erofs-utils: mkfs: Support tar source without data

This improves performance of meta-only image creation in cases where the
source is a tarball stream that is not seekable. The writer may now use
`--tar=headerball` and omit the file data. Previously, the stream writer
was forced to send the file's size worth of null bytes or any data after
each tar header which was simply discarded by mkfs.erofs.

Signed-off-by: Mike Baynton <mike@mbaynton.com>
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Link: https://lore.kernel.org/r/20240227084221.342635-1-hsiangkao@linux.alibaba.com

diff --git a/include/erofs/tar.h b/include/erofs/tar.h
index be03d1bfe1d68430adea2e586b01c503016cf9e1..e45b89553f30ef6dd984cf750f22735a9a84ad30 100644 (file)
--- a/include/erofs/tar.h
+++ b/include/erofs/tar.h
@@ -46,7 +46,7 @@ struct erofs_tarfile {
 
        int fd;
        u64 offset;
-       bool index_mode, aufs;
+       bool index_mode, headeronly_mode, aufs;
 };
 
 void erofs_iostream_close(struct erofs_iostream *ios);

diff --git a/lib/tar.c b/lib/tar.c
index 1d764b24cd176c9b1f7efe2be930f935c32ac319..7c14c06d74b329e2aa26cfe21661e98ea3dc2b67 100644 (file)
--- a/lib/tar.c
+++ b/lib/tar.c
@@ -589,7 +589,7 @@ static int tarerofs_write_file_index(struct erofs_inode *inode,
        ret = tarerofs_write_chunkes(inode, data_offset);
        if (ret)
                return ret;
-       if (erofs_iostream_lskip(&tar->ios, inode->i_size))
+       if (!tar->headeronly_mode && erofs_iostream_lskip(&tar->ios, inode->i_size))
                return -EIO;
        return 0;
 }

diff --git a/man/mkfs.erofs.1 b/man/mkfs.erofs.1
index f32dc26e788cb7447ea2af50a09be8bf1342accb..41eb5fb9658b56ce69041a8e57af4196865a3523 100644 (file)
--- a/man/mkfs.erofs.1
+++ b/man/mkfs.erofs.1
@@ -17,7 +17,7 @@ achieve high performance for embedded devices with limited memory since it has
 unnoticable memory overhead and page cache thrashing.
 .PP
 mkfs.erofs is used to create such EROFS filesystem \fIDESTINATION\fR image file
-from \fISOURCE\fR directory.
+from \fISOURCE\fR directory or tarball.
 .SH OPTIONS
 .TP
 .BI "\-z " compression-algorithm \fR[\fP, # \fR][\fP: ... \fR]\fP
@@ -186,11 +186,22 @@ Use extended inodes instead of compact inodes if the file modification time
 would overflow compact inodes. This is the default. Overrides
 .BR --ignore-mtime .
 .TP
-.B "\-\-tar=f"
-Generate a full EROFS image from a tarball.
-.TP
-.B "\-\-tar=i"
-Generate an meta-only EROFS image from a tarball.
+.BI "\-\-tar, \-\-tar="MODE
+Treat \fISOURCE\fR as a tarball or tarball-like "headerball" rather than as a
+directory.
+
+\fIMODE\fR may be one of \fBf\fR, \fBi\fR, or \fBheaderball\fR.
+
+\fBf\fR: Generate a full EROFS image from a regular tarball. (default)
+
+\fBi\fR: Generate a meta-only EROFS image from a regular tarball. Only
+metadata such as dentries, inodes, and xattrs will be added to the image,
+without file data. Uses for such images include as a layer in an overlay
+filesystem with other data-only layers.
+
+\fBheaderball\fR: Generate a meta-only EROFS image from a stream identical
+to a tarball except that file data is not present after each file header.
+It can improve performance especially when \fISOURCE\fR is not seekable.
 .TP
 .BI "\-\-uid-offset=" UIDOFFSET
 Add \fIUIDOFFSET\fR to all file UIDs.

diff --git a/mkfs/main.c b/mkfs/main.c
index 3191b89f347b236c3f72e9edd9ffd4fda2aff77d..258c1ce4da82dac7a89c5f3f6ed21ef68c4a6804 100644 (file)
--- a/mkfs/main.c
+++ b/mkfs/main.c
@@ -162,7 +162,9 @@ static void usage(int argc, char **argv)
                " --preserve-mtime      keep per-file modification time strictly\n"
                " --offset=#            skip # bytes at the beginning of IMAGE.\n"
                " --aufs                replace aufs special files with overlayfs metadata\n"
-               " --tar=[fi]            generate an image from tarball(s)\n"
+               " --tar=X               generate a full or index-only image from a tarball(-ish) source\n"
+               "                       (X = f|i|headerball; f=full mode, i=index mode,\n"
+               "                                            headerball=file data is omited in the source stream)\n"
                " --ovlfs-strip=<0,1>   strip overlayfs metadata in the target image (e.g. whiteouts)\n"
                " --quiet               quiet execution (do not write anything to standard output.)\n"
 #ifndef NDEBUG
@@ -617,11 +619,13 @@ static int mkfs_parse_options_cfg(int argc, char *argv[])
                        cfg.c_extra_ea_name_prefixes = true;
                        break;
                case 20:
-                       if (optarg && (!strcmp(optarg, "i") ||
-                               !strcmp(optarg, "0") || !memcmp(optarg, "0,", 2))) {
+                       if (optarg && (!strcmp(optarg, "i") || (!strcmp(optarg, "headerball") ||
+                               !strcmp(optarg, "0") || !memcmp(optarg, "0,", 2)))) {
                                erofstar.index_mode = true;
                                if (!memcmp(optarg, "0,", 2))
                                        erofstar.mapfile = strdup(optarg + 2);
+                               if (!strcmp(optarg, "headerball"))
+                                       erofstar.headeronly_mode = true;
                        }
                        tar_mode = true;
                        break;