Re: [PATCH] ipmi: replace call to try_smi_init() with call to add_smi()

[Tony Camuso <tcamuso-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>]

Rescinding.

This hunk is already upstream.

On 01/08/2013 03:12 PM, Tony Camuso wrote:
> Upstream commit 9e368fa0 added function ipmi_pnp_probe(), which
> calls try_smi_init() directly.
>
> try_smi_init() allocates resources for IPMI driver. However
> try_smi_init() can be called again, and the allocated resources
> can be unexpectedly released when the same IPMI resource
> information is found on both ACPI namespace and SMBIOS table.
> ipmi_pnp_probe() should call add_smi() instead of try_smi_init()
>
> For reference, see upstream commit ...
> 2407d77a1a013b88ee3b817f2b934e420e5376f5
>
> Signed-off-by: Tony Camuso <tcamuso-H+wXaHxf7aLQT0dZR+AlfA@public.gmane.org>
>
> ---
>   drivers/char/ipmi/ipmi_si_intf.c |    2 +-
>   1 files changed, 1 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/char/ipmi/ipmi_si_intf.c b/drivers/char/ipmi/ipmi_si_intf.c
> index afb89be..20d13c6 100644
> --- a/drivers/char/ipmi/ipmi_si_intf.c
> +++ b/drivers/char/ipmi/ipmi_si_intf.c
> @@ -2223,7 +2223,7 @@ static int __devinit ipmi_pnp_probe(struct pnp_dev *dev,
>       info->dev = &dev->dev;
>       pnp_set_drvdata(dev, info);
>
> -    return try_smi_init(info);
> +    return add_smi(info);
>
>   err_free:
>       kfree(info);