From: Ackerley Tng HugeTLB subpools currently only track used pages (used_hpages) when a maximum size limit (max_hpages) is configured. This breaks minimum size (min_hpages) guarantees. The subpool guarantee requires that the sum of used pages and remaining subpool reserves satisfies the configured minimum: used_hpages + rsv_hpages >= min_hpages where used_hpages includes both allocated folios and active reservations. Therefore, when pages or reservations are released, the maximum number of reservations the subpool can absorb is: limit = max(0, min_hpages - used_hpages) When used_hpages is untracked and remains zero, hugepage_subpool_put_pages() assumes limit == min_hpages. It falsely restores reservations to the subpool even when existing allocations already satisfy the guarantee. This false restoration leads to: 1. Leaked global reservations: resv_huge_pages is not decremented, reducing the huge pages available for use across the host. 2. Premature freeing: the subpool considers itself unreferenced on unmount (rsv_hpages == min_hpages), freeing the subpool while folios are still active in page cache and leading to a use-after-free. Fix this by unconditionally tracking used_hpages in the subpool, and capping the reservations the subpool can absorb to (min_hpages - used_hpages). With used_hpages always tracked: - subpool_is_free() is simplified to check whether used_hpages is zero. - hugetlbfs_statfs() is updated to explicitly check for max_hpages == -1; previously it relied on (-1 - 0) evaluating to -1, which would now underflow and report negative free block counts. Trace of a false restoration: 1. Mount time: + spool->min_hpages = 1 (user requested min_size=2M) + spool->max_hpages = -1 (no maximum size specified) + spool->rsv_hpages = 1 (reserve min_hpages) + spool->used_hpages = 0 (untracked when max_hpages == -1) + h->resv_huge_pages = 1 (reserved by hugetlb_acct_memory(h, 1)) 2. Shared mapping of 4MB (2 pages) created (mmap with MAP_SHARED): + In hugetlb_reserve_pages(), region_chg() finds chg = 2 (pages 0 and 1 need reservations) + hugepage_subpool_get_pages(spool, 2): + spool->rsv_hpages = 0 (consumed the 1 subpool reservation) + Returns 1 (since this subpool only had 1 reservation) + hugetlb_acct_memory(h, 1): + h->resv_huge_pages = 2 (incremented from 1 to 2 for the global reservation) + region_add() records reservations for pages 0 and 1 in resv_map 3. Process touches and populates Page 0: + hugetlb_no_page() calls alloc_hugetlb_folio() + Page 0 reuses the existing reservation (vma_needs_reservation() returns 0 => map_chg = MAP_CHG_REUSE = 0) + hugepage_subpool_get_pages() is not called (map_chg == 0) + dequeue_hugetlb_folio_nodemask() consumes 1 reservation: h->resv_huge_pages = 1 (decremented from 2 to 1) 4. Process closes the file and exits: + For MAP_SHARED mappings, reservations persist in inode resv_map + Page 1 reservation remains active + h->resv_huge_pages = 1 (retained for Page 1) 5. File is truncated to 2MB (truncate -s 2M): + Truncation invokes remove_inode_hugepages() for range [1, LONG_MAX) + Page 1 was never faulted into page cache => freed = 0 + Calls hugetlb_unreserve_pages(inode, 1, LONG_MAX, freed = 0) + region_del() removes Page 1 from resv_map => chg = 1 6. Inside hugetlb_unreserve_pages(): hugepage_subpool_put_pages(1): + delta = chg - freed = 1 - 0 = 1 + Because spool->max_hpages == -1, spool->used_hpages always = 0 + spool->used_hpages < spool->min_hpages (0 < 1 => true) <<== subpool assumes 0 pages are in use, ignoring allocated Page 0 + spool->rsv_hpages + delta <= spool->min_hpages (0 + 1 <= 1 => true) + spool->rsv_hpages += 1 => spool->rsv_hpages = 1 <<== false reservation restored to subpool + Returns 0 (subpool absorbed the reservation) 7. Back in hugetlb_unreserve_pages(): hugetlb_acct_memory(): + hugetlb_acct_memory(h, -0) does nothing + h->resv_huge_pages remains stuck at 1 even though both reservations have ended (Page 0 allocated, Page 1 truncated) 8. Later during unmounting: + subpool_is_free() checks spool->rsv_hpages == spool->min_hpages (1 == 1 => true) + The subpool is erroneously considered completely free + hugetlb_acct_memory(spool->hstate, -spool->min_hpages) decrements h->resv_huge_pages by 1 (1 - 1 = 