[Vm-dev] VM Maker: Cog-eem.89.mcz

Eliot Miranda eliot.miranda at gmail.com
Fri Sep 20 17:32:14 UTC 2013


On Fri, Sep 20, 2013 at 10:29 AM, Eliot Miranda <eliot.miranda at gmail.com>wrote:

> Frank,
>
>     yes.  This is checking that the bootstrap correctly replaces all quick
> methods that access Character's value inst var with methods that use
> primitive 410 (Character asInteger) instead.
>

Just to be clear, the halt will fire if there is a missing prototype on the
class side of SpurBootstrap.  The relevant Character prototypes there so
far are Character>>asInteger, Character>>asciiValue, Character>>hash
& Character>>identityHash.  I had forgotten Character>>hash.  Conceivably
the bootstrap could be applied to an image that has others; hence the halt.
 I guess I should comment it :-)


> I am *SO* happy that you are reading the code!!  Thanks!!
>
>
>
> On Tue, Sep 17, 2013 at 11:16 PM, Frank Shearar <frank.shearar at gmail.com>wrote:
>
>>
>> On 18 September 2013 02:11,  <commits at source.squeak.org> wrote:
>> >
>> > Eliot Miranda uploaded a new version of Cog to project VM Maker:
>> > http://source.squeak.org/VMMaker/Cog-eem.89.mcz
>> >
>> > ==================== Summary ====================
>> >
>> > Name: Cog-eem.89
>> > Author: eem
>> > Time: 17 September 2013, 6:11:13.622 pm
>> > UUID: d56a6961-bf18-450f-9aaa-427cfb9e753d
>> > Ancestors: Cog-eem.88
>> >
>> > Bring modifyCharacterMethods into withExecutableInterpreter:
>> > scope.
>> >
>> > Fix replacementForCharacterMethod: (was installing oop of
>> > asInteger selector as a SmallInteger, not the oop itself).
>> >
>> > Make sure interpreter:object:perform:withArguments: increments
>> > byteCount even if performed send is primitive.
>> >
>> > Needs VMMaker.oscog-eem.388.
>> >
>> > =============== Diff against Cog-eem.88 ===============
>>
>> Did you mean to leave a selt halt in SpurBootstrap>>
>> replacementForCharacterMethod:?
>>
>> frank
>>
>
>
>
> --
> best,
> Eliot
>



-- 
best,
Eliot
-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.squeakfoundation.org/pipermail/vm-dev/attachments/20130920/98aec6de/attachment-0001.htm


More information about the Vm-dev mailing list