arnd Bergmann <[email protected]>

2021-10-28 14:50:39

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 08/13] arm: imx: rename DEBUG_IMX21_IMX27_UART to DEBUG_IMX27_UART

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Since commit 4b563a066611 ("ARM: imx: Remove imx21 support"), the config
> DEBUG_IMX21_IMX27_UART is really only debug support for IMX27.
>
> So, rename this option to DEBUG_IMX27_UART and adjust dependencies in
> Kconfig and rename the definitions to IMX27 as further clean-up.
>
> This issue was discovered with ./scripts/checkkconfigsymbols.py, which
> reported that DEBUG_IMX21_IMX27_UART depends on the non-existing config
> SOC_IMX21.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>


Reviewed-by: Arnd Bergmann <[email protected]>

2021-10-28 14:51:00

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 07/13] arm: imx: remove dead left-over from i.MX{27,31,35} removal

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> The commits:
>
> commit 879c0e5e0ac7 ("ARM: imx: Remove i.MX27 board files")
> commit c93197b0041d ("ARM: imx: Remove i.MX31 board files")
> commit e1324ece2af4 ("ARM: imx: Remove i.MX35 board files")
>
> remove the config MACH_MX27_3DS, MACH_MX31_3DS and MACH_MX35_3DS.
> Commit a542fc18168c ("ARM: imx31: Remove remaining i.MX31 board code")
> further removes arch/arm/mach-imx/3ds_debugboard.{c,h}. So, only some
> dead left-over in Kconfig and Makefile remains.
>
> Remove this remaining left-over.
>
> This issue was identified with ./scripts/checkkconfigsymbols.py,
> which warns on references to the non-existing configs
> MACH_MX{27,31,35}_3DS in ./arch/arm/mach-imx/Kconfig.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Arnd Bergmann <[email protected]>

2021-10-28 14:51:31

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 04/13] arm: drop an obsolete ifdef with the removed config PCI_HOST_ITE8152

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit 6da5238fa384 ("ARM: 8993/1: remove it8152 PCI controller driver")
> removes the config PCI_HOST_ITE8152, but left a dangling obsolete ifndef
> in ./arch/arm/kernel/bios32.c.
>
> Hence, ./scripts/checkkconfigsymbols.py warns:
>
> PCI_HOST_ITE8152
> Referencing files: arch/arm/kernel/bios32.c
>
> Remove this obsolete ifndef.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Arnd Bergmann <[email protected]>

I wonder if we should just remove this function and use the
(non-empty) default version instead.

96c5590058d7 ("PCI: Pull PCI 'latency timer' setup up into the core")
introduced that generic version, and I suspect the arm version
was left out by mistake, but it's not clear from that patch.

Arnd

2021-10-28 14:53:14

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 06/13] arm: ixp4xx: remove dead configs CPU_IXP43X and CPU_IXP46X

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit 73d04ca5f4ac ("ARM: ixp4xx: Delete Intel reference design
> boardfiles") removes the definition of the configs MACH_IXDP465 and
> MACH_KIXRP435, but misses to remove the configs CPU_IXP43X and CPU_IXP46X
> that depend on those removed configs, and hence are dead now.
>
> Fortunately, ./scripts/checkkconfigsymbols.py warns:
>
> MACH_IXDP465
> Referencing files: arch/arm/mach-ixp4xx/Kconfig
>
> MACH_KIXRP435
> Referencing files: arch/arm/mach-ixp4xx/Kconfig
>
> Remove the dead configs CPU_IXP43X and CPU_IXP46X.
>
> A further quick grep for the name of those two symbols did not show any
> use of the two config symbols; so, there are no further clean-up activities
> beyond this config removal needed.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Arnd Bergmann <[email protected]>

2021-10-28 14:58:05

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 10/13] arm: nomadik: drop selecting obsolete CLKSRC_NOMADIK_MTU_SCHED_CLOCK

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit 85b6fcadcf66 ("clocksource/drivers/ux500: Drop Ux500 custom
> SCHED_CLOCK") removes a sched_clock workaround and its corresponding
> config CLKSRC_NOMADIK_MTU_SCHED_CLOCK. Since then, selecting
> CLKSRC_NOMADIK_MTU_SCHED_CLOCK in ./arch/arm/mach-nomadik/Kconfig has no
> effect and ./scripts/checkkconfigsymbols.py warns:
>
> CLKSRC_NOMADIK_MTU_SCHED_CLOCK
> Referencing files: arch/arm/mach-nomadik/Kconfig
>
> Simply drop selecting the obsolete CLKSRC_NOMADIK_MTU_SCHED_CLOCK.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Arnd Bergmann <[email protected]>

2021-10-28 14:58:46

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 09/13] arm: milbeaut: remove select of non-existing PINCTRL_MILBEAUT

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> The patch series "Add basic support for Socionext Milbeaut M10V SoC" (see
> Link) introduced the config ARCH_MILBEAUT_M10V "Milbeaut SC2000/M10V
> platform" in ./arch/arm/mach-milbeaut/ and intended to introduce timer,
> clock, pinctrl and serial controller drivers.
>
> However, during patch submission in March 2019, the introduction of the
> milbeaut pinctrl driver was dropped from v2 to v3 of the patch series.
> Since then, there was no further patch series to add this pinctrl driver
> later on.
>
> Hence, selecting PINCTRL_MILBEAUT in config is simply dangling and
> referring to a non-existing config symbols.
> Fortunately, ./scripts/checkkconfigsymbols.py warns:
>
> PINCTRL_MILBEAUT
> Referencing files: arch/arm/mach-milbeaut/Kconfig
>
> Remove this select of the non-existing PINCTRL_MILBEAUT for now.
>
> Link: https://lore.kernel.org/linux-arm-kernel/[email protected]/
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

I would take that as an indication that there is no interest in supporting this
platform upstream any more, the version we merged probably never worked
without the rest of the drivers.

