[PATCH] Do not attach a debug location to code inserted by ARC / API for disabling DebugLocations

[Eric Christopher]

Hi Adrian,

Basically I was thinking earlier that this shouldn't be any different
than the cleanups being emitted from a scope and you've only changed
arc and blocks here. I understand that the code you have is very
applicable to the use case of arc and blocks, but I want to make sure
that it'll work for the rest of codegen as well.

Also if you look the cleanups are currently associated with the end of
the block, what's going on that we end up stopping on the particular
arc code at the end of the block first? Is it perhaps not being
associated with a statement that makes sense?

This is why I asked, I think it should be consistent across all of the
various language features. :)

-eric


On Sat, Apr 20, 2013 at 12:33 AM, Adrian Prantl <aprantl at apple.com> wrote:
>
> On Apr 15, 2013, at 11:53 PM, David Blaikie <dblaikie at gmail.com> wrote:
>
>> On Mon, Apr 8, 2013 at 3:07 PM, Adrian Prantl <aprantl at apple.com> wrote:
>>>
>>> Hi Eric,
>>>
>>> On Apr 8, 2013, at 1:23 PM, Eric Christopher <echristo at gmail.com> wrote:
>>>
>>>> So, since you didn't respond to the rest, these are the similar bugs I
>>>> was speaking about:
>>>>
>>>> http://llvm.org/bugs/show_bug.cgi?id=14498
>>>> http://llvm.org/bugs/show_bug.cgi?id=14580
>>>>
>>>> with a little bit due to:
>>>>
>>>> http://llvm.org/bugs/show_bug.cgi?id=14610
>>>> http://llvm.org/bugs/show_bug.cgi?id=14473
>>>
>>> Great! I was hesitant to answer to the rest before I knew more about those PRs :-)
>>>
>>>>
>>>> Before we go forward with this I'd like to discuss the general
>>>> applicability here for line information.
>>>
>>> Sure. My argument for the ARC case is that there is no sensible source line to associate calls to ARC runtime functions (that are mostly transparent to the programmer) with.
>>>
>
>
> I’ve updated the patches to address all comments I’ve got so far.
> The patch removes line table entries associated with code inserted by ARC, on the basis that there is no useful line to point to.
>
> Apart from that, we should definitely come to a consensus on how closely we want to follow GCC’s line table output and/or the GDB testsuite’s expectations. I think we might be able to discuss this in a thread separate from the review of this patch (especially since there is no GCC example to follow for ARC)?
>
> thanks for all the input so far,
> adrian
>
>
>
>