The patch titled
Subject: proc: spread likely/unlikely a bit
has been added to the -mm tree. Its filename is
proc-spread-likely-unlikely-a-bit.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/proc-spread-likely-unlikely-a-bit.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/proc-spread-likely-unlikely-a-bit.patch
Before you just go and hit "reply", please:
a) Consider who else should be cc'ed
b) Prefer to cc a suitable mailing list as well
c) Ideally: find the original patch on the mailing list and do a
reply-to-all to that, adding suitable additional cc's
*** Remember to use Documentation/SubmitChecklist when testing your code ***
The -mm tree is included into linux-next and is updated
there every 3-4 working days
------------------------------------------------------
From: Alexey Dobriyan <adobriyan@xxxxxxxxx>
Subject: proc: spread likely/unlikely a bit
use_pde() is used at every open/read/write/... of every random /proc
file. Negative refcount happens only if PDE is being deleted by module
(read: never). So it gets "likely".
unuse_pde() gets "unlikely" for the same reason.
close_pdeo() gets unlikely as the completion is filled only if there is a
race between PDE removal and close() (read: never ever).
It even saves code on x86_64 defconfig:
add/remove: 0/0 grow/shrink: 1/2 up/down: 2/-20 (-18)
Function old new delta
close_pdeo 183 185 +2
proc_reg_get_unmapped_area 119 111 -8
proc_reg_poll 85 73 -12
Link: http://lkml.kernel.org/r/20180104175657.GA5204@avx2
Signed-off-by: Alexey Dobriyan <adobriyan@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
fs/proc/inode.c | 8 ++++----
1 file changed, 4 insertions(+), 4 deletions(-)
diff -puN fs/proc/inode.c~proc-spread-likely-unlikely-a-bit fs/proc/inode.c
--- a/fs/proc/inode.c~proc-spread-likely-unlikely-a-bit
+++ a/fs/proc/inode.c
@@ -129,12 +129,12 @@ enum {BIAS = -1U<<31};
static inline int use_pde(struct proc_dir_entry *pde)
{
- return atomic_inc_unless_negative(&pde->in_use);
+ return likely(atomic_inc_unless_negative(&pde->in_use));
}
static void unuse_pde(struct proc_dir_entry *pde)
{
- if (atomic_dec_return(&pde->in_use) == BIAS)
+ if (unlikely(atomic_dec_return(&pde->in_use) == BIAS))
complete(pde->pde_unload_completion);
}
@@ -167,7 +167,7 @@ static void close_pdeo(struct proc_dir_e
spin_lock(&pde->pde_unload_lock);
/* After ->release. */
list_del(&pdeo->lh);
- if (pdeo->c)
+ if (unlikely(pdeo->c))
complete(pdeo->c);
kfree(pdeo);
}
@@ -421,7 +421,7 @@ static const char *proc_get_link(struct
struct delayed_call *done)
{
struct proc_dir_entry *pde = PDE(inode);
- if (unlikely(!use_pde(pde)))
+ if (!use_pde(pde))
return ERR_PTR(-EINVAL);
set_delayed_call(done, proc_put_link, pde);
return pde->data;
_
Patches currently in -mm which might be from adobriyan@xxxxxxxxx are
proc-use-%u-for-pid-printing-and-slightly-less-stack.patch
proc-dont-use-read_once-write_once-for-proc-fail-nth.patch
proc-fix-proc-map_files-lookup.patch
proc-simpler-proc-vmcore-cleanup.patch
proc-less-memory-for-proc-map_files-readdir.patch
proc-delete-children_seq_release.patch
proc-rearrange-struct-proc_dir_entry.patch
proc-fixup-comment.patch
proc-spread-__ro_after_init.patch
proc-spread-likely-unlikely-a-bit.patch
uuid-cleanup-uapi-linux-uuidh.patch
seq_file-delete-small-value-optimization.patch
cpumask-make-cpumask_size-return-unsigned-int.patch
--
To unsubscribe from this list: send the line "unsubscribe mm-commits" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
