August 2021 - U-Boot

[PATCH 1/3] fs/erofs: new filesystem
by Huang Jianan 10 Feb '22

10 Feb '22

From: Huang Jianan <huangjianan(a)oppo.com> Add erofs filesystem support. The code is adapted from erofs-utils in order to reduce maintenance burden and keep with the latest feature. This patch mainly deals with uncompressed files. Signed-off-by: Huang Jianan <jnhuang95(a)gmail.com> --- fs/Kconfig | 1 + fs/Makefile | 1 + fs/erofs/Kconfig | 12 ++ fs/erofs/Makefile | 7 + fs/erofs/data.c | 124 ++++++++++++++ fs/erofs/erofs_fs.h | 384 ++++++++++++++++++++++++++++++++++++++++++++ fs/erofs/fs.c | 230 ++++++++++++++++++++++++++ fs/erofs/internal.h | 203 +++++++++++++++++++++++ fs/erofs/namei.c | 238 +++++++++++++++++++++++++++ fs/erofs/super.c | 65 ++++++++ fs/fs.c | 22 +++ include/erofs.h | 19 +++ include/fs.h | 1 + 13 files changed, 1307 insertions(+) create mode 100644 fs/erofs/Kconfig create mode 100644 fs/erofs/Makefile create mode 100644 fs/erofs/data.c create mode 100644 fs/erofs/erofs_fs.h create mode 100644 fs/erofs/fs.c create mode 100644 fs/erofs/internal.h create mode 100644 fs/erofs/namei.c create mode 100644 fs/erofs/super.c create mode 100644 include/erofs.h diff --git a/fs/Kconfig b/fs/Kconfig index 620af7f044..4d04c2c2ce 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -24,4 +24,5 @@ source "fs/yaffs2/Kconfig" source "fs/squashfs/Kconfig" +source "fs/erofs/Kconfig" endmenu diff --git a/fs/Makefile b/fs/Makefile index 937cbcf6e8..f05a21c9e6 100644 --- a/fs/Makefile +++ b/fs/Makefile @@ -25,5 +25,6 @@ obj-$(CONFIG_CMD_UBIFS) += ubifs/ obj-$(CONFIG_YAFFS2) += yaffs2/ obj-$(CONFIG_CMD_ZFS) += zfs/ obj-$(CONFIG_FS_SQUASHFS) += squashfs/ +obj-$(CONFIG_FS_EROFS) += erofs/ endif obj-y += fs_internal.o diff --git a/fs/erofs/Kconfig b/fs/erofs/Kconfig new file mode 100644 index 0000000000..f4b2d51a23 --- /dev/null +++ b/fs/erofs/Kconfig @@ -0,0 +1,12 @@ +config FS_EROFS + bool "Enable EROFS filesystem support" + help + This provides support for reading images from EROFS filesystem. + EROFS (Enhanced Read-Only File System) is a lightweight read-only + file system for scenarios which need high-performance read-only + requirements. + + It also provides fixed-sized output compression support, which + improves storage density, keeps relatively higher compression + ratios, which is more useful to achieve high performance for + embedded devices with limited memory. diff --git a/fs/erofs/Makefile b/fs/erofs/Makefile new file mode 100644 index 0000000000..7398ab7a36 --- /dev/null +++ b/fs/erofs/Makefile @@ -0,0 +1,7 @@ +# SPDX-License-Identifier: GPL-2.0+ +# + +obj-$(CONFIG_$(SPL_)FS_EROFS) = fs.o \ + super.o \ + namei.o \ + data.o diff --git a/fs/erofs/data.c b/fs/erofs/data.c new file mode 100644 index 0000000000..a0f613264a --- /dev/null +++ b/fs/erofs/data.c @@ -0,0 +1,124 @@ +// SPDX-License-Identifier: GPL-2.0+ +#include "internal.h" + +static int erofs_map_blocks_flatmode(struct erofs_inode *inode, + struct erofs_map_blocks *map, + int flags) +{ + int err = 0; + erofs_blk_t nblocks, lastblk; + u64 offset = map->m_la; + struct erofs_inode *vi = inode; + bool tailendpacking = (vi->datalayout == EROFS_INODE_FLAT_INLINE); + + nblocks = DIV_ROUND_UP(inode->i_size, EROFS_BLKSIZ); + lastblk = nblocks - tailendpacking; + + if (offset >= inode->i_size) { + /* leave out-of-bound access unmapped */ + map->m_flags = 0; + map->m_plen = 0; + goto out; + } + + /* there is no hole in flatmode */ + map->m_flags = EROFS_MAP_MAPPED; + + if (offset < blknr_to_addr(lastblk)) { + map->m_pa = blknr_to_addr(vi->u.i_blkaddr) + map->m_la; + map->m_plen = blknr_to_addr(lastblk) - offset; + } else if (tailendpacking) { + /* 2 - inode inline B: inode, [xattrs], inline last blk... */ + map->m_pa = iloc(vi->nid) + vi->inode_isize + + vi->xattr_isize + erofs_blkoff(map->m_la); + map->m_plen = inode->i_size - offset; + + /* inline data should be located in one meta block */ + if (erofs_blkoff(map->m_pa) + map->m_plen > PAGE_SIZE) { + erofs_err("inline data cross block boundary @ nid %" PRIu64, + vi->nid); + DBG_BUGON(1); + err = -EFSCORRUPTED; + goto err_out; + } + + map->m_flags |= EROFS_MAP_META; + } else { + erofs_err("internal error @ nid: %" PRIu64 " (size %llu), m_la 0x%" PRIx64, + vi->nid, (unsigned long long)inode->i_size, map->m_la); + DBG_BUGON(1); + err = -EIO; + goto err_out; + } + +out: + map->m_llen = map->m_plen; + +err_out: + return err; +} + +static int erofs_read_raw_data(struct erofs_inode *inode, char *buffer, + erofs_off_t size, erofs_off_t offset) +{ + struct erofs_map_blocks map = { + .index = UINT_MAX, + }; + int ret; + erofs_off_t ptr = offset; + + while (ptr < offset + size) { + char *const estart = buffer + ptr - offset; + erofs_off_t eend; + + map.m_la = ptr; + ret = erofs_map_blocks_flatmode(inode, &map, 0); + if (ret) + return ret; + + DBG_BUGON(map.m_plen != map.m_llen); + + /* trim extent */ + eend = min(offset + size, map.m_la + map.m_llen); + DBG_BUGON(ptr < map.m_la); + + if (!(map.m_flags & EROFS_MAP_MAPPED)) { + if (!map.m_llen) { + /* reached EOF */ + memset(estart, 0, offset + size - ptr); + ptr = offset + size; + continue; + } + memset(estart, 0, eend - ptr); + ptr = eend; + continue; + } + + if (ptr > map.m_la) { + map.m_pa += ptr - map.m_la; + map.m_la = ptr; + } + + ret = erofs_dev_read(estart, map.m_pa, eend - map.m_la); + if (ret < 0) + return -EIO; + ptr = eend; + } + return 0; +} + +int erofs_pread(struct erofs_inode *inode, char *buf, + erofs_off_t count, erofs_off_t offset) +{ + switch (inode->datalayout) { + case EROFS_INODE_FLAT_PLAIN: + case EROFS_INODE_FLAT_INLINE: + return erofs_read_raw_data(inode, buf, count, offset); + case EROFS_INODE_FLAT_COMPRESSION_LEGACY: + case EROFS_INODE_FLAT_COMPRESSION: + return -EOPNOTSUPP; + default: + break; + } + return -EINVAL; +} diff --git a/fs/erofs/erofs_fs.h b/fs/erofs/erofs_fs.h new file mode 100644 index 0000000000..070087fa60 --- /dev/null +++ b/fs/erofs/erofs_fs.h @@ -0,0 +1,384 @@ +/* SPDX-License-Identifier: GPL-2.0+ OR Apache-2.0 */ +/* + * EROFS (Enhanced ROM File System) on-disk format definition + * + * Copyright (C) 2017-2018 HUAWEI, Inc. + * http://www.huawei.com/ + */ +#ifndef __EROFS_FS_H +#define __EROFS_FS_H + +#include <asm/unaligned.h> +#include <fs.h> +#include <part.h> +#include <stdint.h> + +#define EROFS_SUPER_MAGIC_V1 0xE0F5E1E2 +#define EROFS_SUPER_OFFSET 1024 + +#define EROFS_FEATURE_COMPAT_SB_CHKSUM 0x00000001 + +/* + * Any bits that aren't in EROFS_ALL_FEATURE_INCOMPAT should + * be incompatible with this kernel version. + */ +#define EROFS_FEATURE_INCOMPAT_LZ4_0PADDING 0x00000001 +#define EROFS_FEATURE_INCOMPAT_COMPR_CFGS 0x00000002 +#define EROFS_FEATURE_INCOMPAT_BIG_PCLUSTER 0x00000002 +#define EROFS_ALL_FEATURE_INCOMPAT \ + (EROFS_FEATURE_INCOMPAT_LZ4_0PADDING | \ + EROFS_FEATURE_INCOMPAT_COMPR_CFGS | \ + EROFS_FEATURE_INCOMPAT_BIG_PCLUSTER) + +#define EROFS_SB_EXTSLOT_SIZE 16 + +/* erofs on-disk super block (currently 128 bytes) */ +struct erofs_super_block { + __le32 magic; /* file system magic number */ + __le32 checksum; /* crc32c(super_block) */ + __le32 feature_compat; + __u8 blkszbits; /* support block_size == PAGE_SIZE only */ + __u8 sb_extslots; /* superblock size = 128 + sb_extslots * 16 */ + + __le16 root_nid; /* nid of root directory */ + __le64 inos; /* total valid ino # (== f_files - f_favail) */ + + __le64 build_time; /* inode v1 time derivation */ + __le32 build_time_nsec; /* inode v1 