ceph: ioctls

A few Ceph ioctls for getting and setting file layout (striping) parameters, and learning the identity and network address of the OSD a given region of a file is stored on. Signed-off-by: Sage Weil <sage@newdream.net>

ceph: ioctls
A few Ceph ioctls for getting and setting file layout (striping) parameters, and learning the identity and network address of the OSD a given region of a file is stored on. Signed-off-by: Sage Weil <sage@newdream.net>
Sage Weil
1 parent a8e63b7d51
Showing 3 changed files with 197 additions and 0 deletions Side-by-side Diff
Documentation/ioctl/ioctl-number.txt
fs/ceph/ioctl.c
fs/ceph/ioctl.h
@@ -182,6 +182,7 @@
 					<http://www.proximity.com.au/~brian/winradio/>
 0x90	00	drivers/cdrom/sbpcd.h
 0x93	60-7F	linux/auto_fs.h
+0x97    00-7F   fs/ceph/ioctl.h         Ceph file system
 0x99	00-0F				537-Addinboard driver
 					<mailto:buk@buks.ipn.de>
 0xA0	all	linux/sdp/sdp.h		Industrial Device Project
+#include <linux/in.h>
+
+#include "ioctl.h"
+#include "super.h"
+#include "ceph_debug.h"
+
+
+/*
+ * ioctls
+ */
+
+/*
+ * get and set the file layout
+ */
+static long ceph_ioctl_get_layout(struct file *file, void __user *arg)
+{
+	struct ceph_inode_info *ci = ceph_inode(file->f_dentry->d_inode);
+	struct ceph_ioctl_layout l;
+	int err;
+
+	err = ceph_do_getattr(file->f_dentry->d_inode, CEPH_STAT_CAP_LAYOUT);
+	if (!err) {
+		l.stripe_unit = ceph_file_layout_su(ci->i_layout);
+		l.stripe_count = ceph_file_layout_stripe_count(ci->i_layout);
+		l.object_size = ceph_file_layout_object_size(ci->i_layout);
+		l.data_pool = le32_to_cpu(ci->i_layout.fl_pg_pool);
+		if (copy_to_user(arg, &l, sizeof(l)))
+			return -EFAULT;
+	}
+
+	return err;
+}
+
+static long ceph_ioctl_set_layout(struct file *file, void __user *arg)
+{
+	struct inode *inode = file->f_dentry->d_inode;
+	struct inode *parent_inode = file->f_dentry->d_parent->d_inode;
+	struct ceph_mds_client *mdsc = &ceph_sb_to_client(inode->i_sb)->mdsc;
+	struct ceph_mds_request *req;
+	struct ceph_ioctl_layout l;
+	int err, i;
+
+	/* copy and validate */
+	if (copy_from_user(&l, arg, sizeof(l)))
+		return -EFAULT;
+
+	if ((l.object_size & ~PAGE_MASK) ||
+	    (l.stripe_unit & ~PAGE_MASK) ||
+	    !l.stripe_unit ||
+	    (l.object_size &&
+	     (unsigned)l.object_size % (unsigned)l.stripe_unit))
+		return -EINVAL;
+
+	/* make sure it's a valid data pool */
+	if (l.data_pool > 0) {
+		mutex_lock(&mdsc->mutex);
+		err = -EINVAL;
+		for (i = 0; i < mdsc->mdsmap->m_num_data_pg_pools; i++)
+			if (mdsc->mdsmap->m_data_pg_pools[i] == l.data_pool) {
+				err = 0;
+				break;
+			}
+		mutex_unlock(&mdsc->mutex);
+		if (err)
+			return err;
+	}
+
+	req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_SETLAYOUT,
+				       USE_AUTH_MDS);
+	if (IS_ERR(req))
+		return PTR_ERR(req);
+	req->r_inode = igrab(inode);
+	req->r_inode_drop = CEPH_CAP_FILE_SHARED | CEPH_CAP_FILE_EXCL;
+
+	req->r_args.setlayout.layout.fl_stripe_unit =
+		cpu_to_le32(l.stripe_unit);
+	req->r_args.setlayout.layout.fl_stripe_count =
+		cpu_to_le32(l.stripe_count);
+	req->r_args.setlayout.layout.fl_object_size =
+		cpu_to_le32(l.object_size);
+	req->r_args.setlayout.layout.fl_pg_pool = cpu_to_le32(l.data_pool);
+	req->r_args.setlayout.layout.fl_pg_preferred = cpu_to_le32((s32)-1);
+
+	err = ceph_mdsc_do_request(mdsc, parent_inode, req);
+	ceph_mdsc_put_request(req);
+	return err;
+}
+
+/*
+ * Return object name, size/offset information, and location (OSD
+ * number, network address) for a given file offset.