I've added the original authors of the other drivers to Cc. Should we remove
all of this?

Arnd

2021-10-28 14:59:45

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 11/13] arm: npcm: drop selecting non-existing ARM_ERRATA_794072

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> There is no and never was a Kconfig for ARM_ERRATA_794072 in the kernel
> tree. So, there is no need to select ARM_ERRATA_794072 in
> ./arch/arm/mach-npcm/Kconfig.
>
> Simply drop selecting the non-existing ARM_ERRATA_794072.
>
> This issue was discovered with ./scripts/checkkconfigsymbols.py.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>
> ---

Could this be a typo? Maybe we need to enable a different errata workaround
here, or maybe that code is actually needed and has to get sent.

Arnd

2021-10-28 15:01:16

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 12/13] arm: socfpga: always select PL310_ERRATA_753970

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit fbc125afdc50 ("ARM: socfpga: Turn on ARM errata for L2 cache") adds
> some selects in ./arch/arm/mach-socfpga/Kconfig, with one select being
> conditionally selected on the non-existing Kconfig symbol PL310.
>
> Hence, ./scripts/checkkconfigsymbols.py warns:
>
> PL310
> Referencing files: arch/arm/mach-socfpga/Kconfig
>
> Assuming that this errata should actually be selected for
> ARCH_INTEL_SOCFPGA, simply select this config unconditionally.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

I think it has to be

select PL310_ERRATA_753970 if CACHE_L2X0

otherwise you get a Kconfig warning when building a kernel without the
L2X0 driver.

Arnd

2021-10-28 15:04:38

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 13/13] arm: pgtable: refer to intended CONFIG_ARM_LPAE in comment

On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>
> Commit 5615f69bc209 ("ARM: 9016/2: Initialize the mapping of KASan shadow
> memory") adds some nested ifdef's in ./arch/arm/mm/pgd.c, and follows the
> good practice to annotate the endif's with a comment to indicate the
> corresponding ifdef condition.
>
> One comment annotation refers to CONFIG_LPAE, whereas the config is
> actually called CONFIG_ARM_LPAE. That imprecision in a comment is probably
> tolerable for all human readers.
>
> However, the script ./scripts/checkkconfigsymbols.py, which checks the
> kernel tree for references to non-existing Kconfig symbols, identifies and
> reports that the reference to CONFIG_LPAE is invalid.
>
> The script ./scripts/checkkconfigsymbols.py has been quite useful to
> identify a number of bugs with Kconfig symbols and deserves to be executed
> and checked regularly.
>
> So, repair the comment to reduce the reports from this script and simplify
> to use this script, as new issues are easier to spot when the list of
> reports is shorter.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Arnd Bergmann <[email protected]>

2021-10-28 15:37:40

by Russell King (Oracle)

[permalink] [raw]
Subject: Re: [PATCH 04/13] arm: drop an obsolete ifdef with the removed config PCI_HOST_ITE8152

On Thu, Oct 28, 2021 at 04:46:38PM +0200, Arnd Bergmann wrote:
> On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> >
> > Commit 6da5238fa384 ("ARM: 8993/1: remove it8152 PCI controller driver")
> > removes the config PCI_HOST_ITE8152, but left a dangling obsolete ifndef
> > in ./arch/arm/kernel/bios32.c.
> >
> > Hence, ./scripts/checkkconfigsymbols.py warns:
> >
> > PCI_HOST_ITE8152
> > Referencing files: arch/arm/kernel/bios32.c
> >
> > Remove this obsolete ifndef.
> >
> > Signed-off-by: Lukas Bulwahn <[email protected]>
>
> Reviewed-by: Arnd Bergmann <[email protected]>
>
> I wonder if we should just remove this function and use the
> (non-empty) default version instead.
>
> 96c5590058d7 ("PCI: Pull PCI 'latency timer' setup up into the core")
> introduced that generic version, and I suspect the arm version
> was left out by mistake, but it's not clear from that patch.

That was because PCI_HOST_ITE8152 needed something different from the
"do nothing" default (setting the PCI latency timer to default to 64
as the new generic code did.)

--
RMK's Patch system: https://www.armlinux.org.uk/developer/patches/
FTTP is here! 40Mbps down 10Mbps up. Decent connectivity at last!

2021-10-29 05:37:15

by Lukas Bulwahn

[permalink] [raw]
Subject: Re: [PATCH 01/13] arm: debug: remove obsolete debug code for DEBUG_ZTE_ZX

On Thu, Oct 28, 2021 at 4:38 PM Arnd Bergmann <[email protected]> wrote:
>
> On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> >
> > Commit 89d4f98ae90d ("ARM: remove zte zx platform") removes the config
> > DEBUG_ZTE_ZX. Hence, since then, the "ifdef CONFIG_DEBUG_ZTE_ZX" in
> > ./arch/arm/include/debug/pl01x.S is dead code.
> >
> > Fortunately, ./scripts/checkkconfigsymbols.py detects this and warns:
> >
> > DEBUG_ZTE_ZX
> > Referencing files: arch/arm/include/debug/pl01x.S
> >
> > So, remove the obsolete ifdef CONFIG_DEBUG_ZTE_ZX.
> >
> > Signed-off-by: Lukas Bulwahn <[email protected]>
>
> Reviewed-by: Arnd Bergmann <[email protected]>
>
> I see another copy of these constants in drivers/tty/serial/amba-pl011.c,
> which we should probably clean up as well.
>

Arnd, did you have something like this---see below---for serial
amba-pl011 in mind?

Then, I would adjust the patch to remove all the zte_zx serial
left-over in one commit.

