soc: qcom: socinfo: Add another ID for sc7180

[PATCH] soc: qcom: socinfo: Add another ID for sc7180

Posted by Douglas Anderson 4 years, 1 month ago

It appears the some sc7180 devices, like the one in my
sc7180-trogdor-homestar, report an ID of 407 instead of 425. Add
another ID into the list.

Signed-off-by: Douglas Anderson <dianders@chromium.org>
---

 drivers/soc/qcom/socinfo.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/drivers/soc/qcom/socinfo.c b/drivers/soc/qcom/socinfo.c
index cee579a267a6..2ef1dc2a1dd3 100644
--- a/drivers/soc/qcom/socinfo.c
+++ b/drivers/soc/qcom/socinfo.c
@@ -318,6 +318,7 @@ static const struct soc_id soc_id[] = {
 	{ 396, "IPQ8071A" },
 	{ 402, "IPQ6018" },
 	{ 403, "IPQ6028" },
+	{ 407, "SC7180" },
 	{ 421, "IPQ6000" },
 	{ 422, "IPQ6010" },
 	{ 425, "SC7180" },
-- 
2.36.0.464.gb9c8b46e94-goog

Re: (subset) [PATCH] soc: qcom: socinfo: Add another ID for sc7180

Posted by Bjorn Andersson 4 years, 1 month ago

On Mon, 2 May 2022 17:33:45 -0700, Douglas Anderson wrote:
> It appears the some sc7180 devices, like the one in my
> sc7180-trogdor-homestar, report an ID of 407 instead of 425. Add
> another ID into the list.
> 
> 

Applied, thanks!

[1/1] soc: qcom: socinfo: Add another ID for sc7180
      commit: c35886d605604e0b03cdd835ae3249dc1fe0cc2a

Best regards,
-- 
Bjorn Andersson <bjorn.andersson@linaro.org>

Re: (subset) [PATCH] soc: qcom: socinfo: Add another ID for sc7180

Posted by Doug Anderson 4 years, 1 month ago

Hi,

On Wed, May 4, 2022 at 10:21 AM Bjorn Andersson
<bjorn.andersson@linaro.org> wrote:
>
> On Mon, 2 May 2022 17:33:45 -0700, Douglas Anderson wrote:
> > It appears the some sc7180 devices, like the one in my
> > sc7180-trogdor-homestar, report an ID of 407 instead of 425. Add
> > another ID into the list.
> >
> >
>
> Applied, thanks!
>
> [1/1] soc: qcom: socinfo: Add another ID for sc7180
>       commit: c35886d605604e0b03cdd835ae3249dc1fe0cc2a

Hmm. Did you see the responses from Sai [1] about this? He seemed to
indicate that there might be some issue here because he thought 407
was supposed to be a different SoC. Are we sure we want to land this
patch while we're sorting it out?

[1] https://lore.kernel.org/r/13819b2d-26f0-14f4-9cb9-affb6b18f13d@quicinc.com/

Re: (subset) [PATCH] soc: qcom: socinfo: Add another ID for sc7180

Posted by Bjorn Andersson 4 years, 1 month ago

On Wed 04 May 10:51 PDT 2022, Doug Anderson wrote:

> Hi,
> 
> On Wed, May 4, 2022 at 10:21 AM Bjorn Andersson
> <bjorn.andersson@linaro.org> wrote:
> >
> > On Mon, 2 May 2022 17:33:45 -0700, Douglas Anderson wrote:
> > > It appears the some sc7180 devices, like the one in my
> > > sc7180-trogdor-homestar, report an ID of 407 instead of 425. Add
> > > another ID into the list.
> > >
> > >
> >
> > Applied, thanks!
> >
> > [1/1] soc: qcom: socinfo: Add another ID for sc7180
> >       commit: c35886d605604e0b03cdd835ae3249dc1fe0cc2a
> 
> Hmm. Did you see the responses from Sai [1] about this? He seemed to
> indicate that there might be some issue here because he thought 407
> was supposed to be a different SoC. Are we sure we want to land this
> patch while we're sorting it out?
> 

I stared at the thread but didn't see his reply for some reason.

Congrats on the getting the random SoC in your device ;) As this was the
top patch on drivers-for-5.19, I just popped that off and pushed the
branch again.

Let me know when you've figured the numbering out (and if it turns out
to be the mobile id I wouldn't mind carrying that in the table)

Regards,
Bjorn

> [1] https://lore.kernel.org/r/13819b2d-26f0-14f4-9cb9-affb6b18f13d@quicinc.com/

Re: [PATCH] soc: qcom: socinfo: Add another ID for sc7180

Posted by Sai Prakash Ranjan 4 years, 1 month ago

Hi,

On 5/3/2022 6:03 AM, Douglas Anderson wrote:
> It appears the some sc7180 devices, like the one in my
> sc7180-trogdor-homestar, report an ID of 407 instead of 425. Add
> another ID into the list.
>
> Signed-off-by: Douglas Anderson <dianders@chromium.org>
> ---
>
>   drivers/soc/qcom/socinfo.c | 1 +
>   1 file changed, 1 insertion(+)
>
> diff --git a/drivers/soc/qcom/socinfo.c b/drivers/soc/qcom/socinfo.c
> index cee579a267a6..2ef1dc2a1dd3 100644
> --- a/drivers/soc/qcom/socinfo.c
> +++ b/drivers/soc/qcom/socinfo.c
> @@ -318,6 +318,7 @@ static const struct soc_id soc_id[] = {
>   	{ 396, "IPQ8071A" },
>   	{ 402, "IPQ6018" },
>   	{ 403, "IPQ6028" },
> +	{ 407, "SC7180" },
>   	{ 421, "IPQ6000" },
>   	{ 422, "IPQ6010" },
>   	{ 425, "SC7180" },

