Skip to contentSkip to conversation

platform/chrome: cros_ec_proto: Lock device when updating MKBP version

PATCH Superseded Patryk Duda 2 participants 7 messages Jul 25, 2024 – Jul 30, 2024

Subsystems (2)

v1 of 3 Superseded Latest is v3 →

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 platform/chrome: cros_ec_proto: Lock device when updating MKBP version

The cros_ec_get_host_command_version_mask() function requires that the caller must have ec_dev->lock mutex before calling it. This requirement was not met and as a result it was possible that two commands were sent to the device at the same time.

The problem was observed while using UART backend which doesn't use any additional locks, unlike SPI backend which locks the controller until response is received.

Fixes f74c7557ed0d ("platform/chrome: cros_ec_proto: Update version on GET_NEXT_EVENT failure")

---
The f74c7557ed0d patch was tested with Fingerprint MCU (FPMCU) that uses SPI to communicate with chipset. At that time, UART FPMCU had the same version of GET_NEXT_EVENT command in RO and RW, so the MKBP version was never updated in this case.

Version 3 of GET_NEXT_EVENT command was recently added to CrosEC. As a result MKBP version is also updated when UART FPMCU is used which exposed this problem.

Best regards,
Patryk

 drivers/platform/chrome/cros_ec_proto.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/platform/chrome/cros_ec_proto.c +2 -0
index f776fd42244f..cf7ef5f9db84 100644
@@ -813,6 +813,7 @@ int cros_ec_get_next_event(struct cros_ec_device *ec_dev,
 	if (ret == -ENOPROTOOPT) {
 		dev_dbg(ec_dev->dev,
 			"GET_NEXT_EVENT returned invalid version error.\n");
+		mutex_lock(&ec_dev->lock);
 		ret = cros_ec_get_host_command_version_mask(ec_dev,
 							EC_CMD_GET_NEXT_EVENT,
 							&ver_mask);
@@ -829,6 +830,7 @@ int cros_ec_get_next_event(struct cros_ec_device *ec_dev,
 		ec_dev->mkbp_event_supported = fls(ver_mask);
 		dev_dbg(ec_dev->dev, "MKBP support version changed to %u\n",
 			ec_dev->mkbp_event_supported - 1);
+		mutex_unlock(&ec_dev->lock);
 
 		/* Try to get next event with new MKBP support version set. */
 		ret = get_next_event(ec_dev);
Signature · 1 line
2.45.2.1089.g2a221341d9-goog
maintainer
#1 lore raw
Quoted 4 lines · Patryk Duda

The cros_ec_get_host_command_version_mask() function requires that the caller must have ec_dev->lock mutex before calling it. This requirement was not met and as a result it was possible that two commands were sent to the device at the same time.

To clarify:
- What would happen if multiple cros_ec_get_host_command_version_mask() calls at the same time?
- What are the callees? I'm trying to understand the source of parallelism.

Also, the patch also needs an unlock at [1].

[1]: https://elixir.bootlin.com/linux/v6.10/source/drivers/platform/chrome/cros_ec_proto.c#L819

Quoted 3 lines

The problem was observed while using UART backend which doesn't use any additional locks, unlike SPI backend which locks the controller until response is received.

Is it a general issue if multiple commands send to EC at a time? If yes, it should serialize that in the UART transportation.

#2 lore raw
Quoted 10 lines · Tzung-Bi Shih
Quoted 4 lines · Patryk Duda

The cros_ec_get_host_command_version_mask() function requires that the caller must have ec_dev->lock mutex before calling it. This requirement was not met and as a result it was possible that two commands were sent to the device at the same time.

To clarify:
- What would happen if multiple cros_ec_get_host_command_version_mask() calls at the same time?

In the best case, MCU will receive both commands glued together and will ignore them.
It will result in a timeout in the kernel. In the worst case, request and/or response buffers will be
corrupted.

Quoted 1 line

- What are the callees? I'm trying to understand the source of parallelism.

