Re: [PATCH] platform/chrome: cros_ec_rpmsg: record last_event_time on host event

[Tzung-Bi Shih <tzungbi@kernel.org>]

On Tue, Nov 03, 2026 at 02:28:02PM +0800, Cheng Yueh wrote:
> Transports with a dedicated interrupt line (ec_dev->irq > 0, such as SPI
> and I2C) record ec_dev->last_event_time in the top-half
> cros_ec_irq_handler(), while transports without one (ec_dev->irq == 0,
> such as LPC in cros_ec_lpc_acpi_notify() and ISHTP in ish_event_cb())
> record it in their respective event callbacks.
> 
> cros_ec_rpmsg also operates with ec_dev->irq == 0: when an MKBP event
> notification (HOST_EVENT_MARK) arrives in cros_ec_rpmsg_callback(), it
> schedules host_event_work to invoke the bottom-half cros_ec_irq_thread()
> directly, bypassing cros_ec_irq_handler() and leaving
> ec_dev->last_event_time as 0.
> 
> cros_ec_sensorhub_ring relies on ec_dev->last_event_time (via
> fifo_timestamp) as the AP CLOCK_BOOTTIME reference timestamp to calibrate
> the offset and clock drift between the remote processor's local hardware
> timer and the AP timebase. Without this calibration, IIO sensor sample
> timestamps cannot be synchronized to CLOCK_BOOTTIME
> (SystemClock.elapsedRealtimeNanos()), failing the Android CTS sensor
> specification TimestampClockSourceVerification.
> 
> Record ec_dev->last_event_time = cros_ec_get_time_ns() upon receiving
> HOST_EVENT_MARK in cros_ec_rpmsg_callback(), matching the other ChromeOS
> EC transport drivers.
> 
> Fixes: 2de89fd98958 ("platform/chrome: cros_ec: Add EC host command support using rpmsg")

`last_event_time` was introduced later than this commit, in 05a3c420eaa6
("platform/chrome: cros-ec: Record event timestamp in the hard irq").

Applied to

    https://git.kernel.org/pub/scm/linux/kernel/git/chrome-platform/linux.git for-next

[1/1] platform/chrome: cros_ec_rpmsg: record last_event_time on host event
      commit: 4efef5ee9c41dc7cb9fa8ed67c0831be669702ca

I have shortened the commit message (and updated the Fixes tag) as the
original one was a bit verbose.  Providing extra background context is
helpful for review, but please consider putting it below the "---"
separator in the future so that they will not be in the commit message.

Thanks!