+ */
+static long ceph_ioctl_get_dataloc(struct file *file, void __user *arg)
+{
+	struct ceph_ioctl_dataloc dl;
+	struct inode *inode = file->f_dentry->d_inode;
+	struct ceph_inode_info *ci = ceph_inode(inode);
+	struct ceph_osd_client *osdc = &ceph_client(inode->i_sb)->osdc;
+	u64 len = 1, olen;
+	u64 tmp;
+	struct ceph_object_layout ol;
+	union ceph_pg pgid;
+
+	/* copy and validate */
+	if (copy_from_user(&dl, arg, sizeof(dl)))
+		return -EFAULT;
+
+	down_read(&osdc->map_sem);
+	ceph_calc_file_object_mapping(&ci->i_layout, dl.file_offset, &len,
+				      &dl.object_no, &dl.object_offset, &olen);
+	dl.file_offset -= dl.object_offset;
+	dl.object_size = ceph_file_layout_object_size(ci->i_layout);
+	dl.block_size = ceph_file_layout_su(ci->i_layout);
+
+	/* block_offset = object_offset % block_size */
+	tmp = dl.object_offset;
+	dl.block_offset = do_div(tmp, dl.block_size);
+
+	snprintf(dl.object_name, sizeof(dl.object_name), "%llx.%08llx",
+		 ceph_ino(inode), dl.object_no);
+	ceph_calc_object_layout(&ol, dl.object_name, &ci->i_layout,
+				osdc->osdmap);
+
+	pgid.pg64 = le64_to_cpu(ol.ol_pgid);
+	dl.osd = ceph_calc_pg_primary(osdc->osdmap, pgid);
+	if (dl.osd >= 0) {
+		struct ceph_entity_addr *a =
+			ceph_osd_addr(osdc->osdmap, dl.osd);
+		if (a)
+			memcpy(&dl.osd_addr, &a->in_addr, sizeof(dl.osd_addr));
+	} else {
+		memset(&dl.osd_addr, 0, sizeof(dl.osd_addr));
+	}
+	up_read(&osdc->map_sem);
+
+	/* send result back to user */
+	if (copy_to_user(arg, &dl, sizeof(dl)))
+		return -EFAULT;
+
+	return 0;
+}
+
+long ceph_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
+{
+	dout("ioctl file %p cmd %u arg %lu\n", file, cmd, arg);
+	switch (cmd) {
+	case CEPH_IOC_GET_LAYOUT:
+		return ceph_ioctl_get_layout(file, (void __user *)arg);
+
+	case CEPH_IOC_SET_LAYOUT:
+		return ceph_ioctl_set_layout(file, (void __user *)arg);
+
+	case CEPH_IOC_GET_DATALOC:
+		return ceph_ioctl_get_dataloc(file, (void __user *)arg);
+	}
+	return -ENOTTY;
+}
+#ifndef FS_CEPH_IOCTL_H
+#define FS_CEPH_IOCTL_H
+
+#include <linux/ioctl.h>
+#include <linux/types.h>
+
+#define CEPH_IOCTL_MAGIC 0x97
+
+/* just use u64 to align sanely on all archs */
+struct ceph_ioctl_layout {
+	__u64 stripe_unit, stripe_count, object_size;
+	__u64 data_pool;
+};
+
+#define CEPH_IOC_GET_LAYOUT _IOR(CEPH_IOCTL_MAGIC, 1,		\
+				   struct ceph_ioctl_layout)
+#define CEPH_IOC_SET_LAYOUT _IOW(CEPH_IOCTL_MAGIC, 2,		\
+				   struct ceph_ioctl_layout)
+
+/*
+ * Extract identity, address of the OSD and object storing a given
+ * file offset.