This is a race between interrupt handling and ioctl call from userspace

Handling interrupt path
cros_ec_irq_thread()
cros_ec_handle_event()
cros_ec_get_next_event() - Queries host command version without taking ec_dev->lock mutex first cros_ec_get_host_command_version_mask() cros_ec_send_command()
cros_ec_xfer_command()
cros_ec_uart_pkt_xfer()

Command from userspace
cros_ec_chardev_ioctl()
cros_ec_chardev_ioctl_xcmd()
cros_ec_cmd_xfer() - Locks ec_dev->lock mutex before sending command cros_ec_send_command()
cros_ec_xfer_command()
cros_ec_uart_pkt_xfer()

Quoted 4 lines

Also, the patch also needs an unlock at [1].

[1]: https://elixir.bootlin.com/linux/v6.10/source/drivers/platform/chrome/cros_ec_proto.c#L819

Yeah. I'll fix it in v2

Quoted 7 lines
Quoted 4 lines

The problem was observed while using UART backend which doesn't use any additional locks, unlike SPI backend which locks the controller until response is received.

Is it a general issue if multiple commands send to EC at a time? If yes, it should serialize that in the UART transportation.

Host Commands only support one command at a time. It's enforced by 'lock' mutex from cros_ec_device structure. We just need to use it properly.

maintainer
#3 lore raw
Quoted 39 lines · Patryk Duda
Quoted 10 lines · Tzung-Bi Shih
Quoted 4 lines · Patryk Duda

The cros_ec_get_host_command_version_mask() function requires that the caller must have ec_dev->lock mutex before calling it. This requirement was not met and as a result it was possible that two commands were sent to the device at the same time.

To clarify:
- What would happen if multiple cros_ec_get_host_command_version_mask() calls at the same time?

In the best case, MCU will receive both commands glued together and will ignore them.
It will result in a timeout in the kernel. In the worst case, request and/or response buffers will be
corrupted.

Quoted 1 line

- What are the callees? I'm trying to understand the source of parallelism.

This is a race between interrupt handling and ioctl call from userspace

Handling interrupt path
cros_ec_irq_thread()
cros_ec_handle_event()
cros_ec_get_next_event() - Queries host command version without taking ec_dev->lock mutex first cros_ec_get_host_command_version_mask() cros_ec_send_command()
cros_ec_xfer_command()
cros_ec_uart_pkt_xfer()

Command from userspace
cros_ec_chardev_ioctl()
cros_ec_chardev_ioctl_xcmd()
cros_ec_cmd_xfer() - Locks ec_dev->lock mutex before sending command cros_ec_send_command()
cros_ec_xfer_command()
cros_ec_uart_pkt_xfer()

Quoted 1 line
Quoted 5 lines
Quoted 3 lines

Also, the patch also needs an unlock at [1].

[1]: https://elixir.bootlin.com/linux/v6.10/source/drivers/platform/chrome/cros_ec_proto.c#L819

Yeah. I'll fix it in v2

I'm wondering if it's simpler to just lock and unlock around calling cros_ec_get_host_command_version_mask(). What do you think?

Quoted 9 lines
Quoted 6 lines
Quoted 3 lines

The problem was observed while using UART backend which doesn't use any additional locks, unlike SPI backend which locks the controller until response is received.

Is it a general issue if multiple commands send to EC at a time? If yes, it should serialize that in the UART transportation.

Host Commands only support one command at a time. It's enforced by 'lock' mutex from cros_ec_device structure. We just need to use it properly.

I see. Please use the fixes tag if you'd have chance to send next version:

Fixes f74c7557ed0d ("platform/chrome: cros_ec_proto: Update version on GET_NEXT_EVENT failure")
#4 lore raw
Quoted 41 lines · Tzung-Bi Shih
Quoted 39 lines · Patryk Duda
Quoted 10 lines · Tzung-Bi Shih
Quoted 4 lines · Patryk Duda

The cros_ec_get_host_command_version_mask() function requires that the caller must have ec_dev->lock mutex before calling it. This requirement was not met and as a result it was possible that two commands were sent to the device at the same time.

To clarify:
- What would happen if multiple cros_ec_get_host_command_version_mask() calls at the same time?

