Installing the port for the first time consistently fails for me on 4.0 and 3.4 machines: ===> linux_base-6.1 depends on executable: rpm - found setup-2.0.5-1.noarch.rpm filesystem-1.3.5-1.noarch.rpm basesystem-6.0-4.noarch.rpm ldconfig-1.9.5-15.i386.rpm glibc-2.1.2-11.i386.rpm termcap-9.12.6-15.i386.rpm libtermcap-2.0.8-18.i386.rpm bash-1.14.7-16.i386.rpm ncurses-4.2-25.i386.rpm zlib-1.1.3-5.i386.rpm info-3.12h-2.i386.rpm execution of script failed error: /usr/ports/distfiles/rpm/info-3.12h-2.i386.rpm cannot be installed *** Error code 1 Stop in /3/ports/emulators/linux_base. [...] If I simply repeat ``make install'' again, everything goes smoothly. How-To-Repeat: Try installing the port on a machine for the first time.
Responsible Changed From-To: freebsd-ports->marcel Over to maintainer
Hi Mikhail, The port doesn't fail on bento.freebsd.org. Neither for -current nor for -stable. AFAICT, this means that the port is basicly all right. Can you tell me more about your machines, specifically wrt to anything that has any relation to GNU info. -- Marcel Moolenaar mail: marcel@cup.hp.com / marcel@FreeBSD.org tel: (408) 447-4222
|>Description: | Installing the port for the first time consistently fails | for me on 4.0 and 3.4 machines: | |===> linux_base-6.1 depends on executable: rpm - found |setup-2.0.5-1.noarch.rpm |filesystem-1.3.5-1.noarch.rpm |basesystem-6.0-4.noarch.rpm |ldconfig-1.9.5-15.i386.rpm |glibc-2.1.2-11.i386.rpm |termcap-9.12.6-15.i386.rpm |libtermcap-2.0.8-18.i386.rpm |bash-1.14.7-16.i386.rpm |ncurses-4.2-25.i386.rpm |zlib-1.1.3-5.i386.rpm |info-3.12h-2.i386.rpm |execution of script failed |error: /usr/ports/distfiles/rpm/info-3.12h-2.i386.rpm cannot be installed |*** Error code 1 Exactly the same thing hits me on both 3.4-STABLE and today's 5.0-CURRENT and I can reproduce it and also have a workaround. The principle: Bento often differs from mine in its environment variable. How to repeat: /usr/ports/emulators/linux_base# LC_ALL=ja_JP.EUC make install set LC_ALL variable other than ``C'' and install-info dumps core. and also verify if this works: /usr/ports/emulators/linux_base# LC_ALL=C make install Have the install done without error? fine. Then the workaround is obvious. Please test the following patch. Index: Makefile =================================================================== RCS file: /home/ncvs/ports/emulators/linux_base/Makefile,v retrieving revision 1.41 diff -u -r1.41 Makefile --- Makefile 2000/04/12 04:55:31 1.41 +++ Makefile 2000/04/29 06:22:54 @@ -96,6 +96,7 @@ xpm-3.4k-1.${MACHINE_ARCH}.rpm DBPATH= /var/lib/rpm +RPM= env LC_ALL=C rpm RPMFLAGS= --ignoreos --root ${PREFIX} --dbpath ${DBPATH} \ --nodeps --replacepkgs RPMDIR= ${DISTDIR}/${DIST_SUBDIR} @@ -137,7 +138,7 @@ fi @${MKDIR} ${PREFIX}/${DBPATH} @${MKDIR} ${PREFIX}/var/tmp - @rpm --initdb --root ${PREFIX} --dbpath ${DBPATH} + @${RPM} --initdb --root ${PREFIX} --dbpath ${DBPATH} # # Make sure we have a /dev/null in the chrooted environment. # @@ -155,30 +156,30 @@ # @for R in ${RPM_SET1}; do \ ${ECHO} $$R; \ - rpm -U ${RPMFLAGS} ${RPMDIR}/$$R; \ + ${RPM} -U ${RPMFLAGS} ${RPMDIR}/$$R; \ done @${ECHO} ${RPM_LDCONFIG} - @rpm -U ${RPMFLAGS} --noscripts ${RPMDIR}/${RPM_LDCONFIG} + @${RPM} -U ${RPMFLAGS} --noscripts ${RPMDIR}/${RPM_LDCONFIG} @brandelf -t Linux ${PREFIX}/sbin/ldconfig @${PREFIX}/sbin/ldconfig @${TOUCH} ${PREFIX}/etc/ld.so.conf @for R in ${RPM_SET2}; do \ ${ECHO} $$R; \ - rpm -U ${RPMFLAGS} ${RPMDIR}/$$R; \ + ${RPM} -U ${RPMFLAGS} ${RPMDIR}/$$R; \ done @${ECHO} ${RPM_RPM} - @rpm -U ${RPMFLAGS} --noscripts ${RPMDIR}/${RPM_RPM} + @${RPM} -U ${RPMFLAGS} --noscripts ${RPMDIR}/${RPM_RPM} @brandelf -t Linux ${PREFIX}/bin/rpm @for R in ${RPM_SET3}; do \ ${ECHO} $$R; \ - rpm -U ${RPMFLAGS} ${RPMDIR}/$$R; \ + ${RPM} -U ${RPMFLAGS} ${RPMDIR}/$$R; \ done # # Install updates # @for R in ${PATCHFILES}; do \ ${ECHO} $$R; \ - rpm -U ${RPMFLAGS} ${RPMDIR}/$$R; \ + ${RPM} -U ${RPMFLAGS} ${RPMDIR}/$$R; \ done # # Install yp.conf as a hint to NIS users I suppose you have much cleaner way to solve this. :) -- FUJISHIMA Satsuki
This may be it, for my LANG is set to ru_SU.KOI8-R on all of the machines. (Too bad, there is no Ukrainian locale yet...). [....] =Bento often differs from mine in its environment variable. = =How to repeat: =/usr/ports/emulators/linux_base# LC_ALL=ja_JP.EUC make install = =set LC_ALL variable other than ``C'' and install-info dumps core. [....] -mi
At Sat, 29 Apr 2000 17:55:50 -0400 (EDT), Mikhail Teterin <mi@kot.ne.mediaone.net> wrote: > > This may be it, for my LANG is set to ru_SU.KOI8-R on all of the machines. > (Too bad, there is no Ukrainian locale yet...). That is exactly what I intended. Have you tried my patch? Setting LC_ALL=C temporarily during you run rpm doesn't hurt your environment, it only affects under ``make'' you just typed. -- FUJISHIMA Satsuki
On 3 May, FUJISHIMA Satsuki wrote: = At Sat, 29 Apr 2000 17:55:50 -0400 (EDT), = Mikhail Teterin <mi@kot.ne.mediaone.net> wrote: = > = > This may be it, for my LANG is set to ru_SU.KOI8-R on all of the machines. = > (Too bad, there is no Ukrainian locale yet...). = = That is exactly what I intended. Have you tried my patch? No, all of my machines already have Linux "subsytem" installed. And it seems too simple to worry, that it will be different for me then it is for you :) = Setting LC_ALL=C temporarily during you run rpm doesn't hurt your = environment, it only affects under ``make'' you just typed. I agree. And if that's the fix -- I'm all for it, although it looks more like a workaround... :) I'd say, commit it, but don't close the PR until we find out what's really wrong there. -mi
State Changed From-To: open->closed Patch committed. Thanks!