diff options
Diffstat (limited to 'mm')
| -rw-r--r-- | mm/Kconfig | 1 | ||||
| -rw-r--r-- | mm/Makefile | 3 | ||||
| -rw-r--r-- | mm/debug.c | 4 | ||||
| -rw-r--r-- | mm/fadvise.c | 81 | ||||
| -rw-r--r-- | mm/maccess.c | 6 | ||||
| -rw-r--r-- | mm/readahead.c | 45 | ||||
| -rw-r--r-- | mm/shmem.c | 2 | ||||
| -rw-r--r-- | mm/vmacache.c | 38 | ||||
| -rw-r--r-- | mm/vmscan.c | 11 |
9 files changed, 88 insertions, 103 deletions
diff --git a/mm/Kconfig b/mm/Kconfig index a550635ea5c3..de64ea658716 100644 --- a/mm/Kconfig +++ b/mm/Kconfig @@ -637,6 +637,7 @@ config DEFERRED_STRUCT_PAGE_INIT depends on NO_BOOTMEM depends on SPARSEMEM depends on !NEED_PER_CPU_KM + depends on 64BIT help Ordinarily all struct pages are initialised during early boot in a single thread. On very large machines this can take a considerable diff --git a/mm/Makefile b/mm/Makefile index 8716bdabe1e6..26ef77a3883b 100644 --- a/mm/Makefile +++ b/mm/Makefile @@ -32,7 +32,7 @@ ifdef CONFIG_CROSS_MEMORY_ATTACH mmu-$(CONFIG_MMU) += process_vm_access.o endif -obj-y := filemap.o mempool.o oom_kill.o \ +obj-y := filemap.o mempool.o oom_kill.o fadvise.o \ maccess.o page_alloc.o page-writeback.o \ readahead.o swap.o truncate.o vmscan.o shmem.o \ util.o mmzone.o vmstat.o backing-dev.o \ @@ -49,7 +49,6 @@ else obj-y += bootmem.o endif -obj-$(CONFIG_ADVISE_SYSCALLS) += fadvise.o ifdef CONFIG_MMU obj-$(CONFIG_ADVISE_SYSCALLS) += madvise.o endif diff --git a/mm/debug.c b/mm/debug.c index 38c926520c97..bd10aad8539a 100644 --- a/mm/debug.c +++ b/mm/debug.c @@ -114,7 +114,7 @@ EXPORT_SYMBOL(dump_vma); void dump_mm(const struct mm_struct *mm) { - pr_emerg("mm %px mmap %px seqnum %d task_size %lu\n" + pr_emerg("mm %px mmap %px seqnum %llu task_size %lu\n" #ifdef CONFIG_MMU "get_unmapped_area %px\n" #endif @@ -142,7 +142,7 @@ void dump_mm(const struct mm_struct *mm) "tlb_flush_pending %d\n" "def_flags: %#lx(%pGv)\n", - mm, mm->mmap, mm->vmacache_seqnum, mm->task_size, + mm, mm->mmap, (long long) mm->vmacache_seqnum, mm->task_size, #ifdef CONFIG_MMU mm->get_unmapped_area, #endif diff --git a/mm/fadvise.c b/mm/fadvise.c index 2d8376e3c640..467bcd032037 100644 --- a/mm/fadvise.c +++ b/mm/fadvise.c @@ -27,9 +27,9 @@ * deactivate the pages and clear PG_Referenced. */ -int ksys_fadvise64_64(int fd, loff_t offset, loff_t len, int advice) +static int generic_fadvise(struct file *file, loff_t offset, loff_t len, + int advice) { - struct fd f = fdget(fd); struct inode *inode; struct address_space *mapping; struct backing_dev_info *bdi; @@ -37,22 +37,14 @@ int ksys_fadvise64_64(int fd, loff_t offset, loff_t len, int advice) pgoff_t start_index; pgoff_t end_index; unsigned long nrpages; - int ret = 0; - - if (!f.file) - return -EBADF; - inode = file_inode(f.file); - if (S_ISFIFO(inode->i_mode)) { - ret = -ESPIPE; - goto out; - } + inode = file_inode(file); + if (S_ISFIFO(inode->i_mode)) + return -ESPIPE; - mapping = f.file->f_mapping; - if (!mapping || len < 0) { - ret = -EINVAL; - goto out; - } + mapping = file->f_mapping; + if (!mapping || len < 0) + return -EINVAL; bdi = inode_to_bdi(mapping->host); @@ -67,9 +59,9 @@ int ksys_fadvise64_64(int fd, loff_t offset, loff_t len, int advice) /* no bad return value, but ignore advice */ break; default: - ret = -EINVAL; + return -EINVAL; } - goto out; + return 0; } /* @@ -85,21 +77,21 @@ int ksys_fadvise64_64(int fd, loff_t offset, loff_t len, int advice) switch (advice) { case POSIX_FADV_NORMAL: - f.file->f_ra.ra_pages = bdi->ra_pages; - spin_lock(&f.file->f_lock); - f.file->f_mode &= ~FMODE_RANDOM; - spin_unlock(&f.file->f_lock); + file->f_ra.ra_pages = bdi->ra_pages; + spin_lock(&file->f_lock); + file->f_mode &= ~FMODE_RANDOM; + spin_unlock(&file->f_lock); break; case