Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ 1 file changed, 2 insertions(+)
Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK
boards. Both boards are built from a base board paired with a SoM which
is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs,
LPDDR, eMMC and UFS.
Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com>
---
Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++
1 file changed, 2 insertions(+)
diff --git a/Documentation/devicetree/bindings/arm/qcom.yaml b/Documentation/devicetree/bindings/arm/qcom.yaml
index 2741c07e9f41..28a18254b4b7 100644
--- a/Documentation/devicetree/bindings/arm/qcom.yaml
+++ b/Documentation/devicetree/bindings/arm/qcom.yaml
@@ -63,6 +63,8 @@ properties:
- items:
- enum:
+ - qcom,eliza-cqm-evk
+ - qcom,eliza-cqs-evk
- qcom,eliza-mtp
- const: qcom,eliza
---
base-commit: e98d21c170b01ddef366f023bbfcf6b31509fa83
change-id: 20260514-eliza-bindings-evk-1f05f1c699d5
Best regards,
--
Abel Vesa <abel.vesa@oss.qualcomm.com>On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: > Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK > boards. Both boards are built from a base board paired with a SoM which > is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, > LPDDR, eMMC and UFS. > > Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> > --- > Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/Documentation/devicetree/bindings/arm/qcom.yaml b/Documentation/devicetree/bindings/arm/qcom.yaml > index 2741c07e9f41..28a18254b4b7 100644 > --- a/Documentation/devicetree/bindings/arm/qcom.yaml > +++ b/Documentation/devicetree/bindings/arm/qcom.yaml > @@ -63,6 +63,8 @@ properties: > > - items: > - enum: > + - qcom,eliza-cqm-evk > + - qcom,eliza-cqs-evk When we talked privately, I asked to mirror Shikra approach. But you did not - it's different from what Shikra sent. Shikra received my comments - but you did not mirror these, either. Basically you went with third approach... well, I think this is the same case as in Shikra, thus comments from Shikra apply here (at least what I expect to see). Also, another reason is that this should be consistent. Best regards, Krzysztof
On 26-05-16 11:58:56, Krzysztof Kozlowski wrote: > On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: > > Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK > > boards. Both boards are built from a base board paired with a SoM which > > is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, > > LPDDR, eMMC and UFS. > > > > Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> > > --- > > Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ > > 1 file changed, 2 insertions(+) > > > > diff --git a/Documentation/devicetree/bindings/arm/qcom.yaml b/Documentation/devicetree/bindings/arm/qcom.yaml > > index 2741c07e9f41..28a18254b4b7 100644 > > --- a/Documentation/devicetree/bindings/arm/qcom.yaml > > +++ b/Documentation/devicetree/bindings/arm/qcom.yaml > > @@ -63,6 +63,8 @@ properties: > > > > - items: > > - enum: > > + - qcom,eliza-cqm-evk > > + - qcom,eliza-cqs-evk > > When we talked privately, I asked to mirror Shikra approach. But you did > not - it's different from what Shikra sent. Because I think it is wrong to describe the SoM since it can't be used without a base board. Listing SoMs is useless, IMO. But maybe I'm wrong. So my reason for doing this is basically reducing the list to only describe entire setups: MTP, CQM EVK and CQS EVK. No SoMs. > > Shikra received my comments - but you did not mirror these, either. > Basically you went with third approach... well, I think this is the same > case as in Shikra, thus comments from Shikra apply here (at least what > I expect to see). Also, another reason is that this should be > consistent. I did read go through your comments, but giving the rationale I described above. I did realize just now that I misunderstood your comment on Shikra patchset. Anyway, will rework according to your suggestion. Thanks for reviewing!
