linux/fs/erofs/zmap.c

759 lines
21 KiB
C
Raw Normal View History

// SPDX-License-Identifier: GPL-2.0-only
/*
* Copyright (C) 2018-2019 HUAWEI, Inc.
* https://www.huawei.com/
*/
#include "internal.h"
#include <asm/unaligned.h>
#include <trace/events/erofs.h>
struct z_erofs_maprecorder {
struct inode *inode;
struct erofs_map_blocks *map;
void *kaddr;
unsigned long lcn;
/* compression extent information gathered */
u8 type, headtype;
u16 clusterofs;
u16 delta[2];
erofs_blk_t pblk, compressedblks;
erofs_off_t nextpackoff;
bool partialref;
};
static int z_erofs_load_full_lcluster(struct z_erofs_maprecorder *m,
unsigned long lcn)
{
struct inode *const inode = m->inode;
struct erofs_inode *const vi = EROFS_I(inode);
const erofs_off_t pos = Z_EROFS_FULL_INDEX_ALIGN(erofs_iloc(inode) +
vi->inode_isize + vi->xattr_isize) +
lcn * sizeof(struct z_erofs_lcluster_index);
struct z_erofs_lcluster_index *di;
unsigned int advise;
m->kaddr = erofs_read_metabuf(&m->map->buf, inode->i_sb,
pos, EROFS_KMAP);
if (IS_ERR(m->kaddr))
return PTR_ERR(m->kaddr);
m->nextpackoff = pos + sizeof(struct z_erofs_lcluster_index);
m->lcn = lcn;
di = m->kaddr;
advise = le16_to_cpu(di->di_advise);
m->type = advise & Z_EROFS_LI_LCLUSTER_TYPE_MASK;
if (m->type == Z_EROFS_LCLUSTER_TYPE_NONHEAD) {
m->clusterofs = 1 << vi->z_logical_clusterbits;
m->delta[0] = le16_to_cpu(di->di_u.delta[0]);
if (m->delta[0] & Z_EROFS_LI_D0_CBLKCNT) {
if (!(vi->z_advise & (Z_EROFS_ADVISE_BIG_PCLUSTER_1 |
Z_EROFS_ADVISE_BIG_PCLUSTER_2))) {
DBG_BUGON(1);
return -EFSCORRUPTED;
}
m->compressedblks = m->delta[0] &
~Z_EROFS_LI_D0_CBLKCNT;
m->delta[0] = 1;
}
m->delta[1] = le16_to_cpu(di->di_u.delta[1]);
} else {
m->partialref = !!(advise & Z_EROFS_LI_PARTIAL_REF);
m->clusterofs = le16_to_cpu(di->di_clusterofs);
2023-04-11 01:37:14 +08:00
if (m->clusterofs >= 1 << vi->z_logical_clusterbits) {
DBG_BUGON(1);
return -EFSCORRUPTED;
}
m->pblk = le32_to_cpu(di->di_u.blkaddr);
}
return 0;
}
static unsigned int decode_compactedbits(unsigned int lobits,
u8 *in, unsigned int pos, u8 *type)
{
const unsigned int v = get_unaligned_le32(in + pos / 8) >> (pos & 7);
const unsigned int lo = v & ((1 << lobits) - 1);
*type = (v >> lobits) & 3;
return lo;
}
static int get_compacted_la_distance(unsigned int lobits,
unsigned int encodebits,
unsigned int vcnt, u8 *in, int i)
{
unsigned int lo, d1 = 0;
u8 type;
DBG_BUGON(i >= vcnt);
do {
lo = decode_compactedbits(lobits, in, encodebits * i, &type);
if (type != Z_EROFS_LCLUSTER_TYPE_NONHEAD)
return d1;
++d1;
} while (++i < vcnt);
/* vcnt - 1 (Z_EROFS_LCLUSTER_TYPE_NONHEAD) item */
if (!(lo & Z_EROFS_LI_D0_CBLKCNT))
d1 += lo - 1;
return d1;
}
static int unpack_compacted_index(struct z_erofs_maprecorder *m,
unsigned int amortizedshift,
erofs_off_t pos, bool lookahead)
{
struct erofs_inode *const vi = EROFS_I(m->inode);
const unsigned int lclusterbits = vi->z_logical_clusterbits;
unsigned int vcnt, lo, lobits, encodebits, nblk, bytes;
int i;
u8 *in, type;
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
bool big_pcluster;
if (1 << amortizedshift == 4 && lclusterbits <= 14)
vcnt = 2;
else if (1 << amortizedshift == 2 && lclusterbits <= 12)
vcnt = 16;
else
return -EOPNOTSUPP;
/* it doesn't equal to round_up(..) */
m->nextpackoff = round_down(pos, vcnt << amortizedshift) +
(vcnt << amortizedshift);
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
big_pcluster = vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_1;
lobits = max(lclusterbits, ilog2(Z_EROFS_LI_D0_CBLKCNT) + 1U);
encodebits = ((vcnt << amortizedshift) - sizeof(__le32)) * 8 / vcnt;
bytes = pos & ((vcnt << amortizedshift) - 1);
in = m->kaddr - bytes;
i = bytes >> amortizedshift;
lo = decode_compactedbits(lobits, in, encodebits * i, &type);
m->type = type;
if (type == Z_EROFS_LCLUSTER_TYPE_NONHEAD) {
m->clusterofs = 1 << lclusterbits;
/* figure out lookahead_distance: delta[1] if needed */
if (lookahead)
m->delta[1] = get_compacted_la_distance(lobits,
encodebits, vcnt, in, i);
if (lo & Z_EROFS_LI_D0_CBLKCNT) {
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
if (!big_pcluster) {
DBG_BUGON(1);
return -EFSCORRUPTED;
}
m->compressedblks = lo & ~Z_EROFS_LI_D0_CBLKCNT;
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
m->delta[0] = 1;
return 0;
} else if (i + 1 != (int)vcnt) {
m->delta[0] = lo;
return 0;
}
/*
* since the last lcluster in the pack is special,
* of which lo saves delta[1] rather than delta[0].
