The patch titled
Subject: thp: introduce khugepaged_prealloc_page and khugepaged_alloc_page
has been added to the -mm tree. Its filename is
thp-introduce-khugepaged_prealloc_page-and-khugepaged_alloc_page.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: Xiao Guangrong <xiaoguangrong@xxxxxxxxxxxxxxxxxx>
Subject: thp: introduce khugepaged_prealloc_page and khugepaged_alloc_page
They are used to abstract the difference between NUMA enabled and NUMA
disabled to make the code more readable
Signed-off-by: Xiao Guangrong <xiaoguangrong@xxxxxxxxxxxxxxxxxx>
Cc: Andrea Arcangeli <aarcange@xxxxxxxxxx>
Cc: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: David Rientjes <rientjes@xxxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
mm/huge_memory.c | 166 ++++++++++++++++++++++++++-------------------
1 file changed, 98 insertions(+), 68 deletions(-)
diff -puN mm/huge_memory.c~thp-introduce-khugepaged_prealloc_page-and-khugepaged_alloc_page mm/huge_memory.c
--- a/mm/huge_memory.c~thp-introduce-khugepaged_prealloc_page-and-khugepaged_alloc_page
+++ a/mm/huge_memory.c
@@ -1828,28 +1828,34 @@ static void __collapse_huge_page_copy(pt
}
}
-static void collapse_huge_page(struct mm_struct *mm,
- unsigned long address,
- struct page **hpage,
- struct vm_area_struct *vma,
- int node)
+static void khugepaged_alloc_sleep(void)
{
- pgd_t *pgd;
- pud_t *pud;
- pmd_t *pmd, _pmd;
- pte_t *pte;
- pgtable_t pgtable;
- struct page *new_page;
- spinlock_t *ptl;
- int isolated;
- unsigned long hstart, hend;
+ wait_event_freezable_timeout(khugepaged_wait, false,
+ msecs_to_jiffies(khugepaged_alloc_sleep_millisecs));
+}
- VM_BUG_ON(address & ~HPAGE_PMD_MASK);
-#ifndef CONFIG_NUMA
- up_read(&mm->mmap_sem);
- VM_BUG_ON(!*hpage);
- new_page = *hpage;
-#else
+#ifdef CONFIG_NUMA
+static bool khugepaged_prealloc_page(struct page **hpage, bool *wait)
+{
+ if (IS_ERR(*hpage)) {
+ if (!*wait)
+ return false;
+
+ *wait = false;
+ khugepaged_alloc_sleep();
+ } else if (*hpage) {
+ put_page(*hpage);
+ *hpage = NULL;
+ }
+
+ return true;
+}
+
+static struct page
+*khugepaged_alloc_page(struct page **hpage, struct mm_struct *mm,
+ struct vm_area_struct *vma, unsigned long address,
+ int node)
+{
VM_BUG_ON(*hpage);
/*
* Allocate the page while the vma is still valid and under
@@ -1861,7 +1867,7 @@ static void collapse_huge_page(struct mm
* mmap_sem in read mode is good idea also to allow greater
* scalability.
*/
- new_page = alloc_hugepage_vma(khugepaged_defrag(), vma, address,
+ *hpage = alloc_hugepage_vma(khugepaged_defrag(), vma, address,
node, __GFP_OTHER_NODE);
/*
@@ -1869,15 +1875,81 @@ static void collapse_huge_page(struct mm
* preparation for taking it in write mode.
