Message ID | 1636566415-22750-13-git-send-email-abel.vesa@nxp.com (mailing list archive) |
---|---|
State | New, archived |
Headers | show |
Series | arm64: dts: Add i.MX8DXL initial support | expand |
On Wed, Nov 10, 2021 at 07:46:55PM +0200, Abel Vesa wrote: > Add i.MX8DXL lpuart compatible to the bindings documentation. > > Signed-off-by: Abel Vesa <abel.vesa@nxp.com> > --- > Documentation/devicetree/bindings/serial/fsl-lpuart.yaml | 4 ++++ > 1 file changed, 4 insertions(+) > > diff --git a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > index dc1f0e07cbd4..fa8a602ccb22 100644 > --- a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > +++ b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > @@ -27,6 +27,10 @@ properties: > - items: > - const: fsl,imx8qm-lpuart > - const: fsl,imx8qxp-lpuart > + - items: > + - const: fsl,imx8dxl-lpuart > + - const: fsl,imx8qxp-lpuart > + - const: fsl,imx7ulp-lpuart I'm confused why 8dxl is compatible with 7ulp, but 8qm is not? From the driver, it looks like the difference is clocks. > > reg: > maxItems: 1 > -- > 2.31.1 > >
Am Donnerstag, dem 18.11.2021 um 19:04 -0600 schrieb Rob Herring: > On Wed, Nov 10, 2021 at 07:46:55PM +0200, Abel Vesa wrote: > > Add i.MX8DXL lpuart compatible to the bindings documentation. > > > > Signed-off-by: Abel Vesa < > > abel.vesa@nxp.com > > > > > --- > > Documentation/devicetree/bindings/serial/fsl-lpuart.yaml | 4 ++++ > > 1 file changed, 4 insertions(+) > > > > diff --git a/Documentation/devicetree/bindings/serial/fsl- > > lpuart.yaml b/Documentation/devicetree/bindings/serial/fsl- > > lpuart.yaml > > index dc1f0e07cbd4..fa8a602ccb22 100644 > > --- a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > +++ b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > @@ -27,6 +27,10 @@ properties: > > - items: > > - const: fsl,imx8qm-lpuart > > - const: fsl,imx8qxp-lpuart > > + - items: > > + - const: fsl,imx8dxl-lpuart > > + - const: fsl,imx8qxp-lpuart > > + - const: fsl,imx7ulp-lpuart > > I'm confused why 8dxl is compatible with 7ulp, but 8qm is not? From > the > driver, it looks like the difference is clocks. There is a difference between 8qm and 7ulp regarding the clocks. Are they still considered compatible? Depending on the answer [1] might not be the correct solution for earlycon regression on 8qm. Best regards, Alexander [1] http://lists.infradead.org/pipermail/linux-arm-kernel/2021-November/696770.html
On 21-11-18 19:04:04, Rob Herring wrote: > On Wed, Nov 10, 2021 at 07:46:55PM +0200, Abel Vesa wrote: > > Add i.MX8DXL lpuart compatible to the bindings documentation. > > > > Signed-off-by: Abel Vesa <abel.vesa@nxp.com> > > --- > > Documentation/devicetree/bindings/serial/fsl-lpuart.yaml | 4 ++++ > > 1 file changed, 4 insertions(+) > > > > diff --git a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > index dc1f0e07cbd4..fa8a602ccb22 100644 > > --- a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > +++ b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > @@ -27,6 +27,10 @@ properties: > > - items: > > - const: fsl,imx8qm-lpuart > > - const: fsl,imx8qxp-lpuart > > + - items: > > + - const: fsl,imx8dxl-lpuart > > + - const: fsl,imx8qxp-lpuart > > + - const: fsl,imx7ulp-lpuart > > I'm confused why 8dxl is compatible with 7ulp, but 8qm is not? From the > driver, it looks like the difference is clocks. > There are still things to be upstreamed for the fsl-lpuart driver. Looking at the NXP's tree, 7ulp has some specific fixup that the 8qm doesn't need. Have a look here: https://source.codeaurora.org/external/imx/linux-imx/tree/drivers/tty/serial/fsl_lpuart.c?h=lf-5.10.y#n3242 > > > > reg: > > maxItems: 1 > > -- > > 2.31.1 > > > >
