We are going to need to call shmem_charge() under tree_lock to get accoutning right on collapse of small tmpfs pages into a huge one. The problem is that tree_lock is irq-safe and lockdep is not happy, that we take irq-unsafe lock under irq-safe[1]. Let's convert the lock to irq-safe. [1] https://gist.github.com/kiryl/80c0149e03ed35dfaf26628b8e03cdbc --- ipc/shm.c | 4 ++-- mm/shmem.c | 50 ++++++++++++++++++++++++++------------------------ 2 files changed, 28 insertions(+), 26 deletions(-) diff --git a/ipc/shm.c b/ipc/shm.c index 7fa5cbebbf19..dbac8860c721 100644 --- a/ipc/shm.c +++ b/ipc/shm.c @@ -766,10 +766,10 @@ static void shm_add_rss_swap(struct shmid_kernel *shp, } else { #ifdef CONFIG_SHMEM struct shmem_inode_info *info = SHMEM_I(inode); - spin_lock(&info->lock); + spin_lock_irq(&info->lock); *rss_add += inode->i_mapping->nrpages; *swp_add += info->swapped; - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); #else *rss_add += inode->i_mapping->nrpages; #endif diff --git a/mm/shmem.c b/mm/shmem.c index a3d8469b18a7..6766eeadf48a 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -258,14 +258,15 @@ bool shmem_charge(struct inode *inode, long pages) { struct shmem_inode_info *info = SHMEM_I(inode); struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); + unsigned long flags; if (shmem_acct_block(info->flags, pages)) return false; - spin_lock(&info->lock); + spin_lock_irqsave(&info->lock, flags); info->alloced += pages; inode->i_blocks += pages * BLOCKS_PER_PAGE; shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irqrestore(&info->lock, flags); inode->i_mapping->nrpages += pages; if (!sbinfo->max_blocks) @@ -273,10 +274,10 @@ bool shmem_charge(struct inode *inode, long pages) if (percpu_counter_compare(&sbinfo->used_blocks, sbinfo->max_blocks - pages) > 0) { inode->i_mapping->nrpages -= pages; - spin_lock(&info->lock); + spin_lock_irqsave(&info->lock, flags); info->alloced -= pages; shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irqrestore(&info->lock, flags); return false; } @@ -288,12 +289,13 @@ void shmem_uncharge(struct inode *inode, long pages) { struct shmem_inode_info *info = SHMEM_I(inode); struct shmem_sb_info *sbinfo = SHMEM_SB(inode->i_sb); + unsigned long flags; - spin_lock(&info->lock); + spin_lock_irqsave(&info->lock, flags); info->alloced -= pages; inode->i_blocks -= pages * BLOCKS_PER_PAGE; shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irqrestore(&info->lock, flags); if (sbinfo->max_blocks) percpu_counter_sub(&sbinfo->used_blocks, pages); @@ -818,10 +820,10 @@ static void shmem_undo_range(struct inode *inode, loff_t lstart, loff_t lend, index++; } - spin_lock(&info->lock); + spin_lock_irq(&info->lock); info->swapped -= nr_swaps_freed; shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); } void shmem_truncate_range(struct inode *inode, loff_t lstart, loff_t lend) @@ -838,9 +840,9 @@ static int shmem_getattr(struct vfsmount *mnt, struct dentry *dentry, struct shmem_inode_info *info = SHMEM_I(inode); if (info->alloced - info->swapped != inode->i_mapping->nrpages) { - spin_lock(&info->lock); + spin_lock_irq(&info->lock); shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); } generic_fillattr(inode, stat); return 0; @@ -984,9 +986,9 @@ static int shmem_unuse_inode(struct shmem_inode_info *info, delete_from_swap_cache(*pagep); set_page_dirty(*pagep); if (!error) { - spin_lock(&info->lock); + spin_lock_irq(&info->lock); info->swapped--; - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); swap_free(swap); } } @@ -1134,10 +1136,10 @@ static int shmem_writepage(struct page *page, struct writeback_control *wbc) list_add_tail(&info->swaplist, &shmem_swaplist); if (add_to_swap_cache(page, swap, GFP_ATOMIC) == 0) { - spin_lock(&info->lock); + spin_lock_irq(&info->lock); shmem_recalc_inode(inode); info->swapped++; - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); swap_shmem_alloc(swap); shmem_delete_from_page_cache(page, swp_to_radix_entry(swap)); @@ -1523,10 +1525,10 @@ repeat: mem_cgroup_commit_charge(page, memcg, true, false); - spin_lock(&info->lock); + spin_lock_irq(&info->lock); info->swapped--; shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); if (sgp == SGP_WRITE) mark_page_accessed(page); @@ -1603,11 +1605,11 @@ alloc_nohuge: page = shmem_alloc_and_acct_page(gfp, info, sbinfo, PageTransHuge(page)); lru_cache_add_anon(page); - spin_lock(&info->lock); + spin_lock_irq(&info->lock); info->alloced += 1 << compound_order(page); inode->i_blocks += BLOCKS_PER_PAGE << compound_order(page); shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); alloced = true; /* @@ -1639,9 +1641,9 @@ clear: if (alloced) { ClearPageDirty(page); delete_from_page_cache(page); - spin_lock(&info->lock); + spin_lock_irq(&info->lock); shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); } error = -EINVAL; goto unlock; @@ -1673,9 +1675,9 @@ unlock: } if (error == -ENOSPC && !once++) { info = SHMEM_I(inode); - spin_lock(&info->lock); + spin_lock_irq(&info->lock); shmem_recalc_inode(inode); - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); goto repeat; } if (error == -EEXIST) /* from above or from radix_tree_insert */ @@ -1874,7 +1876,7 @@ int shmem_lock(struct file *file, int lock, struct user_struct *user) struct shmem_inode_info *info = SHMEM_I(inode); int retval = -ENOMEM; - spin_lock(&info->lock); + spin_lock_irq(&info->lock); if (lock && !(info->flags & VM_LOCKED)) { if (!user_shm_lock(inode->i_size, user)) goto out_nomem; @@ -1889,7 +1891,7 @@ int shmem_lock(struct file *file, int lock, struct user_struct *user) retval = 0; out_nomem: - spin_unlock(&info->lock); + spin_unlock_irq(&info->lock); return retval; } -- 2.8.1 -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@xxxxxxxxx. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: <a href=mailto:"dont@xxxxxxxxx";> email@xxxxxxxxx </a>