+Dave, Tony, Keerthy, Hi Stefan, On Tuesday 28 June 2016 09:26 PM, Stefan Eichenberger wrote: > If a module is not available, omap_hwmod will still try to access the > registers if it is present in the hwmod_list. If the module is not > powered at all (e.g. rtc) then linux can't start. It is possible > to disable the module by setting the state to disabled and to assign > ti,hwmods an empty string but this leads to the following warning: > [ 0.188255] omap_hwmod: rtc missing dt data > [ 0.188273] omap_hwmod: rtc: no MPU register target found > [ 0.188287] omap_hwmod: rtc: no dt node > [ 0.188300] ------------[ cut here ]------------ > [ 0.188330] WARNING: CPU: 0 PID: 1 at arch/arm/mach-omap2/omap_hwmod.c:2493 _init+0x1fc/0x45c() > [ 0.188344] omap_hwmod: rtc: doesn't have mpu register target base > [ 0.188357] Modules linked in: > [ 0.188388] CPU: 0 PID: 1 Comm: swapper/0 Not tainted 4.4.0-00004-g17704c9-dirty #24 > [ 0.188402] Hardware name: Generic AM33XX (Flattened Device Tree) > [ 0.188445] [<c0017c24>] (unwind_backtrace) from [<c0013f88>] (show_stack+0x10/0x14) > [ 0.188474] [<c0013f88>] (show_stack) from [<c0347860>] (dump_stack+0x84/0x9c) > [ 0.188508] [<c0347860>] (dump_stack) from [<c003ca40>] (warn_slowpath_common+0x78/0xb4) > [ 0.188534] [<c003ca40>] (warn_slowpath_common) from [<c003caac>] (warn_slowpath_fmt+0x30/0x40) > [ 0.188558] [<c003caac>] (warn_slowpath_fmt) from [<c08b76e0>] (_init+0x1fc/0x45c) > [ 0.188589] [<c08b76e0>] (_init) from [<c0027844>] (omap_hwmod_for_each+0x34/0x60) > [ 0.188613] [<c0027844>] (omap_hwmod_for_each) from [<c08b7d04>] (__omap_hwmod_setup_all+0x24/0x40) > [ 0.188638] [<c08b7d04>] (__omap_hwmod_setup_all) from [<c00098a4>] (do_one_initcall+0x80/0x1e0) > [ 0.188675] [<c00098a4>] (do_one_initcall) from [<c08a9ec0>] (kernel_init_freeable+0x214/0x2e4) > [ 0.188707] [<c08a9ec0>] (kernel_init_freeable) from [<c0642f4c>] (kernel_init+0x8/0xec) > [ 0.188737] [<c0642f4c>] (kernel_init) from [<c000f810>] (ret_from_fork+0x14/0x24) > [ 0.188919] ---[ end trace 9e6a5f1c3af2b60d ]--- > > Unfortunately we cannot check of_device_is_available because timers are > set to disabled. In this case we would not setup the timer module > what leads to a crash during boot. The same issue has been encountered on AM43x-epos evm. Dave has submitted a similar patch[1](Complete Series[2]) to avoid hwmod handling of sudh IPs. Not sure what happened to this thread but later it was decided not to register such IPs to hwmod using compatible properties as done here[3]. Probably Dave can comment more. [1] https://patchwork.kernel.org/patch/5945961/ [2] https://www.mail-archive.com/linux-omap@xxxxxxxxxxxxxxx/msg114645.html [3] https://patchwork.kernel.org/patch/8357011/ Thanks and regards, Lokesh > > For this reason a new ti,hwmod-disabled option was added that prevents > omap_hwmod to setup the module. > > Signed-off-by: Stefan Eichenberger <stefan.eichenberger@xxxxxxxxxxxxx> > --- > Documentation/devicetree/bindings/arm/omap/omap.txt | 3 +++ > arch/arm/mach-omap2/omap_hwmod.c | 15 +++++++++++++++ > arch/arm/mach-omap2/omap_hwmod.h | 1 + > 3 files changed, 19 insertions(+) > > diff --git a/Documentation/devicetree/bindings/arm/omap/omap.txt b/Documentation/devicetree/bindings/arm/omap/omap.txt > index 94b57f2..bc79fe1 100644 > --- a/Documentation/devicetree/bindings/arm/omap/omap.txt > +++ b/Documentation/devicetree/bindings/arm/omap/omap.txt > @@ -21,6 +21,9 @@ Required properties: > Optional properties: > - ti,no_idle_on_suspend: When present, it prevents the PM to idle the module > during suspend. > +- ti,hwmod-disabled: When present, it will not try to initialize the module. > + This is necessary if e.g. the rtc is not powered at all and the registers > + are not accessible. > - ti,no-reset-on-init: When present, the module should not be reset at init > - ti,no-idle-on-init: When present, the module should not be idled at init > - ti,no-idle: When present, the module is never allowed to idle. > diff --git a/arch/arm/mach-omap2/omap_hwmod.c b/arch/arm/mach-omap2/omap_hwmod.c > index 83cb527..476e3ef 100644 > --- a/arch/arm/mach-omap2/omap_hwmod.c > +++ b/arch/arm/mach-omap2/omap_hwmod.c > @@ -2080,6 +2080,9 @@ static int _enable(struct omap_hwmod *oh) > int r; > int hwsup = 0; > > + if (oh->_state == _HWMOD_STATE_NOT_AVAILABLE) > + return 0; > + > pr_debug("omap_hwmod: %s: enabling\n", oh->name); > > /* > @@ -2200,6 +2203,9 @@ static int _enable(struct omap_hwmod *oh) > */ > static int _idle(struct omap_hwmod *oh) > { > + if (oh->_state == _HWMOD_STATE_NOT_AVAILABLE) > + return 0; > + > if (oh->flags & HWMOD_NO_IDLE) { > oh->_int_flags |= _HWMOD_SKIP_ENABLE; > return 0; > @@ -2491,6 +2497,15 @@ static int __init _init(struct omap_hwmod *oh, void *data) > oh->name, np->name); > } > > + if (np) { > + if (of_find_property(np, "ti,hwmod-disabled", NULL)) { > + pr_debug("omap_hwomd: %s is disabled, set to not available\n", > + oh->name); > + oh->_state = _HWMOD_STATE_NOT_AVAILABLE; > + return 0; > + } > + } > + > r = _init_mpu_rt_base(oh, NULL, index, np); > if (r < 0) { > WARN(1, "omap_hwmod: %s: doesn't have mpu register target base\n", > diff --git a/arch/arm/mach-omap2/omap_hwmod.h b/arch/arm/mach-omap2/omap_hwmod.h > index 4041bad..df0e83b 100644 > --- a/arch/arm/mach-omap2/omap_hwmod.h > +++ b/arch/arm/mach-omap2/omap_hwmod.h > @@ -573,6 +573,7 @@ struct omap_hwmod_omap4_prcm { > #define _HWMOD_STATE_ENABLED 4 > #define _HWMOD_STATE_IDLE 5 > #define _HWMOD_STATE_DISABLED 6 > +#define _HWMOD_STATE_NOT_AVAILABLE 7 > > /** > * struct omap_hwmod_class - the type of an IP block > -- 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
