[PATCH] [lld][Core] Implement parallel_for_each
shankare at codeaurora.org
Mon Mar 16 11:50:15 PDT 2015
Agree, we need to debug this, but I am unaware of tools that can show
how threads are being scheduled ? Any pointers ?
In the mean time
On 3/16/2015 12:58 PM, Filipe Cabecinhas wrote:
> But why is it taking longer? Are we ending up waiting for one or two
> threads to do their work? Should we split the work in different ways? Does
> it take noticeably longer on one loop only, etc?
> Can you profile it a bit more and get more info, please? Especially
> relating to thread behaviors.
> On Monday, March 16, 2015, Shankar Kalpathi Easwaran <shankarke at gmail.com>
>> I was thinking of doing it non recursive too, but I thought its easier to
>> experiment or change the way tasks get spawned(by changing few numbers)
>> with a recursive version, no ??
>> The other problem is this approach is not helping and makes the link times
>> longer for building clang/self hosting. Should we move to using the
>> previous implementation ?
>> EMAIL PREFERENCES
>> llvm-commits mailing list
> llvm-commits mailing list
> llvm-commits at cs.uiuc.edu
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by the Linux Foundation
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the llvm-commits