diff mbox

[v2,26/28] drm/i2c: tda998x: code optimization

Message ID 20140109120725.0b0f3e55@armhf (mailing list archive)
State New, archived
Headers show

Commit Message

Jean-Francois Moine Jan. 9, 2014, 11:07 a.m. UTC
This patch reduces the number of I2C exchanges by setting many bits in
one write and removing a useless write.

Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
---
 drivers/gpu/drm/i2c/tda998x_drv.c   | 10 ++++------
 1 file changed, 4 insertions(+), 6 deletions(-)

Comments

Russell King - ARM Linux Jan. 11, 2014, 6:55 p.m. UTC | #1
On Thu, Jan 09, 2014 at 12:07:25PM +0100, Jean-Francois Moine wrote:
> This patch reduces the number of I2C exchanges by setting many bits in
> one write and removing a useless write.
> 
> Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> ---
>  drivers/gpu/drm/i2c/tda998x_drv.c   | 10 ++++------
>  1 file changed, 4 insertions(+), 6 deletions(-)
> 
> diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
> index 6b4f6d2..d3b3f3a 100644
> --- a/drivers/gpu/drm/i2c/tda998x_drv.c
> +++ b/drivers/gpu/drm/i2c/tda998x_drv.c
> @@ -751,10 +751,10 @@ tda998x_configure_audio(struct tda998x_priv *priv,
>  	}
>  
>  	reg_write(priv, REG_AIP_CLKSEL, clksel_aip);
> -	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
> +	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |

This patch clearly hasn't even been build tested, so I doubt there's
much point reviewing this or the following patches.  From a quick scan
of the following patches, this never got fixed so the following patches
can't have been build tested either.

Thanks.
Jean-Francois Moine Jan. 12, 2014, 8:22 a.m. UTC | #2
On Sat, 11 Jan 2014 18:55:09 +0000
Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:

> On Thu, Jan 09, 2014 at 12:07:25PM +0100, Jean-Francois Moine wrote:
> > This patch reduces the number of I2C exchanges by setting many bits in
> > one write and removing a useless write.
> > 
> > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > ---
> >  drivers/gpu/drm/i2c/tda998x_drv.c   | 10 ++++------
> >  1 file changed, 4 insertions(+), 6 deletions(-)
> > 
> > diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
> > index 6b4f6d2..d3b3f3a 100644
> > --- a/drivers/gpu/drm/i2c/tda998x_drv.c
> > +++ b/drivers/gpu/drm/i2c/tda998x_drv.c
> > @@ -751,10 +751,10 @@ tda998x_configure_audio(struct tda998x_priv *priv,
> >  	}
> >  
> >  	reg_write(priv, REG_AIP_CLKSEL, clksel_aip);
> > -	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
> > +	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
> 
> This patch clearly hasn't even been build tested, so I doubt there's
> much point reviewing this or the following patches.  From a quick scan
> of the following patches, this never got fixed so the following patches
> can't have been build tested either.

I don't see what can be the problem with this patch. It does not change
anything in the logic. About testing, it is applied to my Cubox kernel
for more than 4 months and everything works correctly.

I will move the following comment a bit upwards. Maybe the code will be
clearer.
Russell King - ARM Linux Jan. 12, 2014, 9:45 a.m. UTC | #3
On Sun, Jan 12, 2014 at 09:22:01AM +0100, Jean-Francois Moine wrote:
> On Sat, 11 Jan 2014 18:55:09 +0000
> Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:
> 
> > On Thu, Jan 09, 2014 at 12:07:25PM +0100, Jean-Francois Moine wrote:
> > > This patch reduces the number of I2C exchanges by setting many bits in
> > > one write and removing a useless write.
> > > 
> > > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > > ---
> > >  drivers/gpu/drm/i2c/tda998x_drv.c   | 10 ++++------
> > >  1 file changed, 4 insertions(+), 6 deletions(-)
> > > 
> > > diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
> > > index 6b4f6d2..d3b3f3a 100644
> > > --- a/drivers/gpu/drm/i2c/tda998x_drv.c
> > > +++ b/drivers/gpu/drm/i2c/tda998x_drv.c
> > > @@ -751,10 +751,10 @@ tda998x_configure_audio(struct tda998x_priv *priv,
> > >  	}
> > >  
> > >  	reg_write(priv, REG_AIP_CLKSEL, clksel_aip);
> > > -	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
> > > +	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
> > 
> > This patch clearly hasn't even been build tested, so I doubt there's
> > much point reviewing this or the following patches.  From a quick scan
> > of the following patches, this never got fixed so the following patches
> > can't have been build tested either.
> 
> I don't see what can be the problem with this patch. It does not change
> anything in the logic. About testing, it is applied to my Cubox kernel
> for more than 4 months and everything works correctly.
> 
> I will move the following comment a bit upwards. Maybe the code will be
> clearer.

