Merge branch 'linux-next' of git://git.infradead.org/~dedekind/ubi-2.6

* 'linux-next' of git://git.infradead.org/~dedekind/ubi-2.6:
  UBI: fix checkpatch.pl warnings
  UBI: simplify PEB protection code
  UBI: prepare for protection tree improvements
  UBI: return -ENOMEM upon failing vmalloc
  UBI: document UBI ioctls
  UBI: handle write errors in WL worker
  UBI: fix error path
  UBI: some code re-structuring
  UBI: fix deadlock
  UBI: fix warnings when debugging is enabled

Documentation/ioctl/ioctl-number.txt

@@ -97,6 +97,7 @@ Code	Seq#	Include File		Comments
 					<http://linux01.gwdg.de/~alatham/ppdd.html>
 'M'	all	linux/soundcard.h
 'N'	00-1F	drivers/usb/scanner.h
+'O'     00-02   include/mtd/ubi-user.h UBI
 'P'	all	linux/soundcard.h
 'Q'	all	linux/soundcard.h
 'R'	00-1F	linux/random.h
@@ -142,6 +143,9 @@ Code	Seq#	Include File		Comments
 'n'	00-7F	linux/ncp_fs.h
 'n'	E0-FF	video/matrox.h          matroxfb
 'o'	00-1F	fs/ocfs2/ocfs2_fs.h	OCFS2
+'o'     00-03   include/mtd/ubi-user.h  conflict! (OCFS2 and UBI overlaps)
+'o'     40-41   include/mtd/ubi-user.h  UBI
+'o'     01-A1   include/linux/dvb/*.h DVB
 'p'	00-0F	linux/phantom.h		conflict! (OpenHaptics needs this)
 'p'	00-3F	linux/mc146818rtc.h	conflict!
 'p'	40-7F	linux/nvram.h

drivers/mtd/ubi/build.c

@@ -815,6 +815,7 @@ int ubi_attach_mtd_dev(struct mtd_info *mtd, int ubi_num, int vid_hdr_offset)
 	if (err)
 		goto out_free;
 
+	err = -ENOMEM;
 	ubi->peb_buf1 = vmalloc(ubi->peb_size);
 	if (!ubi->peb_buf1)
 		goto out_free;

drivers/mtd/ubi/cdev.c

@@ -721,7 +721,8 @@ static int rename_volumes(struct ubi_device *ubi,
 		 * It seems we need to remove volume with name @re->new_name,
 		 * if it exists.
 		 */
-		desc = ubi_open_volume_nm(ubi->ubi_num, re->new_name, UBI_EXCLUSIVE);
+		desc = ubi_open_volume_nm(ubi->ubi_num, re->new_name,
+					  UBI_EXCLUSIVE);
 		if (IS_ERR(desc)) {
 			err = PTR_ERR(desc);
 			if (err == -ENODEV)

drivers/mtd/ubi/eba.c

@@ -504,12 +504,9 @@ static int recover_peb(struct ubi_device *ubi, int pnum, int vol_id, int lnum,
 	if (!vid_hdr)
 		return -ENOMEM;
 
-	mutex_lock(&ubi->buf_mutex);
-
 retry:
 	new_pnum = ubi_wl_get_peb(ubi, UBI_UNKNOWN);
 	if (new_pnum < 0) {
-		mutex_unlock(&ubi->buf_mutex);
 		ubi_free_vid_hdr(ubi, vid_hdr);
 		return new_pnum;
 	}
@@ -529,20 +526,23 @@ retry:
 		goto write_error;
 
 	data_size = offset + len;
+	mutex_lock(&ubi->buf_mutex);
 	memset(ubi->peb_buf1 + offset, 0xFF, len);
 
 	/* Read everything before the area where the write failure happened */
 	if (offset > 0) {
 		err = ubi_io_read_data(ubi, ubi->peb_buf1, pnum, 0, offset);
 		if (err && err != UBI_IO_BITFLIPS)
-			goto out_put;
+			goto out_unlock;
 	}
 
 	memcpy(ubi->peb_buf1 + offset, buf, len);
 
 	err = ubi_io_write_data(ubi, ubi->peb_buf1, new_pnum, 0, data_size);
-	if (err)
+	if (err) {
+		mutex_unlock(&ubi->buf_mutex);
 		goto write_error;
+	}
 
 	mutex_unlock(&ubi->buf_mutex);
 	ubi_free_vid_hdr(ubi, vid_hdr);
@@ -553,8 +553,9 @@ retry:
 	ubi_msg("data was successfully recovered");
 	return 0;
 
