(In reply to Henri Sivonen (:hsivonen) from comment #27) > How arbitrary is the depth 1000 here? Does it come from RoboHornet? How much > smaller can we make it for this test still to serve its purpose as > reflecting content that Firefox needs to deal with? Note, the crashtest here occasionally triggers stack-overflows in automation: see bug 1825414 (and dupes bug 1866878 and bug 1868113 that I just added for recent instances). In [one log that I looked at](https://firefoxci.taskcluster-artifacts.net/djUUx2-5RtyJ-_DLLooi8g/0/public/logs/live_backing.log), the backtrace has a backtrace that's 3672 levels deep. Not sure if that's surprising or not, but we might want to consider reducing the 1000 depth if that's expected to be directly-proportional to the stacktrace depth.
Bug 1383981 Comment 30 Edit History
Note: The actual edited comment in the bug view page will always show the original commenter’s name and original timestamp.
(In reply to Henri Sivonen (:hsivonen) from comment #27) > How arbitrary is the depth 1000 here? Does it come from RoboHornet? How much > smaller can we make it for this test still to serve its purpose as > reflecting content that Firefox needs to deal with? Note, the crashtest here occasionally triggers stack-overflows in automation: see bug 1825414 (and dupes bug 1866878 and bug 1868113 that I just added for recent instances). In [one log that I looked at](https://firefoxci.taskcluster-artifacts.net/djUUx2-5RtyJ-_DLLooi8g/0/public/logs/live_backing.log), the backtrace has a backtrace that's 3672 levels deep. Not sure if that's surprising or not, but we might want to consider reducing the 1000 depth if that's expected to be directly-proportional to the number of callstack levels.
(In reply to Henri Sivonen (:hsivonen) from comment #27) > How arbitrary is the depth 1000 here? Does it come from RoboHornet? How much > smaller can we make it for this test still to serve its purpose as > reflecting content that Firefox needs to deal with? Note, the crashtest here occasionally triggers stack-overflows in automation: see bug 1825414 (and dupes bug 1866878 and bug 1868113 that I just triaged&duped for recent instances of the same crash). In [one log that I looked at](https://firefoxci.taskcluster-artifacts.net/djUUx2-5RtyJ-_DLLooi8g/0/public/logs/live_backing.log), the backtrace has a backtrace that's 3672 levels deep. Not sure if that's surprising or not, but we might want to consider reducing the 1000 depth if that's expected to be directly-proportional to the number of callstack levels.
(In reply to Henri Sivonen (:hsivonen) from comment #27) > How arbitrary is the depth 1000 here? Does it come from RoboHornet? How much > smaller can we make it for this test still to serve its purpose as > reflecting content that Firefox needs to deal with? Note, the crashtest here occasionally triggers stack-overflows in automation: see bug 1825414 (and dupes bug 1866878 and bug 1868113 that I just triaged&duped for recent instances of the same crash). In [one log that I looked at](https://firefoxci.taskcluster-artifacts.net/djUUx2-5RtyJ-_DLLooi8g/0/public/logs/live_backing.log), we've got a backtrace that's 3672 levels deep when we crash. Not sure if that's surprising or not, but we might want to consider reducing the 1000 depth if that's expected to be directly-proportional to the number of callstack levels.