[kconfig] update: results of some syntactical checks

[kconfig] update: results of some syntactical checks

[Martin Walch]

This is an update to the syntactic results that I sent, back in July.

With kernel 3.12 nearing completion, I would like to point to new sections
in Kconfig files with potential problems:

drivers/media/common/siano/Kconfig:21-26
> config SMS_SIANO_DEBUGFS
>	bool "Enable debugfs for smsdvb"
>	depends on SMS_SIANO_MDTV
>	depends on DEBUG_FS
>	depends on SMS_USB_DRV
>	depends on CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV

The last line adds the dependency CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV.
This expression does not look sound as those two symbols are not declared
anywhere. So, the two strings CONFIG_SMS_USB_DRV and CONFIG_SMS_SDIO_DRV
are compared, yielding always 'n'. As a result, SMS_SIANO_DEBUGFS will never
be enabled.

Probably, it was meant to say something like
>	depends on SMS_USB_DRV = SMS_SDIO_DRV

Two other config sections that probably behave differently than expected:

drivers/staging/rtl8188eu/Kconfig: 13-15
> config 88EU_AP_MODE
>	bool "Realtek RTL8188EU AP mode"
>	default Y

drivers/staging/rtl8188eu/Kconfig: 21-23
> config 88EU_P2P
>	bool "Realtek RTL8188EU Peer-to-peer mode"
>	default Y

The capital Y is different from the lowercase y. While y is an actually
hard coded constant symbol, Y is undeclared and evaluates to n.
The default values are probably only for convenience, so 88EU_AP_MODE and
88EU_P2P are activated together with 8188EU. They still can be turned off.
Anyway, it should probably say "default y" in both cases.


Here the updated full two lists from July:

Actually declared symbols with names of constants: 2
  8260 at
    arch/powerpc/platforms/82xx/Kconfig:55
  8272 at
    arch/powerpc/platforms/82xx/Kconfig:64

