I noticed this vmlinux layout on i686 (where CONFIG_X86_L1_CACHE_SHIFT = 7) :
c06cdab4 d pid_caches_lh
c06cdb00 d qlowmark
c06cdb04 d qhimark
c06cdb08 d blimit
c06cdb80 d rcu_ctrlblk
c06cdc80 d rcu_bh_ctrlblk
This means that qlowmark, qhimark and blimit use a whole 128 bytes cache line.
Linker is not smart enough for us.
Moving these three variables to read_mostly section saves 116 (128-12) bytes.
# size vmlinux vmlinux.before_patch
text data bss dec hex filename
6343966 490818 630784 7465568 71ea60 vmlinux
6343966 490930 630784 7465680 71ead0 vmlinux.before_patch
Signed-off-by: Eric Dumazet <[email protected]>
diff --git a/kernel/rcupdate.c b/kernel/rcupdate.c
index a66d4d1..11c815c 100644
--- a/kernel/rcupdate.c
+++ b/kernel/rcupdate.c
@@ -75,9 +75,9 @@ DEFINE_PER_CPU(struct rcu_data, rcu_bh_data) = { 0L };
/* Fake initialization required by compiler */
static DEFINE_PER_CPU(struct tasklet_struct, rcu_tasklet) = {NULL};
-static int blimit = 10;
-static int qhimark = 10000;
-static int qlowmark = 100;
+static int blimit __read_mostly = 10;
+static int qhimark __read_mostly = 10000;
+static int qlowmark __read_mostly = 100;
static atomic_t rcu_barrier_cpu_count;
static DEFINE_MUTEX(rcu_barrier_mutex);
[Index of Archives]
[Kernel Newbies]
[Netfilter]
[Bugtraq]
[Photo]
[Stuff]
[Gimp]
[Yosemite News]
[MIPS Linux]
[ARM Linux]
[Linux Security]
[Linux RAID]
[Video 4 Linux]
[Linux for the blind]
[Linux Resources]
