Chris Quenelle <[EMAIL PROTECTED]> writes:
> Peter Memishian wrote:
>> > Are there any news about CR #6575435 ("ctf tools cannot handle C99 VLAs
>> > ("variable length arrays")") or is there any ETA when this bug may get
>> > fixed ?
>>
>> Doesn't look like it has an RE yet, so it seems there's no ETA.
>>
>
> I'm not sure what's going on with this problem:
>
> cause known: 6575435 ctf tools cannot handle C99 VLAs ("variable length
> arrays")
> fix delivered: 5035276 ctfconvert needs to support the new VLA stab
> accept: 6379193 ctfconvert needs to support VLA for gcc's DWARF
>
> I'm not sure what the difference is between 6575435 and 5035276.
You guys (Studio) changed the stab format output for C99 VLAs, mildly
re-breaking it. :)
The ctf tools don't like # in a stabs symbol, Studio 11 outputs either
the vla, or the sizor (I forget) as #vlaN where N is an integer.
As I've suggested previously, I think it is enough to make ctf discard
the leading # (as it does with other such characters), but I don't
have the time to test that theory (I tested it very very lightly when
Roland first brought this up, and it 'worked' in that ctfconvert
didn't fail on a simple test case. Whether it worked *properly* I don't
know.)
I have vague recollection of John telling me I was entirely wrong.
Maybe he remembers how.
-- Rich
_______________________________________________
tools-discuss mailing list
tools-discuss@opensolaris.org
