./obj/../soobj/echogs -e .dev -w- -l-dev2 -b -s -l-include -l./obj/../soobj/page -l-obj ./obj/../soobj/pksmraw ./obj/../soobj/gdevpbm.o ./obj/../soobj/gdevppla.o ./obj/../soobj/gdevmpla.o ./obj/../soobj/echogs -e .dev -w- -l-dev2 -b -s -l-include -l./obj/../soobj/page -l-obj ./obj/../soobj/plan9bm ./obj/../soobj/gdevpbm.o ./obj/../soobj/gdevppla.o ./obj/../soobj/gdevmpla.o make[1]: *** No rule to make target `src/png.c', needed by `obj/../soobj/png.o'. Stop. make[1]: Leaving directory `/var/tmp/portage/app-text/ghostscript-gpl-8.70-r1/work/ghostscript-8.70' make: *** [so] Error 2 Reproducible: Always Steps to Reproduce:
Created attachment 219677 [details, diff] ghostscript-gpl-8.70-r1.ebuild.diff quick'n'dirty fix
Adding URL to the upstream bug.
Since it's been reported upstream, I've worked this around in ebuild. + 08 Mar 2010; Samuli Suominen <ssuominen@gentoo.org> + ghostscript-gpl-8.71-r1.ebuild: + Fix libpng14 detection wrt #305111 by Lars Wendler.
Reopening as the problem wasn't yet acknowledged and/or fixed upstream.
there's 3 files to patch configure.ac base/configure.ac jbig2dec/configure.ac so attached fix is here not enough fixed properly in portage
(In reply to comment #5) > there's 3 files to patch Thanks Samuli, I've updated the upstream patch accordingly.
doesn't block bug 305095 anymore. thanks for taking it to upstream
*** Bug 319269 has been marked as a duplicate of this bug. ***
(In reply to comment #8) > *** Bug 319269 has been marked as a duplicate of this bug. *** > Not a duplicate. app-text/ghostscript-gpl-8.64-r3 fails with the same error. ./obj/../soobj/echogs -e .dev -w- -l-dev2 -b -s -l-include -l./obj/../soobj/page -l-obj ./obj/../soobj/plan9bm ./obj/../soobj/gdevpbm.o ./obj/../soobj/gdevppla.o ./obj/../soobj/gdevmpla.o make[1]: *** Keine Regel vorhanden, um das Target »src/png.c«, benötigt von »obj/../soobj/png.o«, zu erstellen. Schluss. make[1]: Leaving directory `/home/tmp/portage/app-text/ghostscript-gpl-8.64-r3/work/ghostscript-8.64' make: *** [so] Fehler 2 * ERROR: app-text/ghostscript-gpl-8.64-r3 failed: * emake failed * * Call stack: * ebuild.sh, line 54: Called src_compile * environment, line 3464: Called die * The specific snippet of code: * emake -j1 so all || die "emake failed"; * * If you need support, post the output of 'emerge --info =app-text/ghostscript-gpl-8.64-r3', * the complete build log and the output of 'emerge -pqv =app-text/ghostscript-gpl-8.64-r3'. * The complete build log is located at '/home/tmp/portage/app-text/ghostscript-gpl-8.64-r3/temp/build.log'. * The ebuild environment file is located at '/home/tmp/portage/app-text/ghostscript-gpl-8.64-r3/temp/environment'. * S: '/home/tmp/portage/app-text/ghostscript-gpl-8.64-r3/work/ghostscript-8.64' sandbox:main signal SIGQUIT already had a handler ...
The same patch applies also cleanly to app-text/ghostscript-gpl-8.64-r3, and I'm not seeing any obvious problems with the resulting binary. http://git.overlays.gentoo.org/gitweb/?p=user/dilfridge.git;a=tree;f=app-text/ghostscript-gpl;hb=HEAD Please consider also applying the patch to this current stable version of ghostscript-gpl, at least before libpng-1.4 becomes stable.
Upstream comment regarding the patch: --- Comment #4 from Hin-Tak Leung <hintak@ghostscript.com> 2010-06-28 03:16:48 UTC --- (In reply to comment #2) > Created an attachment (id=6041) --> (http://bugs.ghostscript.com/attachment.cgi?id=6041) [details] > ghostscript-gpl-8.71-libpng-1.4.0-compile-fix-v2.patch Sorry, the patch is wrong - in 1.4.0, png_check_sig() is removed and replaced by png_sig_cmp() so applying it would break building against linpng 1.2.x . It would be far better to modify configure to check for a function which is common to both 1.2.x and 1.4.x, rather than switching between the two. e.g. check if png_struct is defined or png_create_write_struct() is defined. (these are actually used by the png device so rather more appropriate than either png_check_sig() or png_sig_cmp()).
The bug is fixed in upstream, see: http://bugs.ghostscript.com/show_bug.cgi?id=691111#c4 The upstream guy is wrong though, the patch we apply now (the sed in ebuild) doesn't break building with libpng 1.2.x, the new function was already included in 1.2.x... so there is nothing wrong with the way we fix it now but whatever, both ways work, and we get the upstream version of the fix with next release
Please fix app-text/ghostscript-gpl-8.64-r3 so stable ARM can build.