+ */
+struct ceph_ioctl_dataloc {
+	__u64 file_offset;           /* in+out: file offset */
+	__u64 object_offset;         /* out: offset in object */
+	__u64 object_no;             /* out: object # */
+	__u64 object_size;           /* out: object size */
+	char object_name[64];        /* out: object name */
+	__u64 block_offset;          /* out: offset in block */
+	__u64 block_size;            /* out: block length */
+	__s64 osd;                   /* out: osd # */
+	struct sockaddr_storage osd_addr; /* out: osd address */
+};
+
+#define CEPH_IOC_GET_DATALOC _IOWR(CEPH_IOCTL_MAGIC, 3,	\
+				   struct ceph_ioctl_dataloc)
+
+#endif
...	...	@@ -182,6 +182,7 @@
182	182	<http://www.proximity.com.au/~brian/winradio/>
183	183	0x90 00 drivers/cdrom/sbpcd.h
184	184	0x93 60-7F linux/auto_fs.h
	185	+0x97 00-7F fs/ceph/ioctl.h Ceph file system
185	186	0x99 00-0F 537-Addinboard driver
186	187	<mailto:buk@buks.ipn.de>
187	188	0xA0 all linux/sdp/sdp.h Industrial Device Project
	1	+#include <linux/in.h>
	2	+
	3	+#include "ioctl.h"
	4	+#include "super.h"
	5	+#include "ceph_debug.h"
	6	+
	7	+
	8	+/*
	9	+ * ioctls
	10	+ */
	11	+
	12	+/*
	13	+ * get and set the file layout
	14	+ */
	15	+static long ceph_ioctl_get_layout(struct file file, void __user arg)
	16	+{
	17	+ struct ceph_inode_info *ci = ceph_inode(file->f_dentry->d_inode);
	18	+ struct ceph_ioctl_layout l;
	19	+ int err;
	20	+
	21	+ err = ceph_do_getattr(file->f_dentry->d_inode, CEPH_STAT_CAP_LAYOUT);
	22	+ if (!err) {
	23	+ l.stripe_unit = ceph_file_layout_su(ci->i_layout);
	24	+ l.stripe_count = ceph_file_layout_stripe_count(ci->i_layout);
	25	+ l.object_size = ceph_file_layout_object_size(ci->i_layout);
	26	+ l.data_pool = le32_to_cpu(ci->i_layout.fl_pg_pool);
	27	+ if (copy_to_user(arg, &l, sizeof(l)))
	28	+ return -EFAULT;
	29	+ }
	30	+
	31	+ return err;
	32	+}
	33	+
	34	+static long ceph_ioctl_set_layout(struct file file, void __user arg)
	35	+{
	36	+ struct inode *inode = file->f_dentry->d_inode;
	37	+ struct inode *parent_inode = file->f_dentry->d_parent->d_inode;
	38	+ struct ceph_mds_client *mdsc = &ceph_sb_to_client(inode->i_sb)->mdsc;
	39	+ struct ceph_mds_request *req;
	40	+ struct ceph_ioctl_layout l;
	41	+ int err, i;
	42	+
	43	+ /* copy and validate */
	44	+ if (copy_from_user(&l, arg, sizeof(l)))
	45	+ return -EFAULT;
	46	+
	47	+ if ((l.object_size & ~PAGE_MASK) \|\|
	48	+ (l.stripe_unit & ~PAGE_MASK) \|\|
	49	+ !l.stripe_unit \|\|
	50	+ (l.object_size &&
	51	+ (unsigned)l.object_size % (unsigned)l.stripe_unit))
	52	+ return -EINVAL;
	53	+
	54	+ /* make sure it's a valid data pool */
	55	+ if (l.data_pool > 0) {
	56	+ mutex_lock(&mdsc->mutex);
	57	+ err = -EINVAL;
	58	+ for (i = 0; i < mdsc->mdsmap->m_num_data_pg_pools; i++)
	59	+ if (mdsc->mdsmap->m_data_pg_pools[i] == l.data_pool) {
	60	+ err = 0;
	61	+ break;
	62	+ }
	63	+ mutex_unlock(&mdsc->mutex);
	64	+ if (err)
	65	+ return err;
	66	+ }
	67	+
	68	+ req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_SETLAYOUT,
	69	+ USE_AUTH_MDS);
	70	+ if (IS_ERR(req))
	71	+ return PTR_ERR(req);
	72	+ req->r_inode = igrab(inode);
	73	+ req->r_inode_drop = CEPH_CAP_FILE_SHARED \| CEPH_CAP_FILE_EXCL;
	74	+
	75	+ req->r_args.setlayout.layout.fl_stripe_unit =
	76	+ cpu_to_le32(l.stripe_unit);
	77	+ req->r_args.setlayout.layout.fl_stripe_count =
	78	+ cpu_to_le32(l.stripe_count);
	79	+ req->r_args.setlayout.layout.fl_object_size =
	80	+ cpu_to_le32(l.object_size);
	81	+ req->r_args.setlayout.layout.fl_pg_pool = cpu_to_le32(l.data_pool);
	82	+ req->r_args.setlayout.layout.fl_pg_preferred = cpu_to_le32((s32)-1);
	83	+
	84	+ err = ceph_mdsc_do_request(mdsc, parent_inode, req);
	85	+ ceph_mdsc_put_request(req);
	86	+ return err;
	87	+}
	88	+
	89	+/*
	90	+ * Return object name, size/offset information, and location (OSD
	91	+ * number, network address) for a given file offset.
