[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <YGM5YMjusx214+pl@hovoldconsulting.com>
Date: Tue, 30 Mar 2021 16:44:48 +0200
From: Johan Hovold <johan@...nel.org>
To: Mauro Carvalho Chehab <mchehab+huawei@...nel.org>
Cc: Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
linux-kernel@...r.kernel.org, linux-usb@...r.kernel.org
Subject: Re: [PATCH v2 2/7] USB: serial: xr: use a table for device-specific
settings
On Wed, Mar 24, 2021 at 08:41:06AM +0100, Mauro Carvalho Chehab wrote:
> The same driver is used by a wide range of MaxLinear devices.
>
> Other models are close enough to use the same driver, but they
> use a different register set.
>
> So, instead of having the registers hardcoded at the driver,
> use a table. This will allow further patches to add support for
> other devices.
>
> Signed-off-by: Mauro Carvalho Chehab <mchehab+huawei@...nel.org>
> ---
> drivers/usb/serial/xr_serial.c | 151 ++++++++++++++++++++++++---------
> 1 file changed, 113 insertions(+), 38 deletions(-)
> static int xr_probe(struct usb_serial *serial, const struct usb_device_id *id)
> {
> + struct xr_port_private *port_priv;
> +
> /* Don't bind to control interface */
> if (serial->interface->cur_altsetting->desc.bInterfaceNumber == 0)
> return -ENODEV;
>
> + port_priv = kzalloc(sizeof(*port_priv), GFP_KERNEL);
> + if (!port_priv)
> + return -ENOMEM;
For historical reasons, you cannot allocate memory in probe() directly
(unless using devres) or this can leak on later probe errors.
Instead interface-wide allocations are done in attach() and released in
release(), while port-specific allocations are done in port_probe() and
released in port_remove().
Johan
Powered by blists - more mailing lists