Hmm, this ID maps to SM6250 which is a mobile variant. Not sure we should
use it for SC7180 which already has 425 ID assigned, perks of marketing :)

Thanks,
Sai

Re: [PATCH] soc: qcom: socinfo: Add another ID for sc7180

Posted by Doug Anderson 4 years, 1 month ago

Hi,

On Tue, May 3, 2022 at 7:14 PM Sai Prakash Ranjan
<quic_saipraka@quicinc.com> wrote:
>
> Hi,
>
> On 5/3/2022 6:03 AM, Douglas Anderson wrote:
> > It appears the some sc7180 devices, like the one in my
> > sc7180-trogdor-homestar, report an ID of 407 instead of 425. Add
> > another ID into the list.
> >
> > Signed-off-by: Douglas Anderson <dianders@chromium.org>
> > ---
> >
> >   drivers/soc/qcom/socinfo.c | 1 +
> >   1 file changed, 1 insertion(+)
> >
> > diff --git a/drivers/soc/qcom/socinfo.c b/drivers/soc/qcom/socinfo.c
> > index cee579a267a6..2ef1dc2a1dd3 100644
> > --- a/drivers/soc/qcom/socinfo.c
> > +++ b/drivers/soc/qcom/socinfo.c
> > @@ -318,6 +318,7 @@ static const struct soc_id soc_id[] = {
> >       { 396, "IPQ8071A" },
> >       { 402, "IPQ6018" },
> >       { 403, "IPQ6028" },
> > +     { 407, "SC7180" },
> >       { 421, "IPQ6000" },
> >       { 422, "IPQ6010" },
> >       { 425, "SC7180" },
>
> Hmm, this ID maps to SM6250 which is a mobile variant. Not sure we should
> use it for SC7180 which already has 425 ID assigned, perks of marketing :)

That's super weird. Any idea why my strongbad reports 407 then? Is it
a firmware bug, or someone misprogrammed something? What can we do to
make the CPU recognized in strongbad?

...and just for extra info, I believe strongbad actually has the
slightly faster version of sc7180. It has a higher GPU/CPU speed. In
some places I've heard it called sc7185, but in other places I've
heard it called just sc7180 with a faster clock speed...

-Doug

Re: [PATCH] soc: qcom: socinfo: Add another ID for sc7180

Posted by Sai Prakash Ranjan 4 years, 1 month ago

Hi,

On 5/4/2022 9:00 AM, Doug Anderson wrote:
> Hi,
>
> On Tue, May 3, 2022 at 7:14 PM Sai Prakash Ranjan
> <quic_saipraka@quicinc.com> wrote:
>> Hi,
>>
>> On 5/3/2022 6:03 AM, Douglas Anderson wrote:
>>> It appears the some sc7180 devices, like the one in my
>>> sc7180-trogdor-homestar, report an ID of 407 instead of 425. Add
>>> another ID into the list.
>>>
>>> Signed-off-by: Douglas Anderson <dianders@chromium.org>
>>> ---
>>>
>>>    drivers/soc/qcom/socinfo.c | 1 +
>>>    1 file changed, 1 insertion(+)
>>>
>>> diff --git a/drivers/soc/qcom/socinfo.c b/drivers/soc/qcom/socinfo.c
>>> index cee579a267a6..2ef1dc2a1dd3 100644
>>> --- a/drivers/soc/qcom/socinfo.c
>>> +++ b/drivers/soc/qcom/socinfo.c
>>> @@ -318,6 +318,7 @@ static const struct soc_id soc_id[] = {
>>>        { 396, "IPQ8071A" },
>>>        { 402, "IPQ6018" },
>>>        { 403, "IPQ6028" },
>>> +     { 407, "SC7180" },
>>>        { 421, "IPQ6000" },
>>>        { 422, "IPQ6010" },
>>>        { 425, "SC7180" },
>> Hmm, this ID maps to SM6250 which is a mobile variant. Not sure we should
>> use it for SC7180 which already has 425 ID assigned, perks of marketing :)
> That's super weird. Any idea why my strongbad reports 407 then? Is it
> a firmware bug, or someone misprogrammed something? What can we do to
> make the CPU recognized in strongbad?

Let me check with coretech folks as to why the ID is reused.

> ...and just for extra info, I believe strongbad actually has the
> slightly faster version of sc7180. It has a higher GPU/CPU speed. In
> some places I've heard it called sc7185, but in other places I've
> heard it called just sc7180 with a faster clock speed...

The only other variant of SC7180 I see in the document is SC7180P
which has 495 soc_id and is an APQ variant.

Thanks,
Sai