| Message ID | 20190625001246.685563-7-songliubraving@fb.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
| Series | Enable THP for text section of non-shmem files | expand |
On Mon, 2019-06-24 at 17:12 -0700, Song Liu wrote: > In previous patch, an application could put part of its text section > in > THP via madvise(). These THPs will be protected from writes when the > application is still running (TXTBSY). However, after the application > exits, the file is available for writes. > > This patch avoids writes to file THP by dropping page cache for the > file > when the file is open for write. A new counter nr_thps is added to > struct > address_space. In do_last(), if the file is open for write and > nr_thps > is non-zero, we drop page cache for the whole file. > > Reported-by: kbuild test robot <lkp@intel.com> > Signed-off-by: Song Liu <songliubraving@fb.com> Acked-by: Rik van Riel <riel@surriel.com>
On Mon, Jun 24, 2019 at 05:12:46PM -0700, Song Liu wrote: > In previous patch, an application could put part of its text section in > THP via madvise(). These THPs will be protected from writes when the > application is still running (TXTBSY). However, after the application > exits, the file is available for writes. > > This patch avoids writes to file THP by dropping page cache for the file > when the file is open for write. A new counter nr_thps is added to struct > address_space. In do_last(), if the file is open for write and nr_thps > is non-zero, we drop page cache for the whole file. > > Reported-by: kbuild test robot <lkp@intel.com> > Signed-off-by: Song Liu <songliubraving@fb.com> > --- > fs/inode.c | 3 +++ > fs/namei.c | 23 ++++++++++++++++++++++- > include/linux/fs.h | 32 ++++++++++++++++++++++++++++++++ > mm/filemap.c | 1 + > mm/khugepaged.c | 4 +++- > 5 files changed, 61 insertions(+), 2 deletions(-) > > diff --git a/fs/inode.c b/fs/inode.c > index df6542ec3b88..518113a4e219 100644 > --- a/fs/inode.c > +++ b/fs/inode.c > @@ -181,6 +181,9 @@ int inode_init_always(struct super_block *sb, struct inode *inode) > mapping->flags = 0; > mapping->wb_err = 0; > atomic_set(&mapping->i_mmap_writable, 0); > +#ifdef CONFIG_READ_ONLY_THP_FOR_FS > + atomic_set(&mapping->nr_thps, 0); > +#endif > mapping_set_gfp_mask(mapping, GFP_HIGHUSER_MOVABLE); > mapping->private_data = NULL; > mapping->writeback_index = 0; > diff --git a/fs/namei.c b/fs/namei.c > index 20831c2fbb34..3d95e94029cc 100644 > --- a/fs/namei.c > +++ b/fs/namei.c > @@ -3249,6 +3249,23 @@ static int lookup_open(struct nameidata *nd, struct path *path, > return error; > } > > +/* > + * The file is open for write, so it is not mmapped with VM_DENYWRITE. If > + * it still has THP in page cache, drop the whole file from pagecache > + * before processing writes. This helps us avoid handling write back of > + * THP for now. > + */ > +static inline void release_file_thp(struct file *file) > +{ > + if (IS_ENABLED(CONFIG_READ_ONLY_THP_FOR_FS)) { > + struct inode *inode = file_inode(file); > + > + if (inode_is_open_for_write(inode) && > + filemap_nr_thps(inode->i_mapping)) > + truncate_pagecache(inode, 0); > + } > +} > + > /* > * Handle the last step of open() > */ > @@ -3418,7 +3435,11 @@ static int do_last(struct nameidata *nd, > goto out; > opened: > error = ima_file_check(file, op->acc_mode); > - if (!error && will_truncate) > + if (error) > + goto out; > + > + release_file_thp(file); > + if (will_truncate) > error = handle_truncate(file); This would seem better placed in do_dentry_open(), where we're done with the namespace operation and actually work against the inode. Something roughly like this? diff --git a/fs/open.c b/fs/open.c index b5b80469b93d..cae893edbab6 100644 --- a/fs/open.c +++ b/fs/open.c @@ -799,6 +799,11 @@ static int do_dentry_open(struct file *f, if (!f->f_mapping->a_ops || !f->f_mapping->a_ops->direct_IO) return -EINVAL; } + + /* XXX: Huge page cache doesn't support writing yet */ + if ((f->f_mode & FMODE_WRITE) && filemap_nr_thps(inode->i_mapping)) + truncate_pagecache(inode, 0); + return 0; cleanup_all:
diff --git a/fs/inode.c b/fs/inode.c index df6542ec3b88..518113a4e219 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -181,6 +181,9 @@ int inode_init_always(struct super_block *sb, struct inode *inode) mapping->flags = 0; mapping->wb_err = 0; atomic_set(&mapping->i_mmap_writable, 0); +#ifdef CONFIG_READ_ONLY_THP_FOR_FS + atomic_set(&mapping->nr_thps, 0); +#endif mapping_set_gfp_mask(mapping, GFP_HIGHUSER_MOVABLE); mapping->private_data = NULL; mapping->writeback_index = 0; diff --git a/fs/namei.c b/fs/namei.c index 20831c2fbb34..3d95e94029cc 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -3249,6 +3249,23 @@ static int lookup_open(struct nameidata *nd, struct path *path, return error; } +/* + * The file is open for write, so it is not mmapped with VM_DENYWRITE. If + * it still has THP in page cache, drop the whole file from pagecache + * before processing writes. This helps us avoid handling write back of + * THP for now. + */ +static inline void release_file_thp(struct file *file) +{ + if (IS_ENABLED(CONFIG_READ_ONLY_THP_FOR_FS)) { + struct inode *inode = file_inode(file); + + if (inode_is_open_for_write(inode) && + filemap_nr_thps(inode->i_mapping)) + truncate_pagecache(inode, 0); + } +} + /* * Handle the last step of open() */ @@ -3418,7 +3435,11 @@ static int do_last(struct nameidata *nd, goto out; opened: error = ima_file_check(file, op->acc_mode); - if (!error && will_truncate) + if (error) + goto out; + + release_file_thp(file); + if (will_truncate) error = handle_truncate(file); out: if (unlikely(error > 0)) { diff --git a/include/linux/fs.h b/include/linux/fs.h index f7fdfe93e25d..082fc581c7fc 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -427,6 +427,7 @@ int pagecache_write_end(struct file *, struct address_space *mapping, * @i_pages: Cached pages. * @gfp_mask: Memory allocation flags to use for allocating pages. * @i_mmap_writable: Number of VM_SHARED mappings. + * @nr_thps: Number of THPs in the pagecache (non-shmem only). * @i_mmap: Tree of private and shared mappings. * @i_mmap_rwsem: Protects @i_mmap and @i_mmap_writable. * @nrpages: Number of page entries, protected by the i_pages lock. @@ -444,6 +445,10 @@ struct address_space { struct xarray i_pages; gfp_t gfp_mask; atomic_t i_mmap_writable; +#ifdef CONFIG_READ_ONLY_THP_FOR_FS + /* number of thp, only for non-shmem files */ + atomic_t nr_thps; +#endif struct rb_root_cached i_mmap; struct rw_semaphore i_mmap_rwsem; unsigned long nrpages; @@ -2790,6 +2795,33 @@ static inline errseq_t filemap_sample_wb_err(struct address_space *mapping) return errseq_sample(&mapping->wb_err); } +static inline int filemap_nr_thps(struct address_space *mapping) +{ +#ifdef CONFIG_READ_ONLY_THP_FOR_FS + return atomic_read(&mapping->nr_thps); +#else + return 0; +#endif +} + +static inline void filemap_nr_thps_inc(struct address_space *mapping) +{ +#ifdef CONFIG_READ_ONLY_THP_FOR_FS + atomic_inc(&mapping->nr_thps); +#else + WARN_ON_ONCE(1); +#endif +} + +static inline void filemap_nr_thps_dec(struct address_space *mapping) +{ +#ifdef CONFIG_READ_ONLY_THP_FOR_FS + atomic_dec(&mapping->nr_thps); +#else + WARN_ON_ONCE(1); +#endif +} + extern int vfs_fsync_range(struct file *file, loff_t start, loff_t end, int datasync); extern int vfs_fsync(struct file *file, int datasync); diff --git a/mm/filemap.c b/mm/filemap.c index e79ceccdc6df..a8e86c136381 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -205,6 +205,7 @@ static void unaccount_page_cache_page(struct address_space *mapping, __dec_node_page_state(page, NR_SHMEM_THPS); } else if (PageTransHuge(page)) { __dec_node_page_state(page, NR_FILE_THPS); + filemap_nr_thps_dec(mapping); } /* diff --git a/mm/khugepaged.c b/mm/khugepaged.c index acbbbeaa083c..0bbc6be51197 100644 --- a/mm/khugepaged.c +++ b/mm/khugepaged.c @@ -1503,8 +1503,10 @@ static void collapse_file(struct mm_struct *mm, if (is_shmem) __inc_node_page_state(new_page, NR_SHMEM_THPS); - else + else { __inc_node_page_state(new_page, NR_FILE_THPS); + filemap_nr_thps_inc(mapping); + } if (nr_none) { struct zone *zone = page_zone(new_page);
In previous patch, an application could put part of its text section in THP via madvise(). These THPs will be protected from writes when the application is still running (TXTBSY). However, after the application exits, the file is available for writes. This patch avoids writes to file THP by dropping page cache for the file when the file is open for write. A new counter nr_thps is added to struct address_space. In do_last(), if the file is open for write and nr_thps is non-zero, we drop page cache for the whole file. Reported-by: kbuild test robot <lkp@intel.com> Signed-off-by: Song Liu <songliubraving@fb.com> --- fs/inode.c | 3 +++ fs/namei.c | 23 ++++++++++++++++++++++- include/linux/fs.h | 32 ++++++++++++++++++++++++++++++++ mm/filemap.c | 1 + mm/khugepaged.c | 4 +++- 5 files changed, 61 insertions(+), 2 deletions(-)