HDMI isn't probed on v5.15-rc1 on ULCB+KF

HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kuninori Morimoto]

Hi Laurent

I noticed that ULCB+KF board on v5.15-rc1 doesn't probe HDMI.
I git-bisect:ed and found the first issue.
It re-start works for me if I reverted this commit.
Could you please check it ?

commit d0f44e0dac29aca3f85d238dfbcd2e52450b38dc
Author: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
Date:   Mon May 25 07:05:26 2020 +0300

    drm: rcar-du: dw-hdmi: Set output port number
    
    Report the DT output port number in dw_hdmi_plat_data to connect to the
    next bridge in the dw-hdmi driver.
    
    Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
    Tested-by: Kieran Bingham <kieran.bingham@ideasonboard.com> # On V3U
    Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>



Thank you for your help !!

Best regards
---
Kuninori Morimoto

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Laurent Pinchart]

Hi Morimoto-san,

On Wed, Oct 06, 2021 at 12:05:15PM +0900, Kuninori Morimoto wrote:
> 
> Hi Laurent
> 
> I noticed that ULCB+KF board on v5.15-rc1 doesn't probe HDMI.

Is that the M3 or H3 ULCB ?

Is there any useful information in the kernel log ?

> I git-bisect:ed and found the first issue.
> It re-start works for me if I reverted this commit.
> Could you please check it ?
> 
> commit d0f44e0dac29aca3f85d238dfbcd2e52450b38dc
> Author: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> Date:   Mon May 25 07:05:26 2020 +0300
> 
>     drm: rcar-du: dw-hdmi: Set output port number
>     
>     Report the DT output port number in dw_hdmi_plat_data to connect to the
>     next bridge in the dw-hdmi driver.
>     
>     Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
>     Tested-by: Kieran Bingham <kieran.bingham@ideasonboard.com> # On V3U
>     Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>

I've tested v5.15-rc1 on a M3-N Salvator-XS, and can't reproduce the
issue. If your kernel log doesn't show anything useful, I'll test it on
my ULCB.

-- 
Regards,

Laurent Pinchart

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kuninori Morimoto]

Hi Laurent

Thank you for your reply

> > I noticed that ULCB+KF board on v5.15-rc1 doesn't probe HDMI.
>
> Is that the M3 or H3 ULCB ?

I'm using H3 ULCB + Kingfihser

> > commit d0f44e0dac29aca3f85d238dfbcd2e52450b38dc
> > Author: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > Date:   Mon May 25 07:05:26 2020 +0300
> >
> >     drm: rcar-du: dw-hdmi: Set output port number
> >
> >     Report the DT output port number in dw_hdmi_plat_data to connect to the
> >     next bridge in the dw-hdmi driver.
> >
> >     Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> >     Tested-by: Kieran Bingham <kieran.bingham@ideasonboard.com> # On V3U
> >     Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
>
> I've tested v5.15-rc1 on a M3-N Salvator-XS, and can't reproduce the
> issue. If your kernel log doesn't show anything useful, I'll test it on
> my ULCB.

I'm not sure what is related, but I noticed something.

This is OK case.
DRM failed ? rcar-dw-hdmi was probed
	...
	[    0.499485] gpio_rcar e6055800.gpio: driving 4 GPIOs
	[    0.511240] SuperH (H)SCI(F) driver initialized
	[    0.512316] e6e88000.serial: ttySC0 at MMIO 0xe6e88000 (irq = 123, base_baud = 0) is a scif
	[    1.465894] printk: console [ttySC0] enabled
=>	[    1.476935] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
=>	[    1.485684] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
=>	[    1.496248] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
	[    1.535173] loop: module loaded
	...

This is NG case.
DAM failed. rcar-dw-hdmi is not probed.