Lukas

diff --git a/drivers/tty/serial/amba-pl011.c b/drivers/tty/serial/amba-pl011.c
index d361cd84ff8c..c9534e229166 100644
--- a/drivers/tty/serial/amba-pl011.c
+++ b/drivers/tty/serial/amba-pl011.c
@@ -188,38 +188,6 @@ static struct vendor_data vendor_st = {
.get_fifosize = get_fifosize_st,
};

-static const u16 pl011_zte_offsets[REG_ARRAY_SIZE] = {
- [REG_DR] = ZX_UART011_DR,
- [REG_FR] = ZX_UART011_FR,
- [REG_LCRH_RX] = ZX_UART011_LCRH,
- [REG_LCRH_TX] = ZX_UART011_LCRH,
- [REG_IBRD] = ZX_UART011_IBRD,
- [REG_FBRD] = ZX_UART011_FBRD,
- [REG_CR] = ZX_UART011_CR,
- [REG_IFLS] = ZX_UART011_IFLS,
- [REG_IMSC] = ZX_UART011_IMSC,
- [REG_RIS] = ZX_UART011_RIS,
- [REG_MIS] = ZX_UART011_MIS,
- [REG_ICR] = ZX_UART011_ICR,
- [REG_DMACR] = ZX_UART011_DMACR,
-};
-
-static unsigned int get_fifosize_zte(struct amba_device *dev)
-{
- return 16;
-}
-
-static struct vendor_data vendor_zte = {
- .reg_offset = pl011_zte_offsets,
- .access_32b = true,
- .ifls = UART011_IFLS_RX4_8|UART011_IFLS_TX4_8,
- .fr_busy = ZX_UART01x_FR_BUSY,
- .fr_dsr = ZX_UART01x_FR_DSR,
- .fr_cts = ZX_UART01x_FR_CTS,
- .fr_ri = ZX_UART011_FR_RI,
- .get_fifosize = get_fifosize_zte,
-};
-
/* Deals with DMA transactions */

struct pl011_sgbuf {
@@ -2974,11 +2942,6 @@ static const struct amba_id pl011_ids[] = {
.mask = 0x00ffffff,
.data = &vendor_st,
},
- {
- .id = AMBA_LINUX_ID(0x00, 0x1, 0xffe),
- .mask = 0x00ffffff,
- .data = &vendor_zte,
- },
{ 0, 0 },
};

diff --git a/include/linux/amba/bus.h b/include/linux/amba/bus.h
index edfcf7a14dcd..6c7f47846971 100644
--- a/include/linux/amba/bus.h
+++ b/include/linux/amba/bus.h
@@ -90,14 +90,8 @@ enum amba_vendor {
AMBA_VENDOR_ST = 0x80,
AMBA_VENDOR_QCOM = 0x51,
AMBA_VENDOR_LSI = 0xb6,
- AMBA_VENDOR_LINUX = 0xfe, /* This value is not official */
};

-/* This is used to generate pseudo-ID for AMBA device */
-#define AMBA_LINUX_ID(conf, rev, part) \
- (((conf) & 0xff) << 24 | ((rev) & 0xf) << 20 | \
- AMBA_VENDOR_LINUX << 12 | ((part) & 0xfff))
-
extern struct bus_type amba_bustype;

#define to_amba_device(d) container_of(d, struct amba_device, dev)

2021-10-29 05:56:09

by Lukas Bulwahn

[permalink] [raw]
Subject: Re: [PATCH 04/13] arm: drop an obsolete ifdef with the removed config PCI_HOST_ITE8152

On Thu, Oct 28, 2021 at 5:35 PM Russell King (Oracle)
<[email protected]> wrote:
>
> On Thu, Oct 28, 2021 at 04:46:38PM +0200, Arnd Bergmann wrote:
> > On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> > >
> > > Commit 6da5238fa384 ("ARM: 8993/1: remove it8152 PCI controller driver")
> > > removes the config PCI_HOST_ITE8152, but left a dangling obsolete ifndef
> > > in ./arch/arm/kernel/bios32.c.
> > >
> > > Hence, ./scripts/checkkconfigsymbols.py warns:
> > >
> > > PCI_HOST_ITE8152
> > > Referencing files: arch/arm/kernel/bios32.c
> > >
> > > Remove this obsolete ifndef.
> > >
> > > Signed-off-by: Lukas Bulwahn <[email protected]>
> >
> > Reviewed-by: Arnd Bergmann <[email protected]>
> >
> > I wonder if we should just remove this function and use the
> > (non-empty) default version instead.
> >
> > 96c5590058d7 ("PCI: Pull PCI 'latency timer' setup up into the core")
> > introduced that generic version, and I suspect the arm version
> > was left out by mistake, but it's not clear from that patch.
>
> That was because PCI_HOST_ITE8152 needed something different from the
> "do nothing" default (setting the PCI latency timer to default to 64
> as the new generic code did.)
>

So, can we just drop the empty pcibios_set_master() function in
bios32.c and the pci handling will now (after the removal of
PCI_HOST_ITE8152) just do The Right Thing(TM)?

If you can confirm that, I will send an updated patch here.

Thanks,

Lukas

2021-10-29 06:12:46

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 01/13] arm: debug: remove obsolete debug code for DEBUG_ZTE_ZX

On Fri, Oct 29, 2021 at 7:34 AM Lukas Bulwahn <[email protected]> wrote:
> On Thu, Oct 28, 2021 at 4:38 PM Arnd Bergmann <[email protected]> wrote:
> >
>
> Arnd, did you have something like this---see below---for serial
> amba-pl011 in mind?
>
> Then, I would adjust the patch to remove all the zte_zx serial
> left-over in one commit.
>

Yes, looks good to me.