You're replacing ");" with "|" here, which is not legal C.  Parenthesis
must be balanced and statements must be terminated.
Jean-Francois Moine Jan. 12, 2014, 10:14 a.m. UTC | #4
On Sun, 12 Jan 2014 09:45:33 +0000
Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:

> On Sun, Jan 12, 2014 at 09:22:01AM +0100, Jean-Francois Moine wrote:
> > On Sat, 11 Jan 2014 18:55:09 +0000
> > Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:
> > 
> > > On Thu, Jan 09, 2014 at 12:07:25PM +0100, Jean-Francois Moine wrote:
> > > > This patch reduces the number of I2C exchanges by setting many bits in
> > > > one write and removing a useless write.
> > > > 
> > > > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > > > ---
> > > >  drivers/gpu/drm/i2c/tda998x_drv.c   | 10 ++++------
> > > >  1 file changed, 4 insertions(+), 6 deletions(-)
> > > > 
> > > > diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
> > > > index 6b4f6d2..d3b3f3a 100644
> > > > --- a/drivers/gpu/drm/i2c/tda998x_drv.c
> > > > +++ b/drivers/gpu/drm/i2c/tda998x_drv.c
> > > > @@ -751,10 +751,10 @@ tda998x_configure_audio(struct tda998x_priv *priv,
> > > >  	}
> > > >  
> > > >  	reg_write(priv, REG_AIP_CLKSEL, clksel_aip);
> > > > -	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
> > > > +	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
> > > 
> > > This patch clearly hasn't even been build tested, so I doubt there's
> > > much point reviewing this or the following patches.  From a quick scan
> > > of the following patches, this never got fixed so the following patches
> > > can't have been build tested either.
> > 
> > I don't see what can be the problem with this patch. It does not change
> > anything in the logic. About testing, it is applied to my Cubox kernel
> > for more than 4 months and everything works correctly.
> > 
> > I will move the following comment a bit upwards. Maybe the code will be
> > clearer.
> 
> You're replacing ");" with "|" here, which is not legal C.  Parenthesis
> must be balanced and statements must be terminated.

!? they are:

-	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
+	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
 
 	/* Enable automatic CTS generation */
-	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_ACR_MAN);
+					AIP_CNTRL_0_ACR_MAN);

gives:

