On Sun, Jan 22, 2023 at 09:15:53PM +0200, Dmitry Baryshkov wrote:
> On Sun, 22 Jan 2023 at 20:13, Christian Marangi <ansuelsmth@xxxxxxxxx> wrote:
> >
> > On Sun, Jan 22, 2023 at 07:59:42PM +0200, Dmitry Baryshkov wrote:
> > > On Sun, 22 Jan 2023 at 19:46, Christian Marangi <ansuelsmth@xxxxxxxxx> wrote:
> > > >
> > > > Enlarge opp-supported-hw maximum value. In recent SoC we started
> > > > matching more bit and we currently match mask of 112. The old maximum of
> > > > 7 was good for old SoC that didn't had complex id, but now this is
> > > > limiting and we need to enlarge it to support more variants.
> > > >
> > > > Document all the various mask that can be used and limit them to only
> > > > reasonable values instead of using a generic maximum limit.
> > > >
> > > > Signed-off-by: Christian Marangi <ansuelsmth@xxxxxxxxx>
> > > > ---
> > > > .../bindings/opp/opp-v2-kryo-cpu.yaml | 20 +++++++++++++++++--
> > > > 1 file changed, 18 insertions(+), 2 deletions(-)
> > > >
> > > > diff --git a/Documentation/devicetree/bindings/opp/opp-v2-kryo-cpu.yaml b/Documentation/devicetree/bindings/opp/opp-v2-kryo-cpu.yaml
> > > > index b4947b326773..908cb0d7695a 100644
> > > > --- a/Documentation/devicetree/bindings/opp/opp-v2-kryo-cpu.yaml
> > > > +++ b/Documentation/devicetree/bindings/opp/opp-v2-kryo-cpu.yaml
> > > > @@ -50,12 +50,28 @@ patternProperties:
> > > > opp-supported-hw:
> > > > description: |
> > > > A single 32 bit bitmap value, representing compatible HW.
> > > > - Bitmap:
> > > > + Bitmap for MSM8996 format:
> > > > 0: MSM8996, speedbin 0
> > > > 1: MSM8996, speedbin 1
> > > > 2: MSM8996, speedbin 2
> > > > 3-31: unused
> > > > - maximum: 0x7
> > > > +
> > > > + Bitmap for MSM8996 later revision format:
> > > > + 0: MSM8996, speedbin 0
> > > > + 1: MSM8996, speedbin 1
> > > > + 2: MSM8996, speedbin 2
> > > > + 3: always set
> > >
> > > This is used for speedbin 3
> > >
> >
> > Is it right that 4 bit speedbin is only introduced later? Cause looking
> > at the current opp-supported-hw for MSM8996SG and MSM8996 originally
> > (and based on what this Documentation say) there were only 3 bit and
> > then they started using a 4th bit. Just asking if it's ok to keep the
> > bitmap split or i should just merge it.
>
> I don't think I got the question, please excuse me. Historically
> msm8996.dtsi used 0xff as 'all possible platforms' value for the GPU
> tables. Lately I fixed the CPU tables, added support for speed-bin 3.
> However I left these 0xff in GPU opp table intact. I don't remember if
> there was any reason behind that. Anyway this bit isn't always set, it
> is set only for the entries which should be selected for MSM8996 speed
> bin 3.
>
Question is: Should I merge the 2 format to something like?
A single 32 bit bitmap value, representing compatible HW.
Bitmap for MSM8996 format:
0: MSM8996, speedbin 0
1: MSM8996, speedbin 1
2: MSM8996, speedbin 2
3: MSM8996, speedbin 3
4-31: unused
Bitmap for MSM8996SG format (speedbin shifted of 4 left):
0-3: unused
4: MSM8996SG, speedbin 0
5: MSM8996SG, speedbin 1
6: MSM8996SG, speedbin 2
7-31: unused
Or just replace the 'Always set' to speedbin 3?
(by the looks of it seems they started blowing speedbin 3 only in later
revision and initialy there were only 3 speedbin bit)
> >
> > > > + 4-31: unused
> > > > +
> > > > + Bitmap for MSM8996SG format (speedbin shifted of 4 left):
> > > > + 0-3: unused
> > > > + 4: MSM8996SG, speedbin 0
> > > > + 5: MSM8996SG, speedbin 1
> > > > + 6: MSM8996SG, speedbin 2
> > > > + 7-31: unused
> > > > + enum: [0x1, 0x2, 0x3, 0x4, 0x7,
> > > > + 0x9, 0xd, 0xe, 0xf,
> > > > + 0x10, 0x20, 0x30, 0x70]
> > > >
> > > > clock-latency-ns: true
> > > >
> > > > --
> > > > 2.38.1
> > > >
> >
> > --
> > Ansuel
>
>
>
> --
> With best wishes
> Dmitry
--
Ansuel
