Re: [PATCH 2/2] mkfs.f2fs: add device aliasing feature
From: Jaegeuk Kim
Date: Fri Sep 13 2024 - 18:27:35 EST
On 09/13, Daeho Jeong wrote:
> From: Daeho Jeong <daehojeong@xxxxxxxxxx>
>
> We can add a device aliasing file which can map the whole device with an
> extent, not using node blocks. This mapped area should be pinned and
> normally used for read-only usages. After finished using it, we can
> deallocate the whole area and return it back to use it for other files.
>
> Signed-off-by: Daeho Jeong <daehojeong@xxxxxxxxxx>
> ---
> fsck/dump.c | 13 ++
> fsck/fsck.c | 49 ++++--
> fsck/fsck.h | 4 +-
> fsck/main.c | 5 +
> include/f2fs_fs.h | 11 ++
> mkfs/f2fs_format.c | 335 ++++++++++++++++++++++++++++++++--------
> mkfs/f2fs_format_main.c | 30 +++-
> 7 files changed, 363 insertions(+), 84 deletions(-)
>
> diff --git a/fsck/dump.c b/fsck/dump.c
> index 8d5613e..aa9b5d3 100644
> --- a/fsck/dump.c
> +++ b/fsck/dump.c
> @@ -499,6 +499,19 @@ static int dump_inode_blk(struct f2fs_sb_info *sbi, u32 nid,
> }
>
> c.show_file_map_max_offset = f2fs_max_file_offset(&node_blk->i);
> +
> + if (IS_ALIASING(&node_blk->i)) {
> + u32 blkaddr = le32_to_cpu(node_blk->i.i_ext.blk_addr);
> + u32 len = le32_to_cpu(node_blk->i.i_ext.len);
> + u32 idx;
> +
> + for (idx = 0; idx < len; idx++)
> + dump_data_blk(sbi, idx * F2FS_BLKSIZE, blkaddr++, false);
> + print_extent(true);
> +
> + goto dump_xattr;
> + }
> +
> addr_per_block = ADDRS_PER_BLOCK(&node_blk->i);
>
> /* check data blocks in inode */
> diff --git a/fsck/fsck.c b/fsck/fsck.c
> index f8607fd..6feb294 100644
> --- a/fsck/fsck.c
> +++ b/fsck/fsck.c
> @@ -914,6 +914,7 @@ void fsck_chk_inode_blk(struct f2fs_sb_info *sbi, u32 nid,
> int need_fix = 0;
> int ret;
> u32 cluster_size = 1 << node_blk->i.i_log_cluster_size;
> + bool is_aliasing = IS_ALIASING(&node_blk->i);
>
> if (!compressed)
> goto check_next;
> @@ -1146,6 +1147,33 @@ check_next:
> addrs_per_blk * NIDS_PER_BLOCK *
> NIDS_PER_BLOCK) * F2FS_BLKSIZE;
> }
> +
> + if (is_aliasing) {
> + struct extent_info ei;
> +
> + get_extent_info(&ei, &node_blk->i.i_ext);
> + for (idx = 0; idx < ei.len; idx++, child.pgofs++) {
> + block_t blkaddr = ei.blk + idx;
> +
> + /* check extent info */
> + check_extent_info(&child, blkaddr, 0);
> + ret = fsck_chk_data_blk(sbi, &node_blk->i, blkaddr,
> + &child, (i_blocks == *blk_cnt), ftype, nid,
> + idx, ni->version, node_blk);
> + if (!ret) {
> + *blk_cnt = *blk_cnt + 1;
> + if (cur_qtype != -1)
> + qf_last_blkofs[cur_qtype] = child.pgofs;
> + } else if (c.fix_on) {
> + node_blk->i.i_ext.len = cpu_to_le32(idx);
> + need_fix = 1;
> + break;
> + }
> + }
> +
> + goto check;
> + }
> +
> for (idx = 0; idx < addrs; idx++, child.pgofs++) {
> block_t blkaddr = le32_to_cpu(node_blk->i.i_addr[ofs + idx]);
>
> @@ -1178,11 +1206,11 @@ check_next:
> child.pgofs - cbc->cheader_pgofs < cluster_size)
> cbc->cnt++;
> ret = fsck_chk_data_blk(sbi,
> - IS_CASEFOLDED(&node_blk->i),
> + &node_blk->i,
> blkaddr,
> &child, (i_blocks == *blk_cnt),
> ftype, nid, idx, ni->version,
> - file_is_encrypt(&node_blk->i), node_blk);
> + node_blk);
> if (blkaddr != le32_to_cpu(node_blk->i.i_addr[ofs + idx]))
> need_fix = 1;
> if (!ret) {
> @@ -1376,7 +1404,7 @@ skip_blkcnt_fix:
> }
>
> /* drop extent information to avoid potential wrong access */
> - if (need_fix && f2fs_dev_is_writable())
> + if (need_fix && f2fs_dev_is_writable() && !is_aliasing)
> node_blk->i.i_ext.len = 0;
>
> if ((c.feature & F2FS_FEATURE_INODE_CHKSUM) &&
> @@ -1448,11 +1476,9 @@ int fsck_chk_dnode_blk(struct f2fs_sb_info *sbi, struct f2fs_inode *inode,
