[PATCH] usb: typec: tipd: don't send GAID when probe fails in APP mode

Marian Rotariu <[email protected]> Thu, 6 Aug 2026 11:53:02 +0300
Newsgroups org.kernel.vger.linux-usb,org.kernel.vger.linux-kernel
Message-ID <[email protected]>
When probe fails after the chip was already in APP mode (booted from
EEPROM), the err_reset_controller path unconditionally calls
tps->data->reset(), which for tps25750 sends the GAID 4CC command.
GAID transitions the chip from APP back to PTCH, causing a second
probe attempt that sees PTCH mode and tries to load the firmware from
/lib/firmware, instead of reading the EEPROM as it should.

The err_reset_controller path can be triggered by -EPROBE_DEFER from
the probe sequence. GAID should only be sent when the controller
actually loaded firmware at probe time.

The abnormal behavior was seen on a tps25751d that has an EEPROM
connected to it.

Fixes: d49f90822015 ("usb: typec: tipd: add init and reset functions to tipd_data")
Signed-off-by: Marian Rotariu <[email protected]>
---
 drivers/usb/typec/tipd/core.c | 5 ++++-
 1 file changed, 4 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/typec/tipd/core.c b/drivers/usb/typec/tipd/core.c
index 522f56742aa9..c2b235f3c225 100644
--- a/drivers/usb/typec/tipd/core.c
+++ b/drivers/usb/typec/tipd/core.c
@@ -1783,6 +1783,7 @@ static int tps6598x_probe(struct i2c_client *client)
 	const struct tipd_data *data;
 	struct tps6598x *tps;
 	struct fwnode_handle *fwnode;
+	bool patch_loaded = false;
 	u32 status;
 	u32 vid = 0;
 	int ret;
@@ -1838,6 +1839,7 @@ static int tps6598x_probe(struct i2c_client *client)
 		return ret;
 
 	if (ret == TPS_MODE_PTCH) {
+		patch_loaded = true;
 		ret = tps->data->init(tps);
 		if (ret)
 			return ret;
@@ -1928,7 +1930,8 @@ static int tps6598x_probe(struct i2c_client *client)
 	tps6598x_write64(tps, TPS_REG_INT_MASK1, 0);
 err_reset_controller:
 	/* Reset PD controller to remove any applied patch */
-	tps->data->reset(tps);
+	if (patch_loaded)
+		tps->data->reset(tps);
 
 	return ret;
 }
-- 
2.53.0