3.16.80-rc1 review patch. If anyone has any objections, please let me know.
------------------
From: Johan Hovold johan@kernel.org
commit d98ee2a19c3334e9343df3ce254b496f1fc428eb upstream.
The custom ring-buffer implementation was merged without any locking or explicit memory barriers, but a spinlock was later added by commit 9d33efd9a791 ("USB: ldusb bugfix").
The lock did not cover the update of the tail index once the entry had been processed, something which could lead to memory corruption on weakly ordered architectures or due to compiler optimisations.
Specifically, a completion handler running on another CPU might observe the incremented tail index and update the entry before ld_usb_read() is done with it.
Fixes: 2824bd250f0b ("[PATCH] USB: add ldusb driver") Fixes: 9d33efd9a791 ("USB: ldusb bugfix") Signed-off-by: Johan Hovold johan@kernel.org Link: https://lore.kernel.org/r/20191022143203.5260-2-johan@kernel.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org Signed-off-by: Ben Hutchings ben@decadent.org.uk --- drivers/usb/misc/ldusb.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)
--- a/drivers/usb/misc/ldusb.c +++ b/drivers/usb/misc/ldusb.c @@ -509,11 +509,11 @@ static ssize_t ld_usb_read(struct file * retval = -EFAULT; goto unlock_exit; } - dev->ring_tail = (dev->ring_tail+1) % ring_buffer_size; - retval = bytes_to_read;
spin_lock_irq(&dev->rbsl); + dev->ring_tail = (dev->ring_tail + 1) % ring_buffer_size; + if (dev->buffer_overflow) { dev->buffer_overflow = 0; spin_unlock_irq(&dev->rbsl);