| Message ID | 20240607092140.33112-2-ryan@testtoast.com |
|---|---|
| State | New |
| Headers | show |
| Series | cpufreq: sun50i: add Allwinner H700 speed bin and additional OPPs | expand |
On Fri, 7 Jun 2024 21:20:33 +1200 Ryan Walklin <ryan@testtoast.com> wrote: > Support for the Allwinner H618, H618 and H700 was added to the sun50i > cpufreq-nvmem driver recently [1] however at the time some operating > points supported by the H700 (1.008, 1.032 and 1.512 GHz) and in use by > vendor BSPs were found to be unstable during testing, so the H700 speed > bin and the 1.032 GHz OPP were not included in the mainline driver. > > Retesting with kernel 6.10rc2 (which carries additional fixes for the > driver) now shows stable operation with these points. > > Add the H700 speed bin to the driver. > > Signed-off-by: Ryan Walklin <ryan@testtoast.com> Yes, 0x6c00 is the value for the H700 SoCs in the devices we have seen: Reviewed-by: Andre Przywara <andre.przywara@arm.com> Thanks, Andre > -- > [1] https://lore.kernel.org/linux-sunxi/20240418154408.1740047-1-andre.przywara@arm.com > --- > drivers/cpufreq/sun50i-cpufreq-nvmem.c | 3 +++ > 1 file changed, 3 insertions(+) > > diff --git a/drivers/cpufreq/sun50i-cpufreq-nvmem.c b/drivers/cpufreq/sun50i-cpufreq-nvmem.c > index 0b882765cd66f..969f22aadd950 100644 > --- a/drivers/cpufreq/sun50i-cpufreq-nvmem.c > +++ b/drivers/cpufreq/sun50i-cpufreq-nvmem.c > @@ -91,6 +91,9 @@ static u32 sun50i_h616_efuse_xlate(u32 speedbin) > case 0x5d00: > value = 0; > break; > + case 0x6c00: > + value = 5; > + break; > default: > pr_warn("sun50i-cpufreq-nvmem: unknown speed bin 0x%x, using default bin 0\n", > speedbin & 0xffff);
On 19-06-24, 01:00, Andre Przywara wrote: > On Fri, 7 Jun 2024 21:20:33 +1200 > Ryan Walklin <ryan@testtoast.com> wrote: > > > Support for the Allwinner H618, H618 and H700 was added to the sun50i > > cpufreq-nvmem driver recently [1] however at the time some operating > > points supported by the H700 (1.008, 1.032 and 1.512 GHz) and in use by > > vendor BSPs were found to be unstable during testing, so the H700 speed > > bin and the 1.032 GHz OPP were not included in the mainline driver. > > > > Retesting with kernel 6.10rc2 (which carries additional fixes for the > > driver) now shows stable operation with these points. > > > > Add the H700 speed bin to the driver. > > > > Signed-off-by: Ryan Walklin <ryan@testtoast.com> > > Yes, 0x6c00 is the value for the H700 SoCs in the devices we have seen: > > Reviewed-by: Andre Przywara <andre.przywara@arm.com> Applied. Thanks.
diff --git a/drivers/cpufreq/sun50i-cpufreq-nvmem.c b/drivers/cpufreq/sun50i-cpufreq-nvmem.c index 0b882765cd66f..969f22aadd950 100644 --- a/drivers/cpufreq/sun50i-cpufreq-nvmem.c +++ b/drivers/cpufreq/sun50i-cpufreq-nvmem.c @@ -91,6 +91,9 @@ static u32 sun50i_h616_efuse_xlate(u32 speedbin) case 0x5d00: value = 0; break; + case 0x6c00: + value = 5; + break; default: pr_warn("sun50i-cpufreq-nvmem: unknown speed bin 0x%x, using default bin 0\n", speedbin & 0xffff);
Support for the Allwinner H618, H618 and H700 was added to the sun50i cpufreq-nvmem driver recently [1] however at the time some operating points supported by the H700 (1.008, 1.032 and 1.512 GHz) and in use by vendor BSPs were found to be unstable during testing, so the H700 speed bin and the 1.032 GHz OPP were not included in the mainline driver. Retesting with kernel 6.10rc2 (which carries additional fixes for the driver) now shows stable operation with these points. Add the H700 speed bin to the driver. Signed-off-by: Ryan Walklin <ryan@testtoast.com> -- [1] https://lore.kernel.org/linux-sunxi/20240418154408.1740047-1-andre.przywara@arm.com --- drivers/cpufreq/sun50i-cpufreq-nvmem.c | 3 +++ 1 file changed, 3 insertions(+)