-out_put:
+out_unlock:
 	mutex_unlock(&ubi->buf_mutex);
+out_put:
 	ubi_wl_put_peb(ubi, new_pnum, 1);
 	ubi_free_vid_hdr(ubi, vid_hdr);
 	return err;
@@ -567,7 +568,6 @@ write_error:
 	ubi_warn("failed to write to PEB %d", new_pnum);
 	ubi_wl_put_peb(ubi, new_pnum, 1);
 	if (++tries > UBI_IO_RETRIES) {
-		mutex_unlock(&ubi->buf_mutex);
 		ubi_free_vid_hdr(ubi, vid_hdr);
 		return err;
 	}
@@ -950,9 +950,13 @@ write_error:
  * physical eraseblock @to. The @vid_hdr buffer may be changed by this
  * function. Returns:
  *   o %0 in case of success;
- *   o %1 if the operation was canceled and should be tried later (e.g.,
- *     because a bit-flip was detected at the target PEB);
- *   o %2 if the volume is being deleted and this LEB should not be moved.
+ *   o %1 if the operation was canceled because the volume is being deleted
+ *        or because the PEB was put meanwhile;
+ *   o %2 if the operation was canceled because there was a write error to the
+ *        target PEB;
+ *   o %-EAGAIN if the operation was canceled because a bit-flip was detected
+ *     in the target PEB;
+ *   o a negative error code in case of failure.
  */
 int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to,
 		     struct ubi_vid_hdr *vid_hdr)
@@ -978,7 +982,7 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to,
 	/*
 	 * Note, we may race with volume deletion, which means that the volume
 	 * this logical eraseblock belongs to might be being deleted. Since the
-	 * volume deletion unmaps all the volume's logical eraseblocks, it will
+	 * volume deletion un-maps all the volume's logical eraseblocks, it will
 	 * be locked in 'ubi_wl_put_peb()' and wait for the WL worker to finish.
 	 */
 	vol = ubi->volumes[idx];
@@ -986,7 +990,7 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to,
 		/* No need to do further work, cancel */
 		dbg_eba("volume %d is being removed, cancel", vol_id);
 		spin_unlock(&ubi->volumes_lock);
-		return 2;
+		return 1;
 	}
 	spin_unlock(&ubi->volumes_lock);
 
@@ -1023,7 +1027,7 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to,
 
 	/*
 	 * OK, now the LEB is locked and we can safely start moving it. Since
-	 * this function utilizes thie @ubi->peb1_buf buffer which is shared
+	 * this function utilizes the @ubi->peb1_buf buffer which is shared
 	 * with some other functions, so lock the buffer by taking the
 	 * @ubi->buf_mutex.
 	 */
@@ -1068,8 +1072,11 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to,
 	vid_hdr->sqnum = cpu_to_be64(next_sqnum(ubi));
 
 	err = ubi_io_write_vid_hdr(ubi, to, vid_hdr);
-	if (err)
+	if (err) {
+		if (err == -EIO)
+			err = 2;
 		goto out_unlock_buf;
+	}
 
 	cond_resched();
 
@@ -1079,14 +1086,17 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to,
 		if (err != UBI_IO_BITFLIPS)
 			ubi_warn("cannot read VID header back from PEB %d", to);
 		else
-			err = 1;
+			err = -EAGAIN;
 		goto out_unlock_buf;
 	}
 
 	if (data_size > 0) {
 		err = ubi_io_write_data(ubi, ubi->peb_buf1, to, 0, aldata_size);
-		if (err)
+		if (err) {
+			if (err == -EIO)
+				err = 2;
 			goto out_unlock_buf;
+		}
 
 		cond_resched();
 
@@ -1101,15 +1111,16 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to,
 				ubi_warn("cannot read data back from PEB %d",
 					 to);
 			else
-				err = 1;
+				err = -EAGAIN;
 			goto out_unlock_buf;
 		}
 
 		cond_resched();
 
 		if (memcmp(ubi->peb_buf1, ubi->peb_buf2, aldata_size)) {
-			ubi_warn("read data back from PEB %d - it is different",
-				 to);
+			ubi_warn("read data back from PEB %d and it is "
+				 "different", to);
+			err = -EINVAL;
 			goto out_unlock_buf;
 		}
 	}

