Skip to content

Commit

Permalink
hugetlb_cgroup: add accounting for shared mappings
Browse files Browse the repository at this point in the history 
For shared mappings, the pointer to the hugetlb_cgroup to uncharge lives
in the resv_map entries, in file_region->reservation_counter.

After a call to region_chg, we charge the approprate hugetlb_cgroup, and
if successful, we pass on the hugetlb_cgroup info to a follow up
region_add call.  When a file_region entry is added to the resv_map via
region_add, we put the pointer to that cgroup in
file_region->reservation_counter.  If charging doesn't succeed, we report
the error to the caller, so that the kernel fails the reservation.

On region_del, which is when the hugetlb memory is unreserved, we also
uncharge the file_region->reservation_counter.

[[email protected]: forward declare struct file_region]
Signed-off-by: Mina Almasry <[email protected]>
Signed-off-by: Andrew Morton <[email protected]>
Reviewed-by: Mike Kravetz <[email protected]>
Cc: David Rientjes <[email protected]>
Cc: Greg Thelen <[email protected]>
Cc: Mike Kravetz <[email protected]>
Cc: Sandipan Das <[email protected]>
Cc: Shakeel Butt <[email protected]>
Cc: Shuah Khan <[email protected]>
Link: http://lkml.kernel.org/r/[email protected]
Signed-off-by: Linus Torvalds <[email protected]>
  • Loading branch information
@mina @torvalds
mina authored and torvalds committed Apr 2, 2020
1 parent 0db9d74 commit 075a61d
Show file tree
Hide file tree
Showing 4 changed files with 155 additions and 54 deletions.
35 changes: 35 additions & 0 deletions include/linux/hugetlb.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -57,6 +57,41 @@ struct resv_map {
struct cgroup_subsys_state *css;
#endif
};

/*
* Region tracking -- allows tracking of reservations and instantiated pages
* across the pages in a mapping.
*
* The region data structures are embedded into a resv_map and protected
* by a resv_map's lock. The set of regions within the resv_map represent
* reservations for huge pages, or huge pages that have already been
* instantiated within the map. The from and to elements are huge page
* indicies into the associated mapping. from indicates the starting index
* of the region. to represents the first index past the end of the region.
*
* For example, a file region structure with from == 0 and to == 4 represents
* four huge pages in a mapping. It is important to note that the to element
* represents the first element past the end of the region. This is used in
* arithmetic as 4(to) - 0(from) = 4 huge pages in the region.
*
* Interval notation of the form [from, to) will be used to indicate that
* the endpoint from is inclusive and to is exclusive.
*/
struct file_region {
struct list_head link;
long from;
long to;
#ifdef CONFIG_CGROUP_HUGETLB
/*
* On shared mappings, each reserved region appears as a struct
* file_region in resv_map. These fields hold the info needed to
* uncharge each reservation.
*/
struct page_counter *reservation_counter;
struct cgroup_subsys_state *css;
#endif
};

extern struct resv_map *resv_map_alloc(void);
void resv_map_release(struct kref *ref);

Expand Down
11 changes: 11 additions & 0 deletions include/linux/hugetlb_cgroup.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -19,6 +19,7 @@

struct hugetlb_cgroup;
struct resv_map;
struct file_region;

/*
* Minimum page order trackable by hugetlb cgroup.
Expand Down Expand Up @@ -135,11 +136,21 @@ extern void hugetlb_cgroup_uncharge_counter(struct resv_map *resv,
unsigned long start,
unsigned long end);

extern void hugetlb_cgroup_uncharge_file_region(struct resv_map *resv,
struct file_region *rg,
unsigned long nr_pages);

extern void hugetlb_cgroup_file_init(void) __init;
extern void hugetlb_cgroup_migrate(struct page *oldhpage,
struct page *newhpage);

