From 712177cf1421d6c8551476a783fdf28dfe57b4aa Mon Sep 17 00:00:00 2001 From: Randall Spangler Date: Mon, 21 Oct 2013 13:26:56 -0700 Subject: [PATCH] cleanup: Thermal comments No code changes, just replacing a FIXME from the comments with a more thorough explanation. BUG=chrome-os-partner:20805 BRANCH=none TEST=build falco Change-Id: Ibd98322c2b9fd6e0447771ce5fe43e0283743c60 Signed-off-by: Randall Spangler Reviewed-on: https://chromium-review.googlesource.com/173930 Reviewed-by: Bill Richardson --- common/thermal.c | 14 +++++++++++--- 1 file changed, 11 insertions(+), 3 deletions(-) diff --git a/common/thermal.c b/common/thermal.c index c6c49868d2..26775f62eb 100644 --- a/common/thermal.c +++ b/common/thermal.c @@ -96,9 +96,17 @@ static void thermal_control(void) } if (!num_sensors_read) { - /* If we can't read any sensors, do nothing and hope - * it gets better. - * FIXME: What *should* we do? + /* + * Trigger a SMI event if we can't read any sensors. + * + * In theory we could do something more elaborate like forcing + * the system to shut down if no sensors are available after + * several retries. This is a very unlikely scenario - + * particularly on LM4-based boards, since the LM4 has its own + * internal temp sensor. It's most likely to occur during + * bringup of a new board, where we haven't debugged the I2C + * bus to the sensors; forcing a shutdown in that case would + * merely hamper board bringup. */ smi_sensor_failure_warning(); return;