April 2020 - openSUSE Bugs - openSUSE Mailing Lists

[Bug 1163333] New: LTO: binutils build fails
by bugzilla_noreply＠novell.com 13 Apr '22

13 Apr '22

http://bugzilla.opensuse.org/show_bug.cgi?id=1163333 Bug ID: 1163333 Summary: LTO: binutils build fails Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Basesystem Assignee: bnc-team-screening(a)forge.provo.novell.com Reporter: martin.liska(a)suse.com QA Contact: qa-bugs(a)suse.de Found By: --- Blocker: --- I'm working on binutils 2.34 package update and I noticed quite suspicious thing. We build the package with LTO (and -ffat-lto-objects due to some .a libs), and the built package crashes. More precisely, one needs to built binutils with already built version 2.34. Doing that I end up with: $ LD_LIBRARY_PATH=../bfd/.libs/ .libs/ar r /tmp/x.a /tmp/y Segmentation fault (core dumped) (gdb) bt #0 0x0000000000000000 in ?? () #1 0x00007ffff7afe7dd in make_relative_prefix_1 (progname=0x7fffffffeb09 "/home/abuild/rpmbuild/BUILD/binutils-2.34/build-dir/binutils/.libs/ar", bin_prefix=0x7ffff7ce326b "/usr/bin", prefix=0x7ffff7ce3254 "/usr/lib64/bfd-plugins", resolve_links=1) at ../../libiberty/make-relative-prefix.c:317 #2 0x00007ffff7b1026e in load_plugin (abfd=0x5555555667e0) at ../../bfd/plugin.c:397 #3 bfd_plugin_object_p (abfd=0x5555555667e0) at ../../bfd/plugin.c:451 #4 0x00007ffff7cda4c6 in bfd_check_format_matches (abfd=abfd@entry=0x5555555667e0, format=format@entry=bfd_object, matching=matching@entry=0x0) at ../../bfd/format.c:261 #5 0x00007ffff7cda917 in bfd_check_format (abfd=abfd@entry=0x5555555667e0, format=format@entry=bfd_object) at ../../bfd/format.c:94 #6 0x00007ffff7cdb248 in _bfd_write_archive_contents (arch=0x555555573a90) at ../../bfd/archive.c:2127 #7 0x00007ffff7cd2c4a in bfd_close (abfd=0x555555573a90) at ../../bfd/opncls.c:755 #8 0x000055555555a718 in write_archive (iarch=0x5555555640c0) at ../../binutils/ar.c:1240 #9 0x000055555555a9ea in replace_members (arch=0x5555555640c0, files_to_move=0x555555564080, quick=0) at ../../binutils/ar.c:1482 #10 0x000055555555842e in main (argc=<optimized out>, argv=<optimized out>) at ../../binutils/ar.c:887 Dump of assembler code for function make_relative_prefix_1: 0x00007ffff7afe620 <+0>: push %rbp 0x00007ffff7afe621 <+1>: mov %rsp,%rbp 0x00007ffff7afe624 <+4>: push %r15 0x00007ffff7afe626 <+6>: push %r14 ... 0x00007ffff7afe7d8 <+440>: callq 0x7ffff7af6990 <lrealpath@plt> => 0x00007ffff7afe7dd <+445>: mov %rax,%r15 So it crashes in call to lrealpath (via plt entry). What's maybe problematic is that the link command line: gcc -W -Wall -Wstrict-prototypes -Wmissing-prototypes -Wshadow -Wstack-usage=262144 -O2 -Wall -D_FORTIFY_SOURCE=2 -fstack-protector-strong -funwind-tables -fasynchronous-unwind-tables -fstack-clash-protection -Werror=return-type -flto=auto -g -Wno-error -ffat-lto-objects -o .libs/ar arparse.o arlex.o ar.o not-ranlib.o arsup.o rename.o binemul.o emul_vanilla.o bucomm.o version.o filemode.o ../bfd/.libs/libbfd.so -L/home/abuild/rpmbuild/BUILD/binutils-2.34/build-dir/bfd/../libiberty/pic -liberty -lz ../libiberty/libiberty.a -lfl -ldl -Wl,-rpath -Wl,/usr/lib64 contains: $ nm ../bfd/.libs/libbfd.so | grep lrealpath 00000000000fc510 T lrealpath and LTO resolution file looks like: /home/abuild/rpmbuild/BUILD/binutils-2.34/build-dir/bfd/../libiberty/pic/libiberty.a@0x20eab2 7 ... 269 5c0e5e41e2575a62 RESOLVED_IR lrealpath /home/abuild/rpmbuild/BUILD/binutils-2.34/build-dir/bfd/../libiberty/pic/libiberty.a@0x26afce 2 212 69eadc5b8b30099f PREVAILING_DEF_IRONLY lrealpath ... and the symbol is seen by GCC in LTO as: lrealpath/2736 (lrealpath) @0x7ffff77639d8 Type: function definition analyzed Visibility: externally_visible prevailing_def_ironly public References: Referring: Read from file: /home/abuild/rpmbuild/BUILD/binutils-2.34/build-dir/bfd/../libiberty/pic/libiberty.a Availability: available Function flags: count:1073741824 (estimated locally) Called by: canonical_filename_eq/2611 (1073741824 (estimated locally),1.00 per call) canonical_filename_eq/2611 (1073741824 (estimated locally),1.00 per call) Calls: strdup/2737 (1073741824 (estimated locally),1.00 per call) __realpath_chk/2740 (1073741824 (estimated locally),1.00 per call) But I can't see the symbol being emitted in final executable (nor .s file): $ nm .libs/ar | grep lrealpath 0000000000047e87 W lrealpath.c.60931cf1 Note that adding -fno-lto to the command line helps and the ar is not crashing. The issue seems to me similar to what I've already seen few time for e.g. Inkscape: https://gitlab.com/inkscape/inkscape/issues/235 So it's basically about linking against x.a file and y.so (where y.so contains x.a). -- You are receiving this mail because: You are on the CC list for the bug.

