[llvm] r176136 - Have a way for a target to opt-out of target-independent fast isel

[Eric Christopher]

On Wed, Feb 27, 2013 at 11:55 AM, Michael Ilseman <milseman at apple.com>wrote:

>
> On Feb 27, 2013, at 11:36 AM, Eric Christopher <echristo at gmail.com> wrote:
>
>
>
>
> On Wed, Feb 27, 2013 at 11:32 AM, Michael Ilseman <milseman at apple.com>wrote:
>
>>
>> On Feb 27, 2013, at 11:24 AM, Eric Christopher <echristo at gmail.com>
>> wrote:
>>
>>
>>
>>
>> On Tue, Feb 26, 2013 at 8:37 PM, Michael Ilseman <milseman at apple.com>wrote:
>>
>>>
>>> On Feb 26, 2013, at 3:30 PM, Eli Bendersky <eliben at google.com> wrote:
>>>
>>> > Can you provide more background for this change? Is it being done for
>>> > the sake of debugging? What are the intended use cases?
>>> >
>>>
>>> Sure! Debugging can definitely be improved by this, but this also gives
>>> targets more flexibility. Targets can fast-isel better code with more
>>> coverage in some situations. For example, SelectOperator(), even when it
>>> returns false, has side-effects and can schedule IR instructions to be
>>> fast-iselled that otherwise never would. This becomes particularly
>>> unfortunate when such instructions trigger the DAG slow-path.
>>>
>>>
>> Is there anything that's going to be using this any time soon? Something
>> that doesn't use the target independent fast isel path?
>>
>>
>> There is an out-of-tree backend currently using it.
>>
>
> Seems a bit early to commit effectively dead code unless there's something
> coming down soonish? :)
>
>
> Reverted in r176204
>

FWIW totally appreciate the desire to upstream as much as possible as early
as possible btw and very much appreciate it :)

-eric
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.llvm.org/pipermail/llvm-commits/attachments/20130227/6b50fc6f/attachment.html>