	...
	[    0.490286] gpio_rcar e6055800.gpio: driving 4 GPIOs
	[    0.501956] SuperH (H)SCI(F) driver initialized
	[    0.503027] e6e88000.serial: ttySC0 at MMIO 0xe6e88000 (irq = 123, base_baud = 0) is a scif
	[    1.432046] printk: console [ttySC0] enabled
=>	[    1.443177] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
	[    1.481426] loop: module loaded
	...

Thank you for your help !!

Best regards
---
Kuninori Morimoto

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Laurent Pinchart]

Hi Morimoto-san,

On Thu, Oct 07, 2021 at 01:50:18PM +0900, Kuninori Morimoto wrote:
> 
> Hi Laurent
> 
> Thank you for your reply
> 
> > > I noticed that ULCB+KF board on v5.15-rc1 doesn't probe HDMI.
> >
> > Is that the M3 or H3 ULCB ?
> 
> I'm using H3 ULCB + Kingfihser
> 
> > > commit d0f44e0dac29aca3f85d238dfbcd2e52450b38dc
> > > Author: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > > Date:   Mon May 25 07:05:26 2020 +0300
> > >
> > >     drm: rcar-du: dw-hdmi: Set output port number
> > >
> > >     Report the DT output port number in dw_hdmi_plat_data to connect to the
> > >     next bridge in the dw-hdmi driver.
> > >
> > >     Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > >     Tested-by: Kieran Bingham <kieran.bingham@ideasonboard.com> # On V3U
> > >     Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
> >
> > I've tested v5.15-rc1 on a M3-N Salvator-XS, and can't reproduce the
> > issue. If your kernel log doesn't show anything useful, I'll test it on
> > my ULCB.
> 
> I'm not sure what is related, but I noticed something.
> 
> This is OK case.
> DRM failed ? rcar-dw-hdmi was probed
> 	...
> 	[    0.499485] gpio_rcar e6055800.gpio: driving 4 GPIOs
> 	[    0.511240] SuperH (H)SCI(F) driver initialized
> 	[    0.512316] e6e88000.serial: ttySC0 at MMIO 0xe6e88000 (irq = 123, base_baud = 0) is a scif
> 	[    1.465894] printk: console [ttySC0] enabled
> =>	[    1.476935] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> =>	[    1.485684] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
> =>	[    1.496248] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
> 	[    1.535173] loop: module loaded
> 	...
> 
> This is NG case.
> DAM failed. rcar-dw-hdmi is not probed.
> 
> 	...
> 	[    0.490286] gpio_rcar e6055800.gpio: driving 4 GPIOs
> 	[    0.501956] SuperH (H)SCI(F) driver initialized
> 	[    0.503027] e6e88000.serial: ttySC0 at MMIO 0xe6e88000 (irq = 123, base_baud = 0) is a scif
> 	[    1.432046] printk: console [ttySC0] enabled
> =>	[    1.443177] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> 	[    1.481426] loop: module loaded
> 	...

Strange. I will test it with my H3 ULCB.

-- 
Regards,

Laurent Pinchart

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kuninori Morimoto]

Hi Laurent

> Strange. I will test it with my H3 ULCB.

Thanks.
I could noticed it via kernel log of sound,
because it needs HDMI-Sound.
You can find it almost last.

OK case
	...
	[    4.058402] ALSA device list:
=>	[    4.061416]   #0: rcar-sound
	...

NG case
	...
	[    4.026208] ALSA device list:                                                                                                                            
=>	[    4.029254]   No soundcards found.
	...

Thank you for your help !!

Best regards
---
Kuninori Morimoto

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Geert Uytterhoeven]

