From: Tomi Valkeinen <hidden> Date: 2012-02-29 08:48:20
Hi,
Two small fixes for omapdss.
Florian, if the patches are ok, I guess it's easier if you just apply these
patches from email, instead of me sending a pull request?
Tomi
Tomi Valkeinen (2):
OMAPDSS: panel-dvi: Add Kconfig dependency on I2C
OMAPDSS: APPLY: make ovl_enable/disable synchronous
drivers/video/omap2/displays/Kconfig | 2 +-
drivers/video/omap2/dss/apply.c | 6 ++++++
2 files changed, 7 insertions(+), 1 deletions(-)
--
1.7.4.1
From: Tomi Valkeinen <hidden> Date: 2012-02-29 08:48:21
panel-dvi uses i2c, but the Kconfig didn't have dependency on I2C. Add
it.
Signed-off-by: Tomi Valkeinen <redacted>
---
drivers/video/omap2/displays/Kconfig | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
From: Tomi Valkeinen <hidden> Date: 2012-02-29 08:48:22
ovl->enable/disable are meant to be synchronous so that they can handle
the configuration of fifo sizes. The current kernel doesn't configure
fifo sizes yet, and so the code doesn't need to block to function (from
omapdss driver's perspective).
However, for the users of omapdss a non-blocking ovl->disable is
confusing, because they don't know when the memory area is not used
any more.
Furthermore, when the fifo size configuration is added in the next merge
window, the change from non-blocking to blocking could cause side
effects to the users of omapdss. So by making the functions block
already will keep them behaving in the same manner.
And, while not the main purpose of this patch, this will also remove the
compile warning:
drivers/video/omap2/dss/apply.c:350: warning:
'wait_pending_extra_info_updates' defined but not used
Signed-off-by: Tomi Valkeinen <redacted>
---
drivers/video/omap2/dss/apply.c | 6 ++++++
1 files changed, 6 insertions(+), 0 deletions(-)
@@ -1276,6 +1276,9 @@ int dss_ovl_enable(struct omap_overlay *ovl)spin_unlock_irqrestore(&data_lock,flags);+/* wait for overlay to be enabled */+wait_pending_extra_info_updates();+mutex_unlock(&apply_lock);return0;
@@ -1313,6 +1316,9 @@ int dss_ovl_disable(struct omap_overlay *ovl)spin_unlock_irqrestore(&data_lock,flags);+/* wait for the overlay to be disabled */+wait_pending_extra_info_updates();+mutex_unlock(&apply_lock);return0;
Hi Tomi,
On 02/29/2012 08:48 AM, Tomi Valkeinen wrote:
Hi,
Two small fixes for omapdss.
Florian, if the patches are ok, I guess it's easier if you just apply these
patches from email, instead of me sending a pull request?
Yes, that's correct as those are only a few patches and as we are already in
late -rc's (and therefore I guess, I should take a better look at what I apply).
I'll add some comments to your patches in separate mails.
Best regards,
Florian Tobias Schandinat
Tomi
Tomi Valkeinen (2):
OMAPDSS: panel-dvi: Add Kconfig dependency on I2C
OMAPDSS: APPLY: make ovl_enable/disable synchronous
drivers/video/omap2/displays/Kconfig | 2 +-
drivers/video/omap2/dss/apply.c | 6 ++++++
2 files changed, 7 insertions(+), 1 deletions(-)
panel-dvi uses i2c, but the Kconfig didn't have dependency on I2C. Add
it.
Signed-off-by: Tomi Valkeinen <redacted>
---
drivers/video/omap2/displays/Kconfig | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
It's just a matter of taste, but are you sure you want to "depend" on it and not
"select" it? Other drivers tend to use select for I2C, for me it doesn't really
matter.
Best regards,
Florian Tobias Schandinat
help
Driver for external monitors, connected via DVI. The driver uses i2c
to read EDID information from the monitor.
ovl->enable/disable are meant to be synchronous so that they can handle
the configuration of fifo sizes. The current kernel doesn't configure
fifo sizes yet, and so the code doesn't need to block to function (from
omapdss driver's perspective).
However, for the users of omapdss a non-blocking ovl->disable is
confusing, because they don't know when the memory area is not used
any more.
Furthermore, when the fifo size configuration is added in the next merge
window, the change from non-blocking to blocking could cause side
effects to the users of omapdss. So by making the functions block
already will keep them behaving in the same manner.
Is there any difference to doing it now?
I agree that this should be fixed but if we can't avoid breaking users I'd
prefer to break them in a merge window not in late rc stage. Or did we introduce
these functions just in the last merge window?
