| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <AANLkTi=b8eKuiOPxUiJYaDCzu1nJ40SxNqszxONC3CZ8@mail.gmail.com>
Date: Sat, 26 Feb 2011 09:20:55 +0200
From: Amir Goldstein <amir73il@...il.com>
To: "Theodore Ts'o" <tytso@....edu>
Cc: Ext4 Developers List <linux-ext4@...r.kernel.org>
Subject: Re: [PATCH] Add basic BIGALLOC support for cluster-based allocation
On Sat, Feb 26, 2011 at 4:50 AM, Theodore Ts'o <tytso@....edu> wrote:
> This adds the superblock fields needed so that dumpe2fs works and the
> code points and renames the superblock fields from describing
> fragments to clusters.
>
> Signed-off-by: "Theodore Ts'o" <tytso@....edu>
> ---
> debugfs/set_fields.c | 4 ++--
> e2fsck/super.c | 23 ++++++++---------------
> lib/e2p/feature.c | 2 ++
> lib/e2p/ls.c | 14 ++++++++++++--
> lib/ext2fs/ext2_fs.h | 25 +++++++++++--------------
> lib/ext2fs/ext2fs.h | 4 ++--
> lib/ext2fs/initialize.c | 11 ++++-------
> lib/ext2fs/openfs.c | 2 +-
> lib/ext2fs/swapfs.c | 4 ++--
> lib/ext2fs/tst_super_size.c | 4 ++--
> misc/mke2fs.c | 24 ++++++++++++++++--------
> 11 files changed, 62 insertions(+), 55 deletions(-)
>
> diff --git a/debugfs/set_fields.c b/debugfs/set_fields.c
> index 3b40117..9d7099b 100644
> --- a/debugfs/set_fields.c
> +++ b/debugfs/set_fields.c
> @@ -74,9 +74,9 @@ static struct field_set_info super_fields[] = {
> { "free_inodes_count", &set_sb.s_free_inodes_count, 4, parse_uint },
> { "first_data_block", &set_sb.s_first_data_block, 4, parse_uint },
> { "log_block_size", &set_sb.s_log_block_size, 4, parse_uint },
> - { "log_frag_size", &set_sb.s_log_frag_size, 4, parse_int },
> + { "log_cluster_size", &set_sb.s_log_cluster_size, 4, parse_int },
> { "blocks_per_group", &set_sb.s_blocks_per_group, 4, parse_uint },
> - { "frags_per_group", &set_sb.s_frags_per_group, 4, parse_uint },
> + { "clusters_per_group", &set_sb.s_clusters_per_group, 4, parse_uint },
> { "inodes_per_group", &set_sb.s_inodes_per_group, 4, parse_uint },
> { "mtime", &set_sb.s_mtime, 4, parse_time },
> { "wtime", &set_sb.s_wtime, 4, parse_time },
> diff --git a/e2fsck/super.c b/e2fsck/super.c
> index accc2f1..82c07b2 100644
> --- a/e2fsck/super.c
> +++ b/e2fsck/super.c
> @@ -501,9 +501,12 @@ void check_super_block(e2fsck_t ctx)
> check_super_value(ctx, "log_block_size", sb->s_log_block_size,
> MIN_CHECK | MAX_CHECK, 0,
> EXT2_MAX_BLOCK_LOG_SIZE - EXT2_MIN_BLOCK_LOG_SIZE);
> - check_super_value(ctx, "log_frag_size", sb->s_log_frag_size,
> - MIN_CHECK | MAX_CHECK, 0, sb->s_log_block_size);
> - check_super_value(ctx, "frags_per_group", sb->s_frags_per_group,
> + check_super_value(ctx, "log_cluster_size",
> + sb->s_log_cluster_size,
> + MIN_CHECK | MAX_CHECK, sb->s_log_block_size,