	92	+ */
	93	+static long ceph_ioctl_get_dataloc(struct file file, void __user arg)
	94	+{
	95	+ struct ceph_ioctl_dataloc dl;
	96	+ struct inode *inode = file->f_dentry->d_inode;
	97	+ struct ceph_inode_info *ci = ceph_inode(inode);
	98	+ struct ceph_osd_client *osdc = &ceph_client(inode->i_sb)->osdc;
	99	+ u64 len = 1, olen;
	100	+ u64 tmp;
	101	+ struct ceph_object_layout ol;
	102	+ union ceph_pg pgid;
	103	+
	104	+ /* copy and validate */
	105	+ if (copy_from_user(&dl, arg, sizeof(dl)))
	106	+ return -EFAULT;
	107	+
	108	+ down_read(&osdc->map_sem);
	109	+ ceph_calc_file_object_mapping(&ci->i_layout, dl.file_offset, &len,
	110	+ &dl.object_no, &dl.object_offset, &olen);
	111	+ dl.file_offset -= dl.object_offset;
	112	+ dl.object_size = ceph_file_layout_object_size(ci->i_layout);
	113	+ dl.block_size = ceph_file_layout_su(ci->i_layout);
	114	+
	115	+ /* block_offset = object_offset % block_size */
	116	+ tmp = dl.object_offset;
	117	+ dl.block_offset = do_div(tmp, dl.block_size);
	118	+
	119	+ snprintf(dl.object_name, sizeof(dl.object_name), "%llx.%08llx",
	120	+ ceph_ino(inode), dl.object_no);
	121	+ ceph_calc_object_layout(&ol, dl.object_name, &ci->i_layout,
	122	+ osdc->osdmap);
	123	+
	124	+ pgid.pg64 = le64_to_cpu(ol.ol_pgid);
	125	+ dl.osd = ceph_calc_pg_primary(osdc->osdmap, pgid);
	126	+ if (dl.osd >= 0) {
	127	+ struct ceph_entity_addr *a =
	128	+ ceph_osd_addr(osdc->osdmap, dl.osd);
	129	+ if (a)
	130	+ memcpy(&dl.osd_addr, &a->in_addr, sizeof(dl.osd_addr));
	131	+ } else {
	132	+ memset(&dl.osd_addr, 0, sizeof(dl.osd_addr));
	133	+ }
	134	+ up_read(&osdc->map_sem);
	135	+
	136	+ /* send result back to user */
	137	+ if (copy_to_user(arg, &dl, sizeof(dl)))
	138	+ return -EFAULT;
	139	+
	140	+ return 0;
	141	+}
	142	+
	143	+long ceph_ioctl(struct file *file, unsigned int cmd, unsigned long arg)
	144	+{
	145	+ dout("ioctl file %p cmd %u arg %lu\n", file, cmd, arg);
	146	+ switch (cmd) {
	147	+ case CEPH_IOC_GET_LAYOUT:
	148	+ return ceph_ioctl_get_layout(file, (void __user *)arg);
	149	+
	150	+ case CEPH_IOC_SET_LAYOUT:
	151	+ return ceph_ioctl_set_layout(file, (void __user *)arg);
	152	+
	153	+ case CEPH_IOC_GET_DATALOC:
	154	+ return ceph_ioctl_get_dataloc(file, (void __user *)arg);
	155	+ }
	156	+ return -ENOTTY;
	157	+}
	1	+#ifndef FS_CEPH_IOCTL_H
	2	+#define FS_CEPH_IOCTL_H
	3	+
	4	+#include <linux/ioctl.h>
	5	+#include <linux/types.h>
	6	+
	7	+#define CEPH_IOCTL_MAGIC 0x97
	8	+
	9	+/* just use u64 to align sanely on all archs */
	10	+struct ceph_ioctl_layout {
	11	+ __u64 stripe_unit, stripe_count, object_size;
	12	+ __u64 data_pool;
	13	+};
	14	+
	15	+#define CEPH_IOC_GET_LAYOUT _IOR(CEPH_IOCTL_MAGIC, 1, \
	16	+ struct ceph_ioctl_layout)
	17	+#define CEPH_IOC_SET_LAYOUT _IOW(CEPH_IOCTL_MAGIC, 2, \
	18	+ struct ceph_ioctl_layout)
	19	+
	20	+/*
	21	+ * Extract identity, address of the OSD and object storing a given
	22	+ * file offset.
	23	+ */
	24	+struct ceph_ioctl_dataloc {
	25	+ __u64 file_offset; /* in+out: file offset */
	26	+ __u64 object_offset; /* out: offset in object */
	27	+ __u64 object_no; /* out: object # */
	28	+ __u64 object_size; /* out: object size */
	29	+ char object_name[64]; /* out: object name */
	30	+ __u64 block_offset; /* out: offset in block */
	31	+ __u64 block_size; /* out: block length */
	32	+ __s64 osd; /* out: osd # */
	33	+ struct sockaddr_storage osd_addr; /* out: osd address */
	34	+};
	35	+
	36	+#define CEPH_IOC_GET_DATALOC _IOWR(CEPH_IOCTL_MAGIC, 3, \
	37	+ struct ceph_ioctl_dataloc)
	38	+
	39	+#endif