[llvm-commits] [llvm] r49243 - /llvm/trunk/lib/CodeGen/SelectionDAG/SelectionDAGISel.cpp

[Dale Johannesen]

On Apr 7, 2008, at 9:49 AM, Evan Cheng wrote:

> Ok, can you add a "set root to first pending load and flush" function
> when you have a chance?
>
> Thanks.
>
> Evan

No, as I said, I do not believe adding a new function is a good idea.   
If you want to change something here the right thing is to rename  
getRoot and getControlRoot to something like flushXXXandAdvanceRoot.

> On Apr 5, 2008, at 10:15 AM, Dale Johannesen wrote:
>
>>
>> On Apr 5, 2008, at 1:57 AM, Evan Cheng wrote:
>>
>>> Is this fixing a regression from Dan's patch? Can you separate out
>>> the
>>> flushing functionality from getroot? It seems a bit strange to use
>>> getRoot for its side effects.
>>>
>>> Evan
>>
>> Well, things named getXXX should not have side effects, IMO.  If I
>> were going
>> to work on naming I would probably rename getRoot and getControlRoot
>> to something
>> more descriptive rather than add a new function.  But in general I
>> prefer minimal
>> changes to more invasive ones.  (I wasn't familiar with this code
>> either.)
>>
>> Glad this fixed some Ada regressions, not too surprising.
>>
>> _______________________________________________
>> llvm-commits mailing list
>> llvm-commits at cs.uiuc.edu
>> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits
>
> _______________________________________________
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
> http://lists.cs.uiuc.edu/mailman/listinfo/llvm-commits