From 88bc7d5097a11d9bdcf08ecf85c81ba998353437 Mon Sep 17 00:00:00 2001 From: Niels de Vos Date: Tue, 21 Aug 2018 14:36:31 +0200 Subject: fuse: add support for copy_file_range() There are several FUSE filesystems that can implement server-side copy or other efficient copy/duplication/clone methods. The copy_file_range() syscall is the standard interface that users have access to while not depending on external libraries that bypass FUSE. Signed-off-by: Niels de Vos Signed-off-by: Miklos Szeredi --- include/uapi/linux/fuse.h | 106 ++++++++++++++++++++++++++-------------------- 1 file changed, 60 insertions(+), 46 deletions(-) (limited to 'include/uapi/linux') diff --git a/include/uapi/linux/fuse.h b/include/uapi/linux/fuse.h index 92fa24c24c92..d27b50a44f74 100644 --- a/include/uapi/linux/fuse.h +++ b/include/uapi/linux/fuse.h @@ -116,6 +116,9 @@ * * 7.27 * - add FUSE_ABORT_ERROR + * + * 7.28 + * - add FUSE_COPY_FILE_RANGE */ #ifndef _LINUX_FUSE_H @@ -151,7 +154,7 @@ #define FUSE_KERNEL_VERSION 7 /** Minor version number of this interface */ -#define FUSE_KERNEL_MINOR_VERSION 27 +#define FUSE_KERNEL_MINOR_VERSION 28 /** The node ID of the root inode */ #define FUSE_ROOT_ID 1 @@ -337,53 +340,54 @@ struct fuse_file_lock { #define FUSE_POLL_SCHEDULE_NOTIFY (1 << 0) enum fuse_opcode { - FUSE_LOOKUP = 1, - FUSE_FORGET = 2, /* no reply */ - FUSE_GETATTR = 3, - FUSE_SETATTR = 4, - FUSE_READLINK = 5, - FUSE_SYMLINK = 6, - FUSE_MKNOD = 8, - FUSE_MKDIR = 9, - FUSE_UNLINK = 10, - FUSE_RMDIR = 11, - FUSE_RENAME = 12, - FUSE_LINK = 13, - FUSE_OPEN = 14, - FUSE_READ = 15, - FUSE_WRITE = 16, - FUSE_STATFS = 17, - FUSE_RELEASE = 18, - FUSE_FSYNC = 20, - FUSE_SETXATTR = 21, - FUSE_GETXATTR = 22, - FUSE_LISTXATTR = 23, - FUSE_REMOVEXATTR = 24, - FUSE_FLUSH = 25, - FUSE_INIT = 26, - FUSE_OPENDIR = 27, - FUSE_READDIR = 28, - FUSE_RELEASEDIR = 29, - FUSE_FSYNCDIR = 30, - FUSE_GETLK = 31, - FUSE_SETLK = 32, - FUSE_SETLKW = 33, - FUSE_ACCESS = 34, - FUSE_CREATE = 35, - FUSE_INTERRUPT = 36, - FUSE_BMAP = 37, - FUSE_DESTROY = 38, - FUSE_IOCTL = 39, - FUSE_POLL = 40, - FUSE_NOTIFY_REPLY = 41, - FUSE_BATCH_FORGET = 42, - FUSE_FALLOCATE = 43, - FUSE_READDIRPLUS = 44, - FUSE_RENAME2 = 45, - FUSE_LSEEK = 46, + FUSE_LOOKUP = 1, + FUSE_FORGET = 2, /* no reply */ + FUSE_GETATTR = 3, + FUSE_SETATTR = 4, + FUSE_READLINK = 5, + FUSE_SYMLINK = 6, + FUSE_MKNOD = 8, + FUSE_MKDIR = 9, + FUSE_UNLINK = 10, + FUSE_RMDIR = 11, + FUSE_RENAME = 12, + FUSE_LINK = 13, + FUSE_OPEN = 14, + FUSE_READ = 15, + FUSE_WRITE = 16, + FUSE_STATFS = 17, + FUSE_RELEASE = 18, + FUSE_FSYNC = 20, + FUSE_SETXATTR = 21, + FUSE_GETXATTR = 22, + FUSE_LISTXATTR = 23, + FUSE_REMOVEXATTR = 24, + FUSE_FLUSH = 25, + FUSE_INIT = 26, + FUSE_OPENDIR = 27, + FUSE_READDIR = 28, + FUSE_RELEASEDIR = 29, + FUSE_FSYNCDIR = 30, + FUSE_GETLK = 31, + FUSE_SETLK = 32, + FUSE_SETLKW = 33, + FUSE_ACCESS = 34, + FUSE_CREATE = 35, + FUSE_INTERRUPT = 36, + FUSE_BMAP = 37, + FUSE_DESTROY = 38, + FUSE_IOCTL = 39, + FUSE_POLL = 40, + FUSE_NOTIFY_REPLY = 41, + FUSE_BATCH_FORGET = 42, + FUSE_FALLOCATE = 43, + FUSE_READDIRPLUS = 44, + FUSE_RENAME2 = 45, + FUSE_LSEEK = 46, + FUSE_COPY_FILE_RANGE = 47, /* CUSE specific operations */ - CUSE_INIT = 4096, + CUSE_INIT = 4096, }; enum fuse_notify_code { @@ -792,4 +796,14 @@ struct fuse_lseek_out { uint64_t offset; }; +struct fuse_copy_file_range_in { + uint64_t fh_in; + uint64_t off_in; + uint64_t nodeid_out; + uint64_t fh_out; + uint64_t off_out; + uint64_t len; + uint64_t flags; +}; + #endif /* _LINUX_FUSE_H */ -- cgit v1.3-14-g43fede From 6433b8998a21dc597002731c4ceb4144e856edc4 Mon Sep 17 00:00:00 2001 From: Miklos Szeredi Date: Fri, 28 Sep 2018 16:43:23 +0200 Subject: fuse: add FOPEN_CACHE_DIR Add flag returned by OPENDIR request to allow kernel to cache directory contents in page cache. The effect of FOPEN_CACHE_DIR is twofold: a) if not already cached, it writes entries into the cache b) if already cached, it allows reading entries from the cache The FOPEN_KEEP_CACHE has the same effect as on regular files: unless this flag is given the cache is cleared upon completion of open. So FOPEN_KEEP_CACHE and FOPEN_KEEP_CACHE flags should be used together to make use of the directory caching facility introduced in the following patches. The FUSE_AUTO_INVAL_DATA flag returned in INIT reply also has the same affect on the directory cache as it has on data cache for regular files. Signed-off-by: Miklos Szeredi --- include/uapi/linux/fuse.h | 3 +++ 1 file changed, 3 insertions(+) (limited to 'include/uapi/linux') diff --git a/include/uapi/linux/fuse.h b/include/uapi/linux/fuse.h index d27b50a44f74..31a504f1ee60 100644 --- a/include/uapi/linux/fuse.h +++ b/include/uapi/linux/fuse.h @@ -119,6 +119,7 @@ * * 7.28 * - add FUSE_COPY_FILE_RANGE + * - add FOPEN_CACHE_DIR */ #ifndef _LINUX_FUSE_H @@ -222,10 +223,12 @@ struct fuse_file_lock { * FOPEN_DIRECT_IO: bypass page cache for this open file * FOPEN_KEEP_CACHE: don't invalidate the data cache on open * FOPEN_NONSEEKABLE: the file is not seekable + * FOPEN_CACHE_DIR: allow caching this directory */ #define FOPEN_DIRECT_IO (1 << 0) #define FOPEN_KEEP_CACHE (1 << 1) #define FOPEN_NONSEEKABLE (1 << 2) +#define FOPEN_CACHE_DIR (1 << 3) /** * INIT request/reply flags -- cgit v1.3-14-g43fede From 5da784cce4308ae10a79e3c8c41b13fb9568e4e0 Mon Sep 17 00:00:00 2001 From: Constantine Shulyupin Date: Thu, 6 Sep 2018 15:37:06 +0300 Subject: fuse: add max_pages to init_out Replace FUSE_MAX_PAGES_PER_REQ with the configurable parameter max_pages to improve performance. Old RFC with detailed description of the problem and many fixes by Mitsuo Hayasaka (mitsuo.hayasaka.hu@hitachi.com): - https://lkml.org/lkml/2012/7/5/136 We've encountered performance degradation and fixed it on a big and complex virtual environment. Environment to reproduce degradation and improvement: 1. Add lag to user mode FUSE Add nanosleep(&(struct timespec){ 0, 1000 }, NULL); to xmp_write_buf in passthrough_fh.c 2. patch UM fuse with configurable max_pages parameter. The patch will be provided latter. 3. run test script and perform test on tmpfs fuse_test() { cd /tmp mkdir -p fusemnt passthrough_fh -o max_pages=$1 /tmp/fusemnt grep fuse /proc/self/mounts dd conv=fdatasync oflag=dsync if=/dev/zero of=fusemnt/tmp/tmp \ count=1K bs=1M 2>&1 | grep -v records rm fusemnt/tmp/tmp killall passthrough_fh } Test results: passthrough_fh /tmp/fusemnt fuse.passthrough_fh \ rw,nosuid,nodev,relatime,user_id=0,group_id=0 0 0 1073741824 bytes (1.1 GB) copied, 1.73867 s, 618 MB/s passthrough_fh /tmp/fusemnt fuse.passthrough_fh \ rw,nosuid,nodev,relatime,user_id=0,group_id=0,max_pages=256 0 0 1073741824 bytes (1.1 GB) copied, 1.15643 s, 928 MB/s Obviously with bigger lag the difference between 'before' and 'after' will be more significant. Mitsuo Hayasaka, in 2012 (https://lkml.org/lkml/2012/7/5/136), observed improvement from 400-550 to 520-740. Signed-off-by: Constantine Shulyupin Signed-off-by: Miklos Szeredi --- fs/fuse/dev.c | 5 ++-- fs/fuse/file.c | 59 ++++++++++++++++++++++++----------------------- fs/fuse/fuse_i.h | 10 ++++++-- fs/fuse/inode.c | 8 ++++++- include/uapi/linux/fuse.h | 7 +++++- 5 files changed, 54 insertions(+), 35 deletions(-) (limited to 'include/uapi/linux') diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index fefb9dd8a2f4..69d4df78a417 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -61,6 +61,7 @@ static struct fuse_req *__fuse_request_alloc(unsigned npages, gfp_t flags) struct page **pages = NULL; struct fuse_page_desc *page_descs = NULL; + WARN_ON(npages > FUSE_MAX_MAX_PAGES); if (npages > FUSE_REQ_INLINE_PAGES) { pages = kzalloc(npages * (sizeof(*pages) + sizeof(*page_descs)), flags); @@ -1674,7 +1675,7 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, unsigned int num; unsigned int offset; size_t total_len = 0; - int num_pages; + unsigned int num_pages; offset = outarg->offset & ~PAGE_MASK; file_size = i_size_read(inode); @@ -1686,7 +1687,7 @@ static int fuse_retrieve(struct fuse_conn *fc, struct inode *inode, num = file_size - outarg->offset; num_pages = (num + offset + PAGE_SIZE - 1) >> PAGE_SHIFT; - num_pages = min(num_pages, FUSE_MAX_PAGES_PER_REQ); + num_pages = min(num_pages, fc->max_pages); req = fuse_get_req(fc, num_pages); if (IS_ERR(req)) diff --git a/fs/fuse/file.c b/fs/fuse/file.c index b10d14baeb1f..035843b501fe 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -850,11 +850,11 @@ static int fuse_readpages_fill(void *_data, struct page *page) fuse_wait_on_page_writeback(inode, page->index); if (req->num_pages && - (req->num_pages == FUSE_MAX_PAGES_PER_REQ || + (req->num_pages == fc->max_pages || (req->num_pages + 1) * PAGE_SIZE > fc->max_read || req->pages[req->num_pages - 1]->index + 1 != page->index)) { - int nr_alloc = min_t(unsigned, data->nr_pages, - FUSE_MAX_PAGES_PER_REQ); + unsigned int nr_alloc = min_t(unsigned int, data->nr_pages, + fc->max_pages); fuse_send_readpages(req, data->file); if (fc->async_read) req = fuse_get_req_for_background(fc, nr_alloc); @@ -889,7 +889,7 @@ static int fuse_readpages(struct file *file, struct address_space *mapping, struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_fill_data data; int err; - int nr_alloc = min_t(unsigned, nr_pages, FUSE_MAX_PAGES_PER_REQ); + unsigned int nr_alloc = min_t(unsigned int, nr_pages, fc->max_pages); err = -EIO; if (is_bad_inode(inode)) @@ -1104,12 +1104,13 @@ static ssize_t fuse_fill_write_pages(struct fuse_req *req, return count > 0 ? count : err; } -static inline unsigned fuse_wr_pages(loff_t pos, size_t len) +static inline unsigned int fuse_wr_pages(loff_t pos, size_t len, + unsigned int max_pages) { - return min_t(unsigned, + return min_t(unsigned int, ((pos + len - 1) >> PAGE_SHIFT) - (pos >> PAGE_SHIFT) + 1, - FUSE_MAX_PAGES_PER_REQ); + max_pages); } static ssize_t fuse_perform_write(struct kiocb *iocb, @@ -1131,7 +1132,8 @@ static ssize_t fuse_perform_write(struct kiocb *iocb, do { struct fuse_req *req; ssize_t count; - unsigned nr_pages = fuse_wr_pages(pos, iov_iter_count(ii)); + unsigned int nr_pages = fuse_wr_pages(pos, iov_iter_count(ii), + fc->max_pages); req = fuse_get_req(fc, nr_pages); if (IS_ERR(req)) { @@ -1321,11 +1323,6 @@ static int fuse_get_user_pages(struct fuse_req *req, struct iov_iter *ii, return ret < 0 ? ret : 0; } -static inline int fuse_iter_npages(const struct iov_iter *ii_p) -{ - return iov_iter_npages(ii_p, FUSE_MAX_PAGES_PER_REQ); -} - ssize_t fuse_direct_io(struct fuse_io_priv *io, struct iov_iter *iter, loff_t *ppos, int flags) { @@ -1345,9 +1342,10 @@ ssize_t fuse_direct_io(struct fuse_io_priv *io, struct iov_iter *iter, int err = 0; if (io->async) - req = fuse_get_req_for_background(fc, fuse_iter_npages(iter)); + req = fuse_get_req_for_background(fc, iov_iter_npages(iter, + fc->max_pages)); else - req = fuse_get_req(fc, fuse_iter_npages(iter)); + req = fuse_get_req(fc, iov_iter_npages(iter, fc->max_pages)); if (IS_ERR(req)) return PTR_ERR(req); @@ -1392,9 +1390,10 @@ ssize_t fuse_direct_io(struct fuse_io_priv *io, struct iov_iter *iter, fuse_put_request(fc, req); if (io->async) req = fuse_get_req_for_background(fc, - fuse_iter_npages(iter)); + iov_iter_npages(iter, fc->max_pages)); else - req = fuse_get_req(fc, fuse_iter_npages(iter)); + req = fuse_get_req(fc, iov_iter_npages(iter, + fc->max_pages)); if (IS_ERR(req)) break; } @@ -1823,7 +1822,7 @@ static int fuse_writepages_fill(struct page *page, is_writeback = fuse_page_is_writeback(inode, page->index); if (req && req->num_pages && - (is_writeback || req->num_pages == FUSE_MAX_PAGES_PER_REQ || + (is_writeback || req->num_pages == fc->max_pages || (req->num_pages + 1) * PAGE_SIZE > fc->max_write || data->orig_pages[req->num_pages - 1]->index + 1 != page->index)) { fuse_writepages_send(data); @@ -1851,7 +1850,7 @@ static int fuse_writepages_fill(struct page *page, struct fuse_inode *fi = get_fuse_inode(inode); err = -ENOMEM; - req = fuse_request_alloc_nofs(FUSE_MAX_PAGES_PER_REQ); + req = fuse_request_alloc_nofs(fc->max_pages); if (!req) { __free_page(tmp_page); goto out_unlock; @@ -1908,6 +1907,7 @@ static int fuse_writepages(struct address_space *mapping, struct writeback_control *wbc) { struct inode *inode = mapping->host; + struct fuse_conn *fc = get_fuse_conn(inode); struct fuse_fill_wb_data data; int err; @@ -1920,7 +1920,7 @@ static int fuse_writepages(struct address_space *mapping, data.ff = NULL; err = -ENOMEM; - data.orig_pages = kcalloc(FUSE_MAX_PAGES_PER_REQ, + data.orig_pages = kcalloc(fc->max_pages, sizeof(struct page *), GFP_NOFS); if (!data.orig_pages) @@ -2391,10 +2391,11 @@ static int fuse_copy_ioctl_iovec_old(struct iovec *dst, void *src, } /* Make sure iov_length() won't overflow */ -static int fuse_verify_ioctl_iov(struct iovec *iov, size_t count) +static int fuse_verify_ioctl_iov(struct fuse_conn *fc, struct iovec *iov, + size_t count) { size_t n; - u32 max = FUSE_MAX_PAGES_PER_REQ << PAGE_SHIFT; + u32 max = fc->max_pages << PAGE_SHIFT; for (n = 0; n < count; n++, iov++) { if (iov->iov_len > (size_t) max) @@ -2518,7 +2519,7 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, BUILD_BUG_ON(sizeof(struct fuse_ioctl_iovec) * FUSE_IOCTL_MAX_IOV > PAGE_SIZE); err = -ENOMEM; - pages = kcalloc(FUSE_MAX_PAGES_PER_REQ, sizeof(pages[0]), GFP_KERNEL); + pages = kcalloc(fc->max_pages, sizeof(pages[0]), GFP_KERNEL); iov_page = (struct iovec *) __get_free_page(GFP_KERNEL); if (!pages || !iov_page) goto out; @@ -2557,7 +2558,7 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, /* make sure there are enough buffer pages and init request with them */ err = -ENOMEM; - if (max_pages > FUSE_MAX_PAGES_PER_REQ) + if (max_pages > fc->max_pages) goto out; while (num_pages < max_pages) { pages[num_pages] = alloc_page(GFP_KERNEL | __GFP_HIGHMEM); @@ -2644,11 +2645,11 @@ long fuse_do_ioctl(struct file *file, unsigned int cmd, unsigned long arg, in_iov = iov_page; out_iov = in_iov + in_iovs; - err = fuse_verify_ioctl_iov(in_iov, in_iovs); + err = fuse_verify_ioctl_iov(fc, in_iov, in_iovs); if (err) goto out; - err = fuse_verify_ioctl_iov(out_iov, out_iovs); + err = fuse_verify_ioctl_iov(fc, out_iov, out_iovs); if (err) goto out; @@ -2839,9 +2840,9 @@ static void fuse_do_truncate(struct file *file) fuse_do_setattr(file_dentry(file), &attr, file); } -static inline loff_t fuse_round_up(loff_t off) +static inline loff_t fuse_round_up(struct fuse_conn *fc, loff_t off) { - return round_up(off, FUSE_MAX_PAGES_PER_REQ << PAGE_SHIFT); + return round_up(off, fc->max_pages << PAGE_SHIFT); } static ssize_t @@ -2870,7 +2871,7 @@ fuse_direct_IO(struct kiocb *iocb, struct iov_iter *iter) if (async_dio && iov_iter_rw(iter) != WRITE && offset + count > i_size) { if (offset >= i_size) return 0; - iov_iter_truncate(iter, fuse_round_up(i_size - offset)); + iov_iter_truncate(iter, fuse_round_up(ff->fc, i_size - offset)); count = iov_iter_count(iter); } diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index f5bdce84e766..3d578745c852 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -28,8 +28,11 @@ #include #include -/** Max number of pages that can be used in a single read request */ -#define FUSE_MAX_PAGES_PER_REQ 32 +/** Default max number of pages that can be used in a single read request */ +#define FUSE_DEFAULT_MAX_PAGES_PER_REQ 32 + +/** Maximum of max_pages received in init_out */ +#define FUSE_MAX_MAX_PAGES 256 /** Bias for fi->writectr, meaning new writepages must not be sent */ #define FUSE_NOWRITE INT_MIN @@ -525,6 +528,9 @@ struct fuse_conn { /** Maximum write size */ unsigned max_write; + /** Maxmum number of pages that can be used in a single request */ + unsigned int max_pages; + /** Input queue */ struct fuse_iqueue iq; diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index 82db1ab53420..8cebf4d5f51b 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -928,6 +928,11 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) } if (arg->flags & FUSE_ABORT_ERROR) fc->abort_err = 1; + if (arg->flags & FUSE_MAX_PAGES) { + fc->max_pages = + min_t(unsigned int, FUSE_MAX_MAX_PAGES, + max_t(unsigned int, arg->max_pages, 1)); + } } else { ra_pages = fc->max_read / PAGE_SIZE; fc->no_lock = 1; @@ -959,7 +964,7 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) FUSE_DO_READDIRPLUS | FUSE_READDIRPLUS_AUTO | FUSE_ASYNC_DIO | FUSE_WRITEBACK_CACHE | FUSE_NO_OPEN_SUPPORT | FUSE_PARALLEL_DIROPS | FUSE_HANDLE_KILLPRIV | FUSE_POSIX_ACL | - FUSE_ABORT_ERROR; + FUSE_ABORT_ERROR | FUSE_MAX_PAGES; req->in.h.opcode = FUSE_INIT; req->in.numargs = 1; req->in.args[0].size = sizeof(*arg); @@ -1152,6 +1157,7 @@ static int fuse_fill_super(struct super_block *sb, void *data, int silent) fc->user_id = d.user_id; fc->group_id = d.group_id; fc->max_read = max_t(unsigned, 4096, d.max_read); + fc->max_pages = FUSE_DEFAULT_MAX_PAGES_PER_REQ; /* Used by get_root_inode() */ sb->s_fs_info = fc; diff --git a/include/uapi/linux/fuse.h b/include/uapi/linux/fuse.h index 31a504f1ee60..76f46f159992 100644 --- a/include/uapi/linux/fuse.h +++ b/include/uapi/linux/fuse.h @@ -120,6 +120,7 @@ * 7.28 * - add FUSE_COPY_FILE_RANGE * - add FOPEN_CACHE_DIR + * - add FUSE_MAX_PAGES, add max_pages to init_out */ #ifndef _LINUX_FUSE_H @@ -255,6 +256,7 @@ struct fuse_file_lock { * FUSE_HANDLE_KILLPRIV: fs handles killing suid/sgid/cap on write/chown/trunc * FUSE_POSIX_ACL: filesystem supports posix acls * FUSE_ABORT_ERROR: reading the device after abort returns ECONNABORTED + * FUSE_MAX_PAGES: init_out.max_pages contains the max number of req pages */ #define FUSE_ASYNC_READ (1 << 0) #define FUSE_POSIX_LOCKS (1 << 1) @@ -278,6 +280,7 @@ struct fuse_file_lock { #define FUSE_HANDLE_KILLPRIV (1 << 19) #define FUSE_POSIX_ACL (1 << 20) #define FUSE_ABORT_ERROR (1 << 21) +#define FUSE_MAX_PAGES (1 << 22) /** * CUSE INIT request/reply flags @@ -617,7 +620,9 @@ struct fuse_init_out { uint16_t congestion_threshold; uint32_t max_write; uint32_t time_gran; - uint32_t unused[9]; + uint16_t max_pages; + uint16_t padding; + uint32_t unused[8]; }; #define CUSE_INIT_INFO_MAX 4096 -- cgit v1.3-14-g43fede From 3cdf752506b29ace75b6e1318abac06073d600e4 Mon Sep 17 00:00:00 2001 From: Gerd Hoffmann Date: Fri, 21 Sep 2018 10:30:12 +0200 Subject: vfio: add edid api for display (vgpu) devices. This allows to set EDID monitor information for the vgpu display, for a more flexible display configuration, using a special vfio region. Check the comment describing struct vfio_region_gfx_edid for more details. Signed-off-by: Gerd Hoffmann Signed-off-by: Alex Williamson --- include/uapi/linux/vfio.h | 50 +++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 50 insertions(+) (limited to 'include/uapi/linux') diff --git a/include/uapi/linux/vfio.h b/include/uapi/linux/vfio.h index 1aa7b82e8169..44b66b09c5fe 100644 --- a/include/uapi/linux/vfio.h +++ b/include/uapi/linux/vfio.h @@ -301,6 +301,56 @@ struct vfio_region_info_cap_type { #define VFIO_REGION_SUBTYPE_INTEL_IGD_HOST_CFG (2) #define VFIO_REGION_SUBTYPE_INTEL_IGD_LPC_CFG (3) +#define VFIO_REGION_TYPE_GFX (1) +#define VFIO_REGION_SUBTYPE_GFX_EDID (1) + +/** + * struct vfio_region_gfx_edid - EDID region layout. + * + * Set display link state and EDID blob. + * + * The EDID blob has monitor information such as brand, name, serial + * number, physical size, supported video modes and more. + * + * This special region allows userspace (typically qemu) set a virtual + * EDID for the virtual monitor, which allows a flexible display + * configuration. + * + * For the edid blob spec look here: + * https://en.wikipedia.org/wiki/Extended_Display_Identification_Data + * + * On linux systems you can find the EDID blob in sysfs: + * /sys/class/drm/${card}/${connector}/edid + * + * You can use the edid-decode ulility (comes with xorg-x11-utils) to + * decode the EDID blob. + * + * @edid_offset: location of the edid blob, relative to the + * start of the region (readonly). + * @edid_max_size: max size of the edid blob (readonly). + * @edid_size: actual edid size (read/write). + * @link_state: display link state (read/write). + * VFIO_DEVICE_GFX_LINK_STATE_UP: Monitor is turned on. + * VFIO_DEVICE_GFX_LINK_STATE_DOWN: Monitor is turned off. + * @max_xres: max display width (0 == no limitation, readonly). + * @max_yres: max display height (0 == no limitation, readonly). + * + * EDID update protocol: + * (1) set link-state to down. + * (2) update edid blob and size. + * (3) set link-state to up. + */ +struct vfio_region_gfx_edid { + __u32 edid_offset; + __u32 edid_max_size; + __u32 edid_size; + __u32 max_xres; + __u32 max_yres; + __u32 link_state; +#define VFIO_DEVICE_GFX_LINK_STATE_UP 1 +#define VFIO_DEVICE_GFX_LINK_STATE_DOWN 2 +}; + /* * The MSIX mappable capability informs that MSIX data of a BAR can be mmapped * which allows direct access to non-MSIX registers which happened to be within -- cgit v1.3-14-g43fede From 5571f1e65486be025f73fa6aa30fb03725d362a2 Mon Sep 17 00:00:00 2001 From: Dan Schatzberg Date: Thu, 11 Oct 2018 08:17:00 -0700 Subject: fuse: enable caching of symlinks FUSE file reads are cached in the page cache, but symlink reads are not. This patch enables FUSE READLINK operations to be cached which can improve performance of some FUSE workloads. In particular, I'm working on a FUSE filesystem for access to source code and discovered that about a 10% improvement to build times is achieved with this patch (there are a lot of symlinks in the source tree). Signed-off-by: Miklos Szeredi --- fs/fuse/dir.c | 108 +++++++++++++++++++++++++++++++++++----------- fs/fuse/fuse_i.h | 3 ++ fs/fuse/inode.c | 4 +- include/uapi/linux/fuse.h | 