Hi Morimoto-san,

On Thu, Oct 7, 2021 at 6:55 AM Kuninori Morimoto
<kuninori.morimoto.gx@renesas.com> wrote:
> I'm not sure what is related, but I noticed something.
>
> This is OK case.
> DRM failed ? rcar-dw-hdmi was probed
>         ...
>         [    0.499485] gpio_rcar e6055800.gpio: driving 4 GPIOs
>         [    0.511240] SuperH (H)SCI(F) driver initialized
>         [    0.512316] e6e88000.serial: ttySC0 at MMIO 0xe6e88000 (irq = 123, base_baud = 0) is a scif
>         [    1.465894] printk: console [ttySC0] enabled
> =>      [    1.476935] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> =>      [    1.485684] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
> =>      [    1.496248] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
>         [    1.535173] loop: module loaded
>         ...
>
> This is NG case.
> DAM failed. rcar-dw-hdmi is not probed.
>
>         ...
>         [    0.490286] gpio_rcar e6055800.gpio: driving 4 GPIOs
>         [    0.501956] SuperH (H)SCI(F) driver initialized
>         [    0.503027] e6e88000.serial: ttySC0 at MMIO 0xe6e88000 (irq = 123, base_baud = 0) is a scif
>         [    1.432046] printk: console [ttySC0] enabled
> =>      [    1.443177] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
>         [    1.481426] loop: module loaded
>         ...

Note that in both cases feb00000.display failed to probe first
Is this an fw_devlink issue, due to HDMI linking to DU in DT?
Does it work with fw_devlink=permissive of fw_devlink=off?

Gr{oetje,eeting}s,

                        Geert


--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kuninori Morimoto]

Hi Laurent, again

About this topic, v5.15 was released, but it still
has "No soundcards" issue.
There is no dw-hdmi, thus no soundcards

	...
	[    1.431196] printk: console [ttySC0] enabled
=>	[    1.442432] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
	[    1.481179] loop: module loaded
	...
	[    9.568307] ALSA device list:
=>	[    9.571446]   No soundcards found.
	...

I re-start works for me if I reverted this patch

	--------------------------------
	commit d0f44e0dac29aca3f85d238dfbcd2e52450b38dc
	Author: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
	Date:   Mon May 25 07:05:26 2020 +0300

	    drm: rcar-du: dw-hdmi: Set output port number

	    Report the DT output port number in dw_hdmi_plat_data to connect to the
	    next bridge in the dw-hdmi driver.

	    Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
	    Tested-by: Kieran Bingham <kieran.bingham@ideasonboard.com> # On V3U
	    Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
	--------------------------------

	...
	[    1.431130] printk: console [ttySC0] enabled
=>	[    1.442219] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
=>	[    1.451050] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
=>	[    1.461560] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
	[    1.501043] loop: module loaded
	...
	[    9.180464] ALSA device list:
=>	[    9.183519]   #0: rcar-sound
	...

Thank you for your help !!

Best regards
---
Kuninori Morimoto

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kieran Bingham]

Hi Morimoto-san,

Laurent is away this week, so I've just tried to have a quick look
through this issue in his absence. I'm afraid I don't have a ULCB-H3 to
test locally, so I'm following the code standalone at the moment.

Quoting Geert Uytterhoeven (2021-10-07 08:24:37)
> Hi Morimoto-san,
> 
> On Thu, Oct 7, 2021 at 6:55 AM Kuninori Morimoto
> <kuninori.morimoto.gx@renesas.com> wrote:
> > I'm not sure what is related, but I noticed something.
> >
> > This is OK case.
> > DRM failed ? rcar-dw-hdmi was probed
> >         ...
> >         [    0.499485] gpio_rcar e6055800.gpio: driving 4 GPIOs
> >         [    0.511240] SuperH (H)SCI(F) driver initialized
> >         [    0.512316] e6e88000.serial: ttySC0 at MMIO 0xe6e88000 (irq = 123, base_baud = 0) is a scif
> >         [    1.465894] printk: console [ttySC0] enabled
> > =>      [    1.476935] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)

This comes from calling rcar_du_modeset_init() ... I'll dig down to see
if there's a clear (single) path leading to -6 / -ENXIO.


> > =>      [    1.485684] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
> > =>      [    1.496248] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
> >         [    1.535173] loop: module loaded
> >         ...
> >
> > This is NG case.
> > DAM failed. rcar-dw-hdmi is not probed.

Was something changed to make this happen between this and the above
test?

