Re: [PATCH] power: twl4030_charger: clear IRQs after handling them

From: Tony Lindgren <tony@atomide.com>
To: Nishanth Menon <nm@ti.com>
Cc: Felipe Balbi <balbi@ti.com>,
	Russell King <linux@arm.linux.org.uk>,
	Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
	dbaryshkov@gmail.com,
	Linux OMAP Mailing List <linux-omap@vger.kernel.org>,
	dwmw2@infradead.org,
	Linux ARM Kernel Mailing List 
	<linux-arm-kernel@lists.infradead.org>
Subject: Re: [PATCH] power: twl4030_charger: clear IRQs after handling them
Date: Thu, 8 May 2014 18:03:12 -0700	[thread overview]
Message-ID: <20140509010312.GK2198@atomide.com> (raw)
In-Reply-To: <20140507002421.GA19711@kahuna>

* Nishanth Menon <nm@ti.com> [140506 17:25]:
> Subject: [PATCH] power: twl4030_charger: detect battery presence prior to
>  enabling charger
> 
> TWL4030's Battery Charger seems to be designed for non-hotpluggable
> batteries.
> 
> If battery is not present in the system, BATSTS is always set with the
> expectation that software will take actions to move to a required safe
> state (could be power down or disable various charger paths).
> 
> It does not seem possible even by manipulating the edge detection
> of the event (using BCIEDR2 register) to have a consistent hotplug
> handling. This seems to be the result of BATSTS interrupt generated
> when the thermistor of the battery pack is disconnected from the
> dedicated ADIN1 pin. Clearing the status just results in the status
> being regenerated by the monitoring ADC(MADC) and disabling the
> edges of event just makes hotplug no longer function. The only
> other option is to disable the detection of the MADC by disabling
> BCIMFEN4::BATSTSMCHGEN (battery presence detector) - but then, we can
> never again detect battery reconnection.
> 
> So, detect battery presence based on precharge(which is hardware
> automatic state) or default main charger configuration at the time of
> probe and enable charger logic only if battery was present.
> 
> Reported-by: Russell King <linux@arm.linux.org.uk>
> Signed-off-by: Nishanth Menon <nm@ti.com>

Gets rid of the errors for me if CONFIG_CHARGER_TWL4030=y.
Looks like we don't have that enabled by default in
omap2plus_defconfig which explain why it's taken so long to
notice this one:

Tested-by: Tony Lindgren <tony@atomide.com>