Re: Performance of GC.collect() for single block of `byte`s

Per Nordlöw via Digitalmars-d-learn Fri, 28 Sep 2018 02:35:47 -0700

On Friday, 28 September 2018 at 09:14:18 UTC, Per Nordlöw wrote:

How is it possible for the GC to be 500-1000 times slower thana malloc-free call for a single array containing just byteswith no indirections for such a simple function!!!?
I really don't understand this...


I change the code to not make use of the GC when printing:

import core.stdc.stdio: printf;

void* mallocAndFreeBytes(size_t byteCount)()
{
    import core.memory : pureMalloc, pureFree;
    void* ptr = pureMalloc(byteCount);
    pureFree(ptr);
    return ptr;                 // for side-effects
}

void main(string[] args)
{
    import std.datetime.stopwatch : benchmark;
    import core.time : Duration;

    immutable benchmarkCount = 1;

    // GC
    static foreach (const size_t i; 0 .. 32)
    {
        {
            enum byteCount = 2UL^^i;

const Duration[1] resultsC =benchmark!(mallocAndFreeBytes!(i))(benchmarkCount);

            printf("%ld bytes: mallocAndFreeBytes: %ld nsecs",

byteCount,cast(size_t)(cast(double)resultsC[0].total!"nsecs"/benchmarkCount));


            import core.memory : GC;
            auto dArray = new byte[byteCount]; // one Gig

const Duration[1] resultsD =benchmark!(GC.collect)(benchmarkCount);

            printf("  GC.collect(): %ld nsecs after %p\n",

cast(size_t)(cast(double)resultsD[0].total!"nsecs"/benchmarkCount), dArray.ptr);

            dArray = null;
        }
    }
}

I still get terrible numbers:

1 bytes: mallocAndFreeBytes: 600 nsecs GC.collect(): 29600 nsecsafter 0x7fbab535b0002 bytes: mallocAndFreeBytes: 500 nsecs GC.collect(): 28600 nsecsafter 0x7fbab535b0104 bytes: mallocAndFreeBytes: 400 nsecs GC.collect(): 27700 nsecsafter 0x7fbab535b0008 bytes: mallocAndFreeBytes: 400 nsecs GC.collect(): 27600 nsecsafter 0x7fbab535b01016 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 32100nsecs after 0x7fbab535c00032 bytes: mallocAndFreeBytes: 400 nsecs GC.collect(): 27100nsecs after 0x7fbab535b00064 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 48500nsecs after 0x7fbab535c000128 bytes: mallocAndFreeBytes: 400 nsecs GC.collect(): 23300nsecs after 0x7fbab535b000256 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 22300nsecs after 0x7fbab535c000512 bytes: mallocAndFreeBytes: 400 nsecs GC.collect(): 21800nsecs after 0x7fbab535b0001024 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 21900nsecs after 0x7fbab535c0002048 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 26300nsecs after 0x7fbab3ebe0104096 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 25100nsecs after 0x7fbab3ebf0108192 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 24500nsecs after 0x7fbab3ec101016384 bytes: mallocAndFreeBytes: 400 nsecs GC.collect(): 24700nsecs after 0x7fbab3ec401032768 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 24600nsecs after 0x7fbab3ec901065536 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 24600nsecs after 0x7fbab3ed2010131072 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 25000nsecs after 0x7fbab3ee3010262144 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 25000nsecs after 0x7fbab3f04010524288 bytes: mallocAndFreeBytes: 400 nsecs GC.collect(): 25200nsecs after 0x7fbab3f450101048576 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 25800nsecs after 0x7fbab3fc60102097152 bytes: mallocAndFreeBytes: 300 nsecs GC.collect(): 17200nsecs after 0x7fbab37be0104194304 bytes: mallocAndFreeBytes: 500 nsecs GC.collect(): 25700nsecs after 0x7fbab39bf0108388608 bytes: mallocAndFreeBytes: 400 nsecs GC.collect(): 65500nsecs after 0x7fbab2bbd01016777216 bytes: mallocAndFreeBytes: 1100 nsecs GC.collect():47200 nsecs after 0x7fbab13bc01033554432 bytes: mallocAndFreeBytes: 800 nsecs GC.collect():50300 nsecs after 0x7fbaae3bb01067108864 bytes: mallocAndFreeBytes: 800 nsecs GC.collect():63800 nsecs after 0x7fbaa83ba010134217728 bytes: mallocAndFreeBytes: 600 nsecs GC.collect():100000 nsecs after 0x7fba9c3b9010268435456 bytes: mallocAndFreeBytes: 1000 nsecs GC.collect():176100 nsecs after 0x7fba843b8010536870912 bytes: mallocAndFreeBytes: 1000 nsecs GC.collect():415500 nsecs after 0x7fba543b70101073741824 bytes: mallocAndFreeBytes: 800 nsecs GC.collect():649900 nsecs after 0x7fb9f42f50102147483648 bytes: mallocAndFreeBytes: 1200 nsecs GC.collect():973800 nsecs after 0x7fb934112010

It seems to scale kind linearly with byteCount above 16Mb... Soit seems like its scanning the allocated block of bytes eventhough the element type of array is a value type. Why?

If I zero the pointer just after allocation I get a GC.collect()taking constantly 100ns so it can't be related to the stack.

Re: Performance of GC.collect() for single block of `byte`s

Reply via email to