To be fair, I'm currently being tried by MCUXpressoIDE from NXP. I needed a few too many pins to make the STM's work, so I'm trying out an i.mxrt1176 (the higher speed doesn't hurt, either - even the "slow" core is almost as fast as the ST one...)
Anyway, long story short.. Come back ST, I forgive everything. Well, most things... A lot... Ok, some. But still...
It tries to do a lot for you, because the chip is actually pretty complicated and the pin routing is also complicated, but then when it fails with an unspecified "error" or "warning" and refuses to update the code from the pin-map UI, you're left with not knowing where the issue is. It 'helpfully' says it's a 'tool error' in the "problems" pane - I ought to spell that "pain") but there's no indication of which preference, which tool, which part of the code, what setup, anything that could point you in the right direction.
It's a bit like STMCube just failing to calculate your desired clock speed (as in: the one the chip is advertised to run at) because the default power-envelope is set to "trickle in power as if it were gold-dust". Again, no indication on the clocking screen why you can't have (eg:216MHz) but 16MHz is apparently your limit...
Anyway, I got "blink a led" up and running pretty quickly on the i.mxrt chip, but a quick test of 'put an interrupt handler in place' is failing dismally. I can see the signal going into the pin, I have the correct pin on the eval board (helpfully shown in the UI), I have the GPIO set to be an actual GPIO not a separate function, and I have the interrupt set up the same as their example one AFAICT. It just won't actually trigger. Most of the above I've done by hand, copying and adapting from example source, because the darn UI has thrown a wobbly and won't update the separated-off areas of the code that it's supposed to [sigh]
And, NXP's forums have been down for days now. All you get when you go to sign in is a "Sorry, this sh*t is broke".
Gah.