[PATCH] [libc++abi] Support for __cxa_bad_array_new_length

[Aaron Ballman]

On Thu, Sep 11, 2014 at 1:09 PM, Dan Albert <danalbert at google.com> wrote:
> I've changed the test so that it calls __cxa_throw_bad_array_new_length()
> (I'm mad at whoever gave this function such a long name) directly rather
> than expecting the compiler to call it for bad input to new[]. That check
> can go in the compiler tests.

I'm okay with that approach,  but should this test then be merged back
into test_aux_runtime.cpp instead of being in a separate file?
Regardless, the comments at the top of the file should probably be
corrected, unless we're adding them as a FIXME to a commented-out
version of the original test case.

~Aaron


>
> On Thu, Sep 11, 2014 at 6:39 AM, Aaron Ballman <aaron at aaronballman.com>
> wrote:
>>
>> On Wed, Sep 10, 2014 at 11:52 AM, Dan Albert <danalbert at google.com> wrote:
>> > Sorry, I got used to relying on phabricator to keep track of things, and
>> > forgot about this one.
>>
>> No worries!
>>
>> > Could you split the test in to a separate file and add `// XFAIL: *` at
>> > the
>> > top? That way you can commit it with the guts of the patch and we don't
>> > have
>> > to worry about forgetting to submit the test later.
>> >
>> > Other than that, LGTM.
>>
>> I've split the test out into its own file, and have attached the patch
>> here.
>>
>> Since I don't have a way to test this locally, and no other tests have
>> XFAIL lines, I'm not quite comfortable committing this without someone
>> who can run the tests confirming that it runs cleanly.
>>
>> Thanks!
>>
>> ~Aaron
>
>