linux-kernel - Re: [PATCH v3 2/6] firmware: ti

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives
Hash Suite for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <20241029120302.3twkliytrn5hjufi@sleek>
Date: Tue, 29 Oct 2024 07:03:02 -0500
From: Nishanth Menon <nm@...com>
To: Markus Schneider-Pargmann <msp@...libre.com>
CC: Tero Kristo <kristo@...nel.org>, Santosh Shilimkar <ssantosh@...nel.org>,
        Vignesh Raghavendra <vigneshr@...com>, Rob Herring <robh@...nel.org>,
        Krzysztof Kozlowski <krzk+dt@...nel.org>,
        Conor Dooley <conor+dt@...nel.org>, Anand Gadiyar <gadiyar@...com>,
        <linux-arm-kernel@...ts.infradead.org>, <linux-kernel@...r.kernel.org>,
        <devicetree@...r.kernel.org>, Vishal Mahaveer
	<vishalm@...com>,
        Kevin Hilman <khilman@...libre.com>, Dhruva Gole
	<d-gole@...com>
Subject: Re: [PATCH v3 2/6] firmware: ti_sci: Partial-IO support

On 15:49-20241028, Markus Schneider-Pargmann wrote:
> Hi Nishanth,
> 
> On Fri, Oct 25, 2024 at 12:42:04PM GMT, Nishanth Menon wrote:
> > On 16:39-20241012, Markus Schneider-Pargmann wrote:
> > [...]
> > > 
> > > The possible wakeup devices are found by checking which devices are
> > > powered by the regulator supplying the "vddshv_canuart" line. These are
> > > considered possible wakeup sources. Only wakeup sources that are
> > > actually enabled by the user will be considered as a an active wakeup
> > > source. If none of the wakeup sources are enabled the system will do a
> > > normal poweroff. If at least one wakeup source is enabled it will
> > > instead send a TI_SCI_MSG_PREPARE_SLEEP message from the sys_off
> > > handler. Sending this message will result in an immediate shutdown of
> > > the system. No execution is expected after this point. The code will
> > > wait for 5s and do an emergency_restart afterwards if Partial-IO wasn't
> > > entered at that point.
> > > 
> > [...]
> > 
> > > +static bool tisci_canuart_wakeup_enabled(struct ti_sci_info *info)
> > > +{
> > > +	static const char canuart_name[] = "vddshv_canuart";
> > > +	struct device_node *wakeup_node = NULL;
> > > +
> > > +	for (wakeup_node = of_find_node_with_property(NULL, "vio-supply");
> > > +	     wakeup_node;
> > > +	     wakeup_node = of_find_node_with_property(wakeup_node, "vio-supply")) {
> > > +		struct device_node *supply_node;
> > > +		const char *supply_name;
> > > +		struct platform_device *pdev;
> > > +		int ret;
> > > +
> > > +		supply_node = of_parse_phandle(wakeup_node, "vio-supply", 0);
> > > +		if (!supply_node)
> > > +			continue;
> > > +
> > > +		ret = of_property_read_string(supply_node, "regulator-name", &supply_name);
> > > +		of_node_put(supply_node);
> > > +		if (ret) {
> > > +			dev_warn(info->dev, "Failed to parse vio-supply phandle at %pOF %d\n",
> > > +				 wakeup_node, ret);
> > > +			continue;
> > > +		}
> > > +
> > > +		if (strncmp(canuart_name, supply_name, strlen(canuart_name)))
> > > +			continue;
> > > +
> > > +		pdev = of_find_device_by_node(wakeup_node);
> > > +		if (!pdev)
> > > +			continue;
> > > +
> > > +		if (device_may_wakeup(&pdev->dev)) {
> > > +			dev_dbg(info->dev, "%pOF identified as wakeup source for Partial-IO\n",
> > > +				wakeup_node);
> > > +			put_device(&pdev->dev);
> > > +			of_node_put(wakeup_node);
> > > +			return true;
> > > +		}
> > > +		put_device(&pdev->dev);
> > > +	}
> > > +
> > > +	return false;
> > > +}
> > > +
> > 
> > What is the binding that supports this? I just do not think that
> > scanning the entire tree for vio-supply implies you will get thr right
> > property here.
> > 
> > Just giving an example to illustrate this point:
> > Documentation/devicetree/bindings/net/wireless/ti,wl1251.txt says it
> > needs vio-supply -> so i have a node with the wireless supply as
> > vio-supply -> Since we are scanning from NULL for vio-supply, we hit
> > that, that is a bad choice for enabling io-retention.
> 
> There is no bining that specifically supports this as I think it is not
> needed. The devices that are capable to wakeup the system from
> Partial-IO are all powered through one supply line that is always-on. It
> is called 'vddshv_canuart' and the name of this supply is checked
> in the above code as well. Yes I am using 'vio-supply', but only to

In effect, you are looking for nodes that have vio-supply pointing to
a regulator called vddshv_canuart. Not only is it too specific to a
device, but a board as well. Without documentation in binding, users
who don't have sufficient information is bound to mess this up. Further,
vddshv_canuart may not even be a regulator - there will need to be
checks for that on top of just a strncmp.


> search for the potential consumers of this supply.
> So wl1251 will be skipped in above code at
> 
>   if (strncmp(canuart_name, supply_name, strlen(canuart_name)))

Aah, thanks, but sorry, but I would prefer the drivers handle the
specifics. If a new peripheral comes on to the list for a different
device, or a different regulator name appears for can, we would be
dealing with name mapping etc.

-- 
Regards,
Nishanth Menon
Key (0xDDB5849D1736249D) / Fingerprint: F8A2 8693 54EB 8232 17A3  1A34 DDB5 849D 1736 249D