3 ++ 4 files changed, 92 insertions(+), 26 deletions(-) (limited to 'include/uapi/linux') diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index 7b8f63e7489f..47395b0c3b35 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -1160,38 +1160,78 @@ static int fuse_permission(struct inode *inode, int mask) return err; } -static const char *fuse_get_link(struct dentry *dentry, - struct inode *inode, - struct delayed_call *done) +static int fuse_readlink_page(struct inode *inode, struct page *page) { struct fuse_conn *fc = get_fuse_conn(inode); - FUSE_ARGS(args); - char *link; - ssize_t ret; + struct fuse_req *req; + int err; - if (!dentry) - return ERR_PTR(-ECHILD); + req = fuse_get_req(fc, 1); + if (IS_ERR(req)) + return PTR_ERR(req); + + req->out.page_zeroing = 1; + req->out.argpages = 1; + req->num_pages = 1; + req->pages[0] = page; + req->page_descs[0].length = PAGE_SIZE - 1; + req->in.h.opcode = FUSE_READLINK; + req->in.h.nodeid = get_node_id(inode); + req->out.argvar = 1; + req->out.numargs = 1; + req->out.args[0].size = PAGE_SIZE - 1; + fuse_request_send(fc, req); + err = req->out.h.error; - link = kmalloc(PAGE_SIZE, GFP_KERNEL); - if (!link) - return ERR_PTR(-ENOMEM); + if (!err) { + char *link = page_address(page); + size_t len = req->out.args[0].size; - args.in.h.opcode = FUSE_READLINK; - args.in.h.nodeid = get_node_id(inode); - args.out.argvar = 1; - args.out.numargs = 1; - args.out.args[0].size = PAGE_SIZE - 1; - args.out.args[0].value = link; - ret = fuse_simple_request(fc, &args); - if (ret < 0) { - kfree(link); - link = ERR_PTR(ret); - } else { - link[ret] = '\0'; - set_delayed_call(done, kfree_link, link); + BUG_ON(len >= PAGE_SIZE); + link[len] = '\0'; } + + fuse_put_request(fc, req); fuse_invalidate_atime(inode); - return link; + + return err; +} + +static const char *fuse_get_link(struct dentry *dentry, struct inode *inode, + struct delayed_call *callback) +{ + struct fuse_conn *fc = get_fuse_conn(inode); + struct page *page; + int err; + + err = -EIO; + if (is_bad_inode(inode)) + goto out_err; + + if (fc->cache_symlinks) + return page_get_link(dentry, inode, callback); + + err = -ECHILD; + if (!dentry) + goto out_err; + + page = alloc_page(GFP_KERNEL); + err = -ENOMEM; + if (!page) + goto out_err; + + err = fuse_readlink_page(inode, page); + if (err) { + __free_page(page); + goto out_err; + } + + set_delayed_call(callback, page_put_link, page); + + return page_address(page); + +out_err: + return ERR_PTR(err); } static int fuse_dir_open(struct inode *inode, struct file *file) @@ -1644,7 +1684,25 @@ void fuse_init_dir(struct inode *inode) fi->rdc.version = 0; } +static int fuse_symlink_readpage(struct file *null, struct page *page) +{ + int err = fuse_readlink_page(page->mapping->host, page); + + if (!err) + SetPageUptodate(page); + + unlock_page(page); + + return err; +} + +static const struct address_space_operations fuse_symlink_aops = { + .readpage = fuse_symlink_readpage, +}; + void fuse_init_symlink(struct inode *inode) { inode->i_op = &fuse_symlink_inode_operations; + inode->i_data.a_ops = &fuse_symlink_aops; + inode_nohighmem(inode); } diff --git a/fs/fuse/fuse_i.h b/fs/fuse/fuse_i.h index 0e32524e66bb..e9f712e81c7d 100644 --- a/fs/fuse/fuse_i.h +++ b/fs/fuse/fuse_i.h @@ -613,6 +613,9 @@ struct fuse_conn { /** handle fs handles killing suid/sgid/cap on write/chown/trunc */ unsigned handle_killpriv:1; + /** cache READLINK responses in page cache */ + unsigned cache_symlinks:1; + /* * The following bitfields are only for optimization purposes * and hence races in setting them will not cause malfunction diff --git a/fs/fuse/inode.c b/fs/fuse/inode.c index d5f845aefbc9..0b94b23b02d4 100644 --- a/fs/fuse/inode.c +++ b/fs/fuse/inode.c @@ -928,6 +928,8 @@ static void process_init_reply(struct fuse_conn *fc, struct fuse_req *req) fc->posix_acl = 1; fc->sb->s_xattr = fuse_acl_xattr_handlers; } + if (arg->flags & FUSE_CACHE_SYMLINKS) + fc->cache_symlinks = 1; if (arg->flags & FUSE_ABORT_ERROR) fc->abort_err = 1; if (arg->flags & FUSE_MAX_PAGES) { @@ -966,7 +968,7 @@ static void fuse_send_init(struct fuse_conn *fc, struct fuse_req *req) FUSE_DO_READDIRPLUS | FUSE_READDIRPLUS_AUTO | FUSE_ASYNC_DIO | FUSE_WRITEBACK_CACHE | FUSE_NO_OPEN_SUPPORT | FUSE_PARALLEL_DIROPS | FUSE_HANDLE_KILLPRIV | FUSE_POSIX_ACL | - FUSE_ABORT_ERROR | FUSE_MAX_PAGES; + FUSE_ABORT_ERROR | FUSE_MAX_PAGES | FUSE_CACHE_SYMLINKS; req->in.h.opcode = FUSE_INIT; req->in.numargs = 1; req->in.args[0].size = sizeof(*arg); diff --git a/include/uapi/linux/fuse.h b/include/uapi/linux/fuse.h index 76f46f159992..b4967d48bfda 100644 --- a/include/uapi/linux/fuse.h +++ b/include/uapi/linux/fuse.h @@ -121,6 +121,7 @@ * - add FUSE_COPY_FILE_RANGE * - add FOPEN_CACHE_DIR * - add FUSE_MAX_PAGES, add max_pages to init_out + * - add FUSE_CACHE_SYMLINKS */ #ifndef _LINUX_FUSE_H @@ -257,6 +258,7 @@ struct fuse_file_lock { * FUSE_POSIX_ACL: filesystem supports posix acls * FUSE_ABORT_ERROR: reading the device after abort returns ECONNABORTED * FUSE_MAX_PAGES: init_out.max_pages contains the max number of req pages + * FUSE_CACHE_SYMLINKS: cache READLINK responses */ #define FUSE_ASYNC_READ (1 << 0) #define FUSE_POSIX_LOCKS (1 << 1) @@ -281,6 +283,7 @@ struct fuse_file_lock { #define FUSE_POSIX_ACL (1 << 20) #define FUSE_ABORT_ERROR (1 << 21) #define FUSE_MAX_PAGES (1 << 22) +#define FUSE_CACHE_SYMLINKS (1 << 23) /** * CUSE INIT request/reply flags -- cgit v1.3-14-g43fede From 86a559787e6f5cf662c081363f64a20cad654195 Mon Sep 17 00:00:00 2001 From: Wei Wang Date: Mon, 27 Aug 2018 09:32:17 +0800 Subject: virtio-balloon: VIRTIO_BALLOON_F_FREE_PAGE_HINT Negotiation of the VIRTIO_BALLOON_F_FREE_PAGE_HINT feature indicates the support of reporting hints of guest free pages to host via virtio-balloon. Currenlty, only free page blocks of MAX_ORDER - 1 are reported. They are obtained one by one from the mm free list via the regular allocation function. Host requests the guest to report free page hints by sending a new cmd id to the guest via the free_page_report_cmd_id configuration register. When the guest starts to report, it first sends a start cmd to host via the free page vq, which acks to host the cmd id received. When the guest finishes reporting free pages, a stop cmd is sent to host via the vq. Host may also send a stop cmd id to the guest to stop the reporting. VIRTIO_BALLOON_CMD_ID_STOP: Host sends this cmd to stop the guest reporting. VIRTIO_BALLOON_CMD_ID_DONE: Host sends this cmd to tell the guest that the reported pages are ready to be freed. Why does the guest free the reported pages when host tells it is ready to free? This is because freeing pages appears to be expensive for live migration. free_pages() dirties memory very quickly and makes the live migraion not converge in some cases. So it is good to delay the free_page operation when the migration is done, and host sends a command to guest about that. Why do we need the new VIRTIO_BALLOON_CMD_ID_DONE, instead of reusing VIRTIO_BALLOON_CMD_ID_STOP? This is because live migration is usually done in several rounds. At the end of each round, host needs to send a VIRTIO_BALLOON_CMD_ID_STOP cmd to the guest to stop (or say pause) the reporting. The guest resumes the reporting when it receives a new command id at the beginning of the next round. So we need a new cmd id to distinguish between "stop reporting" and "ready to free the reported pages". TODO: - Add a batch page allocation API to amortize the allocation overhead. Signed-off-by: Wei Wang Signed-off-by: Liang Li Cc: Michael S. Tsirkin Cc: Michal Hocko Cc: Andrew Morton Cc: Linus Torvalds Signed-off-by: Michael S. Tsirkin --- drivers/virtio/virtio_balloon.c | 364 ++++++++++++++++++++++++++++++++---- include/uapi/linux/virtio_balloon.h | 5 + 2 files changed, 336 insertions(+), 33 deletions(-) (limited to 'include/uapi/linux') diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c index d1c1f6283729..a18567889289 100644 --- a/drivers/virtio/virtio_balloon.c +++ b/drivers/virtio/virtio_balloon.c @@ -41,13 +41,34 @@ #define VIRTIO_BALLOON_ARRAY_PFNS_MAX 256 #define VIRTBALLOON_OOM_NOTIFY_PRIORITY 80 +#define VIRTIO_BALLOON_FREE_PAGE_ALLOC_FLAG (__GFP_NORETRY | __GFP_NOWARN | \ + __GFP_NOMEMALLOC) +/* The order of free page blocks to report to host */ +#define VIRTIO_BALLOON_FREE_PAGE_ORDER (MAX_ORDER - 1) +/* The size of a free page block in bytes */ +#define VIRTIO_BALLOON_FREE_PAGE_SIZE \ + (1 << (VIRTIO_BALLOON_FREE_PAGE_ORDER + PAGE_SHIFT)) + #ifdef CONFIG_BALLOON_COMPACTION static struct vfsmount *balloon_mnt; #endif +enum virtio_balloon_vq { + VIRTIO_BALLOON_VQ_INFLATE, + VIRTIO_BALLOON_VQ_DEFLATE, + VIRTIO_BALLOON_VQ_STATS, + VIRTIO_BALLOON_VQ_FREE_PAGE, + VIRTIO_BALLOON_VQ_MAX +}; + struct virtio_balloon { struct virtio_device *vdev; - struct virtqueue *inflate_vq, *deflate_vq, *stats_vq; + struct virtqueue *inflate_vq, *deflate_vq, *stats_vq, *free_page_vq; + + /* Balloon's own wq for cpu-intensive work items */ + struct workqueue_struct *balloon_wq; + /* The free page reporting work item submitted to the balloon wq */ + struct work_struct report_free_page_work; /* The balloon servicing is delegated to a freezable workqueue. */ struct work_struct update_balloon_stats_work; @@ -57,6 +78,18 @@ struct virtio_balloon { spinlock_t stop_update_lock; bool stop_update; + /* The list of allocated free pages, waiting to be given back to mm */ + struct list_head free_page_list; + spinlock_t free_page_list_lock; + /* The number of free page blocks on the above list */ + unsigned long num_free_page_blocks; + /* The cmd id received from host */ + u32 cmd_id_received; + /* The cmd id that is actively in use */ + __virtio32 cmd_id_active; + /* Buffer to store the stop sign */ + __virtio32 cmd_id_stop; + /* Waiting for host to ack the pages we released. */ wait_queue_head_t acked; @@ -320,17 +353,6 @@ static void stats_handle_request(struct virtio_balloon *vb) virtqueue_kick(vq); } -static void virtballoon_changed(struct virtio_device *vdev) -{ - struct virtio_balloon *vb = vdev->priv; - unsigned long flags; - - spin_lock_irqsave(&vb->stop_update_lock, flags); - if (!vb->stop_update) - queue_work(system_freezable_wq, &vb->update_balloon_size_work); - spin_unlock_irqrestore(&vb->stop_update_lock, flags); -} - static inline s64 towards_target(struct virtio_balloon *vb) { s64 target; @@ -347,6 +369,60 @@ static inline s64 towards_target(struct virtio_balloon *vb) return target - vb->num_pages; } +/* Gives back @num_to_return blocks of free pages to mm. */ +static unsigned long return_free_pages_to_mm(struct virtio_balloon *vb, + unsigned long num_to_return) +{ + struct page *page; + unsigned long num_returned; + + spin_lock_irq(&vb->free_page_list_lock); + for (num_returned = 0; num_returned < num_to_return; num_returned++) { + page = balloon_page_pop(&vb->free_page_list); + if (!page) + break; + free_pages((unsigned long)page_address(page), + VIRTIO_BALLOON_FREE_PAGE_ORDER); + } + vb->num_free_page_blocks -= num_returned; + spin_unlock_irq(&vb->free_page_list_lock); + + return num_returned; +} + +static void virtballoon_changed(struct virtio_device *vdev) +{ + struct virtio_balloon *vb = vdev->priv; + unsigned long flags; + s64 diff = towards_target(vb); + + if (diff) { + spin_lock_irqsave(&vb->stop_update_lock, flags); + if (!vb->stop_update) + queue_work(system_freezable_wq, + &vb->update_balloon_size_work); + spin_unlock_irqrestore(&vb->stop_update_lock, flags); + } + + if (virtio_has_feature(vdev, VIRTIO_BALLOON_F_FREE_PAGE_HINT)) { + virtio_cread(vdev, struct virtio_balloon_config, + free_page_report_cmd_id, &vb->cmd_id_received); + if (vb->cmd_id_received == VIRTIO_BALLOON_CMD_ID_DONE) { + /* Pass ULONG_MAX to give back all the free pages */ + return_free_pages_to_mm(vb, ULONG_MAX); + } else if (vb->cmd_id_received != VIRTIO_BALLOON_CMD_ID_STOP && + vb->cmd_id_received != + virtio32_to_cpu(vdev, vb->cmd_id_active)) { + spin_lock_irqsave(&vb->stop_update_lock, flags); + if (!vb->stop_update) { + queue_work(vb->balloon_wq, + &vb->report_free_page_work); + } + spin_unlock_irqrestore(&vb->stop_update_lock, flags); + } + } +} + static void update_balloon_size(struct virtio_balloon *vb) { u32 actual = vb->num_pages; @@ -389,26 +465,44 @@ static void update_balloon_size_func(struct work_struct *work) static int init_vqs(struct virtio_balloon *vb) { - struct virtqueue *vqs[3]; - vq_callback_t *callbacks[] = { balloon_ack, balloon_ack, stats_request }; - static const char * const names[] = { "inflate", "deflate", "stats" }; - int err, nvqs; + struct virtqueue *vqs[VIRTIO_BALLOON_VQ_MAX]; + vq_callback_t *callbacks[VIRTIO_BALLOON_VQ_MAX]; + const char *names[VIRTIO_BALLOON_VQ_MAX]; + int err; /* - * We expect two virtqueues: inflate and deflate, and - * optionally stat. + * Inflateq and deflateq are used unconditionally. The names[] + * will be NULL if the related feature is not enabled, which will + * cause no allocation for the corresponding virtqueue in find_vqs. */ - nvqs = virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_STATS_VQ) ? 3 : 2; - err = virtio_find_vqs(vb->vdev, nvqs, vqs, callbacks, names, NULL); + callbacks[VIRTIO_BALLOON_VQ_INFLATE] = balloon_ack; + names[VIRTIO_BALLOON_VQ_INFLATE] = "inflate"; + callbacks[VIRTIO_BALLOON_VQ_DEFLATE] = balloon_ack; + names[VIRTIO_BALLOON_VQ_DEFLATE] = "deflate"; + names[VIRTIO_BALLOON_VQ_STATS] = NULL; + names[VIRTIO_BALLOON_VQ_FREE_PAGE] = NULL; + + if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_STATS_VQ)) { + names[VIRTIO_BALLOON_VQ_STATS] = "stats"; + callbacks[VIRTIO_BALLOON_VQ_STATS] = stats_request; + } + + if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_FREE_PAGE_HINT)) { + names[VIRTIO_BALLOON_VQ_FREE_PAGE] = "free_page_vq"; + callbacks[VIRTIO_BALLOON_VQ_FREE_PAGE] = NULL; + } + + err = vb->vdev->config->find_vqs(vb->vdev, VIRTIO_BALLOON_VQ_MAX, + vqs, callbacks, names, NULL, NULL); if (err) return err; - vb->inflate_vq = vqs[0]; - vb->deflate_vq = vqs[1]; + vb->inflate_vq = vqs[VIRTIO_BALLOON_VQ_INFLATE]; + vb->deflate_vq = vqs[VIRTIO_BALLOON_VQ_DEFLATE]; if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_STATS_VQ)) { struct scatterlist sg; unsigned int num_stats; - vb->stats_vq = vqs[2]; + vb->stats_vq = vqs[VIRTIO_BALLOON_VQ_STATS]; /* * Prime this virtqueue with one buffer so the hypervisor can @@ -426,9 +520,145 @@ static int init_vqs(struct virtio_balloon *vb) } virtqueue_kick(vb->stats_vq); } + + if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_FREE_PAGE_HINT)) + vb->free_page_vq = vqs[VIRTIO_BALLOON_VQ_FREE_PAGE]; + + return 0; +} + +static int send_cmd_id_start(struct virtio_balloon *vb) +{ + struct scatterlist sg; + struct virtqueue *vq = vb->free_page_vq; + int err, unused; + + /* Detach all the used buffers from the vq */ + while (virtqueue_get_buf(vq, &unused)) + ; + + vb->cmd_id_active = cpu_to_virtio32(vb->vdev, vb->cmd_id_received); + sg_init_one(&sg, &vb->cmd_id_active, sizeof(vb->cmd_id_active)); + err = virtqueue_add_outbuf(vq, &sg, 1, &vb->cmd_id_active, GFP_KERNEL); + if (!err) + virtqueue_kick(vq); + return err; +} + +static int send_cmd_id_stop(struct virtio_balloon *vb) +{ + struct scatterlist sg; + struct virtqueue *vq = vb->free_page_vq; + int err, unused; + + /* Detach all the used buffers from the vq */ + while (virtqueue_get_buf(vq, &unused)) + ; + + sg_init_one(&sg, &vb->cmd_id_stop, sizeof(vb->cmd_id_stop)); + err = virtqueue_add_outbuf(vq, &sg, 1, &vb->cmd_id_stop, GFP_KERNEL); + if (!err) + virtqueue_kick(vq); + return err; +} + +static int get_free_page_and_send(struct virtio_balloon *vb) +{ + struct virtqueue *vq = vb->free_page_vq; + struct page *page; + struct scatterlist sg; + int err, unused; + void *p; + + /* Detach all the used buffers from the vq */ + while (virtqueue_get_buf(vq, &unused)) + ; + + page = alloc_pages(VIRTIO_BALLOON_FREE_PAGE_ALLOC_FLAG, + VIRTIO_BALLOON_FREE_PAGE_ORDER); + /* + * When the allocation returns NULL, it indicates that we have got all + * the possible free pages, so return -EINTR to stop. + */ + if (!page) + return -EINTR; + + p = page_address(page); + sg_init_one(&sg, p, VIRTIO_BALLOON_FREE_PAGE_SIZE); + /* There is always 1 entry reserved for the cmd id to use. */ + if (vq->num_free > 1) { + err = virtqueue_add_inbuf(vq, &sg, 1, p, GFP_KERNEL); + if (unlikely(err)) { + free_pages((unsigned long)p, + VIRTIO_BALLOON_FREE_PAGE_ORDER); + return err; + } + virtqueue_kick(vq); + spin_lock_irq(&vb->free_page_list_lock); + balloon_page_push(&vb->free_page_list, page); + vb->num_free_page_blocks++; + spin_unlock_irq(&vb->free_page_list_lock); + } else { + /* + * The vq has no available entry to add this page block, so + * just free it. + */ + free_pages((unsigned long)p, VIRTIO_BALLOON_FREE_PAGE_ORDER); + } + return 0; } +static int send_free_pages(struct virtio_balloon *vb) +{ + int err; + u32 cmd_id_active; + + while (1) { + /* + * If a stop id or a new cmd id was just received from host, + * stop the reporting. + */ + cmd_id_active = virtio32_to_cpu(vb->vdev, vb->cmd_id_active); + if (cmd_id_active != vb->cmd_id_received) + break; + + /* + * The free page blocks are allocated and sent to host one by + * one. + */ + err = get_free_page_and_send(vb); + if (err == -EINTR) + break; + else if (unlikely(err)) + return err; + } + + return 0; +} + +static void report_free_page_func(struct work_struct *work) +{ + int err; + struct virtio_balloon *vb = container_of(work, struct virtio_balloon, + report_free_page_work); + struct device *dev = &vb->vdev->dev; + + /* Start by sending the received cmd id to host with an outbuf. */ + err = send_cmd_id_start(vb); + if (unlikely(err)) + dev_err(dev, "Failed to send a start id, err = %d\n", err); + + err = send_free_pages(vb); + if (unlikely(err)) + dev_err(dev, "Failed to send a free page, err = %d\n", err); + + /* End by sending a stop id to host with an outbuf. */ + err = send_cmd_id_stop(vb); + if (unlikely(err)) + dev_err(dev, "Failed to send a stop id, err = %d\n", err); +} + #ifdef CONFIG_BALLOON_COMPACTION /* * virtballoon_migratepage - perform the balloon page migration on behalf of @@ -512,14 +742,23 @@ static struct file_system_type balloon_fs = { #endif /* CONFIG_BALLOON_COMPACTION */ -static unsigned long virtio_balloon_shrinker_scan(struct shrinker *shrinker, - struct shrink_control *sc) +static unsigned long shrink_free_pages(struct virtio_balloon *vb, + unsigned long pages_to_free) { - unsigned long pages_to_free, pages_freed = 0; - struct virtio_balloon *vb = container_of(shrinker, - struct virtio_balloon, shrinker); + unsigned long blocks_to_free, blocks_freed; - pages_to_free = sc->nr_to_scan * VIRTIO_BALLOON_PAGES_PER_PAGE; + pages_to_free = round_up(pages_to_free, + 1 << VIRTIO_BALLOON_FREE_PAGE_ORDER); + blocks_to_free = pages_to_free >> VIRTIO_BALLOON_FREE_PAGE_ORDER; + blocks_freed = return_free_pages_to_mm(vb, blocks_to_free); + + return blocks_freed << VIRTIO_BALLOON_FREE_PAGE_ORDER; +} + +static unsigned long shrink_balloon_pages(struct virtio_balloon *vb, + unsigned long pages_to_free) +{ + unsigned long pages_freed = 0; /* * One invocation of leak_balloon can deflate at most @@ -527,12 +766,33 @@ static unsigned long virtio_balloon_shrinker_scan(struct shrinker *shrinker, * multiple times to deflate pages till reaching pages_to_free. */ while (vb->num_pages && pages_to_free) { + pages_freed += leak_balloon(vb, pages_to_free) / + VIRTIO_BALLOON_PAGES_PER_PAGE; pages_to_free -= pages_freed; - pages_freed += leak_balloon(vb, pages_to_free); } update_balloon_size(vb); - return pages_freed / VIRTIO_BALLOON_PAGES_PER_PAGE; + return pages_freed; +} + +static unsigned long virtio_balloon_shrinker_scan(struct shrinker *shrinker, + struct shrink_control *sc) +{ + unsigned long pages_to_free, pages_freed = 0; + struct virtio_balloon *vb = container_of(shrinker, + struct virtio_balloon, shrinker); + + pages_to_free = sc->nr_to_scan * VIRTIO_BALLOON_PAGES_PER_PAGE; + + if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_FREE_PAGE_HINT)) + pages_freed = shrink_free_pages(vb, pages_to_free); + + if (pages_freed >= pages_to_free) + return pages_freed; + + pages_freed += shrink_balloon_pages(vb, pages_to_free - pages_freed); + + return pages_freed; } static unsigned long virtio_balloon_shrinker_count(struct shrinker *shrinker, @@ -540,8 +800,12 @@ static unsigned long virtio_balloon_shrinker_count(struct shrinker *shrinker, { struct virtio_balloon *vb = container_of(shrinker, struct virtio_balloon, shrinker); + unsigned long count; - return vb->num_pages / VIRTIO_BALLOON_PAGES_PER_PAGE; + count = vb->num_pages / VIRTIO_BALLOON_PAGES_PER_PAGE; + count += vb->num_free_page_blocks >> VIRTIO_BALLOON_FREE_PAGE_ORDER; + + return count; } static void virtio_balloon_unregister_shrinker(struct virtio_balloon *vb) @@ -604,6 +868,31 @@ static int virtballoon_probe(struct virtio_device *vdev) } vb->vb_dev_info.inode->i_mapping->a_ops = &balloon_aops; #endif + if (virtio_has_feature(vdev, VIRTIO_BALLOON_F_FREE_PAGE_HINT)) { + /* + * There is always one entry reserved for cmd id, so the ring + * size needs to be at least two to report free page hints. + */ + if (virtqueue_get_vring_size(vb->free_page_vq) < 2) { + err = -ENOSPC; + goto out_del_vqs; + } + vb->balloon_wq = alloc_workqueue("balloon-wq", + WQ_FREEZABLE | WQ_CPU_INTENSIVE, 0); + if (!vb->balloon_wq) { + err = -ENOMEM; + goto out_del_vqs; + } + INIT_WORK(&vb->report_free_page_work, report_free_page_func); + vb->cmd_id_received = VIRTIO_BALLOON_CMD_ID_STOP; + vb->cmd_id_active = cpu_to_virtio32(vb->vdev, + VIRTIO_BALLOON_CMD_ID_STOP); + vb->cmd_id_stop = cpu_to_virtio32(vb->vdev, + VIRTIO_BALLOON_CMD_ID_STOP); + vb->num_free_page_blocks = 0; + spin_lock_init(&vb->free_page_list_lock); + INIT_LIST_HEAD(&vb->free_page_list); + } /* * We continue to use VIRTIO_BALLOON_F_DEFLATE_ON_OOM to decide if a * shrinker needs to be registered to relieve memory pressure. @@ -611,7 +900,7 @@ static int virtballoon_probe(struct virtio_device *vdev) if (virtio_has_feature(vb->vdev, VIRTIO_BALLOON_F_DEFLATE_ON_OOM)) { err = virtio_balloon_register_shrinker(vb); if (err) - goto out_del_vqs; + goto out_del_balloon_wq; } virtio_device_ready(vdev); @@ -619,6 +908,9 @@ static int virtballoon_probe(struct virtio_device *vdev) virtballoon_changed(vdev); return 0; +out_del_balloon_wq: + if (virtio_has_feature(vdev, VIRTIO_BALLOON_F_FREE_PAGE_HINT)) + destroy_workqueue(vb->balloon_wq); out_del_vqs: vdev->config->del_vqs(vdev); out_free_vb: @@ -652,6 +944,11 @@ static void virtballoon_remove(struct virtio_device *vdev) cancel_work_sync(&vb->update_balloon_size_work); cancel_work_sync(&vb->update_balloon_stats_work); + if (virtio_has_feature(vdev, VIRTIO_BALLOON_F_FREE_PAGE_HINT)) { + cancel_work_sync(&vb->report_free_page_work); + destroy_workqueue(vb->balloon_wq); + } + remove_common(vb); #ifdef CONFIG_BALLOON_COMPACTION if (vb->vb_dev_info.inode) @@ -703,6 +1000,7 @@ static unsigned int features[] = { VIRTIO_BALLOON_F_MUST_TELL_HOST, VIRTIO_BALLOON_F_STATS_VQ, VIRTIO_BALLOON_F_DEFLATE_ON_OOM, + VIRTIO_BALLOON_F_FREE_PAGE_HINT, }; static struct virtio_driver virtio_balloon_driver = { diff --git a/include/uapi/linux/virtio_balloon.h b/include/uapi/linux/virtio_balloon.h index 13b8cb563892..47c9eb401c08 100644 --- a/include/uapi/linux/virtio_balloon.h +++ b/include/uapi/linux/virtio_balloon.h @@ -34,15 +34,20 @@ #define VIRTIO_BALLOON_F_MUST_TELL_HOST 0 /* Tell before reclaiming pages */ #define VIRTIO_BALLOON_F_STATS_VQ 1 /* Memory Stats virtqueue */ #define VIRTIO_BALLOON_F_DEFLATE_ON_OOM 2 /* Deflate balloon on OOM */ +#define VIRTIO_BALLOON_F_FREE_PAGE_HINT 3 /* VQ to report free pages */ /* Size of a PFN in the balloon interface. */ #define VIRTIO_BALLOON_PFN_SHIFT 12 +#define VIRTIO_BALLOON_CMD_ID_STOP 0 +#define VIRTIO_BALLOON_CMD_ID_DONE 1 struct virtio_balloon_config { /* Number of pages host wants Guest to give up. */ __u32 num_pages; /* Number of pages we've actually got in balloon. */ __u32 actual; + /* Free page report command id, readonly by guest */ + __u32 free_page_report_cmd_id; }; #define VIRTIO_BALLOON_S_SWAP_IN 0 /* Amount of memory swapped in */ -- cgit v1.3-14-g43fede From 2e991629bcf55a43681aec1ee096eeb03cf81709 Mon Sep 17 00:00:00 2001 From: Wei Wang Date: Mon, 27 Aug 2018 09:32:19 +0800 Subject: virtio-balloon: VIRTIO_BALLOON_F_PAGE_POISON The VIRTIO_BALLOON_F_PAGE_POISON feature bit is used to indicate if the guest is using page poisoning. Guest writes to the poison_val config field to tell host about the page poisoning value that is in use. Suggested-by: Michael S. Tsirkin Signed-off-by: Wei Wang Cc: Michael S. Tsirkin Cc: Michal Hocko Cc: Andrew Morton Signed-off-by: Michael S. Tsirkin --- drivers/virtio/virtio_balloon.c | 10 ++++++++++ include/uapi/linux/virtio_balloon.h | 3 +++ 2 files changed, 13 insertions(+) (limited to 'include/uapi/linux') diff --git a/drivers/virtio/virtio_balloon.c b/drivers/virtio/virtio_balloon.c index a18567889289..728ecd1eea30 100644 --- a/drivers/virtio/virtio_balloon.c +++ b/drivers/virtio/virtio_balloon.c @@ -825,6 +825,7 @@ static int virtio_balloon_register_shrinker(struct virtio_balloon *vb) static int virtballoon_probe(struct virtio_device *vdev) { struct virtio_balloon *vb; + __u32 poison_val; int err; if (!vdev->config->get) { @@ -892,6 +893,11 @@ static int virtballoon_probe(struct virtio_device *vdev) vb->num_free_page_blocks = 0; spin_lock_init(&vb->free_page_list_lock); INIT_LIST_HEAD(&vb->free_page_list); + if (virtio_has_feature(vdev, VIRTIO_BALLOON_F_PAGE_POISON)) { + memset(&poison_val, PAGE_POISON, sizeof(poison_val)); + virtio_cwrite(vb->vdev, struct virtio_balloon_config, + poison_val, &poison_val); + } } /* * We continue to use VIRTIO_BALLOON_F_DEFLATE_ON_OOM to decide if a @@ -992,6 +998,9 @@ static int virtballoon_restore(struct virtio_device *vdev) static int virtballoon_validate(struct virtio_device *vdev) { + if (!page_poisoning_enabled()) + __virtio_clear_bit(vdev, VIRTIO_BALLOON_F_PAGE_POISON); + __virtio_clear_bit(vdev, VIRTIO_F_IOMMU_PLATFORM); return 0; } @@ -1001,6 +1010,7 @@ static unsigned int features[] = { VIRTIO_BALLOON_F_STATS_VQ, VIRTIO_BALLOON_F_DEFLATE_ON_OOM, VIRTIO_BALLOON_F_FREE_PAGE_HINT, + VIRTIO_BALLOON_F_PAGE_POISON, }; static struct virtio_driver virtio_balloon_driver = { diff --git a/include/uapi/linux/virtio_balloon.h b/include/uapi/linux/virtio_balloon.h index 47c9eb401c08..a1966cd7b677 100644 --- a/include/uapi/linux/virtio_balloon.h +++ b/include/uapi/linux/virtio_balloon.h @@ -35,6 +35,7 @@ #define VIRTIO_BALLOON_F_STATS_VQ 1 /* Memory Stats virtqueue */ #define VIRTIO_BALLOON_F_DEFLATE_ON_OOM 2 /* Deflate balloon on OOM */ #define VIRTIO_BALLOON_F_FREE_PAGE_HINT 3 /* VQ to report free pages */ +#define VIRTIO_BALLOON_F_PAGE_POISON 4 /* Guest is using page poisoning */ /* Size of a PFN in the balloon interface. */ #define VIRTIO_BALLOON_PFN_SHIFT 12 @@ -48,6 +49,8 @@ struct virtio_balloon_config { __u32 actual; /* Free page report command id, readonly by guest */ __u32 free_page_report_cmd_id; + /* Stores PAGE_POISON if page poisoning is in use */ + __u32 poison_val; }; #define VIRTIO_BALLOON_S_SWAP_IN 0 /* Amount of memory swapped in */ -- cgit v1.3-14-g43fede From 70025f84e5b79627a6739533c4fe7cef5b605886 Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 9 Oct 2018 17:46:51 +0100 Subject: KEYS: Provide key type operations for asymmetric key ops [ver #2] Provide five new operations in the key_type struct that can be used to provide access to asymmetric key operations. These will be implemented for the asymmetric key type in a later patch and may refer to a key retained in RAM by the kernel or a key retained in crypto hardware. int (*asym_query)(const struct kernel_pkey_params *params, struct kernel_pkey_query *info); int (*asym_eds_op)(struct kernel_pkey_params *params, const void *in, void *out); int (*asym_verify_signature)(struct kernel_pkey_params *params, const void *in, const void *in2); Since encrypt, decrypt and sign are identical in their interfaces, they're rolled together in the asym_eds_op() operation and there's an operation ID in the params argument to distinguish them. Verify is different in that we supply the data and the signature instead and get an error value (or 0) as the only result on the expectation that this may well be how a hardware crypto device may work. Signed-off-by: David Howells Tested-by: Marcel Holtmann Reviewed-by: Marcel Holtmann Reviewed-by: Denis Kenzior Tested-by: Denis Kenzior Signed-off-by: James Morris --- Documentation/security/keys/core.rst | 106 +++++++++++++++++++++++++++++++++++ include/linux/key-type.h | 11 ++++ include/linux/keyctl.h | 46 +++++++++++++++ include/uapi/linux/keyctl.h | 5 ++ 4 files changed, 168 insertions(+) create mode 100644 include/linux/keyctl.h (limited to 'include/uapi/linux') diff --git a/Documentation/security/keys/core.rst b/Documentation/security/keys/core.rst index 9ce7256c6edb..c144978479d5 100644 --- a/Documentation/security/keys/core.rst +++ b/Documentation/security/keys/core.rst @@ -1483,6 +1483,112 @@ The structure has a number of fields, some of which are mandatory: attempted key link operation. If there is no match, -EINVAL is returned. + * ``int (*asym_eds_op)(struct kernel_pkey_params *params, + const void *in, void *out);`` + ``int (*asym_verify_signature)(struct kernel_pkey_params *params, + const void *in, const void *in2);`` + + These methods are optional. If provided the first allows a key to be + used to encrypt, decrypt or sign a blob of data, and the second allows a + key to verify a signature. + + In all cases, the following information is provided in the params block:: + + struct kernel_pkey_params { + struct key *key; + const char *encoding; + const char *hash_algo; + char *info; + __u32 in_len; + union { + __u32 out_len; + __u32 in2_len; + }; + enum kernel_pkey_operation op : 8; + }; + + This includes the key to be used; a string indicating the encoding to use + (for instance, "pkcs1" may be used with an RSA key to indicate + RSASSA-PKCS1-v1.5 or RSAES-PKCS1-v1.5 encoding or "raw" if no encoding); + the name of the hash algorithm used to generate the data for a signature + (if appropriate); the sizes of the input and output (or second input) + buffers; and the ID of the operation to be performed. + + For a given operation ID, the input and output buffers are used as + follows:: + + Operation ID in,in_len out,out_len in2,in2_len + ======================= =============== =============== =============== + kernel_pkey_encrypt Raw data Encrypted data - + kernel_pkey_decrypt Encrypted data Raw data - + kernel_pkey_sign Raw data Signature - + kernel_pkey_verify Raw data - Signature + + asym_eds_op() deals with encryption, decryption and signature creation as + specified by params->op. Note that params->op is also set for + asym_verify_signature(). + + Encrypting and signature creation both take raw data in the input buffer + and return the encrypted result in the output buffer. Padding may have + been added if an encoding was set. In the case of signature creation, + depending on the encoding, the padding created may need to indicate the + digest algorithm - the name of which should be supplied in hash_algo. + + Decryption takes encrypted data in the input buffer and returns the raw + data in the output buffer. Padding will get checked and stripped off if + an encoding was set. + + Verification takes raw data in the input buffer and the signature in the + second input buffer and checks that the one matches the other. Padding + will be validated. Depending on the encoding, the digest algorithm used + to generate the raw data may need to be indicated in hash_algo. + + If successful, asym_eds_op() should return the number of bytes written + into the output buffer. asym_verify_signature() should return 0. + + A variety of errors may be returned, including EOPNOTSUPP if the operation + is not supported; EKEYREJECTED if verification fails; ENOPKG if the + required crypto isn't available. + + + * ``int (*asym_query)(const struct kernel_pkey_params *params, + struct kernel_pkey_query *info);`` + + This method is optional. If provided it allows information about the + public or asymmetric key held in the key to be determined. + + The parameter block is as for asym_eds_op() and co. but in_len and out_len + are unused. The encoding and hash_algo