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: Windows password security audit tool. GUI, reports in PDF.
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <1515408692.12538.10.camel@pengutronix.de>
Date:   Mon, 08 Jan 2018 11:51:32 +0100
From:   Lucas Stach <l.stach@...gutronix.de>
To:     Dong Aisheng <dongas86@...il.com>, Stefan Agner <stefan@...er.ch>
Cc:     shawnguo@...nel.org, kernel@...gutronix.de, fabio.estevam@....com,
        linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org,
        linux-imx@....com
Subject: Re: [PATCH] soc: imx: gpc: de-register power domains only if
 initialized

Am Montag, den 08.01.2018, 18:28 +0800 schrieb Dong Aisheng:
> On Sun, Jan 07, 2018 at 02:49:05PM +0100, Stefan Agner wrote:
> > If power domain information are missing in the device tree, no
> > power domains get initialized. However, imx_gpc_remove tries to
> > remove power domains always in the old DT binding case. Only
> > remove power domains when imx_gpc_probe initialized them in
> > first place.
> > 
> > Fixes: 721cabf6c660 ("soc: imx: move PGC handling to a new GPC
> > driver")
> > Cc: Lucas Stach <l.stach@...gutronix.de>
> > Signed-off-by: Stefan Agner <stefan@...er.ch>
> > ---
> >  drivers/soc/imx/gpc.c | 10 +++++++++-
> >  1 file changed, 9 insertions(+), 1 deletion(-)
> > 
> > diff --git a/drivers/soc/imx/gpc.c b/drivers/soc/imx/gpc.c
> > index 53f7275d6cbd..62bb724726d9 100644
> > --- a/drivers/soc/imx/gpc.c
> > +++ b/drivers/soc/imx/gpc.c
> > @@ -470,13 +470,21 @@ static int imx_gpc_probe(struct
> > platform_device *pdev)
> >  
> >  static int imx_gpc_remove(struct platform_device *pdev)
> >  {
> 
> What's the original purpose of imx_gpc_remove?
> ARM power domain can't be removed.

Why? As long as it stays powered on there is not reason why we wouldn't
be able to remove the driver.

> And why current imx_gpc_remove only remove domains for old DT but not
> for new ones?

With the new binding the power domains will be removed by the sub-
drivers for the domains.

> How about make it un-removable?
> e.g.

I don't see why this would be a good idea. Once more device-dependency
handling is in place we might need to unbind the power domains when the
regulator driver for the domain is unbound. Do you intend to make them
non-removable, too?

Regards,
Lucas

> diff --git a/drivers/soc/imx/gpc.c b/drivers/soc/imx/gpc.c
> index 47e7aa9..7fc6737 100644
> --- a/drivers/soc/imx/gpc.c
> +++ b/drivers/soc/imx/gpc.c
> @@ -454,36 +454,17 @@ static int imx_gpc_probe(struct platform_device
> *pdev)
>         return 0;
>  }
>  
> -static int imx_gpc_remove(struct platform_device *pdev)
> -{
> -       int ret;
> -
> -       /*
> -        * If the old DT binding is used the toplevel driver needs to
> -        * de-register the power domains
> -        */
> -       if (!of_get_child_by_name(pdev->dev.of_node, "pgc")) {
> -               of_genpd_del_provider(pdev->dev.of_node);
> -
> -               ret =
> pm_genpd_remove(&imx_gpc_domains[GPC_PGC_DOMAIN_PU].base);
> -               if (ret)
> -                       return ret;
> -               imx_pgc_put_clocks(&imx_gpc_domains[GPC_PGC_DOMAIN_PU
> ]);
> -
> -               ret =
> pm_genpd_remove(&imx_gpc_domains[GPC_PGC_DOMAIN_ARM].base);
> -               if (ret)
> -                       return ret;
> -       }
> -
> -       return 0;
> -}
> -
>  static struct platform_driver imx_gpc_driver = {
>         .driver = {
>                 .name = "imx-gpc",
>                 .of_match_table = imx_gpc_dt_ids,
> +                /*
> +                 * We can't forcibly eject devices form power
> domain,
> +                 * so we can't really remove power domains once they
> +                 * were added.
> +                 */
> +                .suppress_bind_attrs = true,
>         },
>         .probe = imx_gpc_probe,
> -       .remove = imx_gpc_remove,
>  };
>  builtin_platform_driver(imx_gpc_driver)
> 
> Regards
> Dong Aisheng
> 
> > +	struct device_node *pgc_node;
> >  	int ret;
> >  
> > +	pgc_node = of_get_child_by_name(pdev->dev.of_node, "pgc");
> > +
> > +	/* bail out if DT too old and doesn't provide the
> > necessary info */
> > +	if (!of_property_read_bool(pdev->dev.of_node, "#power-
> > domain-cells") &&
> > +	    !pgc_node)
> > +		return 0;
> > +
> >  	/*
> >  	 * If the old DT binding is used the toplevel driver needs
> > to
> >  	 * de-register the power domains
> >  	 */
> > -	if (!of_get_child_by_name(pdev->dev.of_node, "pgc")) {
> > +	if (!pgc_node) {
> >  		of_genpd_del_provider(pdev->dev.of_node);
> >  
> >  		ret =
> > pm_genpd_remove(&imx_gpc_domains[GPC_PGC_DOMAIN_PU].base);
> > -- 
> > 2.15.1
> > 

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