openjdk team mailing list archive
-
openjdk team
-
Mailing list archive
-
Message #10730
Bug#775760: openjdk-8 rejected due to wrong lintian warning
Hi Matthias! Thanks for the bug report.
* Matthias Klose <doko@xxxxxxxxxx>, 2015-01-19, 19:08:
library-in-debug-or-profile-should-not-be-stripped usr/lib/debug/.build-id/06/06a3353ef3b92e2624bdf24b2a1aa5216a7975dd.debug
This is definitely a bug in Lintian.
Lintian uses file(1) to check whether a binary is stripped or not, and
it says something like this:
ELF 32-bit LSB shared object, Intel 80386, version 1 (GNU/Linux), dynamically linked, BuildID[sha1]=2a7ceaa9fb01ea2953251d707d9251bc3cfd03da, not stripped, too many notes (256)
And this is what Lintian looks for in checks/binaries.pm:
if ($fileinfo =~ m,not stripped\s*$,o) {
The regexp will have to be adjusted to allow for a comma after
"stripped". This should do the trick:
if ($fileinfo =~ m,\bnot stripped\b,o) {
That said, at least for the package in the archive, openjdk-8-dbg
triggers also debug-file-with-no-debug-symbols, which doesn't rely on
file(1) output, so something else might be broken too.
--
Jakub Wilk
References