(resolved) stack overflow in task memory_task?

[mzimmers]

Hi all -

This error crept up this morning. It occurs about 630 ticks into boot.

Code: Select all

***ERROR*** A stack overflow in task memory_task has been detected.

(WROVER, V4.0 IDF)

Any ideas on what to do? Thanks...

[mzimmers]

As additional information, sometimes upon boot, I get this message instead:

Code: Select all

E (30492) phy_init: failed to allocate memory for RF calibration data

The app manages a few calls to heap_caps_get_free_size before it crashes:

Code: Select all

I (7710) memreport: memReport(): there are 8728 MALLOC_CAP_8BIT bytes free.
I (7720) memreport: memReport(): there are 61520 MALLOC_CAP_32BIT bytes free.
I (7720) memreport: memReport(): there are 61520 MALLOC_CAP_INTERNAL bytes free.
I (7730) memreport: memReport(): there are 0 MALLOC_CAP_SPIRAM bytes free.
I (7740) memreport: memReport(): there are 8728 MALLOC_CAP_DMA bytes free.

I'd *greatly* appreciate any assistance on this...thanks.

[mzimmers]

Bump.

Any ideas on how I can chase this down? Thank you.

[vonnieda]

Is memory_task a task you created? If so, just increase the task stack size. Most likely you are using printf and the formatting is taking more stack memory than it originally did.

Jason

[mzimmers]

Hi Jason -

Yes, it's a task I create. It's not an essential task, so I can get rid of it, but then other memory-related issues pop up. Currently, my wifi task throws this error:

Code: Select all

E (32104) phy_init: failed to allocate memory for RF calibration data
abort() was called at PC 0x4018fe46 on core 0

I'm guessing this is heap, not stack, which is why I was running the memcheck in the first place. But...something very squirrelly is happening with my memory management.

[mzimmers]

Marking as resolved. I'm simply running out of memory.