Helllo<br><br><div class="gmail_quote">On Fri, Jan 20, 2012 at 3:53 AM, Linda Walsh <span dir="ltr"><<a href="mailto:xfs@tlinx.org">xfs@tlinx.org</a>></span> wrote:<br><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<div class="im"><br>
<br>
Zheng Da wrote:<br>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
I create a file of 4GB in XFS (the ramdisk has 5GB of space). My test program overwrites 4G of data to the file <br>
</blockquote></div>
----<br>
It sounds like you are asking why multiple threads don't<br>
move memory from one point to another point in memory at a faster rate<br>
than one thread alone.<br>
<br>
I.e. if you had 2 processes doing an assembly instruction, memmov to move<br>
a chunk of memory from 1 area to another, would you expect to do the move<br>
any faster if you had 2 processors doing the move vs. 1??<br></blockquote><div>Yes. Actually, for reading, using multiple threads is faster than a single thread. </div><div>If you try simple memory copy with memcpy in the C library, the overall throughput will still increase if you use multiple processors.</div>
<blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
I think the limiting factor (unless you have a slow processor and some<br>
REALLY fast memory, but stock x86-64 parts, today have memory running about<br>
2-4 times slower than the processor -- so the memory is usually the bottleneck.<br></blockquote><div>Memory bandwidth will eventually become a bottleneck, but it can still scale for a small number of processors. </div><blockquote class="gmail_quote" style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">
<br>
Two processes wouldn't do it any faster, and might actually do it slower due to<br>
resource contention issues -- I would *think*... but I really don't know the<br>
details of how writing from mem2mem and having the target be in the format of<br>
and xfs file system, would cause cpu-bound delays that would be significant to<br>
change the fact that m2m operations are usually mem-bandwidth limited...?<br>
<br>
(I don't know the answers, just clarifying what you are asking)...<br><br></blockquote><div>Da </div></div><br>