time derivation in nano scale */ + __le32 blocks; /* used for statfs */ + __le32 meta_blkaddr; /* start block address of metadata area */ + __le32 xattr_blkaddr; /* start block address of shared xattr area */ + __u8 uuid[16]; /* 128-bit uuid for volume */ + __u8 volume_name[16]; /* volume name */ + __le32 feature_incompat; + union { + /* bitmap for available compression algorithms */ + __le16 available_compr_algs; + /* customized sliding window size instead of 64k by default */ + __le16 lz4_max_distance; + } __packed u1; + __u8 reserved2[42]; +}; + +/* + * erofs inode datalayout (i_format in on-disk inode): + * 0 - inode plain without inline data A: + * inode, [xattrs], ... | ... | no-holed data + * 1 - inode VLE compression B (legacy): + * inode, [xattrs], extents ... | ... + * 2 - inode plain with inline data C: + * inode, [xattrs], last_inline_data, ... | ... | no-holed data + * 3 - inode compression D: + * inode, [xattrs], map_header, extents ... | ... + * 4 - inode chunk-based E: + * inode, [xattrs], chunk indexes ... | ... + * 5~7 - reserved + */ +enum { + EROFS_INODE_FLAT_PLAIN = 0, + EROFS_INODE_FLAT_COMPRESSION_LEGACY = 1, + EROFS_INODE_FLAT_INLINE = 2, + EROFS_INODE_FLAT_COMPRESSION = 3, + EROFS_INODE_CHUNK_BASED = 4, + EROFS_INODE_DATALAYOUT_MAX +}; + +static inline bool erofs_inode_is_data_compressed(unsigned int datamode) +{ + return datamode == EROFS_INODE_FLAT_COMPRESSION || + datamode == EROFS_INODE_FLAT_COMPRESSION_LEGACY; +} + +/* bit definitions of inode i_advise */ +#define EROFS_I_VERSION_BITS 1 +#define EROFS_I_DATALAYOUT_BITS 3 + +#define EROFS_I_VERSION_BIT 0 +#define EROFS_I_DATALAYOUT_BIT 1 + +#define EROFS_I_ALL \ + ((1 << (EROFS_I_DATALAYOUT_BIT + EROFS_I_DATALAYOUT_BITS)) - 1) + +/* indicate chunk blkbits, thus 'chunksize = blocksize << chunk blkbits' */ +#define EROFS_CHUNK_FORMAT_BLKBITS_MASK 0x001F +/* with chunk indexes or just a 4-byte blkaddr array */ +#define EROFS_CHUNK_FORMAT_INDEXES 0x0020 + +#define EROFS_CHUNK_FORMAT_ALL \ + (EROFS_CHUNK_FORMAT_BLKBITS_MASK | EROFS_CHUNK_FORMAT_INDEXES) + +struct erofs_inode_chunk_info { + __le16 format; /* chunk blkbits, etc. */ + __le16 reserved; +}; + +/* 32-byte reduced form of an ondisk inode */ +struct erofs_inode_compact { + __le16 i_format; /* inode format hints */ + +/* 1 header + n-1 * 4 bytes inline xattr to keep continuity */ + __le16 i_xattr_icount; + __le16 i_mode; + __le16 i_nlink; + __le32 i_size; + __le32 i_reserved; + union { + /* file total compressed blocks for data mapping 1 */ + __le32 compressed_blocks; + __le32 raw_blkaddr; + + /* for device files, used to indicate old/new device # */ + __le32 rdev; + + /* for chunk-based files, it contains the summary info */ + struct erofs_inode_chunk_info c; + } i_u; + __le32 i_ino; /* only used for 32-bit stat compatibility */ + __le16 i_uid; + __le16 i_gid; + __le32 i_reserved2; +}; + +/* 32 bytes on-disk inode */ +#define EROFS_INODE_LAYOUT_COMPACT 0 +/* 64 bytes on-disk inode */ +#define EROFS_INODE_LAYOUT_EXTENDED 1 + +/* 64-byte complete form of an ondisk inode */ +struct erofs_inode_extended { + __le16 i_format; /* inode format hints */ + +/* 1 header + n-1 * 4 bytes inline xattr to keep continuity */ + __le16 i_xattr_icount; + __le16 i_mode; + __le16 i_reserved; + __le64 i_size; + union { + /* file total compressed blocks for data mapping 1 */ + __le32 compressed_blocks; + __le32 raw_blkaddr; + + /* for device files, used to indicate old/new device # */ + __le32 rdev; + + /* for chunk-based files, it contains the summary info */ + struct erofs_inode_chunk_info c; + } i_u; + + /* only used for 32-bit stat compatibility */ + __le32 i_ino; + + __le32 i_uid; + __le32 i_gid; + __le64 i_ctime; + __le32 i_ctime_nsec; + __le32 i_nlink; + __u8 i_reserved2[16]; +}; + +#define EROFS_MAX_SHARED_XATTRS (128) +/* h_shared_count between 129 ... 255 are special # */ +#define EROFS_SHARED_XATTR_EXTENT (255) + +/* + * inline xattrs (n == i_xattr_icount): + * erofs_xattr_ibody_header(1) + (n - 1) * 4 bytes + * 12 bytes / \ + * / \ + * /-----------------------\ + * | erofs_xattr_entries+ | + * +-----------------------+ + * inline xattrs must starts in erofs_xattr_ibody_header, + * for read-only fs, no need to introduce h_refcount + */ +struct erofs_xattr_ibody_header { + __le32 h_reserved; + __u8 h_shared_count; + __u8 h_reserved2[7]; + __le32 h_shared_xattrs[0]; /* shared xattr id array */ +}; + +/* Name indexes */ +#define EROFS_XATTR_INDEX_USER 1 +#define EROFS_XATTR_INDEX_POSIX_ACL_ACCESS 2 +#define EROFS_XATTR_INDEX_POSIX_ACL_DEFAULT 3 +#define EROFS_XATTR_INDEX_TRUSTED 4 +#define EROFS_XATTR_INDEX_LUSTRE 5 +#define EROFS_XATTR_INDEX_SECURITY 6 + +/* xattr entry (for both inline & shared xattrs) */ +struct erofs_xattr_entry { + __u8 e_name_len; /* length of name */ + __u8 e_name_index; /* attribute name index */ + __le16 e_value_size; /* size of attribute value */ + /* followed by e_name and e_value */ + char e_name[0]; /* attribute name */ +}; + +static inline unsigned int erofs_xattr_ibody_size(__le16 i_xattr_icount) +{ + if (!i_xattr_icount) + return 0; + + return sizeof(struct erofs_xattr_ibody_header) + + sizeof(__u32) * (le16_to_cpu(i_xattr_icount) - 1); +} + +#define EROFS_XATTR_ALIGN(size) round_up(size, sizeof(struct erofs_xattr_entry)) + +static inline unsigned int erofs_xattr_entry_size(struct erofs_xattr_entry *e) +{ + return EROFS_XATTR_ALIGN(sizeof(struct erofs_xattr_entry) + + e->e_name_len + le16_to_cpu(e->e_value_size)); +} + +/* maximum supported size of a physical compression cluster */ +#define Z_EROFS_PCLUSTER_MAX_SIZE (1024 * 1024) + +/* available compression algorithm types (for h_algorithmtype) */ +enum { + Z_EROFS_COMPRESSION_LZ4 = 0, + Z_EROFS_COMPRESSION_MAX +}; +#define Z_EROFS_ALL_COMPR_ALGS (1 << (Z_EROFS_COMPRESSION_MAX - 1)) + +/* 14 bytes (+ length field = 16 bytes) */ +struct z_erofs_lz4_cfgs { + __le16 max_distance; + __le16 max_pclusterblks; + u8 reserved[10]; +} __packed; + +/* + * bit 0 : COMPACTED_2B indexes (0 - off; 1 - on) + * e.g. for 4k logical cluster size, 4B if compacted 2B is off; + * (4B) + 2B + (4B) if compacted 2B is on. + * bit 1 : HEAD1 big pcluster (0 - off; 1 - on) + * bit 2 : HEAD2 big pcluster (0 - off; 1 - on) + */ +#define Z_EROFS_ADVISE_COMPACTED_2B 0x0001 +#define Z_EROFS_ADVISE_BIG_PCLUSTER_1 0x0002 +#define Z_EROFS_ADVISE_BIG_PCLUSTER_2 0x0004 + +struct z_erofs_map_header { + __le32 h_reserved1; + __le16 h_advise; + /* + * bit 0-3 : algorithm type of head 1 (logical cluster type 01); + * bit 4-7 : algorithm type of head 2 (logical cluster type 11). + */ + __u8 h_algorithmtype; + /* + * bit 0-2 : logical cluster bits - 12, e.g. 0 for 4096; + * bit 3-7 : reserved. + */ + __u8 h_clusterbits; +}; + +#define Z_EROFS_VLE_LEGACY_HEADER_PADDING 8 + +/* + * Fixed-sized output compression ondisk Logical Extent cluster type: + * 0 - literal (uncompressed) cluster + * 1 - compressed cluster (for the head logical cluster) + * 2 - compressed cluster (for the other logical clusters) + * + * In detail, + * 0 - literal (uncompressed) cluster, + * di_advise = 0 + * di_clusterofs = the literal data offset of the cluster + * di_blkaddr = the blkaddr of the literal cluster + * + * 1 - compressed cluster (for the head logical cluster) + * di_advise = 1 + * di_clusterofs = the decompressed data offset of the cluster + * di_blkaddr = the blkaddr of the compressed cluster + * + * 2 - compressed cluster (for the other logical