On Mon, May 14, 2012 at 11:20:58AM +0300, Igor Grinberg wrote:
> Hi Mark,
Hi Igor.
> Thanks for the great work!
>
> On 05/12/12 00:12, Mark A. Greer wrote:
> > From: "Mark A. Greer" <mgreer@xxxxxxxxxxxxxxx>
> >
> > The am35x family of SoCs has a Davinci EMAC ethernet
> > controller on-chip. Unfortunately, the EMAC is unable
> > to wake the PRCM when there is network activity which
> > leads to a hung or extremely slow system when the MPU
> > has executed a 'wfi' instruction (because of pm_idle
> > or CPUidle). To prevent this, add hooks to the EMAC
> > pm_runtime suspend/resume calls so that hlt is disabled
> > whenever the EMAC is in use.
> >
> > Signed-off-by: Mark A. Greer <mgreer@xxxxxxxxxxxxxxx>
> > ---
> > arch/arm/mach-omap2/am35xx-emac.c | 44 +++++++++++++++++++++++++++++----
> > arch/arm/mach-omap2/am35xx-emac.h | 16 +++++++++---
> > arch/arm/mach-omap2/board-am3517evm.c | 3 ++-
> > arch/arm/mach-omap2/board-cm-t3517.c | 3 ++-
> > 4 files changed, 56 insertions(+), 10 deletions(-)
> >
> > diff --git a/arch/arm/mach-omap2/am35xx-emac.c b/arch/arm/mach-omap2/am35xx-emac.c
> > index 3bb5cb3..22ff968 100644
> > --- a/arch/arm/mach-omap2/am35xx-emac.c
> > +++ b/arch/arm/mach-omap2/am35xx-emac.c
> > @@ -23,6 +23,37 @@
> > #include "control.h"
> > #include "am35xx-emac.h"
> >
> > +/*
> > + * Default pm_lats for the am35x.
> > + * The net effect of using am35xx_emac_pm_lats[] is that
> > + * pm_idle or CPUidle won't be called while the emac
> > + * interface is open. This is required because the
> > + * EMAC can't wake up PRCM so if the MPU is executing
> > + * a 'wfi' instruction (e.g., from pm_idle or CPUidle),
> > + * it won't break out of it due to emac activity.
> > + */
> > +static int am35xx_emac_deactivate_func(struct omap_device *od)
> > +{
> > + enable_hlt();
> > + return omap_device_idle_hwmods(od);
> > +}
> > +
> > +static int am35xx_emac_activate_func(struct omap_device *od)
> > +{
> > + disable_hlt();
> > + return omap_device_enable_hwmods(od);
> > +}
> > +
> > +struct omap_device_pm_latency am35xx_emac_pm_lats[] = {
> > + {
> > + .deactivate_func = am35xx_emac_deactivate_func,
> > + .activate_func = am35xx_emac_activate_func,
> > + .flags = OMAP_DEVICE_LATENCY_AUTO_ADJUST,
> > + },
> > +};
> > +
> > +int am35xx_emac_pm_lats_size = ARRAY_SIZE(am35xx_emac_pm_lats);
> > +
> > static void am35xx_enable_emac_int(void)
> > {
> > u32 regval;
> > @@ -61,12 +92,13 @@ static struct emac_platform_data am35xx_emac_pdata = {
> > static struct mdio_platform_data am35xx_mdio_pdata;
> >
> > static int __init omap_davinci_emac_dev_init(struct omap_hwmod *oh,
> > - void *pdata, int pdata_len)
> > + void *pdata, int pdata_len,
> > + struct omap_device_pm_latency *pm_lats, int pm_lats_size)
> > {
> > struct platform_device *pdev;
> >
> > pdev = omap_device_build(oh->class->name, 0, oh, pdata, pdata_len,
> > - NULL, 0, false);
> > + pm_lats, pm_lats_size, false);
> > if (IS_ERR(pdev)) {
> > WARN(1, "Can't build omap_device for %s:%s.\n",
> > oh->class->name, oh->name);
> > @@ -76,7 +108,8 @@ static int __init omap_davinci_emac_dev_init(struct omap_hwmod *oh,
> > return 0;
> > }
> >
> > -void __init am35xx_emac_init(unsigned long mdio_bus_freq, u8 rmii_en)
> > +void __init am35xx_emac_init(unsigned long mdio_bus_freq, u8 rmii_en,
> > + struct omap_device_pm_latency *pm_lats, int pm_lats_size)
> > {
> > struct omap_hwmod *oh;
> > u32 regval;
> > @@ -91,7 +124,7 @@ void __init am35xx_emac_init(unsigned long mdio_bus_freq, u8 rmii_en)
> > am35xx_mdio_pdata.bus_freq = mdio_bus_freq;
