[PATCH 0/5] tty: serial: sh-sci: Hide driver specific config questions

[PATCH 0/5] tty: serial: sh-sci: Hide driver specific config questions

[Geert Uytterhoeven]

	Hi Greg, Simon, Magnus, Sato-san, Rich,

The Renesas (H)SCI(F) driver has several driver-specific config options.
This may confuse users, and leads to inefficient testing and development.
E.g. recently a regression was introduced in SCIF DMA support, and
someone didn't see early con output, both due to the relevant options
being accidentally disabled.

This patch series is an attempt to stop bothering (H)SCI(F) users with
several config questions, while providing sensible defaults for these
options, based on the target platform.  The defaults can still be
overridden if CONFIG_EXPERT is enabled.

The last patch updates shmobile_defconfig to take into account the
defaults.  It is marked RFC, as it depends on the first 4 patches to be
effective, and to avoid feature regressions.

Thanks for your comments!

Geert Uytterhoeven (5):
  tty: serial: sh-sci: Hide number of ports config question
  tty: serial: sh-sci: Hide serial console config question
  tty: serial: sh-sci: Hide earlycon config question
  tty: serial: sh-sci: Hide DMA config question
  [RFC] ARM: shmobile: defconfig: Disable CONFIG_EMBEDDED

 arch/arm/configs/shmobile_defconfig |  6 +-----
 drivers/tty/serial/Kconfig          | 15 ++++++++++-----
 2 files changed, 11 insertions(+), 10 deletions(-)

-- 
2.7.4

Gr{oetje,eeting}s,

						Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

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

[PATCH 1/5] tty: serial: sh-sci: Hide number of ports config question

[Geert Uytterhoeven]

Auto-configure the maximum number of serial ports based on how many can
be present on the architecture:
  - 3 on H8/300,
  - 10 on SuperH,
  - 18 on Reneas ARM.

The default can still be overridden if CONFIG_EXPERT is enabled.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
 drivers/tty/serial/Kconfig | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)

diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
index b788fee54249deff..bd96046d7f94bac0 100644
--- a/drivers/tty/serial/Kconfig
+++ b/drivers/tty/serial/Kconfig
@@ -761,9 +761,11 @@ config SERIAL_SH_SCI
 	select SERIAL_MCTRL_GPIO if GPIOLIB
 
 config SERIAL_SH_SCI_NR_UARTS
-	int "Maximum number of SCI(F) serial ports"
+	int "Maximum number of SCI(F) serial ports" if EXPERT
 	depends on SERIAL_SH_SCI
-	default "2"
+	default "3" if H8300
+	default "10" if SUPERH
+	default "18" if ARCH_RENESAS
 
 config SERIAL_SH_SCI_CONSOLE
 	bool "Support for console on SuperH SCI(F)"
-- 
2.7.4

[PATCH 2/5] tty: serial: sh-sci: Hide serial console config question

[Geert Uytterhoeven]

Most users will want to use a serial console.

Hence make that the default, unless CONFIG_EXPERT is enabled.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
 drivers/tty/serial/Kconfig | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
index bd96046d7f94bac0..0c75562d620feb82 100644
--- a/drivers/tty/serial/Kconfig
+++ b/drivers/tty/serial/Kconfig
@@ -768,9 +768,10 @@ config SERIAL_SH_SCI_NR_UARTS
 	default "18" if ARCH_RENESAS
 
 config SERIAL_SH_SCI_CONSOLE
-	bool "Support for console on SuperH SCI(F)"
+	bool "Support for console on SuperH SCI(F)" if EXPERT
 	depends on SERIAL_SH_SCI=y
 	select SERIAL_CORE_CONSOLE
+	default y
 
 config SERIAL_SH_SCI_EARLYCON
 	bool "Support for early console on SuperH SCI(F)"
-- 
2.7.4

[PATCH 3/5] tty: serial: sh-sci: Hide earlycon config question

[Geert Uytterhoeven]

Renesas H8/300 and ARM platforms use DT and support earlycon, so most
users want earlycon support to be enabled.

On SuperH platforms, earlycon is not yet supported.

