RO and RW of EC may have different EC protocol version. If EC transitions
between RO and RW, but AP does not reboot (this is true for fingerprint
microcontroller / cros_fp, but not true for main ec / cros_ec), the AP
still uses the protocol version queried before transition, which can
cause problems. In the case of fingerprint microcontroller, this causes
AP to send the wrong version of EC_CMD_GET_NEXT_EVENT to RO in the
interrupt handler, which in turn prevents RO to clear the interrupt
line to AP, in an infinite loop.
Once an EC_HOST_EVENT_INTERFACE_READY is received, we know that there
might have been a transition between RO and RW, so re-query the protocol.
Signed-off-by: Yicheng Li <[email protected]>
Change-Id: Ib58032ff4a8e113bdbd07212e8aff42807afff38
Series-to: LKML <[email protected]>
Series-cc: Benson Leung <[email protected]>, Enric Balletbo i Serra <[email protected]>, Gwendal Grignou <[email protected]>
---
drivers/platform/chrome/cros_ec.c | 24 +++++++++++++++++++++
include/linux/platform_data/cros_ec_proto.h | 1 +
2 files changed, 25 insertions(+)
diff --git a/drivers/platform/chrome/cros_ec.c b/drivers/platform/chrome/cros_ec.c
index 9b2d07422e17..0c910846d99d 100644
--- a/drivers/platform/chrome/cros_ec.c
+++ b/drivers/platform/chrome/cros_ec.c
@@ -104,6 +104,23 @@ static int cros_ec_sleep_event(struct cros_ec_device *ec_dev, u8 sleep_event)
return ret;
}
+static int cros_ec_ready_event(struct notifier_block *nb,
+ unsigned long queued_during_suspend, void *_notify)
+{
+ struct cros_ec_device *ec_dev = container_of(nb, struct cros_ec_device,
+ notifier_ready);
+ u32 host_event = cros_ec_get_host_event(ec_dev);
+
+ if (host_event & EC_HOST_EVENT_MASK(EC_HOST_EVENT_INTERFACE_READY)) {
+ mutex_lock(&ec_dev->lock);
+ cros_ec_query_all(ec_dev);
+ mutex_unlock(&ec_dev->lock);
+ return NOTIFY_OK;
+ } else {
+ return NOTIFY_DONE;
+ }
+}
+
/**
* cros_ec_register() - Register a new ChromeOS EC, using the provided info.
* @ec_dev: Device to register.
@@ -201,6 +218,13 @@ int cros_ec_register(struct cros_ec_device *ec_dev)
dev_dbg(ec_dev->dev, "Error %d clearing sleep event to ec",
err);
+ /* Register the notifier for EC_HOST_EVENT_INTERFACE_READY event. */
+ ec_dev->notifier_ready.notifier_call = cros_ec_ready_event;
+ err = blocking_notifier_chain_register(&ec_dev->event_notifier,
+ &ec_dev->notifier_ready);
+ if (err < 0)
+ dev_warn(ec_dev->dev, "Failed to register notifier\n");
+
dev_info(dev, "Chrome EC device registered\n");
return 0;
diff --git a/include/linux/platform_data/cros_ec_proto.h b/include/linux/platform_data/cros_ec_proto.h
index 0d4e4aaed37a..9840408c0b01 100644
--- a/include/linux/platform_data/cros_ec_proto.h
+++ b/include/linux/platform_data/cros_ec_proto.h
@@ -161,6 +161,7 @@ struct cros_ec_device {
int event_size;
u32 host_event_wake_mask;
u32 last_resume_result;
+ struct notifier_block notifier_ready;
/* The platform devices used by the mfd driver */
struct platform_device *ec;
--
2.21.0
Hi Yicheng,
Many thanks for sending the patch upstream.
On 18/11/19 21:00, Yicheng Li wrote:
> RO and RW of EC may have different EC protocol version. If EC transitions
> between RO and RW, but AP does not reboot (this is true for fingerprint
> microcontroller / cros_fp, but not true for main ec / cros_ec), the AP
> still uses the protocol version queried before transition, which can
> cause problems. In the case of fingerprint microcontroller, this causes
> AP to send the wrong version of EC_CMD_GET_NEXT_EVENT to RO in the
> interrupt handler, which in turn prevents RO to clear the interrupt
> line to AP, in an infinite loop.
>
cc'ing: Fabien.
I am wondering if we should be able to reproduce this from userspace, i.e by
setting firmware to RO and issuing a EC_CMD_REBOOT.
What's the actual issue, if the firmware hangs switches to RO and reboots? Is
this happening in products or is something that you detected while playing with
the firmware?