Symbols used, but not declared: 34
  Symbol: ARCH_ARM at
    sound/soc/omap/Kconfig:28
    sound/soc/omap/Kconfig:2
  Symbol: ARCH_EFM32 at
    drivers/spi/Kconfig:167
    drivers/tty/serial/Kconfig:1430
  Symbol: ARCH_HI3xxx at
    drivers/dma/Kconfig:313
    arch/arm/Kconfig.debug:184
    arch/arm/Kconfig.debug:192
  Symbol: ARCH_MOXART at
    drivers/net/ethernet/moxa/Kconfig:22
    drivers/net/ethernet/moxa/Kconfig:7
  Symbol: ARCH_MULTI_V4 at
    arch/arm/Kconfig:916
  Symbol: ATHEROS_AR231X at
    drivers/net/wireless/ath/ath5k/Kconfig:56
    drivers/net/wireless/ath/ath5k/Kconfig:63
    drivers/net/wireless/ath/ath5k/Kconfig:9
    drivers/net/wireless/ath/ath5k/Kconfig:8
    drivers/net/wireless/ath/ath5k/Kconfig:2
  Symbol: CONFIG_SMS_SDIO_DRV at
    drivers/media/common/siano/Kconfig:25
  Symbol: CONFIG_SMS_USB_DRV at
    drivers/media/common/siano/Kconfig:25
  Symbol: CPU_MMP3 at
    drivers/video/mmp/hw/Kconfig:4
    drivers/usb/phy/Kconfig:44
    drivers/video/mmp/Kconfig:2
  Symbol: CPU_PXA988 at
    drivers/video/mmp/hw/Kconfig:4
    drivers/video/mmp/Kconfig:2
  Symbol: CPU_SUBTYPE_SH7764 at
    arch/sh/drivers/dma/Kconfig:14
  Symbol: DEPRECATED at
    arch/mn10300/Kconfig.debug:34
  Symbol: EXYNOS_DEV_SYSMMU at
    drivers/iommu/Kconfig:180
  Symbol: GENERIC_HAS_IOMAP at
    arch/score/Kconfig:39
    arch/score/Kconfig:27
    arch/score/Kconfig:33
  Symbol: GENERIC_TIME at
    arch/arm/mach-bcm/Kconfig:11
  Symbol: GPIO_BCM at
    arch/arm/mach-bcm/Kconfig:12
  Symbol: HAVE_GENERIC_HARDIRQS at
    arch/score/Kconfig:5
  Symbol: LOCAL_TIMERS at
    arch/arm/mach-shmobile/Kconfig:6
    arch/arm/mach-omap2/Kconfig:59
    arch/arm/mach-rockchip/Kconfig:10
    arch/arm/mach-rockchip/Kconfig:7
  Symbol: M at
    drivers/usb/host/Kconfig:547
    drivers/usb/misc/Kconfig:130
  Symbol: MACH_NOKIA_RM696 at
    arch/arm/mach-omap2/Kconfig:321
  Symbol: MACH_OMAP_H4_OTG at
    drivers/usb/gadget/Kconfig:207
  Symbol: MACH_SMDKC210 at
    sound/soc/samsung/Kconfig:138
  Symbol: MACH_SMDKV310 at
    sound/soc/samsung/Kconfig:138
  Symbol: MN10300_PROC_MN2WS0038 at
    arch/mn10300/Kconfig:184
  Symbol: MPILIB_EXTRA at
    crypto/asymmetric_keys/Kconfig:24
  Symbol: MV64360 at
    arch/powerpc/Kconfig:419
  Symbol: N at
    drivers/staging/rtl8187se/Kconfig:8
    drivers/staging/bcm/Kconfig:3
    drivers/staging/usbip/Kconfig:31
    arch/cris/arch-v32/drivers/Kconfig:13
    drivers/staging/rtl8712/Kconfig:14
    drivers/usb/host/Kconfig:340
    drivers/staging/rtl8712/Kconfig:6
    drivers/staging/android/Kconfig:61
    drivers/staging/usbip/Kconfig:42
    drivers/staging/usbip/Kconfig:20
    arch/cris/arch-v32/drivers/Kconfig:118
    drivers/usb/host/Kconfig:317
    drivers/staging/android/Kconfig:4
    drivers/staging/frontier/Kconfig:3
    drivers/usb/host/Kconfig:329
    drivers/staging/rtl8192u/Kconfig:7
    arch/arc/Kconfig:356
    drivers/staging/media/go7007/Kconfig:48
    drivers/staging/media/go7007/Kconfig:15
    drivers/usb/core/Kconfig:12
    drivers/staging/rtl8192e/rtl8192e/Kconfig:7
    drivers/staging/usbip/Kconfig:3
    drivers/staging/media/go7007/Kconfig:26
    drivers/staging/rtl8188eu/Kconfig:5
  Symbol: OF_I2C at
    drivers/i2c/busses/Kconfig:377
  Symbol: OMAP_PM_SRF at
    drivers/staging/tidspbridge/Kconfig:19
  Symbol: PICOXCELL_PC3X3 at
    drivers/char/hw_random/Kconfig:242
  Symbol: PLATFORM_MICROBLAZE_AUTO at
    arch/microblaze/platform/Kconfig.platform:10
  Symbol: PLAT_SPEAR_SINGLE at
    arch/arm/mach-spear/Kconfig:84
    arch/arm/mach-spear/Kconfig:99
    arch/arm/mach-spear/Kconfig:6
    arch/arm/mach-spear/Kconfig:51
    arch/arm/mach-spear/Kconfig:19
  Symbol: RCAR_CLK_ADG at
    sound/soc/sh/Kconfig:40
  Symbol: Y at
    drivers/staging/rtl8188eu/Kconfig:14
    drivers/staging/rtl8188eu/Kconfig:22
-- 

Re: [kconfig] update: results of some syntactical checks

[Paul Bolle]

On Sun, 2013-10-20 at 00:03 +0200, Martin Walch wrote:
> drivers/media/common/siano/Kconfig:21-26
> > config SMS_SIANO_DEBUGFS
> >	bool "Enable debugfs for smsdvb"
> >	depends on SMS_SIANO_MDTV
> >	depends on DEBUG_FS
> >	depends on SMS_USB_DRV
> >	depends on CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV
> 
> The last line adds the dependency CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV.
> This expression does not look sound as those two symbols are not declared
> anywhere. So, the two strings CONFIG_SMS_USB_DRV and CONFIG_SMS_SDIO_DRV
> are compared, yielding always 'n'. As a result, SMS_SIANO_DEBUGFS will never
> be enabled.

Those are obvious typos. Still present in v3.12-rc7. Perhaps you'd like
to send the trivial patch to fix this?

> Probably, it was meant to say something like
> >	depends on SMS_USB_DRV = SMS_SDIO_DRV
> 
> Two other config sections that probably behave differently than expected:
> 
> drivers/staging/rtl8188eu/Kconfig: 13-15
> > config 88EU_AP_MODE
> >	bool "Realtek RTL8188EU AP mode"
> >	default Y
> 
> drivers/staging/rtl8188eu/Kconfig: 21-23
> > config 88EU_P2P
> >	bool "Realtek RTL8188EU Peer-to-peer mode"
> >	default Y
> 
> The capital Y is different from the lowercase y. While y is an actually
> hard coded constant symbol, Y is undeclared and evaluates to n.
> The default values are probably only for convenience, so 88EU_AP_MODE and
> 88EU_P2P are activated together with 8188EU. They still can be turned off.
> Anyway, it should probably say "default y" in both cases.

Ditto. 


