Message ID | 20180422142151.17910-1-contact@paulk.fr (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
On Sun, Apr 22, 2018 at 04:21:51PM +0200, Paul Kocialkowski wrote: > The Solid-Run CuBox-i lower board used in the first generation of > CuBox-i devices feature a hinged micro SD card slot, that does not have > card-detect capability. Since the card-detect GPIO was specified in the > common cubox-i dtsi, it is moved to each device using the 1.5 SOM and is > thus removed from the imx6q-cubox-i dts. This seems incorrect. The 1.5 SOM is a relatively recent thing, and I have one of the first Cubox-i's that were produced which is not hinged, and does not have a 1.5 SOM. There is _no_ correlation between the 1.5 SOM and the SD card slot. Talking to folk at SolidRun, we're all confused about your assertions. We're also confused about "hinged micro SD card slot" - Cubox-i's come with either a push-push slot or a static (push-in, pull-out) slot. (Prototypes were the latter.) Maybe someone's modified yours?
Hi, Le dimanche 22 avril 2018 à 15:39 +0100, Russell King - ARM Linux a écrit : > On Sun, Apr 22, 2018 at 04:21:51PM +0200, Paul Kocialkowski wrote: > > The Solid-Run CuBox-i lower board used in the first generation of > > CuBox-i devices feature a hinged micro SD card slot, that does not > > have > > card-detect capability. Since the card-detect GPIO was specified in > > the > > common cubox-i dtsi, it is moved to each device using the 1.5 SOM > > and is > > thus removed from the imx6q-cubox-i dts. > > This seems incorrect. > > The 1.5 SOM is a relatively recent thing, and I have one of the first > Cubox-i's that were produced which is not hinged, and does not have a > 1.5 SOM. There is _no_ correlation between the 1.5 SOM and the SD > card slot. > > Talking to folk at SolidRun, we're all confused about your assertions. It appears that I was also confused about the different models in the CuBox-i line. I had assumed that the models had been updated to a new revision (carrying that 1.5 SOM) since the time I bought mine (which was thus some sort of first generation). With this scenario in mind, it made sense to me that a CD GPIO required only for the second generation would have been applied by mistake to all the cubox-i devices by having it in the common dtsi. I checked on the leaflet inside the box and models did not change, only the 4x4 model was added in the meantime. > We're also confused about "hinged micro SD card slot" - Cubox-i's > come with either a push-push slot or a static (push-in, pull-out) > slot. > (Prototypes were the latter.) The one I have definitely does not have a spring-loaded mechanism, so I would call it a push/pull static slot. I mentioned a hinged slot because that's the term used to describe the layout in the CuBox-i-Lower schematics, SD/MMC Connector part: "U3 is an internal hinged micro SD". Looking at pictures of what hinged slots look like, that's definitely not what I have here. Are all CuBox-i units that have ever been sold supposed to have a connector with a CD line? I find it hard to believe that it's broken specifically on mine. If that's really needed, I could open up the device, check whether the R8 pull is in place and check the voltage there. Cheers,
Hi, Le dimanche 22 avril 2018 à 15:02 +0000, Jon Nettleton a écrit : > On Sun, Apr 22, 2018 at 4:39 PM Russell King - ARM Linux <linux@armlin > ux.org.uk> wrote: > > On Sun, Apr 22, 2018 at 04:21:51PM +0200, Paul Kocialkowski wrote: > > > The Solid-Run CuBox-i lower board used in the first generation of > > > CuBox-i devices feature a hinged micro SD card slot, that does not > > have > > > card-detect capability. Since the card-detect GPIO was specified > > in the > > > common cubox-i dtsi, it is moved to each device using the 1.5 SOM > > and is > > > thus removed from the imx6q-cubox-i dts. > > > > This seems incorrect. > > > > The 1.5 SOM is a relatively recent thing, and I have one of the > > first > > Cubox-i's that were produced which is not hinged, and does not have > > a > > 1.5 SOM. There is _no_ correlation between the 1.5 SOM and the SD > > card slot. > > > > Talking to folk at SolidRun, we're all confused about your > > assertions. > > > > We're also confused about "hinged micro SD card slot" - Cubox-i's > > come with either a push-push slot or a static (push-in, pull-out) > > slot. > > (Prototypes were the latter.) > > > > Maybe someone's modified yours? > > > Regardless of the physical mechanics of the device, the CD-GPIO is a > mechanism of the carrier and not the SOM. I have one of the few Cubox- > i's that have been physically modified to work with the rev 1.5 som > and have no issues (the heat spreader needs additional cutouts). There was some confusion about the relationship with the 1.5 SOM. It is now clear that the issue is not linked to it at all. > Now if you are still using the old 2013 u-boot then there is possibly > confusion between the eMMC and SDHC slots which can cause > strangeness. Please use the 2018.01 branch on our github and that > will handle this properly. I'm using U-Boot 2018.03, so I think I'm safe on that side. Cheers,