#else
static inline void hugetlb_cgroup_uncharge_file_region(struct resv_map *resv,
struct file_region *rg,
unsigned long nr_pages)
{
}

static inline struct hugetlb_cgroup *hugetlb_cgroup_from_page(struct page *page)
{
return NULL;
Expand Down
148 changes: 94 additions & 54 deletions mm/hugetlb.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -220,31 +220,6 @@ static inline struct hugepage_subpool *subpool_vma(struct vm_area_struct *vma)
return subpool_inode(file_inode(vma->vm_file));
}

/*
* Region tracking -- allows tracking of reservations and instantiated pages
* across the pages in a mapping.
*
* The region data structures are embedded into a resv_map and protected
* by a resv_map's lock. The set of regions within the resv_map represent
* reservations for huge pages, or huge pages that have already been
* instantiated within the map. The from and to elements are huge page
* indicies into the associated mapping. from indicates the starting index
* of the region. to represents the first index past the end of the region.
*
* For example, a file region structure with from == 0 and to == 4 represents
* four huge pages in a mapping. It is important to note that the to element
* represents the first element past the end of the region. This is used in
* arithmetic as 4(to) - 0(from) = 4 huge pages in the region.
*
* Interval notation of the form [from, to) will be used to indicate that
* the endpoint from is inclusive and to is exclusive.
*/
struct file_region {
struct list_head link;
long from;
long to;
};

/* Helper that removes a struct file_region from the resv_map cache and returns
* it for use.
*/
Expand All @@ -266,14 +241,51 @@ get_file_region_entry_from_cache(struct resv_map *resv, long from, long to)
return nrg;
}

static void copy_hugetlb_cgroup_uncharge_info(struct file_region *nrg,
struct file_region *rg)
{
#ifdef CONFIG_CGROUP_HUGETLB
nrg->reservation_counter = rg->reservation_counter;
nrg->css = rg->css;
if (rg->css)
css_get(rg->css);
#endif
}

/* Helper that records hugetlb_cgroup uncharge info. */
static void record_hugetlb_cgroup_uncharge_info(struct hugetlb_cgroup *h_cg,
struct hstate *h,
struct resv_map *resv,
struct file_region *nrg)
{
#ifdef CONFIG_CGROUP_HUGETLB
if (h_cg) {
nrg->reservation_counter =
&h_cg->rsvd_hugepage[hstate_index(h)];
nrg->css = &h_cg->css;
if (!resv->pages_per_hpage)
resv->pages_per_hpage = pages_per_huge_page(h);
/* pages_per_hpage should be the same for all entries in
* a resv_map.
*/
VM_BUG_ON(resv->pages_per_hpage != pages_per_huge_page(h));
} else {
nrg->reservation_counter = NULL;
nrg->css = NULL;
}
#endif
}

/* Must be called with resv->lock held. Calling this with count_only == true
* will count the number of pages to be added but will not modify the linked
* list. If regions_needed != NULL and count_only == true, then regions_needed
* will indicate the number of file_regions needed in the cache to carry out to
* add the regions for this range.
*/
static long add_reservation_in_range(struct resv_map *resv, long f, long t,
long *regions_needed, bool count_only)
struct hugetlb_cgroup *h_cg,
struct hstate *h, long *regions_needed,
bool count_only)
{
long add = 0;
struct list_head *head = &resv->regions;
Expand Down Expand Up @@ -312,6 +324,8 @@ static long add_reservation_in_range(struct resv_map *resv, long f, long t,
if (!count_only) {
nrg = get_file_region_entry_from_cache(
resv, last_accounted_offset, rg->from);
record_hugetlb_cgroup_uncharge_info(h_cg, h,
resv, nrg);
list_add(&nrg->link, rg->link.prev);
} else if (regions_needed)
*regions_needed += 1;
Expand All @@ -328,6 +342,7 @@ static long add_reservation_in_range(struct resv_map *resv, long f, long t,
if (!count_only) {
nrg = get_file_region_entry_from_cache(
resv, last_accounted_offset, t);
record_hugetlb_cgroup_uncharge_info(h_cg, h, resv, nrg);
list_add(&nrg->link, rg->link.prev);
} else if (regions_needed)
*regions_needed += 1;
Expand Down Expand Up @@ -416,15 +431,17 @@ static int allocate_file_region_entries(struct resv_map *resv,
* 1 page will only require at most 1 entry.
*/
static long region_add(struct resv_map *resv, long f, long t,
long in_regions_needed)
long in_regions_needed, struct hstate *h,
struct hugetlb_cgroup *h_cg)
{
long add = 0, actual_regions_needed = 0;

spin_lock(&resv->lock);
retry:

/* Count how many regions are actually needed to execute this add. */
add_reservation_in_range(resv, f, t, &actual_regions_needed, true);
add_reservation_in_range(resv, f, t, NULL, NULL, &actual_regions_needed,
true);

/*
* Check for sufficient descriptors in the cache to accommodate
Expand Down Expand Up @@ -452,7 +469,7 @@ static long region_add(struct resv_map *resv, long f, long t,
goto retry;
}

add = add_reservation_in_range(resv, f, t, NULL, false);
add = add_reservation_in_range(resv, f, t, h_cg, h, NULL, false);

resv->adds_in_progress -= in_regions_needed;

Expand Down Expand Up @@ -489,7 +506,8 @@ static long region_chg(struct resv_map *resv, long f, long t,
spin_lock(&resv->lock);

/* Count how many hugepages in this range are NOT respresented. */
chg = add_reservation_in_range(resv, f, t, out_regions_needed, true);
chg = add_reservation_in_range(resv, f, t, NULL, NULL,
out_regions_needed, true);

if (*out_regions_needed == 0)
*out_regions_needed = 1;
Expand Down Expand Up @@ -589,18 +607,26 @@ static long region_del(struct resv_map *resv, long f, long t)
/* New entry for end of split region */
nrg->from = t;
nrg->to = rg->to;

copy_hugetlb_cgroup_uncharge_info(nrg, rg);

INIT_LIST_HEAD(&nrg->link);

/* Original entry is trimmed */
rg->to = f;

hugetlb_cgroup_uncharge_file_region(
resv, rg, nrg->to - nrg->from);

list_add(&nrg->link, &rg->link);
nrg = NULL;
break;
}

if (f <= rg->from && t >= rg->to) { /* Remove entire region */
del += rg->to - rg->from;
hugetlb_cgroup_uncharge_file_region(resv, rg,
rg->to - rg->from);
list_del(&rg->link);
kfree(rg);
continue;
Expand All @@ -609,9 +635,15 @@ static long region_del(struct resv_map *resv, long f, long t)
if (f <= rg->from) { /* Trim beginning of region */
del += t - rg->from;
rg->from = t;

hugetlb_cgroup_uncharge_file_region(resv, rg,
t - rg->from);
} else { /* Trim end of region */
del += rg->to - f;
rg->to = f;

hugetlb_cgroup_uncharge_file_region(resv, rg,
rg->to - f);
}
}

Expand Down Expand Up @@ -2124,7 +2156,7 @@ static long __vma_reservation_common(struct hstate *h,
VM_BUG_ON(dummy_out_regions_needed != 1);
break;
case VMA_COMMIT_RESV:
ret = region_add(resv, idx, idx + 1, 1);
ret = region_add(resv, idx, idx + 1, 1, NULL, NULL);
/* region_add calls of range 1 should never fail. */
VM_BUG_ON(ret < 0);
break;
Expand All @@ -2134,7 +2166,7 @@ static long __vma_reservation_common(struct hstate *h,
break;
case VMA_ADD_RESV:
if (vma->vm_flags & VM_MAYSHARE) {
ret = region_add(resv, idx, idx + 1, 1);
ret = region_add(resv, idx, idx + 1, 1, NULL, NULL);
/* region_add calls of range 1 should never fail. */
VM_BUG_ON(ret < 0);
} else {
Expand Down Expand Up @@ -4830,7 +4862,7 @@ int hugetlb_reserve_pages(struct inode *inode,
struct hstate *h = hstate_inode(inode);
struct hugepage_subpool *spool = subpool_inode(inode);
struct resv_map *resv_map;
struct hugetlb_cgroup *h_cg;
struct hugetlb_cgroup *h_cg = NULL;
long gbl_reserve, regions_needed = 0;