> Once an EC_HOST_EVENT_INTERFACE_READY is received, we know that there
> might have been a transition between RO and RW, so re-query the protocol.
>
> Signed-off-by: Yicheng Li <[email protected]>
>
> Change-Id: Ib58032ff4a8e113bdbd07212e8aff42807afff38
> Series-to: LKML <[email protected]>
> Series-cc: Benson Leung <[email protected]>, Enric Balletbo i Serra <[email protected]>, Gwendal Grignou <[email protected]>
You should get rid of the three above lines, did you use patman correctly to
send this patch?
> ---
> drivers/platform/chrome/cros_ec.c | 24 +++++++++++++++++++++
> include/linux/platform_data/cros_ec_proto.h | 1 +
> 2 files changed, 25 insertions(+)
>
> diff --git a/drivers/platform/chrome/cros_ec.c b/drivers/platform/chrome/cros_ec.c
> index 9b2d07422e17..0c910846d99d 100644
> --- a/drivers/platform/chrome/cros_ec.c
> +++ b/drivers/platform/chrome/cros_ec.c
> @@ -104,6 +104,23 @@ static int cros_ec_sleep_event(struct cros_ec_device *ec_dev, u8 sleep_event)
> return ret;
> }
>
> +static int cros_ec_ready_event(struct notifier_block *nb,
> + unsigned long queued_during_suspend, void *_notify)
> +{
> + struct cros_ec_device *ec_dev = container_of(nb, struct cros_ec_device,
> + notifier_ready);
> + u32 host_event = cros_ec_get_host_event(ec_dev);
> +
> + if (host_event & EC_HOST_EVENT_MASK(EC_HOST_EVENT_INTERFACE_READY)) {
> + mutex_lock(&ec_dev->lock);
> + cros_ec_query_all(ec_dev);
> + mutex_unlock(&ec_dev->lock);
> + return NOTIFY_OK;
> + } else {
> + return NOTIFY_DONE;
> + }
no else, just
return NOTIFY_DONE;
> +}
> +
> /**
> * cros_ec_register() - Register a new ChromeOS EC, using the provided info.
> * @ec_dev: Device to register.
> @@ -201,6 +218,13 @@ int cros_ec_register(struct cros_ec_device *ec_dev)
> dev_dbg(ec_dev->dev, "Error %d clearing sleep event to ec",
> err);
>
> + /* Register the notifier for EC_HOST_EVENT_INTERFACE_READY event. */
> + ec_dev->notifier_ready.notifier_call = cros_ec_ready_event;
> + err = blocking_notifier_chain_register(&ec_dev->event_notifier,
> + &ec_dev->notifier_ready);
> + if (err < 0)
> + dev_warn(ec_dev->dev, "Failed to register notifier\n");
> +
Is that only a warning and should be ignored if it fails?
BTW, blocking_notfier_chain_register _always_ returns 0. So I'd say that better
do in case at some point starts to returning something different.
if (err)
return err;
> dev_info(dev, "Chrome EC device registered\n");
>
> return 0;
> diff --git a/include/linux/platform_data/cros_ec_proto.h b/include/linux/platform_data/cros_ec_proto.h
> index 0d4e4aaed37a..9840408c0b01 100644
> --- a/include/linux/platform_data/cros_ec_proto.h
> +++ b/include/linux/platform_data/cros_ec_proto.h
> @@ -161,6 +161,7 @@ struct cros_ec_device {
> int event_size;
> u32 host_event_wake_mask;
> u32 last_resume_result;
> + struct notifier_block notifier_ready;
You should add documentations for this new entry.
>
> /* The platform devices used by the mfd driver */
> struct platform_device *ec;
>
Thanks,
Enric
Hi,
I tried to test that patch but I cannot switch to RO firmware.
I do the following steps:
root@debian:/sys/class/chromeos/cros_fp# cat version
RO version: nocturne_fp_v2.2.64-58cf5974e
RW version: nocturne_fp_v2.2.110-b936c0a3c
Firmware copy: RW
Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02 14:16:46
@swarm-cros-461
Chip vendor: stm
Chip name: stm32h7x3
Chip revision:
Board version: EC error 1
root@debian:/sys/class/chromeos/cros_fp# echo ro > reboot
root@debian:/sys/class/chromeos/cros_fp# cat version
RO version: nocturne_fp_v2.2.64-58cf5974e
RW version: nocturne_fp_v2.2.110-b936c0a3c
Firmware copy: RW
Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02 14:16:46
@swarm-cros-461
Chip vendor: stm
Chip name: stm32h7x3
Chip revision:
Board version: EC error 1
root@debian:/sys/class/chromeos/cros_fp#
We see here that cros_fp is still RW.
