Re: lib/59711: "#define HAVE_UNISTD_H 1" breaks 32-bit libz

["Tom Lane via gnats" <gnats-admin%NetBSD.org@localhost>]

The following reply was made to PR lib/59711; it has been noted by GNATS.

From: Tom Lane <tgl%sss.pgh.pa.us@localhost>
To: gnats-bugs%netbsd.org@localhost
Cc: 
Subject: Re: lib/59711: "#define HAVE_UNISTD_H 1" breaks 32-bit libz
Date: Sun, 19 Oct 2025 00:29:09 -0400

 Oh!  I suddenly realized why it appears to be fine on evbarm:
 there's an endianness dependency.  The gzgetc() macro just wants
 to increment the pos field, and if the machine is little-endian
 then it doesn't matter too much whether it thinks pos is 32 or
 64 bits.  (It would matter in the event of overflow of the
 32-bit value, but that won't happen.)  But macppc is big-endian,
 so if gzgetc() thinks pos is 64 bits then it will actually increment
 what the library thinks is the "mode" field, setting off subsequent
 sanity checks.
 
 So this is probably only a live issue on 32-bit big-endian platforms.
 
 			regards, tom lane