Description of problem: Clicked on the "lock icon", at right of the "URL bar" (after clicking to edit the URL). I can not reproduce, however. Version-Release number of selected component: epiphany-runtime-3.16.2-2.fc22 Additional info: reporter: libreport-2.6.0 backtrace_rating: 4 cmdline: epiphany crash_function: ephy_certificate_popover_set_address executable: /usr/bin/epiphany global_pid: 3414 kernel: 4.0.6-300.fc22.x86_64 runlevel: N 5 type: CCpp uid: 1000 Truncated backtrace: Thread no. 1 (10 frames) #0 ephy_certificate_popover_set_address at ephy-certificate-popover.c:69 #1 ephy_certificate_popover_set_property at ephy-certificate-popover.c:177 #2 object_set_property at gobject.c:1415 #3 g_object_new_internal at gobject.c:1808 #4 g_object_new_valist at gobject.c:2033 #6 ephy_certificate_popover_new at ephy-certificate-popover.c:353 #7 open_certificate_popover at ephy-window.c:3168 #8 location_controller_lock_clicked_cb at ephy-window.c:3191 #17 icon_button_press_event_cb at ephy-location-entry.c:884 #22 gtk_entry_event at gtkentry.c:4310
Created attachment 1049572 [details] File: backtrace
Created attachment 1049573 [details] File: cgroup
Created attachment 1049574 [details] File: core_backtrace
Created attachment 1049575 [details] File: dso_list
Created attachment 1049576 [details] File: environ
Created attachment 1049577 [details] File: limits
Created attachment 1049578 [details] File: maps
Created attachment 1049579 [details] File: mountinfo
Created attachment 1049580 [details] File: namespaces
Created attachment 1049581 [details] File: open_fds
Created attachment 1049582 [details] File: proc_pid_status
Created attachment 1049583 [details] File: var_log_messages
Thanks for reporting all these bugs! Program terminated with signal SIGSEGV, Segmentation fault. #0 ephy_certificate_popover_set_address (address=0x273e5d0 "", popover=0x30f82c0) at ephy-certificate-popover.c:69 69 uri_text = g_markup_printf_escaped ("<span weight=\"bold\">%s</span>.", uri->host); Hm, this is a problem with my downstream patchset to display UTF-8 URLs properly. But I am not sure if this code is unsafe, or if the bug is in g_markup_printf_escaped. I would expect the bug to be reproducible on whatever site you were visiting at the time, which presumably has UTF-8 characters in its hostname?
(In reply to Michael Catanzaro from comment #13) > Thanks for reporting all these bugs! C'mon Michael! I'm just the one who "complains" and writes bad English, you are the one who fixes things! > I would expect the bug to be reproducible on whatever site you were visiting > at the time, which presumably has UTF-8 characters in its hostname? I don't remember which site was, sorry :/ Some idea of a UTF-8 hostname to test this? Or how I can "force" UTF-8 in a hostname? (it *needs* to be in the hostname, btw? Cannot be in another part of the URL? And which characters are considered UTF-8 to this? Everything not in ASCII/English?). As an alternative, I will just keep clicking the thing on every website until it crashes again...
(In reply to Diogo Campos from comment #14) > Some idea of a UTF-8 hostname to test this? It seems to be quite hard to find sites with UTF-8 hostnames using internet search engines. :( My recommendation is to look in your browsing history at the sites you viewed yesterday, and look for special characters in the hostname. I just assumed that a UTF-8 hostname is to blame for this; if you don't have any such sites in your browsing history, then I must be wrong. That would be helpful to determine. > Or how I can "force" UTF-8 in a > hostname? (it *needs* to be in the hostname, btw? Cannot be in another part > of the URL? And which characters are considered UTF-8 to this? Everything > not in ASCII/English?). Yeah, if the UTF-8 is in another part of the URL, I don't think it would matter. (I tested on Japanese Wikipedia and it works fine.) Anything that doesn't look like an English character is going to be UTF-8, yes.
Tried to reproduce in all (https) websites visited yesterday. No luck. Also, no UTF-8 character spotted in the hostnames.
*** Bug 1342608 has been marked as a duplicate of this bug. ***