[PATCH v2] ARM: dts: samsung: exynos4412-midas: add USB connector and USB OTG

Henrik Grimler henrik at grimler.se
Tue Jul 25 14:22:09 PDT 2023


Hi Krzysztof,

On Tue, Jul 25, 2023 at 10:08:00PM +0200, Krzysztof Kozlowski wrote:
> On 25/07/2023 21:15, Henrik Grimler wrote:
> > On Sun, Jul 23, 2023 at 04:24:17PM +0200, Krzysztof Kozlowski wrote:
> >> Add full description of USB-MUIC (MAX77693 MUIC) and MUIC-MHL
> >> connections, along with proper USB connector and OTG mode for DWC2 USB
> >> controller.
> >>
> >> This fixes dtc W=1 warnings:
> >>
> >>   Warning (graph_child_address): /i2c-mhl/hdmi-bridge at 39/ports: graph node has single child node 'port at 0', #address-cells/#size-cells are not necessary
> >>
> >> Cc: Marek Szyprowski <m.szyprowski at samsung.com>
> >> Cc: replicant at osuosl.org
> >> Cc: phone-devel at vger.kernel.org
> >> Cc: ~postmarketos/upstreaming at lists.sr.ht
> >> Cc: Martin Jücker <martin.juecker at gmail.com>
> >> Cc: Henrik Grimler <henrik at grimler.se>
> >> Cc: Artur Weber <aweber.kernel at gmail.com>
> >> Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski at linaro.org>
> > 
> > Tested-by: Henrik Grimler <henrik at grimler.se>
> > 
> > * Peripheral mode (still) works
> > * MHL somewhat works
> > * OTG (still) does not work
> > 
> > Further details for those interested:
> > 
> > MHL works reliably if cable is attached before boot, but if cable is
> > attached on already running device it sometimes work after detaching
> > and attaching cable a few times.  On cable attach, when it does not
> > work, we get:
> > 
> >   [  252.831734] max77693-muic max77693-muic: external connector is detached(chg_type:0x0, prev_chg_type:0x0)
> >   [  253.733916] max77693-muic max77693-muic: external connector is attached(chg_type:0x1, prev_chg_type:0x1)
> >   [  253.735232] max77693-muic max77693-muic: CONTROL1 : 0x09, CONTROL2 : 0x04, state : attached
> >   [  254.691877] i2c i2c-15: sendbytes: NAK bailout.
> >   [  254.692150] sii9234 15-0039: writebm:  TPI[0x3d] <- 0x3e
> >   [  255.984336] sii9234 15-0039: RGND is not 1k
> >   [  256.068970] sii9234 15-0039: RSEN_HIGH without RGND_1K
> >   [  256.104457] sii9234 15-0039: discovery failed, no power for MHL?
> >   [  256.118272] i2c i2c-15: sendbytes: NAK bailout.
> >   [  256.118539] sii9234 15-0039: writebm:  TPI[0x3d] <- 0x3e
> >   [  256.375966] sii9234 15-0039: RSEN_HIGH without RGND_1K
> >   [  256.411561] sii9234 15-0039: discovery failed, no power for MHL?
> > 
> > while when it works:
> > 
> >   [  175.348307] max77693-muic max77693-muic: external connector is detached(chg_type:0x1, prev_chg_type:0x0)
> >   [  175.349576] max77693-muic max77693-muic: CONTROL1 : 0x00, CONTROL2 : 0x01, state : detached
> >   [  179.304373] dwc2 12480000.usb: new device is full-speed
> >   [  179.305920] max77693-muic max77693-muic: external connector is detached(chg_type:0x0, prev_chg_type:0x0)
> >   [  180.205735] max77693-muic max77693-muic: external connector is attached(chg_type:0x1, prev_chg_type:0x1)
> >   [  180.207211] max77693-muic max77693-muic: CONTROL1 : 0x09, CONTROL2 : 0x04, state : attached
> >   [  180.309727] i2c i2c-15: sendbytes: NAK bailout.
> >   [  180.309996] sii9234 15-0039: writebm:  TPI[0x3d] <- 0x3e
> > 
> > Not sure if issue here is in muic driver or sii9234 driver.
> > 
> > OTG still does not seem to work, the muic driver detects (only) a
> > detach when OTG cable is attached:
> > 
> >   max77693-muic max77693-muic: external connector is detached(chg_type:0x0, prev_chg_type:0x0)
> > 
> > Same happened when otg cable was attached before these hsotg changes
> > as well. I suppose issue here is with max77693-muic driver.
> > 
> 
> Thanks for testing. I miss here one important thing - did anything
> regressed? Is MHL behaving not worse than before? OTG, as non-working,
> matters less in such case.

No regressions as far as I can tell, only improvements.  We want the
muic node and mhl_to_muic/muic_to_mhl ports so that muic can notify
the mhl chip of cable attachment/detachment (with requires an
additional patch to the sii9234 driver that I will send soon [1]).
The alternative, having the mhl chip always on, has caused issues with
some GUIs [2].

> Best regards,
> Krzysztof

[1] https://gitlab.com/exynos4-mainline/linux/-/commit/742d05797799
[2] https://gitlab.gnome.org/World/Phosh/phosh/-/issues/828#note_1542189

Best regards,
Henrik Grimler



More information about the linux-arm-kernel mailing list