	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |

	/* Enable automatic CTS generation */
					AIP_CNTRL_0_ACR_MAN);
Russell King - ARM Linux Jan. 12, 2014, 10:37 a.m. UTC | #5
On Sun, Jan 12, 2014 at 11:14:20AM +0100, Jean-Francois Moine wrote:
> On Sun, 12 Jan 2014 09:45:33 +0000
> Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:
> 
> > On Sun, Jan 12, 2014 at 09:22:01AM +0100, Jean-Francois Moine wrote:
> > > On Sat, 11 Jan 2014 18:55:09 +0000
> > > Russell King - ARM Linux <linux@arm.linux.org.uk> wrote:
> > > 
> > > > On Thu, Jan 09, 2014 at 12:07:25PM +0100, Jean-Francois Moine wrote:
> > > > > This patch reduces the number of I2C exchanges by setting many bits in
> > > > > one write and removing a useless write.
> > > > > 
> > > > > Signed-off-by: Jean-Francois Moine <moinejf@free.fr>
> > > > > ---
> > > > >  drivers/gpu/drm/i2c/tda998x_drv.c   | 10 ++++------
> > > > >  1 file changed, 4 insertions(+), 6 deletions(-)
> > > > > 
> > > > > diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
> > > > > index 6b4f6d2..d3b3f3a 100644
> > > > > --- a/drivers/gpu/drm/i2c/tda998x_drv.c
> > > > > +++ b/drivers/gpu/drm/i2c/tda998x_drv.c
> > > > > @@ -751,10 +751,10 @@ tda998x_configure_audio(struct tda998x_priv *priv,
> > > > >  	}
> > > > >  
> > > > >  	reg_write(priv, REG_AIP_CLKSEL, clksel_aip);
> > > > > -	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
> > > > > +	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
> > > > 
> > > > This patch clearly hasn't even been build tested, so I doubt there's
> > > > much point reviewing this or the following patches.  From a quick scan
> > > > of the following patches, this never got fixed so the following patches
> > > > can't have been build tested either.
> > > 
> > > I don't see what can be the problem with this patch. It does not change
> > > anything in the logic. About testing, it is applied to my Cubox kernel
> > > for more than 4 months and everything works correctly.
> > > 
> > > I will move the following comment a bit upwards. Maybe the code will be
> > > clearer.
> > 
> > You're replacing ");" with "|" here, which is not legal C.  Parenthesis
> > must be balanced and statements must be terminated.
> 
> !? they are:
> 
> -	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
> +	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
>  
>  	/* Enable automatic CTS generation */
> -	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_ACR_MAN);
> +					AIP_CNTRL_0_ACR_MAN);
> 
> gives:
> 
> 	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
> 
> 	/* Enable automatic CTS generation */
> 					AIP_CNTRL_0_ACR_MAN);

Yuck, that's absolutely horrid.  No wonder I mis-read it.  NAK for bad
and confusing style.
diff mbox

Patch

diff --git a/drivers/gpu/drm/i2c/tda998x_drv.c b/drivers/gpu/drm/i2c/tda998x_drv.c
index 6b4f6d2..d3b3f3a 100644
--- a/drivers/gpu/drm/i2c/tda998x_drv.c
+++ b/drivers/gpu/drm/i2c/tda998x_drv.c
@@ -751,10 +751,10 @@  tda998x_configure_audio(struct tda998x_priv *priv,
 	}
 
 	reg_write(priv, REG_AIP_CLKSEL, clksel_aip);
-	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT);
+	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_LAYOUT |
 
 	/* Enable automatic CTS generation */
-	reg_clear(priv, REG_AIP_CNTRL_0, AIP_CNTRL_0_ACR_MAN);
+					AIP_CNTRL_0_ACR_MAN);
 	reg_write(priv, REG_CTS_N, cts_n);
 
 	/*
@@ -1001,10 +1001,10 @@  tda998x_encoder_mode_set(struct drm_encoder *encoder,
 	reg_write(priv, REG_VIP_CNTRL_5, VIP_CNTRL_5_SP_CNT(0));
 	reg_write(priv, REG_VIP_CNTRL_4, VIP_CNTRL_4_BLANKIT(0) |
 			VIP_CNTRL_4_BLC(0));
-	reg_clear(priv, REG_PLL_SERIAL_3, PLL_SERIAL_3_SRL_CCIR);
 
 	reg_clear(priv, REG_PLL_SERIAL_1, PLL_SERIAL_1_SRL_MAN_IZ);
-	reg_clear(priv, REG_PLL_SERIAL_3, PLL_SERIAL_3_SRL_DE);
+	reg_clear(priv, REG_PLL_SERIAL_3, PLL_SERIAL_3_SRL_CCIR |
+					  PLL_SERIAL_3_SRL_DE);
 	reg_write(priv, REG_SERIALIZER, 0);
 
 	/* video quantization range = 0: full, 1: RGB/YUV, 2: YUV */
@@ -1026,8 +1026,6 @@  tda998x_encoder_mode_set(struct drm_encoder *encoder,
 	/* set BIAS tmds value: */
 	reg_write(priv, REG_ANA_GENERAL, 0x09);
 
-	reg_write(priv, REG_TBG_CNTRL_0, 0);
-
 	/*
 	 * Sync on rising HSYNC/VSYNC
 	 */