Re: [PATCH v2] phy: qualcomm: phy-qcom-eusb2-repeater: Fix SMB Power leakage issue
From: Krishna Kurapati
Date: Thu Nov 12 2026 - 04:29:52 EST
On 11/12/2026 2:37 PM, Vinod Koul wrote:
On 12-11-26, 14:07, Krishna Kurapati wrote:
On Glymur, it is observed that although the exit call is done, there is a
power leakage seen on SMB2370 charger (which embeds the eUSB2 module as
well). As per recommendation of HW team, disable the repeater during the
exit call to prevent this power leakage.
Reviewed-by: Konrad Dybcio <konrad.dybcio@xxxxxxxxxxxxxxxx>
Reviewed-by: Manivannan Sadhasivam <manivannan.sadhasivam@xxxxxxxxxxxxxxxx>
Reviewed-by: Abel Vesa <abel.vesa@xxxxxxxxxxxxxxxx>
Signed-off-by: Krishna Kurapati <krishna.kurapati@xxxxxxxxxxxxxxxx>
---
Changes in v2:
- Rebased on top of phy-next since [1] removes the base variable in rptr struct
- Kept the RB Tags intact since its a rebase
Link to v1:
https://lore.kernel.org/all/20260917-smb-repeater-power-leakage-v1-1-6d7311c3c03e@xxxxxxxxxxxxxxxx/
Compile tested v2.
[1]: https://lore.kernel.org/all/20260721093626.96264-10-angelogioacchino.delregno@xxxxxxxxxxxxx/
---
drivers/phy/qualcomm/phy-qcom-eusb2-repeater.c | 2 ++
1 file changed, 2 insertions(+)
diff --git a/drivers/phy/qualcomm/phy-qcom-eusb2-repeater.c b/drivers/phy/qualcomm/phy-qcom-eusb2-repeater.c
index 89bc1c89fc85..607b5a5a1713 100644
--- a/drivers/phy/qualcomm/phy-qcom-eusb2-repeater.c
+++ b/drivers/phy/qualcomm/phy-qcom-eusb2-repeater.c
@@ -227,7 +227,9 @@ static int eusb2_repeater_set_mode(struct phy *phy,
static int eusb2_repeater_exit(struct phy *phy)
{
struct eusb2_repeater *rptr = phy_get_drvdata(phy);
+ struct regmap *regmap = rptr->regmap;
+ regmap_write(regmap, EUSB2_EN_CTL1, 0);
return regulator_bulk_disable(rptr->cfg->num_vregs, rptr->vregs);
This fails to build:
drivers/phy/qualcomm/phy-qcom-eusb2-repeater.c: In function ‘eusb2_repeater_exit’:
drivers/phy/qualcomm/phy-qcom-eusb2-repeater.c:231:24: error: ‘struct eusb2_repeater’ has no member named ‘base’
Hi Vinod,
When [1] got merged it removed the base pointer. Hence v1 of this patch fails with error saying that base is not present. But v2 has been rebased. This v1 gives the issue but not v2. I see that you have applied v1. Can you revert that and take in v2.
[1]: https://lore.kernel.org/all/20260721093626.96264-10-angelogioacchino.delregno@xxxxxxxxxxxxx/
Regards,
Krishna,