Hi, Le dimanche 22 avril 2018 à 19:22 +0200, Paul Kocialkowski a écrit : > Are all CuBox-i units that have ever been sold supposed to have a > connector with a CD line? I find it hard to believe that it's broken > specifically on mine. Do you have a clue whether there should, in fact, be a CD line on all CuBox-i models out there? Although you seemed to imply that my device might be broken, I am still confused about whether this issue concerns my device only or a specific revision of the CuBox-i. > If that's really needed, I could open up the device, check whether the > R8 pull is in place and check the voltage there. In case there is a doubt, I will proceed with opening the device so we can have a clear idea, but I'd rather avoid it as much as possible. Cheers,
On Sun, Apr 29, 2018 at 3:47 PM, Paul Kocialkowski <contact@paulk.fr> wrote: > Hi, > > Le dimanche 22 avril 2018 à 19:22 +0200, Paul Kocialkowski a écrit : >> Are all CuBox-i units that have ever been sold supposed to have a >> connector with a CD line? I find it hard to believe that it's broken >> specifically on mine. > > Do you have a clue whether there should, in fact, be a CD line on all > CuBox-i models out there? Although you seemed to imply that my device > might be broken, I am still confused about whether this issue concerns > my device only or a specific revision of the CuBox-i. > >> If that's really needed, I could open up the device, check whether the >> R8 pull is in place and check the voltage there. > > In case there is a doubt, I will proceed with opening the device so we > can have a clear idea, but I'd rather avoid it as much as possible. > > Cheers, > > -- > Paul Kocialkowski, > > developer of free digital technology and hardware support. > > Website: https://www.paulk.fr/ > Coding blog: https://code.paulk.fr/ > Git repositories: https://git.paulk.fr/ https://git.code.paulk.fr/ Paul, You can see on our schematics that the hinged sdhc card slot is an assembly option and not the default. https://wiki.solid-run.com/lib/exe/fetch.php?media=imx6:cubox-i:schematics-rev-1.1:cubox-i-lower.pdf You have either an initial early developer unit (pre-production) that was assembled like this for testing, or a unit that someone specifically ordered assembled this way. You can see from the schematic that this option does not have a CD-pin because it is meant to be a more permanent storage option. That being said, this is not a default option and therefore should not be merged into the mainline device-tree. Our soms and carriers are designed to be very flexible regarding the assembly options and we try to mainline device-tree files that represent the most common assembly options. Once there is a common accepted mainlining for device-tree fragments for overlays, that would be the proper place to make this change. Although in this case you would also need a custom u-boot so it is probably best to just handle the change there. If you would like touch base with me off list and I can point you at the required u-boot changes. Thanks, Jon
Hi, Le dimanche 29 avril 2018 à 16:09 +0200, Jon Nettleton a écrit : > You have either an initial early developer unit (pre-production) that > was assembled like this for testing, or a unit that someone > specifically ordered assembled this way. You can see from the > schematic that this option does not have a CD-pin because it is meant > to be a more permanent storage option. It turns out that the card-detection mechanism was somewhat broken on my device (which does have a regular push slot). It would report card- detect only when the card was not fully inserted, and stop asserting card-detect when it was pushed fully. I used a fitting tweezer to move the mechanical parts around and got it to also detect the card when it is pushed all the way. So no change is necessary after all, it was just my device that needed fixing. Thanks for the helpful answers!