POSIX_FADV_RANDOM: - spin_lock(&f.file->f_lock); - f.file->f_mode |= FMODE_RANDOM; - spin_unlock(&f.file->f_lock); + spin_lock(&file->f_lock); + file->f_mode |= FMODE_RANDOM; + spin_unlock(&file->f_lock); break; case POSIX_FADV_SEQUENTIAL: - f.file->f_ra.ra_pages = bdi->ra_pages * 2; - spin_lock(&f.file->f_lock); - f.file->f_mode &= ~FMODE_RANDOM; - spin_unlock(&f.file->f_lock); + file->f_ra.ra_pages = bdi->ra_pages * 2; + spin_lock(&file->f_lock); + file->f_mode &= ~FMODE_RANDOM; + spin_unlock(&file->f_lock); break; case POSIX_FADV_WILLNEED: /* First and last PARTIAL page! */ @@ -115,8 +107,7 @@ int ksys_fadvise64_64(int fd, loff_t offset, loff_t len, int advice) * Ignore return value because fadvise() shall return * success even if filesystem can't retrieve a hint, */ - force_page_cache_readahead(mapping, f.file, start_index, - nrpages); + force_page_cache_readahead(mapping, file, start_index, nrpages); break; case POSIX_FADV_NOREUSE: break; @@ -183,9 +174,32 @@ int ksys_fadvise64_64(int fd, loff_t offset, loff_t len, int advice) } break; default: - ret = -EINVAL; + return -EINVAL; } -out: + return 0; +} + +int vfs_fadvise(struct file *file, loff_t offset, loff_t len, int advice) +{ + if (file->f_op->fadvise) + return file->f_op->fadvise(file, offset, len, advice); + + return generic_fadvise(file, offset, len, advice); +} +EXPORT_SYMBOL(vfs_fadvise); + +#ifdef CONFIG_ADVISE_SYSCALLS + +int ksys_fadvise64_64(int fd, loff_t offset, loff_t len, int advice) +{ + struct fd f = fdget(fd); + int ret; + + if (!f.file) + return -EBADF; + + ret = vfs_fadvise(f.file, offset, len, advice); + fdput(f); return ret; } @@ -203,3 +217,4 @@ SYSCALL_DEFINE4(fadvise64, int, fd, loff_t, offset, size_t, len, int, advice) } #endif +#endif diff --git a/mm/maccess.c b/mm/maccess.c index ec00be51a24f..f3416632e5a4 100644 --- a/mm/maccess.c +++ b/mm/maccess.c @@ -30,8 +30,10 @@ long __probe_kernel_read(void *dst, const void *src, size_t size) set_fs(KERNEL_DS); pagefault_disable(); + current->kernel_uaccess_faults_ok++; ret = __copy_from_user_inatomic(dst, (__force const void __user *)src, size); + current->kernel_uaccess_faults_ok--; pagefault_enable(); set_fs(old_fs); @@ -58,7 +60,9 @@ long __probe_kernel_write(void *dst, const void *src, size_t size) set_fs(KERNEL_DS); pagefault_disable(); + current->kernel_uaccess_faults_ok++; ret = __copy_to_user_inatomic((__force void __user *)dst, src, size); + current->kernel_uaccess_faults_ok--; pagefault_enable(); set_fs(old_fs); @@ -94,11 +98,13 @@ long strncpy_from_unsafe(char *dst, const void *unsafe_addr, long count) set_fs(KERNEL_DS); pagefault_disable(); + current->kernel_uaccess_faults_ok++; do { ret = __get_user(*dst++, (const char __user __force *)src++); } while (dst[-1] && ret == 0 && src - unsafe_addr < count); + current->kernel_uaccess_faults_ok--; dst[-1] = '\0'; pagefault_enable(); set_fs(old_fs); diff --git a/mm/readahead.c b/mm/readahead.c index a59ea70527b9..4e630143a0ba 100644 --- a/mm/readahead.c +++ b/mm/readahead.c @@ -20,6 +20,7 @@ #include <linux/file.h> #include <linux/mm_inline.h> #include <linux/blk-cgroup.h> +#include <linux/fadvise.h> #include "internal.h" @@ -575,24 +576,6 @@ page_cache_async_readahead(struct address_space *mapping, } EXPORT_SYMBOL_GPL(page_cache_async_readahead); -static ssize_t -do_readahead(struct address_space *mapping, struct file *filp, - pgoff_t index, unsigned long nr) -{ - if (!mapping || !mapping->a_ops) - return -EINVAL; - - /* - * Readahead doesn't make sense for DAX inodes, but we don't want it - * to report a failure either. Instead, we just return success and - * don't do any work. - */ - if (dax_mapping(mapping)) - return 0; - - return