pkgsrc-Bugs archive

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]

Re: pkg/41125 (pkgsrc's doesn't work on Solaris)

The following reply was made to PR pkg/41125; it has been noted by GNATS.

From: Tim Zingelman <>
Subject: Re: pkg/41125 (pkgsrc's doesn't work on Solaris)
Date: Mon, 6 Apr 2009 21:12:22 -0500

 Thanks for the additional info.
 I agree that there is a bug here to be fixed, indeed the solaris
 pkgsrc bootstrap would hit the bug except for the MKPROFILE=no in
 pkgsrc/net/libfetch/files/Makefile.  I just want to be sure it gets
 fixed correctly.
 Looking at Simon Gerraty's code... in particular at "$Id:,v
 1.30 2007/11/24 22:12:58 sjg Exp $" it appears that setting an empty
 LD_X has the side effect of skipping the use of LD entirely:
 .if (${LD_X} == "")
         ${COMPILE.c} ${CC_PG} ${.IMPSRC} -o ${.TARGET}
 .cc.po .C.po:
         ${} -pg ${.IMPSRC} -o ${.TARGET}
         ${COMPILE.S} ${PICFLAG} ${CC_PIC} ${CFLAGS:M-[ID]*} ${AINC}
 ${.IMPSRC} -o ${.TARGET}
 ...code that uses the LD_X variable...
 This seems quite similar to the effect of setting COPTS=-g when using
 .if defined(COPTS) && !empty(COPTS:M*-g*)
         ${COMPILE.c} -pg ${.IMPSRC} -o ${.TARGET}
 ... code that does the ${LD} -X ...
 The more I look at that, the more I think the code is more
 correct in handling that... assuming -X is bad for debugging?
 What is the practical difference in the results between doing:
   gcc -O  -Werror   -c MyLib.c -o MyLib.o.o
   ld -r MyLib.o.o -o MyLib.o
   rm -f MyLib.o.o
 and just doing:
   gcc -O -Werror   -c MyLib.c
 and is the answer different using solaris /usr/ccs/bin/ld vs gnu ld?

Home | Main Index | Thread Index | Old Index