*/
up_read(&mm->mmap_sem);
- if (unlikely(!new_page)) {
+ if (unlikely(!*hpage)) {
count_vm_event(THP_COLLAPSE_ALLOC_FAILED);
*hpage = ERR_PTR(-ENOMEM);
- return;
+ return NULL;
}
- *hpage = new_page;
+
count_vm_event(THP_COLLAPSE_ALLOC);
+ return *hpage;
+}
+#else
+static struct page *khugepaged_alloc_hugepage(bool *wait)
+{
+ struct page *hpage;
+
+ do {
+ hpage = alloc_hugepage(khugepaged_defrag());
+ if (!hpage) {
+ count_vm_event(THP_COLLAPSE_ALLOC_FAILED);
+ if (!*wait)
+ return NULL;
+
+ *wait = false;
+ khugepaged_alloc_sleep();
+ } else
+ count_vm_event(THP_COLLAPSE_ALLOC);
+ } while (unlikely(!hpage) && likely(khugepaged_enabled()));
+
+ return hpage;
+}
+
+static bool khugepaged_prealloc_page(struct page **hpage, bool *wait)
+{
+ if (!*hpage)
+ *hpage = khugepaged_alloc_hugepage(wait);
+
+ if (unlikely(!*hpage))
+ return false;
+
+ return true;
+}
+
+static struct page
+*khugepaged_alloc_page(struct page **hpage, struct mm_struct *mm,
+ struct vm_area_struct *vma, unsigned long address,
+ int node)
+{
+ up_read(&mm->mmap_sem);
+ VM_BUG_ON(!*hpage);
+ return *hpage;
+}
#endif
+static void collapse_huge_page(struct mm_struct *mm,
+ unsigned long address,
+ struct page **hpage,
+ struct vm_area_struct *vma,
+ int node)
+{
+ pgd_t *pgd;
+ pud_t *pud;
+ pmd_t *pmd, _pmd;
+ pte_t *pte;
+ pgtable_t pgtable;
+ struct page *new_page;
+ spinlock_t *ptl;
+ int isolated;
+ unsigned long hstart, hend;
+
+ VM_BUG_ON(address & ~HPAGE_PMD_MASK);
+
+ /* release the mmap_sem read lock. */
+ new_page = khugepaged_alloc_page(hpage, mm, vma, address, node);
+ if (!new_page)
+ return;
+
if (unlikely(mem_cgroup_newpage_charge(new_page, mm, GFP_KERNEL)))
return;
@@ -2216,34 +2288,6 @@ static int khugepaged_wait_event(void)
kthread_should_stop();
}
-static void khugepaged_alloc_sleep(void)
-{
- wait_event_freezable_timeout(khugepaged_wait, false,
- msecs_to_jiffies(khugepaged_alloc_sleep_millisecs));
-}
-
-#ifndef CONFIG_NUMA
-static struct page *khugepaged_alloc_hugepage(bool *wait)
-{
- struct page *hpage;
-
- do {
- hpage = alloc_hugepage(khugepaged_defrag());
- if (!hpage) {
- count_vm_event(THP_COLLAPSE_ALLOC_FAILED);
- if (!*wait)
- return NULL;
-
- *wait = false;
- khugepaged_alloc_sleep();
- } else
- count_vm_event(THP_COLLAPSE_ALLOC);
- } while (unlikely(!hpage) && likely(khugepaged_enabled()));
-
- return hpage;
-}
-#endif
-
static void khugepaged_do_scan(void)
{
struct page *hpage = NULL;
@@ -2254,23 +2298,9 @@ static void khugepaged_do_scan(void)
barrier(); /* write khugepaged_pages_to_scan to local stack */
while (progress < pages) {
-#ifndef CONFIG_NUMA
- if (!hpage)
- hpage = khugepaged_alloc_hugepage(&wait);
-
- if (unlikely(!hpage))
+ if (!khugepaged_prealloc_page(&hpage, &wait))
break;
-#else
- if (IS_ERR(hpage)) {
- if (!wait)
- break;
- wait = false;
- khugepaged_alloc_sleep();
- } else if (hpage) {
- put_page(hpage);
- hpage = NULL;
- }
-#endif
+
cond_resched();
if (unlikely(kthread_should_stop() || freezing(current)))
_
Patches currently in -mm which might be from xiaoguangrong@xxxxxxxxxxxxxxxxxx are
linux-next.patch
thp-fix-the-count-of-thp_collapse_alloc.patch
thp-remove-unnecessary-check-in-start_khugepaged.patch
thp-move-khugepaged_mutex-out-of-khugepaged.patch
thp-remove-unnecessary-khugepaged_thread-check.patch
thp-remove-wake_up_interruptible-in-the-exit-path.patch
thp-remove-some-code-depend-on-config_numa.patch
thp-merge-page-pre-alloc-in-khugepaged_loop-into-khugepaged_do_scan.patch
thp-release-page-in-page-pre-alloc-path.patch
thp-introduce-khugepaged_prealloc_page-and-khugepaged_alloc_page.patch
thp-remove-khugepaged_loop.patch
thp-use-khugepaged_enabled-to-remove-duplicate-code.patch
thp-remove-unnecessary-set_recommended_min_free_kbytes.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
