Re: [BUG] Compilation warnings with 29.4
Joel Reicher <[email protected]> Fri, 12 Jul 2024 21:34:36 +1000
| Newsgroups | gmane.emacs.windows |
|---|---|
| Message-ID | <[email protected]> |
Eli Zaretskii <[email protected]> writes: >> From: Joel Reicher <[email protected]> >> Cc: [email protected] >> Date: Fri, 12 Jul 2024 21:10:51 +1000 >> >> I'm not suggesting this is anything other than a (very)=20 >> temporary workaround, and I was already aware of this variable=20 >> from the documentation, but it wasn't clear to me the=20 >> conditions under which Emacs would try to generate a=20 >> trampoline. >> >> It was interesting to see it happen. > > The ELisp manual says about this: > > -- Variable: native-comp-enable-subr-trampolines > This variable controls generation of trampolines. A=20 > trampoline is > a small piece of native code required to allow calling=20 > Lisp > primitives, which were advised or redefined, from Lisp=20 > code that > was natively-compiled with =A1=AEnative-comp-speed=A1=AF set to 2= =20 > or greater. > Emacs stores the generated trampolines on separate=20 > =A1=AE*.eln=A1=AF files. I know, I read this weeks ago, but I could never have guessed that=20 in my usual workflow it wouldn't trigger until I called WoMan. And, to be honest, I still can't see how to connect that=20 documentation with the behaviour I saw, but I haven't yet looked=20 at the WoMan code that triggered this. Thanks and regards, - Joel