Arnd

2021-10-29 06:38:43

by Joel Stanley

[permalink] [raw]
Subject: Re: [PATCH 11/13] arm: npcm: drop selecting non-existing ARM_ERRATA_794072

On Thu, 28 Oct 2021 at 14:57, Arnd Bergmann <[email protected]> wrote:
>
> On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> >
> > There is no and never was a Kconfig for ARM_ERRATA_794072 in the kernel
> > tree. So, there is no need to select ARM_ERRATA_794072 in
> > ./arch/arm/mach-npcm/Kconfig.
> >
> > Simply drop selecting the non-existing ARM_ERRATA_794072.
> >
> > This issue was discovered with ./scripts/checkkconfigsymbols.py.
> >
> > Signed-off-by: Lukas Bulwahn <[email protected]>
> > ---
>
> Could this be a typo? Maybe we need to enable a different errata workaround
> here, or maybe that code is actually needed and has to get sent.

Doing some searching, u-boot had a workaround for something called
ARM_ERRATA_794072.

https://github.com/u-boot/u-boot/commit/f71cbfe3ca5d2ad20159871700e8e248c8818ba8

Lore has the review history for that patch:

https://lore.kernel.org/all/6be32e0b5b454ed7b609317266a8e798@BLUPR03MB358.namprd03.prod.outlook.com/

It looks like it's the same workaround as ARM_ERRATA_742230, which the
kernel does implement.

It would be good to hear from the Nuvoton people, or an Arm person.

Cheers,

Joel

2021-11-01 13:56:05

by Sugaya Taichi

[permalink] [raw]
Subject: Re: [PATCH 09/13] arm: milbeaut: remove select of non-existing PINCTRL_MILBEAUT



On 2021/10/28 23:55, Arnd Bergmann wrote:
> On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
>>
>> The patch series "Add basic support for Socionext Milbeaut M10V SoC" (see
>> Link) introduced the config ARCH_MILBEAUT_M10V "Milbeaut SC2000/M10V
>> platform" in ./arch/arm/mach-milbeaut/ and intended to introduce timer,
>> clock, pinctrl and serial controller drivers.
>>
>> However, during patch submission in March 2019, the introduction of the
>> milbeaut pinctrl driver was dropped from v2 to v3 of the patch series.
>> Since then, there was no further patch series to add this pinctrl driver
>> later on.
>>
>> Hence, selecting PINCTRL_MILBEAUT in config is simply dangling and
>> referring to a non-existing config symbols.
>> Fortunately, ./scripts/checkkconfigsymbols.py warns:
>>
>> PINCTRL_MILBEAUT
>> Referencing files: arch/arm/mach-milbeaut/Kconfig
>>
>> Remove this select of the non-existing PINCTRL_MILBEAUT for now.
>>
>> Link: https://lore.kernel.org/linux-arm-kernel/[email protected]/
>>
>> Signed-off-by: Lukas Bulwahn <[email protected]>
>
> I would take that as an indication that there is no interest in supporting this
> platform upstream any more, the version we merged probably never worked
> without the rest of the drivers.
>
> I've added the original authors of the other drivers to Cc. Should we remove
> all of this?
>
> Arnd
>

It is okay to drop PINCTRL_MILBEAUT. I will add it again when working at the pinctrl
driver. But don`t remove the milbeaut platform. Actually we haven't been doing
maintenance recently, but we have plans to add drivers in the future.

Thanks,
Taichi Sugaya

2021-11-02 07:34:14

by Lukas Bulwahn

[permalink] [raw]
Subject: Re: [PATCH 11/13] arm: npcm: drop selecting non-existing ARM_ERRATA_794072

On Fri, Oct 29, 2021 at 8:36 AM Joel Stanley <[email protected]> wrote:
>
> On Thu, 28 Oct 2021 at 14:57, Arnd Bergmann <[email protected]> wrote:
> >
> > On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> > >
> > > There is no and never was a Kconfig for ARM_ERRATA_794072 in the kernel
> > > tree. So, there is no need to select ARM_ERRATA_794072 in
> > > ./arch/arm/mach-npcm/Kconfig.
> > >
> > > Simply drop selecting the non-existing ARM_ERRATA_794072.
> > >
> > > This issue was discovered with ./scripts/checkkconfigsymbols.py.
> > >
> > > Signed-off-by: Lukas Bulwahn <[email protected]>
> > > ---
> >
> > Could this be a typo? Maybe we need to enable a different errata workaround
> > here, or maybe that code is actually needed and has to get sent.
>
> Doing some searching, u-boot had a workaround for something called
> ARM_ERRATA_794072.
>
> https://github.com/u-boot/u-boot/commit/f71cbfe3ca5d2ad20159871700e8e248c8818ba8
>
> Lore has the review history for that patch:
>
> https://lore.kernel.org/all/6be32e0b5b454ed7b609317266a8e798@BLUPR03MB358.namprd03.prod.outlook.com/
>
> It looks like it's the same workaround as ARM_ERRATA_742230, which the
> kernel does implement.
>
> It would be good to hear from the Nuvoton people, or an Arm person.
>

I will happily update the patch to select ARM_ERRATA_742230 instead of
the dead non-existing ARM_ERRATA_794072.

In contrast to the current patch that basically only cleans up "dead
config" and has no effective functional change, the new patch would
change the behaviour. I cannot test this patch (beyond some basic
compile test) on the hardware; so, we certainly need someone to have
that hardware, knows how to test it or confirm otherwise that we
should select the ARM_ERRATA_742230 fix for this hardware.

The current patch should be subsumed by the new patch; the submission
of the new patch is deferred until that person shows up. Let's see.