> if (!compr_rel && blkaddr == NEW_ADDR && child->pgofs -
> cbc->cheader_pgofs < cluster_size)
> cbc->cnt++;
> - ret = fsck_chk_data_blk(sbi, IS_CASEFOLDED(inode),
> - blkaddr, child,
> + ret = fsck_chk_data_blk(sbi, inode, blkaddr, child,
> le64_to_cpu(inode->i_blocks) == *blk_cnt, ftype,
> - nid, idx, ni->version,
> - file_is_encrypt(inode), node_blk);
> + nid, idx, ni->version, node_blk);
> if (blkaddr != le32_to_cpu(node_blk->dn.addr[idx]))
> need_fix = 1;
> if (!ret) {
> @@ -2056,12 +2082,15 @@ int fsck_chk_dentry_blk(struct f2fs_sb_info *sbi, int casefolded, u32 blk_addr,
> return 0;
> }
>
> -int fsck_chk_data_blk(struct f2fs_sb_info *sbi, int casefolded,
> +int fsck_chk_data_blk(struct f2fs_sb_info *sbi, struct f2fs_inode *inode,
> u32 blk_addr, struct child_info *child, int last_blk,
> enum FILE_TYPE ftype, u32 parent_nid, u16 idx_in_node, u8 ver,
> - int enc_name, struct f2fs_node *node_blk)
> + struct f2fs_node *node_blk)
> {
> struct f2fs_fsck *fsck = F2FS_FSCK(sbi);
> + int casefolded = IS_CASEFOLDED(inode);
> + int enc_name = file_is_encrypt(inode);
> + int aliasing = IS_ALIASING(inode);
>
> /* Is it reserved block? */
> if (blk_addr == NEW_ADDR) {
> @@ -2074,7 +2103,7 @@ int fsck_chk_data_blk(struct f2fs_sb_info *sbi, int casefolded,
> return -EINVAL;
> }
>
> - if (is_valid_ssa_data_blk(sbi, blk_addr, parent_nid,
> + if (!aliasing && is_valid_ssa_data_blk(sbi, blk_addr, parent_nid,
> idx_in_node, ver)) {
> ASSERT_MSG("summary data block is not valid. [0x%x]",
> parent_nid);
> diff --git a/fsck/fsck.h b/fsck/fsck.h
> index 6cac926..d306c22 100644
> --- a/fsck/fsck.h
> +++ b/fsck/fsck.h
> @@ -179,9 +179,9 @@ extern int fsck_chk_idnode_blk(struct f2fs_sb_info *, struct f2fs_inode *,
> extern int fsck_chk_didnode_blk(struct f2fs_sb_info *, struct f2fs_inode *,
> enum FILE_TYPE, struct f2fs_node *, u32 *,
> struct f2fs_compr_blk_cnt *, struct child_info *);
> -extern int fsck_chk_data_blk(struct f2fs_sb_info *, int,
> +extern int fsck_chk_data_blk(struct f2fs_sb_info *, struct f2fs_inode *,
> u32, struct child_info *, int, enum FILE_TYPE, u32, u16, u8,
> - int, struct f2fs_node *);
> + struct f2fs_node *);
> extern int fsck_chk_dentry_blk(struct f2fs_sb_info *, int,
> u32, struct child_info *, int, int, struct f2fs_node *);
> int fsck_chk_inline_dentries(struct f2fs_sb_info *, struct f2fs_node *,
> diff --git a/fsck/main.c b/fsck/main.c
> index c13e287..c4f6167 100644
> --- a/fsck/main.c
> +++ b/fsck/main.c
> @@ -976,6 +976,11 @@ static int do_defrag(struct f2fs_sb_info *sbi)
> return -1;
> }
>
> + if (get_sb(feature) & F2FS_FEATURE_DEVICE_ALIAS) {
> + MSG(0, "Not support on image with device aliasing feature.\n");
> + return -1;
> + }
> +
> if (c.defrag_start > get_sb(block_count))
> goto out_range;
> if (c.defrag_start < SM_I(sbi)->main_blkaddr)
> diff --git a/include/f2fs_fs.h b/include/f2fs_fs.h
> index fdbc25b..108fde3 100644
> --- a/include/f2fs_fs.h
> +++ b/include/f2fs_fs.h
> @@ -442,6 +442,7 @@ struct device_info {
> uint64_t start_blkaddr;
> uint64_t end_blkaddr;
> uint32_t total_segments;
> + char *alias_filename;
>
> /* to handle zone block devices */
> int zoned_model;
> @@ -661,7 +662,13 @@ enum {
> */
> #define F2FS_IMMUTABLE_FL 0x00000010 /* Immutable file */
> #define F2FS_NOATIME_FL 0x00000080 /* do not update atime */
> +#define F2FS_DEVICE_ALIAS_FL 0x80000000 /* File for aliasing a device */
> +#define IS_ALIASING(fi) ((fi)->i_flags & cpu_to_le32(F2FS_DEVICE_ALIAS_FL))
Could you please rebase on the latest f2fs-tools? I got merge conflict around
here.
