6.5-stable review patch. If anyone has any objections, please let me know.
------------------
From: Takashi Iwai <tiwai@xxxxxxx>
commit 86496fd4a2fabb7c978fdaca2d4b718207a96d36 upstream.
The recent fix to clear the padding bytes at
snd_seq_expand_var_event() broke the read to user-space with
in_kernel=0 parameter. For user-space address, it has to use
clear_user() instead of memset().
Fixes: f80e6d60d677 ("ALSA: seq: Clear padded bytes at expanding events")
Reported-and-tested-by: Ash Holland <ash@xxxxxxxxx>
Closes: https://lore.kernel.org/r/8a555319-9f31-4ea2-878f-adc338bc40d4@xxxxxxxxx
Link: https://lore.kernel.org/r/20230905052631.18240-1-tiwai@xxxxxxx
Link: https://lore.kernel.org/r/20230905081210.6731-1-tiwai@xxxxxxx
Signed-off-by: Takashi Iwai <tiwai@xxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
sound/core/seq/seq_memory.c | 9 +++++++--
1 file changed, 7 insertions(+), 2 deletions(-)
--- a/sound/core/seq/seq_memory.c
+++ b/sound/core/seq/seq_memory.c
@@ -187,8 +187,13 @@ int snd_seq_expand_var_event(const struc
err = expand_var_event(event, 0, len, buf, in_kernel);
if (err < 0)
return err;
- if (len != newlen)
- memset(buf + len, 0, newlen - len);
+ if (len != newlen) {
+ if (in_kernel)
+ memset(buf + len, 0, newlen - len);
+ else if (clear_user((__force void __user *)buf + len,
+ newlen - len))
+ return -EFAULT;
+ }
return newlen;
}
EXPORT_SYMBOL(snd_seq_expand_var_event);