Lukas

2021-11-02 07:57:39

by Lukas Bulwahn

[permalink] [raw]
Subject: Re: [PATCH 03/13] arm: Kconfig.debug: drop reference to removed ARCH_MSM

On Thu, Oct 28, 2021 at 4:42 PM Arnd Bergmann <[email protected]> wrote:
>
> On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> >
> > Commit c0c89fafa289 ("ARM: Remove mach-msm and associated ARM architecture
> > code") removes the definition of the config ARCH_MSM. Since then, the
> > reference to ARCH_MSM in the dependencies of UNCOMPRESS_INCLUDE in
> > Kconfig.debug is dead.
> >
> > Fortunately, ./scripts/checkkconfigsymbols.py warns:
> >
> > ARCH_MSM
> > Referencing files: arch/arm/Kconfig.debug
> >
> > Drop the dependency on this removed config.
> >
> > Signed-off-by: Lukas Bulwahn <[email protected]>
> > ---
> > arch/arm/Kconfig.debug | 2 +-
> > 1 file changed, 1 insertion(+), 1 deletion(-)
> >
> > diff --git a/arch/arm/Kconfig.debug b/arch/arm/Kconfig.debug
> > index 83484564b1d9..87aa6e92ee6e 100644
> > --- a/arch/arm/Kconfig.debug
> > +++ b/arch/arm/Kconfig.debug
> > @@ -1886,7 +1886,7 @@ 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
>
> The PLAT_SAMSUNG reference is also misplaced here, I think you just want
> ARCH_S3C24XX instead, since the other samsung ones already require
> ARCH_MULTIPLATFORM.
>

Agree. I can clean up (or better stated: optimize) the dependencies
further in such a way. But that config dependency optimization goes
beyond just removing dead symbols and deserves to be its own patch.
Patch will follow later this week.

Lukas

2021-11-02 08:13:53

by Arnd Bergmann

[permalink] [raw]
Subject: Re: [PATCH 11/13] arm: npcm: drop selecting non-existing ARM_ERRATA_794072

On Tue, Nov 2, 2021 at 8:31 AM Lukas Bulwahn <[email protected]> wrote:
> On Fri, Oct 29, 2021 at 8:36 AM Joel Stanley <[email protected]> wrote:
> > On Thu, 28 Oct 2021 at 14:57, Arnd Bergmann <[email protected]> wrote:
> > > On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> > https://lore.kernel.org/all/6be32e0b5b454ed7b609317266a8e798@BLUPR03MB358.namprd03.prod.outlook.com/
> >
> > It looks like it's the same workaround as ARM_ERRATA_742230, which the
> > kernel does implement.
> >
> > It would be good to hear from the Nuvoton people, or an Arm person.
>
> I will happily update the patch to select ARM_ERRATA_742230 instead of
> the dead non-existing ARM_ERRATA_794072.
>
> In contrast to the current patch that basically only cleans up "dead
> config" and has no effective functional change, the new patch would
> change the behaviour. I cannot test this patch (beyond some basic
> compile test) on the hardware; so, we certainly need someone to have
> that hardware, knows how to test it or confirm otherwise that we
> should select the ARM_ERRATA_742230 fix for this hardware.
>
> The current patch should be subsumed by the new patch; the submission
> of the new patch is deferred until that person shows up. Let's see.

I'd prefer to leave the broken Kconfig symbol in place as a reminder until it
gets fixed properly then.

Arnd

2021-11-02 08:24:44

by Avi Fishman

[permalink] [raw]
Subject: Re: [PATCH 11/13] arm: npcm: drop selecting non-existing ARM_ERRATA_794072

Hi,

At Nuvoton we implied this WA in the past, not because we encountered
a specific problem but since the errata says so and we saw this in
other patches like:
https://patchwork.kernel.org/project/linux-arm-kernel/patch/[email protected]/
But we didn't upstream the arch/arm/mm/proc-v7.S

From the errata document.
"794072 A short loop including a DMB instruction might cause a denial
of service on
another processor which executes a CP15 broadcast operation
Status
Affects:
Fault Type:
Fault Status: Cortex-A9 MPCore.
Programmer Category B
Present in: All r1, r2, r3 and r4 revisions Open
Description
A processor which continuously executes a short loop containing a DMB
instruction might prevent a CP15
operation broadcast by another processor making further progress,
causing a denial of service.
Configurations affected
This erratum affects all Cortex-A9 MPCore processors with two or more
processors.
Conditions
The erratum requires the following conditions:
- Two or more processors are working in SMP mode (ACTLR.SMP=1)
- One of the processors continuously executes a short loop containing
at least one DMB instruction.
- Another processor executes a CP15 maintenance operation that is
broadcast. This requires that this
processor has enabled the broadcasting of CP15 operations (ACTLR.FW=1)
For the erratum to occur, the short loop containing the DMB
instruction must meet all of the following additional
conditions:
- No more than 10 instructions other than the DMB are executed between each DMB
- No non-conditional Load or Store, or conditional Load or Store which
pass the condition code check, are
executed between each DMB
When all the conditions for the erratum are met, the short loop
creates a continuous stream of DMB instructions.
This might cause a denial of service, by preventing the processor
executing the short loop from executing the
received broadcast CP15 operation. As a result, the processor that
originally executed the broadcast CP15
operation is stalled until the execution of the loop is interrupted.
Note that because the process issuing the CP15 broadcast operation
cannot complete operation, it cannot enter
any debug-mode, and cannot take any interrupt. If the processor
executing the short loop also cannot be
interrupted, for example if it has disabled its interrupts, or if no
interrupts are routed to this processor, this
erratum might cause a system livelock.
Implications
The erratum might create performance issues, or in the worst case it
might cause a system livelock if the
processor executing the DMB is in an infinite loop that cannot be interrupted.
Workaround
This erratum can be worked round by setting bit[4] of the undocumented
Diagnostic Control Register to 1. This
register is encoded as CP15 c15 0 c0 1.
This bit can be written in Secure state only, with the following
Read/Modify/Write code sequence:
MRC p15,0,rt,c15,c0,1
ORR rt,rt,#0x10
MCR p15,0,rt,c15,c0,1
When it is set, this bit causes the DMB instruction to be decoded and
executed like a DSB.
Using this software workaround is not expected to have any impact on
the overall performance of the processor
on a typical code base.
Other workarounds are also available for this erratum, to either
prevent or interrupt the continuous stream of
DMB instructions that causes the deadlock. For example:
- Inserting a non-conditional Load or Store instruction in the loop
between each DMB
- Inserting additional instructions in the loop, such as NOPs, to
avoid the processor seeing back to back
DMB instructions.
- Making the processor executing the short loop take regular interrupts."

