The patch titled
Subject: mm: swap: unify cluster-based and vma-based swap readahead
has been added to the -mm tree. Its filename is
mm-swap-unify-cluster-based-and-vma-based-swap-readahead.patch
This patch should soon appear at
http://ozlabs.org/~akpm/mmots/broken-out/mm-swap-unify-cluster-based-and-vma-based-swap-readahead.patch
and later at
http://ozlabs.org/~akpm/mmotm/broken-out/mm-swap-unify-cluster-based-and-vma-based-swap-readahead.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: Minchan Kim <minchan@xxxxxxxxxx>
Subject: mm: swap: unify cluster-based and vma-based swap readahead
This patch makes do_swap_page() not need to be aware of two different swap
readahead algorithms. Just unify cluster-based and vma-based readahead
function call.
Link: http://lkml.kernel.org/r/1509520520-32367-3-git-send-email-minchan@xxxxxxxxxx
Signed-off-by: Minchan Kim <minchan@xxxxxxxxxx>
Cc: Hugh Dickins <hughd@xxxxxxxxxx>
Cc: Huang Ying <ying.huang@xxxxxxxxx>
Signed-off-by: Andrew Morton <akpm@xxxxxxxxxxxxxxxxxxxx>
---
include/linux/swap.h | 27 ++++++-----------------
mm/memory.c | 11 +++------
mm/shmem.c | 5 +++-
mm/swap_state.c | 48 +++++++++++++++++++++++++++++++----------
4 files changed, 53 insertions(+), 38 deletions(-)
diff -puN include/linux/swap.h~mm-swap-unify-cluster-based-and-vma-based-swap-readahead include/linux/swap.h
--- a/include/linux/swap.h~mm-swap-unify-cluster-based-and-vma-based-swap-readahead
+++ a/include/linux/swap.h
@@ -401,7 +401,6 @@ int generic_swapfile_activate(struct swa
#define SWAP_ADDRESS_SPACE_SHIFT 14
#define SWAP_ADDRESS_SPACE_PAGES (1 << SWAP_ADDRESS_SPACE_SHIFT)
extern struct address_space *swapper_spaces[];
-extern bool swap_vma_readahead;
#define swap_address_space(entry) \
(&swapper_spaces[swp_type(entry)][swp_offset(entry) \
>> SWAP_ADDRESS_SPACE_SHIFT])
@@ -423,10 +422,10 @@ extern struct page *read_swap_cache_asyn
extern struct page *__read_swap_cache_async(swp_entry_t, gfp_t,
struct vm_area_struct *vma, unsigned long addr,
bool *new_page_allocated);
-extern struct page *swapin_readahead(swp_entry_t, gfp_t,
- struct vm_area_struct *vma, unsigned long addr);
-extern struct page *do_swap_page_readahead(swp_entry_t fentry, gfp_t gfp_mask,
- struct vm_fault *vmf);
+extern struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t flag,
+ struct vm_fault *vmf);
+extern struct page *swapin_readahead(swp_entry_t entry, gfp_t flag,
+ struct vm_fault *vmf);
/* linux/mm/swapfile.c */
extern atomic_long_t nr_swap_pages;
@@ -434,11 +433,6 @@ extern long total_swap_pages;
extern atomic_t nr_rotate_swap;
extern bool has_usable_swap(void);
-static inline bool swap_use_vma_readahead(void)
-{
- return READ_ONCE(swap_vma_readahead) && !atomic_read(&nr_rotate_swap);
-}
-
/* Swap 50% full? Release swapcache more aggressively.. */
static inline bool vm_swap_full(void)
{
@@ -534,19 +528,14 @@ static inline void put_swap_page(struct
{
}
-static inline struct page *swapin_readahead(swp_entry_t swp, gfp_t gfp_mask,
- struct vm_area_struct *vma, unsigned long addr)
+static inline struct page *swap_cluster_readahead(swp_entry_t entry,
+ gfp_t gfp_mask, struct vm_fault *vmf)
{
return NULL;
}
-static inline bool swap_use_vma_readahead(void)
-{
- return false;
-}
-
-static inline struct page *do_swap_page_readahead(swp_entry_t fentry,
- gfp_t gfp_mask, struct vm_fault *vmf)
+static inline struct page *swapin_readahead(swp_entry_t swp, gfp_t gfp_mask,
+ struct vm_fault *vmf)
{
return NULL;
}
diff -puN mm/memory.c~mm-swap-unify-cluster-based-and-vma-based-swap-readahead mm/memory.c
--- a/mm/memory.c~mm-swap-unify-cluster-based-and-vma-based-swap-readahead
+++ a/mm/memory.c
@@ -2889,7 +2889,8 @@ int do_swap_page(struct vm_fault *vmf)
if (si->flags & SWP_SYNCHRONOUS_IO &&
__swap_count(si, entry) == 1) {
/* skip swapcache */
- page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma, vmf->address);
+ page = alloc_page_vma(GFP_HIGHUSER_MOVABLE, vma,
+ vmf->address);
if (page) {
__SetPageLocked(page);
__SetPageSwapBacked(page);
@@ -2898,12 +2899,8 @@ int do_swap_page(struct vm_fault *vmf)
swap_readpage(page, true);
}
} else {
- if (swap_use_vma_readahead())
- page = do_swap_page_readahead(entry,
- GFP_HIGHUSER_MOVABLE, vmf);
- else
- page = swapin_readahead(entry,
- GFP_HIGHUSER_MOVABLE, vma, vmf->address);
+ page = swapin_readahead(entry, GFP_HIGHUSER_MOVABLE,
+ vmf);
swapcache = page;
}
diff -puN mm/shmem.c~mm-swap-unify-cluster-based-and-vma-based-swap-readahead mm/shmem.c
--- a/mm/shmem.c~mm-swap-unify-cluster-based-and-vma-based-swap-readahead
+++ a/mm/shmem.c
