(j3.2006) Did we intend to prohibit this?
Tue Mar 14 01:09:35 EDT 2017
>Since I can't tell
If you can't tell one might surmise there is no point.
> whether a processor optimizes
> v = f(x)
>to do what
> call move_alloc ( f(x), v )
>would obviously do,
It obviously generates an error message. That's not what an assignment
> because instead of answering the question my request
>got shouted down
No it did not. No-one "shouted". No-one stopped anyone else from replying
to your questions. (Questions that IMO might be better addressed via your
vendors' support channels.)
>, and I assume there aren't command-line options to
>request that a processor do it or suppress it, I can't use a profiler to
>find out if this is true. Does anybody have profiler results, from 1/3
>million-line codes that run fifteen hours on 400 cores, to verify this?
Why on earth would you run a large code for a long time on many cores to
verify what happens in one single statement. It would not take very much
code, time, or cores, to check what happens here.
>reduced the time spent in the storage allocator from 12% to 3%. This
>observation suggested to me that answers to questions about avoiding the
>storage allocator would be useful at least to me.
If avoiding the memory allocator is your aim (and MOVE_ALLOC does not avoid
it, but anyway), you say there is a maximum effect of 3% on your runtime.
That does not sound worth getting all steamed up about. Especially when a
few simple experiments might shed light on the topic.
.............Malcolm Cohen, NAG Oxford/Tokyo.
More information about the J3