On 21-11-19 08:17:11, Alexander Stein wrote: > Am Donnerstag, dem 18.11.2021 um 19:04 -0600 schrieb Rob Herring: > > On Wed, Nov 10, 2021 at 07:46:55PM +0200, Abel Vesa wrote: > > > Add i.MX8DXL lpuart compatible to the bindings documentation. > > > > > > Signed-off-by: Abel Vesa < > > > abel.vesa@nxp.com > > > > > > > --- > > > Documentation/devicetree/bindings/serial/fsl-lpuart.yaml | 4 ++++ > > > 1 file changed, 4 insertions(+) > > > > > > diff --git a/Documentation/devicetree/bindings/serial/fsl- > > > lpuart.yaml b/Documentation/devicetree/bindings/serial/fsl- > > > lpuart.yaml > > > index dc1f0e07cbd4..fa8a602ccb22 100644 > > > --- a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > > +++ b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > > @@ -27,6 +27,10 @@ properties: > > > - items: > > > - const: fsl,imx8qm-lpuart > > > - const: fsl,imx8qxp-lpuart > > > + - items: > > > + - const: fsl,imx8dxl-lpuart > > > + - const: fsl,imx8qxp-lpuart > > > + - const: fsl,imx7ulp-lpuart > > > > I'm confused why 8dxl is compatible with 7ulp, but 8qm is not? From > > the > > driver, it looks like the difference is clocks. > > There is a difference between 8qm and 7ulp regarding the clocks. Are > they still considered compatible? Depending on the answer [1] might not > be the correct solution for earlycon regression on 8qm. > In NXP's tree, they are not compatible. See here: https://source.codeaurora.org/external/imx/linux-imx/tree/arch/arm64/boot/dts/freescale/imx8qm-ss-dma.dtsi?h=lf-5.10.y#n9 > Best regards, > Alexander > > [1] > http://lists.infradead.org/pipermail/linux-arm-kernel/2021-November/696770.html > >
Am Montag, dem 22.11.2021 um 18:45 +0200 schrieb Abel Vesa: > On 21-11-19 08:17:11, Alexander Stein wrote: > > Am Donnerstag, dem 18.11.2021 um 19:04 -0600 schrieb Rob Herring: > > > On Wed, Nov 10, 2021 at 07:46:55PM +0200, Abel Vesa wrote: > > > > Add i.MX8DXL lpuart compatible to the bindings documentation. > > > > > > > > Signed-off-by: Abel Vesa < > > > > abel.vesa@nxp.com > > > > > > > > > > > > --- > > > > Documentation/devicetree/bindings/serial/fsl-lpuart.yaml | 4 > > > > ++++ > > > > 1 file changed, 4 insertions(+) > > > > > > > > diff --git a/Documentation/devicetree/bindings/serial/fsl- > > > > lpuart.yaml b/Documentation/devicetree/bindings/serial/fsl- > > > > lpuart.yaml > > > > index dc1f0e07cbd4..fa8a602ccb22 100644 > > > > --- a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > > > +++ b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > > > @@ -27,6 +27,10 @@ properties: > > > > - items: > > > > - const: fsl,imx8qm-lpuart > > > > - const: fsl,imx8qxp-lpuart > > > > + - items: > > > > + - const: fsl,imx8dxl-lpuart > > > > + - const: fsl,imx8qxp-lpuart > > > > + - const: fsl,imx7ulp-lpuart > > > > > > I'm confused why 8dxl is compatible with 7ulp, but 8qm is not? > > > From > > > the > > > driver, it looks like the difference is clocks. > > > > There is a difference between 8qm and 7ulp regarding the clocks. > > Are > > they