On 16.12.2020 08:29, Heinrich Schuchardt wrote: > On 12/16/20 8:08 AM, Marek Szyprowski wrote: >> On 15.12.2020 20:07, Heinrich Schuchardt wrote: >>> On 12/15/20 5:54 PM, Marek Szyprowski wrote: >>>> Save examined button state in 'button' environment variable to enable >>>> checking button state in the scripts. >>>> >>>> Signed-off-by: Marek Szyprowski <m.szyprow...@samsung.com> >>>> --- >>>> Resend reason: get rid of the Change-Id tag, that was still in v2. >>>> --- >>>> cmd/button.c | 4 +++- >>>> 1 file changed, 3 insertions(+), 1 deletion(-) >>>> >>>> diff --git a/cmd/button.c b/cmd/button.c >>>> index 64c5a8fa04..8da911068a 100644 >>>> --- a/cmd/button.c >>>> +++ b/cmd/button.c >>>> @@ -23,8 +23,10 @@ static int show_button_state(struct udevice *dev) >>>> ret = button_get_state(dev); >>>> if (ret >= BUTTON_COUNT) >>>> ret = -EINVAL; >>>> - if (ret >= 0) >>>> + if (ret >= 0) { >>>> printf("%s\n", state_label[ret]); >>>> + env_set("button", state_label[ret]); >>> >>> Using a hard coded environment variable does not make much sense to me. >>> The button command has a return value. So just use >>> >>> button mybutton; setenv myvar $? >>> >> Thanks for the hint, I wasn't aware that uboot supports '$?'. By setting >> the 'button' env variable I've tried to mimic the behavior of the >> various network and file related commands, which sets 'filesize' env >> variable. >> >> I will need to add the return value propagation to the button command >> anyway to make it usable from the scripts. > > Nothing to be done for the button command. Since > > a6bfd71a96201127836d59736abcb54dc2d5e1a5 > cmd/button: return button status > > the button command returns > > 0 (true) - pressed, on > 1 (false) - not pressed, off > Right, I've missed that commit. I've developed my code on top of v2020.10 release.
Best regards -- Marek Szyprowski, PhD Samsung R&D Institute Poland