Skip to contentSkip to conversation

PWM backlight build warning

DISCUSSION Tony Prisk 3 participants 4 messages Apr 6, 2013

lore
Share & tools
lore

Tony Prisk

author
#0 lore raw

Hi Thierry,

Sorry it's taken so long to report this but I've been tied up with other patches recently. I noticed a build warning on the pwm_bl driver in 3.9-rcX but I'm not sure what the correct resolution would be. This patch:

commit e2c17bc6f717a8847df2a867caec6ba4fe85f3fc
backlight: convert pwm_bl to dev_pm_ops

Should be no functional changes, mainly a reorganisation to support future work.

[[email protected]: fix CONFIG_PM=n build]
Cc Richard Purdie <[email protected]>
Cc Florian Tobias Schandinat <[email protected]>

has introduced a two build warnings in my config:
drivers/video/backlight/pwm_bl.c:278:12: warning: ‘pwm_backlight_suspend’ defined but not used [-Wunused-function] drivers/video/backlight/pwm_bl.c:292:12: warning: ‘pwm_backlight_resume’ defined but not used [-Wunused-function]

Although I noticed it doesn't have your acked-by ;)

Selecting "Run-time PM core functionality" without "Suspend to RAM and standby" causes the unused function warning. With CONFIG_PM=n, no warnings. With CONFIG_PM=y and CONFIG_PM_SLEEP=y, no warnings.

The problem occurs because when CONFIG_PM=y but CONFIG_PM_SLEEP=n:

#define SIMPLE_DEV_PM_OPS(name, suspend_fn, resume_fn) \
const struct dev_pm_ops name = { \
SET_SYSTEM_SLEEP_PM_OPS(syspend_fn, resume_fn) \
}

#ifdef CONFIG_PM_SLEEP
#define SET_SYSTEM_SLEEP_PM_OPS(suspend_fn, resume_fn) \
.suspend = suspend_fn, \
.resume = resume_fn, \
.freeze = suspend_fn, \
.thaw = resume_fn, \
.poweroff = suspend_fn, \
.restore = resume_fn,
#else
#define SET_SYSTEM_SLEEP_PM_OPS(suspend_fn, resume_fn) #endif

I guess this would be a problem on every driver that has been converted to SIMPLE_DEV_PM_OPS. I don't know enough about the PM code to offer a fix, or whethermy chosen config is just 'invalid'.

Regards
Tony P

#1 lore raw
Quoted 6 lines · Tony Prisk

The problem occurs because when CONFIG_PM=y but CONFIG_PM_SLEEP=n:

I guess this would be a problem on every driver that has been converted to SIMPLE_DEV_PM_OPS.
I don't know enough about the PM code to offer a fix, or whethermy chosen config is just 'invalid'.

You just need to put the system suspend functions inside an ifdef
CONFIG_PM_SLEEP instead of CONFIG_PM.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20130406/4475b2d3/attachment.sig>

#2 lore raw
Quoted 11 lines · Mark Brown
Quoted 6 lines · Tony Prisk

The problem occurs because when CONFIG_PM=y but CONFIG_PM_SLEEP=n:

I guess this would be a problem on every driver that has been converted to SIMPLE_DEV_PM_OPS.
I don't know enough about the PM code to offer a fix, or whethermy chosen config is just 'invalid'.

You just need to put the system suspend functions inside an ifdef CONFIG_PM_SLEEP instead of CONFIG_PM.

I carry a fix for that in my for-next branch.

Thierry
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 836 bytes
Desc: not available
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20130406/9787f143/attachment.sig>

author
#3 lore raw
Quoted 8 lines · Thierry Reding
Quoted 7 lines · Mark Brown
Quoted 5 lines · Tony Prisk

The problem occurs because when CONFIG_PM=y but CONFIG_PM_SLEEP=n:
I guess this would be a problem on every driver that has been converted to SIMPLE_DEV_PM_OPS.
I don't know enough about the PM code to offer a fix, or whethermy chosen config is just 'invalid'.

Quoted 5 lines
Quoted 2 lines

You just need to put the system suspend functions inside an ifdef CONFIG_PM_SLEEP instead of CONFIG_PM.

I carry a fix for that in my for-next branch.

Thierry

Ahh ok, all is good then.

Regards
Tony P