still considered compatible? Depending on the answer [1] might > > not > > be the correct solution for earlycon regression on 8qm. > > > > In NXP's tree, they are not compatible. > > See here: > > https://source.codeaurora.org/external/imx/linux-imx/tree/arch/arm64/boot/dts/freescale/imx8qm-ss-dma.dtsi?h=lf-5.10.y#n9 Well, commit https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=b4b844930f27bf7019c0bbd8cc575dde32e00ecc says otherwise. This might be true for earlycon only, because clocks should be setup already. Together with your other mail regarding the imx7ulp fixup, commit b4b844930f27bf7019c0bbd8cc575dde32e00ecc should just be reverted in order to get earlycon running again on imx8qm and imx8qxp, IMHO. Best regards, Alexander
On 21-11-23 08:30:17, Alexander Stein wrote: > Am Montag, dem 22.11.2021 um 18:45 +0200 schrieb Abel Vesa: > > On 21-11-19 08:17:11, Alexander Stein wrote: > > > Am Donnerstag, dem 18.11.2021 um 19:04 -0600 schrieb Rob Herring: > > > > On Wed, Nov 10, 2021 at 07:46:55PM +0200, Abel Vesa wrote: > > > > > Add i.MX8DXL lpuart compatible to the bindings documentation. > > > > > > > > > > Signed-off-by: Abel Vesa < > > > > > abel.vesa@nxp.com > > > > > > > > > > > > > > > --- > > > > > Documentation/devicetree/bindings/serial/fsl-lpuart.yaml | 4 > > > > > ++++ > > > > > 1 file changed, 4 insertions(+) > > > > > > > > > > diff --git a/Documentation/devicetree/bindings/serial/fsl- > > > > > lpuart.yaml b/Documentation/devicetree/bindings/serial/fsl- > > > > > lpuart.yaml > > > > > index dc1f0e07cbd4..fa8a602ccb22 100644 > > > > > --- a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > > > > +++ b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml > > > > > @@ -27,6 +27,10 @@ properties: > > > > > - items: > > > > > - const: fsl,imx8qm-lpuart > > > > > - const: fsl,imx8qxp-lpuart > > > > > + - items: > > > > > + - const: fsl,imx8dxl-lpuart > > > > > + - const: fsl,imx8qxp-lpuart > > > > > + - const: fsl,imx7ulp-lpuart > > > > > > > > I'm confused why 8dxl is compatible with 7ulp, but 8qm is not? > > > > From > > > > the > > > > driver, it looks like the difference is clocks. > > > > > > There is a difference between 8qm and 7ulp regarding the clocks. > > > Are > > > they still considered compatible? Depending on the answer [1] might > > > not > > > be the correct solution for earlycon regression on 8qm. > > > > > > > In NXP's tree, they are not compatible. > > > > See here: > > > > https://source.codeaurora.org/external/imx/linux-imx/tree/arch/arm64/boot/dts/freescale/imx8qm-ss-dma.dtsi?h=lf-5.10.y#n9 > > Well, commit > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=b4b844930f27bf7019c0bbd8cc575dde32e00ecc > says otherwise. > This might be true for earlycon only, because clocks should be setup > already. > > Together with your other mail regarding the imx7ulp fixup, > commit b4b844930f27bf7019c0bbd8cc575dde32e00ecc should just be reverted > in order to get earlycon running again on imx8qm and imx8qxp, IMHO. > I think you're right, the commit b4b844930f27bf7019c should be reverted. You can send the revert yourself, if you want. If not, let me know. > Best regards, > Alexander > >