>
> +/*
> + * On-disk file inline flags (f2fs_inode::i_inline)
> + */
> +#define F2FS_PIN_FILE 0x40 /* file should not be gced */
Do we need the above define?
>
> #define F2FS_ENC_UTF8_12_1 1
> #define F2FS_ENC_STRICT_MODE_FL (1 << 0)
> @@ -694,6 +701,7 @@ enum {
> #define F2FS_FEATURE_CASEFOLD 0x1000
> #define F2FS_FEATURE_COMPRESSION 0x2000
> #define F2FS_FEATURE_RO 0x4000
> +#define F2FS_FEATURE_DEVICE_ALIAS 0x8000
>
> #define MAX_NR_FEATURE 32
>
> @@ -1515,11 +1523,14 @@ struct f2fs_configuration {
> time_t fixed_time;
> int roll_forward;
> bool need_fsync;
> + int aliased_devices;
> + uint32_t aliased_segments;
>
> /* mkfs parameters */
> int fake_seed;
> uint32_t next_free_nid;
> uint32_t lpf_ino;
> + uint32_t first_alias_ino;
> uint32_t root_uid;
> uint32_t root_gid;
> uint32_t blksize;
> diff --git a/mkfs/f2fs_format.c b/mkfs/f2fs_format.c
> index db1bc89..077f620 100644
> --- a/mkfs/f2fs_format.c
> +++ b/mkfs/f2fs_format.c
> @@ -13,6 +13,7 @@
> #include <unistd.h>
> #include <f2fs_fs.h>
> #include <assert.h>
> +#include <stdbool.h>
>
> #ifdef HAVE_SYS_STAT_H
> #include <sys/stat.h>
> @@ -39,10 +40,62 @@ struct f2fs_super_block raw_sb;
> struct f2fs_super_block *sb = &raw_sb;
> struct f2fs_checkpoint *cp;
>
> +static inline bool device_is_aliased(unsigned int dev_num)
> +{
> + if (dev_num >= c.ndevs)
> + return false;
> + return c.devices[dev_num].alias_filename != NULL;
> +}
> +
> +static inline unsigned int target_device_index(uint64_t blkaddr)
> +{
> + int i;
> +
> + for (i = 0; i < c.ndevs; i++)
> + if (c.devices[i].start_blkaddr <= blkaddr &&
> + c.devices[i].end_blkaddr >= blkaddr)
> + return i;
> + return 0;
> +}
> +
> +#define GET_SEGNO(blk_addr) ((blk_addr - get_sb(main_blkaddr)) / \
> + c.blks_per_seg)
> +#define START_BLOCK(segno) (segno * c.blks_per_seg + get_sb(main_blkaddr))
> +
> /* Return first segment number of each area */
> -#define prev_zone(cur) (c.cur_seg[cur] - c.segs_per_zone)
> -#define next_zone(cur) (c.cur_seg[cur] + c.segs_per_zone)
> -#define last_zone(cur) ((cur - 1) * c.segs_per_zone)
> +static inline uint32_t next_zone(int seg_type)
> +{
> + uint32_t next_seg = c.cur_seg[seg_type] + c.segs_per_zone;
> + uint64_t next_blkaddr = START_BLOCK(next_seg);
> + int dev_num;
> +
> + dev_num = target_device_index(next_blkaddr);
> + if (!device_is_aliased(dev_num))
> + return GET_SEGNO(next_blkaddr);
> +
> + while (dev_num < c.ndevs && device_is_aliased(dev_num))
> + dev_num++;
> +
> + return GET_SEGNO(c.devices[dev_num - 1].end_blkaddr + 1);
> +}
> +
> +static inline uint32_t last_zone(uint32_t total_zone)
> +{
> + uint32_t last_seg = (total_zone - 1) * c.segs_per_zone;
> + uint64_t last_blkaddr = START_BLOCK(last_seg);
> + int dev_num;
> +
> + dev_num = target_device_index(last_blkaddr);
> + if (!device_is_aliased(dev_num))
> + return GET_SEGNO(last_blkaddr);
> +
> + while (dev_num > 0 && device_is_aliased(dev_num))
> + dev_num--;
> +
> + return GET_SEGNO(c.devices[dev_num + 1].start_blkaddr) -
> + c.segs_per_zone;
> +}
> +
> #define last_section(cur) (cur + (c.secs_per_zone - 1) * c.segs_per_sec)
>
> /* Return time fixed by the user or current time by default */
> @@ -220,7 +273,7 @@ static int f2fs_prepare_super_block(void)
> uint64_t total_meta_zones, total_meta_segments;
