So if the OMAP_DSS_DPI option is enabled, and the OMAP2_DSS_DSI option is not, you get the following build error with Nico's tree:
drivers/video/omap2/dss/dpi.c: In function 'dpi_set_dsi_clk': drivers/video/omap2/dss/dpi.c:61: error: implicit declaration of function 'dsi_pll_calc_clock_div_pck' drivers/video/omap2/dss/dpi.c:66: error: implicit declaration of function 'dsi_pll_set_clock_div' drivers/video/omap2/dss/dpi.c: In function 'omapdss_dpi_display_enable': drivers/video/omap2/dss/dpi.c:192: error: implicit declaration of function 'dsi_pll_init' drivers/video/omap2/dss/dpi.c:209: error: implicit declaration of function 'dsi_pll_uninit' make[4]: *** [drivers/video/omap2/dss/dpi.o] Error 1
This was reported to the mailing list almost two weeks ago, and still is unfixed.
The following hack works around the issue by making sure DSI is enabled if DPI is enabled. Although I suspect this fix would still break if OMAP3 was disabled, but no worse then it already does today.
Nico: Hopefully a proper fix is on its way, but in the meantime, please consider merging this.
thanks -john
Signed-off-by: John Stultz john.stultz@linaro.org
diff --git a/drivers/video/omap2/dss/Kconfig b/drivers/video/omap2/dss/Kconfig index 7a49a75..ae472a1 100644 --- a/drivers/video/omap2/dss/Kconfig +++ b/drivers/video/omap2/dss/Kconfig @@ -38,6 +38,7 @@ config OMAP2_DSS_COLLECT_IRQ_STATS
config OMAP2_DSS_DPI bool "DPI support" + select OMAP2_DSS_DSI default y help DPI Interface. This is the Parallel Display Interface.
On Tue, 14 Jun 2011, John Stultz wrote:
So if the OMAP_DSS_DPI option is enabled, and the OMAP2_DSS_DSI option is not, you get the following build error with Nico's tree:
drivers/video/omap2/dss/dpi.c: In function 'dpi_set_dsi_clk': drivers/video/omap2/dss/dpi.c:61: error: implicit declaration of function 'dsi_pll_calc_clock_div_pck' drivers/video/omap2/dss/dpi.c:66: error: implicit declaration of function 'dsi_pll_set_clock_div' drivers/video/omap2/dss/dpi.c: In function 'omapdss_dpi_display_enable': drivers/video/omap2/dss/dpi.c:192: error: implicit declaration of function 'dsi_pll_init' drivers/video/omap2/dss/dpi.c:209: error: implicit declaration of function 'dsi_pll_uninit' make[4]: *** [drivers/video/omap2/dss/dpi.o] Error 1
This was reported to the mailing list almost two weeks ago, and still is unfixed.
The following hack works around the issue by making sure DSI is enabled if DPI is enabled. Although I suspect this fix would still break if OMAP3 was disabled, but no worse then it already does today.
Nico: Hopefully a proper fix is on its way, but in the meantime, please consider merging this.
Given this is 1) a hack, and 2) rather trivial to apply separately, I'd prefer not to merge this so to keep the incentive for a real fix to remain alive.
Nicolas
On 06/14/2011 07:42 PM, Somebody in the thread at some point said:
So if the OMAP_DSS_DPI option is enabled, and the OMAP2_DSS_DSI option is not, you get the following build error with Nico's tree:
drivers/video/omap2/dss/dpi.c: In function 'dpi_set_dsi_clk': drivers/video/omap2/dss/dpi.c:61: error: implicit declaration of function 'dsi_pll_calc_clock_div_pck' drivers/video/omap2/dss/dpi.c:66: error: implicit declaration of function 'dsi_pll_set_clock_div' drivers/video/omap2/dss/dpi.c: In function 'omapdss_dpi_display_enable': drivers/video/omap2/dss/dpi.c:192: error: implicit declaration of function 'dsi_pll_init' drivers/video/omap2/dss/dpi.c:209: error: implicit declaration of function 'dsi_pll_uninit' make[4]: *** [drivers/video/omap2/dss/dpi.o] Error 1
This was reported to the mailing list almost two weeks ago, and still is unfixed.
When HDMI works on .39, you will indeed need DSI.
The following hack works around the issue by making sure DSI is enabled if DPI is enabled. Although I suspect this fix would still break if OMAP3 was disabled, but no worse then it already does today.
In linux-omap on kernel.org, they fixed the OMAP3-only dep
config OMAP2_DSS_DSI bool "DSI support" depends on ARCH_OMAP3 || ARCH_OMAP4
Nico: Hopefully a proper fix is on its way, but in the meantime, please consider merging this.
Again, in linux-omap on kernel.org, the put static inline versions of the missing functions in dss.h and fixed the problem that way.
config OMAP2_DSS_DPI bool "DPI support"
- select OMAP2_DSS_DSI
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
-Andy
On 06/14/2011 09:29 PM, Somebody in the thread at some point said:
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
Jassi has indeed fixed the HDMI on .39 today in time for tomorrow, you can get it here
http://git.linaro.org/gitweb?p=people/andygreen/kernel-tilt.git%3Ba=shortlog...
-Andy
On Wed, Jun 22, 2011 at 3:47 PM, Andy Green andy.green@linaro.org wrote:
On 06/14/2011 09:29 PM, Somebody in the thread at some point said:
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
Jassi has indeed fixed the HDMI on .39 today in time for tomorrow, you can get it here
http://git.linaro.org/gitweb?p=people/andygreen/kernel-tilt.git%3Ba=shortlog...
What CONFIG_'s need to be turned on to make it work? I need these for the packaged kernel.