Paul Bolle

Re: [kconfig] update: results of some syntactical checks

[Mauro Carvalho Chehab]

Em Sat, 02 Nov 2013 20:20:54 +0100
Paul Bolle <pebolle@tiscali.nl> escreveu:

> On Sun, 2013-10-20 at 00:03 +0200, Martin Walch wrote:
> > drivers/media/common/siano/Kconfig:21-26
> > > config SMS_SIANO_DEBUGFS
> > >	bool "Enable debugfs for smsdvb"
> > >	depends on SMS_SIANO_MDTV
> > >	depends on DEBUG_FS
> > >	depends on SMS_USB_DRV
> > >	depends on CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV
> > 
> > The last line adds the dependency CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV.
> > This expression does not look sound as those two symbols are not declared
> > anywhere. So, the two strings CONFIG_SMS_USB_DRV and CONFIG_SMS_SDIO_DRV
> > are compared, yielding always 'n'. As a result, SMS_SIANO_DEBUGFS will never
> > be enabled.
> 
> Those are obvious typos. Still present in v3.12-rc7. Perhaps you'd like
> to send the trivial patch to fix this?

Yes, it is a typo...

> 
> > Probably, it was meant to say something like
> > >	depends on SMS_USB_DRV = SMS_SDIO_DRV

But this is not the right thing to do. The Kconfig logic here is that it
should depends on !SMS_SDIO_DRV or SMS_USB_DRV = SMS_SDIO_DRV.

I remember I made a patch like that while testing some things with this
driver, but it seems that I forgot to push. I might have it somewhere on
my test machine.



-- 

Cheers,
Mauro

Re: [kconfig] update: results of some syntactical checks

[Paul Bolle]

On Sat, 2013-11-02 at 17:40 -0200, Mauro Carvalho Chehab wrote:
> Em Sat, 02 Nov 2013 20:20:54 +0100
> Paul Bolle <pebolle@tiscali.nl> escreveu:
> > Those are obvious typos. Still present in v3.12-rc7. Perhaps you'd like
> > to send the trivial patch to fix this?
> 
> Yes, it is a typo...
> 
> > > Probably, it was meant to say something like
> > > >	depends on SMS_USB_DRV = SMS_SDIO_DRV
> 
> But this is not the right thing to do. The Kconfig logic here is that it
> should depends on !SMS_SDIO_DRV or SMS_USB_DRV = SMS_SDIO_DRV.

For the record: my reason for suggesting to Martin to send patches is
that sending a (trivial) patch often turns out to be a quick way to get
issues like that resolved. Maintainers seem to react quite quickly on
patches. Perhaps even faster if a patch is not solving the issue at hand
entirely correct! Messages like Martin's, that basically state that
there are one or more tree-wide problems, seem to be overlooked easily.

> I remember I made a patch like that while testing some things with this
> driver, but it seems that I forgot to push. I might have it somewhere on
> my test machine.


Paul Bolle

Re: [kconfig] update: results of some syntactical checks

[Paul Bolle]

Mauro,

On Sat, 2013-11-02 at 17:40 -0200, Mauro Carvalho Chehab wrote:
> Em Sat, 02 Nov 2013 20:20:54 +0100
> Paul Bolle <pebolle@tiscali.nl> escreveu:
> > On Sun, 2013-10-20 at 00:03 +0200, Martin Walch wrote:
> > > drivers/media/common/siano/Kconfig:21-26
> > > > config SMS_SIANO_DEBUGFS
> > > >	bool "Enable debugfs for smsdvb"
> > > >	depends on SMS_SIANO_MDTV
> > > >	depends on DEBUG_FS
> > > >	depends on SMS_USB_DRV
> > > >	depends on CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV
> > > 
> > > The last line adds the dependency CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV.
> > > This expression does not look sound as those two symbols are not declared
> > > anywhere. So, the two strings CONFIG_SMS_USB_DRV and CONFIG_SMS_SDIO_DRV
> > > are compared, yielding always 'n'. As a result, SMS_SIANO_DEBUGFS will never
> > > be enabled.
>
> [...] The Kconfig logic here is that it
> should depends on !SMS_SDIO_DRV or SMS_USB_DRV = SMS_SDIO_DRV.
> 
> I remember I made a patch like that while testing some things with this
> driver, but it seems that I forgot to push. I might have it somewhere on
> my test machine.

This line is still present in v3.14-rc2. Did you ever find that patch on
your test machine?


Paul Bolle

[PATCH] [media] sms: Remove CONFIG_ prefix from Kconfig symbols

[Paul Bolle]

Remove the CONFIG_ prefix from two Kconfig symbols in a dependency for
SMS_SIANO_DEBUGFS. This prefix is invalid inside Kconfig files.