> uint32_t sit_bitmap_size, max_sit_bitmap_size;
> uint32_t max_nat_bitmap_size, max_nat_segments;
> - uint32_t total_zones, avail_zones;
> + uint32_t total_zones, avail_zones = 0;
> enum quota_type qtype;
> int i;
>
> @@ -314,6 +367,16 @@ static int f2fs_prepare_super_block(void)
> c.devices[i].end_blkaddr = c.devices[i].start_blkaddr +
> c.devices[i].total_segments *
> c.blks_per_seg - 1;
> + if (device_is_aliased(i)) {
> + if (c.devices[i].zoned_model ==
> + F2FS_ZONED_HM) {
> + MSG(1, "\tError: do not support "
> + "device aliasing for device[%d]\n", i);
> + return -1;
> + }
> + c.aliased_segments +=
> + c.devices[i].total_segments;
> + }
> }
> if (c.ndevs > 1) {
> memcpy(sb->devs[i].path, c.devices[i].path, MAX_PATH_LEN);
> @@ -531,10 +594,16 @@ static int f2fs_prepare_super_block(void)
> if (c.feature & F2FS_FEATURE_LOST_FOUND)
> c.lpf_ino = c.next_free_nid++;
>
> + if (c.aliased_devices) {
> + c.first_alias_ino = c.next_free_nid;
> + c.next_free_nid += c.aliased_devices;
> + avail_zones += c.aliased_segments / c.segs_per_zone;
> + }
> +
> if (c.feature & F2FS_FEATURE_RO)
> - avail_zones = 2;
> + avail_zones += 2;
> else
> - avail_zones = 6;
> + avail_zones += 6;
>
> if (total_zones <= avail_zones) {
> MSG(1, "\tError: %d zones: Need more zones "
> @@ -701,6 +770,7 @@ static int f2fs_write_check_point_pack(void)
> char *sum_compact, *sum_compact_p;
> struct f2fs_summary *sum_entry;
> unsigned short vblocks;
> + uint32_t used_segments = c.aliased_segments;
> int ret = -1;
>
> cp = calloc(F2FS_BLKSIZE, 1);
> @@ -752,9 +822,14 @@ static int f2fs_write_check_point_pack(void)
> }
>
> set_cp(cur_node_blkoff[0], c.curseg_offset[CURSEG_HOT_NODE]);
> + set_cp(cur_node_blkoff[2], c.curseg_offset[CURSEG_COLD_NODE]);
> set_cp(cur_data_blkoff[0], c.curseg_offset[CURSEG_HOT_DATA]);
> + set_cp(cur_data_blkoff[2], c.curseg_offset[CURSEG_COLD_DATA]);
> set_cp(valid_block_count, c.curseg_offset[CURSEG_HOT_NODE] +
> - c.curseg_offset[CURSEG_HOT_DATA]);
> + c.curseg_offset[CURSEG_HOT_DATA] +
> + c.curseg_offset[CURSEG_COLD_NODE] +
> + c.curseg_offset[CURSEG_COLD_DATA] +
> + c.aliased_segments * c.blks_per_seg);
> set_cp(rsvd_segment_count, c.reserved_segments);
>
> /*
> @@ -802,15 +877,16 @@ static int f2fs_write_check_point_pack(void)
> c.reserved_segments);
>
> /* main segments - reserved segments - (node + data segments) */
> - if (c.feature & F2FS_FEATURE_RO) {
> - set_cp(free_segment_count, f2fs_get_usable_segments(sb) - 2);
> - set_cp(user_block_count, ((get_cp(free_segment_count) + 2 -
> - get_cp(overprov_segment_count)) * c.blks_per_seg));
> - } else {
> - set_cp(free_segment_count, f2fs_get_usable_segments(sb) - 6);
> - set_cp(user_block_count, ((get_cp(free_segment_count) + 6 -
> - get_cp(overprov_segment_count)) * c.blks_per_seg));
> - }
> + if (c.feature & F2FS_FEATURE_RO)
> + used_segments += 2;
> + else
> + used_segments += 6;
> +
> + set_cp(user_block_count, (f2fs_get_usable_segments(sb) -
> + get_cp(overprov_segment_count)) * c.blks_per_seg);
> + set_cp(free_segment_count, f2fs_get_usable_segments(sb) -
> + used_segments);
> +
> /* cp page (2), data summaries (1), node summaries (3) */
> set_cp(cp_pack_total_block_count, 6 + get_sb(cp_payload));
> flags = CP_UMOUNT_FLAG | CP_COMPACT_SUM_FLAG;
