Fedora Account System
Red Hat Associate
Red Hat Customer
Description of problem: Pyhon crashed when I run command `dnf upgrade --refresh -v --exclude=mutter` Version-Release number of selected component: python3-3.8.0~rc1-1.fc32 Additional info: reporter: libreport-2.10.1 backtrace_rating: 4 cgroup: 0::/user.slice/user-1000.slice/user/gnome-terminal-server.service cmdline: /usr/bin/python3 /usr/bin/dnf upgrade --refresh -v --exclude=mutter crash_function: __memmove_avx_unaligned executable: /usr/bin/python3.8 journald_cursor: s=af85196e12e64a24a411e7dfc5442bbf;i=2be210;b=c4022ea6add5498ea43c164d8284baec;m=6f61a4dfe0;t=59424982718ed;x=8d42724541835a37 kernel: 5.4.0-0.rc0.git6.1.fc32.x86_64 rootdir: / runlevel: N 5 type: CCpp uid: 0 Truncated backtrace: Thread no. 1 (9 frames) #0 __memmove_avx_unaligned at ../sysdeps/x86_64/multiarch/memmove-vec-unaligned-erms.S:144 #3 _IO_new_file_xsputn at fileops.c:1269 #5 _IO_fwrite at libioP.h:948 #6 log_handler(gchar const*, GLogLevelFlags, gchar const*, gpointer) at /usr/src/debug/libdnf-0.35.5-2.fc32.x86_64/python/hawkey/sack-py.cpp:186 #9 dnf_sack_set_arch(DnfSack*, gchar const*, GError**) at /usr/src/debug/libdnf-0.35.5-2.fc32.x86_64/libdnf/dnf-sack.cpp:824 #10 sack_init(_SackObject*, PyObject*, PyObject*) at /usr/src/debug/libdnf-0.35.5-2.fc32.x86_64/python/hawkey/sack-py.cpp:265 #11 stack_effect at /usr/src/debug/python3-3.8.0 #13 chain_next at /usr/src/debug/python3-3.8.0 #14 _PyIO_str_getstate
Created attachment 1622898 [details] File: backtrace
Created attachment 1622899 [details] File: core_backtrace
Created attachment 1622900 [details] File: cpuinfo
Created attachment 1622901 [details] File: dso_list
Created attachment 1622902 [details] File: environ
Created attachment 1622903 [details] File: exploitable
Created attachment 1622904 [details] File: limits
Created attachment 1622905 [details] File: maps
Created attachment 1622906 [details] File: mountinfo
Created attachment 1622907 [details] File: open_fds
Created attachment 1622908 [details] File: proc_pid_status
Do you already have the latest version of python3-hawkey?
python3-hawkey had version 0.35.5-2.fc32.x86_64 at the time of the crash.
Created attachment 1622929 [details] dnf.log
Thanks. Does this happen every time or just once?
(In reply to Miro Hrončok from comment #15) > Thanks. Does this happen every time or just once? This happened just once, but every time when I see a `Segmentation fault` message while dnf executing I'm afraid that I will have to reinstall the system.
In a previous message, I wanted to say that such crashes, even single ones, are not permissible since they can completely break the system.
I agree that this is serious. Being unable to update to the fixed version fo packages would however be critical. This seems like a bug in libdnf, but in case it is in fact related to Python, we are here to help.
*** Bug 1759204 has been marked as a duplicate of this bug. ***
> I agree that this is serious. Being unable to update to the fixed version fo packages would however be critical. I am not able to reproduce the bug neither on freshly updated Fedora Rawhide. Versions: * dnf-4.2.11-2.fc32.noarch * libsolv-0.7.6-2.fc32.x86_64 * libdnf-0.35.5-2.fc32.x86_64 * python3-3.8.0~rc1-1.fc32.x86_64 * python3-hawkey-0.35.5-2.fc32.x86_64 I tried dnf commands: upgrade --refresh -v --exclude=mutter reinstall -y git vim search mutter using the following shell commands: # Regular dnf usage sudo dnf upgrade --refresh -v --exclude=mutter sudo dnf reinstall -y git vim sudo dnf search mutter # Python in release mode with develoment mode enabled (-X dev) sudo /usr/bin/python3 -X dev /usr/bin/dnf-3 upgrade --refresh -v --exclude=mutter sudo /usr/bin/python3 -X dev /usr/bin/dnf-3 reinstall -y git vim sudo /usr/bin/python3 -X dev /usr/bin/dnf-3 search mutter # Python in debug mode sudo /usr/bin/python3-debug /usr/bin/dnf-3 upgrade --refresh -v --exclude=mutter sudo /usr/bin/python3-debug /usr/bin/dnf-3 reinstall -y git vim sudo /usr/bin/python3-debug /usr/bin/dnf-3 search mutter -- It may be a buffer overflow somewhere in a C library. Try maybe the following command as root? PYTHONMALLOC=malloc valgrind /usr/bin/python3 /usr/bin/dnf-3 upgrade --refresh -v --exclude=mutter
> PYTHONMALLOC=malloc valgrind /usr/bin/python3 /usr/bin/dnf-3 upgrade --refresh -v --exclude=mutter Valgrind detects the following error: ==1249== Conditional jump or move depends on uninitialised value(s) ==1249== at 0x4B94073: PyUnicode_Decode (unicodeobject.c:3337) ==1249== by 0x4B94350: PyUnicode_FromEncodedObject (unicodeobject.c:3220) ==1249== by 0x4BA1894: unicode_new (unicodeobject.c:15048) (...) but this is a false alarm, see https://bugs.python.org/issue38118 It is very unlikely to explain this bug.
This is almost certainly a duplicate of https://bugzilla.redhat.com/show_bug.cgi?id=1758737. We know what the problem is and will fix it soon. *** This bug has been marked as a duplicate of bug 1758737 ***