I also tried with:
debian@debian:/src/ec$ sudo build/bds/util/ectool --name=cros_fp
reboot_ec RO
debian@debian:/src/ec$ sudo build/bds/util/ectool --name=cros_fp
version
RO version: nocturne_fp_v2.2.64-58cf5974e
RW version: nocturne_fp_v2.2.110-b936c0a3c
Firmware copy: RW
Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02 14:16:46
@swarm-cros-461
Tool version: v2.0.3074-a5052d4e7 2020-01-16 10:23:05 debian@debian
debian@debian:/src/ec$
with the same result.
Can you decribe us steps you follow to test that patch?
Thanks
Fabien
Le lundi 18 novembre 2019 à 12:00 -0800, Yicheng Li a écrit :
> RO and RW of EC may have different EC protocol version. If EC
> transitions
> between RO and RW, but AP does not reboot (this is true for
> fingerprint
> microcontroller / cros_fp, but not true for main ec / cros_ec), the
> AP
> still uses the protocol version queried before transition, which can
> cause problems. In the case of fingerprint microcontroller, this
> causes
> AP to send the wrong version of EC_CMD_GET_NEXT_EVENT to RO in the
> interrupt handler, which in turn prevents RO to clear the interrupt
> line to AP, in an infinite loop.
>
> Once an EC_HOST_EVENT_INTERFACE_READY is received, we know that there
> might have been a transition between RO and RW, so re-query the
> protocol.
>
> Signed-off-by: Yicheng Li <[email protected]>
>
> Change-Id: Ib58032ff4a8e113bdbd07212e8aff42807afff38
> Series-to: LKML <[email protected]>
> Series-cc: Benson Leung <[email protected]>, Enric Balletbo i Serra
> <[email protected]>, Gwendal Grignou <[email protected]
> >
> ---
> drivers/platform/chrome/cros_ec.c | 24
> +++++++++++++++++++++
> include/linux/platform_data/cros_ec_proto.h | 1 +
> 2 files changed, 25 insertions(+)
>
> diff --git a/drivers/platform/chrome/cros_ec.c
> b/drivers/platform/chrome/cros_ec.c
> index 9b2d07422e17..0c910846d99d 100644
> --- a/drivers/platform/chrome/cros_ec.c
> +++ b/drivers/platform/chrome/cros_ec.c
> @@ -104,6 +104,23 @@ static int cros_ec_sleep_event(struct
> cros_ec_device *ec_dev, u8 sleep_event)
> return ret;
> }
>
> +static int cros_ec_ready_event(struct notifier_block *nb,
> + unsigned long queued_during_suspend, void *_notify)
> +{
> + struct cros_ec_device *ec_dev = container_of(nb, struct
> cros_ec_device,
> + notifier_ready);
> + u32 host_event = cros_ec_get_host_event(ec_dev);
> +
> + if (host_event &
> EC_HOST_EVENT_MASK(EC_HOST_EVENT_INTERFACE_READY)) {
> + mutex_lock(&ec_dev->lock);
> + cros_ec_query_all(ec_dev);
> + mutex_unlock(&ec_dev->lock);
> + return NOTIFY_OK;
> + } else {
> + return NOTIFY_DONE;
> + }
> +}
> +
> /**
> * cros_ec_register() - Register a new ChromeOS EC, using the
> provided info.
> * @ec_dev: Device to register.
> @@ -201,6 +218,13 @@ int cros_ec_register(struct cros_ec_device
> *ec_dev)
> dev_dbg(ec_dev->dev, "Error %d clearing sleep event to
> ec",
> err);
>
> + /* Register the notifier for EC_HOST_EVENT_INTERFACE_READY
> event. */
> + ec_dev->notifier_ready.notifier_call = cros_ec_ready_event;
> + err = blocking_notifier_chain_register(&ec_dev->event_notifier,
> + &ec_dev-
> >notifier_ready);
> + if (err < 0)
> + dev_warn(ec_dev->dev, "Failed to register notifier\n");
> +
> dev_info(dev, "Chrome EC device registered\n");
>
> return 0;
> diff --git a/include/linux/platform_data/cros_ec_proto.h
> b/include/linux/platform_data/cros_ec_proto.h
> index 0d4e4aaed37a..9840408c0b01 100644
> --- a/include/linux/platform_data/cros_ec_proto.h
> +++ b/include/linux/platform_data/cros_ec_proto.h
> @@ -161,6 +161,7 @@ struct cros_ec_device {
> int event_size;
> u32 host_event_wake_mask;
> u32 last_resume_result;
> + struct notifier_block notifier_ready;
>
> /* The platform devices used by the mfd driver */
> struct platform_device *ec;
Hi Yicheng,
Thanks for your quick answer.
