Hi Guenter,
On Thu, 9 Feb 2012 09:17:56 -0800, Guenter Roeck wrote:
> Add support for temperature offset registers for CPUTIN, SYSTIN,
> and AUXTIN temperatures.
>
> Signed-off-by: Guenter Roeck <guenter.roeck@xxxxxxxxxxxx>
> ---
> drivers/hwmon/w83627ehf.c | 73 +++++++++++++++++++++++++++++++++++++++++++++
> 1 files changed, 73 insertions(+), 0 deletions(-)
I can't test that, overall it looks good, except one thing:
> diff --git a/drivers/hwmon/w83627ehf.c b/drivers/hwmon/w83627ehf.c
> index 57a31a0..5273fff 100644
> --- a/drivers/hwmon/w83627ehf.c
> +++ b/drivers/hwmon/w83627ehf.c
> (...)
> @@ -2204,6 +2265,11 @@ static int __devinit w83627ehf_probe(struct platform_device *pdev)
> data->in6_skip = 1; /* No VIN3 */
>
> data->temp_label = w83667hg_b_temp_label;
> + data->have_temp_offset = data->have_temp & 0x03;
> + for (i = 0; i < 2; i++) {
> + if (data->temp_src[i] > 1)
> + data->have_temp_offset &= ~(1 << i);
> + }
This is the W83627UHG case, and while it has only two analog
temperature inputs, it can have up to 3 temperature channels. temp1 is
hard-coded to SYSTIN, temp2 can be set to PECI and temp3 to CPUTIN, in
which case temp3 would support temperature offset. So you should loop
up to i < 3, not 2.
Once this is fixed, you can add:
Acked-by: Jean Delvare <khali@xxxxxxxxxxxx>
While reviewing this patch, I found that the code which sets the source
for temp2 is broken (for W83627UHG only), I'll send a fix now.
Thanks,
--
Jean Delvare
_______________________________________________
lm-sensors mailing list
lm-sensors@xxxxxxxxxxxxxx
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
