I attempted to install devel/statsvn, but bsd.java.mk complained about an environment error with JAVA_HOME. I checked my local shell environment, and JAVA_HOME was not set. I looked through bsd.java.mk, and it appears that JAVA_HOME is set when USE_ANT=yes is in effect, but a check-makevars target then abort the make process because that variable is set. How-To-Repeat: cd /usr/ports/devel/statsvn && make install clean
Responsible Changed From-To: freebsd-ports-bugs->glewis bsd.java.mk is glewis territory (via the GNATS Auto Assign Tool)
State Changed From-To: open->feedback I'm able to build statsvn and other USE_ANT ports without any problems. The check-makevars target has been in bsd.java.mk for a long time, so it seems unlikely that it's suddenly become a problem. Do you definitely get the error from the check-makevars target? Are you sure there isn't a setting somewhere else that is being picked up as part of the build process? Maybe add the value of JAVA_HOME to the error message so you can see what it is.
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Ok, I found the problem, and it was a local issue as you correctly surmised. I forgot that my bsd.port.mk has already incorporated the patch found in http://www.freebsd.org/cgi/query-pr.cgi?pr=ports/164423, and this caused the JAVA_HOME variable to be propagated when it hadn't been before. My main goal with ports/164423 is to pass the make environment to sub-make processes during port configuration, so maybe the other places I pass it should be omitted. On the other hand, is there any reason not to unset the environment variables that should not be set in Stage 2 of bsd.java.mk instead of issuing an error? Thank you, Greg -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk86h8IACgkQ0sRouByUApCfAgCgm3FBUSbRaPvyhDP71nsdgi+t jlcAn1FU7NftB3Nc4npVP9FlLsnlk0ii =BmRE -----END PGP SIGNATURE-----
I tend to agree that we could just unset the troublesome variables. How do you feel about the following patch? Index: bsd.java.mk =================================================================== RCS file: /var/fcvs/ports/Mk/bsd.java.mk,v retrieving revision 1.92 diff -u -r1.92 bsd.java.mk --- bsd.java.mk 9 Feb 2012 06:40:49 -0000 1.92 +++ bsd.java.mk 26 Feb 2012 01:21:12 -0000 @@ -224,8 +224,8 @@ . for variable in JAVA_HOME JAVA_PORT JAVA_PORT_VERSION JAVA_PORT_VENDOR JAVA_PORT_OS . if defined(${variable}) check-makevars:: - @${ECHO_CMD} "${PKGNAME}: Environment error: \"${variable}\" should not be defined." - @${FALSE} + @${ECHO_CMD} "${PKGNAME}: Environment error: \"${variable}\" should not be defined -- clearing." +.undef ${variable} . endif . endfor I'm also cc'ing Herve (co-maintainer) to see if he has any objections. -- Greg Lewis Email : glewis@eyesbeyond.com Eyes Beyond Web : http://www.eyesbeyond.com Information Technology FreeBSD : glewis@FreeBSD.org
On Mon, Mar 12, 2012 at 10:25:53PM -0700, Greg Lewis wrote: > I tend to agree that we could just unset the troublesome variables. > How do you feel about the following patch? I agree, the port should not fail but rather just cope with the set variables. Just a quick note though: > + @${ECHO_CMD} "${PKGNAME}: Environment error: \"${variable}\" should not be defined -- clearing." IMHO "error" is no longer a relevant term here. It's more of a warning message. Other than that, no objection to the patch. Herve
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 On 3/13/12 3:28 AM, Herve Quiroz wrote: > On Mon, Mar 12, 2012 at 10:25:53PM -0700, Greg Lewis wrote: >> I tend to agree that we could just unset the troublesome >> variables. How do you feel about the following patch? > > I agree, the port should not fail but rather just cope with the > set variables. > > Just a quick note though: > >> + @${ECHO_CMD} "${PKGNAME}: Environment error: \"${variable}\" >> should not be defined -- clearing." > > IMHO "error" is no longer a relevant term here. It's more of a > warning message. Other than that, no objection to the patch. > > Herve Thank you both for looking into that. The patch looks good to me! Cheers, Greg - -- Greg Larkin http://www.FreeBSD.org/ - The Power To Serve http://www.sourcehosting.net/ - Ready. Set. Code. http://twitter.com/cpucycle/ - Follow you, follow me -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.11 (Darwin) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org/ iEYEARECAAYFAk9fodQACgkQ0sRouByUApAh7wCeP/c50xe1pMznr83amRzPMQio 46QAnjUFW0IQDw31vOHkLeqlJqqrVUCp =sdGp -----END PGP SIGNATURE-----
glewis 2012-04-13 04:03:42 UTC FreeBSD ports repository Modified files: Mk bsd.java.mk Log: . Instead of bailing when we hit a possibly conflicting variable being defined just clear it and issue a warning that it was set. PR: 165115 Approved by: hq@ (co-maintainer) Revision Changes Path 1.93 +3 -3 ports/Mk/bsd.java.mk _______________________________________________ cvs-all@freebsd.org mailing list http://lists.freebsd.org/mailman/listinfo/cvs-all To unsubscribe, send any mail to "cvs-all-unsubscribe@freebsd.org"
State Changed From-To: feedback->closed Patch committed. Thanks all!