* Hence, get delta[0] by the previous lcluster indirectly.
*/
lo = decode_compactedbits(lobits, in,
encodebits * (i - 1), &type);
if (type != Z_EROFS_LCLUSTER_TYPE_NONHEAD)
lo = 0;
else if (lo & Z_EROFS_LI_D0_CBLKCNT)
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
lo = 1;
m->delta[0] = lo + 1;
return 0;
}
m->clusterofs = lo;
m->delta[0] = 0;
/* figout out blkaddr (pblk) for HEAD lclusters */
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
if (!big_pcluster) {
nblk = 1;
while (i > 0) {
--i;
lo = decode_compactedbits(lobits, in,
encodebits * i, &type);
if (type == Z_EROFS_LCLUSTER_TYPE_NONHEAD)
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
i -= lo;
if (i >= 0)
++nblk;
}
} else {
nblk = 0;
while (i > 0) {
--i;
lo = decode_compactedbits(lobits, in,
encodebits * i, &type);
if (type == Z_EROFS_LCLUSTER_TYPE_NONHEAD) {
if (lo & Z_EROFS_LI_D0_CBLKCNT) {
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
--i;
nblk += lo & ~Z_EROFS_LI_D0_CBLKCNT;
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
continue;
}
/* bigpcluster shouldn't have plain d0 == 1 */
if (lo <= 1) {
DBG_BUGON(1);
return -EFSCORRUPTED;
}
i -= lo - 2;
continue;
}
++nblk;
erofs: support parsing big pcluster compact indexes Different from non-compact indexes, several lclusters are packed as the compact form at once and an unique base blkaddr is stored for each pack, so each lcluster index would take less space on avarage (e.g. 2 bytes for COMPACT_2B.) btw, that is also why BIG_PCLUSTER switch should be consistent for compact head0/1. Prior to big pcluster, the size of all pclusters was 1 lcluster. Therefore, when a new HEAD lcluster was scanned, blkaddr would be bumped by 1 lcluster. However, that way doesn't work anymore for big pcluster since we actually don't know the compressed size of pclusters in advance (before reading CBLKCNT lcluster). So, instead, let blkaddr of each pack be the first pcluster blkaddr with a valid CBLKCNT, in detail, 1) if CBLKCNT starts at the pack, this first valid pcluster is itself, e.g. _____________________________________________________________ |_CBLKCNT0_|_NONHEAD_| .. |_HEAD_|_CBLKCNT1_| ... |_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += CBLKCNT1 2) if CBLKCNT doesn't start at the pack, the first valid pcluster is the next pcluster, e.g. _________________________________________________________ | NONHEAD_| .. |_HEAD_|_CBLKCNT0_| ... |_HEAD_|_HEAD_| ... ^ = blkaddr base ^ += CBLKCNT0 ^ += 1 When a CBLKCNT is found, blkaddr will be increased by CBLKCNT lclusters, or a new HEAD is found immediately, bump blkaddr by 1 instead (see the picture above.) Also noted if CBLKCNT is the end of the pack, instead of storing delta1 (distance of the next HEAD lcluster) as normal NONHEADs, it still uses the compressed block count (delta0) since delta1 can be calculated indirectly but the block count can't. Adjust decoding logic to fit big pcluster compact indexes as well. Link: https://lore.kernel.org/r/20210407043927.10623-9-xiang@kernel.org Acked-by: Chao Yu <yuchao0@huawei.com> Signed-off-by: Gao Xiang <hsiangkao@redhat.com>
2021-04-07 12:39:25 +08:00
}
}
in += (vcnt << amortizedshift) - sizeof(__le32);
m->pblk = le32_to_cpu(*(__le32 *)in) + nblk;
return 0;
}
static int z_erofs_load_compact_lcluster(struct z_erofs_maprecorder *m,
unsigned long lcn, bool lookahead)
{
struct inode *const inode = m->inode;
struct erofs_inode *const vi = EROFS_I(inode);
const erofs_off_t ebase = sizeof(struct z_erofs_map_header) +
ALIGN(erofs_iloc(inode) + vi->inode_isize + vi->xattr_isize, 8);
erofs: avoid hardcoded blocksize for subpage block support As the first step of converting hardcoded blocksize to that specified in on-disk superblock, convert all call sites of hardcoded blocksize to sb->s_blocksize except for: 1) use sbi->blkszbits instead of sb->s_blocksize in erofs_superblock_csum_verify() since sb->s_blocksize has not been updated with the on-disk blocksize yet when the function is called. 2) use inode->i_blkbits instead of sb->s_blocksize in erofs_bread(), since the inode operated on may be an anonymous inode in fscache mode. Currently the anonymous inode is allocated from an anonymous mount maintained in erofs, while in the near future we may allocate anonymous inodes from a generic API directly and thus have no access to the anonymous inode's i_sb. Thus we keep the block size in i_blkbits for anonymous inodes in fscache mode. Be noted that this patch only gets rid of the hardcoded blocksize, in preparation for actually setting the on-disk block size in the following patch. The hard limit of constraining the block size to PAGE_SIZE still exists until the next patch. Signed-off-by: Jingbo Xu <jefflexu@linux.alibaba.com> Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com> Reviewed-by: Yue Hu <huyue2@coolpad.com> Reviewed-by: Chao Yu <chao@kernel.org> Link: https://lore.kernel.org/r/20230313135309.75269-2-jefflexu@linux.alibaba.com [ Gao Xiang: fold a patch to fix incorrect truncated offsets. ] Link: https://lore.kernel.org/r/20230413035734.15457-1-zhujia.zj@bytedance.com Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
2023-03-13 21:53:08 +08:00
