[error] running script BuildSqueakSpurTrunkVMMakerImage.st

Previous Topic Next Topic
 
classic Classic list List threaded Threaded
6 messages Options
Reply | Threaded
Open this post in threaded view
|

[error] running script BuildSqueakSpurTrunkVMMakerImage.st

Robert Withers-2
 
In building the spurtrunkvmmaker64 image, at some point, after loading
Alien, the script BuildSqueakSpurTrunkVMMakerImage.st throws an error...

'FFICallbackThunk cannot be found in the list.'

--- The full stack ---
SmalltalkImage(Object)>>error:
SmalltalkImage>>add:toList:after:
SmalltalkImage>>addToStartUpList:after:
Callback class>>ensureNotifiedAtStartup
Callback class>>initialize
MCMethodDefinition>>postload
MCMethodDefinition(MCDefinition)>>postloadOver:


Reply | Threaded
Open this post in threaded view
|

Re: [error] running script BuildSqueakSpurTrunkVMMakerImage.st

Robert Withers-2
 
As soon as Spur64VMMaker.image is made from trunk6-64.image and starts
the error occur immediately with a dialog which states 'Initializing
Alien-Core-TorstenBergman.101'

On 3/2/20 12:37 AM, Robert wrote:

> In building the spurtrunkvmmaker64 image, at some point, after loading
> Alien, the script BuildSqueakSpurTrunkVMMakerImage.st throws an error...
>
> 'FFICallbackThunk cannot be found in the list.'
>
> --- The full stack ---
> SmalltalkImage(Object)>>error:
> SmalltalkImage>>add:toList:after:
> SmalltalkImage>>addToStartUpList:after:
> Callback class>>ensureNotifiedAtStartup
> Callback class>>initialize
> MCMethodDefinition>>postload
> MCMethodDefinition(MCDefinition)>>postloadOver:
>

Reply | Threaded
Open this post in threaded view
|

Re: [error] running script BuildSqueakSpurTrunkVMMakerImage.st

Nicolas Cellier
 
Hi Robert,
confirmed, thanks for reporting.
It seems that there is a problem with Monticello initialization order.

One should invoke "FFICallbackThunk initialize" before "Callback initialize" because the later wants to be added in startup list after the former...
Unfortunately, MC initialization order seems to be random...
We should find a workaround...

In the interim, you could try to evaluate
FFICallbackThunk initialize
then restart Callback class>>initialize from the debugger...

Le lun. 2 mars 2020 à 06:52, Robert <[hidden email]> a écrit :
 
As soon as Spur64VMMaker.image is made from trunk6-64.image and starts
the error occur immediately with a dialog which states 'Initializing
Alien-Core-TorstenBergman.101'

On 3/2/20 12:37 AM, Robert wrote:
> In building the spurtrunkvmmaker64 image, at some point, after loading
> Alien, the script BuildSqueakSpurTrunkVMMakerImage.st throws an error...
>
> 'FFICallbackThunk cannot be found in the list.'
>
> --- The full stack ---
> SmalltalkImage(Object)>>error:
> SmalltalkImage>>add:toList:after:
> SmalltalkImage>>addToStartUpList:after:
> Callback class>>ensureNotifiedAtStartup
> Callback class>>initialize
> MCMethodDefinition>>postload
> MCMethodDefinition(MCDefinition)>>postloadOver:
>

Reply | Threaded
Open this post in threaded view
|

Re: [error] running script BuildSqueakSpurTrunkVMMakerImage.st

Nicolas Cellier
 
But FFICallbackThunk initialize requires Alien initialize...
I think that I've just fixed it (to be tested).

But then, I get another error when evaluating:
Spur32to64BitBootstrap new bootstrapImage: 'trunk50'.

I see no such class.
Eliot, could you point us to the appropriate MC package?

Le lun. 2 mars 2020 à 10:21, Nicolas Cellier <[hidden email]> a écrit :
Hi Robert,
confirmed, thanks for reporting.
It seems that there is a problem with Monticello initialization order.

One should invoke "FFICallbackThunk initialize" before "Callback initialize" because the later wants to be added in startup list after the former...
Unfortunately, MC initialization order seems to be random...
We should find a workaround...

In the interim, you could try to evaluate
FFICallbackThunk initialize
then restart Callback class>>initialize from the debugger...

