On Tue, Apr 11, 2017 at 09:53:38AM -0500, Alan Tull wrote:
> On Tue, Apr 11, 2017 at 4:22 AM, Tobias Klauser <tklau...@distanz.ch> wrote:
>
> Hi Tobias,
>
> Thanks!
>
> > If either _alt_hps2fpga_enable_set() or fpga_bridge_register() fail in
> > alt_fpga_bridge_probe(), the clock remains enabled and prepared. Also,
> > in the error path for _alt_hps2fpga_enable_set() a call to
> > fpga_bridge_unregister() is made even though the bridge was not
> > registered yet.
> >
> > Remove the unnecessary call to fpga_bridge_unregister() and call
> > clk_disable_unprepare() in both error paths in order to make sure the
> > clock gets properly disabled and unprepared.
> >
> > Signed-off-by: Tobias Klauser <tklau...@distanz.ch>
>
Acked-by: Moritz Fischer <m...@kernel.org>
> Signed-off-by: Alan Tull <at...@kernel.org>
>
> > ---
> > drivers/fpga/altera-hps2fpga.c | 15 +++++++++------
> > 1 file changed, 9 insertions(+), 6 deletions(-)
> >
> > diff --git a/drivers/fpga/altera-hps2fpga.c b/drivers/fpga/altera-hps2fpga.c
> > index 4b354c79be31..3066b805f2d0 100644
> > --- a/drivers/fpga/altera-hps2fpga.c
> > +++ b/drivers/fpga/altera-hps2fpga.c
> > @@ -181,15 +181,18 @@ static int alt_fpga_bridge_probe(struct
> > platform_device *pdev)
> > (enable ? "enabling" : "disabling"));
> >
> > ret = _alt_hps2fpga_enable_set(priv, enable);
> > - if (ret) {
> > - fpga_bridge_unregister(&pdev->dev);
> > - return ret;
> > - }
> > + if (ret)
> > + goto err;
> > }
> > }
> >
> > - return fpga_bridge_register(dev, priv->name,
> > &altera_hps2fpga_br_ops,
> > - priv);
> > + ret = fpga_bridge_register(dev, priv->name, &altera_hps2fpga_br_ops,
> > + priv);
> > +err:
> > + if (ret)
> > + clk_disable_unprepare(priv->clk);
> > +
> > + return ret;
> > }
> >
> > static int alt_fpga_bridge_remove(struct platform_device *pdev)
> > --
> > 2.12.2
> >
> >
