<!DOCTYPE html>
<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=UTF-8">
</head>
<body>
<p><br>
</p>
<div class="moz-cite-prefix">On 11/29/2024 10:22 AM, H.J. Lu wrote:<br>
</div>
<blockquote type="cite"
cite="mid:CAMe9rOoQsqonM-tH66fm4aOGGANsKUNTE+BUKM1MD5MAc64xYg@mail.gmail.com">
<meta http-equiv="content-type" content="text/html; charset=UTF-8">
<div dir="auto">
<div dir="auto">On Fri, Nov 29, 2024, 9:41 AM Guo, Wangyang <<a
href="mailto:wangyang.guo@intel.com" moz-do-not-send="true"
class="moz-txt-link-freetext">wangyang.guo@intel.com</a>>
wrote:</div>
<div class="gmail_quote" dir="auto">
<blockquote class="gmail_quote"
style="margin:0 0 0 .8ex;border-left:1px #ccc solid;padding-left:1ex">On
11/29/2024 6:02 AM, H.J. Lu wrote:<br>
<br>
> On Fri, Nov 29, 2024 at 12:24 AM Wilco Dijkstra <<a
href="mailto:Wilco.Dijkstra@arm.com" target="_blank"
rel="noreferrer" moz-do-not-send="true"
class="moz-txt-link-freetext">Wilco.Dijkstra@arm.com</a>>
wrote:<br>
>> Hi H.J.,<br>
>><br>
>> +static __always_inline void<br>
>> +clear_small_memory (INTERNAL_SIZE_T *mem, unsigned
long nclears)<br>
<br>
>> to avoiding unpredictable branches in this
benchmark.<br>
> Wangang, can you try memset only on Xeon like this?<br>
<br>
only using memset does not work well in Xeon platform.<br>
<br>
Test Platform: Xeon-8380 Bench: bench-calloc-thread Ratio:
New / <br>
Original time_per_iteration (Lower is Better) Threads# |
Ratio <br>
-----------|------ 1 thread | 1.018 4 threads | 1.015<br>
</blockquote>
</div>
<div dir="auto"><br>
</div>
<div dir="auto">Please try my v4 patch with and without ISA
level 3.</div>
</div>
</blockquote>
<p>Which build options do I need to apply?<br>
</p>
</body>
</html>