fields should be used to reduce + the returned buffer/data sizes as appropriate. + + If successful, the following information is filled in:: + + struct kernel_pkey_query { + __u32 supported_ops; + __u32 key_size; + __u16 max_data_size; + __u16 max_sig_size; + __u16 max_enc_size; + __u16 max_dec_size; + }; + + The supported_ops field will contain a bitmask indicating what operations + are supported by the key, including encryption of a blob, decryption of a + blob, signing a blob and verifying the signature on a blob. The following + constants are defined for this:: + + KEYCTL_SUPPORTS_{ENCRYPT,DECRYPT,SIGN,VERIFY} + + The key_size field is the size of the key in bits. max_data_size and + max_sig_size are the maximum raw data and signature sizes for creation and + verification of a signature; max_enc_size and max_dec_size are the maximum + raw data and signature sizes for encryption and decryption. The + max_*_size fields are measured in bytes. + + If successful, 0 will be returned. If the key doesn't support this, + EOPNOTSUPP will be returned. + + Request-Key Callback Service ============================ diff --git a/include/linux/key-type.h b/include/linux/key-type.h index 05d8fb5a06c4..bc9af551fc83 100644 --- a/include/linux/key-type.h +++ b/include/linux/key-type.h @@ -17,6 +17,9 @@ #ifdef CONFIG_KEYS +struct kernel_pkey_query; +struct kernel_pkey_params; + /* * key under-construction record * - passed to the request_key actor if supplied @@ -155,6 +158,14 @@ struct key_type { */ struct key_restriction *(*lookup_restriction)(const char *params); + /* Asymmetric key accessor functions. */ + int (*asym_query)(const struct kernel_pkey_params *params, + struct kernel_pkey_query *info); + int (*asym_eds_op)(struct kernel_pkey_params *params, + const void *in, void *out); + int (*asym_verify_signature)(struct kernel_pkey_params *params, + const void *in, const void *in2); + /* internal fields */ struct list_head link; /* link in types list */ struct lock_class_key lock_class; /* key->sem lock class */ diff --git a/include/linux/keyctl.h b/include/linux/keyctl.h new file mode 100644 index 000000000000..c7c48c79ce0e --- /dev/null +++ b/include/linux/keyctl.h @@ -0,0 +1,46 @@ +/* keyctl kernel bits + * + * Copyright (C) 2016 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public Licence + * as published by the Free Software Foundation; either version + * 2 of the Licence, or (at your option) any later version. + */ + +#ifndef __LINUX_KEYCTL_H +#define __LINUX_KEYCTL_H + +#include + +struct kernel_pkey_query { + __u32 supported_ops; /* Which ops are supported */ + __u32 key_size; /* Size of the key in bits */ + __u16 max_data_size; /* Maximum size of raw data to sign in bytes */ + __u16 max_sig_size; /* Maximum size of signature in bytes */ + __u16 max_enc_size; /* Maximum size of encrypted blob in bytes */ + __u16 max_dec_size; /* Maximum size of decrypted blob in bytes */ +}; + +enum kernel_pkey_operation { + kernel_pkey_encrypt, + kernel_pkey_decrypt, + kernel_pkey_sign, + kernel_pkey_verify, +}; + +struct kernel_pkey_params { + struct key *key; + const char *encoding; /* Encoding (eg. "oaep" or "raw" for none) */ + const char *hash_algo; /* Digest algorithm used (eg. "sha1") or NULL if N/A */ + char *info; /* Modified info string to be released later */ + __u32 in_len; /* Input data size */ + union { + __u32 out_len; /* Output buffer size (enc/dec/sign) */ + __u32 in2_len; /* 2nd input data size (verify) */ + }; + enum kernel_pkey_operation op : 8; +}; + +#endif /* __LINUX_KEYCTL_H */ diff --git a/include/uapi/linux/keyctl.h b/include/uapi/linux/keyctl.h index 0f3cb13db8e9..1d1e9f2877af 100644 --- a/include/uapi/linux/keyctl.h +++ b/include/uapi/linux/keyctl.h @@ -82,4 +82,9 @@ struct keyctl_kdf_params { __u32 __spare[8]; }; +#define KEYCTL_SUPPORTS_ENCRYPT 0x01 +#define KEYCTL_SUPPORTS_DECRYPT 0x02 +#define KEYCTL_SUPPORTS_SIGN 0x04 +#define KEYCTL_SUPPORTS_VERIFY 0x08 + #endif /* _LINUX_KEYCTL_H */ -- cgit v1.3-14-g43fede From 00d60fd3b93219ea854220f0fd264b86398cbc53 Mon Sep 17 00:00:00 2001 From: David Howells Date: Tue, 9 Oct 2018 17:46:59 +0100 Subject: KEYS: Provide keyctls to drive the new key type ops for asymmetric keys [ver #2] Provide five keyctl functions that permit userspace to make use of the new key type ops for accessing and driving asymmetric keys. (*) Query an asymmetric key. long keyctl(KEYCTL_PKEY_QUERY, key_serial_t key, unsigned long reserved, struct keyctl_pkey_query *info); Get information about an asymmetric key. The information is returned in the keyctl_pkey_query struct: __u32 supported_ops; A bit mask of flags indicating which ops are supported. This is constructed from a bitwise-OR of: KEYCTL_SUPPORTS_{ENCRYPT,DECRYPT,SIGN,VERIFY} __u32 key_size; The size in bits of the key. __u16 max_data_size; __u16 max_sig_size; __u16 max_enc_size; __u16 max_dec_size; The maximum sizes in bytes of a blob of data to be signed, a signature blob, a blob to be encrypted and a blob to be decrypted. reserved must be set to 0. This is intended for future use to hand over one or more passphrases needed unlock a key. If successful, 0 is returned. If the key is not an asymmetric key, EOPNOTSUPP is returned. (*) Encrypt, decrypt, sign or verify a blob using an asymmetric key. long keyctl(KEYCTL_PKEY_ENCRYPT, const struct keyctl_pkey_params *params, const char *info, const void *in, void *out); long keyctl(KEYCTL_PKEY_DECRYPT, const struct keyctl_pkey_params *params, const char *info, const void *in, void *out); long keyctl(KEYCTL_PKEY_SIGN, const struct keyctl_pkey_params *params, const char *info, const void *in, void *out); long keyctl(KEYCTL_PKEY_VERIFY, const struct keyctl_pkey_params *params, const char *info, const void *in, const void *in2); Use an asymmetric key to perform a public-key cryptographic operation a blob of data. The parameter block pointed to by params contains a number of integer values: __s32 key_id; __u32 in_len; __u32 out_len; __u32 in2_len; For a given operation, the in and out buffers are used as follows: Operation ID in,in_len out,out_len in2,in2_len ======================= =============== =============== =========== KEYCTL_PKEY_ENCRYPT Raw data Encrypted data - KEYCTL_PKEY_DECRYPT Encrypted data Raw data - KEYCTL_PKEY_SIGN Raw data Signature - KEYCTL_PKEY_VERIFY Raw data - Signature info is a string of key=value pairs that supply supplementary information. The __spare space in the parameter block must be set to 0. This is intended, amongst other things, to allow the passing of passphrases required to unlock a key. If successful, encrypt, decrypt and sign all return the amount of data written into the output buffer. Verification returns 0 on success. Signed-off-by: David Howells Tested-by: Marcel Holtmann Reviewed-by: Marcel Holtmann Reviewed-by: Denis Kenzior Tested-by: Denis Kenzior Signed-off-by: James Morris --- Documentation/security/keys/core.rst | 111 ++++++++++++ include/uapi/linux/keyctl.h | 25 +++ security/keys/Makefile | 1 + security/keys/compat.c | 18 ++ security/keys/internal.h | 39 +++++ security/keys/keyctl.c | 24 +++ security/keys/keyctl_pkey.c | 323 +++++++++++++++++++++++++++++++++++ 7 files changed, 541 insertions(+) create mode 100644 security/keys/keyctl_pkey.c (limited to 'include/uapi/linux') diff --git a/Documentation/security/keys/core.rst b/Documentation/security/keys/core.rst index c144978479d5..9521c4207f01 100644 --- a/Documentation/security/keys/core.rst +++ b/Documentation/security/keys/core.rst @@ -859,6 +859,7 @@ The keyctl syscall functions are: and either the buffer length or the OtherInfo length exceeds the allowed length. + * Restrict keyring linkage:: long keyctl(KEYCTL_RESTRICT_KEYRING, key_serial_t keyring, @@ -890,6 +891,116 @@ The keyctl syscall functions are: applicable to the asymmetric key type. + * Query an asymmetric key:: + + long keyctl(KEYCTL_PKEY_QUERY, + key_serial_t key_id, unsigned long reserved, + struct keyctl_pkey_query *info); + + Get information about an asymmetric key. The information is returned in + the keyctl_pkey_query struct:: + + __u32 supported_ops; + __u32 key_size; + __u16 max_data_size; + __u16 max_sig_size; + __u16 max_enc_size; + __u16 max_dec_size; + __u32 __spare[10]; + + ``supported_ops`` contains a bit mask of flags indicating which ops are + supported. This is constructed from a bitwise-OR of:: + + KEYCTL_SUPPORTS_{ENCRYPT,DECRYPT,SIGN,VERIFY} + + ``key_size`` indicated the size of the key in bits. + + ``max_*_size`` indicate the maximum sizes in bytes of a blob of data to be + signed, a signature blob, a blob to be encrypted and a blob to be + decrypted. + + ``__spare[]`` must be set to 0. This is intended for future use to hand + over one or more passphrases needed unlock a key. + + If successful, 0 is returned. If the key is not an asymmetric key, + EOPNOTSUPP is returned. + + + * Encrypt, decrypt, sign or verify a blob using an asymmetric key:: + + long keyctl(KEYCTL_PKEY_ENCRYPT, + const struct keyctl_pkey_params *params, + const char *info, + const void *in, + void *out); + + long keyctl(KEYCTL_PKEY_DECRYPT, + const struct keyctl_pkey_params *params, + const char *info, + const void *in, + void *out); + + long keyctl(KEYCTL_PKEY_SIGN, + const struct keyctl_pkey_params *params, + const char *info, + const void *in, + void *out); + + long keyctl(KEYCTL_PKEY_VERIFY, + const struct keyctl_pkey_params *params, + const char *info, + const void *in, + const void *in2); + + Use an asymmetric key to perform a public-key cryptographic operation a + blob of data. For encryption and verification, the asymmetric key may + only need the public parts to be available, but for decryption and signing + the private parts are required also. + + The parameter block pointed to by params contains a number of integer + values:: + + __s32 key_id; + __u32 in_len; + __u32 out_len; + __u32 in2_len; + + ``key_id`` is the ID of the asymmetric key to be used. ``in_len`` and + ``in2_len`` indicate the amount of data in the in and in2 buffers and + ``out_len`` indicates the size of the out buffer as appropriate for the + above operations. + + For a given operation, the in and out buffers are used as follows:: + + Operation ID in,in_len out,out_len in2,in2_len + ======================= =============== =============== =============== + KEYCTL_PKEY_ENCRYPT Raw data Encrypted data - + KEYCTL_PKEY_DECRYPT Encrypted data Raw data - + KEYCTL_PKEY_SIGN Raw data Signature - + KEYCTL_PKEY_VERIFY Raw data - Signature + + ``info`` is a string of key=value pairs that supply supplementary + information. These include: + + ``enc=`` The encoding of the encrypted/signature blob. This + can be "pkcs1" for RSASSA-PKCS1-v1.5 or + RSAES-PKCS1-v1.5; "pss" for "RSASSA-PSS"; "oaep" for + "RSAES-OAEP". If omitted