On Tue, 2026-08-18 at 23:20 +0200, Cédric Le Goater wrote:
> On 8/18/26 03:43, Andrew Jeffery wrote:
> > Currently, booting Linux on the ast2600-evb machine via -kernel results
> > in a lot of warnings about failing to apply pinmux configuration when
> > binding drivers:
> >
> > [ 7.339307][ T1] Creating 5 MTD partitions on "bmc":
> > [ 7.340402][ T1] 0x000000000000-0x0000000e0000 : "u-boot"
> > [ 7.513755][ T1] 0x0000000e0000-0x000000100000 : "u-boot-env"
> > [ 7.664219][ T1] 0x000000100000-0x000000a00000 : "kernel"
> > [ 7.826974][ T1] 0x000000a00000-0x000002a00000 : "rofs"
> > [ 8.026666][ T1] 0x000002a00000-0x000004000000 : "rwfs"
> > [ 8.201612][ T1] spi-aspeed-smc 1e630000.spi: error -EPERM:
> > Error applying setting, reverse things back
> > [ 8.370835][ T1] spi-aspeed-smc 1e630000.spi: Calibration area
> > too uniform, using low speed
> > [ 8.371131][ T1] spi-aspeed-smc 1e630000.spi: Force to dts
> > configuration 100000kHz.
> > [ 8.371500][ T1] spi-aspeed-smc 1e630000.spi: CE0 read
> > buswidth:4 [0x406c0741]
> > [ 8.766797][ T1] mdio-aspeed 1e650000.mdio: error -EPERM: Error
> > applying setting, reverse things back
> > [ 8.906008][ T1] mdio-aspeed 1e650008.mdio: error -EPERM: Error
> > applying setting, reverse things back
> > [ 9.039240][ T1] mdio-aspeed 1e650010.mdio: error -EPERM: Error
> > applying setting, reverse things back
> > [ 9.147942][ T1] mdio-aspeed 1e650018.mdio: error -EPERM: Error
> > applying setting, reverse things back
> >
> > The pinmux configuration fails because the SCU is not unlocked.
> >
> > Set both SCU protection keys for the AST2600. Further, set them
> > such that they behave accordance with the read value documented
> > by the datasheet. The new arrangement now upholds the comment in
> > hw/arm/aspeed.c regarding passing -kernel on the qemu command-line.
> >
> > [ 7.568170][ T1] Creating 5 MTD partitions on "bmc":
> > [ 7.569330][ T1] 0x000000000000-0x0000000e0000 : "u-boot"
> > [ 7.740372][ T1] 0x0000000e0000-0x000000100000 : "u-boot-env"
> > [ 7.883570][ T1] 0x000000100000-0x000000a00000 : "kernel"
> > [ 8.041004][ T1] 0x000000a00000-0x000002a00000 : "rofs"
> > [ 8.196824][ T1] 0x000002a00000-0x000004000000 : "rwfs"
> > [ 8.547037][ T1] spi-aspeed-smc 1e630000.spi: Calibration area
> > too uniform, using low speed
> > [ 8.547320][ T1] spi-aspeed-smc 1e630000.spi: Force to dts
> > configuration 100000kHz.
> > [ 8.547642][ T1] spi-aspeed-smc 1e630000.spi: CE0 read
> > buswidth:4 [0x406c0741]
> > [ 9.494235][ T1] ftgmac100 1e660000.ethernet: Read MAC address
> > 52:54:00:12:34:56 from chip
> > [ 9.531186][ T1] RTL8211E Gigabit Ethernet 1e650000.mdio-1:00:
> > attached PHY driver (mii_bus:phy_addr=1e650000.mdio-1:00,
> >
> > Fixes: e09cf36321f6 ("hw: aspeed_scu: Add AST2600 support")
> > Signed-off-by: Andrew Jeffery <[email protected]>
> > ---
> > I ran into this issue while improving test coverage in my patch review
> > process.
> > ---
> > hw/misc/aspeed_scu.c | 3 ++-
> > 1 file changed, 2 insertions(+), 1 deletion(-)
> >
> > diff --git a/hw/misc/aspeed_scu.c b/hw/misc/aspeed_scu.c
> > index ca93c3699d81..1e3db6d03f91 100644
> > --- a/hw/misc/aspeed_scu.c
> > +++ b/hw/misc/aspeed_scu.c
> > @@ -821,7 +821,8 @@ static void aspeed_ast2600_scu_reset_hold(Object *obj,
> > ResetType type)
> > s->regs[AST2600_SILICON_REV2] = s->silicon_rev;
> > s->regs[AST2600_HW_STRAP1] = s->hw_strap1;
> > s->regs[AST2600_HW_STRAP2] = s->hw_strap2;
> > - s->regs[PROT_KEY] = s->hw_prot_key;
> > + s->regs[AST2600_PROT_KEY] = s->hw_prot_key == ASPEED_SCU_PROT_KEY;
> > + s->regs[AST2600_PROT_KEY2] = s->hw_prot_key == ASPEED_SCU_PROT_KEY;
> > }
> >
> > static void aspeed_2600_scu_class_init(ObjectClass *klass, const void
> > *data)
> >
> > ---
> > base-commit: 9696bf5dc5a5bf0b4a9d05b6cdfe5f13990f97aa
> > change-id: 20260818-dev-ast2600-hw-prot-key-reset-e68eb9f49ee9
> >
> > Best regards,
> > --
> > Andrew Jeffery <[email protected]>
> >
>
> Applied to
>
> https://github.com/legoater/qemu aspeed-next
>
> Thanks,
>
> C.
Thanks Cédric!
