| Message ID | 20230823171208.18382-1-athierry@redhat.com |
|---|---|
| Headers | show |
| Series | phy: qcom-qmp-usb: fix initialization of PCS_USB | expand |
On 23/08/2023 18:12, Adrien Thierry wrote: > This series attempts at making sure PCS_USB registers are properly > initialized. Please note I don't have access to the qmp phy datasheet and > only reasoned from the code, so I appreciate a double check to make sure > I'm not messing with the registers. > > Adrien Thierry (2): > phy: qcom-qmp-usb: initialize PCS_USB registers > phy: qcom-qmp-usb: split PCS_USB init table for sc8280xp and sa8775p > > drivers/phy/qualcomm/phy-qcom-qmp-usb.c | 22 +++++++++++++++++----- > 1 file changed, 17 insertions(+), 5 deletions(-) > How is this series tested ? What are the before/after results/effects ? --- bod
Hi Bryan, On Thu, Aug 24, 2023 at 10:15:32AM +0100, Bryan O'Donoghue wrote: > On 23/08/2023 18:12, Adrien Thierry wrote: > > This series attempts at making sure PCS_USB registers are properly > > initialized. Please note I don't have access to the qmp phy datasheet and > > only reasoned from the code, so I appreciate a double check to make sure > > I'm not messing with the registers. > > > > Adrien Thierry (2): > > phy: qcom-qmp-usb: initialize PCS_USB registers > > phy: qcom-qmp-usb: split PCS_USB init table for sc8280xp and sa8775p > > > > drivers/phy/qualcomm/phy-qcom-qmp-usb.c | 22 +++++++++++++++++----- > > 1 file changed, 17 insertions(+), 5 deletions(-) > > > > How is this series tested ? What are the before/after results/effects ? > I tested this series on the sa8775p. AFAICT there's no noticeable change before/after the patch series: lsusb and dmesg output are the same. USB is still working properly. I don't know what those PCS_USB registers do exactly on the qmp PHY and I don't have access to docs, so it's hard for me to tell the impact of them being initialized vs not. > --- > bod Best, Adrien
On Wed, 23 Aug 2023 at 20:12, Adrien Thierry <athierry@redhat.com> wrote: > > Currently, PCS_USB registers that have their initialization data in a > pcs_usb_tbl table are never initialized. Fix that. > > Fixes: fc64623637da ("phy: qcom-qmp-combo,usb: add support for separate PCS_USB region") > Signed-off-by: Adrien Thierry <athierry@redhat.com> > --- > drivers/phy/qualcomm/phy-qcom-qmp-usb.c | 2 ++ > 1 file changed, 2 insertions(+) > > diff --git a/drivers/phy/qualcomm/phy-qcom-qmp-usb.c b/drivers/phy/qualcomm/phy-qcom-qmp-usb.c > index 466f0a56c82e..ccbe64f7897e 100644 > --- a/drivers/phy/qualcomm/phy-qcom-qmp-usb.c > +++ b/drivers/phy/qualcomm/phy-qcom-qmp-usb.c > @@ -2233,6 +2233,7 @@ static int qmp_usb_power_on(struct phy *phy) > void __iomem *tx = qmp->tx; > void __iomem *rx = qmp->rx; > void __iomem *pcs = qmp->pcs; > + void __iomem *pcs_usb = qmp->pcs_usb ?: qmp->pcs; > void __iomem *status; > unsigned int val; > int ret; > @@ -2255,6 +2256,7 @@ static int qmp_usb_power_on(struct phy *phy) > } > > qmp_usb_configure(pcs, cfg->pcs_tbl, cfg->pcs_tbl_num); > + qmp_usb_configure(pcs_usb, cfg->pcs_usb_tbl, cfg->pcs_usb_tbl_num); I think we don't need to fallback to pcs here: if there is a separate pcs_usb_tbl, we need a separate pcs_usb region. Just pass qmp->pcs_usb here. > > if (cfg->has_pwrdn_delay) > usleep_range(10, 20); > -- > 2.41.0 >