wrong type in declaration of ESP.deepSleep()

[5chufti]

Basic Infos

repeated issue as mentioned in "sdk update 2.1.0" branch

Hardware

Hardware: not relevant
Core Version: 2.1.0-rc2

Description

deepsleep argument for µs is uint64 in sdk but only uint32 in esp8266 core
compare definitions in

esp8266/cores/esp8266/Esp.h
vs.
esp8266/tools/sdk/include/user_interface.h

Settings in IDE

not relevant

Sketch

not relevant

Debug Messages

not relevant

messages here

see attached diff
Esp_core.diff.txt

[igrr]

Thanks, will add an overload with uint64_t. What's the cali_proc for in ESP class, by the way?

…

On Sun, Oct 15, 2017, 04:30 5chufti ***@***.***> wrote: Basic Infos repeated issue as mentioned in "sdk update 2.1.0" branch Hardware Hardware: not relevant Core Version: 2.1.0-rc2 Description deepsleep argument for µs is uint64 in sdk but only uint32 in esp8266 core Settings in IDE not relevant Sketch not relevant Debug Messages not relevant messages here see attached diff Esp_core.diff.txt <https://github.com/esp8266/Arduino/files/1385269/Esp_core.diff.txt> — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#3719>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEJcehUfpR2QA8tuZozKktovMefYYzbkks5sscKhgaJpZM4P5q96> .

[5chufti]

according to documentation this value is needed to calculate max. deep_sleep time.
see #3302
but somehow this value even changes during operation, so it is more "orientation" than exact value.

[igrr]

Right, but would it then be more useful to add a function to ESP class to get the max sleep time?

…

On Sun, Oct 15, 2017, 05:07 5chufti ***@***.***> wrote: according to documentation this value is needed to calculate max. deep_sleep time. see #3302 <#3302> but somehow this value even changes during operation, so it is more "orientation" than exact value. — You are receiving this because you commented. Reply to this email directly, view it on GitHub <#3719 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AEJcehH-4SD0gBcBE1AmhqNfUCWRcgPrks5ssctMgaJpZM4P5q96> .

[5chufti]

whatever you like, just wanted to keep it simple.

[5chufti]

@igrr
again not in current release ....

[5chufti]

@igrr @devyte
could you please at least tag a milestone and/or asign it to you so that this topic isn't lost again in next release?

[igrr]

@5chufti Done, tagged as a bug

[devyte]

@igrr I understand from the sdk api doc that the cali_proc is the RTC period in us. It has an odd format: the 12 lower bits are decimal (i.e.: fraction), the higher bits are integer.
The max deep sleep possible is calculated as a function of the returned cali value. AFAICT, you read the cali, then you calculate the max deepsleep possible, then you check the requested deepsleep against that max, and if it's ok you deepsleep.

-edited for correctness