2 6

[Bug 1146017] New: UEFI "Boot Linux System" from installation media screen 2 consistently fails quickly
by bugzilla_noreply＠novell.com 11 Apr '22

11 Apr '22

http://bugzilla.opensuse.org/show_bug.cgi?id=1146017 Bug ID: 1146017 Summary: UEFI "Boot Linux System" from installation media screen 2 consistently fails quickly Classification: openSUSE Product: openSUSE Distribution Version: Leap 15.2 Hardware: Other URL: https://forums.opensuse.org/showthread.php/537185-UEFI -installation-media-gt-Boot-Linux-System?p=2911641#pos t2911641 OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Installation Assignee: yast2-maintainers(a)suse.de Reporter: mrmazda(a)earthlink.net QA Contact: jsrain(a)suse.com Found By: --- Blocker: --- Created attachment 814369 --> http://bugzilla.opensuse.org/attachment.cgi?id=814369&action=edit y2logs .tgz from fresh installation of 15.2a build 481 on Asus Kaby Lake on which "Boot Linux System" last attempted Initial Summary: UEFI "Boot Linux System" from installation media screen 2 consistently fails quickly To reproduce: 1-boot installation media in UEFI mode 2-choose "More ..." 3-choose "Boot Linux System" 4-(optional) append/edit linuxefi line, e.g. add video=1440x900 and/or 3 and/or noresume or remove splash=silent 5-answer the questions asked/make appropriate selections about kernel, options and root filesystem Actual results: 1-instant message in yellow text on red background: "Sorry, system didn't boot" 2-on vtty4: kexec_file: kernel signature verification failed (-129). 3-sh: dmesg: command not found on tty9 4-tty[2,5,6] shell absent 5-sh: systemctl: command not found on tty9 Expected results: 1-installed system boots, or 2-something more than instant failure report Note: forum thread comments report failure also happens with a VM, and with 15.1, and with either OM or USB boot media. Other info from installation on which last attempt made (using NET OM): # efibootmgr BootCurrent: 0000 Timeout: 1 seconds BootOrder: 0000,0007,0006,0003 Boot0000* opensuse Boot0003* CD/DVD Drive Boot0006* Hard Drive Boot0007* UEFI: Optiarc DVD RW AD-7200S # parted -l highlights Disk /dev/sda: 40.0GB Sector size (logical/physical): 512B/512B Partition Table: gpt Number Start End Size File system Name Flags 1 1049kB 337MB 336MB fat16 P01 EFI System (ESP) boot, esp 2 337MB 1202MB 865MB linux-swap(v1) P02 swapper swap 3 1202MB 9590MB 8389MB ext4 P03 linuxroot1 4 9590MB 40.0GB 30.4GB ext4 P04 linuxhome -- You are receiving this mail because: You are on the CC list for the bug.

2 5

[Bug 1134289] New: ghostscript should not have a hard dependency on AppArmor
by bugzilla_noreply＠novell.com 08 Apr '22

08 Apr '22

http://bugzilla.suse.com/show_bug.cgi?id=1134289 Bug ID: 1134289 Summary: ghostscript should not have a hard dependency on AppArmor Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: AppArmor Assignee: suse-beta(a)cboltz.de Reporter: max(a)suse.com QA Contact: qa-bugs(a)suse.de Found By: --- Blocker: --- As a matter of personal preference I ususally set a "don't install" lock on AppArmor in my installations. But with sr#687694 a hard runtime dependency on apparmor-abstractions was added to the ghostscript package, which in turn pulls other AppArmor packages and makes it impossible to have an AppArmor-free system with ghostscript. As AppArmor is optional, I don't think packages like ghostscript should have such a hard dependency on it. Instead it should either be a weak dependency or no dependency at all. If a hard dependency is needed for some reason the AppArmor stuff for GS should be in an optional subpackage to retain the possibility to install GS without AppArmor. -- You are receiving this mail because: You are on the CC list for the bug.