I updated the firmware after my first email on friday.
Now the version is correct:
# ectool version --name=cros_fp
RO version: nocturne_fp_v2.2.64-58cf5974e
RW version: nocturne_fp_v2.2.191-1d529566e
Firmware copy: RW
Build info: nocturne_fp_v2.2.191-1d529566e 2019-11-01 20:30:58
@chromeos-ci-legacy-us-central1-b-x32-72-dltd
Tool version: v2.0.3031-9e0f24fad 2020-01-13 06:11:02 @chromeos-ci-
legacy-us-east1-d-x32-89-i09h
The result does not change. This is the console output after I sent the
reboot command.
$ ectool name=cros_fp output
...
[250664.176204 HC 0xa9]
[250664.176287 HC 0xa9 err 1]
[250664.177413 HC 0x67]
[250664.177489 HC 0x67 err 9]
[250664.178434 HC 0x02]
[250664.180308 HC Suppressed: 0x97=127 0x98=138 0x115=0]
[250664.477987 HC 0x01]
[250892.898421 HC 0x02]
[250892.900447 HC 0x07]
[250892.901286 HC 0x0b]
[250902.139053 HC 0x02]
[250902.141059 HC 0x07]
[250902.141847 HC 0x0b]
[250902.142988 HC 0xd2]
[250902.143072 Executing host reboot command 1]
[250902.143176 HC 0xd2 err 4]
[250908.838254 HC 0x02]
[250908.840474 HC 0x07]
[250908.841391 HC 0x0b]
it seems we have an error when executing the 0zd2 command.
Any idea what is error 4?
Thanks
Fabien
Le jeudi 16 janvier 2020 à 05:04 -0800, Yicheng Li a écrit :
> Hi Fabien,
>
> First, you need the new firmware which has RW 2.2.191 instead of
> 2.2.110. It's available in release M80. The old firmware you were
> using does not have the RO/RW protocol incompatibility problem
> because RW 2.2.110 still only support versions 0 and 1 of
> EC_CMD_GET_NEXT_EVENT, same as RO 2.2.64. RW 2.2.191 has version 2 of
> EC_CMD_GET_NEXT_ EVENT, hence the issue described in the commit
> message.
>
> With RW 2.2.191, if the kernal does not have this patch, rebooting
> cros_fp will leave you in an infinite loop in RO, as described in the
> commit message. This patch solves the problem and you should be in RW
> after the reboot. You can see more details on the commands sent to
> cros_fp during RO if you have cros_fp.log
>
> Let me know if you have other questions. Thanks!
>
> Best,
> Yicheng
>
>
> On Thu, Jan 16, 2020, 4:27 AM Fabien Lahoudere <
> [email protected]> wrote:
> > Hi,
> >
> > I tried to test that patch but I cannot switch to RO firmware.
> > I do the following steps:
> >
> > root@debian:/sys/class/chromeos/cros_fp# cat version
> > RO version: nocturne_fp_v2.2.64-58cf5974e
> > RW version: nocturne_fp_v2.2.110-b936c0a3c
> > Firmware copy: RW
> > Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02 14:16:46
> > @swarm-cros-461
> > Chip vendor: stm
> > Chip name: stm32h7x3
> > Chip revision:
> > Board version: EC error 1
> > root@debian:/sys/class/chromeos/cros_fp# echo ro > reboot
> > root@debian:/sys/class/chromeos/cros_fp# cat version
> > RO version: nocturne_fp_v2.2.64-58cf5974e
> > RW version: nocturne_fp_v2.2.110-b936c0a3c
> > Firmware copy: RW
> > Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02 14:16:46
> > @swarm-cros-461
> > Chip vendor: stm
> > Chip name: stm32h7x3
> > Chip revision:
> > Board version: EC error 1
> > root@debian:/sys/class/chromeos/cros_fp#
> >
> > We see here that cros_fp is still RW.
> >
> > I also tried with:
> >
> > debian@debian:/src/ec$ sudo build/bds/util/ectool --name=cros_fp
> > reboot_ec RO
> > debian@debian:/src/ec$ sudo build/bds/util/ectool --name=cros_fp
> > version
> > RO version: nocturne_fp_v2.2.64-58cf5974e
> > RW version: nocturne_fp_v2.2.110-b936c0a3c
> > Firmware copy: RW
> > Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02 14:16:46
> > @swarm-cros-461
> > Tool version: v2.0.3074-a5052d4e7 2020-01-16 10:23:05
> > debian@debian
> > debian@debian:/src/ec$
> >
> > with the same result.
