diff options
| author | Sergey Bugaev <bugaevc@gmail.com> | 2021-05-26 15:36:28 +0300 |
|---|---|---|
| committer | Samuel Thibault <samuel.thibault@ens-lyon.org> | 2022-08-10 22:11:38 +0200 |
| commit | 7360a092712ad01c5803901df5ca6e0edef4150f (patch) | |
| tree | bb0b78b99b867a239bf8e53dac74d3cae4ee41df /eth-multiplexer | |
| parent | 9393fb33d0405cfb99449241139413f0aae6f4f0 (diff) | |
| download | hurd-7360a092712ad01c5803901df5ca6e0edef4150f.tar.gz hurd-7360a092712ad01c5803901df5ca6e0edef4150f.tar.bz2 hurd-7360a092712ad01c5803901df5ca6e0edef4150f.zip | |
Do not use ports_get_right () for send-one rights
ports_get_right () expects the caller to make a send, not a send-once,
right from the returned receive right, and increments the expected make-send
count accordingly. The kernel, however, does not increment the make-send
count when a send-once right is being made.
The result can be described as a "no-senders leak": libports' idea of the
current make-send count always stays one step ahead of it actual value (or
several steps ahead, if the process is repeated), which makes libports
ignore *all* the subsequent no-senders notifications for the port as outdated.
Diffstat (limited to 'eth-multiplexer')
| -rw-r--r-- | eth-multiplexer/device_impl.c | 2 |
1 files changed, 1 insertions, 1 deletions
diff --git a/eth-multiplexer/device_impl.c b/eth-multiplexer/device_impl.c index a6e6f59c..74f78f22 100644 --- a/eth-multiplexer/device_impl.c +++ b/eth-multiplexer/device_impl.c @@ -189,7 +189,7 @@ ds_device_set_filter (struct vether_device *vdev, mach_port_t receive_port, return D_NO_SUCH_DEVICE; err = mach_port_request_notification (mach_task_self (), receive_port, MACH_NOTIFY_DEAD_NAME, 0, - ports_get_right (notify_pi), + notify_pi->port_right, MACH_MSG_TYPE_MAKE_SEND_ONCE, &tmp); if (err != KERN_SUCCESS) goto out; |