speakup: Turn i18n files utf-8

i18n currently assume latin1 encoding, which is not enough for most
languages.

This separates out the utf-8 processing of /dev/synthu, and uses it for
a new synth_writeu, which we make synth_printf now use. This has the
effect of making all the i18 messages processed in utf-8.

Signed-off-by: Samuel Thibault <samuel.thibault@ens-lyon.org>
Link: https://lore.kernel.org/r/20240327115051.ng7xqnhozyii4ik2@begin
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>

1 files changed, 13 insertions, 46 deletions

diff --git a/drivers/accessibility/speakup/devsynth.c b/drivers/accessibility/speakup/devsynth.c
index cb7e1114e8eb..e3d909bd0480 100644
--- a/drivers/accessibility/speakup/devsynth.c
+++ b/drivers/accessibility/speakup/devsynth.c
@@ -39,13 +39,13 @@ static ssize_t speakup_file_write(struct file *fp, const char __user *buffer,
 static ssize_t speakup_file_writeu(struct file *fp, const char __user *buffer,
 				   size_t nbytes, loff_t *ppos)
 {
-	size_t count = nbytes, want;
+	size_t count = nbytes, consumed, want;
 	const char __user *ptr = buffer;
 	size_t bytes;
 	unsigned long flags;
 	unsigned char buf[256];
 	u16 ubuf[256];
-	size_t in, in2, out;
+	size_t in, out;
 
 	if (!synth)
 		return -ENODEV;
@@ -58,57 +58,24 @@ static ssize_t speakup_file_writeu(struct file *fp, const char __user *buffer,
 			return -EFAULT;
 
 		/* Convert to u16 */
-		for (in = 0, out = 0; in < bytes; in++) {
-			unsigned char c = buf[in];
-			int nbytes = 8 - fls(c ^ 0xff);
-			u32 value;
-
-			switch (nbytes) {
-			case 8: /* 0xff */
-			case 7: /* 0xfe */
-			case 1: /* 0x80 */
-				/* Invalid, drop */
-				goto drop;
-
-			case 0:
-				/* ASCII, copy */
-				ubuf[out++] = c;
-				continue;
+		for (in = 0, out = 0; in < bytes; in += consumed) {
+			s32 value;
 
-			default:
-				/* 2..6-byte UTF-8 */
+			value = synth_utf8_get(buf + in, bytes - in, &consumed, &want);
+			if (value == -1) {
+				/* Invalid or incomplete */
 
-				if (bytes - in < nbytes) {
+				if (want > bytes - in)
 					/* We don't have it all yet, stop here
 					 * and wait for the rest
 					 */
 					bytes = in;
-					want = nbytes;
-					continue;
-				}
-
-				/* First byte */
-				value = c & ((1u << (7 - nbytes)) - 1);
-
-				/* Other bytes */
-				for (in2 = 2; in2 <= nbytes; in2++) {
-					c = buf[in + 1];
-					if ((c & 0xc0) != 0x80)	{
-						/* Invalid, drop the head */
-						want = 1;
-						goto drop;
-					}
-					value = (value << 6) | (c & 0x3f);
-					in++;
-				}
-
-				if (value < 0x10000)
-					ubuf[out++] = value;
-				want = 1;
-				break;
+
+				continue;
 			}
-drop:
-			/* empty statement */;
+
+			if (value < 0x10000)
+				ubuf[out++] = value;
 		}
 
 		count -= bytes;