| Message ID | 20201104134514.28413-1-matthias.schiffer@ew.tq-group.com |
|---|---|
| State | Accepted |
| Delegated to: | Tom Rini |
| Headers | show |
| Series | net: eth-uclass: avoid running start() twice without stop() | expand |
On Wed, Nov 04, 2020 at 02:45:14PM +0100, Matthias Schiffer wrote: > Running the start() handler twice without a stop() inbetween completely > breaks communication for some ethernet drivers like fec_mxc. > > eth_halt() is called before each eth_init(). Due to the switch to > eth_is_active() in commit 68acb51f442f ("net: Only call halt on a driver > that has been init'ed"), this is not sufficient anymore when netconsole > is active: eth_init_state_only()/eth_halt_state_only() manipulate the > state check that is performed by eth_is_active() without actually > calling into the driver. > > The issue can be triggered by starting a network operation (e.g. ping or > tftp) while netconsole is active. > > Add an additional "running" flag that reflects the actual state of the > driver and use it to ensure that eth_halt() actually stops the device as > it is supposed to. > > Fixes: 68acb51f442f ("net: Only call halt on a driver that has been init'ed") > Signed-off-by: Matthias Schiffer <matthias.schiffer@ew.tq-group.com> Applied to u-boot/master, thanks!
diff --git a/net/eth-uclass.c b/net/eth-uclass.c index e14695c0f1..7c9278f3a9 100644 --- a/net/eth-uclass.c +++ b/net/eth-uclass.c @@ -26,6 +26,7 @@ DECLARE_GLOBAL_DATA_PTR; */ struct eth_device_priv { enum eth_state_t state; + bool running; }; /** @@ -293,6 +294,7 @@ int eth_init(void) current->uclass_priv; priv->state = ETH_STATE_ACTIVE; + priv->running = true; return 0; } } else { @@ -322,13 +324,16 @@ void eth_halt(void) struct eth_device_priv *priv; current = eth_get_dev(); - if (!current || !eth_is_active(current)) + if (!current) return; - eth_get_ops(current)->stop(current); priv = current->uclass_priv; - if (priv) - priv->state = ETH_STATE_PASSIVE; + if (!priv || !priv->running) + return; + + eth_get_ops(current)->stop(current); + priv->state = ETH_STATE_PASSIVE; + priv->running = false; } int eth_is_active(struct udevice *dev) @@ -537,6 +542,7 @@ static int eth_post_probe(struct udevice *dev) #endif priv->state = ETH_STATE_INIT; + priv->running = false; /* Check if the device has a valid MAC address in device tree */ if (!eth_dev_get_mac_address(dev, pdata->enetaddr) ||
Running the start() handler twice without a stop() inbetween completely breaks communication for some ethernet drivers like fec_mxc. eth_halt() is called before each eth_init(). Due to the switch to eth_is_active() in commit 68acb51f442f ("net: Only call halt on a driver that has been init'ed"), this is not sufficient anymore when netconsole is active: eth_init_state_only()/eth_halt_state_only() manipulate the state check that is performed by eth_is_active() without actually calling into the driver. The issue can be triggered by starting a network operation (e.g. ping or tftp) while netconsole is active. Add an additional "running" flag that reflects the actual state of the driver and use it to ensure that eth_halt() actually stops the device as it is supposed to. Fixes: 68acb51f442f ("net: Only call halt on a driver that has been init'ed") Signed-off-by: Matthias Schiffer <matthias.schiffer@ew.tq-group.com> --- net/eth-uclass.c | 14 ++++++++++---- 1 file changed, 10 insertions(+), 4 deletions(-)