clusters) + * di_advise = 2 + * di_clusterofs = + * the decompressed data offset in its own head cluster + * di_u.delta[0] = distance to its corresponding head cluster + * di_u.delta[1] = distance to its corresponding tail cluster + * (di_advise could be 0, 1 or 2) + */ +enum { + Z_EROFS_VLE_CLUSTER_TYPE_PLAIN = 0, + Z_EROFS_VLE_CLUSTER_TYPE_HEAD = 1, + Z_EROFS_VLE_CLUSTER_TYPE_NONHEAD = 2, + Z_EROFS_VLE_CLUSTER_TYPE_RESERVED = 3, + Z_EROFS_VLE_CLUSTER_TYPE_MAX +}; + +#define Z_EROFS_VLE_DI_CLUSTER_TYPE_BITS 2 +#define Z_EROFS_VLE_DI_CLUSTER_TYPE_BIT 0 + +/* + * D0_CBLKCNT will be marked _only_ at the 1st non-head lcluster to store the + * compressed block count of a compressed extent (in logical clusters, aka. + * block count of a pcluster). + */ +#define Z_EROFS_VLE_DI_D0_CBLKCNT (1 << 11) + +struct z_erofs_vle_decompressed_index { + __le16 di_advise; + /* where to decompress in the head cluster */ + __le16 di_clusterofs; + + union { + /* for the head cluster */ + __le32 blkaddr; + /* + * for the rest clusters + * eg. for 4k page-sized cluster, maximum 4K*64k = 256M) + * [0] - pointing to the head cluster + * [1] - pointing to the tail cluster + */ + __le16 delta[2]; + } di_u; +}; + +#define Z_EROFS_VLE_LEGACY_INDEX_ALIGN(size) \ + (round_up(size, sizeof(struct z_erofs_vle_decompressed_index)) + \ + sizeof(struct z_erofs_map_header) + Z_EROFS_VLE_LEGACY_HEADER_PADDING) + +/* dirent sorts in alphabet order, thus we can do binary search */ +struct erofs_dirent { + __le64 nid; /* node number */ + __le16 nameoff; /* start offset of file name */ + __u8 file_type; /* file type */ + __u8 reserved; /* reserved */ +} __packed; + +/* file types used in inode_info->flags */ +enum { + EROFS_FT_UNKNOWN, + EROFS_FT_REG_FILE, + EROFS_FT_DIR, + EROFS_FT_CHRDEV, + EROFS_FT_BLKDEV, + EROFS_FT_FIFO, + EROFS_FT_SOCK, + EROFS_FT_SYMLINK, + EROFS_FT_MAX +}; + +#define EROFS_NAME_LEN 255 + +/* check the EROFS on-disk layout strictly at compile time */ +static inline void erofs_check_ondisk_layout_definitions(void) +{ + BUILD_BUG_ON(sizeof(struct erofs_super_block) != 128); + BUILD_BUG_ON(sizeof(struct erofs_inode_compact) != 32); + BUILD_BUG_ON(sizeof(struct erofs_inode_extended) != 64); + BUILD_BUG_ON(sizeof(struct erofs_xattr_ibody_header) != 12); + BUILD_BUG_ON(sizeof(struct erofs_xattr_entry) != 4); + BUILD_BUG_ON(sizeof(struct z_erofs_map_header) != 8); + BUILD_BUG_ON(sizeof(struct z_erofs_vle_decompressed_index) != 8); + BUILD_BUG_ON(sizeof(struct erofs_dirent) != 12); + + BUILD_BUG_ON(BIT(Z_EROFS_VLE_DI_CLUSTER_TYPE_BITS) < + Z_EROFS_VLE_CLUSTER_TYPE_MAX - 1); +} + +#endif diff --git a/fs/erofs/fs.c b/fs/erofs/fs.c new file mode 100644 index 0000000000..2c8d20b50d --- /dev/null +++ b/fs/erofs/fs.c @@ -0,0 +1,230 @@ +#include "internal.h" +#include <fs_internal.h> + +struct erofs_sb_info erofs_sbi; + +static struct erofs_ctxt { + struct disk_partition cur_part_info; + struct blk_desc *cur_dev; +} ctxt; + +int erofs_dev_read(void *buf, u64 offset, size_t len) +{ + lbaint_t sect = offset >> ctxt.cur_dev->log2blksz; + int off = offset & (ctxt.cur_dev->blksz - 1); + + if (!ctxt.cur_dev) + return -EIO; + + if (fs_devread(ctxt.cur_dev, &ctxt.cur_part_info, sect, + off, len, buf)) + return 0; + return -EIO; +} + +int erofs_blk_read(void *buf, erofs_blk_t start, u32 nblocks) +{ + return erofs_dev_read(buf, blknr_to_addr(start), + blknr_to_addr(nblocks)); +} + +int erofs_probe(struct blk_desc *fs_dev_desc, + struct disk_partition *fs_partition) +{ + int ret; + + ctxt.cur_dev = fs_dev_desc; + ctxt.cur_part_info = *fs_partition; + + ret = erofs_read_superblock(); + if (ret) + goto error; + + return 0; +error: + ctxt.cur_dev = NULL; + return ret; +} + +struct erofs_dir_stream { + struct fs_dir_stream fs_dirs; + struct fs_dirent dirent; + + struct erofs_inode inode; + char dblk[EROFS_BLKSIZ]; + unsigned int maxsize, de_end; + erofs_off_t pos; +}; + +int erofs_opendir(const char *filename, struct fs_dir_stream **dirsp) +{ + struct erofs_dir_stream *dirs; + int err; + + dirs = calloc(1, sizeof(*dirs)); + if (!dirs) + return -ENOMEM; + + err = erofs_ilookup(filename, &dirs->inode); + if (err) + goto err_out; + + if (!S_ISDIR(dirs->inode.i_mode)) { + err = -ENOTDIR; + goto err_out; + } + *dirsp = (struct fs_dir_stream *)dirs; + return 0; +err_out: + free(dirs); + return err; +} + +int erofs_readdir(struct fs_dir_stream *fs_dirs, struct fs_dirent **dentp) +{ + struct erofs_dir_stream *dirs = (struct erofs_dir_stream *)fs_dirs; + struct fs_dirent *dent = &dirs->dirent; + erofs_off_t pos = dirs->pos; + unsigned int nameoff, de_namelen; + struct erofs_dirent *de; + char *de_name; + int err; + + if (pos >= dirs->inode.i_size) + return 1; + + if (!dirs->maxsize) { + dirs->maxsize = min_t(unsigned int, EROFS_BLKSIZ, + dirs->inode.i_size - pos); + + err = erofs_pread(&dirs->inode, dirs->dblk, + dirs->maxsize, pos); + if (err) + return err; + + de = (struct erofs_dirent *)dirs->dblk; + dirs->de_end = le16_to_cpu(de->nameoff); + if (dirs->de_end < sizeof(struct erofs_dirent) || + dirs->de_end >= EROFS_BLKSIZ) { + erofs_err("invalid de[0].nameoff %u @ nid %llu", + nameoff, de->nid | 0ULL); + return -EFSCORRUPTED; + } + } + + de = (struct erofs_dirent *)(dirs->dblk + erofs_blkoff(pos)); + nameoff = le16_to_cpu(de->nameoff); + de_name = (char *)dirs->dblk + nameoff; + + /* the last dirent in the block? */ + if (de + 1 >= (struct erofs_dirent *)(dirs->dblk + dirs->de_end)) + de_namelen = strnlen(de_name, dirs->maxsize - nameoff); + else + de_namelen = le16_to_cpu(de[1].nameoff) - nameoff; + + /* a corrupted entry is found */ + if (nameoff + de_namelen > dirs->maxsize || + de_namelen > EROFS_NAME_LEN) { + erofs_err("bogus dirent @ nid %llu", de->nid | 0ULL); + DBG_BUGON(1); + return -EFSCORRUPTED; + } + + memcpy(dent->name, de_name, de_namelen); + dent->name[de_namelen] = '\0'; + + if (de->file_type == EROFS_FT_DIR) { + dent->type = FS_DT_DIR; + } else if (de->file_type == EROFS_FT_SYMLINK) { + dent->type = FS_DT_LNK; + } else { + struct erofs_inode vi; + + dent->type = FS_DT_REG; + vi.nid = de->nid; + + err = erofs_read_inode_from_disk(&vi); + if (err) + return err; + dent->size = vi.i_size; + } + *dentp = dent; + + pos += sizeof(*de); + if (erofs_blkoff(pos) >= dirs->de_end) { + pos = blknr_to_addr(erofs_blknr(pos) + 1); + dirs->maxsize = 0; + } + dirs->pos = pos; + return 0; +} + +void erofs_closedir(struct fs_dir_stream *fs_dirs) +{ + free(fs_dirs); +} + +int erofs_exists(const char *filename) +{ + struct erofs_inode vi; + int err; + + err = erofs_ilookup(filename, &vi); + return err == 0; +} + +int erofs_size(const char *filename, loff_t *size) +{ + struct erofs_inode vi; + int err; + + err = erofs_ilookup(filename, &vi); + if (err) + return err; + *size = vi.i_size; + return 0; +} + +int erofs_read(const char *filename, void *buf, loff_t offset, loff_t len, + loff_t *actread) +{ + struct erofs_inode vi; + int err; + + err = erofs_ilookup(filename, &vi); + if (err) + return err; + + if (!len) + len = vi.i_size; + + err = erofs_pread(&vi, buf, len, offset); + if (err) { + *actread = 0; + return err; + } + + if (offset >= vi.i_size) + *actread = 0; + else if (offset + len > vi.i_size) + *actread = vi.i_size - offset; + else + *actread = len; + return 0; +} + +void erofs_close(void) +{ + ctxt.cur_dev = NULL; +} + +int erofs_uuid(char *uuid_str) +{ +#ifdef CONFIG_LIB_UUID + if (ctxt.cur_dev) + uuid_bin_to_str(erofs_sbi.uuid, uuid_str, + UUID_STR_FORMAT_STD); + return 