Red Hat Bugzilla – Bug 1483942
Emacs crashing while connecting via Windows with X11 forwarding
Last modified: 2018-05-17 10:39:45 EDT
Description of problem: Emacs is crashing over X11 forwarding when connected using Windows. Following is the backtrace : [root@localhost ~]# emacs & [1] 18679 [root@localhost ~]# X protocol error: BadRequest (invalid request code or no such operation) on protocol request 130 When compiled with GTK, Emacs cannot recover from X disconnects. This is a GTK bug: https://bugzilla.gnome.org/show_bug.cgi?id=85715 For details, see etc/PROBLEMS. Backtrace: emacs[0x4f98c3] emacs[0x4deef1] emacs[0x4f9903] emacs[0x4b339b] emacs[0x4b560c] emacs[0x4b566d] /lib64/libX11.so.6(_XError+0x12b)[0x7fccda6f107b] /lib64/libX11.so.6(+0x420d7)[0x7fccda6ee0d7] /lib64/libX11.so.6(+0x42195)[0x7fccda6ee195] /lib64/libX11.so.6(_XReply+0x208)[0x7fccda6ef088] /lib64/libX11.so.6(XQueryPointer+0x8e)[0x7fccda6e50be] ... You have new mail in /var/spool/mail/root [1]+ Aborted (core dumped) emacs Version-Release number of selected component (if applicable): gtk3-3.22.10-4 How reproducible: Always Steps to Reproduce: 1. Connect to RHEL-7.4 via Windows system with X11 forwarding enabled 2. Run $ emacs & 3. Move the mouse cursor in the emacs window/ wait for a few seconds Actual results: Emacs crashes Expected results: Emacs should not crash Additional info: 1) This wasn't reproducible with RHEL-7.3, so it seems to be a regression bug. 2) While crashing, following bug was suggested : https://bugzilla.gnome.org/show_bug.cgi?id=85715
I have discovered the same issue, with the same error results. Is there a workaround for this until the issue is fixed?
I can confirm this bug on RHEL-7.4 and emacs 24.3-20.el7_4. It's affecting many of our developers.
The upstream bug is not relevant - closing displays has not worked for a very long time, it is certainly not a 7.3-74 regression. The description "crashing over X11 forwarding when connected using Windows." is a little unclear: - What is the role of Windows here ? - Is this a Windows X server ? If so, which one ? - Can we get the output of xdpyinfo --ext all, please ? The stacktrace is not detailed enough to say much.
Created attachment 1347519 [details] Sample xdpyinfo output requested by M. Clasen
(In reply to Matthias Clasen from comment #7) > The description "crashing over X11 forwarding when connected using Windows." > is a little unclear: > > - What is the role of Windows here ? Display. > - Is this a Windows X server ? If so, which one ? Yes. VcXsrv.exe (X2Go/Arctica Builds) v1.17.0.0-3. Also, same issue reproduced by jigar using the Xming Windows X server. > - Can we get the output of xdpyinfo --ext all, please ? See attachment.
I am using it in the same manner, but using XMing. No screen shot available.
Emacs crashes with X11 tunneling through XMing when our servers run 3.10.0-693.11.1.el7.x86_64 kernel. After we rolled back to 3.10.0-514.10.2.el7.x86_64, emacs runs fine.
(In reply to Yu Wang from comment #18) > Emacs crashes with X11 tunneling through XMing when our servers run > 3.10.0-693.11.1.el7.x86_64 kernel. After we rolled back to > 3.10.0-514.10.2.el7.x86_64, emacs runs fine. What is gtk and emacs version at your end? not working on my end even after downgrading to 3.10.0-514.10.2.el7.x86_64,
Issue seems confined to xming ; mobaxterm does not result any crash
(In reply to Piyush Bhoot from comment #21) > Issue seems confined to xming ; mobaxterm does not result any crash I was about to post the same finding: "I spoke too soon. Emacs crashed yesterday on servers I downgraded to 3.10.0-514.10..el7.x86_64. So I tried a few other things and found one that works: I replaced Xming with MobaXterm. Emacs has been running for a good two hours."
Yes, it seems to be running well with MobaXterm.
Hi, I'd like to get an update of the ETA on this. Its listed as Critical but was opened back in August. I'm about to be upgrading 1000s of hosts up to RHEL7.4, as early as this weekend and so this could impact our users. The workaround above of using MobaXterm is not an option for us. I've tried an alias : alias emacstest='XLIB_SKIP_ARGB_VISUALS=1 emacs' but thats not constantly helped. So far only a private build with "--with-x-toolkit=gtk2" as per 1349412 seems to of worked for us, but thats not a great supportable solution. Just talking to our redhat engineer in our case 02030623 they've recommended we stay at RHEL7.3, however thats not practical for us now, nor does it help with other vulnerabilities.
Created attachment 1395089 [details] x11: Don't call XInput API for core events Benjamin did this patch, which should hopefully address the issue.
Since the problem described in this bug report should be resolved in a recent advisory, it has been closed with a resolution of ERRATA. For information on the advisory, and where to find the updated files, follow the link below. If the solution does not work for you, open a new bug report. https://access.redhat.com/errata/RHBA-2018:0770