> + (EXT2_MAX_CLUSTER_LOG_SIZE -
> + EXT2_MIN_CLUSTER_LOG_SIZE));
> + check_super_value(ctx, "clusters_per_group", sb->s_clusters_per_group,
> MIN_CHECK | MAX_CHECK, sb->s_blocks_per_group,
> bpg_max);
> check_super_value(ctx, "blocks_per_group", sb->s_blocks_per_group,
> @@ -541,24 +544,14 @@ void check_super_block(e2fsck_t ctx)
> }
> }
>
> - if (sb->s_log_block_size != (__u32) sb->s_log_frag_size) {
> + if (sb->s_log_block_size != (__u32) sb->s_log_cluster_size) {
> pctx.blk = EXT2_BLOCK_SIZE(sb);
> - pctx.blk2 = EXT2_FRAG_SIZE(sb);
> + pctx.blk2 = EXT2_CLUSTER_SIZE(sb);
> fix_problem(ctx, PR_0_NO_FRAGMENTS, &pctx);
> ctx->flags |= E2F_FLAG_ABORT;
> return;
> }
>
> - should_be = sb->s_frags_per_group >>
> - (sb->s_log_block_size - sb->s_log_frag_size);
> - if (sb->s_blocks_per_group != should_be) {
> - pctx.blk = sb->s_blocks_per_group;
> - pctx.blk2 = should_be;
> - fix_problem(ctx, PR_0_BLOCKS_PER_GROUP, &pctx);
> - ctx->flags |= E2F_FLAG_ABORT;
> - return;
> - }
> -
> should_be = (sb->s_log_block_size == 0) ? 1 : 0;
> if (sb->s_first_data_block != should_be) {
> pctx.blk = sb->s_first_data_block;
> diff --git a/lib/e2p/feature.c b/lib/e2p/feature.c
> index 9324199..16fba53 100644
> --- a/lib/e2p/feature.c
> +++ b/lib/e2p/feature.c
> @@ -57,6 +57,8 @@ static struct feature feature_list[] = {
> "extra_isize" },
> { E2P_FEATURE_RO_INCOMPAT, EXT4_FEATURE_RO_COMPAT_QUOTA,
> "quota" },
> + { E2P_FEATURE_RO_INCOMPAT, EXT4_FEATURE_RO_COMPAT_BIGALLOC,
> + "bigalloc"},
>
> { E2P_FEATURE_INCOMPAT, EXT2_FEATURE_INCOMPAT_COMPRESSION,
> "compression" },
> diff --git a/lib/e2p/ls.c b/lib/e2p/ls.c
> index 8b8360a..04a2c71 100644
> --- a/lib/e2p/ls.c
> +++ b/lib/e2p/ls.c
> @@ -229,12 +229,22 @@ void list_super2(struct ext2_super_block * sb, FILE *f)
> fprintf(f, "Free inodes: %u\n", sb->s_free_inodes_count);
> fprintf(f, "First block: %u\n", sb->s_first_data_block);
> fprintf(f, "Block size: %u\n", EXT2_BLOCK_SIZE(sb));
> - fprintf(f, "Fragment size: %u\n", EXT2_FRAG_SIZE(sb));
> + if (sb->s_feature_ro_compat & EXT4_FEATURE_RO_COMPAT_BIGALLOC)
> + fprintf(f, "Cluster size: %u\n",
> + EXT2_CLUSTER_SIZE(sb));
> + else
> + fprintf(f, "Fragment size: %u\n",
> + EXT2_CLUSTER_SIZE(sb));
> if (sb->s_reserved_gdt_blocks)
> fprintf(f, "Reserved GDT blocks: %u\n",
> sb->s_reserved_gdt_blocks);
> fprintf(f, "Blocks per group: %u\n", sb->s_blocks_per_group);
> - fprintf(f, "Fragments per group: %u\n", sb->s_frags_per_group);
> + if (sb->s_feature_ro_compat & EXT4_FEATURE_RO_COMPAT_BIGALLOC)
> + fprintf(f, "Clusters per group: %u\n",
> + sb->s_clusters_per_group);
> + else
> + fprintf(f, "Fragments per group: %u\n",
> + sb->s_clusters_per_group);
> fprintf(f, "Inodes per group: %u\n", sb->s_inodes_per_group);