On 16/05/2026 12:27, Abel Vesa wrote: > On 26-05-16 11:58:56, Krzysztof Kozlowski wrote: >> On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: >>> Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK >>> boards. Both boards are built from a base board paired with a SoM which >>> is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, >>> LPDDR, eMMC and UFS. >>> >>> Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> >>> --- >>> Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ >>> 1 file changed, 2 insertions(+) >>> >>> diff --git a/Documentation/devicetree/bindings/arm/qcom.yaml b/Documentation/devicetree/bindings/arm/qcom.yaml >>> index 2741c07e9f41..28a18254b4b7 100644 >>> --- a/Documentation/devicetree/bindings/arm/qcom.yaml >>> +++ b/Documentation/devicetree/bindings/arm/qcom.yaml >>> @@ -63,6 +63,8 @@ properties: >>> >>> - items: >>> - enum: >>> + - qcom,eliza-cqm-evk >>> + - qcom,eliza-cqs-evk >> >> When we talked privately, I asked to mirror Shikra approach. But you did >> not - it's different from what Shikra sent. > > Because I think it is wrong to describe the SoM since it can't be used > without a base board. Listing SoMs is useless, IMO. But maybe I'm wrong. > > So my reason for doing this is basically reducing the list to only > describe entire setups: MTP, CQM EVK and CQS EVK. No SoMs. That's really unique statement. All or most of other vendors, including the most popular for SoMs - NXP, use it. But of course existing practice is not the sole reason. The reason is that it is the same useful as having SoC compatible. You cannot use SoC without base board, then why do you have SoC compatible? To allow quirks for the SoC. And here to allow quirks for the SoM, since it is a dedicated and specific hardware component. Best regards, Krzysztof
On 26-05-16 12:30:41, Krzysztof Kozlowski wrote: > On 16/05/2026 12:27, Abel Vesa wrote: > > On 26-05-16 11:58:56, Krzysztof Kozlowski wrote: > >> On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: > >>> Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK > >>> boards. Both boards are built from a base board paired with a SoM which > >>> is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, > >>> LPDDR, eMMC and UFS. > >>> > >>> Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> > >>> --- > >>> Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ > >>> 1 file changed, 2 insertions(+) > >>> > >>> diff --git a/Documentation/devicetree/bindings/arm/qcom.yaml b/Documentation/devicetree/bindings/arm/qcom.yaml > >>> index 2741c07e9f41..28a18254b4b7 100644 > >>> --- a/Documentation/devicetree/bindings/arm/qcom.yaml > >>> +++ b/Documentation/devicetree/bindings/arm/qcom.yaml > >>> @@ -63,6 +63,8 @@ properties: > >>> > >>> - items: > >>> - enum: > >>> + - qcom,eliza-cqm-evk > >>> + - qcom,eliza-cqs-evk > >> > >> When we talked privately, I asked to mirror Shikra approach. But you did > >> not - it's different from what Shikra sent. > > > > Because I think it is wrong to describe the SoM since it can't be used > > without a base board. Listing SoMs is useless, IMO. But maybe I'm wrong. > > > > So my reason for doing this is basically reducing the list to only > > describe entire setups: MTP, CQM EVK and CQS EVK. No SoMs. > > That's really unique statement. All or most of other vendors, including > the most popular for SoMs - NXP, use it. But of course existing practice > is not the sole reason. > > The reason is that it is the same useful as having SoC compatible. You > cannot use SoC without base board, then why do you have SoC compatible? > To allow quirks for the SoC. And here to allow quirks for the SoM, since > it is a dedicated and specific hardware component. Yeah, makes sense.
On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: > Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK > boards. Both boards are built from a base board paired with a SoM which > is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, > LPDDR, eMMC and UFS. > > Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> > --- > Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ > 1 file changed, 2 insertions(+) Heh, I noticed only after sending email - where is any user of that? I expressed it many times - we do not take bindings without users. Best regards, Krzysztof
On 26-05-16 11:59:37, Krzysztof Kozlowski wrote: > On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: > > Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK > > boards. Both boards are built from a base board paired with a SoM which > > is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, > > LPDDR, eMMC and UFS. > > > > Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> > > --- > > Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ > > 1 file changed, 2 insertions(+) > > Heh, I noticed only after sending email - where is any user of that? > > I expressed it many times - we do not take bindings without users. Well, my intention for the DT patchset was to include the eliza-cqm-som.dtsi and eliza-cqm-ek.dts as well, basically having the eliza-cqm-som.dtsi include the eliza-cqs-som.dtsi and later on add the modem (and the rest of the CQM) in there. But then, I realized that I don't know exactly the differences between CQM and CQS, specifically the SoM related ones. So I decided to send without the CQM entirely. Anyway, I'll respin this patch with CQM dropped. Thanks for reviewing!
On 16/05/2026 12:34, Abel Vesa wrote: > On 26-05-16 11:59:37, Krzysztof Kozlowski wrote: >> On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: >>> Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK >>> boards. Both boards are built from a base board paired with a SoM which >>> is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, >>> LPDDR, eMMC and UFS. >>> >>> Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> >>> --- >>> Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ >>> 1 file changed, 2 insertions(+) >> >> Heh, I noticed only after sending email - where is any user of that? >> >> I expressed it many times - we do not take bindings without users. > > Well, my intention for the DT patchset was to include the eliza-cqm-som.dtsi > and eliza-cqm-ek.dts as well, basically having the eliza-cqm-som.dtsi > include the eliza-cqs-som.dtsi and later on add the modem (and the rest > of the CQM) in there. > > But then, I realized that I don't know exactly the differences between > CQM and CQS, specifically the SoM related ones. > > So I decided to send without the CQM entirely. But it is not about CQM. There is no user of any of these - neither CQS nor CQM. And still whatever intention you had, it cannot bypass the rules - we do not take bindings without users. They make almost no sense to the project, because the goal is not to document entire world's hardware in the DT. Best regards, Krzysztof
On 26-05-16 12:37:03, Krzysztof Kozlowski wrote: > On 16/05/2026 12:34, Abel Vesa wrote: > > On 26-05-16 11:59:37, Krzysztof Kozlowski wrote: > >> On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: > >>> Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK > >>> boards. Both boards are built from a base board paired with a SoM which > >>> is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, > >>> LPDDR, eMMC and UFS. > >>> > >>> Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> > >>> --- > >>> Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ > >>> 1 file changed, 2 insertions(+) > >> > >> Heh, I noticed only after sending email - where is any user of that? > >> > >> I expressed it many times - we do not take bindings without users. > > > > Well, my intention for the DT patchset was to include the eliza-cqm-som.dtsi > > and eliza-cqm-ek.dts as well, basically having the eliza-cqm-som.dtsi > > include the eliza-cqs-som.dtsi and later on add the modem (and the rest > > of the CQM) in there. > > > > But then, I realized that I don't know exactly the differences between > > CQM and CQS, specifically the SoM related ones. > > > > So I decided to send without the CQM entirely. > > But it is not about CQM. There is no user of any of these - neither CQS > nor CQM. I'm not sure I understand this one. Here is the CQS user: https://lore.kernel.org/all/20260515-eliza-dts-qcs-evk-v1-2-7169d78a33e1@oss.qualcomm.com/ > > And still whatever intention you had, it cannot bypass the rules - we do > not take bindings without users. They make almost no sense to the > project, because the goal is not to document entire world's hardware in > the DT. I agree. I should've replied back to this thread to say it should be ignored as I should send a v2 without the CQM.
