From a583a2bc59a4ee2b067e5520f6c5bc0c61852c32 Mon Sep 17 00:00:00 2001
From: Eelco Dolstra <eelco.dolstra@logicblox.com>
Date: Wed, 7 Aug 2013 11:51:55 +0000
Subject: Run the daemon worker on the same CPU as the client

On a system with multiple CPUs, running Nix operations through the
daemon is significantly slower than "direct" mode:

$ NIX_REMOTE= nix-instantiate '<nixos>' -A system
real    0m0.974s
user    0m0.875s
sys     0m0.088s

$ NIX_REMOTE=daemon nix-instantiate '<nixos>' -A system
real    0m2.118s
user    0m1.463s
sys     0m0.218s

The main reason seems to be that the client and the worker get moved
to a different CPU after every call to the worker.  This patch adds a
hack to lock them to the same CPU.  With this, the overhead of going
through the daemon is very small:

$ NIX_REMOTE=daemon nix-instantiate '<nixos>' -A system
real    0m1.074s
user    0m0.809s
sys     0m0.098s
---
 src/libutil/affinity.hh | 9 +++++++++
 1 file changed, 9 insertions(+)
 create mode 100644 src/libutil/affinity.hh

(limited to 'src/libutil/affinity.hh')

diff --git a/src/libutil/affinity.hh b/src/libutil/affinity.hh
new file mode 100644
index 000000000000..c1bd28e1367a
--- /dev/null
+++ b/src/libutil/affinity.hh
@@ -0,0 +1,9 @@
+#pragma once
+
+namespace nix {
+
+void setAffinityTo(int cpu);
+int lockToCurrentCPU();
+void restoreAffinity();
+
+}
-- 
cgit 1.4.1