> fprintf(f, "Inode blocks per group: %u\n", inode_blocks_per_group);
> if (sb->s_raid_stride)
> diff --git a/lib/ext2fs/ext2_fs.h b/lib/ext2fs/ext2_fs.h
> index f396a10..b4fc6d9 100644
> --- a/lib/ext2fs/ext2_fs.h
> +++ b/lib/ext2fs/ext2_fs.h
> @@ -98,18 +98,15 @@
> #define EXT2_ADDR_PER_BLOCK(s) (EXT2_BLOCK_SIZE(s) / sizeof(__u32))
>
> /*
> - * Macro-instructions used to manage fragments
> + * Macro-instructions used to manage allocation clusters
> */
> -#define EXT2_MIN_FRAG_SIZE EXT2_MIN_BLOCK_SIZE
> -#define EXT2_MAX_FRAG_SIZE EXT2_MAX_BLOCK_SIZE
> -#define EXT2_MIN_FRAG_LOG_SIZE EXT2_MIN_BLOCK_LOG_SIZE
> -#ifdef __KERNEL__
> -# define EXT2_FRAG_SIZE(s) (EXT2_SB(s)->s_frag_size)
> -# define EXT2_FRAGS_PER_BLOCK(s) (EXT2_SB(s)->s_frags_per_block)
> -#else
> -# define EXT2_FRAG_SIZE(s) (EXT2_MIN_FRAG_SIZE << (s)->s_log_frag_size)
> -# define EXT2_FRAGS_PER_BLOCK(s) (EXT2_BLOCK_SIZE(s) / EXT2_FRAG_SIZE(s))
> -#endif
> +#define EXT2_MIN_CLUSTER_LOG_SIZE EXT2_MIN_BLOCK_LOG_SIZE
> +#define EXT2_MAX_CLUSTER_LOG_SIZE 29 /* 512MB */
> +#define EXT2_MIN_CLUSTER_SIZE EXT2_MIN_BLOCK_SIZE
> +#define EXT2_MAX_CLUSTER_SIZE (1 << EXT2_MAX_CLUSTER_LOG_SIZE)
> +#define EXT2_CLUSTER_SIZE(s) (EXT2_MIN_BLOCK_SIZE << \
> + (s)->s_log_cluster_size)
> +#define EXT2_CLUSTER_SIZE_BITS(s) ((s)->s_log_cluster_size + 10)
>
> /*
> * ACL structures
> @@ -518,9 +515,9 @@ struct ext2_super_block {
> __u32 s_free_inodes_count; /* Free inodes count */
> __u32 s_first_data_block; /* First Data Block */
> __u32 s_log_block_size; /* Block size */
> - __s32 s_log_frag_size; /* Fragment size */
> + __s32 s_log_cluster_size; /* Allocation cluster size */
> __u32 s_blocks_per_group; /* # Blocks per group */
> - __u32 s_frags_per_group; /* # Fragments per group */
> + __u32 s_clusters_per_group; /* # Fragments per group */
> __u32 s_inodes_per_group; /* # Inodes per group */
Just to be clear, my alternative suggestion to on-disk format change was:
__u32 s_log_block_size; /* Block size */
- __s32 s_log_frag_size; /* Fragment size */
+ __s32 s_log_cluster_size; /* Allocation cluster size */
- __u32 s_blocks_per_group; /* # Blocks per group */
+ __u32 s_clusters_per_group; /* # Clusters per group */
- __u32 s_frags_per_group; /* # Fragments per group */
+ __u32 s_blocks_per_group; /* # Blocks per group */
__u32 s_inodes_per_group; /* # Inodes per group */
This way, old kernels see a sane value in what used to be
s_blocks_per_group (32K)
and new kernels (even without the BIGALLOC feature) make sure to write
correct values
in both new s_blocks_per_group and s_clusters_per_group (old s_blocks_per_group)
> __u32 s_mtime; /* Mount time */
> __u32 s_wtime; /* Write time */
> @@ -675,6 +672,7 @@ struct ext2_super_block {