Le lun. 2 mars 2020 à 06:52, Robert <[hidden email]> a écrit :
 
As soon as Spur64VMMaker.image is made from trunk6-64.image and starts
the error occur immediately with a dialog which states 'Initializing
Alien-Core-TorstenBergman.101'

On 3/2/20 12:37 AM, Robert wrote:
> In building the spurtrunkvmmaker64 image, at some point, after loading
> Alien, the script BuildSqueakSpurTrunkVMMakerImage.st throws an error...
>
> 'FFICallbackThunk cannot be found in the list.'
>
> --- The full stack ---
> SmalltalkImage(Object)>>error:
> SmalltalkImage>>add:toList:after:
> SmalltalkImage>>addToStartUpList:after:
> Callback class>>ensureNotifiedAtStartup
> Callback class>>initialize
> MCMethodDefinition>>postload
> MCMethodDefinition(MCDefinition)>>postloadOver:
>

Reply | Threaded
Open this post in threaded view
|

Re: [error] running script BuildSqueakSpurTrunkVMMakerImage.st

Nicolas Cellier
 
Ah, I guess that it should now be:
Spur32to64BitImageConverter new bootstrapImage: 'trunk50'.

But maybe I also mixed up the scripts...

Le lun. 2 mars 2020 à 10:48, Nicolas Cellier <[hidden email]> a écrit :
But FFICallbackThunk initialize requires Alien initialize...
I think that I've just fixed it (to be tested).

But then, I get another error when evaluating:
Spur32to64BitBootstrap new bootstrapImage: 'trunk50'.

I see no such class.
Eliot, could you point us to the appropriate MC package?

Le lun. 2 mars 2020 à 10:21, Nicolas Cellier <[hidden email]> a écrit :
Hi Robert,
confirmed, thanks for reporting.
It seems that there is a problem with Monticello initialization order.

One should invoke "FFICallbackThunk initialize" before "Callback initialize" because the later wants to be added in startup list after the former...
Unfortunately, MC initialization order seems to be random...
We should find a workaround...

In the interim, you could try to evaluate
FFICallbackThunk initialize
then restart Callback class>>initialize from the debugger...

Le lun. 2 mars 2020 à 06:52, Robert <[hidden email]> a écrit :
 
As soon as Spur64VMMaker.image is made from trunk6-64.image and starts
the error occur immediately with a dialog which states 'Initializing
Alien-Core-TorstenBergman.101'

On 3/2/20 12:37 AM, Robert wrote:
> In building the spurtrunkvmmaker64 image, at some point, after loading
> Alien, the script BuildSqueakSpurTrunkVMMakerImage.st throws an error...
>
> 'FFICallbackThunk cannot be found in the list.'
>
> --- The full stack ---
> SmalltalkImage(Object)>>error:
> SmalltalkImage>>add:toList:after:
> SmalltalkImage>>addToStartUpList:after:
> Callback class>>ensureNotifiedAtStartup
> Callback class>>initialize
> MCMethodDefinition>>postload
> MCMethodDefinition(MCDefinition)>>postloadOver:
>

Reply | Threaded
Open this post in threaded view
|

Re: [error] running script BuildSqueakSpurTrunkVMMakerImage.st

Robert Withers-2
 

Hi Nicolas,

All is well now. I was able to build the SpurTrunkVmmaker64 image, cleanly. Thank you!

K, r

On 3/2/20 6:31 AM, Nicolas Cellier wrote:
Ah, I guess that it should now be:
Spur32to64BitImageConverter new bootstrapImage: 'trunk50'.

But maybe I also mixed up the scripts...

Le lun. 2 mars 2020 à 10:48, Nicolas Cellier <[hidden email]> a écrit :
But FFICallbackThunk initialize requires Alien initialize...
I think that I've just fixed it (to be tested).

But then, I get another error when evaluating:
Spur32to64BitBootstrap new bootstrapImage: 'trunk50'.

I see no such class.
Eliot, could you point us to the appropriate MC package?

Le lun. 2 mars 2020 à 10:21, Nicolas Cellier <[hidden email]> a écrit :
Hi Robert,
confirmed, thanks for reporting.
It seems that there is a problem with Monticello initialization order.

One should invoke "FFICallbackThunk initialize" before "Callback initialize" because the later wants to be added in startup list after the former...
Unfortunately, MC initialization order seems to be random...
We should find a workaround...

In the interim, you could try to evaluate
FFICallbackThunk initialize
then restart Callback class>>initialize from the debugger...

Le lun. 2 mars 2020 à 06:52, Robert <[hidden email]> a écrit :
 
As soon as Spur64VMMaker.image is made from trunk6-64.image and starts
the error occur immediately with a dialog which states 'Initializing
Alien-Core-TorstenBergman.101'

On 3/2/20 12:37 AM, Robert wrote:
> In building the spurtrunkvmmaker64 image, at some point, after loading
> Alien, the script BuildSqueakSpurTrunkVMMakerImage.st throws an error...
>
> 'FFICallbackThunk cannot be found in the list.'
>
> --- The full stack ---
> SmalltalkImage(Object)>>error:
> SmalltalkImage>>add:toList:after:
> SmalltalkImage>>addToStartUpList:after:
> Callback class>>ensureNotifiedAtStartup
> Callback class>>initialize
> MCMethodDefinition>>postload
> MCMethodDefinition(MCDefinition)>>postloadOver:
>