> >
> > Can you decribe us steps you follow to test that patch?
> >
> > Thanks
> >
> > Fabien
> >
> > Le lundi 18 novembre 2019 à 12:00 -0800, Yicheng Li a écrit :
> > > RO and RW of EC may have different EC protocol version. If EC
> > > transitions
> > > between RO and RW, but AP does not reboot (this is true for
> > > fingerprint
> > > microcontroller / cros_fp, but not true for main ec / cros_ec),
> > the
> > > AP
> > > still uses the protocol version queried before transition, which
> > can
> > > cause problems. In the case of fingerprint microcontroller, this
> > > causes
> > > AP to send the wrong version of EC_CMD_GET_NEXT_EVENT to RO in
> > the
> > > interrupt handler, which in turn prevents RO to clear the
> > interrupt
> > > line to AP, in an infinite loop.
> > >
> > > Once an EC_HOST_EVENT_INTERFACE_READY is received, we know that
> > there
> > > might have been a transition between RO and RW, so re-query the
> > > protocol.
> > >
> > > Signed-off-by: Yicheng Li <[email protected]>
> > >
> > > Change-Id: Ib58032ff4a8e113bdbd07212e8aff42807afff38
> > > Series-to: LKML <[email protected]>
> > > Series-cc: Benson Leung <[email protected]>, Enric Balletbo i
> > Serra
> > > <[email protected]>, Gwendal Grignou <
> > [email protected]
> > > >
> > > ---
> > > drivers/platform/chrome/cros_ec.c | 24
> > > +++++++++++++++++++++
> > > include/linux/platform_data/cros_ec_proto.h | 1 +
> > > 2 files changed, 25 insertions(+)
> > >
> > > diff --git a/drivers/platform/chrome/cros_ec.c
> > > b/drivers/platform/chrome/cros_ec.c
> > > index 9b2d07422e17..0c910846d99d 100644
> > > --- a/drivers/platform/chrome/cros_ec.c
> > > +++ b/drivers/platform/chrome/cros_ec.c
> > > @@ -104,6 +104,23 @@ static int cros_ec_sleep_event(struct
> > > cros_ec_device *ec_dev, u8 sleep_event)
> > > return ret;
> > > }
> > >
> > > +static int cros_ec_ready_event(struct notifier_block *nb,
> > > + unsigned long queued_during_suspend, void *_notify)
> > > +{
> > > + struct cros_ec_device *ec_dev = container_of(nb, struct
> > > cros_ec_device,
> > > +
> > notifier_ready);
> > > + u32 host_event = cros_ec_get_host_event(ec_dev);
> > > +
> > > + if (host_event &
> > > EC_HOST_EVENT_MASK(EC_HOST_EVENT_INTERFACE_READY)) {
> > > + mutex_lock(&ec_dev->lock);
> > > + cros_ec_query_all(ec_dev);
> > > + mutex_unlock(&ec_dev->lock);
> > > + return NOTIFY_OK;
> > > + } else {
> > > + return NOTIFY_DONE;
> > > + }
> > > +}
> > > +
> > > /**
> > > * cros_ec_register() - Register a new ChromeOS EC, using the
> > > provided info.
> > > * @ec_dev: Device to register.
> > > @@ -201,6 +218,13 @@ int cros_ec_register(struct cros_ec_device
> > > *ec_dev)
> > > dev_dbg(ec_dev->dev, "Error %d clearing sleep event
> > to
> > > ec",
> > > err);
> > >
> > > + /* Register the notifier for EC_HOST_EVENT_INTERFACE_READY
> > > event. */
> > > + ec_dev->notifier_ready.notifier_call = cros_ec_ready_event;
> > > + err = blocking_notifier_chain_register(&ec_dev-
> > >event_notifier,
> > > + &ec_dev-
> > > >notifier_ready);
> > > + if (err < 0)
> > > + dev_warn(ec_dev->dev, "Failed to register
> > notifier\n");
> > > +
> > > dev_info(dev, "Chrome EC device registered\n");
> > >
> > > return 0;
> > > diff --git a/include/linux/platform_data/cros_ec_proto.h
> > > b/include/linux/platform_data/cros_ec_proto.h
> > > index 0d4e4aaed37a..9840408c0b01 100644
> > > --- a/include/linux/platform_data/cros_ec_proto.h
> > > +++ b/include/linux/platform_data/cros_ec_proto.h
> > > @@ -161,6 +161,7 @@ struct cros_ec_device {
> > > int event_size;
> > > u32 host_event_wake_mask;
> > > u32 last_resume_result;
> > > + struct notifier_block notifier_ready;
> > >
> > > /* The platform devices used by the mfd driver */
> > > struct platform_device *ec;
> >
Hi,
With the patch applied,
root@debian:/src/ec# ./build/bds/util/ectool --name=cros_fp
reboot_ec
ioctl -1, errno 110 (Connection timed out), EC result 255
(<unknown>)
root@debian:/src/ec# ./build/bds/util/ectool --name=cros_fp version
RO version: nocturne_fp_v2.2.64-58cf5974e
RW version: nocturne_fp_v2.2.191-1d529566e
Firmware copy: RW
Build info: nocturne_fp_v2.2.191-1d529566e 2019-11-01 20:30:58
@chromeos-ci-legacy-us-central1-b-x32-72-dltd
Tool version: v2.0.3074-a5052d4e7 2020-01-16 10:23:05 debian@debian
in that case we see that the firmware copy is RW after the reboot.