unsigned int totalidx = erofs_iblks(inode);
unsigned int compacted_4b_initial, compacted_2b;
unsigned int amortizedshift;
erofs_off_t pos;
if (lcn >= totalidx)
return -EINVAL;
m->lcn = lcn;
/* used to align to 32-byte (compacted_2b) alignment */
compacted_4b_initial = (32 - ebase % 32) / 4;
if (compacted_4b_initial == 32 / 4)
compacted_4b_initial = 0;
if ((vi->z_advise & Z_EROFS_ADVISE_COMPACTED_2B) &&
compacted_4b_initial < totalidx)
compacted_2b = rounddown(totalidx - compacted_4b_initial, 16);
else
compacted_2b = 0;
pos = ebase;
if (lcn < compacted_4b_initial) {
amortizedshift = 2;
goto out;
}
pos += compacted_4b_initial * 4;
lcn -= compacted_4b_initial;
if (lcn < compacted_2b) {
amortizedshift = 1;
goto out;
}
pos += compacted_2b * 2;
lcn -= compacted_2b;
amortizedshift = 2;
out:
pos += lcn * (1 << amortizedshift);
m->kaddr = erofs_read_metabuf(&m->map->buf, inode->i_sb,
pos, EROFS_KMAP);
if (IS_ERR(m->kaddr))
return PTR_ERR(m->kaddr);
return unpack_compacted_index(m, amortizedshift, pos, lookahead);
}
static int z_erofs_load_lcluster_from_disk(struct z_erofs_maprecorder *m,
unsigned int lcn, bool lookahead)
{
switch (EROFS_I(m->inode)->datalayout) {
case EROFS_INODE_COMPRESSED_FULL:
return z_erofs_load_full_lcluster(m, lcn);
case EROFS_INODE_COMPRESSED_COMPACT:
return z_erofs_load_compact_lcluster(m, lcn, lookahead);
default:
return -EINVAL;
}
}
static int z_erofs_extent_lookback(struct z_erofs_maprecorder *m,
unsigned int lookback_distance)
{
struct super_block *sb = m->inode->i_sb;
struct erofs_inode *const vi = EROFS_I(m->inode);
const unsigned int lclusterbits = vi->z_logical_clusterbits;
while (m->lcn >= lookback_distance) {
unsigned long lcn = m->lcn - lookback_distance;
int err;
err = z_erofs_load_lcluster_from_disk(m, lcn, false);
if (err)
return err;
switch (m->type) {
case Z_EROFS_LCLUSTER_TYPE_NONHEAD:
lookback_distance = m->delta[0];
if (!lookback_distance)
goto err_bogus;
continue;
case Z_EROFS_LCLUSTER_TYPE_PLAIN:
case Z_EROFS_LCLUSTER_TYPE_HEAD1:
case Z_EROFS_LCLUSTER_TYPE_HEAD2:
m->headtype = m->type;
m->map->m_la = (lcn << lclusterbits) | m->clusterofs;
return 0;
default:
erofs_err(sb, "unknown type %u @ lcn %lu of nid %llu",
m->type, lcn, vi->nid);
DBG_BUGON(1);
return -EOPNOTSUPP;
}
}
err_bogus:
erofs_err(sb, "bogus lookback distance %u @ lcn %lu of nid %llu",
lookback_distance, m->lcn, vi->nid);
DBG_BUGON(1);
return -EFSCORRUPTED;
}
static int z_erofs_get_extent_compressedlen(struct z_erofs_maprecorder *m,
unsigned int initial_lcn)
{
erofs: avoid hardcoded blocksize for subpage block support As the first step of converting hardcoded blocksize to that specified in on-disk superblock, convert all call sites of hardcoded blocksize to sb->s_blocksize except for: 1) use sbi->blkszbits instead of sb->s_blocksize in erofs_superblock_csum_verify() since sb->s_blocksize has not been updated with the on-disk blocksize yet when the function is called. 2) use inode->i_blkbits instead of sb->s_blocksize in erofs_bread(), since the inode operated on may be an anonymous inode in fscache mode. Currently the anonymous inode is allocated from an anonymous mount maintained in erofs, while in the near future we may allocate anonymous inodes from a generic API directly and thus have no access to the anonymous inode's i_sb. Thus we keep the block size in i_blkbits for anonymous inodes in fscache mode. Be noted that this patch only gets rid of the hardcoded blocksize, in preparation for actually setting the on-disk block size in the following patch. The hard limit of constraining the block size to PAGE_SIZE still exists until the next patch. Signed-off-by: Jingbo Xu <jefflexu@linux.alibaba.com> Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com> Reviewed-by: Yue Hu <huyue2@coolpad.com> Reviewed-by: Chao Yu <chao@kernel.org> Link: https://lore.kernel.org/r/20230313135309.75269-2-jefflexu@linux.alibaba.com [ Gao Xiang: fold a patch to fix incorrect truncated offsets. ] Link: https://lore.kernel.org/r/20230413035734.15457-1-zhujia.zj@bytedance.com Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
2023-03-13 21:53:08 +08:00
struct super_block *sb = m->inode->i_sb;
struct erofs_inode *const vi = EROFS_I(m->inode);
struct erofs_map_blocks *const map = m->map;
const unsigned int lclusterbits = vi->z_logical_clusterbits;
unsigned long lcn;
int err;
DBG_BUGON(m->type != Z_EROFS_LCLUSTER_TYPE_PLAIN &&
m->type != Z_EROFS_LCLUSTER_TYPE_HEAD1 &&
m->type != Z_EROFS_LCLUSTER_TYPE_HEAD2);
DBG_BUGON(m->type != m->headtype);
if (m->headtype == Z_EROFS_LCLUSTER_TYPE_PLAIN ||
((m->headtype == Z_EROFS_LCLUSTER_TYPE_HEAD1) &&
!(vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_1)) ||
((m->headtype == Z_EROFS_LCLUSTER_TYPE_HEAD2) &&
!(vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_2))) {
map->m_plen = 1ULL << lclusterbits;
return 0;
}
lcn = m->lcn + 1;
if (m->compressedblks)
goto out;
err = z_erofs_load_lcluster_from_disk(m, lcn, false);
if (err)
return err;
/*
* If the 1st NONHEAD lcluster has already been handled initially w/o
* valid compressedblks, which means at least it mustn't be CBLKCNT, or
* an internal implemenatation error is detected.
