On Sun, Jan 30, 2022 at 01:33:36AM +0200, Nir Soffer wrote:
Add benchmark for coping a buffer using 3 strategies - reusing same buffer, making a new uninitialized buffer per copy, and using a zeroed buffer per copy. This benchmark is the worst possible case, coping a
copying
buffer to memory. Any real I/O will be much slower hiding the overhead of allocating buffer or zeroed buffered.
will be much slower, hiding the overhead of allocating or zeroing buffers.
$ go test -run=AioBuffer -bench=Copy -benchtime=5s goos: linux goarch: amd64 pkg: libguestfs.org/libnbd cpu: Intel(R) Core(TM) i7-10850H CPU @ 2.70GHz BenchmarkAioBufferCopyBaseline-12 1142508 4523 ns/op BenchmarkAioBufferCopyMake-12 1000000 5320 ns/op BenchmarkAioBufferCopyMakeZero-12 728940 8218 ns/op
Signed-off-by: Nir Soffer <nsoffer@redhat.com> --- golang/libnbd_620_aio_buffer_test.go | 32 ++++++++++++++++++++++++++++ 1 file changed, 32 insertions(+)
diff --git a/golang/libnbd_620_aio_buffer_test.go b/golang/libnbd_620_aio_buffer_test.go index 8f65b558..9ab411b3 100644 --- a/golang/libnbd_620_aio_buffer_test.go +++ b/golang/libnbd_620_aio_buffer_test.go @@ -195,10 +195,42 @@ func BenchmarkAioBufferBytes(b *testing.B) { func BenchmarkAioBufferSlice(b *testing.B) { buf := MakeAioBuffer(bufferSize) defer buf.Free() var r int
b.ResetTimer()
Looks like you had a blank line with trailing whitespace in an earlier patch. -- Eric Blake, Principal Software Engineer Red Hat, Inc. +1-919-301-3266 Virtualization: qemu.org | libvirt.org