force_page_cache_readahead(mapping, filp, index, nr); -} - ssize_t ksys_readahead(int fd, loff_t offset, size_t count) { ssize_t ret; @@ -600,16 +583,22 @@ ssize_t ksys_readahead(int fd, loff_t offset, size_t count) ret = -EBADF; f = fdget(fd); - if (f.file) { - if (f.file->f_mode & FMODE_READ) { - struct address_space *mapping = f.file->f_mapping; - pgoff_t start = offset >> PAGE_SHIFT; - pgoff_t end = (offset + count - 1) >> PAGE_SHIFT; - unsigned long len = end - start + 1; - ret = do_readahead(mapping, f.file, start, len); - } - fdput(f); - } + if (!f.file || !(f.file->f_mode & FMODE_READ)) + goto out; + + /* + * The readahead() syscall is intended to run only on files + * that can execute readahead. If readahead is not possible + * on this file, then we must return -EINVAL. + */ + ret = -EINVAL; + if (!f.file->f_mapping || !f.file->f_mapping->a_ops || + !S_ISREG(file_inode(f.file)->i_mode)) + goto out; + + ret = vfs_fadvise(f.file, offset, count, POSIX_FADV_WILLNEED); +out: + fdput(f); return ret; } diff --git a/mm/shmem.c b/mm/shmem.c index 0376c124b043..446942677cd4 100644 --- a/mm/shmem.c +++ b/mm/shmem.c @@ -2227,6 +2227,8 @@ static struct inode *shmem_get_inode(struct super_block *sb, const struct inode mpol_shared_policy_init(&info->policy, NULL); break; } + + lockdep_annotate_inode_mutex_key(inode); } else shmem_free_inode(sb); return inode; diff --git a/mm/vmacache.c b/mm/vmacache.c index ea517bef7dc5..cdc32a3b02fa 100644 --- a/mm/vmacache.c +++ b/mm/vmacache.c @@ -20,44 +20,6 @@ #define VMACACHE_HASH(addr) ((addr >> VMACACHE_SHIFT) & VMACACHE_MASK) /* - * Flush vma caches for threads that share a given mm. - * - * The operation is safe because the caller holds the mmap_sem - * exclusively and other threads accessing the vma cache will - * have mmap_sem held at least for read, so no extra locking - * is required to maintain the vma cache. - */ -void vmacache_flush_all(struct mm_struct *mm) -{ - struct task_struct *g, *p; - - count_vm_vmacache_event(VMACACHE_FULL_FLUSHES); - - /* - * Single threaded tasks need not iterate the entire - * list of process. We can avoid the flushing as well - * since the mm's seqnum was increased and don't have - * to worry about other threads' seqnum. Current's - * flush will occur upon the next lookup. - */ - if (atomic_read(&mm->mm_users) == 1) - return; - - rcu_read_lock(); - for_each_process_thread(g, p) { - /* - * Only flush the vmacache pointers as the - * mm seqnum is already set and curr's will - * be set upon invalidation when the next - * lookup is done. - */ - if (mm == p->mm) - vmacache_flush(p); - } - rcu_read_unlock(); -} - -/* * This task may be accessing a foreign mm via (for example) * get_user_pages()->find_vma(). The vmacache is task-local and this * task's vmacache pertains to a different mm (ie, its own). There is diff --git a/mm/vmscan.c b/mm/vmscan.c index 7e7d25504651..c7ce2c161225 100644 --- a/mm/vmscan.c +++ b/mm/vmscan.c @@ -476,6 +476,17 @@ static unsigned long do_shrink_slab(struct shrink_control *shrinkctl, delta = freeable >> priority; delta *= 4; do_div(delta, shrinker->seeks); + + /* + * Make sure we apply some minimal pressure on default priority + * even on small cgroups. Stale objects are not only consuming memory + * by themselves, but can also hold a reference to a dying cgroup, + * preventing it from being reclaimed. A dying cgroup with all + * corresponding structures like per-cpu stats and kmem caches + * can be really big, so it may lead to a significant waste of memory. + */ + delta = max_t(unsigned long long, delta, min(freeable, batch_size)); + total_scan += delta; if (total_scan < 0) { pr_err("shrink_slab: %pF negative objects to delete nr=%ld\n", |