Summary: | =app-crypt/pinentry-0.8.2 fails with sys-libs/ncurses[tinfo] | ||
---|---|---|---|
Product: | Gentoo Linux | Reporter: | Justin Lecher (RETIRED) <jlec> |
Component: | Current packages | Assignee: | Crypto team [DISABLED] <crypto+disabled> |
Status: | RESOLVED FIXED | ||
Severity: | normal | CC: | alonbl, mgmadden |
Priority: | Normal | ||
Version: | unspecified | ||
Hardware: | All | ||
OS: | Linux | ||
Whiteboard: | |||
Package list: | Runtime testing required: | --- | |
Bug Depends on: | |||
Bug Blocks: | 457530 | ||
Attachments: |
pinentry-0.8.2.ebuild.diff
pinentry-0.8.2-ncurses.patch pinentry-0.8.2-ncurses.patch build log for my x86_64 unit automake.out file in addition to my build-log |
Description
Justin Lecher (RETIRED)
2013-05-03 13:19:26 UTC
Hi, I cannot reproduce... any hint? sys-libs/ncurses-5.9-r2 was built with the following: USE="gpm tinfo unicode -ada -cxx -debug -doc -minimal -profile -static-libs -trace" ABI_X86="64" --- make[2]: Leaving directory `/var/tmp/portage/app-crypt/pinentry-0.8.2/work/pinentry-0.8.2/pinentry' Making all in curses make[2]: Entering directory `/var/tmp/portage/app-crypt/pinentry-0.8.2/work/pinentry-0.8.2/curses' x86_64-pc-linux-gnu-gcc -DHAVE_CONFIG_H -I. -I.. -I/usr/include/ncursesw -I../pinentry -Wall -O2 -march=native -fomit-frame-pointer -pipe -Wall -Wcast-align -Wshadow -Wstrict-prototypes -Wno-pointer-sign -c pinentry-curses.c x86_64-pc-linux-gnu-gcc -O2 -march=native -fomit-frame-pointer -pipe -Wall -Wcast-align -Wshadow -Wstrict-prototypes -Wno-pointer-sign -Wl,-O1 -Wl,--as-needed -o pinentry-curses pinentry-curses.o ../pinentry/libpinentry.a ../pinentry/libpinentry-curses.a ../assuan/libassuan.a ../secmem/libsecmem.a -lcap -lncursesw make[2]: Leaving directory `/var/tmp/portage/app-crypt/pinentry-0.8.2/work/pinentry-0.8.2/curses' (In reply to comment #1) > Hi, > > I cannot reproduce... any hint? > Perhaps, it is a gold triggered one. Going blindly... Can you please test these patches? Created attachment 347402 [details, diff]
pinentry-0.8.2.ebuild.diff
Created attachment 347404 [details, diff]
pinentry-0.8.2-ncurses.patch
Created attachment 347410 [details, diff]
pinentry-0.8.2-ncurses.patch
I would sugest to use this kind of patch. You simply use pkg-config to detect what is right.
(In reply to comment #6) > Created attachment 347410 [details, diff] [details, diff] > pinentry-0.8.2-ncurses.patch > > I would sugest to use this kind of patch. You simply use pkg-config to > detect what is right. As I know upstream, the chance that pkg-config will be merge is low. So I prefer something that might be accepted. (In reply to comment #7) > As I know upstream, the chance that pkg-config will be merge is low. So I > prefer something that might be accepted. I think it should be fine with upstream. The configure script is using pkg-config at several places $ grep PKG configure.ac AC_PATH_PROG(PKG_CONFIG, pkg-config, no) if test x"${PKG_CONFIG}" = xno ; then "${PKG_CONFIG}" --exists gtk+-2.0 modvers=`"${PKG_CONFIG}" --modversion gtk+-2.0` "${PKG_CONFIG}" --atleast-version=2.4.0 gtk+-2.0 GTK2CFLAGS=`"${PKG_CONFIG}" --cflags gtk+-2.0` GTK2LIBS=`"${PKG_CONFIG}" --libs gtk+-2.0` PKG_CHECK_MODULES(QT4_CORE, QtCore,, PKG_CHECK_MODULES(QT4_GUI, QtGui,, there are only two solutions, ncurses-config or pkg-config. And if upstream doesn't like it, then we need to maintain it here. (In reply to comment #8) > there are only two solutions, ncurses-config or pkg-config. And if upstream > doesn't like it, then we need to maintain it here. 'We' is I and I do not wish to maintain anything outside of upstream domain. I will present these two (or three) options in upstream bug, but will provide the simplest most likely to be accepted for now. Please check the variant I posted, and report if it works for you. Thanks. (In reply to comment #9) > 'We' is I and I do not wish to maintain anything outside of upstream domain. "we" should mean gentoo developers who have the intention to support a working ebuild for users. > Please check the variant I posted, and report if it works for you. That doesn't work. the problem is that some symbols are moved from libncurses(w).so to libtinfo(w).so. Therefore when ever you link against libncurses you most probably also need to link with libtinfo.so. (In reply to comment #10) > (In reply to comment #9) > > 'We' is I and I do not wish to maintain anything outside of upstream domain. > > "we" should mean gentoo developers who have the intention to support a > working ebuild for users. Do you want to maintain this package? No problem... really. > > > Please check the variant I posted, and report if it works for you. > > That doesn't work. the problem is that some symbols are moved from > libncurses(w).so to libtinfo(w).so. Therefore when ever you link against > libncurses you most probably also need to link with libtinfo.so. This is what I have done in attachment#347404 [details, diff], please tell me where it fails. Thanks. (In reply to comment #11) > This is what I have done in attachment#347404 [details, diff] [details, diff], please tell > me where it fails. Strange, now it works. I can swear that my first test was negativ. SO lets go with your patch, it works. x86_64-pc-linux-gnu-gcc -O2 -pipe -ftracer -march=native -frecord-gcc-switches -g -Wimplicit-function-declaration -Wall -Wcast-align -Wshadow -Wstrict-prototypes -Wno-pointer-sign -Wl,-O1 -Wl,--as-needed -Wl,-O1 -Wl,--hash-style=gnu -Wl,--sort-common -o pinentry-gtk-2 pinentry-gtk-2.o gtksecentry.o ../pinentry/libpinentry.a ../assuan/libassuan.a ../secmem/libsecmem.a -lcap -lgtk-x11-2.0 -lgdk-x11-2.0 -lpangocairo-1.0 -latk-1.0 -lcairo -lgdk_pixbuf-2.0 -lgio-2.0 -lpangoft2-1.0 -lpango-1.0 -lgobject-2.0 -lglib-2.0 -lfreetype -lfontconfig ../pinentry/libpinentry-curses.a -lncursesw -ltinfow Thanks, fixed. Created attachment 350258 [details]
build log for my x86_64 unit
Created attachment 350260 [details]
automake.out file in addition to my build-log
Geoff Madden, Please do not post anything in resolved bugs, unless you are sure that the issue is the exact one. Your issue is pinentry and newer gettext compatibility BTW. Thanks, |