0; +#endif + return -ENOSYS; +} diff --git a/fs/erofs/internal.h b/fs/erofs/internal.h new file mode 100644 index 0000000000..02dfd7ead4 --- /dev/null +++ b/fs/erofs/internal.h @@ -0,0 +1,203 @@ +/* SPDX-License-Identifier: GPL-2.0+ */ +#ifndef __EROFS_INTERNAL_H +#define __EROFS_INTERNAL_H + +#define __packed __attribute__((__packed__)) + +#include <linux/stat.h> +#include <linux/bug.h> +#include <linux/err.h> +#include <linux/printk.h> +#include <inttypes.h> +#include "erofs_fs.h" + +#define erofs_err(fmt, ...) \ + pr_err(fmt "\n", ##__VA_ARGS__) + +#define erofs_info(fmt, ...) \ + pr_info(fmt "\n", ##__VA_ARGS__) + +#define erofs_dbg(fmt, ...) \ + pr_debug(fmt "\n", ##__VA_ARGS__) + +#define DBG_BUGON(condition) BUG_ON(condition) + +/* no obvious reason to support explicit PAGE_SIZE != 4096 for now */ +#if PAGE_SIZE != 4096 +#error incompatible PAGE_SIZE is already defined +#endif + +#define PAGE_MASK (~(PAGE_SIZE-1)) + +#define LOG_BLOCK_SIZE (12) +#define EROFS_BLKSIZ (1U << LOG_BLOCK_SIZE) + +#define EROFS_ISLOTBITS 5 +#define EROFS_SLOTSIZE (1U << EROFS_ISLOTBITS) + +typedef u64 erofs_off_t; +typedef u64 erofs_nid_t; +/* data type for filesystem-wide blocks number */ +typedef u32 erofs_blk_t; + +#define NULL_ADDR ((unsigned int)-1) +#define NULL_ADDR_UL ((unsigned long)-1) + +#define erofs_blknr(addr) ((addr) / EROFS_BLKSIZ) +#define erofs_blkoff(addr) ((addr) % EROFS_BLKSIZ) +#define blknr_to_addr(nr) ((erofs_off_t)(nr) * EROFS_BLKSIZ) + +#define BLK_ROUND_UP(addr) DIV_ROUND_UP(addr, EROFS_BLKSIZ) + +struct erofs_sb_info { + u64 blocks; + + erofs_blk_t meta_blkaddr; + erofs_blk_t xattr_blkaddr; + + u32 feature_compat; + u32 feature_incompat; + u64 build_time; + u32 build_time_nsec; + + unsigned char islotbits; + + /* what we really care is nid, rather than ino.. */ + erofs_nid_t root_nid; + /* used for statfs, f_files - f_favail */ + u64 inos; + + u8 uuid[16]; + + u16 available_compr_algs; + u16 lz4_max_distance; +}; + +/* global sbi */ +extern struct erofs_sb_info erofs_sbi; + +static inline erofs_off_t iloc(erofs_nid_t nid) +{ + return blknr_to_addr(erofs_sbi.meta_blkaddr) + + (nid << erofs_sbi.islotbits); +} + +#define EROFS_FEATURE_FUNCS(name, compat, feature) \ +static inline bool erofs_sb_has_##name(void) \ +{ \ + return erofs_sbi.feature_##compat & EROFS_FEATURE_##feature; \ +} + +EROFS_FEATURE_FUNCS(lz4_0padding, incompat, INCOMPAT_LZ4_0PADDING) +EROFS_FEATURE_FUNCS(compr_cfgs, incompat, INCOMPAT_COMPR_CFGS) +EROFS_FEATURE_FUNCS(big_pcluster, incompat, INCOMPAT_BIG_PCLUSTER) +EROFS_FEATURE_FUNCS(sb_chksum, compat, COMPAT_SB_CHKSUM) + +#define EROFS_I_EA_INITED (1 << 0) +#define EROFS_I_Z_INITED (1 << 1) + +struct erofs_inode { + unsigned int flags; + umode_t i_mode; + erofs_off_t i_size; + + u64 i_ino[2]; + u32 i_uid; + u32 i_gid; + u64 i_ctime; + u32 i_ctime_nsec; + u32 i_nlink; + + union { + u32 i_blkaddr; + u32 i_blocks; + u32 i_rdev; + } u; + + unsigned char datalayout; + unsigned char inode_isize; + unsigned int xattr_isize; + unsigned int extent_isize; + + erofs_nid_t nid; + + struct { + uint16_t z_advise; + uint8_t z_algorithmtype[2]; + uint8_t z_logical_clusterbits; + }; +}; + +static inline bool is_inode_layout_compression(struct erofs_inode *inode) +{ + return erofs_inode_is_data_compressed(inode->datalayout); +} + +static inline unsigned int erofs_bitrange(unsigned int value, unsigned int bit, + unsigned int bits) +{ + return (value >> bit) & ((1 << bits) - 1); +} + +static inline unsigned int erofs_inode_version(unsigned int value) +{ + return erofs_bitrange(value, EROFS_I_VERSION_BIT, + EROFS_I_VERSION_BITS); +} + +static inline unsigned int erofs_inode_datalayout(unsigned int value) +{ + return erofs_bitrange(value, EROFS_I_DATALAYOUT_BIT, + EROFS_I_DATALAYOUT_BITS); +} + +#define IS_ROOT(x) ((x) == (x)->i_parent) + +static inline bool is_dot_dotdot(const char *name) +{ + if (name[0] != '.') + return false; + + return name[1] == '\0' || (name[1] == '.' && name[2] == '\0'); +} + +enum { + BH_Meta, + BH_Mapped, + BH_Zipped, + BH_FullMapped, +}; + +/* Has a disk mapping */ +#define EROFS_MAP_MAPPED (1 << BH_Mapped) +/* Located in metadata (could be copied from bd_inode) */ +#define EROFS_MAP_META (1 << BH_Meta) +/* The extent has been compressed */ +#define EROFS_MAP_ZIPPED (1 << BH_Zipped) +/* The length of extent is full */ +#define EROFS_MAP_FULL_MAPPED (1 << BH_FullMapped) + +struct erofs_map_blocks { + char mpage[EROFS_BLKSIZ]; + + erofs_off_t m_pa, m_la; + u64 m_plen, m_llen; + + unsigned int m_flags; + erofs_blk_t index; +}; + +int erofs_blk_read(void *buf, erofs_blk_t start, u32 nblocks); +int erofs_dev_read(void *buf, u64 offset, size_t len); + +int erofs_read_superblock(void); +int erofs_read_inode_from_disk(struct erofs_inode *vi); +int erofs_ilookup(const char *path, struct erofs_inode *vi); +int erofs_pread(struct erofs_inode *inode, char *buf, + erofs_off_t count, erofs_off_t offset); +int z_erofs_fill_inode(struct erofs_inode *vi); +int z_erofs_map_blocks_iter(struct erofs_inode *vi, + struct erofs_map_blocks *map); + +#define EFSCORRUPTED EUCLEAN /* Filesystem is corrupted */ +#endif diff --git a/fs/erofs/namei.c b/fs/erofs/namei.c new file mode 100644 index 0000000000..4c5d614483 --- /dev/null +++ b/fs/erofs/namei.c @@ -0,0 +1,238 @@ +// SPDX-License-Identifier: GPL-2.0+ +#include "internal.h" + +int erofs_read_inode_from_disk(struct erofs_inode *vi) +{ + int ret, ifmt; + char buf[sizeof(struct erofs_inode_extended)]; + struct erofs_inode_compact *dic; + struct erofs_inode_extended *die; + const erofs_off_t inode_loc = iloc(vi->nid); + + ret = erofs_dev_read(buf, inode_loc, sizeof(*dic)); + if (ret < 0) + return -EIO; + + dic = (struct erofs_inode_compact *)buf; + ifmt = le16_to_cpu(dic->i_format); + + vi->datalayout = erofs_inode_datalayout(ifmt); + if (vi->datalayout >= EROFS_INODE_DATALAYOUT_MAX) { + erofs_err("unsupported datalayout %u of nid %llu", + vi->datalayout, vi->nid | 0ULL); + return -EOPNOTSUPP; + } + switch (erofs_inode_version(ifmt)) { + case EROFS_INODE_LAYOUT_EXTENDED: + vi->inode_isize = sizeof(struct erofs_inode_extended); + + ret = erofs_dev_read(buf + sizeof(*dic), inode_loc + sizeof(*dic), + sizeof(*die) - sizeof(*dic)); + if (ret < 0) + return -EIO; + + die = (struct erofs_inode_extended *)buf; + vi->xattr_isize = erofs_xattr_ibody_size(die->i_xattr_icount); + vi->i_mode = le16_to_cpu(die->i_mode); + + switch (vi->i_mode & S_IFMT) { + case S_IFREG: + case S_IFDIR: + case S_IFLNK: + vi->u.i_blkaddr = le32_to_cpu(die->i_u.raw_blkaddr); + break; + case S_IFCHR: + case S_IFBLK: + vi->u.i_rdev = 0; + break; + case S_IFIFO: + case S_IFSOCK: + vi->u.i_rdev = 0; + break; + default: + goto bogusimode; + } + + vi->i_uid = le32_to_cpu(die->i_uid); + vi->i_gid = le32_to_cpu(die->i_gid); + vi->i_nlink = le32_to_cpu(die->i_nlink); + + vi->i_ctime = le64_to_cpu(die->i_ctime); + vi->i_ctime_nsec = le64_to_cpu(die->i_ctime_nsec); + vi->i_size = le64_to_cpu(die->i_size); + break; + case EROFS_INODE_LAYOUT_COMPACT: + vi->inode_isize = sizeof(struct erofs_inode_compact); + vi->xattr_isize = erofs_xattr_ibody_size(dic->i_xattr_icount); + vi->i_mode = le16_to_cpu(dic->i_mode); + + switch (vi->i_mode & S_IFMT) { + case S_IFREG: + case S_IFDIR: + case S_IFLNK: + vi->u.i_blkaddr = le32_to_cpu(dic->i_u.raw_blkaddr); + break; + case S_IFCHR: + case S_IFBLK: + vi->u.i_rdev = 0; + break; + case S_IFIFO: + case S_IFSOCK: + vi->u.i_rdev = 0; + break; + default: + goto bogusimode; + } + + vi->i_uid = le16_to_cpu(dic->i_uid); + vi->i_gid = le16_to_cpu(dic->i_gid); + vi->i_nlink = le16_to_cpu(dic->i_nlink); + + vi->i_ctime = erofs_sbi.build_time; + vi->i_ctime_nsec = erofs_sbi.build_time_nsec; + + vi->i_size = le32_to_cpu(dic->i_size); + break; + default: + erofs_err("unsupported on-disk inode version %u of nid %llu", + erofs_inode_version(ifmt), vi->nid | 0ULL); + return -EOPNOTSUPP; + } + + vi->flags = 0; + if (erofs_inode_is_data_compressed(vi->datalayout)) + return -EOPNOTSUPP; + return 0; +bogusimode: + erofs_err("bogus i_mode (%o) @ nid %llu", vi->i_mode, vi->nid | 0ULL); + return -EFSCORRUPTED; +} + +struct erofs_dirent *find_target_dirent(erofs_nid_t pnid, + void *dentry_blk, + const char *name, unsigned int len, + unsigned int nameoff, + unsigned int maxsize) +{ + struct erofs_dirent *de = dentry_blk; + const struct erofs_dirent *end = dentry_blk + nameoff; + + while (de < end) { + const char *de_name; + unsigned int de_namelen; + + nameoff = le16_to_cpu(de->nameoff); + de_name = (char *)dentry_blk + nameoff; + + /* the last dirent in the block? */ + if (de + 1 >= end) + de_namelen = strnlen(de_name, maxsize - nameoff); + else + de_namelen = le16_to_cpu(de[1].nameoff) - nameoff; + + /* a corrupted entry is found */ + if (nameoff + de_namelen > maxsize || + de_namelen > EROFS_NAME_LEN) { + erofs_err("bogus dirent @ nid %llu", pnid | 0ULL); + DBG_BUGON(1); + return ERR_PTR(-EFSCORRUPTED); + } + + if (len == de_namelen && !memcmp(de_name, name, de_namelen)) + return de; + ++de; + } + return NULL; +} + +struct nameidata { + erofs_nid_t nid; + unsigned int ftype; +}; + +int erofs_namei(struct nameidata *nd, + const char *name, unsigned int len) +{ + erofs_nid_t nid = nd->nid; + int ret; + char buf[EROFS_BLKSIZ]; + struct erofs_inode vi = { .nid = nid }; + erofs_off_t offset; + + ret = erofs_read_inode_from_disk(&vi); + if (ret) + return ret; + + offset = 0; + while (offset < vi.i_size) { + erofs_off_t maxsize = min_t(erofs_off_t, + vi.i_size - offset, EROFS_BLKSIZ); + struct erofs_dirent *de = (void *)buf; + unsigned int nameoff; + + ret = erofs_pread(&vi, buf, maxsize, offset); + if (ret) + return ret; + + nameoff = le16_to_cpu(de->nameoff); + if (nameoff < sizeof(struct erofs_dirent) || + nameoff >= PAGE_SIZE) { + erofs_err("invalid de[0].nameoff %u @ nid %llu", + nameoff, nid | 0ULL); + return -EFSCORRUPTED; + } + + de = find_target_dirent(nid, buf, name, len, + nameoff, maxsize); + if (IS_ERR(de)) + return PTR_ERR(de); + + if (de) { + nd->nid = le64_to_cpu(de->nid); + return 0; + } + offset += maxsize; + } + return -ENOENT; +} + +static int link_path_walk(const char *name, struct nameidata *nd) +{ + nd->nid = erofs_sbi.root_nid; + + while (*name == '/') + name++; + + /* At this point we know we have a real path component. */ + while (*name != '\0') { + const char *p = name; + int ret; + + do { + ++p; + } while (*p != '\0' && *p != '/'); + + DBG_BUGON(p <= name); + ret = erofs_namei(nd, name, p - name); + if (ret) + return ret; + + name = p; + /* Skip until no more slashes. */ + for (name = p; *name == '/'; ++name); + } + return 0; +} + +int erofs_ilookup(const char *path, struct erofs_inode *vi) +{ + int ret; + struct nameidata nd; + + ret = link_path_walk(path, &nd); + if (ret) + return ret; + + vi->nid = nd.nid; + return erofs_read_inode_from_disk(vi); +} diff --git a/fs/erofs/super.c b/fs/erofs/super.c new file mode 100644 index 0000000000..c58a826b4b --- /dev/null +++ b/fs/erofs/super.c @@ -0,0 +1,65 @@ +// SPDX-License-Identifier: GPL-2.0+ +#include "internal.h" + +static bool check_layout_compatibility(struct erofs_sb_info *sbi, + struct erofs_super_block *dsb) +{ + const unsigned int feature = le32_to_cpu(dsb->feature_incompat); + + sbi->feature_incompat = feature; + + /* check if current kernel meets all mandatory requirements */ + if (feature & (~EROFS_ALL_FEATURE_INCOMPAT)) { + erofs_err("unidentified incompatible feature %x, please upgrade kernel version", + feature & ~EROFS_ALL_FEATURE_INCOMPAT); + return false; + } + return true; +} + +int erofs_read_superblock(void) +{ + char data[EROFS_BLKSIZ]; + struct erofs_super_block *dsb; + unsigned int blkszbits; + int ret; + + ret = erofs_blk_read(data, 0, 1); + if (ret < 0) { + erofs_err("cannot read erofs superblock: %d", ret); + return -EIO; + } + dsb = (struct erofs_super_block *)(data + EROFS_SUPER_OFFSET); + + ret = -EINVAL; + if (le32_to_cpu(dsb->magic) != EROFS_SUPER_MAGIC_V1) { + erofs_err("cannot find valid erofs superblock"); + return ret; + } + + erofs_sbi.feature_compat = le32_to_cpu(dsb->feature_compat); + + blkszbits = dsb->blkszbits; + /* 9(512 bytes) + LOG_SECTORS_PER_BLOCK == LOG_BLOCK_SIZE */ + if (blkszbits != LOG_BLOCK_SIZE) { + erofs_err("blksize %u isn't supported on this platform", + 1 << blkszbits); + return ret; + } + + if (!check_layout_compatibility(&erofs_sbi, dsb)) + return ret; + + erofs_sbi.blocks = le32_to_cpu(dsb->blocks); + erofs_sbi.meta_blkaddr = le32_to_cpu(dsb->meta_blkaddr); + erofs_sbi.xattr_blkaddr = le32_to_cpu(dsb->xattr_blkaddr); + erofs_sbi.islotbits = EROFS_ISLOTBITS; + erofs_sbi.root_nid = le16_to_cpu(dsb->root_nid); + erofs_sbi.inos = le64_to_cpu(dsb->inos); + + erofs_sbi.build_time = le64_to_cpu(dsb->build_time); + erofs_sbi.build_time_nsec = le32_to_cpu(dsb->build_time_nsec); + + memcpy(&erofs_sbi.uuid, dsb->uuid, sizeof(dsb->uuid)); + return 0; +} diff --git a/fs/fs.c b/fs/fs.c index 7c682582c8..3c06b0252b 100644 --- a/fs/fs.c +++ b/fs/fs.c @@ -26,6 +26,7 @@ #include <linux/math64.h> #include <efi_loader.h> #include <squashfs.h> +#include <erofs.h> DECLARE_GLOBAL_DATA_PTR; @@ -304,6 +305,27 @@ static struct fstype_info fstypes[] = { .unlink = fs_unlink_unsupported, .mkdir = fs_mkdir_unsupported, }, +#endif +#if IS_ENABLED(CONFIG_FS_EROFS) + { + .fstype = FS_TYPE_EROFS, + .name = "erofs", + .null_dev_desc_ok = false, + .probe = erofs_probe, + .opendir = erofs_opendir, + .readdir = erofs_readdir, + .ls = fs_ls_generic, + .read = erofs_read, + .size = erofs_size, + .close = erofs_close, + .closedir = erofs_closedir, + .exists = erofs_exists, + .uuid = fs_uuid_unsupported, + .write = fs_write_unsupported, + .ln = fs_ln_unsupported, + .unlink = fs_unlink_unsupported, + .mkdir = fs_mkdir_unsupported, + }, #endif { .fstype = FS_TYPE_ANY, diff --git a/include/erofs.h b/include/erofs.h new file mode 100644 index 0000000000..79de78257f --- /dev/null +++ b/include/erofs.h @@ -0,0 +1,19 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +#ifndef _EROFS_H_ +#define _EROFS_H_ + +struct disk_partition; + +int erofs_opendir(const char *filename, struct fs_dir_stream **dirsp); +int erofs_readdir(struct fs_dir_stream *dirs, struct fs_dirent **dentp); +int erofs_probe(struct blk_desc *fs_dev_desc, + struct disk_partition *fs_partition); +int erofs_read(const char *filename, void *buf, loff_t offset, + loff_t len, loff_t *actread); +int erofs_size(const char *filename, loff_t *size); +int erofs_exists(const char *filename); +void erofs_close(void); +void erofs_closedir(struct fs_dir_stream *dirs); +int erofs_uuid(char *uuid_str); + +#endif /* _EROFS_H */ diff --git a/include/fs.h b/include/fs.h index 1c79e299fd..f96afc4c89 100644 --- a/include/fs.h +++ b/include/fs.h @@ -17,6 +17,7 @@ struct cmd_tbl; #define FS_TYPE_UBIFS 4 #define FS_TYPE_BTRFS 5 #define FS_TYPE_SQUASHFS 6 +#define FS_TYPE_EROFS 7 struct blk_desc; -- 2.25.1

