<div dir="ltr">Fixed most of the issues you pointed out. Except I didn't quite understand the "nid is optional unless we do a fsck." part. Not sure how we can implement the iterate dir function w/o nid. Can you provide more context?<br></div><br><div class="gmail_quote"><div dir="ltr" class="gmail_attr">On Mon, Dec 13, 2021 at 6:20 PM Kelvin Zhang <<a href="mailto:zhangkelvin@google.com">zhangkelvin@google.com</a>> wrote:<br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">Change-Id: Ia35708080a72ee204eaaddfc670d3cb8023a078c<br>
Signed-off-by: Kelvin Zhang <<a href="mailto:zhangkelvin@google.com" target="_blank">zhangkelvin@google.com</a>><br>
---<br>
include/erofs/iterate.h | 40 +++++++++++<br>
include/erofs_fs.h | 4 +-<br>
lib/Makefile.am | 2 +-<br>
lib/iterate.c | 148 ++++++++++++++++++++++++++++++++++++++++<br>
4 files changed, 191 insertions(+), 3 deletions(-)<br>
create mode 100644 include/erofs/iterate.h<br>
create mode 100644 lib/iterate.c<br>
<br>
diff --git a/include/erofs/iterate.h b/include/erofs/iterate.h<br>
new file mode 100644<br>
index 0000000..af29e14<br>
--- /dev/null<br>
+++ b/include/erofs/iterate.h<br>
@@ -0,0 +1,40 @@<br>
+// SPDX-License-Identifier: Apache-2.0<br>
+<br>
+#ifndef ITERATE_ITERATE<br>
+#define ITERATE_ITERATE<br>
+<br>
+#ifdef __cplusplus<br>
+extern "C"<br>
+{<br>
+#endif<br>
+<br>
+<br>
+#include "erofs/io.h"<br>
+#include "erofs/print.h"<br>
+<br>
+<br>
+struct erofs_inode_info {<br>
+ const char* name;<br>
+ enum erofs_ftype ftype;<br>
+ struct erofs_inode* inode;<br>
+ void* arg;<br>
+};<br>
+// Callback function for iterating over inodes of EROFS<br>
+<br>
+typedef bool (*erofs_readdir_cb)(struct erofs_inode_info*);<br>
+<br>
+int erofs_iterate_dir(const struct erofs_sb_info* sbi,<br>
+ erofs_nid_t nid,<br>
+ erofs_nid_t parent_nid,<br>
+ erofs_readdir_cb cb,<br>
+ void* arg);<br>
+int erofs_iterate_root_dir(const struct erofs_sb_info* sbi,<br>
+ erofs_readdir_cb cbg,<br>
+ void* arg);<br>
+int erofs_get_occupied_size(struct erofs_inode* inode, erofs_off_t* size);<br>
+<br>
+#ifdef __cplusplus<br>
+}<br>
+#endif<br>
+<br>
+#endif // ITERATE_ITERATE<br>
diff --git a/include/erofs_fs.h b/include/erofs_fs.h<br>
index 9a91877..7ee8251 100644<br>
--- a/include/erofs_fs.h<br>
+++ b/include/erofs_fs.h<br>
@@ -389,8 +389,8 @@ struct erofs_dirent {<br>
} __packed;<br>
<br>
/* file types used in inode_info->flags */<br>
-enum {<br>
- EROFS_FT_UNKNOWN,<br>
+enum erofs_ftype {<br>
+ EROFS_FT_UNKNOWN = 0,<br>
EROFS_FT_REG_FILE,<br>
EROFS_FT_DIR,<br>
EROFS_FT_CHRDEV,<br>
diff --git a/lib/Makefile.am b/lib/Makefile.am<br>
index 67ba798..20c0e4f 100644<br>
--- a/lib/Makefile.am<br>
+++ b/lib/Makefile.am<br>