*
* The following code can also handle it properly anyway, but let's
* BUG_ON in the debugging mode only for developers to notice that.
*/
DBG_BUGON(lcn == initial_lcn &&
m->type == Z_EROFS_LCLUSTER_TYPE_NONHEAD);
switch (m->type) {
case Z_EROFS_LCLUSTER_TYPE_PLAIN:
case Z_EROFS_LCLUSTER_TYPE_HEAD1:
case Z_EROFS_LCLUSTER_TYPE_HEAD2:
/*
* if the 1st NONHEAD lcluster is actually PLAIN or HEAD type
* rather than CBLKCNT, it's a 1 lcluster-sized pcluster.
*/
erofs: avoid hardcoded blocksize for subpage block support As the first step of converting hardcoded blocksize to that specified in on-disk superblock, convert all call sites of hardcoded blocksize to sb->s_blocksize except for: 1) use sbi->blkszbits instead of sb->s_blocksize in erofs_superblock_csum_verify() since sb->s_blocksize has not been updated with the on-disk blocksize yet when the function is called. 2) use inode->i_blkbits instead of sb->s_blocksize in erofs_bread(), since the inode operated on may be an anonymous inode in fscache mode. Currently the anonymous inode is allocated from an anonymous mount maintained in erofs, while in the near future we may allocate anonymous inodes from a generic API directly and thus have no access to the anonymous inode's i_sb. Thus we keep the block size in i_blkbits for anonymous inodes in fscache mode. Be noted that this patch only gets rid of the hardcoded blocksize, in preparation for actually setting the on-disk block size in the following patch. The hard limit of constraining the block size to PAGE_SIZE still exists until the next patch. Signed-off-by: Jingbo Xu <jefflexu@linux.alibaba.com> Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com> Reviewed-by: Yue Hu <huyue2@coolpad.com> Reviewed-by: Chao Yu <chao@kernel.org> Link: https://lore.kernel.org/r/20230313135309.75269-2-jefflexu@linux.alibaba.com [ Gao Xiang: fold a patch to fix incorrect truncated offsets. ] Link: https://lore.kernel.org/r/20230413035734.15457-1-zhujia.zj@bytedance.com Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
2023-03-13 21:53:08 +08:00
m->compressedblks = 1 << (lclusterbits - sb->s_blocksize_bits);
break;
case Z_EROFS_LCLUSTER_TYPE_NONHEAD:
if (m->delta[0] != 1)
goto err_bonus_cblkcnt;
if (m->compressedblks)
break;
fallthrough;
default:
erofs_err(sb, "cannot found CBLKCNT @ lcn %lu of nid %llu", lcn,
vi->nid);
DBG_BUGON(1);
return -EFSCORRUPTED;
}
out:
erofs: avoid hardcoded blocksize for subpage block support As the first step of converting hardcoded blocksize to that specified in on-disk superblock, convert all call sites of hardcoded blocksize to sb->s_blocksize except for: 1) use sbi->blkszbits instead of sb->s_blocksize in erofs_superblock_csum_verify() since sb->s_blocksize has not been updated with the on-disk blocksize yet when the function is called. 2) use inode->i_blkbits instead of sb->s_blocksize in erofs_bread(), since the inode operated on may be an anonymous inode in fscache mode. Currently the anonymous inode is allocated from an anonymous mount maintained in erofs, while in the near future we may allocate anonymous inodes from a generic API directly and thus have no access to the anonymous inode's i_sb. Thus we keep the block size in i_blkbits for anonymous inodes in fscache mode. Be noted that this patch only gets rid of the hardcoded blocksize, in preparation for actually setting the on-disk block size in the following patch. The hard limit of constraining the block size to PAGE_SIZE still exists until the next patch. Signed-off-by: Jingbo Xu <jefflexu@linux.alibaba.com> Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com> Reviewed-by: Yue Hu <huyue2@coolpad.com> Reviewed-by: Chao Yu <chao@kernel.org> Link: https://lore.kernel.org/r/20230313135309.75269-2-jefflexu@linux.alibaba.com [ Gao Xiang: fold a patch to fix incorrect truncated offsets. ] Link: https://lore.kernel.org/r/20230413035734.15457-1-zhujia.zj@bytedance.com Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
2023-03-13 21:53:08 +08:00
map->m_plen = erofs_pos(sb, m->compressedblks);
return 0;
err_bonus_cblkcnt:
erofs_err(sb, "bogus CBLKCNT @ lcn %lu of nid %llu", lcn, vi->nid);
DBG_BUGON(1);
return -EFSCORRUPTED;
}
static int z_erofs_get_extent_decompressedlen(struct z_erofs_maprecorder *m)
{
struct inode *inode = m->inode;
struct erofs_inode *vi = EROFS_I(inode);
struct erofs_map_blocks *map = m->map;
unsigned int lclusterbits = vi->z_logical_clusterbits;
u64 lcn = m->lcn, headlcn = map->m_la >> lclusterbits;
int err;
do {
/* handle the last EOF pcluster (no next HEAD lcluster) */
if ((lcn << lclusterbits) >= inode->i_size) {
map->m_llen = inode->i_size - map->m_la;
return 0;
}
err = z_erofs_load_lcluster_from_disk(m, lcn, true);
if (err)
return err;
if (m->type == Z_EROFS_LCLUSTER_TYPE_NONHEAD) {