Avi

On Tue, Nov 2, 2021 at 9:31 AM Lukas Bulwahn <[email protected]> wrote:
>
> On Fri, Oct 29, 2021 at 8:36 AM Joel Stanley <[email protected]> wrote:
> >
> > On Thu, 28 Oct 2021 at 14:57, Arnd Bergmann <[email protected]> wrote:
> > >
> > > On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> > > >
> > > > There is no and never was a Kconfig for ARM_ERRATA_794072 in the kernel
> > > > tree. So, there is no need to select ARM_ERRATA_794072 in
> > > > ./arch/arm/mach-npcm/Kconfig.
> > > >
> > > > Simply drop selecting the non-existing ARM_ERRATA_794072.
> > > >
> > > > This issue was discovered with ./scripts/checkkconfigsymbols.py.
> > > >
> > > > Signed-off-by: Lukas Bulwahn <[email protected]>
> > > > ---
> > >
> > > Could this be a typo? Maybe we need to enable a different errata workaround
> > > here, or maybe that code is actually needed and has to get sent.
> >
> > Doing some searching, u-boot had a workaround for something called
> > ARM_ERRATA_794072.
> >
> > https://github.com/u-boot/u-boot/commit/f71cbfe3ca5d2ad20159871700e8e248c8818ba8
> >
> > Lore has the review history for that patch:
> >
> > https://lore.kernel.org/all/6be32e0b5b454ed7b609317266a8e798@BLUPR03MB358.namprd03.prod.outlook.com/
> >
> > It looks like it's the same workaround as ARM_ERRATA_742230, which the
> > kernel does implement.
> >
> > It would be good to hear from the Nuvoton people, or an Arm person.
> >
>
> I will happily update the patch to select ARM_ERRATA_742230 instead of
> the dead non-existing ARM_ERRATA_794072.
>
> In contrast to the current patch that basically only cleans up "dead
> config" and has no effective functional change, the new patch would
> change the behaviour. I cannot test this patch (beyond some basic
> compile test) on the hardware; so, we certainly need someone to have
> that hardware, knows how to test it or confirm otherwise that we
> should select the ARM_ERRATA_742230 fix for this hardware.
>
> The current patch should be subsumed by the new patch; the submission
> of the new patch is deferred until that person shows up. Let's see.
>
> Lukas



--
Regards,
Avi

2021-11-02 08:35:26

by Avi Fishman

[permalink] [raw]
Subject: Re: [PATCH 11/13] arm: npcm: drop selecting non-existing ARM_ERRATA_794072

