Code size report:

Reference:  unix/README: Update the supported targets list. [d901e98]
Comparison: Revert "rp2: Build with nano.specs for newlib-nano.". [merge of c2ceed1]
  mpy-cross:    +0 +0.000% 
   bare-arm:    +0 +0.000% 
minimal x86:    +0 +0.000% 
   unix x64:    +0 +0.000% standard
      stm32:    +0 +0.000% PYBV10
      esp32:    +0 +0.000% ESP32_GENERIC
     mimxrt:    +0 +0.000% TEENSY40
        rp2: +1180 +0.128% RPI_PICO_W[incl +720(bss)]
       samd:    +0 +0.000% ADAFRUIT_ITSYBITSY_M4_EXPRESS
  qemu rv32:    +0 +0.000% VIRT_RV32

Yeah, I feel like this is the right thing to do. Keeps things simple which is always a good thing.

(For the mbdetls change, gtime_r, probably that's a good change that can stay because it eliminates an extra path through mbedtls to do an unnecessary memcpy. But I guess that can be done separately, and also done for all other ports.)

I am forced to agree. Though am continuing to cut firmware builds with this change, which is something I'll probably want to address in retrospect.

I carried this technique over from RP2040 and didn't consider its implications on RP2350. Sorry!

Edit: A quick investigation suggests roughly a 30% performance hit to bm_pidigits. No wonder I was seeing such a huge gain when I found those double-zero-init bugs.

I carried this technique over from RP2040 and didn't consider its implications on RP2350. Sorry!

No worries. We are all leaning something here 😄

A quick investigation suggests roughly a 30% performance hit to bm_pidigits. No wonder I was seeing such a huge gain when I found those double-zero-init bugs.

Ah, that makes a lot more sense now! (see #18771 for reference)