drivers/mtd/ubi/io.c

@@ -637,8 +637,6 @@ int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum,
 
 	dbg_io("read EC header from PEB %d", pnum);
 	ubi_assert(pnum >= 0 && pnum < ubi->peb_count);
-	if (UBI_IO_DEBUG)
-		verbose = 1;
 
 	err = ubi_io_read(ubi, ec_hdr, pnum, 0, UBI_EC_HDR_SIZE);
 	if (err) {
@@ -685,6 +683,9 @@ int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum,
 			if (verbose)
 				ubi_warn("no EC header found at PEB %d, "
 					 "only 0xFF bytes", pnum);
+			else if (UBI_IO_DEBUG)
+				dbg_msg("no EC header found at PEB %d, "
+					"only 0xFF bytes", pnum);
 			return UBI_IO_PEB_EMPTY;
 		}
 
@@ -696,7 +697,9 @@ int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum,
 			ubi_warn("bad magic number at PEB %d: %08x instead of "
 				 "%08x", pnum, magic, UBI_EC_HDR_MAGIC);
 			ubi_dbg_dump_ec_hdr(ec_hdr);
-		}
+		} else if (UBI_IO_DEBUG)
+			dbg_msg("bad magic number at PEB %d: %08x instead of "
+				"%08x", pnum, magic, UBI_EC_HDR_MAGIC);
 		return UBI_IO_BAD_EC_HDR;
 	}
 
@@ -708,7 +711,9 @@ int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum,
 			ubi_warn("bad EC header CRC at PEB %d, calculated "
 				 "%#08x, read %#08x", pnum, crc, hdr_crc);
 			ubi_dbg_dump_ec_hdr(ec_hdr);
-		}
+		} else if (UBI_IO_DEBUG)
+			dbg_msg("bad EC header CRC at PEB %d, calculated "
+				"%#08x, read %#08x", pnum, crc, hdr_crc);
 		return UBI_IO_BAD_EC_HDR;
 	}
 
@@ -912,8 +917,6 @@ int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum,
 
 	dbg_io("read VID header from PEB %d", pnum);
 	ubi_assert(pnum >= 0 &&  pnum < ubi->peb_count);
-	if (UBI_IO_DEBUG)
-		verbose = 1;
 
 	p = (char *)vid_hdr - ubi->vid_hdr_shift;
 	err = ubi_io_read(ubi, p, pnum, ubi->vid_hdr_aloffset,
@@ -960,6 +963,9 @@ int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum,
 			if (verbose)
 				ubi_warn("no VID header found at PEB %d, "
 					 "only 0xFF bytes", pnum);
+			else if (UBI_IO_DEBUG)
+				dbg_msg("no VID header found at PEB %d, "
+					"only 0xFF bytes", pnum);
 			return UBI_IO_PEB_FREE;
 		}
 
@@ -971,7 +977,9 @@ int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum,
 			ubi_warn("bad magic number at PEB %d: %08x instead of "
 				 "%08x", pnum, magic, UBI_VID_HDR_MAGIC);
 			ubi_dbg_dump_vid_hdr(vid_hdr);
-		}
+		} else if (UBI_IO_DEBUG)
+			dbg_msg("bad magic number at PEB %d: %08x instead of "
+				"%08x", pnum, magic, UBI_VID_HDR_MAGIC);
 		return UBI_IO_BAD_VID_HDR;
 	}
 
@@ -983,7 +991,9 @@ int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum,
 			ubi_warn("bad CRC at PEB %d, calculated %#08x, "
 				 "read %#08x", pnum, crc, hdr_crc);
 			ubi_dbg_dump_vid_hdr(vid_hdr);
-		}
+		} else if (UBI_IO_DEBUG)
+			dbg_msg("bad CRC at PEB %d, calculated %#08x, "
+				"read %#08x", pnum, crc, hdr_crc);
 		return UBI_IO_BAD_VID_HDR;
 	}
 
@@ -1024,7 +1034,7 @@ int ubi_io_write_vid_hdr(struct ubi_device *ubi, int pnum,
 
 	err = paranoid_check_peb_ec_hdr(ubi, pnum);
 	if (err)
-		return err > 0 ? -EINVAL: err;
+		return err > 0 ? -EINVAL : err;
 
 	vid_hdr->magic = cpu_to_be32(UBI_VID_HDR_MAGIC);
 	vid_hdr->version = UBI_VERSION;

drivers/mtd/ubi/ubi.h

@@ -73,6 +73,13 @@
  */
 #define UBI_IO_RETRIES 3
 
+/*
+ * Length of the protection queue. The length is effectively equivalent to the
+ * number of (global) erase cycles PEBs are protected from the wear-leveling
+ * worker.
+ */
+#define UBI_PROT_QUEUE_LEN 10
+
 /*
  * Error codes returned by the I/O sub-system.
  *
@@ -95,7 +102,8 @@ enum {
 