Am Donnerstag, dem 25.11.2021 um 15:19 +0200 schrieb Abel Vesa: > On 21-11-23 08:30:17, Alexander Stein wrote: > > Am Montag, dem 22.11.2021 um 18:45 +0200 schrieb Abel Vesa: > > > On 21-11-19 08:17:11, Alexander Stein wrote: > > > > Am Donnerstag, dem 18.11.2021 um 19:04 -0600 schrieb Rob > > > > Herring: > > > > > On Wed, Nov 10, 2021 at 07:46:55PM +0200, Abel Vesa wrote: > > > > > > Add i.MX8DXL lpuart compatible to the bindings > > > > > > documentation. > > > > > > > > > > > > Signed-off-by: Abel Vesa < > > > > > > abel.vesa@nxp.com > > > > > > > > > > > > > > > > > > > > > > > > --- > > > > > > Documentation/devicetree/bindings/serial/fsl-lpuart.yaml | > > > > > > 4 > > > > > > ++++ > > > > > > 1 file changed, 4 insertions(+) > > > > > > > > > > > > diff --git a/Documentation/devicetree/bindings/serial/fsl- > > > > > > lpuart.yaml b/Documentation/devicetree/bindings/serial/fsl- > > > > > > lpuart.yaml > > > > > > index dc1f0e07cbd4..fa8a602ccb22 100644 > > > > > > --- a/Documentation/devicetree/bindings/serial/fsl- > > > > > > lpuart.yaml > > > > > > +++ b/Documentation/devicetree/bindings/serial/fsl- > > > > > > lpuart.yaml > > > > > > @@ -27,6 +27,10 @@ properties: > > > > > > - items: > > > > > > - const: fsl,imx8qm-lpuart > > > > > > - const: fsl,imx8qxp-lpuart > > > > > > + - items: > > > > > > + - const: fsl,imx8dxl-lpuart > > > > > > + - const: fsl,imx8qxp-lpuart > > > > > > + - const: fsl,imx7ulp-lpuart > > > > > > > > > > I'm confused why 8dxl is compatible with 7ulp, but 8qm is > > > > > not? > > > > > From > > > > > the > > > > > driver, it looks like the difference is clocks. > > > > > > > > There is a difference between 8qm and 7ulp regarding the > > > > clocks. > > > > Are > > > > they still considered compatible? Depending on the answer [1] > > > > might > > > > not > > > > be the correct solution for earlycon regression on 8qm. > > > > > > > > > > In NXP's tree, they are not compatible. > > > > > > See here: > > > > > > https://source.codeaurora.org/external/imx/linux-imx/tree/arch/arm64/boot/dts/freescale/imx8qm-ss-dma.dtsi?h=lf-5.10.y#n9 > > > > > > > Well, commit > > https://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git/commit/?id=b4b844930f27bf7019c0bbd8cc575dde32e00ecc > > > > says otherwise. > > This might be true for earlycon only, because clocks should be > > setup > > already. > > > > Together with your other mail regarding the imx7ulp fixup, > > commit b4b844930f27bf7019c0bbd8cc575dde32e00ecc should just be > > reverted > > in order to get earlycon running again on imx8qm and imx8qxp, IMHO. > > > > I think you're right, the commit b4b844930f27bf7019c should be > reverted. > > You can send the revert yourself, if you want. If not, let me know. It's already there: https://lore.kernel.org/all/20211124073109.805088-1-alexander.stein@ew.tq-group.com/T/ Thanks, Alexander
diff --git a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml index dc1f0e07cbd4..fa8a602ccb22 100644 --- a/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml +++ b/Documentation/devicetree/bindings/serial/fsl-lpuart.yaml @@ -27,6 +27,10 @@ properties: - items: - const: fsl,imx8qm-lpuart - const: fsl,imx8qxp-lpuart + - items: + - const: fsl,imx8dxl-lpuart + - const: fsl,imx8qxp-lpuart + - const: fsl,imx7ulp-lpuart reg: maxItems: 1
Add i.MX8DXL lpuart compatible to the bindings documentation. Signed-off-by: Abel Vesa <abel.vesa@nxp.com> --- Documentation/devicetree/bindings/serial/fsl-lpuart.yaml | 4 ++++ 1 file changed, 4 insertions(+)