or is "raw", the raw output + of the encryption function is specified. + + ``hash=`` If the data buffer contains the output of a hash + function and the encoding includes some indication of + which hash function was used, the hash function can be + specified with this, eg. "hash=sha256". + + The ``__spare[]`` space in the parameter block must be set to 0. This is + intended, amongst other things, to allow the passing of passphrases + required to unlock a key. + + If successful, encrypt, decrypt and sign all return the amount of data + written into the output buffer. Verification returns 0 on success. + + Kernel Services =============== diff --git a/include/uapi/linux/keyctl.h b/include/uapi/linux/keyctl.h index 1d1e9f2877af..f45ee0f69c0c 100644 --- a/include/uapi/linux/keyctl.h +++ b/include/uapi/linux/keyctl.h @@ -61,6 +61,11 @@ #define KEYCTL_INVALIDATE 21 /* invalidate a key */ #define KEYCTL_GET_PERSISTENT 22 /* get a user's persistent keyring */ #define KEYCTL_DH_COMPUTE 23 /* Compute Diffie-Hellman values */ +#define KEYCTL_PKEY_QUERY 24 /* Query public key parameters */ +#define KEYCTL_PKEY_ENCRYPT 25 /* Encrypt a blob using a public key */ +#define KEYCTL_PKEY_DECRYPT 26 /* Decrypt a blob using a public key */ +#define KEYCTL_PKEY_SIGN 27 /* Create a public key signature */ +#define KEYCTL_PKEY_VERIFY 28 /* Verify a public key signature */ #define KEYCTL_RESTRICT_KEYRING 29 /* Restrict keys allowed to link to a keyring */ /* keyctl structures */ @@ -87,4 +92,24 @@ struct keyctl_kdf_params { #define KEYCTL_SUPPORTS_SIGN 0x04 #define KEYCTL_SUPPORTS_VERIFY 0x08 +struct keyctl_pkey_query { + __u32 supported_ops; /* Which ops are supported */ + __u32 key_size; /* Size of the key in bits */ + __u16 max_data_size; /* Maximum size of raw data to sign in bytes */ + __u16 max_sig_size; /* Maximum size of signature in bytes */ + __u16 max_enc_size; /* Maximum size of encrypted blob in bytes */ + __u16 max_dec_size; /* Maximum size of decrypted blob in bytes */ + __u32 __spare[10]; +}; + +struct keyctl_pkey_params { + __s32 key_id; /* Serial no. of public key to use */ + __u32 in_len; /* Input data size */ + union { + __u32 out_len; /* Output buffer size (encrypt/decrypt/sign) */ + __u32 in2_len; /* 2nd input data size (verify) */ + }; + __u32 __spare[7]; +}; + #endif /* _LINUX_KEYCTL_H */ diff --git a/security/keys/Makefile b/security/keys/Makefile index ef1581b337a3..9cef54064f60 100644 --- a/security/keys/Makefile +++ b/security/keys/Makefile @@ -22,6 +22,7 @@ obj-$(CONFIG_PROC_FS) += proc.o obj-$(CONFIG_SYSCTL) += sysctl.o obj-$(CONFIG_PERSISTENT_KEYRINGS) += persistent.o obj-$(CONFIG_KEY_DH_OPERATIONS) += dh.o +obj-$(CONFIG_ASYMMETRIC_KEY_TYPE) += keyctl_pkey.o # # Key types diff --git a/security/keys/compat.c b/security/keys/compat.c index e87c89c0177c..9482df601dc3 100644 --- a/security/keys/compat.c +++ b/security/keys/compat.c @@ -141,6 +141,24 @@ COMPAT_SYSCALL_DEFINE5(keyctl, u32, option, return keyctl_restrict_keyring(arg2, compat_ptr(arg3), compat_ptr(arg4)); + case KEYCTL_PKEY_QUERY: + if (arg3 != 0) + return -EINVAL; + return keyctl_pkey_query(arg2, + compat_ptr(arg4), + compat_ptr(arg5)); + + case KEYCTL_PKEY_ENCRYPT: + case KEYCTL_PKEY_DECRYPT: + case KEYCTL_PKEY_SIGN: + return keyctl_pkey_e_d_s(option, + compat_ptr(arg2), compat_ptr(arg3), + compat_ptr(arg4), compat_ptr(arg5)); + + case KEYCTL_PKEY_VERIFY: + return keyctl_pkey_verify(compat_ptr(arg2), compat_ptr(arg3), + compat_ptr(arg4), compat_ptr(arg5)); + default: return -EOPNOTSUPP; } diff --git a/security/keys/internal.h b/security/keys/internal.h index 9f8208dc0e55..74cb0ff42fed 100644 --- a/security/keys/internal.h +++ b/security/keys/internal.h @@ -298,6 +298,45 @@ static inline long compat_keyctl_dh_compute( #endif #endif +#ifdef CONFIG_ASYMMETRIC_KEY_TYPE +extern long keyctl_pkey_query(key_serial_t, + const char __user *, + struct keyctl_pkey_query __user *); + +extern long keyctl_pkey_verify(const struct keyctl_pkey_params __user *, + const char __user *, + const void __user *, const void __user *); + +extern long keyctl_pkey_e_d_s(int, + const struct keyctl_pkey_params __user *, + const char __user *, + const void __user *, void __user *); +#else +static inline long keyctl_pkey_query(key_serial_t id, + const char __user *_info, + struct keyctl_pkey_query __user *_res) +{ + return -EOPNOTSUPP; +} + +static inline long keyctl_pkey_verify(const struct keyctl_pkey_params __user *params, + const char __user *_info, + const void __user *_in, + const void __user *_in2) +{ + return -EOPNOTSUPP; +} + +static inline long keyctl_pkey_e_d_s(int op, + const struct keyctl_pkey_params __user *params, + const char __user *_info, + const void __user *_in, + void __user *_out) +{ + return -EOPNOTSUPP; +} +#endif + /* * Debugging key validation */ diff --git a/security/keys/keyctl.c b/security/keys/keyctl.c index 1ffe60bb2845..18619690ce77 100644 --- a/security/keys/keyctl.c +++ b/security/keys/keyctl.c @@ -1747,6 +1747,30 @@ SYSCALL_DEFINE5(keyctl, int, option, unsigned long, arg2, unsigned long, arg3, (const char __user *) arg3, (const char __user *) arg4); + case KEYCTL_PKEY_QUERY: + if (arg3 != 0) + return -EINVAL; + return keyctl_pkey_query((key_serial_t)arg2, + (const char __user *)arg4, + (struct keyctl_pkey_query *)arg5); + + case KEYCTL_PKEY_ENCRYPT: + case KEYCTL_PKEY_DECRYPT: + case KEYCTL_PKEY_SIGN: + return keyctl_pkey_e_d_s( + option, + (const struct keyctl_pkey_params __user *)arg2, + (const char __user *)arg3, + (const void __user *)arg4, + (void __user *)arg5); + + case KEYCTL_PKEY_VERIFY: + return keyctl_pkey_verify( + (const struct keyctl_pkey_params __user *)arg2, + (const char __user *)arg3, + (const void __user *)arg4, + (const void __user *)arg5); + default: return -EOPNOTSUPP; } diff --git a/security/keys/keyctl_pkey.c b/security/keys/keyctl_pkey.c new file mode 100644 index 000000000000..783978842f13 --- /dev/null +++ b/security/keys/keyctl_pkey.c @@ -0,0 +1,323 @@ +/* Public-key operation keyctls + * + * Copyright (C) 2016 Red Hat, Inc. All Rights Reserved. + * Written by David Howells (dhowells@redhat.com) + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public Licence + * as published by the Free Software Foundation; either version + * 2 of the Licence, or (at your option) any later version. + */ + +#include +#include +#include +#include +#include +#include +#include +#include "internal.h" + +static void keyctl_pkey_params_free(struct kernel_pkey_params *params) +{ + kfree(params->info); + key_put(params->key); +} + +enum { + Opt_err = -1, + Opt_enc, /* "enc=" eg. "enc=oaep" */ + Opt_hash, /* "hash=" eg. "hash=sha1" */ +}; + +static const match_table_t param_keys = { + { Opt_enc, "enc=%s" }, + { Opt_hash, "hash=%s" }, + { Opt_err, NULL } +}; + +/* + * Parse the information string which consists of key=val pairs. + */ +static int keyctl_pkey_params_parse(struct kernel_pkey_params *params) +{ + unsigned long token_mask = 0; + substring_t args[MAX_OPT_ARGS]; + char *c = params->info, *p, *q; + int token; + + while ((p = strsep(&c, " \t"))) { + if (*p == '\0' || *p == ' ' || *p == '\t') + continue; + token = match_token(p, param_keys, args); + if (__test_and_set_bit(token, &token_mask)) + return -EINVAL; + q = args[0].from; + if (!q[0]) + return -EINVAL; + + switch (token) { + case Opt_enc: + params->encoding = q; + break; + + case Opt_hash: + params->hash_algo = q; + break; + + default: + return -EINVAL; + } + } + + return 0; +} + +/* + * Interpret parameters. Callers must always call the free function + * on params, even if an error is returned. + */ +static int keyctl_pkey_params_get(key_serial_t id, + const char __user *_info, + struct kernel_pkey_params *params) +{ + key_ref_t key_ref; + void *p; + int ret; + + memset(params, 0, sizeof(*params)); + params->encoding = "raw"; + + p = strndup_user(_info, PAGE_SIZE); + if (IS_ERR(p)) + return PTR_ERR(p); + params->info = p; + + ret = keyctl_pkey_params_parse(params); + if (ret < 0) + return ret; + + key_ref = lookup_user_key(id, 0, KEY_NEED_SEARCH); + if (IS_ERR(key_ref)) + return PTR_ERR(key_ref); + params->key = key_ref_to_ptr(key_ref); + + if (!params->key->type->asym_query) + return -EOPNOTSUPP; + + return 0; +} + +/* + * Get parameters from userspace. Callers must always call the free function + * on params, even if an error is returned. + */ +static int keyctl_pkey_params_get_2(const struct keyctl_pkey_params __user *_params, + const char __user *_info, + int op, + struct kernel_pkey_params *params) +{ + struct keyctl_pkey_params uparams; + struct kernel_pkey_query info; + int ret; + + memset(params, 0, sizeof(*params)); + params->encoding = "raw"; + + if (copy_from_user(&uparams, _params, sizeof(uparams)) != 0) + return -EFAULT; + + ret = keyctl_pkey_params_get(uparams.key_id, _info, params); + if (ret < 0) + return ret; + + ret = params->key->type->asym_query(params, &info); + if (ret < 0) + return ret; + + switch (op) { + case KEYCTL_PKEY_ENCRYPT: + case KEYCTL_PKEY_DECRYPT: + if (uparams.in_len > info.max_enc_size || + uparams.out_len > info.max_dec_size) + return -EINVAL; + break; + case KEYCTL_PKEY_SIGN: + case KEYCTL_PKEY_VERIFY: + if (uparams.in_len > info.max_sig_size || + uparams.out_len > info.max_data_size) + return -EINVAL; + break; + default: + BUG(); + } + + params->in_len = uparams.in_len; + params->out_len = uparams.out_len; + return 0; +} + +/* + * Query information about an asymmetric key. + */ +long keyctl_pkey_query(key_serial_t id, + const char __user *_info, + struct keyctl_pkey_query __user *_res) +{ + struct kernel_pkey_params params; + struct kernel_pkey_query res; + long ret; + + memset(¶ms, 0, sizeof(params)); + + ret = keyctl_pkey_params_get(id, _info, ¶ms); + if (ret < 0) + goto error; + + ret = params.key->type->asym_query(¶ms, &res); + if (ret < 0) + goto error; + + ret = -EFAULT; + if (copy_to_user(_res, &res, sizeof(res)) == 0 && + clear_user(_res->__spare, sizeof(_res->__spare)) == 0) + ret = 0; + +error: + keyctl_pkey_params_free(¶ms); + return ret; +} + +/* + * Encrypt/decrypt/sign + * + * Encrypt data, decrypt data or sign data using a public key. + * + * _info is a string of supplementary information in key=val format. For + * instance, it might contain: + * + * "enc=pkcs1 hash=sha256" + * + * where enc= specifies the encoding and hash= selects the OID to go in that + * particular encoding if required. If enc= isn't supplied, it's assumed that + * the caller is supplying raw values. + * + * If successful, the amount of data written into the output buffer is + * returned. + */ +long keyctl_pkey_e_d_s(int op, + const struct keyctl_pkey_params __user *_params, + const char __user *_info, + const void __user *_in, + void __user *_out) +{ + struct kernel_pkey_params params; + void *in, *out; + long