> >
> >         ...
> >         [    0.490286] gpio_rcar e6055800.gpio: driving 4 GPIOs
> >         [    0.501956] SuperH (H)SCI(F) driver initialized
> >         [    0.503027] e6e88000.serial: ttySC0 at MMIO 0xe6e88000 (irq = 123, base_baud = 0) is a scif
> >         [    1.432046] printk: console [ttySC0] enabled
> > =>      [    1.443177] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> >         [    1.481426] loop: module loaded
> >         ...
> 
> Note that in both cases feb00000.display failed to probe first
> Is this an fw_devlink issue, due to HDMI linking to DU in DT?
> Does it work with fw_devlink=permissive of fw_devlink=off?

The DT is all connected up with the ports as I would expect them on
ulcb+kf, but I saw the unanswered question from Geert above. Could you
check that please?

Is there anything else specific to replicating this issue? Is it only on
the ULCB+KF? Or does it occur on Salvator-XS too? (And does it need a
specific kernel config to reproduce?)

--
Kieran


> Gr{oetje,eeting}s,
> 
>                         Geert
> 
> 
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org
> 
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
>                                 -- Linus Torvalds

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kieran Bingham]

Quoting Kuninori Morimoto (2021-11-03 23:23:15)
> 
> Hi Kieran
> 
> > > > =>      [    1.485684] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
> > > > =>      [    1.496248] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
> > > >         [    1.535173] loop: module loaded
> > > >         ...
> > > >
> > > > This is NG case.
> > > > DAM failed. rcar-dw-hdmi is not probed.
> > 
> > Was something changed to make this happen between this and the above
> > test?
> 
> Only reverting the issue patch.

Aha, ok - that was clearer when I re-read that you bisected to identify
the patch. Thanks for identifying the root cause.

> > > Note that in both cases feb00000.display failed to probe first
> > > Is this an fw_devlink issue, due to HDMI linking to DU in DT?
> > > Does it work with fw_devlink=permissive of fw_devlink=off?
> > 
> > The DT is all connected up with the ports as I would expect them on
> > ulcb+kf, but I saw the unanswered question from Geert above. Could you
> > check that please?
> 
> In my understanding, Laurent answered it at private meeting.
> He said it is not related, if my memory was correct.

Ah, sorry - I missed that.

> 
> > Is there anything else specific to replicating this issue? Is it only on
> > the ULCB+KF? Or does it occur on Salvator-XS too? (And does it need a
> > specific kernel config to reproduce?)
> 
> I could reproduce the same issue on Salvator-XS.
> I attached my .config on this mail.

Great, that's something I can try to reproduce then. Thanks, I'll boot
that up next.

> Thank you for your help !!
> 
> Best regards
> ---
> Kuninori Morimoto
>

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kuninori Morimoto]

Hi Laurent

You back !

This is gentle ping mail not to buried in a large amount of emails,
it had happen before ;P

> About this topic, v5.15 was released, but it still
> has "No soundcards" issue.
> There is no dw-hdmi, thus no soundcards
> 
> 	...
> 	[    1.431196] printk: console [ttySC0] enabled
> =>	[    1.442432] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> 	[    1.481179] loop: module loaded
> 	...
> 	[    9.568307] ALSA device list:
> =>	[    9.571446]   No soundcards found.
> 	...
> 
> I re-start works for me if I reverted this patch
> 
> 	--------------------------------
> 	commit d0f44e0dac29aca3f85d238dfbcd2e52450b38dc
> 	Author: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> 	Date:   Mon May 25 07:05:26 2020 +0300
> 
> 	    drm: rcar-du: dw-hdmi: Set output port number
> 
> 	    Report the DT output port number in dw_hdmi_plat_data to connect to the
> 	    next bridge in the dw-hdmi driver.
> 
> 	    Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> 	    Tested-by: Kieran Bingham <kieran.bingham@ideasonboard.com> # On V3U
> 	    Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
> 	--------------------------------
> 
> 	...
> 	[    1.431130] printk: console [ttySC0] enabled
> =>	[    1.442219] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> =>	[    1.451050] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
> =>	[    1.461560] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
> 	[    1.501043] loop: module loaded
> 	...
> 	[    9.180464] ALSA device list:
> =>	[    9.183519]   #0: rcar-sound
> 	...
> 
> Thank you for your help !!
> 
> Best regards
> ---
> Kuninori Morimoto

