On Oct 2, 2012, at 3:37 PM, Jan Rękorajski <[email protected]> wrote:
> On Tue, 02 Oct 2012, Jeffrey Johnson wrote: > >> >> On Oct 2, 2012, at 1:51 PM, Jan Rękorajski <[email protected]> wrote: >> >>> >>> That package is built from the, quite large, qt4.spec so it's hard to >>> spot anything there :( >> >> Yes. I needed some understanding why some but not all packages are affected. >> >>> Maybe %files will give a hint? >>> >> >> Yes. I predict that if you move >> %attr(755,root,root) %{_qtdir}/bin/xmlpatternsvalidator >> to the end of %files, then the problem will disappear. > > Just FYI - moving that to the end of %files did not helped at all. > I still see poisoned digests. > Thanks for checking. This patch SHOULD ensure ascii digest are terminated with a trailing \0 Index: rpmdb/legacy.c =================================================================== RCS file: /v/rpm/cvs/rpm/rpmdb/legacy.c,v retrieving revision 1.44.4.3 diff -p -u -w -r1.44.4.3 legacy.c --- rpmdb/legacy.c 16 Apr 2012 23:43:34 -0000 1.44.4.3 +++ rpmdb/legacy.c 2 Oct 2012 20:17:21 -0000 @@ -282,8 +282,10 @@ int dodigest(int dalgo, const char * fn, exit: if (fsizep) *fsizep = fsize; - if (!rc) + if (!rc) { memcpy(digest, dsum, dlen); + if (asAscii) digest[dlen] = '\0'; + } dsum = _free(dsum); return rc; Is something else needed for the missing symlink end-point entry? 73 de Jeff _______________________________________________ pld-devel-en mailing list [email protected] http://lists.pld-linux.org/mailman/listinfo/pld-devel-en