ret; + + ret = keyctl_pkey_params_get_2(_params, _info, op, ¶ms); + if (ret < 0) + goto error_params; + + ret = -EOPNOTSUPP; + if (!params.key->type->asym_eds_op) + goto error_params; + + switch (op) { + case KEYCTL_PKEY_ENCRYPT: + params.op = kernel_pkey_encrypt; + break; + case KEYCTL_PKEY_DECRYPT: + params.op = kernel_pkey_decrypt; + break; + case KEYCTL_PKEY_SIGN: + params.op = kernel_pkey_sign; + break; + default: + BUG(); + } + + in = memdup_user(_in, params.in_len); + if (IS_ERR(in)) { + ret = PTR_ERR(in); + goto error_params; + } + + ret = -ENOMEM; + out = kmalloc(params.out_len, GFP_KERNEL); + if (!out) + goto error_in; + + ret = params.key->type->asym_eds_op(¶ms, in, out); + if (ret < 0) + goto error_out; + + if (copy_to_user(_out, out, ret) != 0) + ret = -EFAULT; + +error_out: + kfree(out); +error_in: + kfree(in); +error_params: + keyctl_pkey_params_free(¶ms); + return ret; +} + +/* + * Verify a signature. + * + * Verify a public key signature using the given key, or if not given, search + * for a matching key. + * + * _info is a string of supplementary information in key=val format. For + * instance, it might contain: + * + * "enc=pkcs1 hash=sha256" + * + * where enc= specifies the signature blob encoding and hash= selects the OID + * to go in that particular encoding. If enc= isn't supplied, it's assumed + * that the caller is supplying raw values. + * + * If successful, 0 is returned. + */ +long keyctl_pkey_verify(const struct keyctl_pkey_params __user *_params, + const char __user *_info, + const void __user *_in, + const void __user *_in2) +{ + struct kernel_pkey_params params; + void *in, *in2; + long ret; + + ret = keyctl_pkey_params_get_2(_params, _info, KEYCTL_PKEY_VERIFY, + ¶ms); + if (ret < 0) + goto error_params; + + ret = -EOPNOTSUPP; + if (!params.key->type->asym_verify_signature) + goto error_params; + + in = memdup_user(_in, params.in_len); + if (IS_ERR(in)) { + ret = PTR_ERR(in); + goto error_params; + } + + in2 = memdup_user(_in2, params.in2_len); + if (IS_ERR(in2)) { + ret = PTR_ERR(in2); + goto error_in; + } + + params.op = kernel_pkey_verify; + ret = params.key->type->asym_verify_signature(¶ms, in, in2); + + kfree(in2); +error_in: + kfree(in); +error_params: + keyctl_pkey_params_free(¶ms); + return ret; +} -- cgit v1.3-14-g43fede From 4f8f382e635707ddaddf8269a116e4f8cc8835c0 Mon Sep 17 00:00:00 2001 From: David Miller Date: Tue, 30 Oct 2018 22:24:04 -0700 Subject: perf tools: Don't clone maps from parent when synthesizing forks When synthesizing FORK events, we are trying to create thread objects for the already running tasks on the machine. Normally, for a kernel FORK event, we want to clone the parent's maps because that is what the kernel just did. But when synthesizing, this should not be done. If we do, we end up with overlapping maps as we process the sythesized MMAP2 events that get delivered shortly thereafter. Use the FORK event misc flags in an internal way to signal this situation, so we can elide the map clone when appropriate. Signed-off-by: David S. Miller Cc: Don Zickus Cc: Jiri Olsa Cc: Joe Mario Link: http://lkml.kernel.org/r/20181030.222404.2085088822877051075.davem@davemloft.net [ Added comment about flag use in machine__process_fork_event(), use ternary op in thread__clone_map_groups() as suggested by Jiri ] Signed-off-by: Arnaldo Carvalho de Melo --- include/uapi/linux/perf_event.h | 2 ++ tools/include/uapi/linux/perf_event.h | 2 ++ tools/perf/util/event.c | 1 + tools/perf/util/machine.c | 19 ++++++++++++++++++- tools/perf/util/thread.c | 13 +++++-------- tools/perf/util/thread.h | 2 +- 6 files changed, 29 insertions(+), 10 deletions(-) (limited to 'include/uapi/linux') diff --git a/include/uapi/linux/perf_event.h b/include/uapi/linux/perf_event.h index f35eb72739c0..9de8780ac8d9 100644 --- a/include/uapi/linux/perf_event.h +++ b/include/uapi/linux/perf_event.h @@ -646,10 +646,12 @@ struct perf_event_mmap_page { * * PERF_RECORD_MISC_MMAP_DATA - PERF_RECORD_MMAP* events * PERF_RECORD_MISC_COMM_EXEC - PERF_RECORD_COMM event + * PERF_RECORD_MISC_FORK_EXEC - PERF_RECORD_FORK event (perf internal) * PERF_RECORD_MISC_SWITCH_OUT - PERF_RECORD_SWITCH* events */ #define PERF_RECORD_MISC_MMAP_DATA (1 << 13) #define PERF_RECORD_MISC_COMM_EXEC (1 << 13) +#define PERF_RECORD_MISC_FORK_EXEC (1 << 13) #define PERF_RECORD_MISC_SWITCH_OUT (1 << 13) /* * These PERF_RECORD_MISC_* flags below are safely reused diff --git a/tools/include/uapi/linux/perf_event.h b/tools/include/uapi/linux/perf_event.h index f35eb72739c0..9de8780ac8d9 100644 --- a/tools/include/uapi/linux/perf_event.h +++ b/tools/include/uapi/linux/perf_event.h @@ -646,10 +646,12 @@ struct perf_event_mmap_page { * * PERF_RECORD_MISC_MMAP_DATA - PERF_RECORD_MMAP* events * PERF_RECORD_MISC_COMM_EXEC - PERF_RECORD_COMM event + * PERF_RECORD_MISC_FORK_EXEC - PERF_RECORD_FORK event (perf internal) * PERF_RECORD_MISC_SWITCH_OUT - PERF_RECORD_SWITCH* events */ #define PERF_RECORD_MISC_MMAP_DATA (1 << 13) #define PERF_RECORD_MISC_COMM_EXEC (1 << 13) +#define PERF_RECORD_MISC_FORK_EXEC (1 << 13) #define PERF_RECORD_MISC_SWITCH_OUT (1 << 13) /* * These PERF_RECORD_MISC_* flags below are safely reused diff --git a/tools/perf/util/event.c b/tools/perf/util/event.c index bc646185f8d9..e9c108a6b1c3 100644 --- a/tools/perf/util/event.c +++ b/tools/perf/util/event.c @@ -308,6 +308,7 @@ static int perf_event__synthesize_fork(struct perf_tool *tool, event->fork.pid = tgid; event->fork.tid = pid; event->fork.header.type = PERF_RECORD_FORK; + event->fork.header.misc = PERF_RECORD_MISC_FORK_EXEC; event->fork.header.size = (sizeof(event->fork) + machine->id_hdr_size); diff --git a/tools/perf/util/machine.c b/tools/perf/util/machine.c index 8ee8ab39d8ac..8f36ce813bc5 100644 --- a/tools/perf/util/machine.c +++ b/tools/perf/util/machine.c @@ -1708,6 +1708,7 @@ int machine__process_fork_event(struct machine *machine, union perf_event *event struct thread *parent = machine__findnew_thread(machine, event->fork.ppid, event->fork.ptid); + bool do_maps_clone = true; int err = 0; if (dump_trace) @@ -1736,9 +1737,25 @@ int machine__process_fork_event(struct machine *machine, union perf_event *event thread = machine__findnew_thread(machine, event->fork.pid, event->fork.tid); + /* + * When synthesizing FORK events, we are trying to create thread + * objects for the already running tasks on the machine. + * + * Normally, for a kernel FORK event, we want to clone the parent's + * maps because that is what the kernel just did. + * + * But when synthesizing, this should not be done. If we do, we end up + * with overlapping maps as we process the sythesized MMAP2 events that + * get delivered shortly thereafter. + * + * Use the FORK event misc flags in an internal way to signal this + * situation, so we can elide the map clone when appropriate. + */ + if (event->fork.header.misc & PERF_RECORD_MISC_FORK_EXEC) + do_maps_clone = false; if (thread == NULL || parent == NULL || - thread__fork(thread, parent, sample->time) < 0) { + thread__fork(thread, parent, sample->time, do_maps_clone) < 0) { dump_printf("problem processing PERF_RECORD_FORK, skipping event.\n"); err = -1; } diff --git a/tools/perf/util/thread.c b/tools/perf/util/thread.c index 2048d393ece6..3d9ed7d0e281 100644 --- a/tools/perf/util/thread.c +++ b/tools/perf/util/thread.c @@ -330,7 +330,8 @@ static int thread__prepare_access(struct thread *thread) } static int thread__clone_map_groups(struct thread *thread, - struct thread *parent) + struct thread *parent, + bool do_maps_clone) { /* This is new thread, we share map groups for process. */ if (thread->pid_ == parent->pid_) @@ -341,15 +342,11 @@ static int thread__clone_map_groups(struct thread *thread, thread->pid_, thread->tid, parent->pid_, parent->tid); return 0; } - /* But this one is new process, copy maps. */ - if (map_groups__clone(thread, parent->mg) < 0) - return -ENOMEM; - - return 0; + return do_maps_clone ? map_groups__clone(thread, parent->mg) : 0; } -int thread__fork(struct thread *thread, struct thread *parent, u64 timestamp) +int thread__fork(struct thread *thread, struct thread *parent, u64 timestamp, bool do_maps_clone) { if (parent->comm_set) { const char *comm = thread__comm_str(parent); @@ -362,7 +359,7 @@ int thread__fork(struct thread *thread, struct thread *parent, u64 timestamp) } thread->ppid = parent->tid; - return thread__clone_map_groups(thread, parent); + return thread__clone_map_groups(thread, parent, do_maps_clone); } void thread__find_cpumode_addr_location(struct thread *thread, u64 addr, diff --git a/tools/perf/util/thread.h b/tools/perf/util/thread.h index 36c09a9904e6..30e2b4c165fe 100644 --- a/tools/perf/util/thread.h +++ b/tools/perf/util/thread.h @@ -89,7 +89,7 @@ struct comm *thread__comm(const struct thread *thread); struct comm *thread__exec_comm(const struct thread *thread); const char *thread__comm_str(const struct thread *thread); int thread__insert_map(struct thread *thread, struct map *map); -int thread__fork(struct thread *thread, struct thread *parent, u64 timestamp); +int thread__fork(struct thread *thread, struct thread *parent, u64 timestamp, bool do_maps_clone); size_t thread__fprintf(struct thread *thread, FILE *fp); struct thread *thread__main_thread(struct machine *machine, struct thread *thread); -- cgit v1.3-14-g43fede From 9b4789eacb654b7bbc806c831bcebd799ae0e2f5 Mon Sep 17 00:00:00 2001 From: Palmer Dabbelt Date: Mon, 25 Jun 2018 13:23:12 -0700 Subject: Move EM_RISCV into elf-em.h This should never have been inside our arch port to begin with, it's just a relic from when we were maintaining out of tree patches. Reviewed-by: Kees Cook Reviewed-by: Paul Walmsley Reviewed-by: Christoph Hellwig Tested-by: David Abdurachmanov Signed-off-by: Palmer Dabbelt --- arch/riscv/include/asm/elf.h | 3 --- include/uapi/linux/elf-em.h | 1 + 2 files changed, 1 insertion(+), 3 deletions(-) (limited to 'include/uapi/linux') diff --git a/arch/riscv/include/asm/elf.h b/arch/riscv/include/asm/elf.h index a1ef503d616e..697fc23b0d5a 100644 --- a/arch/riscv/include/asm/elf.h +++ b/arch/riscv/include/asm/elf.h @@ -16,9 +16,6 @@ #include #include -/* TODO: Move definition into include/uapi/linux/elf-em.h */ -#define EM_RISCV 0xF3 - /* * These are used to set parameters in the core dumps. */ diff --git a/include/uapi/linux/elf-em.h b/include/uapi/linux/elf-em.h index 31aa10178335..93722e60204c 100644 --- a/include/uapi/linux/elf-em.h +++ b/include/uapi/linux/elf-em.h @@ -41,6 +41,7 @@ #define EM_TILEPRO 188 /* Tilera TILEPro */ #define EM_MICROBLAZE 189 /* Xilinx MicroBlaze */ #define EM_TILEGX 191 /* Tilera TILE-Gx */ +#define EM_RISCV 243 /* RISC-V */ #define EM_BPF 247 /* Linux BPF - in-kernel virtual machine */ #define EM_FRV 0x5441 /* Fujitsu FR-V */ -- cgit v1.3-14-g43fede