diff --git a/arch/arm/boot/dts/imx6dl-cubox-i-emmc-som-v15.dts b/arch/arm/boot/dts/imx6dl-cubox-i-emmc-som-v15.dts index 2b2fc360b865..90a15f3b4fc0 100644 --- a/arch/arm/boot/dts/imx6dl-cubox-i-emmc-som-v15.dts +++ b/arch/arm/boot/dts/imx6dl-cubox-i-emmc-som-v15.dts @@ -50,3 +50,7 @@ model = "SolidRun Cubox-i Solo/DualLite (1.5som+emmc)"; compatible = "solidrun,cubox-i/dl", "fsl,imx6dl"; }; + +&usdhc2 { + cd-gpios = <&gpio1 4 GPIO_ACTIVE_LOW>; +}; diff --git a/arch/arm/boot/dts/imx6dl-cubox-i-som-v15.dts b/arch/arm/boot/dts/imx6dl-cubox-i-som-v15.dts index e09c565d1d1f..9bdec07fefe3 100644 --- a/arch/arm/boot/dts/imx6dl-cubox-i-som-v15.dts +++ b/arch/arm/boot/dts/imx6dl-cubox-i-som-v15.dts @@ -49,3 +49,7 @@ model = "SolidRun Cubox-i Solo/DualLite (1.5som)"; compatible = "solidrun,cubox-i/dl", "fsl,imx6dl"; }; + +&usdhc2 { + cd-gpios = <&gpio1 4 GPIO_ACTIVE_LOW>; +}; diff --git a/arch/arm/boot/dts/imx6q-cubox-i-emmc-som-v15.dts b/arch/arm/boot/dts/imx6q-cubox-i-emmc-som-v15.dts index 3e59ebbb3608..679a5431cbdb 100644 --- a/arch/arm/boot/dts/imx6q-cubox-i-emmc-som-v15.dts +++ b/arch/arm/boot/dts/imx6q-cubox-i-emmc-som-v15.dts @@ -58,3 +58,7 @@ fsl,transmit-atten-16ths = <9>; fsl,no-spread-spectrum; }; + +&usdhc2 { + cd-gpios = <&gpio1 4 GPIO_ACTIVE_LOW>; +}; diff --git a/arch/arm/boot/dts/imx6q-cubox-i-som-v15.dts b/arch/arm/boot/dts/imx6q-cubox-i-som-v15.dts index dab70d1230a2..b9b8733f13c3 100644 --- a/arch/arm/boot/dts/imx6q-cubox-i-som-v15.dts +++ b/arch/arm/boot/dts/imx6q-cubox-i-som-v15.dts @@ -57,3 +57,7 @@ fsl,transmit-atten-16ths = <9>; fsl,no-spread-spectrum; }; + +&usdhc2 { + cd-gpios = <&gpio1 4 GPIO_ACTIVE_LOW>; +}; diff --git a/arch/arm/boot/dts/imx6qdl-cubox-i.dtsi b/arch/arm/boot/dts/imx6qdl-cubox-i.dtsi index d1cfdc264126..170b3d27f875 100644 --- a/arch/arm/boot/dts/imx6qdl-cubox-i.dtsi +++ b/arch/arm/boot/dts/imx6qdl-cubox-i.dtsi @@ -254,7 +254,6 @@ pinctrl-names = "default"; pinctrl-0 = <&pinctrl_cubox_i_usdhc2_aux &pinctrl_cubox_i_usdhc2>; vmmc-supply = <&vcc_3v3>; - cd-gpios = <&gpio1 4 GPIO_ACTIVE_LOW>; status = "okay"; };
The Solid-Run CuBox-i lower board used in the first generation of CuBox-i devices feature a hinged micro SD card slot, that does not have card-detect capability. Since the card-detect GPIO was specified in the common cubox-i dtsi, it is moved to each device using the 1.5 SOM and is thus removed from the imx6q-cubox-i dts. Enabling the card-detect GPIO on the first generation CuBox-i devices results in waiting for a card-detect event (that never occurs) and thus not detecting the mmc card. Since the CD pin is still dedicated to the purpose on the SOM board even on the first generation CuBox-i devices, the pin control properties are kept untouched regardless of the slot capability. Signed-off-by: Paul Kocialkowski <contact@paulk.fr> --- arch/arm/boot/dts/imx6dl-cubox-i-emmc-som-v15.dts | 4 ++++ arch/arm/boot/dts/imx6dl-cubox-i-som-v15.dts | 4 ++++ arch/arm/boot/dts/imx6q-cubox-i-emmc-som-v15.dts | 4 ++++ arch/arm/boot/dts/imx6q-cubox-i-som-v15.dts | 4 ++++ arch/arm/boot/dts/imx6qdl-cubox-i.dtsi | 1 - 5 files changed, 16 insertions(+), 1 deletion(-)