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Laurent Pinchart]

Hello Morimoto-san,

On Thu, Nov 11, 2021 at 09:35:10AM +0900, Kuninori Morimoto wrote:
> 
> Hi Laurent
> 
> You back !

And now I'm even back in Finland. I've found my H3 ULCB and tried to
boot it up, but I got stuck updating the firmware :-S

NOTICE:  BL2: R-Car Gen3 Initial Program Loader(CA57) Rev.2.0.6
NOTICE:  BL2: PRR is R-Car H3 Ver.2.0
NOTICE:  BL2: Board is Starter Kit Premier Rev.1.0
NOTICE:  BL2: Boot device is HyperFlash(80MHz)
NOTICE:  BL2: LCM state is CM
NOTICE:  AVS setting succeeded. DVFS_SetVID=0x53
NOTICE:  BL2: DDR3200(rev.0.40)
NOTICE:  BL2: [COLD_BOOT]
NOTICE:  BL2: Failed to DRAM initialize (-1).

This is on the ATF abloader branch from Marek. I've also tried the
master branch from
https://github.com/ARM-software/arm-trusted-firmware.git and the
rcar_gen3_v2.5 branch from
git://github.com/renesas-rcar/arm-trusted-firmware.git. They fail the
same way.

I have compiled ATF with

make PLAT=rcar LSI=H3 RCAR_DRAM_SPLIT=1 RCAR_GEN3_ULCB=1 SPD=opteed \
	CROSS_COMPILE=aarch64-buildroot-linux-gnu- \
	MBEDTLS_DIR=$PWD/../mbedtls/ clean bl2 bl31 rcar

and have flashed the four files to the addresses listed in
https://elinux.org/R-Car/Boards/H3SK#Flashing_firmware for the 4GiB
model (my board is a RTP0RC77951SKBX010SA00, S/N 2298)..

For reference, before trying to update the firmware, I had

[    0.000190] NOTICE:  BL2: R-Car Gen3 Initial Program Loader(CA57) Rev.1.0.13
[    0.005757] NOTICE:  BL2: PRR is R-Car H3 Ver2.0
[    0.010336] NOTICE:  BL2: Boot device is HyperFlash(80MHz)
[    0.015776] NOTICE:  BL2: LCM state is CM
[    0.019752] NOTICE:  BL2: AVS setting succeeded. DVFS_SetVID=0x52
[    0.025942] NOTICE:  BL2: DDR3200(rev.0.22)[    0.055637] NOTICE:  [COLD_BOOT][    0.087833] NOTICE:  ..0
[    0.088885] NOTICE:  BL2: DRAM Split is 4ch
[    0.093031] NOTICE:  BL2: QoS is default setting(rev.0.12)
[    0.098531] NOTICE:  BL2: Lossy Decomp areas
[    0.102708] NOTICE:       Entry 0: DCMPAREACRAx:0x80000540 DCMPAREACRBx:0x570
[    0.109793] NOTICE:       Entry 1: DCMPAREACRAx:0x40000000 DCMPAREACRBx:0x0
[    0.116705] NOTICE:       Entry 2: DCMPAREACRAx:0x20000000 DCMPAREACRBx:0x0
[    0.123620] NOTICE:  BL2: v1.3(release):867d841
[    0.128110] NOTICE:  BL2: Built : 16:02:18, Mar 16 2017
[    0.133297] NOTICE:  BL2: Normal boot
[    0.136938] NOTICE:  BL2: dst=0xe6320190 src=0x8180000 len=512(0x200)
[    0.143484] NOTICE:  BL2: dst=0x43f00000 src=0x8180400 len=6144(0x1800)
[    0.149946] NOTICE:  BL2: dst=0x44000000 src=0x81c0000 len=65536(0x10000)
[    0.157172] NOTICE:  BL2: dst=0x44100000 src=0x8200000 len=524288(0x80000)
[    0.167652] NOTICE:  BL2: dst=0x50000000 src=0x8640000 len=1048576(0x100000)