Hence follow the above rationale to configure the default, unless
CONFIG_EXPERT is enabled.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
 drivers/tty/serial/Kconfig | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
index 0c75562d620feb82..952a2c6a9da08fdd 100644
--- a/drivers/tty/serial/Kconfig
+++ b/drivers/tty/serial/Kconfig
@@ -774,10 +774,11 @@ config SERIAL_SH_SCI_CONSOLE
 	default y
 
 config SERIAL_SH_SCI_EARLYCON
-	bool "Support for early console on SuperH SCI(F)"
+	bool "Support for early console on SuperH SCI(F)" if EXPERT
 	depends on SERIAL_SH_SCI=y
 	select SERIAL_CORE_CONSOLE
 	select SERIAL_EARLYCON
+	default ARCH_RENESAS || H8300
 
 config SERIAL_SH_SCI_DMA
 	bool "DMA support"
-- 
2.7.4

Re: [PATCH 3/5] tty: serial: sh-sci: Hide earlycon config question

[Rich Felker]

On Thu, Nov 30, 2017 at 02:12:00PM +0100, Geert Uytterhoeven wrote:
> Renesas H8/300 and ARM platforms use DT and support earlycon, so most
> users want earlycon support to be enabled.
> 
> On SuperH platforms, earlycon is not yet supported.
> 
> Hence follow the above rationale to configure the default, unless
> CONFIG_EXPERT is enabled.
> 
> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
> ---
>  drivers/tty/serial/Kconfig | 3 ++-
>  1 file changed, 2 insertions(+), 1 deletion(-)
> 
> diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
> index 0c75562d620feb82..952a2c6a9da08fdd 100644
> --- a/drivers/tty/serial/Kconfig
> +++ b/drivers/tty/serial/Kconfig
> @@ -774,10 +774,11 @@ config SERIAL_SH_SCI_CONSOLE
>  	default y
>  
>  config SERIAL_SH_SCI_EARLYCON
> -	bool "Support for early console on SuperH SCI(F)"
> +	bool "Support for early console on SuperH SCI(F)" if EXPERT
>  	depends on SERIAL_SH_SCI=y
>  	select SERIAL_CORE_CONSOLE
>  	select SERIAL_EARLYCON
> +	default ARCH_RENESAS || H8300
>  
>  config SERIAL_SH_SCI_DMA
>  	bool "DMA support"
> -- 

Can you clarify what the claim that SuperH does not support earlycon
is based on? My understanding is that users were successfully using
this option on Renesas SH systems, and I'm using it on J2 with the
uartlite earlycon support which I added in 7cdcc29e49. I think if you
want to omit the question it should always default to enabled.

Rich

Re: [PATCH 3/5] tty: serial: sh-sci: Hide earlycon config question

[Geert Uytterhoeven]

Hi Rich,

On Sun, Apr 15, 2018 at 2:34 AM, Rich Felker <dalias@libc.org> wrote:
> On Thu, Nov 30, 2017 at 02:12:00PM +0100, Geert Uytterhoeven wrote:
>> Renesas H8/300 and ARM platforms use DT and support earlycon, so most
>> users want earlycon support to be enabled.
>>
>> On SuperH platforms, earlycon is not yet supported.
>>
>> Hence follow the above rationale to configure the default, unless
>> CONFIG_EXPERT is enabled.
>>
>> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
>> ---
>>  drivers/tty/serial/Kconfig | 3 ++-
>>  1 file changed, 2 insertions(+), 1 deletion(-)
>>
>> diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
>> index 0c75562d620feb82..952a2c6a9da08fdd 100644
>> --- a/drivers/tty/serial/Kconfig
>> +++ b/drivers/tty/serial/Kconfig
>> @@ -774,10 +774,11 @@ config SERIAL_SH_SCI_CONSOLE
>>       default y
>>
>>  config SERIAL_SH_SCI_EARLYCON
>> -     bool "Support for early console on SuperH SCI(F)"
>> +     bool "Support for early console on SuperH SCI(F)" if EXPERT
>>       depends on SERIAL_SH_SCI=y
>>       select SERIAL_CORE_CONSOLE
>>       select SERIAL_EARLYCON
>> +     default ARCH_RENESAS || H8300
>>
>>  config SERIAL_SH_SCI_DMA
>>       bool "DMA support"
>> --
>
> Can you clarify what the claim that SuperH does not support earlycon
> is based on? My understanding is that users were successfully using
> this option on Renesas SH systems, and I'm using it on J2 with the
> uartlite earlycon support which I added in 7cdcc29e49. I think if you
> want to omit the question it should always default to enabled.