> #define EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE 0x0040
> #define EXT4_FEATURE_RO_COMPAT_HAS_SNAPSHOT 0x0080
> #define EXT4_FEATURE_RO_COMPAT_QUOTA 0x0100
> +#define EXT4_FEATURE_RO_COMPAT_BIGALLOC 0x0200
>
> #define EXT2_FEATURE_INCOMPAT_COMPRESSION 0x0001
> #define EXT2_FEATURE_INCOMPAT_FILETYPE 0x0002
> @@ -688,7 +686,6 @@ struct ext2_super_block {
> #define EXT4_FEATURE_INCOMPAT_EA_INODE 0x0400
> #define EXT4_FEATURE_INCOMPAT_DIRDATA 0x1000
>
> -
> #define EXT2_FEATURE_COMPAT_SUPP 0
> #define EXT2_FEATURE_INCOMPAT_SUPP (EXT2_FEATURE_INCOMPAT_FILETYPE)
> #define EXT2_FEATURE_RO_COMPAT_SUPP (EXT2_FEATURE_RO_COMPAT_SPARSE_SUPER| \
> diff --git a/lib/ext2fs/ext2fs.h b/lib/ext2fs/ext2fs.h
> index cf76562..516eb1a 100644
> --- a/lib/ext2fs/ext2fs.h
> +++ b/lib/ext2fs/ext2fs.h
> @@ -195,7 +195,7 @@ struct struct_ext2_filsys {
> char * device_name;
> struct ext2_super_block * super;
> unsigned int blocksize;
> - int fragsize;
> + int clustersize;
> dgrp_t group_desc_count;
> unsigned long desc_blocks;
> struct ext2_group_desc * group_desc;
> @@ -545,7 +545,7 @@ typedef struct ext2_icount *ext2_icount_t;
> * to ext2fs_openfs()
> */
> #define EXT2_LIB_SOFTSUPP_INCOMPAT (0)
> -#define EXT2_LIB_SOFTSUPP_RO_COMPAT (0)
> +#define EXT2_LIB_SOFTSUPP_RO_COMPAT (EXT4_FEATURE_RO_COMPAT_BIGALLOC)
>
> /*
> * function prototypes
> diff --git a/lib/ext2fs/initialize.c b/lib/ext2fs/initialize.c
> index cba2453..c109d08 100644
> --- a/lib/ext2fs/initialize.c
> +++ b/lib/ext2fs/initialize.c
> @@ -83,7 +83,6 @@ errcode_t ext2fs_initialize(const char *name, int flags,
> ext2_filsys fs;
> errcode_t retval;
> struct ext2_super_block *super;
> - int frags_per_block;
> unsigned int rem;
> unsigned int overhead = 0;
> unsigned int ipg;
> @@ -136,7 +135,7 @@ errcode_t ext2fs_initialize(const char *name, int flags,
> super->s_state = EXT2_VALID_FS;
>
> set_field(s_log_block_size, 0); /* default blocksize: 1024 bytes */
> - set_field(s_log_frag_size, 0); /* default fragsize: 1024 bytes */
> + set_field(s_log_cluster_size, 0);
> set_field(s_first_data_block, super->s_log_block_size ? 0 : 1);
> set_field(s_max_mnt_count, 0);
> set_field(s_errors, EXT2_ERRORS_DEFAULT);
> @@ -179,14 +178,13 @@ errcode_t ext2fs_initialize(const char *name, int flags,
> super->s_creator_os = CREATOR_OS;
>
> fs->blocksize = EXT2_BLOCK_SIZE(super);
> - fs->fragsize = EXT2_FRAG_SIZE(super);
> - frags_per_block = fs->blocksize / fs->fragsize;
> + fs->clustersize = EXT2_CLUSTER_SIZE(super);
>
> /* default: (fs->blocksize*8) blocks/group, up to 2^16 (GDT limit) */
> set_field(s_blocks_per_group, fs->blocksize * 8);
> if (super->s_blocks_per_group > EXT2_MAX_BLOCKS_PER_GROUP(super))