0), masking the leak on unmount 9. If the folio outlives the inode: + When Page 0 is freed, free_huge_folio() attempts to access the freed subpool pointer, leading to a use-after-free and double-free. Fixes: 09a95e29cb30 ("mm/hugetlb: optimize minimum size (min_size) accounting") Fixes: 1c5ecae3a93fa ("hugetlbfs: add minimum size accounting to subpools") Signed-off-by: Ackerley Tng Reviewed-by: Joshua Hahn Cc: stable@vger.kernel.org --- Documentation/mm/hugetlbfs_reserv.rst | 17 +---- .../translations/zh_CN/mm/hugetlbfs_reserv.rst | 11 +--- fs/hugetlbfs/inode.c | 8 ++- include/linux/hugetlb.h | 4 +- mm/hugetlb.c | 73 +++++++++++++--------- 5 files changed, 55 insertions(+), 58 deletions(-) diff --git a/Documentation/mm/hugetlbfs_reserv.rst b/Documentation/mm/hugetlbfs_reserv.rst index a49115db18c76..d244583fdcbc3 100644 --- a/Documentation/mm/hugetlbfs_reserv.rst +++ b/Documentation/mm/hugetlbfs_reserv.rst @@ -314,21 +314,8 @@ huge pages. If they can not be reserved, the mount fails. The routines hugepage_subpool_get/put_pages() are called when pages are obtained from or released back to a subpool. They perform all subpool accounting, and track any reservations associated with the subpool. -hugepage_subpool_get/put_pages are passed the number of huge pages by which -to adjust the subpool 'used page' count (down for get, up for put). Normally, -they return the same value that was passed or an error if not enough pages -exist in the subpool. - -However, if reserves are associated with the subpool a return value less -than the passed value may be returned. This return value indicates the -number of additional global pool adjustments which must be made. For example, -suppose a subpool contains 3 reserved huge pages and someone asks for 5. -The 3 reserved pages associated with the subpool can be used to satisfy part -of the request. But, 2 pages must be obtained from the global pools. To -relay this information to the caller, the value 2 is returned. The caller -is then responsible for attempting to obtain the additional two pages from -the global pools. - +hugepage_subpool_get/put_pages() use the number of huge pages passed to adjust +the subpool 'used page' count. COW and Reservations ==================== diff --git a/Documentation/translations/zh_CN/mm/hugetlbfs_reserv.rst b/Documentation/translations/zh_CN/mm/hugetlbfs_reserv.rst index 20947f8bd0654..ae1f1f31477fc 100644 --- a/Documentation/translations/zh_CN/mm/hugetlbfs_reserv.rst +++ b/Documentation/translations/zh_CN/mm/hugetlbfs_reserv.rst @@ -246,15 +246,8 @@ hugepage_subpool的min_hpages字段中被跟踪。在挂载时，hugetlb_acct_me 被调用以预留指定数量的巨页。如果它们不能被预留，挂载就会失败。 当从子池中获取或释放页面时，会调用hugepage_subpool_get/put_pages()函数。 -hugepage_subpool_get/put_pages被传递给巨页数量，以此来调整子池的 “已用页面” 计数 -（get为下降，put为上升）。通常情况下，如果子池中没有足够的页面，它们会返回与传递的相同的值或 -一个错误。 - -然而，如果预留与子池相关联，可能会返回一个小于传递值的返回值。这个返回值表示必须进行的额外全局 -池调整的数量。例如，假设一个子池包含3个预留的巨页，有人要求5个。与子池相关的3个预留页可以用来 -满足部分请求。但是，必须从全局池中获得2个页面。为了向调用者转达这一信息，将返回值2。然后，调用 -者要负责从全局池中获取另外两个页面。 - +它们负责所有子池的统计核算，并跟踪与子池相关联的预留。 +hugepage_subpool_get/put_pages()函数使用传入的巨页数量来调整子池的“已用页面”计数。 COW和预留 ========== diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index 7611a8470ea26..5113f743f6fc7 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -1109,8 +1109,12 @@ static int hugetlbfs_statfs(struct dentry *dentry, struct kstatfs *buf) spin_lock_irq(&sbinfo->spool->lock); buf->f_blocks = sbinfo->spool->max_hpages; - free_pages = sbinfo->spool->max_hpages - - sbinfo->spool->used_hpages; + if (sbinfo->spool->max_hpages == -1) { + free_pages = -1; + } else { + free_pages = sbinfo->spool->max_hpages - + sbinfo->spool->used_hpages; + } buf->f_bavail = buf->f_bfree = free_pages; spin_unlock_irq(&sbinfo->spool->lock); buf->f_files = sbinfo->max_inodes; diff --git a/include/linux/hugetlb.h b/include/linux/hugetlb.h index 16c4c4caa126c..4551ff3023640 100644 --- a/include/linux/hugetlb.h +++ b/include/linux/hugetlb.h @@ -39,8 +39,8 @@ struct hugepage_subpool { spinlock_t lock; long count; long max_hpages; /* Maximum huge pages or -1 if no maximum. */ - long used_hpages; /* Used count against maximum, includes */ - /* both allocated and reserved pages. */ + long used_hpages; /* Used page count, includes both */ + /* allocated and reserved