+ thp-extract-khugepaged-from-mm-huge_memoryc-fix.patch added to -mm tree

+ thp-extract-khugepaged-from-mm-huge_memoryc-fix.patch added to -mm tree

[akpm]

The patch titled
     Subject: thp-extract-khugepaged-from-mm-huge_memoryc-fix
has been added to the -mm tree.  Its filename is
     thp-extract-khugepaged-from-mm-huge_memoryc-fix.patch

This patch should soon appear at
    http://ozlabs.org/~akpm/mmots/broken-out/thp-extract-khugepaged-from-mm-huge_memoryc-fix.patch
and later at
    http://ozlabs.org/~akpm/mmotm/broken-out/thp-extract-khugepaged-from-mm-huge_memoryc-fix.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: Andrew Morton <akpm@linux-foundation.org>
Subject: thp-extract-khugepaged-from-mm-huge_memoryc-fix

khugepaged_mutex can be static to mm/huge_memory.c:enabled_store()

Makes me wonder what it actually does...

Cc: "Kirill A. Shutemov" <kirill.shutemov@linux.intel.com>
Cc: Andrea Arcangeli <aarcange@redhat.com>
Cc: Andres Lagar-Cavilla <andreslc@google.com>
Cc: Aneesh Kumar K.V <aneesh.kumar@linux.vnet.ibm.com>
Cc: Christoph Lameter <cl@gentwo.org>
Cc: Dave Hansen <dave.hansen@intel.com>
Cc: Hugh Dickins <hughd@google.com>
Cc: Jerome Marchand <jmarchan@redhat.com>
Cc: Kirill A. Shutemov <kirill.shutemov@linux.intel.com>
Cc: Naoya Horiguchi <n-horiguchi@ah.jp.nec.com>
Cc: Ning Qu <quning@gmail.com>
Cc: Peter Zijlstra (Intel) <peterz@infradead.org>
Cc: Sasha Levin <sasha.levin@oracle.com>
Cc: Vlastimil Babka <vbabka@suse.cz>
Cc: Yang Shi <yang.shi@linaro.org>
Cc: neha agarwal <neha.agbk@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
---

 include/linux/khugepaged.h |    1 -
 mm/huge_memory.c           |    1 +
 mm/khugepaged.c            |    1 -
 3 files changed, 1 insertion(+), 2 deletions(-)

diff -puN include/linux/khugepaged.h~thp-extract-khugepaged-from-mm-huge_memoryc-fix include/linux/khugepaged.h
--- a/include/linux/khugepaged.h~thp-extract-khugepaged-from-mm-huge_memoryc-fix
+++ a/include/linux/khugepaged.h
@@ -4,7 +4,6 @@
 #include <linux/sched.h> /* MMF_VM_HUGEPAGE */
 
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
-extern struct mutex khugepaged_mutex;
 extern struct attribute_group khugepaged_attr_group;
 
 extern int khugepaged_init(void);
diff -puN mm/huge_memory.c~thp-extract-khugepaged-from-mm-huge_memoryc-fix mm/huge_memory.c
--- a/mm/huge_memory.c~thp-extract-khugepaged-from-mm-huge_memoryc-fix
+++ a/mm/huge_memory.c
@@ -174,6 +174,7 @@ static ssize_t enabled_store(struct kobj
 			     const char *buf, size_t count)
 {
 	ssize_t ret;
+	static DEFINE_MUTEX(khugepaged_mutex);
 
 	ret = triple_flag_store(kobj, attr, buf, count,
 				TRANSPARENT_HUGEPAGE_FLAG,
diff -puN mm/khugepaged.c~thp-extract-khugepaged-from-mm-huge_memoryc-fix mm/khugepaged.c
--- a/mm/khugepaged.c~thp-extract-khugepaged-from-mm-huge_memoryc-fix
+++ a/mm/khugepaged.c
@@ -57,7 +57,6 @@ static unsigned int khugepaged_scan_slee
 static unsigned int khugepaged_alloc_sleep_millisecs __read_mostly = 60000;
 static unsigned long khugepaged_sleep_expire;
 static struct task_struct *khugepaged_thread __read_mostly;
-DEFINE_MUTEX(khugepaged_mutex);
 static DEFINE_SPINLOCK(khugepaged_mm_lock);
 static DECLARE_WAIT_QUEUE_HEAD(khugepaged_wait);
 /*
_

