[PATCH] erofs-utils: mkfs: Support tar source without data

[Mike Baynton]

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 at mbaynton.com>
---
Sorry for the delay getting to this, it fulfills a feature request
discussed a week or so ago:
<https://lore.kernel.org/linux-erofs/5db6f727-00e1-4625-9914-3ffabfc8c43f@linux.alibaba.com>
It sounded like we wanted the new option to be more verbose than just
`h`, and I thought "headerball" hinted at what it is more than "header",
but happy to change it if that's too weird!

Thank you for your careful review, it's a simple change but this was
motivated by necessity and isn't really my wheelhouse.

 include/erofs/tar.h |  2 +-
 lib/tar.c           |  2 +-
 man/mkfs.erofs.1    | 23 +++++++++++++++++------
 mkfs/main.c         |  9 ++++++---
 4 files changed, 25 insertions(+), 11 deletions(-)

diff --git a/include/erofs/tar.h b/include/erofs/tar.h
index a76f740..3d40a0f 100644
--- 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 8204939..e916395 100644
--- a/lib/tar.c
+++ b/lib/tar.c
@@ -584,7 +584,7 @@ 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 00ac2ac..d488983 100644
--- 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
@@ -180,11 +180,22 @@ 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.
+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 6d2b700..36e248a 100644
--- a/mkfs/main.c
+++ b/mkfs/main.c
@@ -122,7 +122,8 @@ static void usage(void)
 	      " --max-extent-bytes=#  set maximum decompressed extent size # in bytes\n"
 	      " --preserve-mtime      keep per-file modification time strictly\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, i=index, headerball=tar w/o file data\n"
 	      " --ovlfs-strip=[01]    strip overlayfs metadata in the target image (e.g. whiteouts)\n"
 	      " --quiet               quiet execution (do not write anything to standard output.)\n"
 #ifndef NDEBUG
@@ -514,11 +515,13 @@ 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;
-- 
2.34.1