Re: [PATCH 2/3] drm/bridge: ti-sn65dsi86: add debugfs

[Laurent Pinchart]

Hi Rob,

On Thu, Jul 04, 2019 at 06:56:56AM -0700, Rob Clark wrote:
> On Thu, Jul 4, 2019 at 5:35 AM Laurent Pinchartwrote:
> > On Thu, Jul 04, 2019 at 02:31:20PM +0200, Andrzej Hajda wrote:
> > > On 02.07.2019 17:44, Rob Clark wrote:
> > > > From: Rob Clark <robdclark@xxxxxxxxxxxx>
> > > >
> > > > Add a debugfs file to show status registers.
> > > >
> > > > Signed-off-by: Rob Clark <robdclark@xxxxxxxxxxxx>
> > > > ---
> > > >  drivers/gpu/drm/bridge/ti-sn65dsi86.c | 42 +++++++++++++++++++++++++++
> > > >  1 file changed, 42 insertions(+)
> > > >
> > > > diff --git a/drivers/gpu/drm/bridge/ti-sn65dsi86.c b/drivers/gpu/drm/bridge/ti-sn65dsi86.c
> > > > index f1a2493b86d9..a6f27648c015 100644
> > > > --- a/drivers/gpu/drm/bridge/ti-sn65dsi86.c
> > > > +++ b/drivers/gpu/drm/bridge/ti-sn65dsi86.c
> > > > @@ -5,6 +5,7 @@
> > > >   */
> > > >
> > > >  #include <linux/clk.h>
> > > > +#include <linux/debugfs.h>
> > > >  #include <linux/gpio/consumer.h>
> > > >  #include <linux/i2c.h>
> > > >  #include <linux/iopoll.h>
> > > > @@ -109,6 +110,7 @@ struct ti_sn_bridge {
> > > >     struct drm_dp_aux               aux;
> > > >     struct drm_bridge               bridge;
> > > >     struct drm_connector            connector;
> > > > +   struct dentry                   *debugfs;
> > > >     struct device_node              *host_node;
> > > >     struct mipi_dsi_device          *dsi;
> > > >     struct clk                      *refclk;
> > > > @@ -178,6 +180,42 @@ static const struct dev_pm_ops ti_sn_bridge_pm_ops = {
> > > >     SET_RUNTIME_PM_OPS(ti_sn_bridge_suspend, ti_sn_bridge_resume, NULL)
> > > >  };
> > > >
> > > > +static int status_show(struct seq_file *s, void *data)
> > > > +{
> > > > +   struct ti_sn_bridge *pdata = s->private;
> > > > +   unsigned int reg, val;
> > > > +
> > > > +   seq_puts(s, "STATUS REGISTERS:\n");
> >
> > NO NEED TO SHOUT :-)
> >
> > > > +
> > > > +   pm_runtime_get_sync(pdata->dev);
> > > > +
> > > > +   /* IRQ Status Registers, see Table 31 in datasheet */
> > > > +   for (reg = 0xf0; reg <= 0xf8; reg++) {
> > > > +           regmap_read(pdata->regmap, reg, &val);
> > > > +           seq_printf(s, "[0x%02x] = 0x%08x\n", reg, val);
> > > > +   }
> > > > +
> > > > +   pm_runtime_put(pdata->dev);
> > > > +
> > > > +   return 0;
> > > > +}
> > > > +
> > > > +DEFINE_SHOW_ATTRIBUTE(status);
> > > > +
> > > > +static void ti_sn_debugfs_init(struct ti_sn_bridge *pdata)
> > > > +{
> > > > +   pdata->debugfs = debugfs_create_dir("ti_sn65dsi86", NULL);
> > >
> > > If some day we will have board with two such bridges there will be a
> > > problem.
> >
> > Could we use the platform device name for this ?
> 
> hmm, yeah, that would solve the 2x bridges issue
> 
> > > Anyway:
> > >
> > > Reviewed-by: Andrzej Hajda <a.hajda@xxxxxxxxxxx>
> > >
> > > > +
> > > > +   debugfs_create_file("status", 0600, pdata->debugfs, pdata,
> > > > +                   &status_fops);
> > > > +}
> > > > +
> > > > +static void ti_sn_debugfs_remove(struct ti_sn_bridge *pdata)
> > > > +{
> > > > +   debugfs_remove_recursive(pdata->debugfs);
> > > > +   pdata->debugfs = NULL;
> > > > +}
> > > > +
> >
> > You need to conditionally-compile this based on CONFIG_DEBUG_FS.
> 
> Hmm, is that really true?  Debugfs appears to be sufficently stub'd w/
> inline no-ops in the !CONFIG_DEBUG_FS case

You're right, my bad. I wonder if the compiler will optimise the above
two functions out. It might warrant a CONFIG_DEBUG_FS check here for
that reason, but that's really bikeshedding. So with the 2x bridges
issue addressed, I think the patch will be good.

-- 
Regards,

Laurent Pinchart