Description of problem: It happening when I try list content of archive retromod_digital_1_demo_content_v1_install_script.zip Version-Release number of selected component: mc-1:4.8.28-3.fc37 Additional info: reporter: libreport-2.17.6 type: CCpp reason: mc killed by SIGSEGV journald_cursor: s=b4a46d32f9cb494fb977313b87857e14;i=3338d0;b=071ed2dcf12d44d9b4ba8c0ca0c7c07e;m=c39d90807;t=5f16f1711f975;x=7fa52feff52d116b executable: /usr/bin/mc cmdline: /usr/bin/mc -P /var/tmp/mc-mikhail/mc.pwd.39807 cgroup: 0::/user.slice/user-1000.slice/user/app.slice/app-org.gnome.Terminal.slice/vte-spawn-8c7cfe1c-ea5c-4051-94d8-75595ef3b594.scope rootdir: / uid: 1000 kernel: 6.2.0-0.rc2.20230103git69b41ac87e4a.19.fc38.x86_64 package: mc-1:4.8.28-3.fc37 runlevel: N 5 backtrace_rating: 4 crash_function: extfs_find_entry_int comment: It happening when I try list content of archive retromod_digital_1_demo_content_v1_install_script.zip Truncated backtrace: Thread no. 1 (27 frames) #0 extfs_find_entry_int at vfs/extfs/extfs.c:257 #1 extfs_find_entry at vfs/extfs/extfs.c:314 #2 extfs_add_file at vfs/extfs/extfs.c:434 #3 extfs_read_archive at vfs/extfs/extfs.c:683 #4 extfs_open_and_read_archive at vfs/extfs/extfs.c:753 #5 extfs_get_path at vfs/extfs/extfs.c:803 #6 extfs_opendir at vfs/extfs/extfs.c:1151 #7 extfs_chdir at vfs/extfs/extfs.c:1466 #8 mc_chdir at ../lib/vfs/interface.c:678 #9 panel_do_cd_int at filemanager/panel.c:3271 #10 panel_do_cd at filemanager/panel.c:4649 #11 panel_cd at filemanager/panel.c:5053 #12 exec_extension_cd at filemanager/ext.c:411 #13 exec_extension at filemanager/ext.c:446 #14 regex_command_for at filemanager/ext.c:1033 #15 regex_command at filemanager/ext.h:30 #16 do_enter_on_file_entry.isra.0 at filemanager/panel.c:2813 #17 do_enter at filemanager/panel.c:2864 #18 panel_execute_cmd at filemanager/panel.c:3456 #19 send_message at ../lib/widget/../../lib/widget/widget-common.h:254 #20 group_handle_key at ../lib/widget/group.c:508 #21 group_default_callback at ../lib/widget/group.c:640 #22 dlg_key_event at ../lib/widget/dialog.c:251 #23 dlg_process_event at ../lib/widget/dialog.c:566 #24 frontend_dlg_run at ../lib/widget/dialog.c:320 #25 dlg_run at ../lib/widget/dialog.c:600 #26 do_nc at filemanager/filemanager.c:1829
Created attachment 1935707 [details] File: proc_pid_status
Created attachment 1935708 [details] File: maps
Created attachment 1935709 [details] File: limits
Created attachment 1935710 [details] File: environ
Created attachment 1935711 [details] File: open_fds
Created attachment 1935712 [details] File: mountinfo
Created attachment 1935713 [details] File: os_info
Created attachment 1935714 [details] File: cpuinfo
Created attachment 1935715 [details] File: core_backtrace
Created attachment 1935716 [details] File: exploitable
Created attachment 1935717 [details] File: dso_list
Created attachment 1935718 [details] File: backtrace
Created attachment 1935719 [details] archive which caused this crash
Hi Mikhail, does a file in your archive start with ".." - two dots in its name?