> @@ -826,8 +902,10 @@ static int f2fs_write_check_point_pack(void)
>
> set_cp(ckpt_flags, flags);
> set_cp(cp_pack_start_sum, 1 + get_sb(cp_payload));
> - set_cp(valid_node_count, c.curseg_offset[CURSEG_HOT_NODE]);
> - set_cp(valid_inode_count, c.curseg_offset[CURSEG_HOT_NODE]);
> + set_cp(valid_node_count, c.curseg_offset[CURSEG_HOT_NODE] +
> + c.curseg_offset[CURSEG_COLD_NODE]);
> + set_cp(valid_inode_count, c.curseg_offset[CURSEG_HOT_NODE] +
> + c.curseg_offset[CURSEG_COLD_NODE]);
> set_cp(next_free_nid, c.next_free_nid);
> set_cp(sit_ver_bitmap_bytesize, ((get_sb(segment_count_sit) / 2) <<
> get_sb(log_blocks_per_seg)) / 8);
> @@ -975,9 +1053,12 @@ static int f2fs_write_check_point_pack(void)
> goto free_cp_payload;
> }
>
> - /* Fill segment summary for COLD_NODE to zero. */
> + /* Prepare and write Segment summary for COLD_NODE */
> memset(sum, 0, F2FS_BLKSIZE);
> SET_SUM_TYPE(sum, SUM_TYPE_NODE);
> + memcpy(sum->entries, c.sum[CURSEG_COLD_NODE],
> + sizeof(struct f2fs_summary) * MAX_CACHE_SUMS);
> +
> cp_seg_blk++;
> DBG(1, "\tWriting Segment summary for COLD_NODE, at offset 0x%08"PRIx64"\n",
> cp_seg_blk);
> @@ -1210,10 +1291,40 @@ void update_summary_entry(int curseg_type, nid_t nid,
> sum->ofs_in_node = cpu_to_le16(ofs_in_node);
> }
>
> +static void add_dentry(struct f2fs_dentry_block *dent_blk, unsigned int *didx,
> + const char *name, uint32_t ino, u8 type)
> +{
> + int len = strlen(name);
> + f2fs_hash_t hash;
> +
> + if (name[0] == '.' && (len == 1 || (len == 2 && name[1] == '.')))
> + hash = 0;
> + else
> + hash = f2fs_dentry_hash(0, 0, (unsigned char *)name, len);
> +
> + F2FS_DENTRY_BLOCK_DENTRY(dent_blk, *didx).hash_code = cpu_to_le32(hash);
> + F2FS_DENTRY_BLOCK_DENTRY(dent_blk, *didx).ino = cpu_to_le32(ino);
> + F2FS_DENTRY_BLOCK_DENTRY(dent_blk, *didx).name_len = cpu_to_le16(len);
> + F2FS_DENTRY_BLOCK_DENTRY(dent_blk, *didx).file_type = type;
> +
> + while (len > F2FS_SLOT_LEN) {
> + memcpy(F2FS_DENTRY_BLOCK_FILENAME(dent_blk, *didx), name,
> + F2FS_SLOT_LEN);
> + test_and_set_bit_le(*didx, dent_blk->dentry_bitmap);
> + len -= (int)F2FS_SLOT_LEN;
> + name += F2FS_SLOT_LEN;
> + (*didx)++;
> + }
> + memcpy(F2FS_DENTRY_BLOCK_FILENAME(dent_blk, *didx), name, len);
> + test_and_set_bit_le(*didx, dent_blk->dentry_bitmap);
> + (*didx)++;
> +}
> +
> static block_t f2fs_add_default_dentry_root(void)
> {
> struct f2fs_dentry_block *dent_blk = NULL;
> block_t data_blkaddr;
> + unsigned int didx = 0;
>
> dent_blk = calloc(F2FS_BLKSIZE, 1);
> if(dent_blk == NULL) {
> @@ -1221,37 +1332,26 @@ static block_t f2fs_add_default_dentry_root(void)
> return 0;
> }
>
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 0).hash_code = 0;
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 0).ino = sb->root_ino;
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 0).name_len = cpu_to_le16(1);
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 0).file_type = F2FS_FT_DIR;
> - memcpy(F2FS_DENTRY_BLOCK_FILENAME(dent_blk, 0), ".", 1);
> + add_dentry(dent_blk, &didx, ".",
> + le32_to_cpu(sb->root_ino), F2FS_FT_DIR);
> + add_dentry(dent_blk, &didx, "..",
> + le32_to_cpu(sb->root_ino), F2FS_FT_DIR);
>
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 1).hash_code = 0;
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 1).ino = sb->root_ino;
