By selectively collecting an allocation, we can skip scanning many
allocations for pointers because we know up front they won't have
them. This helps a ton when large buffers are being used and memory is
slow (PSRAM). In one Fruit Jam example GC times drop from 80+ms to
~25ms. The example uses a number of bitmaps that are now no longer
scanned.
* Use 0 bytes for QSTR hash. Fixes#10151
* Update libgcc used on SAMD21 to 14.2. Reduces test build by over 100 bytes.
* Turn off 'd' typecode support on non-full builds.
This port is meant to grow to encompass all existing boards. For
now, it is a port while we transition over.
It is named `zephyr-cp` to differentiate it from the MicroPython
`zephyr` port. They are separate implementations.
This simplifies allocating outside of the VM because the VM doesn't
take up all remaining memory by default.
On ESP we delegate to the IDF for allocations. For all other ports,
we use TLSF to manage an outer "port" heap. The IDF uses TLSF
internally and we use their fork for the other ports.
This also removes the dynamic C stack sizing. It wasn't often used
and is not possible with a fixed outer heap.
Fixes#8512. Fixes#7334.
This reduces the stack frame size of mp_builtin___import__ by
limiting the support path length of files from 256 to 96. This
function can be called recursively for nested imports so it adds up.
Also reduce mp_execute_bytecode (vm.c) from 206 a bc call to 124.
This too is recursive and adds up. It is reduced by preventing
some inlining. It may decrease performance slightly when importing
and unpacking.
Adds two new scripts for debugging. One is used from gdb to print
frame sizes in a backtrace. The other prints what pcs use a
particular stack offset. This helps find infrequently used stack
space.
Fixes#8053.
In #7497 port_background_task was renamed to port_background_tick
but the actual call site wasn't changed. This meant that it was
no longer called!
Rename more functions from task to tick to make it clearer which is
which.
