From: Kory Maincent (Dent Project) <kory.maincent@bootlin.com>
Detect when PSE hardware is already configured (user byte == 42) and
skip hardware initialization to prevent power interruption to connected
devices during system reboots.
Previously, the driver would always reconfigure the PSE hardware on
probe, causing a port matrix reflash that resulted in temporary power
loss to all connected devices. This change maintains power continuity
by preserving existing configuration when the PSE has been previously
initialized.
Signed-off-by: Kory Maincent <kory.maincent@bootlin.com>
---
drivers/net/pse-pd/pd692x0.c | 35 ++++++++++++++++++++++++++++++++---
1 file changed, 32 insertions(+), 3 deletions(-)
diff --git a/drivers/net/pse-pd/pd692x0.c b/drivers/net/pse-pd/pd692x0.c
index 782b1abf94cb..134435e90073 100644
--- a/drivers/net/pse-pd/pd692x0.c
+++ b/drivers/net/pse-pd/pd692x0.c
@@ -30,6 +30,8 @@
#define PD692X0_FW_MIN_VER 5
#define PD692X0_FW_PATCH_VER 5
+#define PD692X0_USER_BYTE 42
+
enum pd692x0_fw_state {
PD692X0_FW_UNKNOWN,
PD692X0_FW_OK,
@@ -80,6 +82,7 @@ enum {
PD692X0_MSG_GET_PORT_PARAM,
PD692X0_MSG_GET_POWER_BANK,
PD692X0_MSG_SET_POWER_BANK,
+ PD692X0_MSG_SET_USER_BYTE,
/* add new message above here */
PD692X0_MSG_CNT
@@ -103,6 +106,7 @@ struct pd692x0_priv {
bool last_cmd_key;
unsigned long last_cmd_key_time;
+ bool cfg_saved;
enum ethtool_c33_pse_admin_state admin_state[PD692X0_MAX_PIS];
struct regulator_dev *manager_reg[PD692X0_MAX_MANAGERS];
int manager_pw_budget[PD692X0_MAX_MANAGERS];
@@ -193,6 +197,12 @@ static const struct pd692x0_msg pd692x0_msg_template_list[PD692X0_MSG_CNT] = {
.key = PD692X0_KEY_CMD,
.sub = {0x07, 0x0b, 0x57},
},
+ [PD692X0_MSG_SET_USER_BYTE] = {
+ .key = PD692X0_KEY_PRG,
+ .sub = {0x41, PD692X0_USER_BYTE},
+ .data = {0x4e, 0x4e, 0x4e, 0x4e,
+ 0x4e, 0x4e, 0x4e, 0x4e},
+ },
};
static u8 pd692x0_build_msg(struct pd692x0_msg *msg, u8 echo)
@@ -1233,6 +1243,15 @@ static void pd692x0_managers_free_pw_budget(struct pd692x0_priv *priv)
}
}
+static int
+pd692x0_save_user_byte(struct pd692x0_priv *priv)
+{
+ struct pd692x0_msg msg, buf;
+
+ msg = pd692x0_msg_template_list[PD692X0_MSG_SET_USER_BYTE];
+ return pd692x0_sendrecv_msg(priv, &msg, &buf);
+}
+
static int pd692x0_setup_pi_matrix(struct pse_controller_dev *pcdev)
{
struct pd692x0_priv *priv = to_pd692x0_priv(pcdev);
@@ -1268,9 +1287,16 @@ static int pd692x0_setup_pi_matrix(struct pse_controller_dev *pcdev)
if (ret)
goto err_managers_req_pw;
- ret = pd692x0_hw_conf_init(priv);
- if (ret)
- goto err_managers_req_pw;
+ /* Do not init the conf if it is already saved */
+ if (!priv->cfg_saved) {
+ ret = pd692x0_hw_conf_init(priv);
+ if (ret)
+ goto err_managers_req_pw;
+
+ ret = pd692x0_save_user_byte(priv);
+ if (ret)
+ goto err_managers_req_pw;
+ }
pd692x0_of_put_managers(priv, manager);
kfree(manager);
@@ -1793,6 +1819,9 @@ static int pd692x0_i2c_probe(struct i2c_client *client)
}
}
+ if (buf.data[2] == PD692X0_USER_BYTE)
+ priv->cfg_saved = true;
+
priv->np = dev->of_node;
priv->pcdev.nr_lines = PD692X0_MAX_PIS;
priv->pcdev.owner = THIS_MODULE;
--
2.43.0On Mon, Oct 13, 2025 at 04:05:33PM +0200, Kory Maincent wrote: > From: Kory Maincent (Dent Project) <kory.maincent@bootlin.com> > > Detect when PSE hardware is already configured (user byte == 42) and > skip hardware initialization to prevent power interruption to connected > devices during system reboots. > > Previously, the driver would always reconfigure the PSE hardware on > probe, causing a port matrix reflash that resulted in temporary power > loss to all connected devices. This change maintains power continuity > by preserving existing configuration when the PSE has been previously > initialized. > > Signed-off-by: Kory Maincent <kory.maincent@bootlin.com> Hi Kory, Perhaps I'm over thinking things here. But I'm wondering what provision there is for a situation whereby: 1. The driver configures the device 2. A reboot occurs 2. The (updated) driver wants to (re)configure the device with a different configuration, say because it turns out there was a bug in or enhancement to the procedure at 1. ...
