Trunk test status

On Thu, 26 Jul 2012, Frank Shearar wrote:

In a freshly updated trunk we have 3184 out of 3157 tests passing. We
have 16 expected failures, and 11 failures, the latter being:
* BecomeTest>>#testBecomeIdentityHash

This is failing due to a VM bug. There's a fix for it somewhere, but it seems like it's not integrated into Cog yet. Explore this to see that two consecutive objects share the same identityHash:

Array new: 10 streamContents: [ :stream |
        1 to: 10 do: [ :e | stream nextPut: Object new identityHash ] ]

* ExceptionTests>>#testHandlerFromAction

This is more like a feature request, the current exception handling mechanism doesn't work like this.

* LocaleTest>>#testLocaleChanged

A bug introduced during the GetText integration.

* MCFileInTest>>#testStWriter
* MCMczInstallerTest>>#testInstallFromFile
* MCMczInstallerTest>>#testInstallFromStream

Some old and funky MC issues, if you run the tests one or two more times,
they somehow pass.

* PackageDependencyTest>>#testMultilingual

Looks okay, just update the dependencies.

* PackageDependencyTest>>#testSystem

This one is a side effect of the GetText integration. Not sure if NaturalLanguageTranslator should use TextDomainManager or if it should be part of the System package. Also not sure about other dependencies here either.

* RWBinaryOrTextStreamTest>>#testExisiting

Another feature request, which changes the semantics of ReadWriteStream. Not implemented yet, because it breaks some code, so it requires a larger rewrite.

* ReleaseTest>>#testNoObsoleteClasses

As you described, some TestCases hold references to obsolete classes and the TestRunner holds the reference to the TestCases.

* SocketTest>>#testSendTimeout

Either a bug or just some network problem, works for me on windows.


Levente

ReleaseTest>>#testNoObsoleteClasses lists a bunch of what look to be
test artifacts: things like
AnObsoleteAutoGeneratedClassForTestingSystemChanges. ChangeHooksTest
and ClassRenameFixTest look to be the culprits. How do we make these
two guys not interfere with ReleaseTest? (Or, how do we nuke the
obsolete classes created by these guys?)

PackageDependencyTest>>#testExisiting [sic] and #testMultilingual
looks like they need updating. I don't know if these dependencies
SHOULD be there, but they ARE there. I'll fix them to reflect reality,
but if they shouldn't be there, please raise a Mantis report to remove
the dependency!

I suspect that SocketTest might well involve platform dependencies: it
fails on my machine, an Ubuntu Lucid Lynx running the latest Cog VM.

I'd really like to have a green test status before we ship 4.4, so I'm
asking for folks to take a look at the above tests (I'll do the
PackageDependencyTest ones) and see if anyone can make them green? And
then maybe, for bonus brownie points, see if we can reduce the number
of expected failures?

Thanks!

frank

Ah OK, I was not concentrated enough when I read Eliot post :)

Eliot, this is not a bug, and normal code would statistically create
other objects in between...
But isn't there a risk  that tools like Fuel populate a small
collection of simple objects (say an IdentitySet of Symbol), sharing
same identityHash for each pair?
It would noticeably increase the number of collisions well before the 4096 wall.

Nicolas

2012/7/27 Bert Freudenberg <[hidden email]>:

>
> On 27.07.2012, at 05:18, Nicolas Cellier wrote:
>
>> 2012/7/27 Frank Shearar <[hidden email]>:
>>> On 26 July 2012 22:35, Eliot Miranda <[hidden email]> wrote:
>>>>
>>>>
>>>> On Thu, Jul 26, 2012 at 3:59 AM, Levente Uzonyi <[hidden email]> wrote:
>>>>>
>>>>> On Thu, 26 Jul 2012, Frank Shearar wrote:
>>>>>
>>>>>> In a freshly updated trunk we have 3184 out of 3157 tests passing. We
>>>>>> have 16 expected failures, and 11 failures, the latter being:
>>>>>> * BecomeTest>>#testBecomeIdentityHash
>>>>>
>>>>>
>>>>> This is failing due to a VM bug. There's a fix for it somewhere, but it
>>>>> seems like it's not integrated into Cog yet. Explore this to see that two
>>>>> consecutive objects share the same identityHash:
>>>>>
>>>>> Array new: 10 streamContents: [ :stream |
>>>>>        1 to: 10 do: [ :e | stream nextPut: Object new identityHash ] ]
>>>>
>>>>
>>>> IMO this isn't a bug.  The identity hash changes at least every other
>>>> object.  Hashes don't have to be unique.  But they do have to be
>>>> well-distributed.  With 12 bits of identityHash Cog does fine basing its
>>>> identityHash on the allocation pointer.  The above will wrap around after
>>>> 8192 allocations, and provide 4096 distinct hashes (the maximum available).
>>>> So the test needs rewriting to be more statistical.  The rationale for this
>>>> is to speed up allocation.  Instead of a read-modify-write cycle to turn the
>>>> crank of a pseudo-random generator there's a masking of the allocation
>>>> pointer, which has to be read anyway to allocate an object.  BTW, the
>>>> *right* way to implement this is to lazily allocate hashes, but for that
>>>> there needs to be a flag (e.g. an identityHash of 0) to mark an object as
>>>> not yet having a hash but existing Squeak images (because of the old
>>>> definition) use 0 as a valid hash, so lazy hashes requires either a header
>>>> bit (not enough of those) or an image change (which is my plan, as part of
>>>> the new object representation).
>>>
>>> If it's not a bug, let's nuke the test. We need to get to a position
>>> where we have a green light.
>>>
>>> frank
>>>
>>
>> If I understood Eliot correctly, it would suffice to keep a pointer
>> alive to the created objects...
>>
>> preserveObjectsFromGarbageCollection := IdentitySet new.
>> Array new: 10 streamContents: [ :stream |
>>         1 to: 10 do: [ :e | stream nextPut:
>> (preserveObjectsFromGarbageCollection add: Object new) identityHash ]
>> ]
>>
>> Nicolas
>
>
> Makes no difference. GC does not happen after each allocation. Here's one that would work with Cog because each allocation is larger:
>
> (1 to: 10) collect: [ :e | (Array new: 4) identityHash ]
>
> But as Eliot said the test is somewhat meaningless in its current form.
>
> - Bert -
>
>
>