> >
> > ret = omap_davinci_emac_dev_init(oh, &am35xx_mdio_pdata,
> > - sizeof(am35xx_mdio_pdata));
> > + sizeof(am35xx_mdio_pdata), NULL, 0);
> > if (ret) {
> > pr_err("Could not build davinci_mdio hwmod device\n");
> > return;
> > @@ -106,7 +139,8 @@ void __init am35xx_emac_init(unsigned long mdio_bus_freq, u8 rmii_en)
> > am35xx_emac_pdata.rmii_en = rmii_en;
> >
> > ret = omap_davinci_emac_dev_init(oh, &am35xx_emac_pdata,
> > - sizeof(am35xx_emac_pdata));
> > + sizeof(am35xx_emac_pdata),
> > + pm_lats, pm_lats_size);
> > if (ret) {
> > pr_err("Could not build davinci_emac hwmod device\n");
> > return;
> > diff --git a/arch/arm/mach-omap2/am35xx-emac.h b/arch/arm/mach-omap2/am35xx-emac.h
> > index 15c6f9c..7c23808 100644
> > --- a/arch/arm/mach-omap2/am35xx-emac.h
> > +++ b/arch/arm/mach-omap2/am35xx-emac.h
> > @@ -6,10 +6,20 @@
> > * published by the Free Software Foundation.
> > */
> >
> > +#include <plat/omap_device.h>
> > +
> > #define AM35XX_DEFAULT_MDIO_FREQUENCY 1000000
> >
> > -#if defined(CONFIG_TI_DAVINCI_EMAC) || defined(CONFIG_TI_DAVINCI_EMAC_MODULE)
> > -void am35xx_emac_init(unsigned long mdio_bus_freq, u8 rmii_en);
> > +#if IS_ENABLED(CONFIG_TI_DAVINCI_EMAC)
> > +extern struct omap_device_pm_latency am35xx_emac_pm_lats[];
> > +extern int am35xx_emac_pm_lats_size;
> > +
> > +void am35xx_emac_init(unsigned long mdio_bus_freq, u8 rmii_en,
> > + struct omap_device_pm_latency *pm_lats, int pm_lats_size);
> > #else
> > -static inline void am35xx_emac_init(unsigned long mdio_bus_freq, u8 rmii_en) {}
> > +#define am35xx_emac_pm_lats NULL
> > +#define am35xx_emac_pm_lats_size 0
> > +
> > +static inline void am35xx_emac_init(unsigned long mdio_bus_freq, u8 rmii_en,
> > + struct omap_device_pm_latency *pm_lats, int pm_lats_size) {}
> > #endif
> > diff --git a/arch/arm/mach-omap2/board-am3517evm.c b/arch/arm/mach-omap2/board-am3517evm.c
> > index 3645285..fe3a304 100644
> > --- a/arch/arm/mach-omap2/board-am3517evm.c
> > +++ b/arch/arm/mach-omap2/board-am3517evm.c
> > @@ -385,7 +385,8 @@ static void __init am3517_evm_init(void)
> > i2c_register_board_info(1, am3517evm_i2c1_boardinfo,
> > ARRAY_SIZE(am3517evm_i2c1_boardinfo));
> > /*Ethernet*/
> > - am35xx_emac_init(AM35XX_DEFAULT_MDIO_FREQUENCY, 1);
> > + am35xx_emac_init(AM35XX_DEFAULT_MDIO_FREQUENCY, 1,
> > + am35xx_emac_pm_lats, am35xx_emac_pm_lats_size);
>
> Why is it essential for board file to pass these pm structures?
> Is it something board specific?
> Can't you just use them inside the am35xx-emac.c file?
Great question. I went back & forth on this mayself (I have 3 different
versions sitting in my repo :). I ended up passing the pm structures in
case there is a variant that comes along that doesn't need this fixup (or
a board that wants to do this plus additional things).
The variants that I have are:
1) Leave am35xx_emac_init() interface the way it is and automatically
add the necessary things. The problem is lack of flexibility if & when
some other variant comes along.
2) Add pm_lats & pm_lats_size to am35xx_emac_init() but if they're NULL
or 0, then use the defaults that are the same as what's in this patch.
More flexible but doesn't easily allow a board file to NULL out the pm
structure (have to make up and pass NULL version).
3) Pass the pm struct via am35xx_emac_init() and just make the default
available for use. This is what I submitted since it was the most flexible.
Which one do you prefer or do you have something else in mind?
Thanks,
Mark
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
