[Vm-dev] [Pharo-dev] Random corrupted data when copying from very large byte array

Alistair Grant akgrant0710 at gmail.com
Wed Jan 24 22:35:27 UTC 2018


Hi Stu,

On 24 January 2018 at 22:24, Stuart Cassoff <aa72 at bell.net> wrote:
>
> Yes, all my VMs are built as OpenBSD ports.
> Thanks for the tip on updateSCCSVersions - I suspect I'll get nicer version strings now. :)
> That's great though, now that I understand. The tip-off was from Pharo.
> What do you want me to test on?
> With your instructions I'm now running the test in Squeak-5.1 but I have 6.0alpha, and Pharo 5,6 and 7.

I assume you're compiling with gcc.  It would be great if you could
try to reproduce the problem with gcc 4.8.x and >= 5.4.x.

The other important attributes:

- 32 bit VM
- VM from commit >= Fri Jan 19 13:17:57 2018 -0800

Other than that, whether you use Squeak or Pharo is up to you.

I've found the problem tends to appear more often if the test is run
immediately after starting Squeak / Pharo.

Thanks!
Alistair



> Stu
>
>
>> ---------- Original Message ----------
>> From: Alistair Grant <akgrant0710 at gmail.com>
>> Date: January 24, 2018 at 4:04 PM
>>
>>
>> On 24 January 2018 at 21:58, Stuart Cassoff <aa72 at bell.net> wrote:
>> >
>> > Right now Pharo sez my vm is too old but it turns out that's because I'm getting a 'subscript out of bounds' when I do a 'Smalltalk vm interpreterSourceDate'. *shrug*
>>
>> Is this a vm you compiled yourself?
>>
>> If so, you need to run scripts/updateSCCSVersions before compiling.
>>
>> Cheers,
>> Alistair


More information about the Vm-dev mailing list