Without the patch :
root@debian:/src/ec# ./build/bds/util/ectool --name=cros_fp
reboot_ec
ioctl -1, errno 110 (Connection timed out), EC result 255
(<unknown>)
root@debian:/src/ec# ./build/bds/util/ectool --name=cros_fp version
RO version: nocturne_fp_v2.2.64-58cf5974e
RW version: nocturne_fp_v2.2.191-1d529566e
Firmware copy: RO
Build info: nocturne_fp_v2.2.64-58cf5974e 2018-10-01 19:11:56
@cros-beefy532-c2
Tool version: v2.0.3074-a5052d4e7 2020-01-16 10:23:05 debian@debian
root@debian:/src/ec#
The firmware copy stays RO
For me the patch works fine
Tested-by: Fabien Lahoudere <[email protected]>
Thanks
Le mardi 21 janvier 2020 à 10:13 -0800, Yicheng Li a écrit :
> Hi Fabien,
>
> error 4 is EC_RES_ACCESS_DENIED. I think it's because you are trying
> to directly reboot to RO, which issues the command EC_CMD_REBOOT_EC
> (0xD2)
> https://chromium.git.corp.google.com/chromiumos/platform/ec/+/15edeaabe90c554d83faa4ff0c9143c66248e0e4/common/system.c#1495
> and is ultimately disallowed since system_is_locked() is true:
> https://chromium.git.corp.google.com/chromiumos/platform/ec/+/15edeaabe90c554d83faa4ff0c9143c66248e0e4/common/system.c#551
>
> The way I test it is to run "ectool --name=cros_fp reboot_ec" without
> the "RO" argument. This issues EC_CMD_REBOOT
> https://chromium.git.corp.google.com/chromiumos/platform/ec/+/15edeaabe90c554d83faa4ff0c9143c66248e0e4/util/ectool.c#1055
> which is 0xD1. (Sorry I didn't notice the RO argument in the first
> pass...)
>
> With "ectool --name=cros_fp reboot_ec", we are telling cros_fp to do
> an immediate reboot (to RW), however because of the issue outlined in
> the commit description, it would hang in RO without this patch.
>
> Thanks,
> Yicheng
>
>
>
> On Mon, Jan 20, 2020 at 3:16 AM Fabien Lahoudere <
> [email protected]> wrote:
> > Hi Yicheng,
> >
> > Thanks for your quick answer.
> >
> > I updated the firmware after my first email on friday.
> > Now the version is correct:
> >
> > # ectool version --name=cros_fp
> > RO version: nocturne_fp_v2.2.64-58cf5974e
> > RW version: nocturne_fp_v2.2.191-1d529566e
> > Firmware copy: RW
> > Build info: nocturne_fp_v2.2.191-1d529566e 2019-11-01
> > 20:30:58
> > @chromeos-ci-legacy-us-central1-b-x32-72-dltd
> > Tool version: v2.0.3031-9e0f24fad 2020-01-13 06:11:02
> > @chromeos-ci-
> > legacy-us-east1-d-x32-89-i09h
> >
> > The result does not change. This is the console output after I sent
> > the
> > reboot command.
> >
> > $ ectool name=cros_fp output
> > ...
> > [250664.176204 HC 0xa9]
> > [250664.176287 HC 0xa9 err 1]
> > [250664.177413 HC 0x67]
> > [250664.177489 HC 0x67 err 9]
> > [250664.178434 HC 0x02]
> > [250664.180308 HC Suppressed: 0x97=127 0x98=138 0x115=0]
> > [250664.477987 HC 0x01]
> > [250892.898421 HC 0x02]
> > [250892.900447 HC 0x07]
> > [250892.901286 HC 0x0b]
> > [250902.139053 HC 0x02]
> > [250902.141059 HC 0x07]
> > [250902.141847 HC 0x0b]
> > [250902.142988 HC 0xd2]
> > [250902.143072 Executing host reboot command 1]
> > [250902.143176 HC 0xd2 err 4]
> > [250908.838254 HC 0x02]
> > [250908.840474 HC 0x07]
> > [250908.841391 HC 0x0b]
> >
> > it seems we have an error when executing the 0zd2 command.
