[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Public WebGL] questions about vertexAttribPointer / getVertexAttrib offsets



On Fri, Apr 27, 2012 at 5:53 PM, Boris Zbarsky <bzbarsky@mit.edu> wrote:
> On 4/27/12 8:40 PM, Kenneth Russell wrote:
>>>
>>> That sounds fine.  The test suite needs updating accordingly: right now
>>> it
>>> has a test that tests that no error is generated when offset is
>>> -8000000000.
>>
>>
>> Gregg should be consulted; I agree the test needs to be changed, but
>> that particular test was added in order to catch a crash, so we should
>> make sure we aren't significantly decreasing test coverage.
>
>
> I think that simply changing the test to expect an error to be generated but
> continuing to pass the large value would keep the original "catch a crash"
> test coverage, right?

Yes, I agree. The test is relying on a bug in WebKit's bindings where
the large negative value is being truncated to 32 bits.

I am not completely sure that WebKit can even handle passing "long
long" values from JavaScript to C++. It will almost certainly not have
proper Web IDL behavior, where TypeError is thrown for out-of-range
values.

I've fixed the bug in the test in the TOT conformance suite, and would
like to backport this test bug fix to the 1.0.1 suite, even though it
will cause Chrome and Safari to start failing the test. Tim, in
particular, do you have any objections to this?

-Ken

-----------------------------------------------------------
You are currently subscribed to public_webgl@khronos.org.
To unsubscribe, send an email to majordomo@khronos.org with
the following command in the body of your email:
unsubscribe public_webgl
-----------------------------------------------------------