Created attachment 1935750 [details] screenshot (In reply to Jindrich Novy from comment #14) > Hi Mikhail, does a file in your archive start with ".." - two dots in its > name? Yes, but not idea how it's possible. It not mine archive, I downloaded it from tracktion.com site. https://cdn.tracktion.com/file/tracktiondownloader/retromod/digital1/content/10/demo/retromod_digital_1_demo_content_v1_install_script.zip?Authorization=3_20230104143111_0859a4d4a662579e1ec353a5_d037d232b5e580141e0b5c6fca1a4ef02b500125_001_20230105023111_0033_dnld
Wrote a simple patch to "fix" this: diff -up mc-4.8.28/src/vfs/extfs/extfs.c.jnovy mc-4.8.28/src/vfs/extfs/extfs.c --- mc-4.8.28/src/vfs/extfs/extfs.c.jnovy 2022-03-27 13:54:06.000000000 +0200 +++ mc-4.8.28/src/vfs/extfs/extfs.c 2023-01-04 15:55:08.081396288 +0100 @@ -254,8 +254,15 @@ extfs_find_entry_int (struct vfs_s_inode *q = '\0'; if (DIR_IS_DOTDOT (p)) - pent = pent->dir->ent; - else + { + if (pent->dir) + { + pent = pent->dir->ent; + } else + { + pent = NULL; + } + } else { GList *pl; But ultimately I suggest to contact upstream - and report it there given it's so easily reproducible: http://midnight-commander.org/report
Just created the upstream report here: https://midnight-commander.org/ticket/4422
Created attachment 1935826 [details] PowerZip screenshot (In reply to Jindrich Novy from comment #16) > Wrote a simple patch to "fix" this: > > diff -up mc-4.8.28/src/vfs/extfs/extfs.c.jnovy > mc-4.8.28/src/vfs/extfs/extfs.c > --- mc-4.8.28/src/vfs/extfs/extfs.c.jnovy 2022-03-27 13:54:06.000000000 +0200 > +++ mc-4.8.28/src/vfs/extfs/extfs.c 2023-01-04 15:55:08.081396288 +0100 > @@ -254,8 +254,15 @@ extfs_find_entry_int (struct vfs_s_inode > *q = '\0'; > > if (DIR_IS_DOTDOT (p)) > - pent = pent->dir->ent; > - else > + { > + if (pent->dir) > + { > + pent = pent->dir->ent; > + } else > + { > + pent = NULL; > + } > + } else > { > GList *pl; After applying this patch I am unable to enter in this archive with mc. So it still looks like an error. If displaying ".." is problematic, you can replace them with the word "__Parent__". This is what some archivators on MacOS do, for example PowerZip.
(In reply to Jindrich Novy from comment #16) > Wrote a simple patch to "fix" this: > > diff -up mc-4.8.28/src/vfs/extfs/extfs.c.jnovy > mc-4.8.28/src/vfs/extfs/extfs.c > --- mc-4.8.28/src/vfs/extfs/extfs.c.jnovy 2022-03-27 13:54:06.000000000 +0200 > +++ mc-4.8.28/src/vfs/extfs/extfs.c 2023-01-04 15:55:08.081396288 +0100 > @@ -254,8 +254,15 @@ extfs_find_entry_int (struct vfs_s_inode > *q = '\0'; > > if (DIR_IS_DOTDOT (p)) > - pent = pent->dir->ent; > - else > + { > + if (pent->dir) > + { > + pent = pent->dir->ent; > + } else > + { > + pent = NULL; > + } > + } else > { > GList *pl; > > But ultimately I suggest to contact upstream - and report it there given > it's so easily reproducible: http://midnight-commander.org/report {leasse submit that patch as PR on https://github.com/MidnightCommander/mc/
(In reply to Tomasz Kłoczko from comment #19) > > But ultimately I suggest to contact upstream - and report it there given > > it's so easily reproducible: http://midnight-commander.org/report > > {leasse submit that patch as PR on https://github.com/MidnightCommander/mc/ Please don't do this, we ask to use our bug tracker instead of PRs on GitHub.
> Please don't do this, we ask to use our bug tracker instead of PRs on GitHub. This is not about tracking bug but about evaluating proposed fix by mc maintainer.
(In reply to Tomasz Kłoczko from comment #21) > > Please don't do this, we ask to use our bug tracker instead of PRs on GitHub. > > This is not about tracking bug but about evaluating proposed fix by mc > maintainer. I *am* the maintainer and I'm kindly asking you to please NOT encourage others send us PRs, but send patches to our tracker instead if necessary instead.
This bug appears to have been reported against 'rawhide' during the Fedora Linux 38 development cycle. Changing version to 38.