Hi Sergio, It looks like a problem on the GC (or so I can tell by just looking at the stack). I'd suggest you to first try exporting your data using some serialization mechanism (STON or JSON for a textual representation, or fuel for a binary one), to avoid losing any of it. What version are you on? Is that a unix or osx? Your vm seems from last june. Date: Thu Jun 28 14:56:30 2018 CommitHash: a8a1dc1 I put in CC the VM list, maybe someone there is aware of a bug fixed between june and latest versions. Cheers, Guille On Mon, Jan 21, 2019 at 2:59 AM sergio ruiz <[hidden email]> wrote:
|
Hi Guille, can you guide Sergio through the process of applying the assert vm and the leak checker to his image to find out where the corruption is?
_,,,^..^,,,_ (phone)
|
Hi Sergio, I'm forwarding Eliot's answer ^^. He suggests you compile a VM with assert flags to be able to check the output. Maybe first you could try with the latest VM for pharo7 and see if the same problem still happens? Then, if you want to compile the VM in assert I can give you a hand, It should not be too difficult (probably the only boring/time-consuming thing is installing XCode if you don't have it). Usually it should be as easy as going to the build.macos32x86/pharo.cog.spur/ directory and execute ./mvm -a On Mon, Jan 21, 2019 at 4:07 PM Eliot Miranda <[hidden email]> wrote:
Eliot, I assume from your answer that no big changes were done on the GC since june? Also, I've never seen/used the leak checker before. I remember having listened to somebody naming it, but that's all. Do you have some pointer? Now I've found only some old emails from the archive. Maybe the following is enough? @Sergio, could you try running the vm with this additional argument?
|
Free forum by Nabble | Edit this page |