This is a patch for a Kconfig option for the Renesas sh-sci driver, which
supports the SCI, SCIF, SCIFA, SCIFB, and HSCIF uarts found on various
Renesas SoCs.

Earlycon is used with DT only. While you are using earlycon on J2, you do
use it with a different uart (uartlite). Currently there's no upstream support
for using DT on Renesas SuperH SoCs. If this changes, the default for
SERIAL_SH_SCI_EARLYCON has to be changed.

So none of my patch applies to the current state of SuperH Linux support.

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

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

Re: [PATCH 3/5] tty: serial: sh-sci: Hide earlycon config question

[Rich Felker]

On Sun, Apr 15, 2018 at 08:58:42PM +0200, Geert Uytterhoeven wrote:
> Hi Rich,
> 
> On Sun, Apr 15, 2018 at 2:34 AM, Rich Felker <dalias@libc.org> wrote:
> > On Thu, Nov 30, 2017 at 02:12:00PM +0100, Geert Uytterhoeven wrote:
> >> Renesas H8/300 and ARM platforms use DT and support earlycon, so most
> >> users want earlycon support to be enabled.
> >>
> >> On SuperH platforms, earlycon is not yet supported.
> >>
> >> Hence follow the above rationale to configure the default, unless
> >> CONFIG_EXPERT is enabled.
> >>
> >> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
> >> ---
> >>  drivers/tty/serial/Kconfig | 3 ++-
> >>  1 file changed, 2 insertions(+), 1 deletion(-)
> >>
> >> diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
> >> index 0c75562d620feb82..952a2c6a9da08fdd 100644
> >> --- a/drivers/tty/serial/Kconfig
> >> +++ b/drivers/tty/serial/Kconfig
> >> @@ -774,10 +774,11 @@ config SERIAL_SH_SCI_CONSOLE
> >>       default y
> >>
> >>  config SERIAL_SH_SCI_EARLYCON
> >> -     bool "Support for early console on SuperH SCI(F)"
> >> +     bool "Support for early console on SuperH SCI(F)" if EXPERT
> >>       depends on SERIAL_SH_SCI=y
> >>       select SERIAL_CORE_CONSOLE
> >>       select SERIAL_EARLYCON
> >> +     default ARCH_RENESAS || H8300
> >>
> >>  config SERIAL_SH_SCI_DMA
> >>       bool "DMA support"
> >> --
> >
> > Can you clarify what the claim that SuperH does not support earlycon
> > is based on? My understanding is that users were successfully using
> > this option on Renesas SH systems, and I'm using it on J2 with the
> > uartlite earlycon support which I added in 7cdcc29e49. I think if you
> > want to omit the question it should always default to enabled.
> 
> This is a patch for a Kconfig option for the Renesas sh-sci driver, which
> supports the SCI, SCIF, SCIFA, SCIFB, and HSCIF uarts found on various
> Renesas SoCs.
> 
> Earlycon is used with DT only. While you are using earlycon on J2, you do
> use it with a different uart (uartlite). Currently there's no upstream support
> for using DT on Renesas SuperH SoCs. If this changes, the default for
> SERIAL_SH_SCI_EARLYCON has to be changed.
> 
> So none of my patch applies to the current state of SuperH Linux support.

OK, I was under the impression (from users) that it worked on Renesas
SH devices without DT. If it really doesn't then it doesn't matter
until DT support for them is added. I've got some hardware to
experiment with now so I'll see what can be done.

