auto: Add a check for Linux's sched_getaffinity(2)

This will help to better determine the number of router threads to
create in certain situations.

Unlike sysconf(_SC_NPROCESSORS_ONLN) this takes into account per-process
cpu allowed masks as set by sched_setaffinity(2)/cpusets etc.

So while a system may have 64 on-line cpu's, Unit itself may be limited
to using just four of them in which case we should create four extra
router threads, not sixty-four!

Signed-off-by: Andrew Clayton <a.clayton@nginx.com>

2 files changed, 20 insertions, 0 deletions

diff --git a/auto/sched b/auto/sched
new file mode 100644
index 00000000..a6a4499b
--- /dev/null
+++ b/auto/sched
@@ -0,0 +1,19 @@
+# Copyright (C) Andrew Clayton
+# Copyright (C) NGINX, Inc.
+
+
+nxt_feature="Linux sched_getaffinity()"
+nxt_feature_name=NXT_HAVE_LINUX_SCHED_GETAFFINITY
+nxt_feature_run=no
+nxt_feature_incs=
+nxt_feature_libs=
+nxt_feature_test="#define _GNU_SOURCE
+                  #include <sched.h>
+
+                  int main(void) {
+                      cpu_set_t set;
+
+                      sched_getaffinity(0, sizeof(set), &set);
+                      return 0;
+                  }"
+. auto/feature
diff --git a/configure b/configure
index 4ad0dcdc..6929d41d 100755
--- a/configure
+++ b/configure
@@ -135,6 +135,7 @@ fi
 . auto/cgroup
 . auto/isolation
 . auto/capability
+. auto/sched
 
 
 case "$NXT_SYSTEM_PLATFORM" in