> super->s_blocks_per_group = EXT2_MAX_BLOCKS_PER_GROUP(super);
> - super->s_frags_per_group = super->s_blocks_per_group * frags_per_block;
> + super->s_clusters_per_group = super->s_blocks_per_group;
>
> super->s_blocks_count = param->s_blocks_count;
> super->s_r_blocks_count = param->s_r_blocks_count;
> @@ -239,8 +237,7 @@ retry:
> /* Try again with slightly different parameters */
> super->s_blocks_per_group -= 8;
> super->s_blocks_count = param->s_blocks_count;
> - super->s_frags_per_group = super->s_blocks_per_group *
> - frags_per_block;
> + super->s_clusters_per_group = super->s_blocks_per_group;
> goto retry;
> } else {
> retval = EXT2_ET_TOO_MANY_INODES;
> diff --git a/lib/ext2fs/openfs.c b/lib/ext2fs/openfs.c
> index d638da0..a20b923 100644
> --- a/lib/ext2fs/openfs.c
> +++ b/lib/ext2fs/openfs.c
> @@ -244,7 +244,7 @@ errcode_t ext2fs_open2(const char *name, const char *io_options,
> retval = EXT2_ET_CORRUPT_SUPERBLOCK;
> goto cleanup;
> }
> - fs->fragsize = EXT2_FRAG_SIZE(fs->super);
> + fs->clustersize = EXT2_CLUSTER_SIZE(fs->super);
> fs->inode_blocks_per_group = ((EXT2_INODES_PER_GROUP(fs->super) *
> EXT2_INODE_SIZE(fs->super) +
> EXT2_BLOCK_SIZE(fs->super) - 1) /
> diff --git a/lib/ext2fs/swapfs.c b/lib/ext2fs/swapfs.c
> index 15ff6fc..52b0fa9 100644
> --- a/lib/ext2fs/swapfs.c
> +++ b/lib/ext2fs/swapfs.c
> @@ -31,9 +31,9 @@ void ext2fs_swap_super(struct ext2_super_block * sb)
> sb->s_free_inodes_count = ext2fs_swab32(sb->s_free_inodes_count);
> sb->s_first_data_block = ext2fs_swab32(sb->s_first_data_block);
> sb->s_log_block_size = ext2fs_swab32(sb->s_log_block_size);
> - sb->s_log_frag_size = ext2fs_swab32(sb->s_log_frag_size);
> + sb->s_log_cluster_size = ext2fs_swab32(sb->s_log_cluster_size);
> sb->s_blocks_per_group = ext2fs_swab32(sb->s_blocks_per_group);
> - sb->s_frags_per_group = ext2fs_swab32(sb->s_frags_per_group);
> + sb->s_clusters_per_group = ext2fs_swab32(sb->s_clusters_per_group);
> sb->s_inodes_per_group = ext2fs_swab32(sb->s_inodes_per_group);
> sb->s_mtime = ext2fs_swab32(sb->s_mtime);
> sb->s_wtime = ext2fs_swab32(sb->s_wtime);
> diff --git a/lib/ext2fs/tst_super_size.c b/lib/ext2fs/tst_super_size.c
> index a3a5b44..6fffcfd 100644
> --- a/lib/ext2fs/tst_super_size.c
> +++ b/lib/ext2fs/tst_super_size.c
> @@ -50,9 +50,9 @@ void check_superblock_fields()
> check_field(s_free_inodes_count);
> check_field(s_first_data_block);
> check_field(s_log_block_size);
> - check_field(s_log_frag_size);
> + check_field(s_log_cluster_size);
> check_field(s_blocks_per_group);
> - check_field(s_frags_per_group);
> + check_field(s_clusters_per_group);
> check_field(s_inodes_per_group);
> check_field(s_mtime);
> check_field(s_wtime);
