I should have looked at the attachment… What you’re seeing is a problem with UFFI, Esteban is trying to hunt down the cause. For some reason the font loading mechanism uses an FFI call that doesn’t always do what it’s supposed to. Usually this will lead to a playground or browser displaying a red “square for death” after boot but I’ve also seen segfaults myself.
This (very likely) has nothing to do with image size, as I suggested earlier. Unfortunately, I don’t think there is a fix at the moment other than trying save a new copy of the image. Cheers, Max > On 15 Mar 2016, at 01:54, Ben Coman <[email protected]> wrote: > > On Tue, Mar 15, 2016 at 1:25 AM, Jan Kurš <[email protected]> wrote: >> Hi, >> >> I am trying to open an image I saved a few days ago and I am getting this >> message (the whole log is attached): >> >> stack page bytes 4096 available headroom 2788 minimum unused headroom 3020 >> >> (Segmentation fault) >> >> Anyone experiencing something similar? I am running the latest pharo5 VM on >> Linux. >> >> Cheers, >> Jan > > Something similar happened to me for the first time last night, about > three times in quick succession. Seemed to be when I started an Image > from a shell and later killed it with control-C. Incidentally, these > was with a recent Tracker image only an hour old for me, which had > locked while I was doing risky changes to DelayScheduler, but I was > surprised that it affected the Image on disk. I was doing temporary > checkpoint saves just prior to Accepting risky code changes. Then upon > restart I got the error. But I could not reliably reproduce the > problem. > > Certainly it was not a large Image. > > VM: http://files.pharo.org/vm/pharo-spur32/linux/463.zip > Image: > https://ci.inria.fr/pharo/view/5.0-Analysis/job/Pharo-5.0-Issue-Tracker-Image/ > Platform: Debian 8 Jessie 32-bit > > cheers -ben >