2 7

[Bug 1162121] New: Installcheck issues with xml-commons-resolver10
by bugzilla_noreply＠novell.com 07 Apr '22

07 Apr '22

http://bugzilla.suse.com/show_bug.cgi?id=1162121 Bug ID: 1162121 Summary: Installcheck issues with xml-commons-resolver10 Classification: openSUSE Product: openSUSE Distribution Version: Leap 15.2 Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Other Assignee: bnc-team-screening(a)forge.provo.novell.com Reporter: lubos.kocman(a)suse.com QA Contact: qa-bugs(a)suse.de Found By: --- Blocker: --- Hello current xml-commons-resolver submission seems to run into following issues https://build.opensuse.org/request/show/762685 can't install xml-commons-resolver10-1.4.01-lp152.8.35.noarch: package xerces-j2-2.12.0.noarch requires osgi(org.apache.xml.resolver), but none of the providers can be installed package xml-commons-resolver10-1.4.01.noarch requires jaxp_parser_impl, but none of the providers can be installed package xml-commons-resolver-1.2.noarch obsoletes xml-commons-resolver10 provided by xml-commons-resolver10-1.4.01.noarch can't install xml-commons-resolver11-1.4.01-lp152.8.35.noarch: package xerces-j2-2.12.0.noarch requires osgi(org.apache.xml.resolver), but none of the providers can be installed package xml-commons-resolver11-1.4.01.noarch requires jaxp_parser_impl, but none of the providers can be installed package xml-commons-resolver11-1.4.01.noarch obsoletes xml-commons-resolver <= 1.3.02 provided by xml-commons-resolver-1.2.noarch package xml-commons-resolver-1.2.noarch obsoletes xml-commons-resolver11 provided by xml-commons-resolver11-1.4.01.noarch can't install xml-commons-resolver12-1.4.01-lp152.8.35.noarch: package xerces-j2-2.12.0.noarch requires osgi(org.apache.xml.resolver), but none of the providers can be installed package xml-commons-resolver12-1.4.01.noarch requires jaxp_parser_impl, but none of the providers can be installed package xml-commons-resolver-1.2.noarch obsoletes xml-commons-resolver12 provided by xml-commons-resolver12-1.4.01.noarch -- You are receiving this mail because: You are on the CC list for the bug.

2 4

[Bug 1115942] New: Fonts in all applications are green and magenta or have a cast of that
by bugzilla_noreply＠novell.com 07 Apr '22

07 Apr '22

http://bugzilla.suse.com/show_bug.cgi?id=1115942 Bug ID: 1115942 Summary: Fonts in all applications are green and magenta or have a cast of that Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Basesystem Assignee: bnc-team-screening(a)forge.provo.novell.com Reporter: asn(a)cryptomilk.org QA Contact: qa-bugs(a)suse.de Found By: --- Blocker: --- I've updated to Tumbleweed yesterday morning (2018-11-13) and then all fonts started to have a green an magenta case or are mostly green. See https://xor.cryptomilk.org/pics/fonts.png -- You are receiving this mail because: You are on the CC list for the bug.

2 5

[Bug 1101663] New: The %_jvmjardir macro is missing from the javapackages-tools package
by bugzilla_noreply＠novell.com 07 Apr '22

07 Apr '22

http://bugzilla.suse.com/show_bug.cgi?id=1101663 Bug ID: 1101663 Summary: The %_jvmjardir macro is missing from the javapackages-tools package Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Java Assignee: tchvatal(a)suse.com Reporter: mls(a)suse.com QA Contact: qa-bugs(a)suse.de Found By: --- Blocker: --- While cleaning up obsolete macros in rpm I found that all javapackage macros are now in the javapackages-tools package except %_jvmjardir. %_jvmjardir seems to be used in a couple of packages, so it would be good if that macro also is included in javapackages-tools. The current definition is: %_jvmjardir %{_libdir}/jvm-exports -- You are receiving this mail because: You are on the CC list for the bug.