> diff --git a/misc/mke2fs.c b/misc/mke2fs.c
> index a5cf16b..eace0d3 100644
> --- a/misc/mke2fs.c
> +++ b/misc/mke2fs.c
> @@ -616,8 +616,13 @@ static void show_stats(ext2_filsys fs)
> printf("\n");
> printf(_("Block size=%u (log=%u)\n"), fs->blocksize,
> s->s_log_block_size);
> - printf(_("Fragment size=%u (log=%u)\n"), fs->fragsize,
> - s->s_log_frag_size);
> + if (EXT2_HAS_RO_COMPAT_FEATURE(fs->super,
> + EXT4_FEATURE_RO_COMPAT_BIGALLOC))
> + printf(_("Cluster size=%u (log=%u)\n"),
> + fs->clustersize, s->s_log_cluster_size);
> + else
> + printf(_("Fragment size=%u (log=%u)\n"), fs->clustersize,
> + s->s_log_cluster_size);
> printf(_("Stride=%u blocks, Stripe width=%u blocks\n"),
> s->s_raid_stride, s->s_raid_stripe_width);
> printf(_("%u inodes, %u blocks\n"), s->s_inodes_count,
> @@ -634,8 +639,13 @@ static void show_stats(ext2_filsys fs)
> printf(_("%u block groups\n"), fs->group_desc_count);
> else
> printf(_("%u block group\n"), fs->group_desc_count);
> - printf(_("%u blocks per group, %u fragments per group\n"),
> - s->s_blocks_per_group, s->s_frags_per_group);
> + if (EXT2_HAS_RO_COMPAT_FEATURE(fs->super,
> + EXT4_FEATURE_RO_COMPAT_BIGALLOC))
> + printf(_("%u blocks per group, %u clusters per group\n"),
> + s->s_blocks_per_group, s->s_clusters_per_group);
> + else
> + printf(_("%u blocks per group, %u fragments per group\n"),
> + s->s_blocks_per_group, s->s_clusters_per_group);
> printf(_("%u inodes per group\n"), s->s_inodes_per_group);
>
> if (fs->group_desc_count == 1) {
> @@ -1309,8 +1319,6 @@ static void PRS(int argc, char *argv[])
> optarg);
> exit(1);
> }
> - fs_param.s_log_frag_size =
> - int_log2(size >> EXT2_MIN_BLOCK_LOG_SIZE);
> fprintf(stderr, _("Warning: fragments not supported. "
> "Ignoring -f option\n"));
> break;
> @@ -1542,7 +1550,7 @@ static void PRS(int argc, char *argv[])
> check_plausibility(device_name);
> check_mount(device_name, force, _("filesystem"));
>
> - fs_param.s_log_frag_size = fs_param.s_log_block_size;
> + fs_param.s_log_cluster_size = fs_param.s_log_block_size;
>
> if (noaction && fs_param.s_blocks_count) {
> dev_size = fs_param.s_blocks_count;
> @@ -1780,7 +1788,7 @@ got_size:
> inode_ratio = blocksize;
> }
>
> - fs_param.s_log_frag_size = fs_param.s_log_block_size =
> + fs_param.s_log_cluster_size = fs_param.s_log_block_size =
> int_log2(blocksize >> EXT2_MIN_BLOCK_LOG_SIZE);
>
> #ifdef HAVE_BLKID_PROBE_GET_TOPOLOGY
> --
> 1.7.3.1
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
> the body of a message to majordomo@...r.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
To unsubscribe from this list: send the line "unsubscribe linux-ext4" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Powered by blists - more mailing lists