| Message ID | 6238a78e32fa21f0c795406b6cba7bce7af92577.1708513940.git.geert+renesas@glider.be |
|---|---|
| State | New |
| Headers | show |
| Series | pinctrl: renesas: Allow the compiler to optimize away sh_pfc_pm | expand |
On Wed, Feb 21, 2024 at 12:13:59PM +0100, Geert Uytterhoeven wrote: > The conversion to DEFINE_NOIRQ_DEV_PM_OPS() lost the ability of the > compiler to optimize away the struct dev_pm_ops object when it is not > needed. > > Fix this by replacing the use of pm_sleep_ptr() by a custom wrapper. > > Fixes: 727eb02eb753375e ("pinctrl: renesas: Switch to use DEFINE_NOIRQ_DEV_PM_OPS() helper") Is this a new requirement (I mean more than 12 digits of SHA)? ... > --- > To be queued in renesas-pinctrl for v6.9. > > Alternatively, one could add a unified definition: > > #define pm_psci_sleep_ptr(_ptr) PTR_IF(IS_ENABLED(CONFIG_PM_SLEEP) && \ > IS_ENABLED(CONFIG_ARM_PSCI_FW), (_ptr)) > > Since there are already separate sections for CONFIG_ARM_PSCI_FW enabled > vs. disabled, I split in two and simplified the definition. I'm fine with current approach, IIUC we discussed this that time and the 727eb02eb753 was not the first version (or I can be mistaken, I don't remember exact driver that the discussion was about during the DEFINE_NOIRQ_DEV_PM_OPS() conversions). FWIW, Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Hi Andy, Cc Paul On Wed, Feb 21, 2024 at 2:19 PM Andy Shevchenko <andriy.shevchenko@linux.intel.com> wrote: > On Wed, Feb 21, 2024 at 12:13:59PM +0100, Geert Uytterhoeven wrote: > > The conversion to DEFINE_NOIRQ_DEV_PM_OPS() lost the ability of the > > compiler to optimize away the struct dev_pm_ops object when it is not > > needed. > > > > Fix this by replacing the use of pm_sleep_ptr() by a custom wrapper. > > > > Fixes: 727eb02eb753375e ("pinctrl: renesas: Switch to use DEFINE_NOIRQ_DEV_PM_OPS() helper") > > Is this a new requirement (I mean more than 12 digits of SHA)? Nah, better safe than sorry, as 12 is becoming a bit small. (Yes, I have to rework/resubmit patches related to that) > > --- > > To be queued in renesas-pinctrl for v6.9. > > > > Alternatively, one could add a unified definition: > > > > #define pm_psci_sleep_ptr(_ptr) PTR_IF(IS_ENABLED(CONFIG_PM_SLEEP) && \ > > IS_ENABLED(CONFIG_ARM_PSCI_FW), (_ptr)) > > > > Since there are already separate sections for CONFIG_ARM_PSCI_FW enabled > > vs. disabled, I split in two and simplified the definition. > > I'm fine with current approach, IIUC we discussed this that time and the > 727eb02eb753 was not the first version (or I can be mistaken, I don't remember > exact driver that the discussion was about during the DEFINE_NOIRQ_DEV_PM_OPS() > conversions). Thanks, I had forgotten about that discussion... > FWIW, > Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Thanks! Gr{oetje,eeting}s, Geert
diff --git a/drivers/pinctrl/renesas/core.c b/drivers/pinctrl/renesas/core.c index 78331d7f7cca9d8c..96d6040a8871419b 100644 --- a/drivers/pinctrl/renesas/core.c +++ b/drivers/pinctrl/renesas/core.c @@ -737,10 +737,12 @@ static int sh_pfc_resume_noirq(struct device *dev) sh_pfc_walk_regs(pfc, sh_pfc_restore_reg); return 0; } +#define pm_psci_sleep_ptr(_ptr) pm_sleep_ptr(_ptr) #else static int sh_pfc_suspend_init(struct sh_pfc *pfc) { return 0; } static int sh_pfc_suspend_noirq(struct device *dev) { return 0; } static int sh_pfc_resume_noirq(struct device *dev) { return 0; } +#define pm_psci_sleep_ptr(_ptr) PTR_IF(false, (_ptr)) #endif /* CONFIG_ARM_PSCI_FW */ static DEFINE_NOIRQ_DEV_PM_OPS(sh_pfc_pm, sh_pfc_suspend_noirq, sh_pfc_resume_noirq); @@ -1423,7 +1425,7 @@ static struct platform_driver sh_pfc_driver = { .driver = { .name = DRV_NAME, .of_match_table = of_match_ptr(sh_pfc_of_table), - .pm = pm_sleep_ptr(&sh_pfc_pm), + .pm = pm_psci_sleep_ptr(&sh_pfc_pm), }, };
The conversion to DEFINE_NOIRQ_DEV_PM_OPS() lost the ability of the compiler to optimize away the struct dev_pm_ops object when it is not needed. Fix this by replacing the use of pm_sleep_ptr() by a custom wrapper. Fixes: 727eb02eb753375e ("pinctrl: renesas: Switch to use DEFINE_NOIRQ_DEV_PM_OPS() helper") Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> --- To be queued in renesas-pinctrl for v6.9. Alternatively, one could add a unified definition: #define pm_psci_sleep_ptr(_ptr) PTR_IF(IS_ENABLED(CONFIG_PM_SLEEP) && \ IS_ENABLED(CONFIG_ARM_PSCI_FW), (_ptr)) Since there are already separate sections for CONFIG_ARM_PSCI_FW enabled vs. disabled, I split in two and simplified the definition. --- drivers/pinctrl/renesas/core.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-)