Damn it..... I'm so mad

5 days ago someone made this claim :

Because I was able to crash phoc 2-3 times and trigger many freezes in less than 10 minutes, the first thing going through my mind was: CHALLENGE ACCEPTED !

I spend my yesterdey evening (around 2-3 hours) hunting and crash testing byzantium with phosh 0.7.0 in Librem5 VM, here is what I could trigger :

  • 0 crash
  • 2 small temporary freezes

Because I didn’t wanted to come back empty-handed from my hunt, I decided to be more creative, I was able to trigger 2 minor bugs I could not even reproduce

  • Nautilus showed me a message saying something like ‘gnome extension could not be loaded’, I had to restart it
  • For few minutes, I was in a state were I was not able to close the settings app by swiping to top, nor clicking the cross : the preview of the app was disapearing …theeen… coming back from the top (but it was working for any other apps)

I have another small minor bug in my pocket, maybe related to xwayland, that I need to confirm, I will hunt another night soon !

Did I wrote ‘mad’ ? … I meant ‘Glad’ !

Big :+1: for the stability improvements ! :partying_face:


Good hunt.

1 Like

Click bait title is click bait Seriously though, glad to hear the improvements.

1 Like

Sorry, I’m totally guilty, your honor ! :sweat_smile:


FYI: phoc 0.7.0 has just migrated from staging into amber-phone, so everyone should see it now as an available update - enjoy! :slight_smile:


Damn it…… I’m so mad…
You know that feeling ? when you search something for hours and do not find it, then 2 days later, the thing you search just falls before your feet, you didn’t even ask

well … I crashed it … Now the title fits ! :nerd_face:
The good news is the probabilities it happens to you in an everyday regular use seems to me close to zero

It doesn’t count! :stuck_out_tongue: It works as expected:

mars 27 13:07:08 pureos phoc[3557]: [backend/session/logind.c:76] Failed to take device '/dev/dri/card0': Device or resource busy
mars 27 13:07:08 pureos phoc[3557]: [backend/backend.c:235] failed to start backend 'drm'
mars 27 13:07:08 pureos phoc[3557]: Could not create backend

Looks like sddm is running first and in turn phoc can’t get access to the DRI node because it’s already taken. sudo systemctl disable --now sddm :wink:


Damn it …
You know that feeling ? When you think you have found something, but you didn’t… :sweat_smile:

More seriousely, I did your command now it’s starting again :+1:

But I still find odd that, in that case, the process ends up with SIGTRAP producing a coredump
Shouldn’t this stop the process with a regular exit or end of main ?

Haven’t looked, but I’d guess it’s g_error's doing.

1 Like