Re: [MacPorts] #65034: rsync fails to upgrade to @3.2.4 on 10.6.8 when building as ppc using gcc-4.2 error: pointer targets in passing argument 2 of ‘getgroups’ differ in signedness

[MacPorts]

#65034: rsync fails to upgrade to @3.2.4 on 10.6.8 when building as ppc using
gcc-4.2 error: pointer targets in passing argument 2 of ‘getgroups’ differ
in signedness
---------------------------+-------------------------------------------
  Reporter:  barracuda156  |      Owner:  ryandesign
      Type:  defect        |     Status:  assigned
  Priority:  Normal        |  Milestone:
 Component:  ports         |    Version:  2.7.2
Resolution:                |   Keywords:  powerpc, snowleopard, rosetta
      Port:  rsync         |
---------------------------+-------------------------------------------

Comment (by ryandesign):

 Ok, the log shows this configure output:
 {{{
 :info:configure checking type of array argument to getgroups... int
 }}}
 That's suspicious: `getgroups` actually takes a `gid_t` which is a
 `__darwin_gid_t` which is a `__uint32_t` which is an unsigned integer, vs.
 `int` which is a signed integer, hence the message about differing
 signedness.

 The configure script gets the correct result on my Catalina system and on
 the buildbot Snow Leopard machines too. Something about your cross-
 compiling scenario seems to be throwing off the detection. Could you
 attach your config.log so we can see what happened?

 This determination of the type of arguments for `getgroups` is implemented
 in a standard autoconf macro `AC_TYPE_GETGROUPS`. There is a comment in
 [https://github.com/autotools-
 mirror/autoconf/blob/64df9b4523fe3b3c5d6e444fb65b1d92fc5073ce/lib/autoconf/types.m4#L258-L300
 the code for that macro] that talks about how it will use `int` when cross
 compiling on old systems that lack prototypes. That explanation doesn't
 make sense for Mac OS X, all versions of which are new enough to have
 prototypes, so I think some other failure, which the config.log should
 tell us about, is confusing this check into giving the wrong answer.

-- 
Ticket URL: <https://trac.macports.org/ticket/65034#comment:4>
MacPorts <https://www.macports.org/>
Ports system for macOS