2 2

[Bug 1071230] New: java installs symlinks that are not owned by a package
by bugzilla_noreply＠novell.com 07 Apr '22

07 Apr '22

http://bugzilla.suse.com/show_bug.cgi?id=1071230 Bug ID: 1071230 Summary: java installs symlinks that are not owned by a package Classification: openSUSE Product: openSUSE Distribution Version: Leap 15.0 Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Java Assignee: bnc-team-java(a)forge.provo.novell.com Reporter: lnussel(a)suse.com QA Contact: qa-bugs(a)suse.de Found By: --- Blocker: --- java packages install e.g /usr/bin/keytool managed by update-alternatives but those files are not owned by any package. Since the symlink in /usr does not actually change it should be owned by a package. -- You are receiving this mail because: You are on the CC list for the bug.

2 4

[Bug 994084] New: Tracker bug: make packages own their symlinks in /etc/alternatives/*
by bugzilla_noreply＠novell.com 07 Apr '22

07 Apr '22

http://bugzilla.suse.com/show_bug.cgi?id=994084 Bug ID: 994084 Summary: Tracker bug: make packages own their symlinks in /etc/alternatives/* Classification: openSUSE Product: openSUSE Tumbleweed Version: Current Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Basesystem Assignee: bnc-team-screening(a)forge.provo.novell.com Reporter: vcuadradojuan(a)suse.com QA Contact: qa-bugs(a)suse.de Found By: --- Blocker: --- This is a tracker bug for making all packages using update-alternatives own their symlinks at `/etc/alternatives/*`. Problem: Currently there is a mix of owned and not owned files by packages in `/etc/alternatives/*` (just look at `rpm -qf /etc/alternatives/*`). Currently there seems to be a sizeable chunk of packages not declaring `%ghost %_sysconfdir/alternatives/<binary>`, as specified at [1]. Also, see discussion on opensuse-factory m-l [2]. Once the mass bug filling for the issue is done, I will wait until 1 October for maintainers to action on the bugs, and afterwards I will start fixing each package by hand. Cheers, [1] https://en.opensuse.org/openSUSE:Packaging_Multiple_Version_guidelines [2] https://lists.opensuse.org/opensuse-factory/2016-08/msg00218.html -- You are receiving this mail because: You are on the CC list for the bug.

2 37

[Bug 931978] New: VUL-0: CVE-2014-9721: zeromq: protocol downgrade attack on sockets using the ZMTP v3 protocol
by bugzilla_noreply＠novell.com 07 Apr '22

07 Apr '22

http://bugzilla.suse.com/show_bug.cgi?id=931978 Bug ID: 931978 Summary: VUL-0: CVE-2014-9721: zeromq: protocol downgrade attack on sockets using the ZMTP v3 protocol Classification: openSUSE Product: openSUSE Distribution Version: 13.2 Hardware: Other OS: Other Status: NEW Severity: Major Priority: P5 - None Component: Security Assignee: tchvatal(a)suse.com Reporter: abergmann(a)suse.com QA Contact: qa-bugs(a)suse.de Found By: Security Response Team Blocker: --- rh#1221666 ------------------------------------- It was discovered that zeromq, a lightweight messaging kernel, is susceptible to a protocol downgrade attack on sockets using the ZMTP v3 protocol. This could allow remote attackers to bypass ZMTP v3 security mechanisms by sending ZMTP v2 or earlier headers. CVE request: http://openwall.com/lists/oss-security/2015/05/07/8 Upstream bug report: https://github.com/zeromq/libzmq/issues/1273 Upstream fix: https://github.com/zeromq/zeromq4-x/commit/b6e3e0f601e2c1ec1f3aac880ed6a3fe… ------------------------------------- References: https://bugzilla.redhat.com/show_bug.cgi?id=1221666 http://web.nvd.nist.gov/view/vuln/detail?vulnId=CVE-2014-9721 http://seclists.org/oss-sec/2015/q2/515 http://people.canonical.com/~ubuntu-security/cve/2014/CVE-2014-9721.html -- You are receiving this mail because: You are on the CC list for the bug.

2 12

[Bug 1105305] New: pencil doesn't run on Leap 15
by bugzilla_noreply＠novell.com 06 Apr '22

06 Apr '22

http://bugzilla.opensuse.org/show_bug.cgi?id=1105305 Bug ID: 1105305 Summary: pencil doesn't run on Leap 15 Classification: openSUSE Product: openSUSE Distribution Version: Leap 15.0 Hardware: Other OS: Other Status: NEW Severity: Normal Priority: P5 - None Component: Other Assignee: aj(a)ajaissle.de Reporter: lnussel(a)suse.com QA Contact: qa-bugs(a)suse.de Found By: --- Blocker: --- $ pencil Error: Platform version '60.1.0' is not compatible with minVersion >= 4.0 maxVersion <= 52.0 -- You are receiving this mail because: You are on the CC list for the bug.

2 3