/* This should never happen */
Expand Down Expand Up @@ -4871,19 +4903,6 @@ int hugetlb_reserve_pages(struct inode *inode,

chg = to - from;

if (hugetlb_cgroup_charge_cgroup_rsvd(
hstate_index(h), chg * pages_per_huge_page(h),
&h_cg)) {
kref_put(&resv_map->refs, resv_map_release);
return -ENOMEM;
}

/*
* Since this branch handles private mappings, we attach the
* counter to uncharge for this reservation off resv_map.
*/
resv_map_set_hugetlb_cgroup_uncharge_info(resv_map, h_cg, h);

set_vma_resv_map(vma, resv_map);
set_vma_resv_flags(vma, HPAGE_RESV_OWNER);
}
Expand All @@ -4893,6 +4912,21 @@ int hugetlb_reserve_pages(struct inode *inode,
goto out_err;
}

ret = hugetlb_cgroup_charge_cgroup_rsvd(
hstate_index(h), chg * pages_per_huge_page(h), &h_cg);

if (ret < 0) {
ret = -ENOMEM;
goto out_err;
}

if (vma && !(vma->vm_flags & VM_MAYSHARE) && h_cg) {
/* For private mappings, the hugetlb_cgroup uncharge info hangs
* of the resv_map.
*/
resv_map_set_hugetlb_cgroup_uncharge_info(resv_map, h_cg, h);
}

/*
* There must be enough pages in the subpool for the mapping. If
* the subpool has a minimum size, there may be some global
Expand All @@ -4901,7 +4935,7 @@ int hugetlb_reserve_pages(struct inode *inode,
gbl_reserve = hugepage_subpool_get_pages(spool, chg);
if (gbl_reserve < 0) {
ret = -ENOSPC;
goto out_err;
goto out_uncharge_cgroup;
}

/*
Expand All @@ -4910,9 +4944,7 @@ int hugetlb_reserve_pages(struct inode *inode,
*/
ret = hugetlb_acct_memory(h, gbl_reserve);
if (ret < 0) {
/* put back original number of pages, chg */
(void)hugepage_subpool_put_pages(spool, chg);
goto out_err;
goto out_put_pages;
}

/*
Expand All @@ -4927,13 +4959,11 @@ int hugetlb_reserve_pages(struct inode *inode,
* else has to be done for private mappings here
*/
if (!vma || vma->vm_flags & VM_MAYSHARE) {
add = region_add(resv_map, from, to, regions_needed);
add = region_add(resv_map, from, to, regions_needed, h, h_cg);

if (unlikely(add < 0)) {
hugetlb_acct_memory(h, -gbl_reserve);
/* put back original number of pages, chg */
(void)hugepage_subpool_put_pages(spool, chg);
goto out_err;
goto out_put_pages;
} else if (unlikely(chg > add)) {
/*
* pages in this range were added to the reserve
Expand All @@ -4944,12 +4974,22 @@ int hugetlb_reserve_pages(struct inode *inode,
*/
long rsv_adjust;

hugetlb_cgroup_uncharge_cgroup_rsvd(
hstate_index(h),
(chg - add) * pages_per_huge_page(h), h_cg);

rsv_adjust = hugepage_subpool_put_pages(spool,
chg - add);
hugetlb_acct_memory(h, -rsv_adjust);
}
}
return 0;
out_put_pages:
/* put back original number of pages, chg */
(void)hugepage_subpool_put_pages(spool, chg);
out_uncharge_cgroup:
hugetlb_cgroup_uncharge_cgroup_rsvd(hstate_index(h),
chg * pages_per_huge_page(h), h_cg);
out_err:
if (!vma || vma->vm_flags & VM_MAYSHARE)
/* Only call region_abort if the region_chg succeeded but the
Expand Down
Loading

0 comments on commit 075a61d

Please sign in to comment.