[3.11] gh-91162: Fix substitution of unpacked tuples in generic aliases (GH-92335)

[miss-islington]

(cherry picked from commit 9d25db9)

Co-authored-by: Serhiy Storchaka storchaka@gmail.com

[miss-islington]

@serhiy-storchaka: Status check is done, and it's a success ✅ .

[miss-islington]

@serhiy-storchaka: Status check is done, and it's a success ✅ .

[miss-islington]

@serhiy-storchaka: Status check is done, and it's a success ✅ .

[Fidget-Spinner]

Huh, the strings are tied to runtime state, which is tied to the ABI, so the ABI test is now failing. Should we regenerate the ABI data file or just change the code to lookup the C const char *(which will be slightly slower)?

CC @pablogsal

[pablogsal]

Huh, the strings are tied to runtime state, which is tied to the ABI, so the ABI test is now failing. Should we regenerate the ABI data file or just change the code to look up the C const char *(which will be slightly slower)?

In this case, you can regenerate the ABI, but this will not be possible in the RC state

[miss-islington]

@serhiy-storchaka: Status check is done, and it's a failure ❌ .

[serhiy-storchaka]

How to merge this PR?

[Fidget-Spinner]

@serhiy-storchaka please see this message, or the whole bpo issue.

In the past, I had to get @pablogsal to regenerate it, because we don't have a devguide article on this and I was confused.

[pablogsal]

I will merge this today. Will also try to add some information to the devguide, but in general the RM needs to approve ABI regenerations.