Rich

Re: [PATCH 3/5] tty: serial: sh-sci: Hide earlycon config question

[Rob Herring]

On Sun, Apr 15, 2018 at 2:09 PM, Rich Felker <dalias@libc.org> wrote:
> On Sun, Apr 15, 2018 at 08:58:42PM +0200, Geert Uytterhoeven wrote:
>> Hi Rich,
>>
>> On Sun, Apr 15, 2018 at 2:34 AM, Rich Felker <dalias@libc.org> wrote:
>> > On Thu, Nov 30, 2017 at 02:12:00PM +0100, Geert Uytterhoeven wrote:
>> >> Renesas H8/300 and ARM platforms use DT and support earlycon, so most
>> >> users want earlycon support to be enabled.
>> >>
>> >> On SuperH platforms, earlycon is not yet supported.
>> >>
>> >> Hence follow the above rationale to configure the default, unless
>> >> CONFIG_EXPERT is enabled.
>> >>
>> >> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
>> >> ---
>> >>  drivers/tty/serial/Kconfig | 3 ++-
>> >>  1 file changed, 2 insertions(+), 1 deletion(-)
>> >>
>> >> diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
>> >> index 0c75562d620feb82..952a2c6a9da08fdd 100644
>> >> --- a/drivers/tty/serial/Kconfig
>> >> +++ b/drivers/tty/serial/Kconfig
>> >> @@ -774,10 +774,11 @@ config SERIAL_SH_SCI_CONSOLE
>> >>       default y
>> >>
>> >>  config SERIAL_SH_SCI_EARLYCON
>> >> -     bool "Support for early console on SuperH SCI(F)"
>> >> +     bool "Support for early console on SuperH SCI(F)" if EXPERT
>> >>       depends on SERIAL_SH_SCI=y
>> >>       select SERIAL_CORE_CONSOLE
>> >>       select SERIAL_EARLYCON
>> >> +     default ARCH_RENESAS || H8300
>> >>
>> >>  config SERIAL_SH_SCI_DMA
>> >>       bool "DMA support"
>> >> --
>> >
>> > Can you clarify what the claim that SuperH does not support earlycon
>> > is based on? My understanding is that users were successfully using
>> > this option on Renesas SH systems, and I'm using it on J2 with the
>> > uartlite earlycon support which I added in 7cdcc29e49. I think if you
>> > want to omit the question it should always default to enabled.
>>
>> This is a patch for a Kconfig option for the Renesas sh-sci driver, which
>> supports the SCI, SCIF, SCIFA, SCIFB, and HSCIF uarts found on various
>> Renesas SoCs.
>>
>> Earlycon is used with DT only. While you are using earlycon on J2, you do
>> use it with a different uart (uartlite). Currently there's no upstream support
>> for using DT on Renesas SuperH SoCs. If this changes, the default for
>> SERIAL_SH_SCI_EARLYCON has to be changed.
>>
>> So none of my patch applies to the current state of SuperH Linux support.
>
> OK, I was under the impression (from users) that it worked on Renesas
> SH devices without DT. If it really doesn't then it doesn't matter
> until DT support for them is added. I've got some hardware to
> experiment with now so I'll see what can be done.

Yes, it works without DT (but maybe that is UART specific). It was
originally an x86 8250 thing.

The main thing you need is either fixmap support or ioremap has to
work before paging_init when early_params are processed.

Rob

Re: [PATCH 3/5] tty: serial: sh-sci: Hide earlycon config question

[Geert Uytterhoeven]

