drivers/usb/typec/tcpm/tcpm.c | 3 +++ 1 file changed, 3 insertions(+)
From: Xu Yang <xu.yang_2@nxp.com>
Commit fa4d0a7cd036 ("usb: typec: tcpm: fix Discover Identity retry loop
on Control Message reject") advances vdm_discovery_state to
VDM_DISCOVERY_COMPLETE when a CMDT_RSP_NAK is received, but it does not
advance the state when a non-modal port returns PD_CTRL_NOT_SUPP.
This can re-arm another vdm_discovery_work if the port is going to
respond with PD_CTRL_NOT_SUPP to the partner.
tcpm_pd_handle_msg(port, PD_MSG_CTRL_NOT_SUPP, NONE_AMS);
tcpm_queue_message();
mod_tcpm_delayed_work();
tcpm_state_machine_work()
tcpm_send_queued_message() -> send PD_MSG_CTRL_NOT_SUPP
run_state_machine()
case SRC_READY:
case SNK_READY:
mod_vdm_discovery_delayed_work(port, 0);
If both the port and its partner do not support modal operation, they
will send DISCOVER_IDENTITY to each other infinitely.
Advance vdm_discovery_state to VDM_DISCOVERY_COMPLETE in this path as
well, mirroring the existing handling for the SVDM CMDT_RSP_NAK case.
Fixes: fa4d0a7cd036 ("usb: typec: tcpm: fix Discover Identity retry loop on Control Message reject")
Signed-off-by: Xu Yang <xu.yang_2@nxp.com>
---
drivers/usb/typec/tcpm/tcpm.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/usb/typec/tcpm/tcpm.c b/drivers/usb/typec/tcpm/tcpm.c
index 2d6b14aa2085..cc9ea5d5e78a 100644
--- a/drivers/usb/typec/tcpm/tcpm.c
+++ b/drivers/usb/typec/tcpm/tcpm.c
@@ -3891,6 +3891,9 @@ static void tcpm_pd_ctrl_request(struct tcpm_port *port,
case SRC_READY:
case SNK_READY:
if (port->vdm_state > VDM_STATE_READY) {
+ if (port->ams == DISCOVER_IDENTITY)
+ tcpm_update_vdm_discovery_state(port,
+ VDM_DISCOVERY_COMPLETE);
port->vdm_state = VDM_STATE_DONE;
if (tcpm_vdm_ams(port))
tcpm_ams_finish(port);
--
2.34.1
On Mon, Sep 14, 2026 at 05:38:28PM +0800, Xu Yang wrote:
> From: Xu Yang <xu.yang_2@nxp.com>
>
> Commit fa4d0a7cd036 ("usb: typec: tcpm: fix Discover Identity retry loop
> on Control Message reject") advances vdm_discovery_state to
> VDM_DISCOVERY_COMPLETE when a CMDT_RSP_NAK is received, but it does not
> advance the state when a non-modal port returns PD_CTRL_NOT_SUPP.
>
> This can re-arm another vdm_discovery_work if the port is going to
> respond with PD_CTRL_NOT_SUPP to the partner.
>
> tcpm_pd_handle_msg(port, PD_MSG_CTRL_NOT_SUPP, NONE_AMS);
> tcpm_queue_message();
> mod_tcpm_delayed_work();
> tcpm_state_machine_work()
> tcpm_send_queued_message() -> send PD_MSG_CTRL_NOT_SUPP
> run_state_machine()
> case SRC_READY:
> case SNK_READY:
> mod_vdm_discovery_delayed_work(port, 0);
>
> If both the port and its partner do not support modal operation, they
> will send DISCOVER_IDENTITY to each other infinitely.
>
> Advance vdm_discovery_state to VDM_DISCOVERY_COMPLETE in this path as
> well, mirroring the existing handling for the SVDM CMDT_RSP_NAK case.
>
> Fixes: fa4d0a7cd036 ("usb: typec: tcpm: fix Discover Identity retry loop on Control Message reject")
I can't find that commit from anywhere?
Thanks,
> Signed-off-by: Xu Yang <xu.yang_2@nxp.com>
> ---
> drivers/usb/typec/tcpm/tcpm.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/drivers/usb/typec/tcpm/tcpm.c b/drivers/usb/typec/tcpm/tcpm.c
> index 2d6b14aa2085..cc9ea5d5e78a 100644
> --- a/drivers/usb/typec/tcpm/tcpm.c
> +++ b/drivers/usb/typec/tcpm/tcpm.c
> @@ -3891,6 +3891,9 @@ static void tcpm_pd_ctrl_request(struct tcpm_port *port,
> case SRC_READY:
> case SNK_READY:
> if (port->vdm_state > VDM_STATE_READY) {
> + if (port->ams == DISCOVER_IDENTITY)
> + tcpm_update_vdm_discovery_state(port,
> + VDM_DISCOVERY_COMPLETE);
> port->vdm_state = VDM_STATE_DONE;
> if (tcpm_vdm_ams(port))
> tcpm_ams_finish(port);
> --
> 2.34.1
--
heikki
Hi Heikki,
On Fri, Sep 18, 2026 at 09:53:17AM +0200, Heikki Krogerus wrote:
> On Mon, Sep 14, 2026 at 05:38:28PM +0800, Xu Yang wrote:
> > From: Xu Yang <xu.yang_2@nxp.com>
> >
> > Commit fa4d0a7cd036 ("usb: typec: tcpm: fix Discover Identity retry loop
> > on Control Message reject") advances vdm_discovery_state to
> > VDM_DISCOVERY_COMPLETE when a CMDT_RSP_NAK is received, but it does not
> > advance the state when a non-modal port returns PD_CTRL_NOT_SUPP.
> >
> > This can re-arm another vdm_discovery_work if the port is going to
> > respond with PD_CTRL_NOT_SUPP to the partner.
> >
> > tcpm_pd_handle_msg(port, PD_MSG_CTRL_NOT_SUPP, NONE_AMS);
> > tcpm_queue_message();
> > mod_tcpm_delayed_work();
> > tcpm_state_machine_work()
> > tcpm_send_queued_message() -> send PD_MSG_CTRL_NOT_SUPP
> > run_state_machine()
> > case SRC_READY:
> > case SNK_READY:
> > mod_vdm_discovery_delayed_work(port, 0);
> >
> > If both the port and its partner do not support modal operation, they
> > will send DISCOVER_IDENTITY to each other infinitely.
> >
> > Advance vdm_discovery_state to VDM_DISCOVERY_COMPLETE in this path as
> > well, mirroring the existing handling for the SVDM CMDT_RSP_NAK case.
> >
> > Fixes: fa4d0a7cd036 ("usb: typec: tcpm: fix Discover Identity retry loop on Control Message reject")
>
> I can't find that commit from anywhere?
It's a mistake, and I sent v2 to correct it:
https://lore.kernel.org/linux-usb/20260916043957.3362995-1-xu.yang_2@oss.nxp.com/
Thanks,
Xu Yang
© 2016 - 2026 Red Hat, Inc.