Best regards,
Florian Tobias Schandinat
quoted hunk
And, while not the main purpose of this patch, this will also remove the
compile warning:
drivers/video/omap2/dss/apply.c:350: warning:
'wait_pending_extra_info_updates' defined but not used
Signed-off-by: Tomi Valkeinen <redacted>
---
drivers/video/omap2/dss/apply.c | 6 ++++++
1 files changed, 6 insertions(+), 0 deletions(-)
@@ -1276,6 +1276,9 @@ int dss_ovl_enable(struct omap_overlay *ovl)spin_unlock_irqrestore(&data_lock,flags);+/* wait for overlay to be enabled */+wait_pending_extra_info_updates();+mutex_unlock(&apply_lock);return0;
@@ -1313,6 +1316,9 @@ int dss_ovl_disable(struct omap_overlay *ovl)spin_unlock_irqrestore(&data_lock,flags);+/* wait for the overlay to be disabled */+wait_pending_extra_info_updates();+mutex_unlock(&apply_lock);return0;
From: Tomi Valkeinen <hidden> Date: 2012-02-29 10:21:42
On Wed, 2012-02-29 at 10:03 +0000, Florian Tobias Schandinat wrote:
On 02/29/2012 08:48 AM, Tomi Valkeinen wrote:
quoted
panel-dvi uses i2c, but the Kconfig didn't have dependency on I2C. Add
it.
Signed-off-by: Tomi Valkeinen <redacted>
---
drivers/video/omap2/displays/Kconfig | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
It's just a matter of taste, but are you sure you want to "depend" on it and not
"select" it? Other drivers tend to use select for I2C, for me it doesn't really
matter.
Well, I'd like to "select", but I don't think that's correct. From
Documentation/kbuild/kconfig-language.txt:
In general use select only for non-visible symbols
(no prompts anywhere) and for symbols with no dependencies.
But I do see quite many selects for I2C, so I'm not sure if all those
are wrong, or has it just been decided that I2C is a valid target for
select.
Using depend is in line with the other panel drivers in the same
directory. I've been thinking about the same thing from time to time,
and I'd rather select I2C, SPI and BACKLIGHT_CLASS_DEVICE than use
depend.
But, for example, using select to BACKLIGHT_CLASS_DEVICE is broken: if I
change a panel driver to select BACKLIGHT_CLASS_DEVICE, but then I
manually disable CONFIG_BACKLIGHT_LCD_SUPPORT (which the
BACKLIGHT_CLASS_DEVICE depends on) from the kernel config, this leads to
BACKLIGHT_CLASS_DEVICE being enabled, but CONFIG_BACKLIGHT_LCD_SUPPORT
being disabled, which is clearly broken.
So... As I see it, depending is a bit awkward, but it's correct.
Tomi
From: Tomi Valkeinen <hidden> Date: 2012-02-29 10:30:38
On Wed, 2012-02-29 at 10:13 +0000, Florian Tobias Schandinat wrote:
On 02/29/2012 08:48 AM, Tomi Valkeinen wrote:
quoted
ovl->enable/disable are meant to be synchronous so that they can handle
the configuration of fifo sizes. The current kernel doesn't configure
fifo sizes yet, and so the code doesn't need to block to function (from
omapdss driver's perspective).
However, for the users of omapdss a non-blocking ovl->disable is
confusing, because they don't know when the memory area is not used
any more.
Furthermore, when the fifo size configuration is added in the next merge
window, the change from non-blocking to blocking could cause side
effects to the users of omapdss. So by making the functions block
already will keep them behaving in the same manner.
Is there any difference to doing it now?
I agree that this should be fixed but if we can't avoid breaking users I'd
prefer to break them in a merge window not in late rc stage. Or did we introduce
these functions just in the last merge window?
Yes, these were introduced in the merge window. And I explicitly said
the functions are blocking so that they can perform their job. And just
to clarify, the functions already use a mutex, so in that sense they are
blocking. They just don't currently wait until the HW has finished with
the overlay.
The problem was raised by Rob Clark, who's writing the omapdrm driver,
as he doesn't have a way to ensure that the overlay has been truly
disabled and the memory is is no longer in use.
(I forgot to cc him for the patch, adding him now).
Tomi
Hi Tomi,
On 02/29/2012 10:30 AM, Tomi Valkeinen wrote:
On Wed, 2012-02-29 at 10:13 +0000, Florian Tobias Schandinat wrote:
quoted
On 02/29/2012 08:48 AM, Tomi Valkeinen wrote:
quoted
ovl->enable/disable are meant to be synchronous so that they can handle
the configuration of fifo sizes. The current kernel doesn't configure
fifo sizes yet, and so the code doesn't need to block to function (from
omapdss driver's perspective).
However, for the users of omapdss a non-blocking ovl->disable is
confusing, because they don't know when the memory area is not used
any more.
Furthermore, when the fifo size configuration is added in the next merge
window, the change from non-blocking to blocking could cause side
effects to the users of omapdss. So by making the functions block
already will keep them behaving in the same manner.
Is there any difference to doing it now?
I agree that this should be fixed but if we can't avoid breaking users I'd
prefer to break them in a merge window not in late rc stage. Or did we introduce
these functions just in the last merge window?
Yes, these were introduced in the merge window. And I explicitly said
the functions are blocking so that they can perform their job. And just
to clarify, the functions already use a mutex, so in that sense they are
blocking. They just don't currently wait until the HW has finished with
the overlay.
okay, than I'll apply this patch as is. I was just worried about asking Linus to
pull something that is labled as "Breaks existing users" now, but that doesn't
look like an issue here.
