Re: Remove useless casts to (void *)
Peter Eisentraut <peter@eisentraut.org>
From: Peter Eisentraut <peter@eisentraut.org>
To: Tom Lane <tgl@sss.pgh.pa.us>
Cc: pgsql-hackers <pgsql-hackers@postgresql.org>
Date: 2024-11-14T08:59:07Z
Lists: pgsql-hackers
On 29.10.24 15:20, Tom Lane wrote: > Peter Eisentraut <peter@eisentraut.org> writes: >> There are a bunch of (void *) casts in the code that don't make sense to >> me. I think some of these were once necessary because char * was used >> in place of void * for some function arguments. And some of these were >> probably just copied around without further thought. I went through and >> cleaned up most of these. I didn't find any redeeming value in these. >> They are just liable to hide actual problems such as incompatible types. >> But maybe there are other opinions. > > I don't recall details, but I'm fairly sure some of these prevented > compiler warnings on some (old?) compilers. Hard to be sure if said > compilers are all gone. > > Looking at the sheer size of the patch, I'm kind of -0.1, just > because I'm afraid it's going to create back-patching gotchas. > I don't really find that it's improving readability, though > clearly that's a matter of opinion. I did a bit of archeological research on these. None of these casts were ever necessary, and in many cases even the original patch that introduced an API used the coding style inconsistently. So I'm very confident that there are no significant backward compatibility or backpatching gotchas here. I'm more concerned that many of these just keep getting copied around indiscriminately, and this is liable to hide actual type mismatches or silently discard qualifiers. So I'm arguing in favor of a more restrictive style in this matter.
Commits
-
Remove PointerIsValid()
- a5b35fcedb54 19 (unreleased) landed
-
Define __EXTENSIONS__ on Solaris, too.
- 3b08d5224d7d 18.0 landed
-
Define _POSIX_C_SOURCE as 200112L on Solaris.
- 32a7deb2a02b 18.0 landed
-
Remove useless casts to (void *)
- 7f798aca1d5d 18.0 landed