https://devzone.nordicsemi.com/f/nordic-q-a/64723/problem-exposing-nrfx-watchdog-codeHi! I was defining NRFX_WDT_ENABLED in sdk_config.h, but found that apply_old_config.h was undefining it. Even though I had WDT_ENABLED defined to zero, the test in apply_old_config.h is for if it was defined AT ALL, not if it was defined to a non-zeroen-USTelligent Community 13Wed, 12 Aug 2020 12:56:39 GMThttps://devzone.nordicsemi.com/thread/264272?ContentTypeID=1Wed, 12 Aug 2020 12:56:39 GMT137ad170-7792-4731-bb38-c0d22fbe4515:aee75235-2f64-4d99-88dc-39f17e6c5e76BenBurch<p>A comment header in sdk_config.h would likely be a good option here.</p><div style="clear:both;"></div>https://devzone.nordicsemi.com/thread/264222?ContentTypeID=1Wed, 12 Aug 2020 10:33:53 GMT137ad170-7792-4731-bb38-c0d22fbe4515:cbcfbf96-e0d2-46e6-b00c-e4616b92c280Susheel Nuguru<p>Ben,</p> <p>The actual purpose initially for the apply_old_config.h was to take the settings from sdk_config.h and make it work with the NRFX layer.</p> <p>defining anything in sdk_config.h should ideally overseed other defines that are underneath it. So defining _WDT_ENABLED to 0 in sdk_config.h file and defining&nbsp;NRFX_WDT_ENABLED to 1 is wrong.</p> <p>But this is not your mistake, I guess it is not clearly documented on how these integration of macros is happening.</p> <p>If you are intending to directly use nrfx_wdt then do not define the _WDT_ENABLED at all. I think that the check on WDT_ENABLED is correct, but it seems that the documentation on what was attempted to do here could be improved.</p><div style="clear:both;"></div>https://devzone.nordicsemi.com/thread/264125?ContentTypeID=1Tue, 11 Aug 2020 17:11:14 GMT137ad170-7792-4731-bb38-c0d22fbe4515:2eb608d7-064e-4b0d-a085-d4f755032e46BenBurch<p>This would also fail in the legacy case that apply_old_config.h was hoping to solve, as defining WDT_ENABLED prevents compilation of the new nrf wdt code</p><div style="clear:both;"></div>