pages. */ struct hstate *hstate; long min_hpages; /* Minimum huge pages or -1 if no minimum. */ long rsv_hpages; /* Pages reserved against global pool to */ diff --git a/mm/hugetlb.c b/mm/hugetlb.c index 4f6f58bf3db6c..e72e22f887478 100644 --- a/mm/hugetlb.c +++ b/mm/hugetlb.c @@ -130,12 +130,8 @@ static inline bool subpool_is_free(struct hugepage_subpool *spool) { if (spool->count) return false; - if (spool->max_hpages != -1) - return spool->used_hpages == 0; - if (spool->min_hpages != -1) - return spool->rsv_hpages == spool->min_hpages; - return true; + return spool->used_hpages == 0; } static inline void unlock_or_release_subpool(struct hugepage_subpool *spool, @@ -193,13 +189,18 @@ void hugepage_put_subpool(struct hugepage_subpool *spool) unlock_or_release_subpool(spool, flags); } -/* - * Subpool accounting for allocating and reserving pages. - * Return -ENOMEM if there are not enough resources to satisfy the - * request. Otherwise, return the number of pages by which the - * global pools must be adjusted (upward). The returned value may - * only be different than the passed value (delta) in the case where - * a subpool minimum size must be maintained. +/** + * hugepage_subpool_get_pages - Get pages from a subpool + * @spool: pointer to subpool structure (may be NULL) + * @delta: number of pages to allocate or reserve + * + * Check and update subpool page usage counts when allocating or + * reserving @delta hugepages. + * + * Context: Takes spool->lock using spin_lock_irq(). + * Return: Non-negative number of reservations that cannot be + * satisfied by the subpool, or -ENOMEM if the subpool maximum + * limit would be exceeded. */ static long hugepage_subpool_get_pages(struct hugepage_subpool *spool, long delta) @@ -211,15 +212,14 @@ static long hugepage_subpool_get_pages(struct hugepage_subpool *spool, spin_lock_irq(&spool->lock); - if (spool->max_hpages != -1) { /* maximum size accounting */ - if ((spool->used_hpages + delta) <= spool->max_hpages) - spool->used_hpages += delta; - else { - ret = -ENOMEM; - goto unlock_ret; - } + if (spool->max_hpages != -1 && + spool->used_hpages + delta > spool->max_hpages) { + ret = -ENOMEM; + goto unlock_ret; } + spool->used_hpages += delta; + /* minimum size accounting */ if (spool->min_hpages != -1 && spool->rsv_hpages) { if (delta > spool->rsv_hpages) { @@ -240,11 +240,19 @@ static long hugepage_subpool_get_pages(struct hugepage_subpool *spool, return ret; } -/* - * Subpool accounting for freeing and unreserving pages. - * Return the number of global page reservations that must be dropped. - * The return value may only be different than the passed value (delta) - * in the case where a subpool minimum size must be maintained. +/** + * hugepage_subpool_put_pages - Release pages back to a subpool + * @spool: pointer to subpool structure (may be NULL) + * @delta: number of pages to free or unreserve + * + * Check and update subpool page usage counts when freeing or + * unreserving @delta hugepages. + * + * Context: Takes spool->lock using spin_lock_irqsave(). May release + * and free @spool if its usage count and references reach + * zero. + * Return: Non-negative number of reservations that the subpool cannot + * absorb. */ static long hugepage_subpool_put_pages(struct hugepage_subpool *spool, long delta) @@ -257,19 +265,24 @@ static long hugepage_subpool_put_pages(struct hugepage_subpool *spool, spin_lock_irqsave(&spool->lock, flags); - if (spool->max_hpages != -1) /* maximum size accounting */ - spool->used_hpages -= delta; + spool->used_hpages -= delta; /* minimum size accounting */ if (spool->min_hpages != -1 && spool->used_hpages < spool->min_hpages) { - if (spool->rsv_hpages + delta <= spool->min_hpages) + /* + * limit is the maximum number of reservations that + * can be restored to this subpool. + */ + long limit = spool->min_hpages - spool->used_hpages; + + if (spool->rsv_hpages + delta <= limit) ret = 0; else - ret = spool->rsv_hpages + delta - spool->min_hpages; + ret = spool->rsv_hpages + delta - limit; spool->rsv_hpages += delta; - if (spool->rsv_hpages > spool->min_hpages) - spool->rsv_hpages = spool->min_hpages; + if (spool->rsv_hpages > limit) + spool->rsv_hpages = limit; } /* -- 2.55.0.1082.g2b9226bbc0-goog