Hi Rob,

On Wed, May 2, 2018 at 4:32 AM, Rob Herring <robherring2@gmail.com> wrote:
> On Sun, Apr 15, 2018 at 2:09 PM, Rich Felker <dalias@libc.org> wrote:
>> On Sun, Apr 15, 2018 at 08:58:42PM +0200, Geert Uytterhoeven wrote:
>>> On Sun, Apr 15, 2018 at 2:34 AM, Rich Felker <dalias@libc.org> wrote:
>>> > On Thu, Nov 30, 2017 at 02:12:00PM +0100, Geert Uytterhoeven wrote:
>>> >> Renesas H8/300 and ARM platforms use DT and support earlycon, so most
>>> >> users want earlycon support to be enabled.
>>> >>
>>> >> On SuperH platforms, earlycon is not yet supported.
>>> >>
>>> >> Hence follow the above rationale to configure the default, unless
>>> >> CONFIG_EXPERT is enabled.
>>> >>
>>> >> Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
>>> >> ---
>>> >>  drivers/tty/serial/Kconfig | 3 ++-
>>> >>  1 file changed, 2 insertions(+), 1 deletion(-)
>>> >>
>>> >> diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
>>> >> index 0c75562d620feb82..952a2c6a9da08fdd 100644
>>> >> --- a/drivers/tty/serial/Kconfig
>>> >> +++ b/drivers/tty/serial/Kconfig
>>> >> @@ -774,10 +774,11 @@ config SERIAL_SH_SCI_CONSOLE
>>> >>       default y
>>> >>
>>> >>  config SERIAL_SH_SCI_EARLYCON
>>> >> -     bool "Support for early console on SuperH SCI(F)"
>>> >> +     bool "Support for early console on SuperH SCI(F)" if EXPERT
>>> >>       depends on SERIAL_SH_SCI=y
>>> >>       select SERIAL_CORE_CONSOLE
>>> >>       select SERIAL_EARLYCON
>>> >> +     default ARCH_RENESAS || H8300
>>> >>
>>> >>  config SERIAL_SH_SCI_DMA
>>> >>       bool "DMA support"
>>> >> --
>>> >
>>> > Can you clarify what the claim that SuperH does not support earlycon
>>> > is based on? My understanding is that users were successfully using
>>> > this option on Renesas SH systems, and I'm using it on J2 with the
>>> > uartlite earlycon support which I added in 7cdcc29e49. I think if you
>>> > want to omit the question it should always default to enabled.
>>>
>>> This is a patch for a Kconfig option for the Renesas sh-sci driver, which
>>> supports the SCI, SCIF, SCIFA, SCIFB, and HSCIF uarts found on various
>>> Renesas SoCs.
>>>
>>> Earlycon is used with DT only. While you are using earlycon on J2, you do
>>> use it with a different uart (uartlite). Currently there's no upstream support
>>> for using DT on Renesas SuperH SoCs. If this changes, the default for
>>> SERIAL_SH_SCI_EARLYCON has to be changed.
>>>
>>> So none of my patch applies to the current state of SuperH Linux support.
>>
>> OK, I was under the impression (from users) that it worked on Renesas
>> SH devices without DT. If it really doesn't then it doesn't matter
>> until DT support for them is added. I've got some hardware to
>> experiment with now so I'll see what can be done.
>
> Yes, it works without DT (but maybe that is UART specific). It was
> originally an x86 8250 thing.
>
> The main thing you need is either fixmap support or ioremap has to
> work before paging_init when early_params are processed.

Without DT and its chosen/stdout-path, you need a way to specify which
UART to use. That's typically done through a UART-specific earlyconextension.

The sh-sci.c driver uses:

    OF_EARLYCON_DECLARE(scif, "renesas,scif", scif_early_console_setup);
    (+ for sci, scifa, scifb, hscif)

So it may actually work with earlycon=scif,0x<addr>?

Anyone who can try on a non-DT SH system?

if it works, the default should be changed to

    default ARCH_RENESAS || H8300 || SUPERH

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

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

[PATCH 4/5] tty: serial: sh-sci: Hide DMA config question

[Geert Uytterhoeven]

On most Renesas ARM platforms, the SCIF serial ports can be used with
DMA, so most users will want DMA support to be enabled.

On SuperH platforms, SCI(F) serial ports cannot be used with DMA yet
(see also commit 219fb0c1436e4893 ("serial: sh-sci: Remove the platform
data dma slave rx/tx channel IDs")), so users will want it disabled to
reduce kernel size.