@@ -27,7 +27,7 @@ noinst_HEADERS = $(top_srcdir)/include/erofs_fs.h \<br>
noinst_HEADERS += compressor.h<br>
liberofs_la_SOURCES = config.c io.c cache.c super.c inode.c xattr.c exclude.c \<br>
namei.c data.c compress.c compressor.c zmap.c decompress.c \<br>
- compress_hints.c hashmap.c sha256.c blobchunk.c<br>
+ compress_hints.c hashmap.c sha256.c blobchunk.c iterate.c<br>
liberofs_la_CFLAGS = -Wall -Werror -I$(top_srcdir)/include<br>
if ENABLE_LZ4<br>
liberofs_la_CFLAGS += ${LZ4_CFLAGS}<br>
diff --git a/lib/iterate.c b/lib/iterate.c<br>
new file mode 100644<br>
index 0000000..ec44217<br>
--- /dev/null<br>
+++ b/lib/iterate.c<br>
@@ -0,0 +1,148 @@<br>
+// SPDX-License-Identifier: Apache-2.0<br>
+<br>
+#include "erofs/internal.h"<br>
+#include "erofs_fs.h"<br>
+#include "erofs/print.h"<br>
+#include "erofs/iterate.h"<br>
+<br>
+static int erofs_read_dirent(const struct erofs_sb_info* sbi,<br>
+ const struct erofs_dirent* de,<br>
+ erofs_nid_t nid,<br>
+ erofs_nid_t parent_nid,<br>
+ const char* dname,<br>
+ erofs_readdir_cb cb,<br>
+ void* arg) {<br>
+ int err;<br>
+ erofs_off_t occupied_size = 0;<br>
+ struct erofs_inode inode = {.nid = de->nid};<br>
+ err = erofs_read_inode_from_disk(&inode);<br>
+ if (err) {<br>
+ erofs_err("read file inode from disk failed!");<br>
+ return err;<br>
+ }<br>
+ char buf[PATH_MAX + 1];<br>
+ erofs_get_inode_name(sbi, de->nid, buf, PATH_MAX + 1);<br>
+ struct erofs_inode_info info = {<br>
+ .name = buf,<br>
+ .ftype = de->file_type,<br>
+ .inode = &inode,<br>
+ .arg = arg};<br>
+ cb(&info);<br>
+ if ((de->file_type == EROFS_FT_DIR) && de->nid != nid &&<br>
+ de->nid != parent_nid) {<br>
+ err = erofs_iterate_dir(sbi, de->nid, nid, cb, arg);<br>
+ if (err) {<br>
+ erofs_err("parse dir nid %u error occurred\n",<br>
+ (unsigned int)(de->nid));<br>
+ return err;<br>
+ }<br>
+ }<br>
+ return 0;<br>
+}<br>
+<br>
+static inline int erofs_checkdirent(const struct erofs_dirent* de,<br>
+ const struct erofs_dirent* last_de,<br>
+ u32 maxsize,<br>
+ const char* dname) {<br>
+ int dname_len;<br>
+ unsigned int nameoff = le16_to_cpu(de->nameoff);<br>
+ if (nameoff < sizeof(struct erofs_dirent) || nameoff >= PAGE_SIZE) {<br>
+ erofs_err("invalid de[0].nameoff %u @ nid %llu", nameoff, de->nid | 0ULL);<br>
+ return -EFSCORRUPTED;<br>
+ }<br>
+ dname_len = (de + 1 >= last_de) ? strnlen(dname, maxsize - nameoff)<br>
+ : le16_to_cpu(de[1].nameoff) - nameoff;<br>
+ /* a corrupted entry is found */<br>
+ if (nameoff + dname_len > maxsize || dname_len > EROFS_NAME_LEN) {<br>
+ erofs_err("bogus dirent @ nid %llu", le64_to_cpu(de->nid) | 0ULL);<br>
+ DBG_BUGON(1);<br>
+ return -EFSCORRUPTED;<br>
+ }<br>
+ if (de->file_type >= EROFS_FT_MAX) {<br>
