Re: [Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Chris Wilson]

On Mon, 17 Oct 2011 11:12:29 -0200, Eugeni Dodonov  
wrote:
> This allows to avoid talking to a non-existent bus repeatedly until we
> finally timeout. The non-existent bus is signalled by -ENXIO error,
> provided by i2c_algo_bit:bit_doAddress call.
> 
> As the advantage of such change, all the other routines which use
> drm_get_edid would benefit for this timeout.
> 
> This change should fix
> https://bugs.freedesktop.org/show_bug.cgi?id=41059 and improve overall
> edid detection timing by 10-30% in most cases, and by a much larger margin
> in case of phantom outputs.
> 
> Signed-off-by: Eugeni Dodonov 

Looks like we have reached the conclusion that this simple patch is the
least likely to cause problems and easiest to fix if it does. :)
Reviewed-by: Chris Wilson 
-Chris

-- 
Chris Wilson, Intel Open Source Technology Centre
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Eugeni Dodonov]

On Mon, Oct 17, 2011 at 20:41, Keith Packard  wrote:

> On Mon, 17 Oct 2011 19:07:51 -0200, Eugeni Dodonov 
> wrote:
>
> > From what I've checked, the other return error value in this context
> could
> > be -EREMOTEIO, which could be caused by transmission error so it should
> be
> > retried.
>
> Oh, there's -ENOMEM, -EINVAL and probably a few others down in the
> bowels of the kernel i2c bits. Starting with the obvious (ENXIO) seems
> safest to me.
>

Yes, of course, but I was referring to the values which could be returned
through the i2c-algo-bit call used in this edid detection call.

-- 
Eugeni Dodonov
 
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Keith Packard]

On Mon, 17 Oct 2011 19:07:51 -0200, Eugeni Dodonov  wrote:

> From what I've checked, the other return error value in this context could
> be -EREMOTEIO, which could be caused by transmission error so it should be
> retried.

Oh, there's -ENOMEM, -EINVAL and probably a few others down in the
bowels of the kernel i2c bits. Starting with the obvious (ENXIO) seems
safest to me.

-- 
keith.pack...@intel.com


pgpmzcIOhOLho.pgp
Description: PGP signature
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Eugeni Dodonov]

On Mon, Oct 17, 2011 at 18:41, Keith Packard  wrote:

> On Mon, 17 Oct 2011 11:12:29 -0200, Eugeni Dodonov <
> eugeni.dodo...@intel.com> wrote:
>
> > + if (ret == -ENXIO) {
> > + DRM_DEBUG_KMS("drm: skipping non-existent adapter
> %s\n",
> > + adapter->name);
> > + break;
> > + }
>
> This seems good to me; are there additional error values which should
> also be considered fatal and not subject to retry?
>

>From what I've checked, the other return error value in this context could
be -EREMOTEIO, which could be caused by transmission error so it should be
retried.

-- 
Eugeni Dodonov

___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Keith Packard]

On Mon, 17 Oct 2011 11:12:29 -0200, Eugeni Dodonov  
wrote:

> + if (ret == -ENXIO) {
> + DRM_DEBUG_KMS("drm: skipping non-existent adapter %s\n",
> + adapter->name);
> + break;
> + }

This seems good to me; are there additional error values which should
also be considered fatal and not subject to retry?

Reviewed-by: Keith Packard 

-keith


pgpejJTI3U2AP.pgp
Description: PGP signature
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

[Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Eugeni Dodonov]

This allows to avoid talking to a non-existent bus repeatedly until we
finally timeout. The non-existent bus is signalled by -ENXIO error,
provided by i2c_algo_bit:bit_doAddress call.

As the advantage of such change, all the other routines which use
drm_get_edid would benefit for this timeout.

This change should fix
https://bugs.freedesktop.org/show_bug.cgi?id=41059 and improve overall
edid detection timing by 10-30% in most cases, and by a much larger margin
in case of phantom outputs.

Signed-off-by: Eugeni Dodonov 
---
 drivers/gpu/drm/drm_edid.c |5 +
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 7425e5c..1bca6d7 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -265,6 +265,11 @@ drm_do_probe_ddc_edid(struct i2c_adapter *adapter, 
unsigned char *buf,
}
};
ret = i2c_transfer(adapter, msgs, 2);
+   if (ret == -ENXIO) {
+   DRM_DEBUG_KMS("drm: skipping non-existent adapter %s\n",
+   adapter->name);
+   break;
+   }
} while (ret != 2 && --retries);
 
return ret == 2 ? 0 : -1;
-- 
1.7.7

___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

[Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Eugeni Dodonov]

This allows to avoid talking to a non-existent bus repeatedly until we
finally timeout. The non-existent bus is signalled by -ENXIO error,
provided by i2c_algo_bit:bit_doAddress call.

As the advantage of such change, all the other routines which use
drm_get_edid would benefit for this timeout.

