- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
With the 20719 we notice that only, roughly, 1800 bytes of the 16KB AON is available for our use.
The comments in make_cortex-m4_gcc.mk mention that AON is being utilized by patches; in addition to the 64KB already taken from main RAM for use by Patches.
If there a map that shows memory that comes available when certain patches are not loaded?
I expect that since we are not using WIFI, for example, none of the wifi patches would be loaded, and that should free up some memory in main "Patch" RAM or AON.
Another note mentions that Pairing information may be written to AON memory. Is that operation taking from the 1800 bytes "available" or is some amount reserved in the 14+K already taken by the system?
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi todd_3568211,
As given in make_cortex-m4_gcc.mk, some memory is reserved for firmware patches. We have around 2KB memory, which is available for application code. The pairing information is stored within this 2KB only.
/20719-B1_Bluetooth/WICED/internal/20719B1/patches/patch.lst gives the AON free memory starting address.
Please check following blog post for memory layout: CYW20719 Memory Layout
We do not have any memory map that shows the memory structure with and without certain patches.
CYW20719 is a BT chip, WIFI does not come in the scenario here.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
But, what is the purpose of needing so much AON? What sort of persistent data is being maintained?
We'd like to store some persistent data but the patches are effectively reserving all of it. Couldn't the patch use regular RAM? (Although we'd like more of that back as well.)
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content