[Cuis-dev] Cuis won't launch on CogSpur Arm64 MacOS when not using mixed arithmetic

Juan Vuletich juan at cuis.st
Tue Nov 15 06:55:09 PST 2022


That's great!

Thanks,
On 11/5/2022 2:53 AM, Alan Dao via Cuis-dev wrote:
> Looks like you were right, I reported it to vm-dev yesterday and Eliot 
> Miranda pushed a fix for it today.
>
> Best,
> Alan
>
> On Thu, Nov 3, 2022 at 6:19 AM Juan Vuletich <juan at cuis.st 
> <mailto:juan at cuis.st>> wrote:
>
>     Hi Alan,
>
>     On 11/3/2022 4:20 AM, Alan Dao via Cuis-dev wrote:
>>     Hey all,
>>
>>     I've recently switched to an Arm64 Macbook. One of the first
>>     things I did on it was launch the latest base image of Cuis
>>     (6.0-5542). To my surprise, the Arm64 CogSpur VM crashes when
>>     launching the base image. However, the Stack VM successfully
>>     launches the base image. I looked at the crash.dmp and saw that
>>     the CogSpur crash was happening in
>>     SystemDictionary>doMixedArithmetic. In the base image, this
>>     method is called with a False value from SystemDictionary>doStartUp.
>>
>>     I used the Stack VM to change the image such that
>>     SystemDictionary>doMixedArithmetic is called with a True value.
>>     After this, I was able to successfully launch the MixedArithmetic
>>     image on my Arm64 Macbook.
>
>     This looks like a VM bug. You can confirm this by using Squeak or
>     Pharo, and doing `Smalltalk doMixedArithmethic: false`. It should
>     also crash. Same with the Cuis image you were able to start by
>     setting it to true. If so, please report the bug in vm-dev:
>     http://lists.squeakfoundation.org/mailman/listinfo/vm-dev
>
>>     Does anyone know why using mixed arithmetic isn't the default? To
>>     my surprise, the base images for Pharo and Squeak were able to
>>     launch without any modifications, so I'm assuming they're by
>>     default using mixed arithmetic.
>
>     The explanation is in the comment at #doMixedArithmetic. Please
>     take a look. It is about letting the image (and not the VM) handle
>     arguably decisions WRT conversions between number types. If you
>     find it cryptic or have further questions, please ask here.
>
>>     I've attached my crash.dmp to this email for those who're curious.
>>
>>     Best,
>>     Alan
>
>     I'm pretty sure you are the first to exercise this corner of the
>     VM. Please report in vm-dev. I'm sure they'll fix it.
>
>     Thanks,
>
>     -- 
>     Juan Vuletich
>     cuis.st  <http://cuis.st>
>     github.com/jvuletich  <http://github.com/jvuletich>
>     researchgate.net/profile/Juan-Vuletich  <http://researchgate.net/profile/Juan-Vuletich>
>     independent.academia.edu/JuanVuletich  <http://independent.academia.edu/JuanVuletich>
>     patents.justia.com/inventor/juan-manuel-vuletich  <http://patents.justia.com/inventor/juan-manuel-vuletich>
>     linkedin.com/in/juan-vuletich-75611b3  <http://linkedin.com/in/juan-vuletich-75611b3>
>     twitter.com/JuanVuletich  <http://twitter.com/JuanVuletich>
>


-- 
Juan Vuletich
cuis.st
github.com/jvuletich
researchgate.net/profile/Juan-Vuletich
independent.academia.edu/JuanVuletich
patents.justia.com/inventor/juan-manuel-vuletich
linkedin.com/in/juan-vuletich-75611b3
twitter.com/JuanVuletich

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.cuis.st/mailman/archives/cuis-dev/attachments/20221115/e06684d8/attachment.htm>


More information about the Cuis-dev mailing list