DBG_BUGON(!m->delta[1] &&
m->clusterofs != 1 << lclusterbits);
} else if (m->type == Z_EROFS_LCLUSTER_TYPE_PLAIN ||
m->type == Z_EROFS_LCLUSTER_TYPE_HEAD1 ||
m->type == Z_EROFS_LCLUSTER_TYPE_HEAD2) {
/* go on until the next HEAD lcluster */
if (lcn != headlcn)
break;
m->delta[1] = 1;
} else {
erofs_err(inode->i_sb, "unknown type %u @ lcn %llu of nid %llu",
m->type, lcn, vi->nid);
DBG_BUGON(1);
return -EOPNOTSUPP;
}
lcn += m->delta[1];
} while (m->delta[1]);
map->m_llen = (lcn << lclusterbits) + m->clusterofs - map->m_la;
return 0;
}
static int z_erofs_do_map_blocks(struct inode *inode,
struct erofs_map_blocks *map, int flags)
{
struct erofs_inode *const vi = EROFS_I(inode);
bool ztailpacking = vi->z_advise & Z_EROFS_ADVISE_INLINE_PCLUSTER;
bool fragment = vi->z_advise & Z_EROFS_ADVISE_FRAGMENT_PCLUSTER;
struct z_erofs_maprecorder m = {
.inode = inode,
.map = map,
};
int err = 0;
unsigned int lclusterbits, endoff, afmt;
unsigned long initial_lcn;
unsigned long long ofs, end;
lclusterbits = vi->z_logical_clusterbits;
ofs = flags & EROFS_GET_BLOCKS_FINDTAIL ? inode->i_size - 1 : map->m_la;
initial_lcn = ofs >> lclusterbits;
endoff = ofs & ((1 << lclusterbits) - 1);
err = z_erofs_load_lcluster_from_disk(&m, initial_lcn, false);
if (err)
goto unmap_out;
if (ztailpacking && (flags & EROFS_GET_BLOCKS_FINDTAIL))
vi->z_idataoff = m.nextpackoff;
map->m_flags = EROFS_MAP_MAPPED | EROFS_MAP_ENCODED;
end = (m.lcn + 1ULL) << lclusterbits;
switch (m.type) {
case Z_EROFS_LCLUSTER_TYPE_PLAIN:
case Z_EROFS_LCLUSTER_TYPE_HEAD1:
case Z_EROFS_LCLUSTER_TYPE_HEAD2:
if (endoff >= m.clusterofs) {
m.headtype = m.type;
map->m_la = (m.lcn << lclusterbits) | m.clusterofs;
/*
* For ztailpacking files, in order to inline data more
* effectively, special EOF lclusters are now supported
* which can have three parts at most.
*/
if (ztailpacking && end > inode->i_size)
end = inode->i_size;
break;
}
/* m.lcn should be >= 1 if endoff < m.clusterofs */
if (!m.lcn) {
erofs_err(inode->i_sb,
"invalid logical cluster 0 at nid %llu",
vi->nid);
err = -EFSCORRUPTED;
goto unmap_out;
}
end = (m.lcn << lclusterbits) | m.clusterofs;
map->m_flags |= EROFS_MAP_FULL_MAPPED;
m.delta[0] = 1;
fallthrough;
case Z_EROFS_LCLUSTER_TYPE_NONHEAD:
/* get the corresponding first chunk */
err = z_erofs_extent_lookback(&m, m.delta[0]);
if (err)
goto unmap_out;
break;
default:
erofs_err(inode->i_sb,
"unknown type %u @ offset %llu of nid %llu",
m.type, ofs, vi->nid);
err = -EOPNOTSUPP;
goto unmap_out;
}
if (m.partialref)
map->m_flags |= EROFS_MAP_PARTIAL_REF;
map->m_llen = end - map->m_la;
if (flags & EROFS_GET_BLOCKS_FINDTAIL) {
vi->z_tailextent_headlcn = m.lcn;
/* for non-compact indexes, fragmentoff is 64 bits */
if (fragment && vi->datalayout == EROFS_INODE_COMPRESSED_FULL)
vi->z_fragmentoff |= (u64)m.pblk << 32;
}
if (ztailpacking && m.lcn == vi->z_tailextent_headlcn) {
map->m_flags |= EROFS_MAP_META;
map->m_pa = vi->z_idataoff;
map->m_plen = vi->z_idata_size;
} else if (fragment && m.lcn == vi->z_tailextent_headlcn) {
map->m_flags |= EROFS_MAP_FRAGMENT;
} else {
erofs: avoid hardcoded blocksize for subpage block support As the first step of converting hardcoded blocksize to that specified in on-disk superblock, convert all call sites of hardcoded blocksize to sb->s_blocksize except for: 1) use sbi->blkszbits instead of sb->s_blocksize in erofs_superblock_csum_verify() since sb->s_blocksize has not been updated with the on-disk blocksize yet when the function is called. 2) use inode->i_blkbits instead of sb->s_blocksize in erofs_bread(), since the inode operated on may be an anonymous inode in fscache mode. Currently the anonymous inode is allocated from an anonymous mount maintained in erofs, while in the near future we may allocate anonymous inodes from a generic API directly and thus have no access to the anonymous inode's i_sb. Thus we keep the block size in i_blkbits for anonymous inodes in fscache mode. Be noted that this patch only gets rid of the hardcoded blocksize, in preparation for actually setting the on-disk block size in the following patch. The hard limit of constraining the block size to PAGE_SIZE still exists until the next patch. Signed-off-by: Jingbo Xu <jefflexu@linux.alibaba.com> Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com> Reviewed-by: Yue Hu <huyue2@coolpad.com> Reviewed-by: Chao Yu <chao@kernel.org> Link: https://lore.kernel.org/r/20230313135309.75269-2-jefflexu@linux.alibaba.com [ Gao Xiang: fold a patch to fix incorrect truncated offsets. ] Link: https://lore.kernel.org/r/20230413035734.15457-1-zhujia.zj@bytedance.com Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
2023-03-13 21:53:08 +08:00
map->m_pa = erofs_pos(inode->i_sb, m.pblk);
err = z_erofs_get_extent_compressedlen(&m, initial_lcn);
