From a583a2bc59a4ee2b067e5520f6c5bc0c61852c32 Mon Sep 17 00:00:00 2001 From: Eelco Dolstra 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 '' -A system real 0m0.974s user 0m0.875s sys 0m0.088s $ NIX_REMOTE=daemon nix-instantiate '' -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 '' -A system real 0m1.074s user 0m0.809s sys 0m0.098s --- configure.ac | 4 ++++ 1 file changed, 4 insertions(+) (limited to 'configure.ac') diff --git a/configure.ac b/configure.ac index 9ffefa914..89f009923 100644 --- a/configure.ac +++ b/configure.ac @@ -127,6 +127,10 @@ AC_CHECK_HEADERS([sys/mount.h], [], [], AC_CHECK_FUNCS([lutimes]) +# Check for sched_setaffinity. +AC_CHECK_FUNCS([sched_setaffinity]) + + # Check whether the store optimiser can optimise symlinks. AC_MSG_CHECKING([whether it is possible to create a link to a symlink]) ln -s bla tmp_link -- cgit v1.2.3