4 29

converting IMX6 board to DM_ETH and DM_USB - usb_ether gadget support
by Tim Harvey 10 Feb '22

10 Feb '22

Greetings, I'm trying to convert the gwventana board support to DM_ETH and DM_USB and one item I have not resolved yet is USB Ethernet gadget support. For non-dm a call to 'usb_eth_initialize' creates a usb_ether gadget that can be used for ethernet communication between the IMX6 OTG controller in device mode to a USB host. How is that accomplished via dm? I find that drivers/usb/gadget/ether.c has dm support but it would appear that a controller of UCLASS_USB_GADGET_GENERIC must bind for it to work yet I don't see how this works with CI_UDC. Other IMX6 boards that have gone through dm conversion have simply dropped the call to usb_eth_initialize and I don't see what else they enable that provides this functionality. Any ideas? Best regards, Tim

2 6

[PATCH] fastboot: fix fastboot_set_reboot_flag()
by Roman Stratiienko 28 Jan '22

28 Jan '22

In case CONFIG_FASTBOOT_FLASH_MMC_DEV == 0, compile-time condition is not met and fastboot_set_reboot_flag() fails. Fixes: a362ce214f03 ("fastboot: Implement generic fastboot_set_reboot_flag") Signed-off-by: Roman Stratiienko <r.stratiienko(a)gmail.com> --- drivers/fastboot/fb_common.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/fastboot/fb_common.c b/drivers/fastboot/fb_common.c index cbcc3683c47..ef399d0c4ab 100644 --- a/drivers/fastboot/fb_common.c +++ b/drivers/fastboot/fb_common.c @@ -91,7 +91,7 @@ void fastboot_okay(const char *reason, char *response) */ int __weak fastboot_set_reboot_flag(enum fastboot_reboot_reason reason) { -#if CONFIG_IS_ENABLED(FASTBOOT_FLASH_MMC_DEV) +#ifdef CONFIG_FASTBOOT_FLASH_MMC_DEV static const char * const boot_cmds[] = { [FASTBOOT_REBOOT_REASON_BOOTLOADER] = "bootonce-bootloader", [FASTBOOT_REBOOT_REASON_FASTBOOTD] = "boot-fastboot", -- 2.27.0

