From d583f0f96691a3edcddef92a6066ee9f191610d1 Mon Sep 17 00:00:00 2001
From: Thomas Schwinge <thomas@codesourcery.com>
Date: Fri, 25 Jul 2014 15:19:54 +0200
Subject: open_issues/performance: Configuration required for huge pages and
CPU sets.
---
open_issues/performance.mdwn | 27 +++++++++++++++++++++++----
1 file changed, 23 insertions(+), 4 deletions(-)
(limited to 'open_issues')
diff --git a/open_issues/performance.mdwn b/open_issues/performance.mdwn
index 64b245f2..ab08dd11 100644
--- a/open_issues/performance.mdwn
+++ b/open_issues/performance.mdwn
@@ -241,13 +241,26 @@ call|/glibc/fork]]'s case.
<teythoon> i especially like how general it is
-# Measurement
+# <a name="measure">Measure</a>
+
+On some pages, we're filing information about performace measurements.
+
+
+## kepler.SCHWINGE
+
+Debian GNU/Linux, x86. Running as a Xen domU, the system is not reserved
+exclusively for measurement purposes, so it's a best-effort service.
+
+
+## laplace.SCHWINGE
+
+Debian GNU/Hurd, x86. Running as a QEMU/KVM instance, the system is not
+reserved exclusively for measurement purposes, so it's a best-effort service.
-## coulomb
### [[!message-id "87wqghouoc.fsf@schwinge.name"]]
-## IRC, freenode, #hurd, 2014-02-27
+### IRC, freenode, #hurd, 2014-02-27
<braunr> tschwinge: about your concern with regard to performance
measurements, you could run kvm with hugetlbfs and cpuset
@@ -257,4 +270,10 @@ call|/glibc/fork]]'s case.
<braunr> hugetlbs reduces the overhead of page faults, and also implies
locked memory while cpuset isolates the vm from global scheduling
<braunr> hugetlbfs*
- <tschwinge> Thanks, will look into that.
+
+
+### 2014-07-25, tschwinge
+
+Support for [huge pages](https://wiki.debian.org/Hugepages) as well as [CPU
+sets](https://code.google.com/p/cpuset/) requires special setup; not doing that
+at the moment.
--
cgit v1.2.3