 /**
  * struct ubi_wl_entry - wear-leveling entry.
- * @rb: link in the corresponding RB-tree
+ * @u.rb: link in the corresponding (free/used) RB-tree
+ * @u.list: link in the protection queue
  * @ec: erase counter
  * @pnum: physical eraseblock number
  *
@@ -104,7 +112,10 @@ enum {
  * RB-trees. See WL sub-system for details.
  */
 struct ubi_wl_entry {
+	union {
 		struct rb_node rb;
+		struct list_head list;
+	} u;
 	int ec;
 	int pnum;
 };
@@ -288,7 +299,7 @@ struct ubi_wl_entry;
  * @beb_rsvd_level: normal level of PEBs reserved for bad PEB handling
  *
  * @autoresize_vol_id: ID of the volume which has to be auto-resized at the end
- *                     of UBI ititializetion
+ *                     of UBI initialization
  * @vtbl_slots: how many slots are available in the volume table
  * @vtbl_size: size of the volume table in bytes
  * @vtbl: in-RAM volume table copy
@@ -306,18 +317,17 @@ struct ubi_wl_entry;
  * @used: RB-tree of used physical eraseblocks
  * @free: RB-tree of free physical eraseblocks
  * @scrub: RB-tree of physical eraseblocks which need scrubbing
- * @prot: protection trees
- * @prot.pnum: protection tree indexed by physical eraseblock numbers
- * @prot.aec: protection tree indexed by absolute erase counter value
- * @wl_lock: protects the @used, @free, @prot, @lookuptbl, @abs_ec, @move_from,
- *           @move_to, @move_to_put @erase_pending, @wl_scheduled, and @works
+ * @pq: protection queue (contain physical eraseblocks which are temporarily
+ *      protected from the wear-leveling worker)
+ * @pq_head: protection queue head
+ * @wl_lock: protects the @used, @free, @pq, @pq_head, @lookuptbl, @move_from,
+ * 	     @move_to, @move_to_put @erase_pending, @wl_scheduled and @works
  * 	     fields
  * @move_mutex: serializes eraseblock moves
- * @work_sem: sycnhronizes the WL worker with use tasks
+ * @work_sem: synchronizes the WL worker with use tasks
  * @wl_scheduled: non-zero if the wear-leveling was scheduled
  * @lookuptbl: a table to quickly find a &struct ubi_wl_entry object for any
  *             physical eraseblock
- * @abs_ec: absolute erase counter
  * @move_from: physical eraseblock from where the data is being moved
  * @move_to: physical eraseblock where the data is being moved to
  * @move_to_put: if the "to" PEB was put
@@ -351,11 +361,11 @@ struct ubi_wl_entry;
  *
  * @peb_buf1: a buffer of PEB size used for different purposes
  * @peb_buf2: another buffer of PEB size used for different purposes
- * @buf_mutex: proptects @peb_buf1 and @peb_buf2
+ * @buf_mutex: protects @peb_buf1 and @peb_buf2
  * @ckvol_mutex: serializes static volume checking when opening
- * @mult_mutex: serializes operations on multiple volumes, like re-nameing
+ * @mult_mutex: serializes operations on multiple volumes, like re-naming
  * @dbg_peb_buf: buffer of PEB size used for debugging
- * @dbg_buf_mutex: proptects @dbg_peb_buf
+ * @dbg_buf_mutex: protects @dbg_peb_buf
  */
 struct ubi_device {
 	struct cdev cdev;
@@ -392,16 +402,13 @@ struct ubi_device {
 	struct rb_root used;
 	struct rb_root free;
 	struct rb_root scrub;
-	struct {
-		struct rb_root pnum;
-		struct rb_root aec;
-	} prot;
+	struct list_head pq[UBI_PROT_QUEUE_LEN];
+	int pq_head;
 	spinlock_t wl_lock;
 	struct mutex move_mutex;
 	struct rw_semaphore work_sem;
 	int wl_scheduled;
 	struct ubi_wl_entry **lookuptbl;
-	unsigned long long abs_ec;
 	struct ubi_wl_entry *move_from;
 	struct ubi_wl_entry *move_to;
 	int move_to_put;