if (err)
goto unmap_out;
}
if (m.headtype == Z_EROFS_LCLUSTER_TYPE_PLAIN) {
if (map->m_llen > map->m_plen) {
DBG_BUGON(1);
err = -EFSCORRUPTED;
goto unmap_out;
}
afmt = vi->z_advise & Z_EROFS_ADVISE_INTERLACED_PCLUSTER ?
Z_EROFS_COMPRESSION_INTERLACED :
Z_EROFS_COMPRESSION_SHIFTED;
} else {
afmt = m.headtype == Z_EROFS_LCLUSTER_TYPE_HEAD2 ?
vi->z_algorithmtype[1] : vi->z_algorithmtype[0];
if (!(EROFS_I_SB(inode)->available_compr_algs & (1 << afmt))) {
erofs_err(inode->i_sb, "inconsistent algorithmtype %u for nid %llu",
afmt, vi->nid);
err = -EFSCORRUPTED;
goto unmap_out;
}
}
map->m_algorithmformat = afmt;
if ((flags & EROFS_GET_BLOCKS_FIEMAP) ||
((flags & EROFS_GET_BLOCKS_READMORE) &&
erofs: DEFLATE compression support Add DEFLATE compression as the 3rd supported algorithm. DEFLATE is a popular generic-purpose compression algorithm for quite long time (many advanced formats like gzip, zlib, zip, png are all based on that) as Apple documentation written "If you require interoperability with non-Apple devices, use COMPRESSION_ZLIB. [1]". Due to its popularity, there are several hardware on-market DEFLATE accelerators, such as (s390) DFLTCC, (Intel) IAA/QAT, (HiSilicon) ZIP accelerator, etc. In addition, there are also several high-performence IP cores and even open-source FPGA approches available for DEFLATE. Therefore, it's useful to support DEFLATE compression in order to find a way to utilize these accelerators for asynchronous I/Os and get benefits from these later. Besides, it's a good choice to trade off between compression ratios and performance compared to LZ4 and LZMA. The DEFLATE core format is simple as well as easy to understand, therefore the code size of its decompressor is small even for the bootloader use cases. The runtime memory consumption is quite limited too (e.g. 32K + ~7K for each zlib stream). As usual, EROFS ourperforms similar approaches too. Alternatively, DEFLATE could still be used for some specific files since EROFS supports multiple compression algorithms in one image. [1] https://developer.apple.com/documentation/compression/compression_algorithm Reviewed-by: Chao Yu <chao@kernel.org> Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com> Link: https://lore.kernel.org/r/20230810154859.118330-1-hsiangkao@linux.alibaba.com
2023-08-10 23:48:59 +08:00
(map->m_algorithmformat == Z_EROFS_COMPRESSION_LZMA ||
map->m_algorithmformat == Z_EROFS_COMPRESSION_DEFLATE ||
map->m_algorithmformat == Z_EROFS_COMPRESSION_ZSTD) &&
erofs: DEFLATE compression support Add DEFLATE compression as the 3rd supported algorithm. DEFLATE is a popular generic-purpose compression algorithm for quite long time (many advanced formats like gzip, zlib, zip, png are all based on that) as Apple documentation written "If you require interoperability with non-Apple devices, use COMPRESSION_ZLIB. [1]". Due to its popularity, there are several hardware on-market DEFLATE accelerators, such as (s390) DFLTCC, (Intel) IAA/QAT, (HiSilicon) ZIP accelerator, etc. In addition, there are also several high-performence IP cores and even open-source FPGA approches available for DEFLATE. Therefore, it's useful to support DEFLATE compression in order to find a way to utilize these accelerators for asynchronous I/Os and get benefits from these later. Besides, it's a good choice to trade off between compression ratios and performance compared to LZ4 and LZMA. The DEFLATE core format is simple as well as easy to understand, therefore the code size of its decompressor is small even for the bootloader use cases. The runtime memory consumption is quite limited too (e.g. 32K + ~7K for each zlib stream). As usual, EROFS ourperforms similar approaches too. Alternatively, DEFLATE could still be used for some specific files since EROFS supports multiple compression algorithms in one image. [1] https://developer.apple.com/documentation/compression/compression_algorithm Reviewed-by: Chao Yu <chao@kernel.org> Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com> Link: https://lore.kernel.org/r/20230810154859.118330-1-hsiangkao@linux.alibaba.com
2023-08-10 23:48:59 +08:00
map->m_llen >= i_blocksize(inode))) {
err = z_erofs_get_extent_decompressedlen(&m);
if (!err)
map->m_flags |= EROFS_MAP_FULL_MAPPED;
}
unmap_out:
erofs_unmap_metabuf(&m.map->buf);
return err;
}
static int z_erofs_fill_inode_lazy(struct inode *inode)
{
struct erofs_inode *const vi = EROFS_I(inode);
struct super_block *const sb = inode->i_sb;
int err, headnr;
erofs_off_t pos;
struct erofs_buf buf = __EROFS_BUF_INITIALIZER;
struct z_erofs_map_header *h;
if (test_bit(EROFS_I_Z_INITED_BIT, &vi->flags)) {
/*
* paired with smp_mb() at the end of the function to ensure
* fields will only be observed after the bit is set.
