BUG: unable to handle kernel NULL pointer dereference at 0000000000000028
IP: [<ffffffffa02c66b4>] hidraw_ioctl+0xfc/0x32c [hid]
[...]
This is reproducible by disconnecting the device while userspace does ioctl in
a loop and doesn't check return values in order to exit the loop, like in the
following test program:
int main(int argc, char *argv[])
{
int fd = -1;
unsigned char name[256];
int ret;
if (argc != 2) {
fprintf(stderr, "usage: %s </dev/hidrawX>\n",
argv[0]);
exit(1);
}
fd = open(argv[1], O_RDWR);
if (fd < 0) {
perror("hidraw open");
exit(1);
}
while (1) {
ret = ioctl(fd, HIDIOCGRAWNAME(256), name);
printf("ret: %d name: %s\n", ret, name);
}
close(fd);
exit(0);
}
Signed-off-by: Antonio Ospite <ospite@studenti.unina.it>
---
Should this be applied to older stable kernels too?
Alan, Jiri,
there is a similar problem when _writing_ to the device, but Alan's
changes in that area are shuffling the code a bit, should I send a patch
[to hidraw_send_report()] on top of Alan's work for that, or a fix for
current mainline [in hidraw_write()] on which Alan should rebase his
work would be better?
The same pattern of unchecked hidraw_table[minor] is also present in
hidraw_get_report but this function is called only after the NULL check
in hidraw_ioctl _for_now_, so that is currently safe.
Thanks,
Antonio Ospite
http://ao2.it
drivers/hid/hidraw.c | 5 +++++
1 files changed, 5 insertions(+), 0 deletions(-)
diff --git a/drivers/hid/hidraw.c b/drivers/hid/hidraw.c
index 47d70c5..9eaf6ae 100644
--- a/drivers/hid/hidraw.c
+++ b/drivers/hid/hidraw.c
@@ -244,6 +244,10 @@ static long hidraw_ioctl(struct file *file, unsigned int cmd,
mutex_lock(&minors_lock);
dev = hidraw_table[minor];
+ if (!dev) {
+ ret = -ENODEV;
+ goto out;
+ }
switch (cmd) {
case HIDIOCGRDESCSIZE:
@@ -317,6 +321,7 @@ static long hidraw_ioctl(struct file *file, unsigned int cmd,
ret = -ENOTTY;
}
+out:
...