Hence follow the above rationale to configure the default, unless
CONFIG_EXPERT is enabled.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
 drivers/tty/serial/Kconfig | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/tty/serial/Kconfig b/drivers/tty/serial/Kconfig
index 952a2c6a9da08fdd..4e6dfb0a762b5807 100644
--- a/drivers/tty/serial/Kconfig
+++ b/drivers/tty/serial/Kconfig
@@ -781,8 +781,9 @@ config SERIAL_SH_SCI_EARLYCON
 	default ARCH_RENESAS || H8300
 
 config SERIAL_SH_SCI_DMA
-	bool "DMA support"
+	bool "DMA support" if EXPERT
 	depends on SERIAL_SH_SCI && DMA_ENGINE
+	default ARCH_RENESAS
 
 config SERIAL_PNX8XXX
 	bool "Enable PNX8XXX SoCs' UART Support"
-- 
2.7.4

[PATCH/RFC 5/5] ARM: shmobile: defconfig: Disable CONFIG_EMBEDDED

[Geert Uytterhoeven]

CONFIG_EXPERT exposes too many config options that do not matter for
development. However, CONFIG_EMBEDDED selects CONFIG_EXPERT, so it
cannot be disabled.

Hence disable CONFIG_EMBEDDED, and compensate for the loss of
CONFIG_DEBUG_KERNEL by enabling the latter.

Actual impact:
  - CONFIG_NAMESPACES=y,
  - CONFIG_SYSCTL_SYSCALL=n,
  - CONFIG_SERIAL_SH_SCI_NR_UARTS changed from 20 to 18,
  - CONFIG_SERIAL_SH_SCI_EARLYCON=y,
  - Some HID support became enabled,
  - CONFIG_DEBUG_MEMORY_INIT=y,

Refresh the result.

Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
---
 arch/arm/configs/shmobile_defconfig | 6 +-----
 1 file changed, 1 insertion(+), 5 deletions(-)

diff --git a/arch/arm/configs/shmobile_defconfig b/arch/arm/configs/shmobile_defconfig
index 7b4fc01431483f21..41eccc1b6cef7251 100644
--- a/arch/arm/configs/shmobile_defconfig
+++ b/arch/arm/configs/shmobile_defconfig
@@ -5,8 +5,6 @@ CONFIG_IKCONFIG_PROC=y
 CONFIG_CGROUPS=y
 CONFIG_BLK_DEV_INITRD=y
 CONFIG_CC_OPTIMIZE_FOR_SIZE=y
-CONFIG_SYSCTL_SYSCALL=y
-CONFIG_EMBEDDED=y
 CONFIG_PERF_EVENTS=y
 CONFIG_SLAB=y
 CONFIG_ARCH_RENESAS=y
@@ -34,7 +32,6 @@ CONFIG_SMP=y
 CONFIG_SCHED_MC=y
 CONFIG_HAVE_ARM_ARCH_TIMER=y
 CONFIG_NR_CPUS=8
-CONFIG_AEABI=y
 CONFIG_HIGHMEM=y
 CONFIG_CMA=y
 CONFIG_ZBOOT_ROM_TEXT=0x0
@@ -106,7 +103,6 @@ CONFIG_SERIAL_8250_EM=y
 CONFIG_SERIAL_SH_SCI=y
 CONFIG_SERIAL_SH_SCI_NR_UARTS 
 CONFIG_SERIAL_SH_SCI_CONSOLE=y
-CONFIG_SERIAL_SH_SCI_DMA=y
 CONFIG_I2C_CHARDEV=y
 CONFIG_I2C_MUX=y
 CONFIG_I2C_DEMUX_PINCTRL=y
@@ -166,7 +162,6 @@ CONFIG_FB_SH_MOBILE_MERAM=y
 # CONFIG_BACKLIGHT_GENERIC is not set
 CONFIG_BACKLIGHT_PWM=y
 CONFIG_BACKLIGHT_AS3711=y
-CONFIG_FRAMEBUFFER_CONSOLE=y
 CONFIG_SOUND=y
 CONFIG_SND=y
 CONFIG_SND_SOC=y
@@ -225,4 +220,5 @@ CONFIG_NLS_ISO8859_1=y
 CONFIG_PRINTK_TIME=y
 # CONFIG_ENABLE_WARN_DEPRECATED is not set
 # CONFIG_ENABLE_MUST_CHECK is not set
+CONFIG_DEBUG_KERNEL=y
 # CONFIG_ARM_UNWIND is not set
-- 
2.7.4