In the best case, MCU will receive both commands glued together and will ignore them.
It will result in a timeout in the kernel. In the worst case, request and/or response buffers will be
corrupted.

Quoted 1 line

- What are the callees? I'm trying to understand the source of parallelism.

This is a race between interrupt handling and ioctl call from userspace

Handling interrupt path
cros_ec_irq_thread()
cros_ec_handle_event()
cros_ec_get_next_event() - Queries host command version without taking ec_dev->lock mutex first cros_ec_get_host_command_version_mask() cros_ec_send_command()
cros_ec_xfer_command()
cros_ec_uart_pkt_xfer()

Command from userspace
cros_ec_chardev_ioctl()
cros_ec_chardev_ioctl_xcmd()
cros_ec_cmd_xfer() - Locks ec_dev->lock mutex before sending command cros_ec_send_command()
cros_ec_xfer_command()
cros_ec_uart_pkt_xfer()

Quoted 1 line
Quoted 9 lines
Quoted 5 lines
Quoted 3 lines

Also, the patch also needs an unlock at [1].

[1]: https://elixir.bootlin.com/linux/v6.10/source/drivers/platform/chrome/cros_ec_proto.c#L819

Yeah. I'll fix it in v2

I'm wondering if it's simpler to just lock and unlock around calling cros_ec_get_host_command_version_mask(). What do you think?

Initially, I thought it would be good to keep ec_dev->mkbp_event_supported update under the mutex (similar to cros_ec_query_all() which is called with locked mutex), but mkbp_event_supported is also used without locked mutex.

I don't see any obvious risks with updating the MKBP version outside mutex. Do you want me to change it?

Quoted 12 lines
Quoted 9 lines
Quoted 6 lines
Quoted 3 lines

The problem was observed while using UART backend which doesn't use any additional locks, unlike SPI backend which locks the controller until response is received.

Is it a general issue if multiple commands send to EC at a time? If yes, it should serialize that in the UART transportation.

Host Commands only support one command at a time. It's enforced by 'lock' mutex from cros_ec_device structure. We just need to use it properly.

I see. Please use the fixes tag if you'd have chance to send next version:

Fixes f74c7557ed0d ("platform/chrome: cros_ec_proto: Update version on GET_NEXT_EVENT failure")
maintainer
#5 lore raw
Quoted 12 lines · Patryk Duda
Quoted 11 lines · Tzung-Bi Shih
Quoted 6 lines · Patryk Duda
Quoted 3 lines · Tzung-Bi Shih

Also, the patch also needs an unlock at [1].

[1]: https://elixir.bootlin.com/linux/v6.10/source/drivers/platform/chrome/cros_ec_proto.c#L819

Yeah. I'll fix it in v2

I'm wondering if it's simpler to just lock and unlock around calling cros_ec_get_host_command_version_mask(). What do you think?

Quoted 6 lines

Initially, I thought it would be good to keep ec_dev->mkbp_event_supported update under the mutex (similar to cros_ec_query_all() which is called with locked mutex), but mkbp_event_supported is also used without locked mutex.

I don't see any obvious risks with updating the MKBP version outside mutex. Do you want me to change it?

Yes.

#6 lore raw
Quoted 14 lines · Tzung-Bi Shih
Quoted 12 lines · Patryk Duda
Quoted 11 lines · Tzung-Bi Shih
Quoted 6 lines · Patryk Duda
Quoted 3 lines · Tzung-Bi Shih

Also, the patch also needs an unlock at [1].

[1]: https://elixir.bootlin.com/linux/v6.10/source/drivers/platform/chrome/cros_ec_proto.c#L819

Yeah. I'll fix it in v2

I'm wondering if it's simpler to just lock and unlock around calling cros_ec_get_host_command_version_mask(). What do you think?

Quoted 8 lines
Quoted 6 lines

Initially, I thought it would be good to keep ec_dev->mkbp_event_supported update under the mutex (similar to cros_ec_query_all() which is called with locked mutex), but mkbp_event_supported is also used without locked mutex.

I don't see any obvious risks with updating the MKBP version outside mutex. Do you want me to change it?

Yes.

Fixed in v3: https://lore.kernel.org/lkml/[email protected]