Patches currently in -mm which might be from akpm@linux-foundation.org are

i-need-old-gcc.patch
arch-alpha-kernel-systblss-remove-debug-check.patch
revert-mm-memcontrol-fix-possible-css-ref-leak-on-oom.patch
arm-arch-arm-include-asm-pageh-needs-personalityh.patch
mm.patch
mm-reorganize-slab-freelist-randomization-fix.patch
mm-frontswap-convert-frontswap_enabled-to-static-key-checkpatch-fixes.patch
mm-add-nr_zsmalloc-to-vmstat-fix.patch
mm-memcg-use-consistent-gfp-flags-during-readahead-checkpatch-fixes.patch
mm-thp-avoid-unnecessary-swapin-in-khugepaged-fix.patch
mm-thp-make-swapin-readahead-under-down_read-of-mmap_sem-fix-2-fix.patch
thp-extract-khugepaged-from-mm-huge_memoryc-fix.patch
mm-kasan-switch-slub-to-stackdepot-enable-memory-quarantine-for-slub-fix.patch
mm-kasan-switch-slub-to-stackdepot-enable-memory-quarantine-for-slub-fix-fix.patch
nvme-dont-allocate-unused-nvme_major-fix.patch
bitmap-bitmap_equal-memcmp-optimization-fix.patch
kexec-ensure-user-memory-sizes-do-not-wrap-fix.patch
kexec-allow-architectures-to-override-boot-mapping-fix.patch
drivers-net-wireless-intel-iwlwifi-dvm-calibc-fix-min-warning.patch
kernel-forkc-export-kernel_thread-to-modules.patch
slab-leaks3-default-y.patch

Re: + thp-extract-khugepaged-from-mm-huge_memoryc-fix.patch added to -mm tree

[Kirill A. Shutemov]

On Wed, Jun 08, 2016 at 02:55:01PM -0700, akpm@linux-foundation.org wrote:
> 
> The patch titled
>      Subject: thp-extract-khugepaged-from-mm-huge_memoryc-fix
> has been added to the -mm tree.  Its filename is
>      thp-extract-khugepaged-from-mm-huge_memoryc-fix.patch
> 
> This patch should soon appear at
>     http://ozlabs.org/~akpm/mmots/broken-out/thp-extract-khugepaged-from-mm-huge_memoryc-fix.patch
> and later at
>     http://ozlabs.org/~akpm/mmotm/broken-out/thp-extract-khugepaged-from-mm-huge_memoryc-fix.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: Andrew Morton <akpm@linux-foundation.org>
> Subject: thp-extract-khugepaged-from-mm-huge_memoryc-fix
> 
> khugepaged_mutex can be static to mm/huge_memory.c:enabled_store()
> 
> Makes me wonder what it actually does...

IIUC, it protects khugepaged_thread.

What about something like this:

diff --git a/include/linux/khugepaged.h b/include/linux/khugepaged.h
index f7b4b7dc97bc..1e032a1ddb3e 100644
--- a/include/linux/khugepaged.h
+++ b/include/linux/khugepaged.h
@@ -4,7 +4,6 @@
 #include <linux/sched.h> /* MMF_VM_HUGEPAGE */
 
 #ifdef CONFIG_TRANSPARENT_HUGEPAGE
-extern struct mutex khugepaged_mutex;
 extern struct attribute_group khugepaged_attr_group;
 
 extern int khugepaged_init(void);
diff --git a/mm/huge_memory.c b/mm/huge_memory.c
index eaf3a4a655a6..522cf178f118 100644
--- a/mm/huge_memory.c
+++ b/mm/huge_memory.c
@@ -181,12 +181,7 @@ static ssize_t enabled_store(struct kobject *kobj,
 				TRANSPARENT_HUGEPAGE_REQ_MADV_FLAG);
 
 	if (ret > 0) {
-		int err;
-
-		mutex_lock(&khugepaged_mutex);
-		err = start_stop_khugepaged();
-		mutex_unlock(&khugepaged_mutex);
-
+		int err = start_stop_khugepaged();
 		if (err)
 			ret = err;
 	}
diff --git a/mm/khugepaged.c b/mm/khugepaged.c
index 71a6f9bedfdc..6f5815488c6c 100644
--- a/mm/khugepaged.c
+++ b/mm/khugepaged.c
@@ -58,8 +58,6 @@ static unsigned int khugepaged_scan_sleep_millisecs __read_mostly = 10000;
 /* during fragmentation poll the hugepage allocator once every minute */
 static unsigned int khugepaged_alloc_sleep_millisecs __read_mostly = 60000;
 static unsigned long khugepaged_sleep_expire;
-static struct task_struct *khugepaged_thread __read_mostly;
-DEFINE_MUTEX(khugepaged_mutex);
 static DEFINE_SPINLOCK(khugepaged_mm_lock);
 static DECLARE_WAIT_QUEUE_HEAD(khugepaged_wait);
 /*
@@ -1885,7 +1883,11 @@ static void set_recommended_min_free_kbytes(void)
 
 int start_stop_khugepaged(void)
 {
+	static struct task_struct *khugepaged_thread __read_mostly;
+	static DEFINE_MUTEX(khugepaged_mutex);
 	int err = 0;
+
+	mutex_lock(&khugepaged_mutex);
 	if (khugepaged_enabled()) {
 		if (!khugepaged_thread)
 			khugepaged_thread = kthread_run(khugepaged, NULL,
@@ -1906,5 +1908,6 @@ int start_stop_khugepaged(void)
 		khugepaged_thread = NULL;
 	}
 fail:
+	mutex_unlock(&khugepaged_mutex);
 	return err;
 }
-- 
 Kirill A. Shutemov