llvm dev - Oct 2017 - Empty implementation of SchedulingPriorityQueue::dump

Hi All,

  While reading SchedulePostRATDList::ListScheduleTopDown() [1], I
find SchedulingPriorityQueue::dump has an empty implementation. Therefore,
the following debug dump basically outputs nothing.

      DEBUG(dbgs() << "\n*** Examining Available\n";
AvailableQueue.dump(this));

  Not sure why we want this. Is it intended?

[1] http://llvm.org/doxygen/PostRASchedulerList_8cpp_source.html

Regards,
chenwj

-- 
Wei-Ren Chen (陳韋任)
Homepage: https://people.cs.nctu.edu.tw/~chenwj
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20171025/a11ce777/attachment.html>

Hi,

On 25/10/2017 15:34, 陳韋任 via llvm-dev wrote:
> Hi All,
> 
>    While reading SchedulePostRATDList::ListScheduleTopDown() [1], I 
> find SchedulingPriorityQueue::dump has an empty implementation. 
> Therefore, the following debug dump basically outputs nothing.
> 
>        DEBUG(dbgs() << "\n*** Examining Available\n"; 
> AvailableQueue.dump(this));
>    Not sure why we want this. Is it intended?
> 
It seems like no-one bothered implementing it so far. It should be 
fairly straight forward to implement for LatencyPriorityQueue though. 
AFAIK, there is not much work on PostRAscheudlerList going on, as there 
is the newer PostMachineScheduler.

Cheers,
Florian

2017-10-25 22:59 GMT+08:00 Florian Hahn <florian.hahn at arm.com>:
> Hi,
>
> On 25/10/2017 15:34, 陳韋任 via llvm-dev wrote:
>
>> Hi All,
>>
>>    While reading SchedulePostRATDList::ListScheduleTopDown() [1], I
>> find SchedulingPriorityQueue::dump has an empty implementation.
>> Therefore, the following debug dump basically outputs nothing.
>>
>>        DEBUG(dbgs() << "\n*** Examining Available\n";
>> AvailableQueue.dump(this));
>>    Not sure why we want this. Is it intended?
>>
>>
> It seems like no-one bothered implementing it so far. It should be fairly
> straight forward to implement for LatencyPriorityQueue though. AFAIK, there
> is not much work on PostRAscheudlerList going on, as there is the newer
> PostMachineScheduler.
>
​I am playing around Hexagon test case, and I find it still
run PostRAscheudlerList under O2, unless I add -misched-postra option.

-- 
Wei-Ren Chen (陳韋任)
Homepage: https://people.cs.nctu.edu.tw/~chenwj
-------------- next part --------------
An HTML attachment was scrubbed...
URL:
<http://lists.llvm.org/pipermail/llvm-dev/attachments/20171025/724fc2b4/attachment.html>