Re: [PATCH/RFC 5/5] ARM: shmobile: defconfig: Disable CONFIG_EMBEDDED

[Simon Horman]

On Thu, Nov 30, 2017 at 02:12:02PM +0100, Geert Uytterhoeven wrote:
> CONFIG_EXPERT exposes too many config options that do not matter for
> development. However, CONFIG_EMBEDDED selects CONFIG_EXPERT, so it
> cannot be disabled.
> 
> Hence disable CONFIG_EMBEDDED, and compensate for the loss of
> CONFIG_DEBUG_KERNEL by enabling the latter.
> 
> Actual impact:
>   - CONFIG_NAMESPACES=y,
>   - CONFIG_SYSCTL_SYSCALL=n,
>   - CONFIG_SERIAL_SH_SCI_NR_UARTS changed from 20 to 18,
>   - CONFIG_SERIAL_SH_SCI_EARLYCON=y,
>   - Some HID support became enabled,
>   - CONFIG_DEBUG_MEMORY_INIT=y,
> 
> Refresh the result.

I'm sorry but I don't understand the motivation for this change.
I understand that it might make the image smaller or build times faster.
But it seems to be a risky change for those benefits.
What are the side-effects of the changes you describe above?

Re: [PATCH/RFC 5/5] ARM: shmobile: defconfig: Disable CONFIG_EMBEDDED

[Geert Uytterhoeven]

Hi Simon,

On Fri, Dec 1, 2017 at 9:33 AM, Simon Horman <horms@verge.net.au> wrote:
> On Thu, Nov 30, 2017 at 02:12:02PM +0100, Geert Uytterhoeven wrote:
>> CONFIG_EXPERT exposes too many config options that do not matter for
>> development. However, CONFIG_EMBEDDED selects CONFIG_EXPERT, so it
>> cannot be disabled.
>>
>> Hence disable CONFIG_EMBEDDED, and compensate for the loss of
>> CONFIG_DEBUG_KERNEL by enabling the latter.
>>
>> Actual impact:
>>   - CONFIG_NAMESPACES=y,
>>   - CONFIG_SYSCTL_SYSCALL=n,
>>   - CONFIG_SERIAL_SH_SCI_NR_UARTS changed from 20 to 18,
>>   - CONFIG_SERIAL_SH_SCI_EARLYCON=y,
>>   - Some HID support became enabled,
>>   - CONFIG_DEBUG_MEMORY_INIT=y,
>>
>> Refresh the result.
>
> I'm sorry but I don't understand the motivation for this change.
> I understand that it might make the image smaller or build times faster.
> But it seems to be a risky change for those benefits.

My main motivation is to start using the default values for the various SH_SCI
options from the first 4 patches, but that requires disabling CONFIG_EXPERT.
Sorry for not making that clearer.

> What are the side-effects of the changes you describe above?

Enabling CONFIG_NAMESPACES shouldn't hurt (isn't that also needed for
systemd anyway?), just like the HID devices, or CONFIG_DEBUG_MEMORY_INIT
(it does a few more checks, and defaults to yes).

Disabling CONFIG_SYSCTL_SYSCALL also shouldn't hurt (it defaults to no),
except for very old userland not yet using /proc/sys, which might have been
broken along the way by changed binary paths anyway.

Thanks!

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

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

Re: [PATCH/RFC 5/5] ARM: shmobile: defconfig: Disable CONFIG_EMBEDDED

[Simon Horman]

