Julius Werner has submitted this change and it was merged. ( https://review.coreboot.org/21727 )
Change subject: google/gru: Fix pmu1833_volsel handling ......................................................................
google/gru: Fix pmu1833_volsel handling
RK3399 has a pin that can decide whether GPIO port 1 is driven with 1.8V or 3.0V. We thought this mechanism was disabled by default, but it turns out it wasn't. We want to use that pin as an output GPIO on Scarlet so we need to reconfigure the respective SoC controls before we do that. It seems that we also need to explicitly pinmux the pin away from that special function (to normal GPIO) or weird things happen on some boards.
BUG=b:66534913 TEST=Sprinkled several long udelays, poked test points with a multi-meter on Scarlet.
Change-Id: Ia02cbb4f3b2f14b0d958b84adcddb0c5f4259efa Signed-off-by: Julius Werner jwerner@chromium.org Reviewed-on: https://review.coreboot.org/21727 Tested-by: build bot (Jenkins) no-reply@coreboot.org --- M src/mainboard/google/gru/bootblock.c 1 file changed, 6 insertions(+), 0 deletions(-)
Approvals: Julius Werner: Looks good to me, approved build bot (Jenkins): Verified
diff --git a/src/mainboard/google/gru/bootblock.c b/src/mainboard/google/gru/bootblock.c index a3d3121..162a7e5 100644 --- a/src/mainboard/google/gru/bootblock.c +++ b/src/mainboard/google/gru/bootblock.c @@ -48,6 +48,12 @@ if (IS_ENABLED(CONFIG_BOARD_GOOGLE_SCARLET)) write32(&rk3399_grf->io_vsel, RK_SETBITS(1 << 3));
+ /* Reconfigure GPIO1 from dynamic voltage selection through GPIO0_B1 to + hardcoded 1.8V, and change that pin to a normal GPIO. The TRM says + this is already the power-on reset, but we all know that TRMs lie. */ + write32(&rk3399_pmugrf->soc_con0, RK_SETBITS(1 << 9 | 1 << 8)); + write32(&rk3399_pmugrf->gpio0b_iomux, RK_CLRBITS(3 << 2)); + /* Enable rails powering GPIO blocks, among other things. */ gpio_output(GPIO_P30V_EN, 1); if (!IS_ENABLED(CONFIG_BOARD_GOOGLE_SCARLET))