> >
> > Any idea what is error 4?
> >
> > Thanks
> >
> > Fabien
> >
> > Le jeudi 16 janvier 2020 à 05:04 -0800, Yicheng Li a écrit :
> > > Hi Fabien,
> > >
> > > First, you need the new firmware which has RW 2.2.191 instead of
> > > 2.2.110. It's available in release M80. The old firmware you were
> > > using does not have the RO/RW protocol incompatibility problem
> > > because RW 2.2.110 still only support versions 0 and 1 of
> > > EC_CMD_GET_NEXT_EVENT, same as RO 2.2.64. RW 2.2.191 has version
> > 2 of
> > > EC_CMD_GET_NEXT_ EVENT, hence the issue described in the commit
> > > message.
> > >
> > > With RW 2.2.191, if the kernal does not have this patch,
> > rebooting
> > > cros_fp will leave you in an infinite loop in RO, as described in
> > the
> > > commit message. This patch solves the problem and you should be
> > in RW
> > > after the reboot. You can see more details on the commands sent
> > to
> > > cros_fp during RO if you have cros_fp.log
> > >
> > > Let me know if you have other questions. Thanks!
> > >
> > > Best,
> > > Yicheng
> > >
> > >
> > > On Thu, Jan 16, 2020, 4:27 AM Fabien Lahoudere <
> > > [email protected]> wrote:
> > > > Hi,
> > > >
> > > > I tried to test that patch but I cannot switch to RO firmware.
> > > > I do the following steps:
> > > >
> > > > root@debian:/sys/class/chromeos/cros_fp# cat version
> > > > RO version: nocturne_fp_v2.2.64-58cf5974e
> > > > RW version: nocturne_fp_v2.2.110-b936c0a3c
> > > > Firmware copy: RW
> > > > Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02
> > 14:16:46
> > > > @swarm-cros-461
> > > > Chip vendor: stm
> > > > Chip name: stm32h7x3
> > > > Chip revision:
> > > > Board version: EC error 1
> > > > root@debian:/sys/class/chromeos/cros_fp# echo ro > reboot
> > > > root@debian:/sys/class/chromeos/cros_fp# cat version
> > > > RO version: nocturne_fp_v2.2.64-58cf5974e
> > > > RW version: nocturne_fp_v2.2.110-b936c0a3c
> > > > Firmware copy: RW
> > > > Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02
> > 14:16:46
> > > > @swarm-cros-461
> > > > Chip vendor: stm
> > > > Chip name: stm32h7x3
> > > > Chip revision:
> > > > Board version: EC error 1
> > > > root@debian:/sys/class/chromeos/cros_fp#
> > > >
> > > > We see here that cros_fp is still RW.
> > > >
> > > > I also tried with:
> > > >
> > > > debian@debian:/src/ec$ sudo build/bds/util/ectool --
> > name=cros_fp
> > > > reboot_ec RO
> > > > debian@debian:/src/ec$ sudo build/bds/util/ectool --
> > name=cros_fp
> > > > version
> > > > RO version: nocturne_fp_v2.2.64-58cf5974e
> > > > RW version: nocturne_fp_v2.2.110-b936c0a3c
> > > > Firmware copy: RW
> > > > Build info: nocturne_fp_v2.2.110-b936c0a3c 2018-11-02
> > 14:16:46
> > > > @swarm-cros-461
> > > > Tool version: v2.0.3074-a5052d4e7 2020-01-16 10:23:05
> > > > debian@debian
> > > > debian@debian:/src/ec$
> > > >
> > > > with the same result.
> > > >
> > > > Can you decribe us steps you follow to test that patch?
> > > >
> > > > Thanks
> > > >
> > > > Fabien
> > > >
> > > > Le lundi 18 novembre 2019 à 12:00 -0800, Yicheng Li a écrit :
> > > > > RO and RW of EC may have different EC protocol version. If EC
> > > > > transitions
> > > > > between RO and RW, but AP does not reboot (this is true for
> > > > > fingerprint
> > > > > microcontroller / cros_fp, but not true for main ec /
> > cros_ec),
> > > > the
> > > > > AP
> > > > > still uses the protocol version queried before transition,
> > which
> > > > can
> > > > > cause problems. In the case of fingerprint microcontroller,
> > this
> > > > > causes
> > > > > AP to send the wrong version of EC_CMD_GET_NEXT_EVENT to RO
> > in
> > > > the
> > > > > interrupt handler, which in turn prevents RO to clear the
> > > > interrupt
> > > > > line to AP, in an infinite loop.