*/
smp_mb();
return 0;
}
if (wait_on_bit_lock(&vi->flags, EROFS_I_BL_Z_BIT, TASK_KILLABLE))
return -ERESTARTSYS;
err = 0;
if (test_bit(EROFS_I_Z_INITED_BIT, &vi->flags))
goto out_unlock;
pos = ALIGN(erofs_iloc(inode) + vi->inode_isize + vi->xattr_isize, 8);
h = erofs_read_metabuf(&buf, sb, pos, EROFS_KMAP);
if (IS_ERR(h)) {
err = PTR_ERR(h);
goto out_unlock;
}
/*
* if the highest bit of the 8-byte map header is set, the whole file
* is stored in the packed inode. The rest bits keeps z_fragmentoff.
*/
if (h->h_clusterbits >> Z_EROFS_FRAGMENT_INODE_BIT) {
vi->z_advise = Z_EROFS_ADVISE_FRAGMENT_PCLUSTER;
vi->z_fragmentoff = le64_to_cpu(*(__le64 *)h) ^ (1ULL << 63);
vi->z_tailextent_headlcn = 0;
goto done;
}
vi->z_advise = le16_to_cpu(h->h_advise);
vi->z_algorithmtype[0] = h->h_algorithmtype & 15;
vi->z_algorithmtype[1] = h->h_algorithmtype >> 4;
headnr = 0;
if (vi->z_algorithmtype[0] >= Z_EROFS_COMPRESSION_MAX ||
vi->z_algorithmtype[++headnr] >= Z_EROFS_COMPRESSION_MAX) {
erofs_err(sb, "unknown HEAD%u format %u for nid %llu, please upgrade kernel",
headnr + 1, vi->z_algorithmtype[headnr], vi->nid);
err = -EOPNOTSUPP;
goto out_put_metabuf;
}
erofs: avoid hardcoded blocksize for subpage block support As the first step of converting hardcoded blocksize to that specified in on-disk superblock, convert all call sites of hardcoded blocksize to sb->s_blocksize except for: 1) use sbi->blkszbits instead of sb->s_blocksize in erofs_superblock_csum_verify() since sb->s_blocksize has not been updated with the on-disk blocksize yet when the function is called. 2) use inode->i_blkbits instead of sb->s_blocksize in erofs_bread(), since the inode operated on may be an anonymous inode in fscache mode. Currently the anonymous inode is allocated from an anonymous mount maintained in erofs, while in the near future we may allocate anonymous inodes from a generic API directly and thus have no access to the anonymous inode's i_sb. Thus we keep the block size in i_blkbits for anonymous inodes in fscache mode. Be noted that this patch only gets rid of the hardcoded blocksize, in preparation for actually setting the on-disk block size in the following patch. The hard limit of constraining the block size to PAGE_SIZE still exists until the next patch. Signed-off-by: Jingbo Xu <jefflexu@linux.alibaba.com> Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com> Reviewed-by: Yue Hu <huyue2@coolpad.com> Reviewed-by: Chao Yu <chao@kernel.org> Link: https://lore.kernel.org/r/20230313135309.75269-2-jefflexu@linux.alibaba.com [ Gao Xiang: fold a patch to fix incorrect truncated offsets. ] Link: https://lore.kernel.org/r/20230413035734.15457-1-zhujia.zj@bytedance.com Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
2023-03-13 21:53:08 +08:00
vi->z_logical_clusterbits = sb->s_blocksize_bits + (h->h_clusterbits & 7);
if (!erofs_sb_has_big_pcluster(EROFS_SB(sb)) &&
vi->z_advise & (Z_EROFS_ADVISE_BIG_PCLUSTER_1 |
Z_EROFS_ADVISE_BIG_PCLUSTER_2)) {
erofs_err(sb, "per-inode big pcluster without sb feature for nid %llu",
vi->nid);
err = -EFSCORRUPTED;
goto out_put_metabuf;
}
if (vi->datalayout == EROFS_INODE_COMPRESSED_COMPACT &&
!(vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_1) ^
!(vi->z_advise & Z_EROFS_ADVISE_BIG_PCLUSTER_2)) {
erofs_err(sb, "big pcluster head1/2 of compact indexes should be consistent for nid %llu",
vi->nid);
err = -EFSCORRUPTED;
goto out_put_metabuf;
}
if (vi->z_advise & Z_EROFS_ADVISE_INLINE_PCLUSTER) {
struct erofs_map_blocks map = {
.buf = __EROFS_BUF_INITIALIZER
};
vi->z_idata_size = le16_to_cpu(h->h_idata_size);
err = z_erofs_do_map_blocks(inode, &map,
EROFS_GET_BLOCKS_FINDTAIL);
erofs_put_metabuf(&map.buf);
if (!map.m_plen ||