+ erofs_err("invalid file type %u", (unsigned int)(de->nid));<br>
+ return -EFSCORRUPTED;<br>
+ }<br>
+ return dname_len;<br>
+}<br>
+<br>
+int erofs_iterate_dir(const struct erofs_sb_info* sbi,<br>
+ erofs_nid_t nid,<br>
+ erofs_nid_t parent_nid,<br>
+ erofs_readdir_cb cb,<br>
+ void* arg) {<br>
+ int err;<br>
+ erofs_off_t offset;<br>
+ char buf[EROFS_BLKSIZ];<br>
+ struct erofs_inode vi = {.nid = nid};<br>
+ err = erofs_read_inode_from_disk(&vi);<br>
+ if (err)<br>
+ return err;<br>
+ struct erofs_inode_info inode_info = {<br>
+ .name = buf,<br>
+ .ftype = EROFS_FT_DIR,<br>
+ .inode = &vi,<br>
+ .arg = arg,<br>
+ };<br>
+ err = erofs_get_inode_name(sbi, nid, buf, EROFS_BLKSIZ);<br>
+ cb(&inode_info);<br>
+ if (err) {<br>
+ return err;<br>
+ }<br>
+ offset = 0;<br>
+ while (offset < vi.i_size) {<br>
+ erofs_off_t maxsize = min_t(erofs_off_t, vi.i_size - offset, EROFS_BLKSIZ);<br>
+ const struct erofs_dirent* de = (const struct erofs_dirent*)(buf);<br>
+ struct erofs_dirent* end;<br>
+ unsigned int nameoff;<br>
+ err = erofs_pread(&vi, buf, maxsize, offset);<br>
+ if (err)<br>
+ return err;<br>
+ nameoff = le16_to_cpu(de->nameoff);<br>
+ end = (struct erofs_dirent*)(buf + nameoff);<br>
+ while (de < end) {<br>
+ const char* dname;<br>
+ int ret;<br>
+ /* skip "." and ".." dentry */<br>
+ if (de->nid == nid || de->nid == parent_nid) {<br>
+ de++;<br>
+ continue;<br>
+ }<br>
+ dname = (char*)buf + nameoff;<br>
+ ret = erofs_checkdirent(de, end, maxsize, dname);<br>
+ if (ret < 0)<br>
+ return ret;<br>
+ ret = erofs_read_dirent(sbi, de, nid, parent_nid, dname, cb, arg);<br>
+ if (ret < 0)<br>
+ return ret;<br>
+ ++de;<br>
+ }<br>
+ offset += maxsize;<br>
+ }<br>
+ return 0;<br>
+}<br>
+<br>
+int erofs_get_occupied_size(struct erofs_inode* inode, erofs_off_t* size) {<br>
+ *size = 0;<br>
+ switch (inode->datalayout) {<br>
+ case EROFS_INODE_FLAT_INLINE:<br>
+ case EROFS_INODE_FLAT_PLAIN:<br>
+ case EROFS_INODE_CHUNK_BASED:<br>
+ *size = inode->i_size;<br>
+ break;<br>
+ case EROFS_INODE_FLAT_COMPRESSION_LEGACY:<br>
+ case EROFS_INODE_FLAT_COMPRESSION:<br>
+ *size = inode->u.i_blocks * EROFS_BLKSIZ;<br>
+ break;<br>
+ default:<br>
+ erofs_err("unknown datalayout");<br>
+ return -1;<br>
+ }<br>
+ return 0;<br>
+}<br>
+<br>
+int erofs_iterate_root_dir(const struct erofs_sb_info* sbi,<br>
+ erofs_readdir_cb cb,<br>
+ void* arg) {<br>
+ return erofs_iterate_dir(sbi, sbi->root_nid, sbi->root_nid, cb, arg);<br>
+}<br>
+<br>
-- <br>
2.34.1.173.g76aa8bc2d0-goog<br>
<br>
</blockquote></div><br clear="all"><div><br></div>-- <br><div dir="ltr" class="gmail_signature"><div dir="ltr">Sincerely,<div><br></div><div>Kelvin Zhang</div></div></div>