On 16/05/2026 13:12, Abel Vesa wrote: > On 26-05-16 12:37:03, Krzysztof Kozlowski wrote: >> On 16/05/2026 12:34, Abel Vesa wrote: >>> On 26-05-16 11:59:37, Krzysztof Kozlowski wrote: >>>> On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: >>>>> Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK >>>>> boards. Both boards are built from a base board paired with a SoM which >>>>> is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, >>>>> LPDDR, eMMC and UFS. >>>>> >>>>> Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> >>>>> --- >>>>> Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ >>>>> 1 file changed, 2 insertions(+) >>>> >>>> Heh, I noticed only after sending email - where is any user of that? >>>> >>>> I expressed it many times - we do not take bindings without users. >>> >>> Well, my intention for the DT patchset was to include the eliza-cqm-som.dtsi >>> and eliza-cqm-ek.dts as well, basically having the eliza-cqm-som.dtsi >>> include the eliza-cqs-som.dtsi and later on add the modem (and the rest >>> of the CQM) in there. >>> >>> But then, I realized that I don't know exactly the differences between >>> CQM and CQS, specifically the SoM related ones. >>> >>> So I decided to send without the CQM entirely. >> >> But it is not about CQM. There is no user of any of these - neither CQS >> nor CQM. > > I'm not sure I understand this one. Here is the CQS user: > https://lore.kernel.org/all/20260515-eliza-dts-qcs-evk-v1-2-7169d78a33e1@oss.qualcomm.com/ Why is that posted separately? This breaks every practice we had, including all my recent comments complains on internal chats and in public to Nord, Shikra and IPQ upstreaming about grouping patches. :/ Best regards, Krzysztof
On 26-05-16 13:22:19, Krzysztof Kozlowski wrote: > On 16/05/2026 13:12, Abel Vesa wrote: > > On 26-05-16 12:37:03, Krzysztof Kozlowski wrote: > >> On 16/05/2026 12:34, Abel Vesa wrote: > >>> On 26-05-16 11:59:37, Krzysztof Kozlowski wrote: > >>>> On Fri, May 15, 2026 at 02:13:31PM +0300, Abel Vesa wrote: > >>>>> Document the compatible strings for the Qualcomm Eliza CQM and CQS EVK > >>>>> boards. Both boards are built from a base board paired with a SoM which > >>>>> is populated with either CQ7790M or CQ7790S (Eliza variants), PMICs, > >>>>> LPDDR, eMMC and UFS. > >>>>> > >>>>> Signed-off-by: Abel Vesa <abel.vesa@oss.qualcomm.com> > >>>>> --- > >>>>> Documentation/devicetree/bindings/arm/qcom.yaml | 2 ++ > >>>>> 1 file changed, 2 insertions(+) > >>>> > >>>> Heh, I noticed only after sending email - where is any user of that? > >>>> > >>>> I expressed it many times - we do not take bindings without users. > >>> > >>> Well, my intention for the DT patchset was to include the eliza-cqm-som.dtsi > >>> and eliza-cqm-ek.dts as well, basically having the eliza-cqm-som.dtsi > >>> include the eliza-cqs-som.dtsi and later on add the modem (and the rest > >>> of the CQM) in there. > >>> > >>> But then, I realized that I don't know exactly the differences between > >>> CQM and CQS, specifically the SoM related ones. > >>> > >>> So I decided to send without the CQM entirely. > >> > >> But it is not about CQM. There is no user of any of these - neither CQS > >> nor CQM. > > > > I'm not sure I understand this one. Here is the CQS user: > > https://lore.kernel.org/all/20260515-eliza-dts-qcs-evk-v1-2-7169d78a33e1@oss.qualcomm.com/ > > Why is that posted separately? This breaks every practice we had, > including all my recent comments complains on internal chats and in > public to Nord, Shikra and IPQ upstreaming about grouping patches. :/ Yeah, my bad. It should've been sent together with the DT. Will include it there when respinning that patchset. Sorry for the mess.
© 2016 - 2026 Red Hat, Inc.