@@ -1413,9 +1413,12 @@ static struct page *shmem_swapin(swp_ent
{
struct vm_area_struct pvma;
struct page *page;
+ struct vm_fault vmf;
shmem_pseudo_vma_init(&pvma, info, index);
- page = swapin_readahead(swap, gfp, &pvma, 0);
+ vmf.vma = &pvma;
+ vmf.address = 0;
+ page = swap_cluster_readahead(swap, gfp, &vmf);
shmem_pseudo_vma_destroy(&pvma);
return page;
diff -puN mm/swap_state.c~mm-swap-unify-cluster-based-and-vma-based-swap-readahead mm/swap_state.c
--- a/mm/swap_state.c~mm-swap-unify-cluster-based-and-vma-based-swap-readahead
+++ a/mm/swap_state.c
@@ -38,7 +38,7 @@ static const struct address_space_operat
struct address_space *swapper_spaces[MAX_SWAPFILES] __read_mostly;
static unsigned int nr_swapper_spaces[MAX_SWAPFILES] __read_mostly;
-bool swap_vma_readahead __read_mostly = true;
+bool enable_vma_readahead __read_mostly = true;
#define SWAP_RA_WIN_SHIFT (PAGE_SHIFT / 2)
#define SWAP_RA_HITS_MASK ((1UL << SWAP_RA_WIN_SHIFT) - 1)
@@ -322,6 +322,11 @@ void free_pages_and_swap_cache(struct pa
release_pages(pagep, nr);
}
+static inline bool swap_use_vma_readahead(void)
+{
+ return READ_ONCE(enable_vma_readahead) && !atomic_read(&nr_rotate_swap);
+}
+
/*
* Lookup a swap entry in the swap cache. A found page will be returned
* unlocked and with its refcount incremented - we rely on the kernel
@@ -539,11 +544,10 @@ static unsigned long swapin_nr_pages(uns
}
/**
- * swapin_readahead - swap in pages in hope we need them soon
+ * swap_cluster_readahead - swap in pages in hope we need them soon
* @entry: swap entry of this memory
* @gfp_mask: memory allocation flags
- * @vma: user vma this address belongs to
- * @addr: target address for mempolicy
+ * @vmf: fault information
*
* Returns the struct page for entry and addr, after queueing swapin.
*
@@ -555,10 +559,10 @@ static unsigned long swapin_nr_pages(uns
* This has been extended to use the NUMA policies from the mm triggering
* the readahead.
*
- * Caller must hold down_read on the vma->vm_mm if vma is not NULL.
+ * Caller must hold down_read on the vma->vm_mm if vmf->vma is not NULL.
*/
-struct page *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask,
- struct vm_area_struct *vma, unsigned long addr)
+struct page *swap_cluster_readahead(swp_entry_t entry, gfp_t gfp_mask,
+ struct vm_fault *vmf)
{
struct page *page;
unsigned long entry_offset = swp_offset(entry);
@@ -568,6 +572,8 @@ struct page *swapin_readahead(swp_entry_
struct swap_info_struct *si = swp_swap_info(entry);
struct blk_plug plug;
bool do_poll = true, page_allocated;
+ struct vm_area_struct *vma = vmf->vma;
+ unsigned long addr = vmf->address;
mask = swapin_nr_pages(offset) - 1;
if (!mask)
@@ -723,7 +729,7 @@ static void swap_ra_info(struct vm_fault
pte_unmap(orig_pte);
}
-struct page *do_swap_page_readahead(swp_entry_t fentry, gfp_t gfp_mask,
+struct page *swap_vma_readahead(swp_entry_t fentry, gfp_t gfp_mask,
struct vm_fault *vmf)
{
struct blk_plug plug;
@@ -771,20 +777,40 @@ skip:
ra_info.win == 1);
}
+/**
+ * swapin_readahead - swap in pages in hope we need them soon
+ * @entry: swap entry of this memory
+ * @gfp_mask: memory allocation flags
+ * @vmf: fault information
+ *
+ * Returns the struct page for entry and addr, after queueing swapin.
+ *
+ * It's a main entry function for swap readahead. By the configuration,
+ * it will read ahead blocks by cluster-based(ie, physical disk based)
+ * or vma-based(ie, virtual address based on faulty address) readahead.
+ */
+struct page *swapin_readahead(swp_entry_t entry, gfp_t gfp_mask,
+ struct vm_fault *vmf)
+{
+ return swap_use_vma_readahead() ?
+ swap_vma_readahead(entry, gfp_mask, vmf) :
+ swap_cluster_readahead(entry, gfp_mask, vmf);
+}
+
#ifdef CONFIG_SYSFS
static ssize_t vma_ra_enabled_show(struct kobject *kobj,
struct kobj_attribute *attr, char *buf)
{
- return sprintf(buf, "%s\n", swap_vma_readahead ? "true" : "false");
+ return sprintf(buf, "%s\n", enable_vma_readahead ? "true" : "false");
}
static ssize_t vma_ra_enabled_store(struct kobject *kobj,
struct kobj_attribute *attr,
const char *buf, size_t count)
{
if (!strncmp(buf, "true", 4) || !strncmp(buf, "1", 1))
- swap_vma_readahead = true;
+ enable_vma_readahead = true;
else if (!strncmp(buf, "false", 5) || !strncmp(buf, "0", 1))
- swap_vma_readahead = false;
+ enable_vma_readahead = false;
else
return -EINVAL;
_
Patches currently in -mm which might be from minchan@xxxxxxxxxx are
mm-swap-clean-up-swap-readahead.patch
mm-swap-unify-cluster-based-and-vma-based-swap-readahead.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