I don't have firmware binaries I can use to boot the board, so I can't
debug the HDMI probe issue :-(

I wonder if someone could help me. Has the latest ATF been tested on H3
ULCB ?

> This is gentle ping mail not to buried in a large amount of emails,
> it had happen before ;P
> 
> > About this topic, v5.15 was released, but it still
> > has "No soundcards" issue.
> > There is no dw-hdmi, thus no soundcards
> > 
> > 	...
> > 	[    1.431196] printk: console [ttySC0] enabled
> > =>	[    1.442432] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> > 	[    1.481179] loop: module loaded
> > 	...
> > 	[    9.568307] ALSA device list:
> > =>	[    9.571446]   No soundcards found.
> > 	...
> > 
> > I re-start works for me if I reverted this patch
> > 
> > 	--------------------------------
> > 	commit d0f44e0dac29aca3f85d238dfbcd2e52450b38dc
> > 	Author: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > 	Date:   Mon May 25 07:05:26 2020 +0300
> > 
> > 	    drm: rcar-du: dw-hdmi: Set output port number
> > 
> > 	    Report the DT output port number in dw_hdmi_plat_data to connect to the
> > 	    next bridge in the dw-hdmi driver.
> > 
> > 	    Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> > 	    Tested-by: Kieran Bingham <kieran.bingham@ideasonboard.com> # On V3U
> > 	    Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
> > 	--------------------------------
> > 
> > 	...
> > 	[    1.431130] printk: console [ttySC0] enabled
> > =>	[    1.442219] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> > =>	[    1.451050] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
> > =>	[    1.461560] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
> > 	[    1.501043] loop: module loaded
> > 	...
> > 	[    9.180464] ALSA device list:
> > =>	[    9.183519]   #0: rcar-sound
> > 	...

-- 
Regards,

Laurent Pinchart

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Laurent Pinchart]

Hello Morimoto-san,

Until I can get my ULCB to boot to test this here, I'll try to help you
debug the problem.

On Mon, Nov 01, 2021 at 03:49:24PM +0900, Kuninori Morimoto wrote:
> 
> Hi Laurent, again
> 
> About this topic, v5.15 was released, but it still
> has "No soundcards" issue.
> There is no dw-hdmi, thus no soundcards
> 
> 	...
> 	[    1.431196] printk: console [ttySC0] enabled
> =>	[    1.442432] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> 	[    1.481179] loop: module loaded
> 	...
> 	[    9.568307] ALSA device list:
> =>	[    9.571446]   No soundcards found.
> 	...

Do you get any error message from the dw-hdmi driver ? In particular,
I'd like to know if the dw_hdmi_parse_dt() function fails. There are
error messages in all error paths but the last one, the following patch
may help debugging the problem.

diff --git a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
index 54d8fdad395f..8be3a1042ede 100644
--- a/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
+++ b/drivers/gpu/drm/bridge/synopsys/dw-hdmi.c
@@ -3209,10 +3209,14 @@ static int dw_hdmi_parse_dt(struct dw_hdmi *hdmi)
 	}

 	hdmi->next_bridge = of_drm_find_bridge(remote);
-	of_node_put(remote);
-	if (!hdmi->next_bridge)
+	if (!hdmi->next_bridge) {
+		dev_err(hdmi->dev, "no bridge found for port@%u (%pOF)\n",
+			hdmi->plat_data->output_port, remote);
+		of_node_put(remote);
 		return -EPROBE_DEFER;
+	}

+	of_node_put(remote);
 	return 0;
 }

Is the CONFIG_DRM_DISPLAY_CONNECTOR option enabled in your kernel ?

