[PATCH 7/8] Support non-BMP characters in NTFS.

[Vladimir 'Ï-coder/phcoder' Serbinenko]

ntfs-3g supports non-BMP characters just fine. This patch make standard read-only driver work as well.
Patch is bigger than required since checkpatch.pl asked me to restructure the code to avoid overlong lines.

Signed-off-by: Vladimir Serbinenko <phcoder@xxxxxxxxx>
---
 fs/ntfs/unistr.c |   70 ++++++++++++++++++++++++++++++++----------------------
 1 file changed, 41 insertions(+), 29 deletions(-)

diff --git a/fs/ntfs/unistr.c b/fs/ntfs/unistr.c
index 0d371e7..79f82d2 100644
--- a/fs/ntfs/unistr.c
+++ b/fs/ntfs/unistr.c
@@ -265,33 +265,35 @@ int ntfs_nlstoucs(const ntfs_volume *vol, const char *ins,
 	int i, o, wc_len;
 
 	/* We do not trust outside sources. */
-	if (likely(ins)) {
-		ucs = kmem_cache_alloc(ntfs_name_cache, GFP_NOFS);
-		if (likely(ucs)) {
-			for (i = o = 0; i < ins_len; i += wc_len) {
-				wc_len = nls->char2uni(ins + i, ins_len - i,
-						&wc);
-				if (likely(wc_len >= 0 && wc <= 0xffff
-					   && o < NTFS_MAX_NAME_LEN)) {
-					if (likely(wc)) {
-						ucs[o++] = cpu_to_le16(wc);
-						continue;
-					} /* else if (!wc) */
-					break;
-				} /* else if (wc_len < 0 ||
-						o >= NTFS_MAX_NAME_LEN) */
-				goto name_err;
-			}
-			ucs[o] = 0;
-			*outs = ucs;
-			return o;
-		} /* else if (!ucs) */
-		ntfs_error(vol->sb, "Failed to allocate buffer for converted "
-				"name from ntfs_name_cache.");
+	if (unlikely(!ins)) {
+		ntfs_error(vol->sb, "Received NULL pointer.");
+		return -EINVAL;
+	}
+
+	ucs = kmem_cache_alloc(ntfs_name_cache, GFP_NOFS);
+	if (unlikely(!ucs)) {
+		ntfs_error(vol->sb,
+			   "Failed to allocate buffer for converted "
+			   "name from ntfs_name_cache.");
 		return -ENOMEM;
-	} /* else if (!ins) */
-	ntfs_error(vol->sb, "Received NULL pointer.");
-	return -EINVAL;
+	}
+
+	for (i = o = 0; i < ins_len; i += wc_len) {
+		int s;
+		wc_len = nls->char2uni(ins + i, ins_len - i, &wc);
+		if (unlikely(wc_len < 0 || o >= NTFS_MAX_NAME_LEN))
+			goto name_err;
+		if (unlikely(!wc))
+			break;
+		s = unicode_to_utf16s(wc, UTF16_LITTLE_ENDIAN, ucs + o,
+				      NTFS_MAX_NAME_LEN - o);
+		if (s <= 0)
+			break;
+		o += s;
+	}
+	ucs[o] = 0;
+	*outs = ucs;
+	return o;
 name_err:
 	kmem_cache_free(ntfs_name_cache, ucs);
 	if (wc_len < 0) {
@@ -342,6 +344,7 @@ int ntfs_ucstonls(const ntfs_volume *vol, const ntfschar *ins,
 
 	/* We don't trust outside sources. */
 	if (ins) {
+		int step;
 		ns = *outs;
 		ns_len = outs_len;
 		if (ns && !ns_len) {
@@ -354,9 +357,18 @@ int ntfs_ucstonls(const ntfs_volume *vol, const ntfschar *ins,
 			if (!ns)
 				goto mem_err_out;
 		}
-		for (i = o = 0; i < ins_len; i++) {
-retry:			wc = nls->uni2char(le16_to_cpu(ins[i]), ns + o,
-					ns_len - o);
+		for (i = o = 0; i < ins_len; i += step) {
+			unicode_t uni;
+retry:
+			step = utf16s_to_unicode(ins + i, ins_len - i,
+						 UTF16_LITTLE_ENDIAN,
+						 &uni);
+			if (step <= 0) {
+				uni = le16_to_cpu(ins[i]);
+				step = 1;
+			}
+
+			wc = nls->uni2char(uni, ns + o, ns_len - o);
 			if (wc > 0) {
 				o += wc;
 				continue;
-- 
1.7.10

-- 
Regards
Vladimir 'Ï-coder/phcoder' Serbinenko

Attachment: signature.asc
Description: OpenPGP digital signature