erofs: avoid hardcoded blocksize for subpage block support As the first step of converting hardcoded blocksize to that specified in on-disk superblock, convert all call sites of hardcoded blocksize to sb->s_blocksize except for: 1) use sbi->blkszbits instead of sb->s_blocksize in erofs_superblock_csum_verify() since sb->s_blocksize has not been updated with the on-disk blocksize yet when the function is called. 2) use inode->i_blkbits instead of sb->s_blocksize in erofs_bread(), since the inode operated on may be an anonymous inode in fscache mode. Currently the anonymous inode is allocated from an anonymous mount maintained in erofs, while in the near future we may allocate anonymous inodes from a generic API directly and thus have no access to the anonymous inode's i_sb. Thus we keep the block size in i_blkbits for anonymous inodes in fscache mode. Be noted that this patch only gets rid of the hardcoded blocksize, in preparation for actually setting the on-disk block size in the following patch. The hard limit of constraining the block size to PAGE_SIZE still exists until the next patch. Signed-off-by: Jingbo Xu <jefflexu@linux.alibaba.com> Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com> Reviewed-by: Yue Hu <huyue2@coolpad.com> Reviewed-by: Chao Yu <chao@kernel.org> Link: https://lore.kernel.org/r/20230313135309.75269-2-jefflexu@linux.alibaba.com [ Gao Xiang: fold a patch to fix incorrect truncated offsets. ] Link: https://lore.kernel.org/r/20230413035734.15457-1-zhujia.zj@bytedance.com Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
2023-03-13 21:53:08 +08:00
erofs_blkoff(sb, map.m_pa) + map.m_plen > sb->s_blocksize) {
erofs_err(sb, "invalid tail-packing pclustersize %llu",
map.m_plen);
err = -EFSCORRUPTED;
}
if (err < 0)
goto out_put_metabuf;
}
if (vi->z_advise & Z_EROFS_ADVISE_FRAGMENT_PCLUSTER &&
!(h->h_clusterbits >> Z_EROFS_FRAGMENT_INODE_BIT)) {
struct erofs_map_blocks map = {
.buf = __EROFS_BUF_INITIALIZER
};
vi->z_fragmentoff = le32_to_cpu(h->h_fragmentoff);
err = z_erofs_do_map_blocks(inode, &map,
EROFS_GET_BLOCKS_FINDTAIL);
erofs_put_metabuf(&map.buf);
if (err < 0)
goto out_put_metabuf;
}
done:
/* paired with smp_mb() at the beginning of the function */
smp_mb();
set_bit(EROFS_I_Z_INITED_BIT, &vi->flags);
out_put_metabuf:
erofs_put_metabuf(&buf);
out_unlock:
clear_and_wake_up_bit(EROFS_I_BL_Z_BIT, &vi->flags);
return err;
}
int z_erofs_map_blocks_iter(struct inode *inode, struct erofs_map_blocks *map,
int flags)
{
struct erofs_inode *const vi = EROFS_I(inode);
int err = 0;
trace_erofs_map_blocks_enter(inode, map, flags);
if (map->m_la >= inode->i_size) { /* post-EOF unmapped extent */
map->m_llen = map->m_la + 1 - inode->i_size;
map->m_la = inode->i_size;
map->m_flags = 0;
} else {
err = z_erofs_fill_inode_lazy(inode);
if (!err) {
if ((vi->z_advise & Z_EROFS_ADVISE_FRAGMENT_PCLUSTER) &&
!vi->z_tailextent_headlcn) {
map->m_la = 0;
map->m_llen = inode->i_size;
map->m_flags = EROFS_MAP_MAPPED |
EROFS_MAP_FULL_MAPPED | EROFS_MAP_FRAGMENT;
} else {
err = z_erofs_do_map_blocks(inode, map, flags);
}
}
if (!err && (map->m_flags & EROFS_MAP_ENCODED) &&
unlikely(map->m_plen > Z_EROFS_PCLUSTER_MAX_SIZE ||
map->m_llen > Z_EROFS_PCLUSTER_MAX_DSIZE))
err = -EOPNOTSUPP;
if (err)
map->m_llen = 0;
}
trace_erofs_map_blocks_exit(inode, map, flags, err);
return err;
}
static int z_erofs_iomap_begin_report(struct inode *inode, loff_t offset,
loff_t length, unsigned int flags,
struct iomap *iomap, struct iomap *srcmap)
{
int ret;
struct erofs_map_blocks map = { .m_la = offset };
ret = z_erofs_map_blocks_iter(inode, &map, EROFS_GET_BLOCKS_FIEMAP);
erofs_put_metabuf(&map.buf);
if (ret < 0)
return ret;
iomap->bdev = inode->i_sb->s_bdev;
iomap->offset = map.m_la;
iomap->length = map.m_llen;
if (map.m_flags & EROFS_MAP_MAPPED) {
iomap->type = IOMAP_MAPPED;
iomap->addr = map.m_flags & EROFS_MAP_FRAGMENT ?
IOMAP_NULL_ADDR : map.m_pa;
} else {
iomap->type = IOMAP_HOLE;
iomap->addr = IOMAP_NULL_ADDR;
/*
* No strict rule on how to describe extents for post EOF, yet
* we need to do like below. Otherwise, iomap itself will get
* into an endless loop on post EOF.
*
* Calculate the effective offset by subtracting extent start
* (map.m_la) from the requested offset, and add it to length.
* (NB: offset >= map.m_la always)
*/
if (iomap->offset >= inode->i_size)
iomap->length = length + offset - map.m_la;
}
iomap->flags = 0;
return 0;
}
const struct iomap_ops z_erofs_iomap_report_ops = {
.iomap_begin = z_erofs_iomap_begin_report,
};