On Tue, 14 Oct 2025 09:58:56 +0100 Simon Horman <horms@kernel.org> wrote: > On Mon, Oct 13, 2025 at 04:05:33PM +0200, Kory Maincent wrote: > > From: Kory Maincent (Dent Project) <kory.maincent@bootlin.com> > > > > Detect when PSE hardware is already configured (user byte == 42) and > > skip hardware initialization to prevent power interruption to connected > > devices during system reboots. > > > > Previously, the driver would always reconfigure the PSE hardware on > > probe, causing a port matrix reflash that resulted in temporary power > > loss to all connected devices. This change maintains power continuity > > by preserving existing configuration when the PSE has been previously > > initialized. > > > > Signed-off-by: Kory Maincent <kory.maincent@bootlin.com> > > Hi Kory, > > Perhaps I'm over thinking things here. But I'm wondering > what provision there is for a situation whereby: > > 1. The driver configures the device > 2. A reboot occurs > 2. The (updated) driver wants to (re)configure the device > with a different configuration, say because it turns > out there was a bug in or enhancement to the procedure at 1. > > ... You have to find a way to turn off the power supply of the PSE controller. As adding a devlink uAPI for this was not accepted, a hard reset for the PSE controller is the only way to clean the user byte register and (re)configure the controller at boot time. Regards, -- Köry Maincent, Bootlin Embedded Linux and kernel engineering https://bootlin.com
On Tue, Oct 14, 2025 at 11:56:59AM +0200, Kory Maincent wrote: > On Tue, 14 Oct 2025 09:58:56 +0100 > Simon Horman <horms@kernel.org> wrote: > > > On Mon, Oct 13, 2025 at 04:05:33PM +0200, Kory Maincent wrote: > > > From: Kory Maincent (Dent Project) <kory.maincent@bootlin.com> > > > > > > Detect when PSE hardware is already configured (user byte == 42) and > > > skip hardware initialization to prevent power interruption to connected > > > devices during system reboots. > > > > > > Previously, the driver would always reconfigure the PSE hardware on > > > probe, causing a port matrix reflash that resulted in temporary power > > > loss to all connected devices. This change maintains power continuity > > > by preserving existing configuration when the PSE has been previously > > > initialized. > > > > > > Signed-off-by: Kory Maincent <kory.maincent@bootlin.com> > > > > Hi Kory, > > > > Perhaps I'm over thinking things here. But I'm wondering > > what provision there is for a situation whereby: > > > > 1. The driver configures the device > > 2. A reboot occurs > > 2. The (updated) driver wants to (re)configure the device > > with a different configuration, say because it turns > > out there was a bug in or enhancement to the procedure at 1. > > > > ... > > You have to find a way to turn off the power supply of the PSE > controller. As adding a devlink uAPI for this was not accepted, a hard > reset for the PSE controller is the only way to clean the user byte > register and (re)configure the controller at boot time. Ok, as long as there is a way. Maybe worth mentioning in the commit message. But either way this patch looks good to me. Reviewed-by: Simon Horman <horms@kernel.org>
© 2016 - 2025 Red Hat, Inc.