4 3

[PATCH 00/27] Add F1C100s based PocketGo handheld support
by Yifan Gu 20 Jan '22

20 Jan '22

This series add support for the PocketGo handheld, a low cost portable gaming console that is based on the Allwinner F1C100s chip. A significant portion of this patchset is authored by: - Icenowy Zheng <icenowy(a)aosc.io> - George Hilliard <thirtythreeforty(a)gmail.com> This patchset includes 4 parts necessary to make PocketGo work: - suniv + f1c100s support by Icenowy Zheng <icenowy(a)aosc.io>, rebased by George Hilliard <thirtythreeforty(a)gmail.com> and then by me. See: https://github.com/Icenowy/u-boot/commits/f1c100s-spiflash - additional support for suniv clock and MMC controller by George Hilliard, rebased. See: https://github.com/thirtythreeforty/u-boot/tree/f1c100s-v2019.04 - additional support for serial uart1, and misc fixes for compilation, SPL -> U-Boot proper load on F1C100s. - DTS and defconfig for PocketGo I have accidentally CC'ed a few people while testing git send-email, I am awfully sorry for my lack of experience in email-patch based workflows. This is my first time attempting to contribute to U-Boot. Cc: Jagan Teki <jagan(a)amarulasolutions.com> Cc: Andre Przywara <andre.przywara(a)arm.com> Cc: Icenowy Zheng <icenowy(a)aosc.io> Cc: George Hilliard <thirtythreeforty(a)gmail.com> George Hilliard (6): sunxi: Don't provide enable_cache() on suniv sunxi: implement clock driver for suniv f1c100s sunxi: gpio: Add support for suniv-f1c100s sunxi: spi: restore bus speed and mode after reset sunxi: spi: Add suniv pin controller support sunxi: suniv: add device tree nodes for f1c100s MMC controllers Icenowy Zheng (12): arm: arm926ejs: start.S: port save_boot_params support from armv7 code arm: arm926ej-s: add sunxi code sunxi: add support for suniv architecture sunxi: suniv: add support for Lichee Pi Nano sunxi: suniv: add boot sequence for SPL to try sunxi: spi-spl: add support for SUNIV sunxi: add SPI0 node for suniv sunxi: enable SPI NOR on Lichee Pi Nano sunxi: add defconfig for Lichee Pi Nano with SPI Flash support sunxi: add support for UART at PF for suniv sunxi: do not set PF MMC0 pinmux when PF uart is used sunxi: allow to enable MMC driver when using PF UART0 Yifan Gu (9): arm: dts: sunxi: do not renumber if mmc2 does not exist board: licheepi_nano: set CONFIG_MMC_SUNXI_SLOT_EXTRA sunxi: suniv: add missing header include for udelay sunxi: suniv: allow serial connection on uart1 sunxi: suniv: do not detect boot sector on suniv sunxi: suniv: set SYS_TEXT_BASE default sunxi: suniv: disable mmc optimization for suniv arm: dts: suniv: define uart1 pins arm: dts: add pocketgo handheld support arch/arm/cpu/arm926ejs/Makefile | 1 + arch/arm/cpu/arm926ejs/start.S | 19 + arch/arm/cpu/arm926ejs/sunxi/Makefile | 16 + arch/arm/cpu/arm926ejs/sunxi/config.mk | 6 + arch/arm/cpu/arm926ejs/sunxi/fel_utils.S | 38 ++ arch/arm/cpu/arm926ejs/sunxi/lowlevel_init.S | 68 +++ arch/arm/cpu/arm926ejs/sunxi/start.c | 1 + arch/arm/cpu/arm926ejs/sunxi/timer.c | 114 ++++ arch/arm/cpu/arm926ejs/sunxi/u-boot-spl.lds | 63 +++ arch/arm/dts/Makefile | 3 + arch/arm/dts/suniv-f1c100s-licheepi-nano.dts | 58 ++ arch/arm/dts/suniv-f1c100s-pocketgo.dts | 48 ++ arch/arm/dts/suniv-f1c100s.dtsi | 6 + arch/arm/dts/suniv.dtsi | 253 +++++++++ arch/arm/dts/sunxi-u-boot.dtsi | 2 + arch/arm/include/asm/arch-sunxi/clock.h | 2 +- arch/arm/include/asm/arch-sunxi/clock_sun6i.h | 21 + arch/arm/include/asm/arch-sunxi/cpu_sun4i.h | 8 + arch/arm/include/asm/arch-sunxi/dram.h | 2 + arch/arm/include/asm/arch-sunxi/dram_suniv.h | 47 ++ arch/arm/include/asm/arch-sunxi/gpio.h | 3 + arch/arm/mach-sunxi/Kconfig | 16 +- arch/arm/mach-sunxi/Makefile | 2 + arch/arm/mach-sunxi/board.c | 37 +- arch/arm/mach-sunxi/clock.c | 3 +- arch/arm/mach-sunxi/clock_sun6i.c | 47 +- arch/arm/mach-sunxi/cpu_info.c | 2 + arch/arm/mach-sunxi/dram_helpers.c | 4 + arch/arm/mach-sunxi/dram_suniv.c | 497 ++++++++++++++++++ arch/arm/mach-sunxi/spl_spi_sunxi.c | 13 + board/sunxi/board.c | 8 +- common/Kconfig.boot | 3 +- configs/licheepi_nano_defconfig | 10 + configs/licheepi_nano_spiflash_defconfig | 26 + configs/pocketgo_defconfig | 12 + drivers/clk/sunxi/Kconfig | 7 + drivers/clk/sunxi/Makefile | 1 + drivers/clk/sunxi/clk_f1c100s.c | 66 +++ drivers/gpio/sunxi_gpio.c | 6 + drivers/mmc/Kconfig | 2 +- drivers/mmc/sunxi_mmc.c | 6 +- drivers/spi/spi-sunxi.c | 12 +- include/configs/suniv.h | 15 + include/configs/sunxi-common.h | 69 ++- include/dt-bindings/clock/suniv-ccu.h | 69 +++ include/dt-bindings/reset/suniv-ccu.h | 37 ++ 46 files changed, 1715 insertions(+), 34 deletions(-) create mode 100644 arch/arm/cpu/arm926ejs/sunxi/Makefile create mode 100644 arch/arm/cpu/arm926ejs/sunxi/config.mk create mode 100644 arch/arm/cpu/arm926ejs/sunxi/fel_utils.S create mode 100644 arch/arm/cpu/arm926ejs/sunxi/lowlevel_init.S create mode 100644 arch/arm/cpu/arm926ejs/sunxi/start.c create mode 100644 arch/arm/cpu/arm926ejs/sunxi/timer.c create mode 100644 arch/arm/cpu/arm926ejs/sunxi/u-boot-spl.lds create mode 100644 arch/arm/dts/suniv-f1c100s-licheepi-nano.dts create mode 100644 arch/arm/dts/suniv-f1c100s-pocketgo.dts create mode 100644 arch/arm/dts/suniv-f1c100s.dtsi create mode 100644 arch/arm/dts/suniv.dtsi create mode 100644 arch/arm/include/asm/arch-sunxi/dram_suniv.h create mode 100644 arch/arm/mach-sunxi/dram_suniv.c create mode 100644 configs/licheepi_nano_defconfig create mode 100644 configs/licheepi_nano_spiflash_defconfig create mode 100644 configs/pocketgo_defconfig create mode 100644 drivers/clk/sunxi/clk_f1c100s.c create mode 100644 include/configs/suniv.h create mode 100644 include/dt-bindings/clock/suniv-ccu.h create mode 100644 include/dt-bindings/reset/suniv-ccu.h -- 2.25.1

3 29

[PATCH] gpio: mxc_gpio: Fix i.MX8M GPIO output status read
by Harm Berntsen 18 Jan '22

18 Jan '22

Currently the driver gets value from PSR register, but this register is only for input mode. For output mode, it always returns 0, not the value we set for output. This patch changes to use DR register, which returns the DR value for output mode, and PSR value for input mode. This patch is based on code from Ye Li <ye.li(a)nxp.com> Signed-off-by: Robert Krikke <robert.krikke(a)nedap.com> Signed-off-by: Harm Berntsen <harm.berntsen(a)nedap.com> CC: Ye Li <ye.li(a)nxp.com> CC: Stefano Babic <sbabic(a)denx.de> --- drivers/gpio/mxc_gpio.c | 20 +++++++++++++------- 1 file changed, 13 insertions(+), 7 deletions(-) diff --git a/drivers/gpio/mxc_gpio.c b/drivers/gpio/mxc_gpio.c index 06e6b2279f..c30246399c 100644 --- a/drivers/gpio/mxc_gpio.c +++ b/drivers/gpio/mxc_gpio.c @@ -36,6 +36,17 @@ struct mxc_bank_info { struct gpio_regs *regs; }; +static int gpio_bank_get_value(struct gpio_regs *regs, int offset) +{ + u32 gpio_val; + + if (IS_ENABLED(CONFIG_ARCH_IMX8M)) + gpio_val = readl(&regs->gpio_dr); + else + gpio_val = readl(&regs->gpio_psr); + return (gpio_val >> offset) & 0x01; +} + #if !CONFIG_IS_ENABLED(DM_GPIO) #define GPIO_TO_PORT(n) ((n) / 32) @@ -125,18 +136,13 @@ int gpio_get_value(unsigned gpio) { unsigned int port = GPIO_TO_PORT(gpio); struct gpio_regs *regs; - u32 val; - if (port >= ARRAY_SIZE(gpio_ports)) return -1; gpio &= 0x1f; regs = (struct gpio_regs *)gpio_ports[port]; - - val = (readl(&regs->gpio_psr) >> gpio) & 0x01; - - return val; + return gpio_bank_get_value(regs, gpio); } int gpio_request(unsigned gpio, const char *label) @@ -211,7 +217,7 @@ static void mxc_gpio_bank_set_value(struct gpio_regs *regs, int offset, static int mxc_gpio_bank_get_value(struct gpio_regs *regs, int offset) { - return (readl(&regs->gpio_psr) >> offset) & 0x01; + return gpio_bank_get_value(regs, offset); } /* set GPIO pin 'gpio' as an input */ -- 2.32.0

2 5

[PATCH 0/6] arm: sunxi: v3s: add ethernet support
by Andreas Rehn 15 Jan '22

15 Jan '22

This patchset enables ethernet on Allwinner V3/Sochip S3 SoCs. It is tested with an LicheePi Zero Dock board in combination with different switches and one direct link. Test scenario: Download mainline kernel and device-tree over tftp and start kernel with nfs rootfs. Without [PATCH 6/6], i wasn't able to get a stable connection on 100 Mb full duplex switches with autonegation enabled. Maybe the internal phy has a different behavior on softreset then others which results in a delayed established link. Andreas Rehn (6): dts: sunxi: add licheepi-zero-dock clk: sunxi: v3s: Implement EMAC clocks/resets clk: sunxi: v3s: fix tabs / spaces net: sun8i-emac: add v3s pinmux setting dts: sunxi: v3s: enable emac support net: sun8i-emac: v3s: fix soft reset timeout arch/arm/dts/Makefile | 3 ++- arch/arm/dts/sun8i-v3s-licheepi-zero-dock.dts | 11 +++++++++++ arch/arm/dts/sun8i-v3s.dtsi | 10 +++++++++- board/sunxi/MAINTAINERS | 5 +++++ configs/LicheePi_Zero_dock_defconfig | 7 +++++++ drivers/clk/sunxi/clk_v3s.c | 10 ++++++++-- drivers/net/sun8i_emac.c | 5 ++++- 7 files changed, 46 insertions(+), 5 deletions(-) create mode 100644 configs/LicheePi_Zero_dock_defconfig -- 2.25.1