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 1).name_len = cpu_to_le16(2);
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 1).file_type = F2FS_FT_DIR;
> - memcpy(F2FS_DENTRY_BLOCK_FILENAME(dent_blk, 1), "..", 2);
> -
> - /* bitmap for . and .. */
> - test_and_set_bit_le(0, dent_blk->dentry_bitmap);
> - test_and_set_bit_le(1, dent_blk->dentry_bitmap);
> -
> - if (c.lpf_ino) {
> - int len = strlen(LPF);
> - f2fs_hash_t hash = f2fs_dentry_hash(0, 0, (unsigned char *)LPF, len);
> + if (c.lpf_ino)
> + add_dentry(dent_blk, &didx, LPF, c.lpf_ino, F2FS_FT_DIR);
>
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 2).hash_code = cpu_to_le32(hash);
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 2).ino = cpu_to_le32(c.lpf_ino);
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 2).name_len = cpu_to_le16(len);
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 2).file_type = F2FS_FT_DIR;
> - memcpy(F2FS_DENTRY_BLOCK_FILENAME(dent_blk, 2), LPF, F2FS_SLOT_LEN);
> + if (c.aliased_devices) {
> + int i, dev_off = 0;
>
> - memcpy(F2FS_DENTRY_BLOCK_FILENAME(dent_blk, 3), &LPF[F2FS_SLOT_LEN],
> - len - F2FS_SLOT_LEN);
> + for (i = 1; i < c.ndevs; i++) {
> + if (!device_is_aliased(i))
> + continue;
>
> - test_and_set_bit_le(2, dent_blk->dentry_bitmap);
> - test_and_set_bit_le(3, dent_blk->dentry_bitmap);
> + add_dentry(dent_blk, &didx, c.devices[i].alias_filename,
> + c.first_alias_ino + dev_off,
> + F2FS_FT_REG_FILE);
> + dev_off++;
> + }
> }
>
> data_blkaddr = alloc_next_free_block(CURSEG_HOT_DATA);
> @@ -1324,6 +1424,7 @@ static int f2fs_write_default_quota(int qtype, __le32 raw_id)
> struct v2_disk_dqinfo ddqinfo;
> struct v2r1_disk_dqblk dqblk;
> block_t blkaddr;
> + uint64_t icnt = 1, bcnt = 1;
> int i;
>
> if (filebuf == NULL) {
> @@ -1359,16 +1460,18 @@ static int f2fs_write_default_quota(int qtype, __le32 raw_id)
> dqblk.dqb_pad = cpu_to_le32(0);
> dqblk.dqb_ihardlimit = cpu_to_le64(0);
> dqblk.dqb_isoftlimit = cpu_to_le64(0);
> - if (c.lpf_ino)
> - dqblk.dqb_curinodes = cpu_to_le64(2);
> - else
> - dqblk.dqb_curinodes = cpu_to_le64(1);
> + if (c.lpf_ino) {
> + icnt++;
> + bcnt++;
> + }
> + if (c.aliased_devices) {
> + icnt += c.aliased_devices;
> + bcnt += c.aliased_segments * c.blks_per_seg;
> + }
> + dqblk.dqb_curinodes = cpu_to_le64(icnt);
> dqblk.dqb_bhardlimit = cpu_to_le64(0);
> dqblk.dqb_bsoftlimit = cpu_to_le64(0);
> - if (c.lpf_ino)
> - dqblk.dqb_curspace = cpu_to_le64(F2FS_BLKSIZE * 2);
> - else
> - dqblk.dqb_curspace = cpu_to_le64(F2FS_BLKSIZE);
> + dqblk.dqb_curspace = cpu_to_le64(F2FS_BLKSIZE * bcnt);
> dqblk.dqb_btime = cpu_to_le64(0);
> dqblk.dqb_itime = cpu_to_le64(0);
>
> @@ -1491,6 +1594,7 @@ static block_t f2fs_add_default_dentry_lpf(void)
> {
> struct f2fs_dentry_block *dent_blk;
> block_t data_blkaddr;
> + unsigned int didx = 0;
>
> dent_blk = calloc(F2FS_BLKSIZE, 1);
> if (dent_blk == NULL) {
> @@ -1498,20 +1602,8 @@ static block_t f2fs_add_default_dentry_lpf(void)
> return 0;
> }
>
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 0).hash_code = 0;
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 0).ino = cpu_to_le32(c.lpf_ino);
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 0).name_len = cpu_to_le16(1);