On Mon, Dec 04, 2017 at 10:46:22AM +0100, Geert Uytterhoeven wrote:
> Hi Simon,
> 
> On Fri, Dec 1, 2017 at 9:33 AM, Simon Horman <horms@verge.net.au> wrote:
> > On Thu, Nov 30, 2017 at 02:12:02PM +0100, Geert Uytterhoeven wrote:
> >> CONFIG_EXPERT exposes too many config options that do not matter for
> >> development. However, CONFIG_EMBEDDED selects CONFIG_EXPERT, so it
> >> cannot be disabled.
> >>
> >> Hence disable CONFIG_EMBEDDED, and compensate for the loss of
> >> CONFIG_DEBUG_KERNEL by enabling the latter.
> >>
> >> Actual impact:
> >>   - CONFIG_NAMESPACES=y,
> >>   - CONFIG_SYSCTL_SYSCALL=n,
> >>   - CONFIG_SERIAL_SH_SCI_NR_UARTS changed from 20 to 18,
> >>   - CONFIG_SERIAL_SH_SCI_EARLYCON=y,
> >>   - Some HID support became enabled,
> >>   - CONFIG_DEBUG_MEMORY_INIT=y,
> >>
> >> Refresh the result.
> >
> > I'm sorry but I don't understand the motivation for this change.
> > I understand that it might make the image smaller or build times faster.
> > But it seems to be a risky change for those benefits.
> 
> My main motivation is to start using the default values for the various SH_SCI
> options from the first 4 patches, but that requires disabling CONFIG_EXPERT.
> Sorry for not making that clearer.

Thanks for explaining, that helps me significantly.

> > What are the side-effects of the changes you describe above?
> 
> Enabling CONFIG_NAMESPACES shouldn't hurt (isn't that also needed for
> systemd anyway?), just like the HID devices, or CONFIG_DEBUG_MEMORY_INIT
> (it does a few more checks, and defaults to yes).
> 
> Disabling CONFIG_SYSCTL_SYSCALL also shouldn't hurt (it defaults to no),
> except for very old userland not yet using /proc/sys, which might have been
> broken along the way by changed binary paths anyway.

Understood. I (no longer) have any particular objections to this,
though I would appreciate an enhanced changelog.

Re: [PATCH/RFC 5/5] ARM: shmobile: defconfig: Disable CONFIG_EMBEDDED

[Rob Landley]

On 12/05/2017 02:28 AM, Simon Horman wrote:
> On Mon, Dec 04, 2017 at 10:46:22AM +0100, Geert Uytterhoeven wrote:
>> Hi Simon,
>>
>> On Fri, Dec 1, 2017 at 9:33 AM, Simon Horman <horms@verge.net.au> wrote:
>>> On Thu, Nov 30, 2017 at 02:12:02PM +0100, Geert Uytterhoeven wrote:
>>>> CONFIG_EXPERT exposes too many config options that do not matter for
>>>> development. However, CONFIG_EMBEDDED selects CONFIG_EXPERT, so it
>>>> cannot be disabled.
>>>>
>>>> Hence disable CONFIG_EMBEDDED, and compensate for the loss of
>>>> CONFIG_DEBUG_KERNEL by enabling the latter.

Didn't I already submit a patch to fix this in 2015?

https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?idV0b8c0ed45a

If it came back, it's a regression.

Rob

Re: [PATCH/RFC 5/5] ARM: shmobile: defconfig: Disable CONFIG_EMBEDDED

[Geert Uytterhoeven]

Hi Rob,

On Sat, Dec 9, 2017 at 3:18 AM, Rob Landley <rob@landley.net> wrote:
> On 12/05/2017 02:28 AM, Simon Horman wrote:
>> On Mon, Dec 04, 2017 at 10:46:22AM +0100, Geert Uytterhoeven wrote:
>>> On Fri, Dec 1, 2017 at 9:33 AM, Simon Horman <horms@verge.net.au> wrote:
>>>> On Thu, Nov 30, 2017 at 02:12:02PM +0100, Geert Uytterhoeven wrote:
>>>>> CONFIG_EXPERT exposes too many config options that do not matter for
>>>>> development. However, CONFIG_EMBEDDED selects CONFIG_EXPERT, so it
>>>>> cannot be disabled.
>>>>>
>>>>> Hence disable CONFIG_EMBEDDED, and compensate for the loss of
>>>>> CONFIG_DEBUG_KERNEL by enabling the latter.
>
> Didn't I already submit a patch to fix this in 2015?
>
> https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?idV0b8c0ed45a
>
> If it came back, it's a regression.

Your patch was for arch/sh/Kconfig, not for (arm32) shmobile_defconfig.

Gr{oetje,eeting}s,

                        Geert

--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

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