On Tue, Nov 2, 2021 at 10:22 AM Avi Fishman <[email protected]> wrote:
>
> Hi,
>
> At Nuvoton we implied this WA in the past, not because we encountered
> a specific problem but since the errata says so and we saw this in
> other patches like:
> https://patchwork.kernel.org/project/linux-arm-kernel/patch/[email protected]/
> But we didn't upstream the arch/arm/mm/proc-v7.S
>
> From the errata document.
> "794072 A short loop including a DMB instruction might cause a denial
> of service on
> another processor which executes a CP15 broadcast operation
> Status
> Affects:
> Fault Type:
> Fault Status: Cortex-A9 MPCore.
> Programmer Category B
> Present in: All r1, r2, r3 and r4 revisions Open
> Description
> A processor which continuously executes a short loop containing a DMB
> instruction might prevent a CP15
> operation broadcast by another processor making further progress,
> causing a denial of service.
> Configurations affected
> This erratum affects all Cortex-A9 MPCore processors with two or more
> processors.
> Conditions
> The erratum requires the following conditions:
> - Two or more processors are working in SMP mode (ACTLR.SMP=1)
> - One of the processors continuously executes a short loop containing
> at least one DMB instruction.
> - Another processor executes a CP15 maintenance operation that is
> broadcast. This requires that this
> processor has enabled the broadcasting of CP15 operations (ACTLR.FW=1)
> For the erratum to occur, the short loop containing the DMB
> instruction must meet all of the following additional
> conditions:
> - No more than 10 instructions other than the DMB are executed between each DMB
> - No non-conditional Load or Store, or conditional Load or Store which
> pass the condition code check, are
> executed between each DMB
> When all the conditions for the erratum are met, the short loop
> creates a continuous stream of DMB instructions.
> This might cause a denial of service, by preventing the processor
> executing the short loop from executing the
> received broadcast CP15 operation. As a result, the processor that
> originally executed the broadcast CP15
> operation is stalled until the execution of the loop is interrupted.
> Note that because the process issuing the CP15 broadcast operation
> cannot complete operation, it cannot enter
> any debug-mode, and cannot take any interrupt. If the processor
> executing the short loop also cannot be
> interrupted, for example if it has disabled its interrupts, or if no
> interrupts are routed to this processor, this
> erratum might cause a system livelock.
> Implications
> The erratum might create performance issues, or in the worst case it
> might cause a system livelock if the
> processor executing the DMB is in an infinite loop that cannot be interrupted.
> Workaround
> This erratum can be worked round by setting bit[4] of the undocumented
> Diagnostic Control Register to 1. This
> register is encoded as CP15 c15 0 c0 1.
> This bit can be written in Secure state only, with the following
> Read/Modify/Write code sequence:
> MRC p15,0,rt,c15,c0,1
> ORR rt,rt,#0x10
> MCR p15,0,rt,c15,c0,1
> When it is set, this bit causes the DMB instruction to be decoded and
> executed like a DSB.
> Using this software workaround is not expected to have any impact on
> the overall performance of the processor
> on a typical code base.
> Other workarounds are also available for this erratum, to either
> prevent or interrupt the continuous stream of
> DMB instructions that causes the deadlock. For example:
> - Inserting a non-conditional Load or Store instruction in the loop
> between each DMB
> - Inserting additional instructions in the loop, such as NOPs, to
> avoid the processor seeing back to back
> DMB instructions.
> - Making the processor executing the short loop take regular interrupts."
>
> Avi
>
> On Tue, Nov 2, 2021 at 9:31 AM Lukas Bulwahn <[email protected]> wrote:
> >
> > On Fri, Oct 29, 2021 at 8:36 AM Joel Stanley <[email protected]> wrote:
> > >
> > > On Thu, 28 Oct 2021 at 14:57, Arnd Bergmann <[email protected]> wrote:
> > > >
> > > > On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> > > > >
> > > > > There is no and never was a Kconfig for ARM_ERRATA_794072 in the kernel
> > > > > tree. So, there is no need to select ARM_ERRATA_794072 in
> > > > > ./arch/arm/mach-npcm/Kconfig.
> > > > >
> > > > > Simply drop selecting the non-existing ARM_ERRATA_794072.
> > > > >
> > > > > This issue was discovered with ./scripts/checkkconfigsymbols.py.
> > > > >
> > > > > Signed-off-by: Lukas Bulwahn <[email protected]>
> > > > > ---
> > > >
> > > > Could this be a typo? Maybe we need to enable a different errata workaround
> > > > here, or maybe that code is actually needed and has to get sent.
> > >
> > > Doing some searching, u-boot had a workaround for something called
> > > ARM_ERRATA_794072.
> > >
> > > https://github.com/u-boot/u-boot/commit/f71cbfe3ca5d2ad20159871700e8e248c8818ba8
> > >
> > > Lore has the review history for that patch:
> > >
> > > https://lore.kernel.org/all/6be32e0b5b454ed7b609317266a8e798@BLUPR03MB358.namprd03.prod.outlook.com/
> > >
> > > It looks like it's the same workaround as ARM_ERRATA_742230, which the
> > > kernel does implement.
> > >
> > > It would be good to hear from the Nuvoton people, or an Arm person.
> > >
> >
> > I will happily update the patch to select ARM_ERRATA_742230 instead of
> > the dead non-existing ARM_ERRATA_794072.
> >
> > In contrast to the current patch that basically only cleans up "dead
> > config" and has no effective functional change, the new patch would
> > change the behaviour. I cannot test this patch (beyond some basic
> > compile test) on the hardware; so, we certainly need someone to have
> > that hardware, knows how to test it or confirm otherwise that we
> > should select the ARM_ERRATA_742230 fix for this hardware.

Note that ARM_ERRATA_742230 is applied in code up-to CORTEX-A9 r2p2
but while ARM_ERRATA_794072 exist also in CORTEX-A9 r4p1
https://github.com/torvalds/linux/blob/322a3b843d7f475b857646ed8f95b40431d3ecd0/arch/arm/mm/proc-v7.S#L347
> >
> > The current patch should be subsumed by the new patch; the submission
> > of the new patch is deferred until that person shows up. Let's see.
> >
> > Lukas
>
>
>
> --
> Regards,
> Avi



--
Regards,
Avi

2021-11-02 12:17:02

by Lukas Bulwahn

[permalink] [raw]
Subject: Re: [PATCH 11/13] arm: npcm: drop selecting non-existing ARM_ERRATA_794072

On Tue, Nov 2, 2021 at 9:11 AM Arnd Bergmann <[email protected]> wrote:
>
> On Tue, Nov 2, 2021 at 8:31 AM Lukas Bulwahn <[email protected]> wrote:
> > On Fri, Oct 29, 2021 at 8:36 AM Joel Stanley <[email protected]> wrote:
> > > On Thu, 28 Oct 2021 at 14:57, Arnd Bergmann <[email protected]> wrote:
> > > > On Thu, Oct 28, 2021 at 4:19 PM Lukas Bulwahn <[email protected]> wrote:
> > > https://lore.kernel.org/all/6be32e0b5b454ed7b609317266a8e798@BLUPR03MB358.namprd03.prod.outlook.com/
> > >
> > > It looks like it's the same workaround as ARM_ERRATA_742230, which the
> > > kernel does implement.
> > >
> > > It would be good to hear from the Nuvoton people, or an Arm person.
> >
> > I will happily update the patch to select ARM_ERRATA_742230 instead of
> > the dead non-existing ARM_ERRATA_794072.
> >
> > In contrast to the current patch that basically only cleans up "dead
> > config" and has no effective functional change, the new patch would
> > change the behaviour. I cannot test this patch (beyond some basic
> > compile test) on the hardware; so, we certainly need someone to have
> > that hardware, knows how to test it or confirm otherwise that we
> > should select the ARM_ERRATA_742230 fix for this hardware.
> >
> > The current patch should be subsumed by the new patch; the submission
> > of the new patch is deferred until that person shows up. Let's see.
>
> I'd prefer to leave the broken Kconfig symbol in place as a reminder until it
> gets fixed properly then.
>

Agree, this patch here should not be integrated. I rather expect that
Avi or others at Nuvoton will provide a proper patch to act
appropriately for the ARM_ERRATA_794072, or after proper analysis can
determine that there is no change in the kernel required.