7 33

IMX8MM SD UHS support
by Tim Harvey 14 Jan '22

14 Jan '22

Greetings, In 50b1a69cee0d ("ARM: dts: imx8m: add UHS or HS400/HS400ES properties") u-boot dt props were added to enable UHS and HS400 on a couple of IMX8MM boards including the imx8mm-evk and in the subsequent patch enabled the config items. While I see this making a huge difference for eMMC performance in U-Boot I find it doesn't do anything for microSD performance. The issue appears to be that sd_get_capabilities() is not appropriately detecting UHS speeds on cards that appropriately detect as SDR104/DDR50 in Linux: u-boot=> mmc info Device: FSL_SDHC Manufacturer ID: 1b OEM: 534d Name: 00000 Bus Speed: 50000000 Mode: SD High Speed (50MHz) card capabilities: widths [4, 1] modes [MMC legacy, SD High Speed (50MHz)] ^^^^ no SDR104 detected for this card host capabilities: widths [4, 1] modes [MMC legacy, MMC High Speed (26MHz), SD High Speed (50MHz), MMC High Speed (52MHz), UHS DDR50 (50MHz), UHS SDR104 (208MHz)] Rd Block Len: 512 SD version 3.0 High Capacity: Yes Capacity: 14.9 GiB Bus Width: 4-bit Erase Group Size: 512 Bytes The same card in Linux shows the following upon insertion and performance tests show that it is operating at SDR104 speeds: mmc1: new ultra high speed SDR104 SDHC card at address 0001 I haven't found very good documentation on the SD switch settings to understand if something is wrong in the U-Boot implementation of of sd_get_capabilities() and I suppose it also could be an issue in sdhci-esdhc-imx.c. Any ideas where to look? Anyone see SDR104 cards being detected properly for other platforms? Best regards, Tim

6 18

[PATCH] net: emaclite: fix possible underflow bug
by Tianrui Wei 10 Jan '22

10 Jan '22

This commit fixes a corner case when length < first_read. which would cause the last argument of xemaclite_alignedread to be a very large unsigned integer. Signed-off-by: Tianrui Wei <tianrui-wei(a)outlook.com> --- drivers/net/xilinx_emaclite.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/xilinx_emaclite.c b/drivers/net/xilinx_emaclite.c index 43fc36dc6a..3b8f0f4678 100644 --- a/drivers/net/xilinx_emaclite.c +++ b/drivers/net/xilinx_emaclite.c @@ -526,7 +526,7 @@ try_again: } /* Read the rest of the packet which is longer then first read */ - if (length != first_read) + if (length >= first_read) xemaclite_alignedread(addr + first_read, etherrxbuff + first_read, length - first_read); -- 2.32.0

1 1

[PATCH] engicam: Rename board dir, px30_core to px30
by Jagan Teki 24 Dec '21

24 Dec '21

Engicam hardware design solutions are based on System On Modules(SoM) with various SoCs like i.MX6, i.MX8 and PX30 etc. The current directory structure uses the simplest way to understand these designs by listing each of these SoC in board/engicam/SoC and inside SoC directory has SoM file name as a board file name. Right now all of the SoC's follow a similar approach except px30. So rename px30_core to px30. Signed-off-by: Jagan Teki <jagan(a)amarulasolutions.com> --- Note: This patch is on top of https://patchwork.ozlabs.org/project/uboot/cover/20210804090551.90819-1-jag… arch/arm/mach-rockchip/px30/Kconfig | 2 +- board/engicam/{px30_core => px30}/Kconfig | 2 +- board/engicam/{px30_core => px30}/MAINTAINERS | 0 board/engicam/{px30_core => px30}/Makefile | 0 board/engicam/{px30_core => px30}/px30_core.c | 0 5 files changed, 2 insertions(+), 2 deletions(-) rename board/engicam/{px30_core => px30}/Kconfig (90%) rename board/engicam/{px30_core => px30}/MAINTAINERS (100%) rename board/engicam/{px30_core => px30}/Makefile (100%) rename board/engicam/{px30_core => px30}/px30_core.c (100%) diff --git a/arch/arm/mach-rockchip/px30/Kconfig b/arch/arm/mach-rockchip/px30/Kconfig index 98e09019a2..7b9b380dfa 100644 --- a/arch/arm/mach-rockchip/px30/Kconfig +++ b/arch/arm/mach-rockchip/px30/Kconfig @@ -68,7 +68,7 @@ config DEBUG_UART_CHANNEL For using the UART for early debugging the route to use needs to be declared (0 or 1). -source "board/engicam/px30_core/Kconfig" +source "board/engicam/px30/Kconfig" source "board/hardkernel/odroid_go2/Kconfig" source "board/rockchip/evb_px30/Kconfig" diff --git a/board/engicam/px30_core/Kconfig b/board/engicam/px30/Kconfig similarity index 90% rename from board/engicam/px30_core/Kconfig rename to board/engicam/px30/Kconfig index a03be78369..10a7ec203f 100644 --- a/board/engicam/px30_core/Kconfig +++ b/board/engicam/px30/Kconfig @@ -1,7 +1,7 @@ if TARGET_PX30_CORE config SYS_BOARD - default "px30_core" + default "px30" config SYS_VENDOR default "engicam" diff --git a/board/engicam/px30_core/MAINTAINERS b/board/engicam/px30/MAINTAINERS similarity index 100% rename from board/engicam/px30_core/MAINTAINERS rename to board/engicam/px30/MAINTAINERS diff --git a/board/engicam/px30_core/Makefile b/board/engicam/px30/Makefile similarity index 100% rename from board/engicam/px30_core/Makefile rename to board/engicam/px30/Makefile diff --git a/board/engicam/px30_core/px30_core.c b/board/engicam/px30/px30_core.c similarity index 100% rename from board/engicam/px30_core/px30_core.c rename to board/engicam/px30/px30_core.c -- 2.25.1

2 1

Re: Broken arc toolchain
by Alexey Brodkin 18 Dec '21

18 Dec '21

Hi Tom and Simon, So there're a couple of things involved indeed. 1. As Tom correctly mentioned we used to use SNSP pre-built tools for ARC in TravisCI previously. And frankly the reason was we needed something GCC 8.x based while on some other arches such a new GCC triggered extra problems, see https://lists.denx.de/pipermail/u-boot/2018-May/329577.html. So I updated ARC pre-built versions and we kept using it. 2. Now, that particular "mcpu" was never added into upstream GCC because we wanted to re-implement it a bit nicer, thus any up-to-date vanilla GCC won't accept it as opposed to tools built from SNPS' GCC port [which has it https://github.com/foss-for-synopsys-dwc-arc-processors/gcc/commit/84a3ee38…]. 3. IIRC "-mcpu=hs4x_rel31" is basically an umbrella option for some options combination based on generic "-mcpu=hs4x" so we may probably update HSDK-4xD flags to be buildable by vanialla GGC. But anyways the simples short-term solution is to use our pre-built tools, latest of which are avaialble here https://github.com/foss-for-synopsys-dwc-arc-processors/toolchain/releases/…. And for U-Boot elf32 version is what I would suggest to use, so please use https://github.com/foss-for-synopsys-dwc-arc-processors/toolchain/releases/…. Moreover since by default for ARC we use U-Boot built-in libgcc the same one little-endian toolchain might be safely used for building our big-endian configs equally well. If more info or actions needed, please let me know. -Alexey ________________________________ From: Tom Rini Sent: Saturday, July 10, 2021 11:22 PM To: Simon Glass Cc: U-Boot Mailing List; Alexey Brodkin; Eugeniy Paltsev Subject: Re: Broken arc toolchain On Sat, Jul 10, 2021 at 02:11:30PM -0600, Simon Glass wrote: > Hi, > > I see this: > > $ buildman --fetch-arch arc > > Downloading toolchain for arch 'arc' > Checking: https://www.kernel.org/pub/tools/crosstool/files/bin/x86_64/9.2.0/ > Downloading: https://www.kernel.org/pub/tools/crosstool/files/bin/x86_64/9.2.0/x86_64-gc… > Unpacking to: /home/sglass/.buildman-toolchains > Testing > - looking in > '/home/sglass/.buildman-toolchains/gcc-9.2.0-nolibc/arc-linux/.' > - looking in > '/home/sglass/.buildman-toolchains/gcc-9.2.0-nolibc/arc-linux/bin' > - found > '/home/sglass/.buildman-toolchains/gcc-9.2.0-nolibc/arc-linux/bin/arc-linux-gcc' > - looking in > '/home/sglass/.buildman-toolchains/gcc-9.2.0-nolibc/arc-linux/usr/bin' > Tool chain test: OK, arch='arc', priority 5 > > $ buildman -A hsdk_4xd > /home/sglass/.buildman-toolchains/gcc-9.2.0-nolibc/arc-linux/bin/arc-linux- > $ buildman hsdk_4xd > ... > arc-linux-gcc: error: unrecognized argument in option '-mcpu=hs4x_rel31 > > > Do you have any thoughts on what is wrong, please? There's two parts. The first part is I keep forgetting to update buildman to fetch the latest toolchains from kernel.org, it should be grabbing 11.1.0 now, not 9.2.0. The second part is that arc, nds32 and xtensa have non-standard toolchain locations, see tools/docker/Dockerfile for where to fetch. -- Tom

2 1