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 0).file_type = F2FS_FT_DIR;
> - memcpy(F2FS_DENTRY_BLOCK_FILENAME(dent_blk, 0), ".", 1);
> -
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 1).hash_code = 0;
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 1).ino = sb->root_ino;
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 1).name_len = cpu_to_le16(2);
> - F2FS_DENTRY_BLOCK_DENTRY(dent_blk, 1).file_type = F2FS_FT_DIR;
> - memcpy(F2FS_DENTRY_BLOCK_FILENAME(dent_blk, 1), "..", 2);
> -
> - test_and_set_bit_le(0, dent_blk->dentry_bitmap);
> - test_and_set_bit_le(1, dent_blk->dentry_bitmap);
> + add_dentry(dent_blk, &didx, ".", c.lpf_ino, F2FS_FT_DIR);
> + add_dentry(dent_blk, &didx, "..", c.lpf_ino, F2FS_FT_DIR);
>
> data_blkaddr = alloc_next_free_block(CURSEG_HOT_DATA);
>
> @@ -1579,6 +1671,104 @@ exit:
> return err;
> }
>
> +static void allocate_blocks_for_aliased_device(struct f2fs_node *raw_node,
> + unsigned int dev_num)
> +{
> + uint32_t start_segno = (c.devices[dev_num].start_blkaddr -
> + get_sb(main_blkaddr)) / c.blks_per_seg;
> + uint32_t end_segno = (c.devices[dev_num].end_blkaddr -
> + get_sb(main_blkaddr) + 1) / c.blks_per_seg;
> + uint32_t segno;
> + uint64_t blkcnt;
> + struct f2fs_sit_block *sit_blk = calloc(F2FS_BLKSIZE, 1);
> +
> + ASSERT(sit_blk);
> +
> + for (segno = start_segno; segno < end_segno; segno++) {
> + struct f2fs_sit_entry *sit;
> + uint64_t sit_blk_addr = get_sb(sit_blkaddr) +
> + (segno / SIT_ENTRY_PER_BLOCK);
> +
> + ASSERT(dev_read_block(sit_blk, sit_blk_addr) >= 0);
> + sit = &sit_blk->entries[segno % SIT_ENTRY_PER_BLOCK];
> + memset(&sit->valid_map, 0xFF, SIT_VBLOCK_MAP_SIZE);
> + sit->vblocks = cpu_to_le16((CURSEG_COLD_DATA <<
> + SIT_VBLOCKS_SHIFT) | c.blks_per_seg);
> + sit->mtime = cpu_to_le64(mkfs_time);
> + ASSERT(dev_write_block(sit_blk, sit_blk_addr) >= 0);
> + }
> +
> + blkcnt = (end_segno - start_segno) * c.blks_per_seg;
> + raw_node->i.i_size = cpu_to_le64(blkcnt << get_sb(log_blocksize));
> + raw_node->i.i_blocks = cpu_to_le64(blkcnt + 1);
> +
> + raw_node->i.i_ext.fofs = cpu_to_le32(0);
> + raw_node->i.i_ext.blk_addr =
> + cpu_to_le32(c.devices[dev_num].start_blkaddr);
> + raw_node->i.i_ext.len = cpu_to_le32(blkcnt);
> +
> + free(sit_blk);
> +}
> +
> +static int f2fs_write_alias_inodes(void)
> +{
> + struct f2fs_node *raw_node;
> + block_t node_blkaddr;
> + int err = 0;
> + unsigned int i, dev_off = 0;
> +
> + ASSERT(c.aliased_devices);
> +
> + raw_node = calloc(F2FS_BLKSIZE, 1);
> + if (raw_node == NULL) {
> + MSG(1, "\tError: Calloc Failed for raw_node!!!\n");
> + return -1;
> + }
> +
> + for (i = 1; i < c.ndevs; i++) {
> + const char *filename;
> + nid_t ino;
> +
> + if (!device_is_aliased(i))
> + continue;
> +
> + ino = c.first_alias_ino + dev_off;
> + dev_off++;
> + f2fs_init_inode(sb, raw_node, ino, mkfs_time, 0x81c0);
> +
> + raw_node->i.i_flags = cpu_to_le32(F2FS_IMMUTABLE_FL |
> + F2FS_DEVICE_ALIAS_FL);
> + raw_node->i.i_inline = F2FS_PIN_FILE;
> + raw_node->i.i_pino = sb->root_ino;
> + filename = c.devices[i].alias_filename;
> + raw_node->i.i_namelen = cpu_to_le32(strlen(filename));
> + memcpy(raw_node->i.i_name, filename, strlen(filename));
> +
> + node_blkaddr = alloc_next_free_block(CURSEG_COLD_NODE);