Best regards,
Florian Tobias Schandinat
The problem was raised by Rob Clark, who's writing the omapdrm driver,
as he doesn't have a way to ensure that the overlay has been truly
disabled and the memory is is no longer in use.
(I forgot to cc him for the patch, adding him now).
Tomi
Hi Tomi,
On 02/29/2012 10:21 AM, Tomi Valkeinen wrote:
On Wed, 2012-02-29 at 10:03 +0000, Florian Tobias Schandinat wrote:
quoted
On 02/29/2012 08:48 AM, Tomi Valkeinen wrote:
quoted
panel-dvi uses i2c, but the Kconfig didn't have dependency on I2C. Add
it.
Signed-off-by: Tomi Valkeinen <redacted>
---
drivers/video/omap2/displays/Kconfig | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
It's just a matter of taste, but are you sure you want to "depend" on it and not
"select" it? Other drivers tend to use select for I2C, for me it doesn't really
matter.
Well, I'd like to "select", but I don't think that's correct. From
Documentation/kbuild/kconfig-language.txt:
In general use select only for non-visible symbols
(no prompts anywhere) and for symbols with no dependencies.
But I do see quite many selects for I2C, so I'm not sure if all those
are wrong, or has it just been decided that I2C is a valid target for
select.
I'd say that the above is only a guideline but not absolute rule. For I2C I'd
argue that I probably wouldn't even know that my hardware has it if I weren't
the one writing the driver using it.
But it's true that select should be used with care as it is much easier to get
the config messed up by using it.
Using depend is in line with the other panel drivers in the same
directory. I've been thinking about the same thing from time to time,
and I'd rather select I2C, SPI and BACKLIGHT_CLASS_DEVICE than use
depend.
But, for example, using select to BACKLIGHT_CLASS_DEVICE is broken: if I
change a panel driver to select BACKLIGHT_CLASS_DEVICE, but then I
manually disable CONFIG_BACKLIGHT_LCD_SUPPORT (which the
BACKLIGHT_CLASS_DEVICE depends on) from the kernel config, this leads to
BACKLIGHT_CLASS_DEVICE being enabled, but CONFIG_BACKLIGHT_LCD_SUPPORT
being disabled, which is clearly broken.
Yes, as far as I understand the problem here is that we don't have any mechanism
to handle transitive dependencies (probably because we don't want it). So if one
wants to do select something I think the right way to do it would be to inherit
all its select/depend statements in the option.
So... As I see it, depending is a bit awkward, but it's correct.
Yes, it certainly is correct.
Best regards,
Florian Tobias Schandinat
From: Rob Clark <hidden> Date: 2012-02-29 14:52:18
On Wed, Feb 29, 2012 at 4:48 AM, Florian Tobias Schandinat
[off-list ref] wrote:
Hi Tomi,
On 02/29/2012 10:30 AM, Tomi Valkeinen wrote:
quoted
On Wed, 2012-02-29 at 10:13 +0000, Florian Tobias Schandinat wrote:
quoted
On 02/29/2012 08:48 AM, Tomi Valkeinen wrote:
quoted
ovl->enable/disable are meant to be synchronous so that they can handle
the configuration of fifo sizes. The current kernel doesn't configure
fifo sizes yet, and so the code doesn't need to block to function (from
omapdss driver's perspective).
However, for the users of omapdss a non-blocking ovl->disable is
confusing, because they don't know when the memory area is not used
any more.
Furthermore, when the fifo size configuration is added in the next merge
window, the change from non-blocking to blocking could cause side
effects to the users of omapdss. So by making the functions block
already will keep them behaving in the same manner.
Is there any difference to doing it now?
I agree that this should be fixed but if we can't avoid breaking users I'd
prefer to break them in a merge window not in late rc stage. Or did we introduce
these functions just in the last merge window?
Yes, these were introduced in the merge window. And I explicitly said
the functions are blocking so that they can perform their job. And just
to clarify, the functions already use a mutex, so in that sense they are
blocking. They just don't currently wait until the HW has finished with
the overlay.
okay, than I'll apply this patch as is. I was just worried about asking Linus to
pull something that is labled as "Breaks existing users" now, but that doesn't
look like an issue here.
I don't expect this change should break any existing users.. I think
it is safe to call this a bug fix
BR,
-R
Best regards,
Florian Tobias Schandinat
quoted
The problem was raised by Rob Clark, who's writing the omapdrm driver,
as he doesn't have a way to ensure that the overlay has been truly
disabled and the memory is is no longer in use.
(I forgot to cc him for the patch, adding him now).
Tomi
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Florian, if the patches are ok, I guess it's easier if you just apply these
patches from email, instead of me sending a pull request?
Tomi
Tomi Valkeinen (2):
OMAPDSS: panel-dvi: Add Kconfig dependency on I2C
OMAPDSS: APPLY: make ovl_enable/disable synchronous
drivers/video/omap2/displays/Kconfig | 2 +-
drivers/video/omap2/dss/apply.c | 6 ++++++
2 files changed, 7 insertions(+), 1 deletions(-)