benpicco
benpicco
This now changes the API of `coap_block_finish()` so the caller can account for the removed bytes. I couldn't find any case where the old return value would be used, so...
What I don't like about that patch it that is basically is a per-application workaround :confused: #20855 provides a much simpler fix.
For that special case I could add a new `THREAD_CREATE_NO_STACKTEST` flag
IIRC we have some form of confirm send now?
This needs a rebase.
`tests/net/gnrc_sock_{ip, udp}` are still failing
Thank you, that explanation makes perfect sense! So the (long) option gets first written by `coap_opt_add_block2()`, then it is overwritten later by `coap_block2_finish()` with a shorter option, leaving two garbage...
Very nice! Can you please split up the commit into one for the CPU changes, one for each new board and then another one for the `Makefile.ci` updates? Thank you!
Can we also add `CFLAGS += -Wno-error=deprecated-declarations` so users don't have to fix those immediately? (Which is the whole point of the deprecation period)
I agree with @maribu - why not upstream the CPU code? It's true that upstreaming comes with additional work, but you will benefit in the long run as you don't...