2003-09-01 16:35:56

by Tom Rini

[permalink] [raw]
Subject: Re: [PATCH] Firmware loading depends on hotplug support

On Fri, Aug 15, 2003 at 04:27:11PM +0000, [email protected] wrote:
> ChangeSet 1.1065.2.3, 2003/08/15 18:27:11+02:00, [email protected]
>
> [PATCH] Firmware loading depends on hotplug support
>
> This patch makes the firmware loading support only selectable if the
> hotplug support is also enabled.
>
>
> # This patch includes the following deltas:
> # ChangeSet 1.1065.2.2 -> 1.1065.2.3
> # lib/Config.in 1.4 -> 1.5
> #
>
> Config.in | 3 ++-
> 1 files changed, 2 insertions(+), 1 deletion(-)
>
>
> diff -Nru a/lib/Config.in b/lib/Config.in
> --- a/lib/Config.in Mon Sep 1 08:02:24 2003
> +++ b/lib/Config.in Mon Sep 1 08:02:24 2003
> @@ -41,7 +41,8 @@
> fi
> fi
>
> -if [ "$CONFIG_EXPERIMENTAL" = "y" ]; then
> +if [ "$CONFIG_EXPERIMENTAL" = "y" -a \
> + "$CONFIG_HOTPLUG" = "y" ]; then
> tristate 'Hotplug firmware loading support (EXPERIMENTAL)' CONFIG_FW_LOADER
> fi

Is this really a good idea, given that USB devices may be hot-plugged,
without CONFIG_HOTPLUG set (drivers are compiled in, for example) ?

--
Tom Rini
http://gate.crashing.org/~trini/


2003-09-01 16:59:42

by Marcel Holtmann

[permalink] [raw]
Subject: Re: [PATCH] Firmware loading depends on hotplug support

Hi Tom,

> > [PATCH] Firmware loading depends on hotplug support
> >
> > This patch makes the firmware loading support only selectable if the
> > hotplug support is also enabled.
> >
> >
> > # This patch includes the following deltas:
> > # ChangeSet 1.1065.2.2 -> 1.1065.2.3
> > # lib/Config.in 1.4 -> 1.5
> > #
> >
> > Config.in | 3 ++-
> > 1 files changed, 2 insertions(+), 1 deletion(-)
> >
> >
> > diff -Nru a/lib/Config.in b/lib/Config.in
> > --- a/lib/Config.in Mon Sep 1 08:02:24 2003
> > +++ b/lib/Config.in Mon Sep 1 08:02:24 2003
> > @@ -41,7 +41,8 @@
> > fi
> > fi
> >
> > -if [ "$CONFIG_EXPERIMENTAL" = "y" ]; then
> > +if [ "$CONFIG_EXPERIMENTAL" = "y" -a \
> > + "$CONFIG_HOTPLUG" = "y" ]; then
> > tristate 'Hotplug firmware loading support (EXPERIMENTAL)' CONFIG_FW_LOADER
> > fi
>
> Is this really a good idea, given that USB devices may be hot-plugged,
> without CONFIG_HOTPLUG set (drivers are compiled in, for example) ?

this option is only provided for drivers outside the kernel. If a driver
inside the kernel needs it, it is selected automaticly like the CRC32
routines.

If you disable hotplug support the request_firmware() routine will
return an error and this must be handled by the driver.

Regards

Marcel