Lukas

2021-11-09 14:03:13

by Linus Walleij

[permalink] [raw]
Subject: Re: [PATCH 13/13] arm: pgtable: refer to intended CONFIG_ARM_LPAE in comment

On Thu, Oct 28, 2021 at 4:20 PM Lukas Bulwahn <[email protected]> wrote:

> Commit 5615f69bc209 ("ARM: 9016/2: Initialize the mapping of KASan shadow
> memory") adds some nested ifdef's in ./arch/arm/mm/pgd.c, and follows the
> good practice to annotate the endif's with a comment to indicate the
> corresponding ifdef condition.
>
> One comment annotation refers to CONFIG_LPAE, whereas the config is
> actually called CONFIG_ARM_LPAE. That imprecision in a comment is probably
> tolerable for all human readers.
>
> However, the script ./scripts/checkkconfigsymbols.py, which checks the
> kernel tree for references to non-existing Kconfig symbols, identifies and
> reports that the reference to CONFIG_LPAE is invalid.
>
> The script ./scripts/checkkconfigsymbols.py has been quite useful to
> identify a number of bugs with Kconfig symbols and deserves to be executed
> and checked regularly.
>
> So, repair the comment to reduce the reports from this script and simplify
> to use this script, as new issues are easier to spot when the list of
> reports is shorter.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Reviewed-by: Linus Walleij <[email protected]>

Please put this patch into Russell's patch tracker!

Yours,
Linus Walleij

2021-11-10 08:29:30

by Linus Walleij

[permalink] [raw]
Subject: Re: [PATCH 06/13] arm: ixp4xx: remove dead configs CPU_IXP43X and CPU_IXP46X

On Thu, Oct 28, 2021 at 4:20 PM Lukas Bulwahn <[email protected]> wrote:

> Commit 73d04ca5f4ac ("ARM: ixp4xx: Delete Intel reference design
> boardfiles") removes the definition of the configs MACH_IXDP465 and
> MACH_KIXRP435, but misses to remove the configs CPU_IXP43X and CPU_IXP46X
> that depend on those removed configs, and hence are dead now.
>
> Fortunately, ./scripts/checkkconfigsymbols.py warns:
>
> MACH_IXDP465
> Referencing files: arch/arm/mach-ixp4xx/Kconfig
>
> MACH_KIXRP435
> Referencing files: arch/arm/mach-ixp4xx/Kconfig
>
> Remove the dead configs CPU_IXP43X and CPU_IXP46X.
>
> A further quick grep for the name of those two symbols did not show any
> use of the two config symbols; so, there are no further clean-up activities
> beyond this config removal needed.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Patch applied to my IXP4xx tree!

Yours,
Linus Walleij

2021-11-10 08:32:23

by Linus Walleij

[permalink] [raw]
Subject: Re: [PATCH 10/13] arm: nomadik: drop selecting obsolete CLKSRC_NOMADIK_MTU_SCHED_CLOCK

On Thu, Oct 28, 2021 at 4:20 PM Lukas Bulwahn <[email protected]> wrote:

> Commit 85b6fcadcf66 ("clocksource/drivers/ux500: Drop Ux500 custom
> SCHED_CLOCK") removes a sched_clock workaround and its corresponding
> config CLKSRC_NOMADIK_MTU_SCHED_CLOCK. Since then, selecting
> CLKSRC_NOMADIK_MTU_SCHED_CLOCK in ./arch/arm/mach-nomadik/Kconfig has no
> effect and ./scripts/checkkconfigsymbols.py warns:
>
> CLKSRC_NOMADIK_MTU_SCHED_CLOCK
> Referencing files: arch/arm/mach-nomadik/Kconfig
>
> Simply drop selecting the obsolete CLKSRC_NOMADIK_MTU_SCHED_CLOCK.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Patch applied to my Nomadik tree!

Yours,
Linus Walleij

2021-11-21 08:44:08

by Shawn Guo

[permalink] [raw]
Subject: Re: [PATCH 07/13] arm: imx: remove dead left-over from i.MX{27,31,35} removal

On Thu, Oct 28, 2021 at 04:19:32PM +0200, Lukas Bulwahn wrote:
> The commits:
>
> commit 879c0e5e0ac7 ("ARM: imx: Remove i.MX27 board files")
> commit c93197b0041d ("ARM: imx: Remove i.MX31 board files")
> commit e1324ece2af4 ("ARM: imx: Remove i.MX35 board files")
>
> remove the config MACH_MX27_3DS, MACH_MX31_3DS and MACH_MX35_3DS.
> Commit a542fc18168c ("ARM: imx31: Remove remaining i.MX31 board code")
> further removes arch/arm/mach-imx/3ds_debugboard.{c,h}. So, only some
> dead left-over in Kconfig and Makefile remains.
>
> Remove this remaining left-over.
>
> This issue was identified with ./scripts/checkkconfigsymbols.py,
> which warns on references to the non-existing configs
> MACH_MX{27,31,35}_3DS in ./arch/arm/mach-imx/Kconfig.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Applied, thanks!

2021-11-21 08:47:59

by Shawn Guo

[permalink] [raw]
Subject: Re: [PATCH 08/13] arm: imx: rename DEBUG_IMX21_IMX27_UART to DEBUG_IMX27_UART

On Thu, Oct 28, 2021 at 04:19:33PM +0200, Lukas Bulwahn wrote:
> Since commit 4b563a066611 ("ARM: imx: Remove imx21 support"), the config
> DEBUG_IMX21_IMX27_UART is really only debug support for IMX27.
>
> So, rename this option to DEBUG_IMX27_UART and adjust dependencies in
> Kconfig and rename the definitions to IMX27 as further clean-up.
>
> This issue was discovered with ./scripts/checkkconfigsymbols.py, which
> reported that DEBUG_IMX21_IMX27_UART depends on the non-existing config
> SOC_IMX21.
>
> Signed-off-by: Lukas Bulwahn <[email protected]>

Applied, thanks!