On Wed, Jun 22, 2011 at 3:52 PM, John Rigby john.rigby@linaro.org wrote:
On Wed, Jun 22, 2011 at 3:47 PM, Andy Green andy.green@linaro.org wrote:
On 06/14/2011 09:29 PM, Somebody in the thread at some point said:
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
Jassi has indeed fixed the HDMI on .39 today in time for tomorrow, you can get it here
http://git.linaro.org/gitweb?p=people/andygreen/kernel-tilt.git%3Ba=shortlog...
What CONFIG_'s need to be turned on to make it work? I need these for the packaged kernel.
Nevermind I see this is the tilt kernel, I thought it was a pull request for npitre.
So for tilt once you have tagged it please send a config to me so I can do a packaged version for inclusion in hwpacks.
Thanks, John
On 06/22/2011 10:55 PM, Somebody in the thread at some point said:
On Wed, Jun 22, 2011 at 3:52 PM, John Rigbyjohn.rigby@linaro.org wrote:
On Wed, Jun 22, 2011 at 3:47 PM, Andy Greenandy.green@linaro.org wrote:
On 06/14/2011 09:29 PM, Somebody in the thread at some point said:
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
Jassi has indeed fixed the HDMI on .39 today in time for tomorrow, you can get it here
http://git.linaro.org/gitweb?p=people/andygreen/kernel-tilt.git%3Ba=shortlog...
What CONFIG_'s need to be turned on to make it work? I need these for the packaged kernel.
Nevermind I see this is the tilt kernel, I thought it was a pull request for npitre.
So for tilt once you have tagged it please send a config to me so I can do a packaged version for inclusion in hwpacks.
All the branches that are modified in that repo maintain their own config in the tree directly, so the config patchlevel matches whatever patchset is in there.
It's "omap4_defconfig" for vanilla kernels and "android_omap4_defconfig" for the androidized branches.
If you do ( (c) npitre ^^)
make ARCH=arm omap4_defconfig
it will spit out a .config file from the tree that should always be suitable for what's working in the tree.
-Andy
On 06/22/2011 10:52 PM, Somebody in the thread at some point said:
On Wed, Jun 22, 2011 at 3:47 PM, Andy Greenandy.green@linaro.org wrote:
On 06/14/2011 09:29 PM, Somebody in the thread at some point said:
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
Jassi has indeed fixed the HDMI on .39 today in time for tomorrow, you can get it here
http://git.linaro.org/gitweb?p=people/andygreen/kernel-tilt.git%3Ba=shortlog...
What CONFIG_'s need to be turned on to make it work? I need these for the packaged kernel.
Inside the tree it maintains "omap4_defconfig" which is the one to use.
I added a patch series modifying "omap2plus_defconfig" so that it is actually usable for Panda, but I was unable to get BT working on it for example whereas it does work on omap4_defconfig.
-Andy
On Wed, 2011-06-22 at 22:47 +0100, Andy Green wrote:
On 06/14/2011 09:29 PM, Somebody in the thread at some point said:
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
Jassi has indeed fixed the HDMI on .39 today in time for tomorrow, you can get it here
http://git.linaro.org/gitweb?p=people/andygreen/kernel-tilt.git%3Ba=shortlog...
Cool!
Is this something that is planned to be pushed to Nico? I'd rather not carry any non-android specific changes in my tree.
thanks -john
On 06/22/2011 10:55 PM, Somebody in the thread at some point said:
On Wed, 2011-06-22 at 22:47 +0100, Andy Green wrote:
On 06/14/2011 09:29 PM, Somebody in the thread at some point said:
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
Jassi has indeed fixed the HDMI on .39 today in time for tomorrow, you can get it here
http://git.linaro.org/gitweb?p=people/andygreen/kernel-tilt.git%3Ba=shortlog...
Cool!
Is this something that is planned to be pushed to Nico? I'd rather not carry any non-android specific changes in my tree.
Well although the earlier block of DSS patches are from linux-omap and probably safe (up to f29733759ced243ecd3efb3a4937ca59b5cbf5e7), Jassi's series at the end vs OMAP3 / Beagle are an unknown again; they're only tested on Panda so far.
-Andy
On Wed, 22 Jun 2011, Andy Green wrote:
On 06/22/2011 10:55 PM, Somebody in the thread at some point said:
On Wed, 2011-06-22 at 22:47 +0100, Andy Green wrote:
On 06/14/2011 09:29 PM, Somebody in the thread at some point said:
I think what we should do is make an effort on uplevelling from linux-omap dss to solve that with upstreamfulness, if there's any broken deps coming then we can solve them here and upstream at the same time.
In addition it might help the HDMI patches with what's gone on there in the meantime.
Unfortunately as you know from other mail I will be looking at PVR / SGX mainly tomorrow, but Jassi has offered to help with this dss stuff when he's finished what he's working on in the next day or two.
Jassi has indeed fixed the HDMI on .39 today in time for tomorrow, you can get it here
http://git.linaro.org/gitweb?p=people/andygreen/kernel-tilt.git%3Ba=shortlog...
Cool!
Is this something that is planned to be pushed to Nico? I'd rather not carry any non-android specific changes in my tree.
Well although the earlier block of DSS patches are from linux-omap and probably safe (up to f29733759ced243ecd3efb3a4937ca59b5cbf5e7), Jassi's series at the end vs OMAP3 / Beagle are an unknown again; they're only tested on Panda so far.
I merged it all. Let's see how it goes.
The linaro-2.6.39 tree is effectively frozen wrt new features from now until June 30th. We therefore have one week to make it bug free! :-)
Nicolas
linaro-kernel@lists.linaro.org
-
Andy Green -
John Rigby -
John Stultz -
Nicolas Pitre