Note that the current (common sense) dependency on SMS_USB_DRV and
SMS_SDIO_DRV being equal ensures that SMS_SIANO_DEBUGFS will not
violate its constraints. These constraint are that:
- it should only be built if SMS_USB_DRV is set;
- it can't be builtin if USB support is modular.

So drop the dependency on SMS_USB_DRV, as it is unneeded.

Fixes: 6c84b214284e ("[media] sms: fix randconfig building error")
Reported-by: Martin Walch <walch.martin@web.de>
Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
---
This is not runtime tested, as I don't have the hardware.

A matrix of the three cases in which this symbol can be set, to aid
review and to see whether I actually understood the constraints:

USB  SMS_USB_DRV  SMS_SDIO_DRV  SMS_SIANO_MDTV  SMS_SIANO_DEBUGFS
m    m            m             m               y 
y    m            m             m               y 
y    y            y             y               y 

By the way, I found myself staring at the entries in this file for quite
some time. Perhaps things would have been easier to understand if
SMS_USB_DRV and SMS_SDIO_DRV both selected SMS_SIANO_MDTV. But I didn't
dare to test that idea.

 drivers/media/common/siano/Kconfig | 3 +--
 1 file changed, 1 insertion(+), 2 deletions(-)

diff --git a/drivers/media/common/siano/Kconfig b/drivers/media/common/siano/Kconfig
index f953d33ee151..4bfbd5f463d1 100644
--- a/drivers/media/common/siano/Kconfig
+++ b/drivers/media/common/siano/Kconfig
@@ -22,8 +22,7 @@ config SMS_SIANO_DEBUGFS
 	bool "Enable debugfs for smsdvb"
 	depends on SMS_SIANO_MDTV
 	depends on DEBUG_FS
-	depends on SMS_USB_DRV
-	depends on CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV
+	depends on SMS_USB_DRV = SMS_SDIO_DRV
 
 	---help---
 	  Choose Y to enable visualizing a dump of the frontend
-- 
1.9.0

Re: [PATCH] [media] sms: Remove CONFIG_ prefix from Kconfig symbols

[Paul Bolle]

On Wed, 2014-04-16 at 17:47 +0200, Paul Bolle wrote:
> Remove the CONFIG_ prefix from two Kconfig symbols in a dependency for
> SMS_SIANO_DEBUGFS. This prefix is invalid inside Kconfig files.
> 
> Note that the current (common sense) dependency on SMS_USB_DRV and
> SMS_SDIO_DRV being equal ensures that SMS_SIANO_DEBUGFS will not
> violate its constraints. These constraint are that:
> - it should only be built if SMS_USB_DRV is set;
> - it can't be builtin if USB support is modular.
> 
> So drop the dependency on SMS_USB_DRV, as it is unneeded.
> 
> Fixes: 6c84b214284e ("[media] sms: fix randconfig building error")
> Reported-by: Martin Walch <walch.martin@web.de>
> Signed-off-by: Paul Bolle <pebolle@tiscali.nl>
> ---
> This is not runtime tested, as I don't have the hardware.
> 
> A matrix of the three cases in which this symbol can be set, to aid
> review and to see whether I actually understood the constraints:
> 
> USB  SMS_USB_DRV  SMS_SDIO_DRV  SMS_SIANO_MDTV  SMS_SIANO_DEBUGFS
> m    m            m             m               y 
> y    m            m             m               y 
> y    y            y             y               y 
> 
> By the way, I found myself staring at the entries in this file for quite
> some time. Perhaps things would have been easier to understand if
> SMS_USB_DRV and SMS_SDIO_DRV both selected SMS_SIANO_MDTV. But I didn't
> dare to test that idea.

This can still be seen in v3.16-rc1 and in next-20140616.

Debugfs for smsdvb has been unbuildable since v3.12. That problem has
been reported even before v3.12 was released. Does no one care and can
it be removed?

>  drivers/media/common/siano/Kconfig | 3 +--
>  1 file changed, 1 insertion(+), 2 deletions(-)
> 
> diff --git a/drivers/media/common/siano/Kconfig b/drivers/media/common/siano/Kconfig
> index f953d33ee151..4bfbd5f463d1 100644
> --- a/drivers/media/common/siano/Kconfig
> +++ b/drivers/media/common/siano/Kconfig
> @@ -22,8 +22,7 @@ config SMS_SIANO_DEBUGFS
>  	bool "Enable debugfs for smsdvb"
>  	depends on SMS_SIANO_MDTV
>  	depends on DEBUG_FS
> -	depends on SMS_USB_DRV
> -	depends on CONFIG_SMS_USB_DRV = CONFIG_SMS_SDIO_DRV
> +	depends on SMS_USB_DRV = SMS_SDIO_DRV
>  
>  	---help---
>  	  Choose Y to enable visualizing a dump of the frontend


Paul Bolle