> I re-start works for me if I reverted this patch
> 
> 	--------------------------------
> 	commit d0f44e0dac29aca3f85d238dfbcd2e52450b38dc
> 	Author: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> 	Date:   Mon May 25 07:05:26 2020 +0300
> 
> 	    drm: rcar-du: dw-hdmi: Set output port number
> 
> 	    Report the DT output port number in dw_hdmi_plat_data to connect to the
> 	    next bridge in the dw-hdmi driver.
> 
> 	    Signed-off-by: Laurent Pinchart <laurent.pinchart+renesas@ideasonboard.com>
> 	    Tested-by: Kieran Bingham <kieran.bingham@ideasonboard.com> # On V3U
> 	    Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com>
> 	--------------------------------
> 
> 	...
> 	[    1.431130] printk: console [ttySC0] enabled
> =>	[    1.442219] rcar-du feb00000.display: failed to initialize DRM/KMS (-6)
> =>	[    1.451050] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
> =>	[    1.461560] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
> 	[    1.501043] loop: module loaded
> 	...
> 	[    9.180464] ALSA device list:
> =>	[    9.183519]   #0: rcar-sound
> 	...

-- 
Regards,

Laurent Pinchart

Re: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Laurent Pinchart]

Hello Morimoto-san,

For some reason this mail hadn't made it to the list, nor to my inbox.

On Thu, Nov 04, 2021 at 08:23:15AM +0900, Kuninori Morimoto wrote:
> 
> Hi Kieran
> 
> > > > =>      [    1.485684] rcar-dw-hdmi fead0000.hdmi: Detected HDMI TX controller v2.01a with HDCP (DWC HDMI 2.0 TX PHY)
> > > > =>      [    1.496248] rcar-dw-hdmi fead0000.hdmi: registered DesignWare HDMI I2C bus driver
> > > >         [    1.535173] loop: module loaded
> > > >         ...
> > > >
> > > > This is NG case.
> > > > DAM failed. rcar-dw-hdmi is not probed.
> > 
> > Was something changed to make this happen between this and the above
> > test?
> 
> Only reverting the issue patch.
> 
> > > Note that in both cases feb00000.display failed to probe first
> > > Is this an fw_devlink issue, due to HDMI linking to DU in DT?
> > > Does it work with fw_devlink=permissive of fw_devlink=off?
> > 
> > The DT is all connected up with the ports as I would expect them on
> > ulcb+kf, but I saw the unanswered question from Geert above. Could you
> > check that please?
> 
> In my understanding, Laurent answered it at private meeting.
> He said it is not related, if my memory was correct.
> 
> > Is there anything else specific to replicating this issue? Is it only on
> > the ULCB+KF? Or does it occur on Salvator-XS too? (And does it need a
> > specific kernel config to reproduce?)
> 
> I could reproduce the same issue on Salvator-XS.
> I attached my .config on this mail.

[snip]

> # CONFIG_DRM_DISPLAY_CONNECTOR is not set

Enabling this option should help.

-- 
Regards,

Laurent Pinchart

RE: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kuninori Morimoto]

Hi Laurent

>> # CONFIG_DRM_DISPLAY_CONNECTOR is not set
>
> Enabling this option should help.

Thank you for help me.
I'm sorry but I'm busy in these days.
Will try and check above after that.

--
Best Regards,

Kuninori Morimoto

RE: HDMI isn't probed on v5.15-rc1 on ULCB+KF

[Kuninori Morimoto]

Hi Laurent

Thank you for your help,
and sorry for my late response.
It was PC replace week

>>> # CONFIG_DRM_DISPLAY_CONNECTOR is not set
>>
>> Enabling this option should help.

Thank you. I could confirm that HDMI sound could be probed.

Best regards
---
Kuninori Morimoto