> > > > >
> > > > > Once an EC_HOST_EVENT_INTERFACE_READY is received, we know
> > that
> > > > there
> > > > > might have been a transition between RO and RW, so re-query
> > the
> > > > > protocol.
> > > > >
> > > > > Signed-off-by: Yicheng Li <[email protected]>
> > > > >
> > > > > Change-Id: Ib58032ff4a8e113bdbd07212e8aff42807afff38
> > > > > Series-to: LKML <[email protected]>
> > > > > Series-cc: Benson Leung <[email protected]>, Enric Balletbo
> > i
> > > > Serra
> > > > > <[email protected]>, Gwendal Grignou <
> > > > [email protected]
> > > > > >
> > > > > ---
> > > > > drivers/platform/chrome/cros_ec.c | 24
> > > > > +++++++++++++++++++++
> > > > > include/linux/platform_data/cros_ec_proto.h | 1 +
> > > > > 2 files changed, 25 insertions(+)
> > > > >
> > > > > diff --git a/drivers/platform/chrome/cros_ec.c
> > > > > b/drivers/platform/chrome/cros_ec.c
> > > > > index 9b2d07422e17..0c910846d99d 100644
> > > > > --- a/drivers/platform/chrome/cros_ec.c
> > > > > +++ b/drivers/platform/chrome/cros_ec.c
> > > > > @@ -104,6 +104,23 @@ static int cros_ec_sleep_event(struct
> > > > > cros_ec_device *ec_dev, u8 sleep_event)
> > > > > return ret;
> > > > > }
> > > > >
> > > > > +static int cros_ec_ready_event(struct notifier_block *nb,
> > > > > + unsigned long queued_during_suspend, void *_notify)
> > > > > +{
> > > > > + struct cros_ec_device *ec_dev = container_of(nb, struct
> > > > > cros_ec_device,
> > > > > +
> > > > notifier_ready);
> > > > > + u32 host_event = cros_ec_get_host_event(ec_dev);
> > > > > +
> > > > > + if (host_event &
> > > > > EC_HOST_EVENT_MASK(EC_HOST_EVENT_INTERFACE_READY)) {
> > > > > + mutex_lock(&ec_dev->lock);
> > > > > + cros_ec_query_all(ec_dev);
> > > > > + mutex_unlock(&ec_dev->lock);
> > > > > + return NOTIFY_OK;
> > > > > + } else {
> > > > > + return NOTIFY_DONE;
> > > > > + }
> > > > > +}
> > > > > +
> > > > > /**
> > > > > * cros_ec_register() - Register a new ChromeOS EC, using
> > the
> > > > > provided info.
> > > > > * @ec_dev: Device to register.
> > > > > @@ -201,6 +218,13 @@ int cros_ec_register(struct
> > cros_ec_device
> > > > > *ec_dev)
> > > > > dev_dbg(ec_dev->dev, "Error %d clearing sleep
> > event
> > > > to
> > > > > ec",
> > > > > err);
> > > > >
> > > > > + /* Register the notifier for
> > EC_HOST_EVENT_INTERFACE_READY
> > > > > event. */
> > > > > + ec_dev->notifier_ready.notifier_call =
> > cros_ec_ready_event;
> > > > > + err = blocking_notifier_chain_register(&ec_dev-
> > > > >event_notifier,
> > > > > + &ec_dev-
> > > > > >notifier_ready);
> > > > > + if (err < 0)
> > > > > + dev_warn(ec_dev->dev, "Failed to register
> > > > notifier\n");
> > > > > +
> > > > > dev_info(dev, "Chrome EC device registered\n");
> > > > >
> > > > > return 0;
> > > > > diff --git a/include/linux/platform_data/cros_ec_proto.h
> > > > > b/include/linux/platform_data/cros_ec_proto.h
> > > > > index 0d4e4aaed37a..9840408c0b01 100644
> > > > > --- a/include/linux/platform_data/cros_ec_proto.h
> > > > > +++ b/include/linux/platform_data/cros_ec_proto.h
> > > > > @@ -161,6 +161,7 @@ struct cros_ec_device {
> > > > > int event_size;
> > > > > u32 host_event_wake_mask;
> > > > > u32 last_resume_result;
> > > > > + struct notifier_block notifier_ready;
> > > > >
> > > > > /* The platform devices used by the mfd driver */
> > > > > struct platform_device *ec;
> > > >
> >