> + F2FS_NODE_FOOTER(raw_node)->next_blkaddr =
> + cpu_to_le32(node_blkaddr + 1);
> +
> + allocate_blocks_for_aliased_device(raw_node, i);
> +
> + DBG(1, "\tWriting aliased device inode (cold node), "
> + "offset 0x%x\n", node_blkaddr);
> + if (write_inode(raw_node, node_blkaddr) < 0) {
> + MSG(1, "\tError: While writing the raw_node to "
> + "disk!!!\n");
> + err = -1;
> + goto exit;
> + }
> +
> + update_nat_journal(ino, node_blkaddr);
> + update_sit_journal(CURSEG_COLD_NODE);
> + update_summary_entry(CURSEG_COLD_NODE, ino, 0);
> + }
> +
> +exit:
> + free(raw_node);
> + return err;
> +}
> +
> static int f2fs_create_root_dir(void)
> {
> enum quota_type qtype;
> @@ -1608,6 +1798,15 @@ static int f2fs_create_root_dir(void)
> }
> }
>
> + if (c.aliased_devices) {
> + err = f2fs_write_alias_inodes();
> + if (err < 0) {
> + MSG(1, "\tError: Failed to write aliased device "
> + "inodes!!!\n");
> + goto exit;
> + }
> + }
> +
> #ifndef WITH_ANDROID
> err = f2fs_discard_obsolete_dnode();
> if (err < 0) {
> diff --git a/mkfs/f2fs_format_main.c b/mkfs/f2fs_format_main.c
> index 2ba1c21..b113bbc 100644
> --- a/mkfs/f2fs_format_main.c
> +++ b/mkfs/f2fs_format_main.c
> @@ -50,7 +50,7 @@ static void mkfs_usage()
> MSG(0, "\nUsage: mkfs.f2fs [options] device [sectors]\n");
> MSG(0, "[options]:\n");
> MSG(0, " -b filesystem block size [default:4096]\n");
> - MSG(0, " -c [device_name] up to 7 additional devices, except meta device\n");
> + MSG(0, " -c [device_name[@alias_filename]] up to 7 additional devices, except meta device\n");
> MSG(0, " -d debug level [default:0]\n");
> MSG(0, " -e [cold file ext list] e.g. \"mp3,gif,mov\"\n");
> MSG(0, " -E [hot file ext list] e.g. \"db\"\n");
> @@ -105,6 +105,9 @@ static void f2fs_show_info()
>
> if (c.feature & F2FS_FEATURE_COMPRESSION)
> MSG(0, "Info: Enable Compression\n");
> +
> + if (c.feature & F2FS_FEATURE_DEVICE_ALIAS)
> + MSG(0, "Info: Enable device aliasing\n");
> }
>
> #if defined(ANDROID_TARGET) && defined(HAVE_SYS_UTSNAME_H)
> @@ -181,6 +184,7 @@ static void f2fs_parse_options(int argc, char *argv[])
> int32_t option=0;
> int val;
> char *token;
> + int dev_num;
>
> while ((option = getopt_long(argc,argv,option_string,long_opts,NULL)) != EOF) {
> switch (option) {
> @@ -200,17 +204,35 @@ static void f2fs_parse_options(int argc, char *argv[])
> }
> break;
> case 'c':
> - if (c.ndevs >= MAX_DEVICES) {
> + dev_num = c.ndevs;
> +
> + if (dev_num >= MAX_DEVICES) {
> MSG(0, "Error: Too many devices\n");
> mkfs_usage();
> }
>
> - if (strlen(optarg) > MAX_PATH_LEN) {
> + token = strtok(optarg, "@");
> + if (strlen(token) > MAX_PATH_LEN) {
> MSG(0, "Error: device path should be less than "
> "%d characters\n", MAX_PATH_LEN);
> mkfs_usage();
> }
> - c.devices[c.ndevs++].path = strdup(optarg);
> + c.devices[dev_num].path = strdup(token);
> + token = strtok(NULL, "");
> + if (token) {
> + if (strlen(token) > MAX_PATH_LEN) {
> + MSG(0, "Error: alias_filename should "
> + "be less than %d characters\n",
> + MAX_PATH_LEN);
> + mkfs_usage();
> + }
> + c.devices[dev_num].alias_filename =
> + strdup(token);
> + if (!c.aliased_devices)
> + c.feature |= F2FS_FEATURE_DEVICE_ALIAS;
> + c.aliased_devices++;
> + }
> + c.ndevs++;
> break;
> case 'd':
> c.dbg_lv = atoi(optarg);
> --
> 2.46.0.662.g92d0881bb0-goog
>