As the disadvantage comes the fact that I only tested it on Intel cards,
so I am not sure whether it would work on nouveau and radeon.

This change should potentially fix
https://bugs.freedesktop.org/show_bug.cgi?id=41059

v2: change printk level to KERN_DEBUG to avoid filling up dmesg

Signed-off-by: Eugeni Dodonov 
---
 drivers/gpu/drm/drm_edid.c |5 +
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 7425e5c..5ed34f2 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -265,6 +265,11 @@ drm_do_probe_ddc_edid(struct i2c_adapter *adapter, 
unsigned char *buf,
}
};
ret = i2c_transfer(adapter, msgs, 2);
+   if (ret == -ENXIO) {
+   printk(KERN_DEBUG "drm: skipping non-existent adapter 
%s\n",
+   adapter->name);
+   break;
+   }
} while (ret != 2 && --retries);
 
return ret == 2 ? 0 : -1;
-- 
1.7.6.4

___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Eugeni Dodonov]

On Fri, Oct 7, 2011 at 11:08, Jesse Barnes  wrote:

> On Thu,  6 Oct 2011 20:30:35 -0300
> Eugeni Dodonov  wrote:
>
> > From: Eugeni Dodonov 
> >
> > This allows to avoid talking to a non-existent bus repeatedly until we
> > finally timeout. The non-existent bus is signalled by -ENXIO error,
> > provided by i2c_algo_bit:bit_doAddress call.
> >
> > As the advantage of such change, all the other routines which use
> > drm_get_edid would benefit for this timeout.
> >
> > As the disadvantage comes the fact that I only tested it on Intel
> > cards, so I am not sure whether it would work on nouveau and radeon.
> >
> > This change should potentially fix
> > https://bugs.freedesktop.org/show_bug.cgi?id=41059
>
> I think I like this, assuming i2c doesn't lie to us.  But won't we spam
> the log quite a bit?  We do detection a lot because userspace often
> polls it and performs detection at app startup a lot.
>

This extra logging is there just for making it easy to see when the outputs
come and go. For the final version, we could use a KERN_DEBUG instead.

-- 
Eugeni Dodonov

___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

Re: [Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Jesse Barnes]

On Thu,  6 Oct 2011 20:30:35 -0300
Eugeni Dodonov  wrote:

> From: Eugeni Dodonov 
> 
> This allows to avoid talking to a non-existent bus repeatedly until we
> finally timeout. The non-existent bus is signalled by -ENXIO error,
> provided by i2c_algo_bit:bit_doAddress call.
> 
> As the advantage of such change, all the other routines which use
> drm_get_edid would benefit for this timeout.
> 
> As the disadvantage comes the fact that I only tested it on Intel
> cards, so I am not sure whether it would work on nouveau and radeon.
> 
> This change should potentially fix
> https://bugs.freedesktop.org/show_bug.cgi?id=41059

I think I like this, assuming i2c doesn't lie to us.  But won't we spam
the log quite a bit?  We do detection a lot because userspace often
polls it and performs detection at app startup a lot.

Jesse
___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx

[Intel-gfx] [PATCH 1/2] Give up on edid retries when i2c tells us that bus is not there

[Eugeni Dodonov]

From: Eugeni Dodonov 

This allows to avoid talking to a non-existent bus repeatedly until we
finally timeout. The non-existent bus is signalled by -ENXIO error,
provided by i2c_algo_bit:bit_doAddress call.

As the advantage of such change, all the other routines which use
drm_get_edid would benefit for this timeout.

As the disadvantage comes the fact that I only tested it on Intel cards,
so I am not sure whether it would work on nouveau and radeon.

This change should potentially fix
https://bugs.freedesktop.org/show_bug.cgi?id=41059

Signed-off-by: Eugeni Dodonov 
---
 drivers/gpu/drm/drm_edid.c |5 +
 1 files changed, 5 insertions(+), 0 deletions(-)

diff --git a/drivers/gpu/drm/drm_edid.c b/drivers/gpu/drm/drm_edid.c
index 7425e5c..475eff3 100644
--- a/drivers/gpu/drm/drm_edid.c
+++ b/drivers/gpu/drm/drm_edid.c
@@ -265,6 +265,11 @@ drm_do_probe_ddc_edid(struct i2c_adapter *adapter, 
unsigned char *buf,
}
};
ret = i2c_transfer(adapter, msgs, 2);
+   if (ret == -ENXIO) {
+   printk(KERN_WARNING "drm: i2c told us that device %s is 
not there\n",
+   adapter->name);
+   break;
+   }
} while (ret != 2 && --retries);
 
return ret == 2 ? 0 : -1;
-- 
1.7.6.3

___
Intel-gfx mailing list
Intel-gfx@lists.freedesktop.org
http://lists.freedesktop.org/mailman/listinfo/intel-gfx