| Message ID | 20231121191435.4049995-3-Raju.Rangoju@amd.com (mailing list archive) |
|---|---|
| State | Accepted |
| Commit | 7121205d5330c6a3cb3379348886d47c77b78d06 |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | amd-xgbe: fixes to handle corner-cases | expand |
On 21.11.2023 20:14, Raju Rangoju wrote: > The existing implementation uses software logic to accumulate tx > completions until the specified time (1ms) is met and then poll them. > However, there exists a tiny gap which leads to a race between > resetting and checking the tx_activate flag. Due to this the tx > completions are not reported to upper layer and tx queue timeout > kicks-in restarting the device. > > To address this, introduce a tx cleanup mechanism as part of the > periodic maintenance process. > > Fixes: c5aa9e3b8156 ("amd-xgbe: Initial AMD 10GbE platform driver") > Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com> > Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com> > --- Reviewed-by: Wojciech Drewek <wojciech.drewek@intel.com> > drivers/net/ethernet/amd/xgbe/xgbe-drv.c | 14 ++++++++++++++ > 1 file changed, 14 insertions(+) > > diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c > index 614c0278419b..6b73648b3779 100644 > --- a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c > +++ b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c > @@ -682,10 +682,24 @@ static void xgbe_service(struct work_struct *work) > static void xgbe_service_timer(struct timer_list *t) > { > struct xgbe_prv_data *pdata = from_timer(pdata, t, service_timer); > + struct xgbe_channel *channel; > + unsigned int i; > > queue_work(pdata->dev_workqueue, &pdata->service_work); > > mod_timer(&pdata->service_timer, jiffies + HZ); > + > + if (!pdata->tx_usecs) > + return; > + > + for (i = 0; i < pdata->channel_count; i++) { > + channel = pdata->channel[i]; > + if (!channel->tx_ring || channel->tx_timer_active) > + break; > + channel->tx_timer_active = 1; > + mod_timer(&channel->tx_timer, > + jiffies + usecs_to_jiffies(pdata->tx_usecs)); > + } > } > > static void xgbe_init_timers(struct xgbe_prv_data *pdata)
On 11/21/23 13:14, Raju Rangoju wrote: > The existing implementation uses software logic to accumulate tx > completions until the specified time (1ms) is met and then poll them. > However, there exists a tiny gap which leads to a race between > resetting and checking the tx_activate flag. Due to this the tx > completions are not reported to upper layer and tx queue timeout > kicks-in restarting the device. > > To address this, introduce a tx cleanup mechanism as part of the > periodic maintenance process. This looks to just be a work-around that happens to work (for now) and the actual race condition should be fixed. Thanks, Tom > > Fixes: c5aa9e3b8156 ("amd-xgbe: Initial AMD 10GbE platform driver") > Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com> > Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com> > --- > drivers/net/ethernet/amd/xgbe/xgbe-drv.c | 14 ++++++++++++++ > 1 file changed, 14 insertions(+) > > diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c > index 614c0278419b..6b73648b3779 100644 > --- a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c > +++ b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c > @@ -682,10 +682,24 @@ static void xgbe_service(struct work_struct *work) > static void xgbe_service_timer(struct timer_list *t) > { > struct xgbe_prv_data *pdata = from_timer(pdata, t, service_timer); > + struct xgbe_channel *channel; > + unsigned int i; > > queue_work(pdata->dev_workqueue, &pdata->service_work); > > mod_timer(&pdata->service_timer, jiffies + HZ); > + > + if (!pdata->tx_usecs) > + return; > + > + for (i = 0; i < pdata->channel_count; i++) { > + channel = pdata->channel[i]; > + if (!channel->tx_ring || channel->tx_timer_active) > + break; > + channel->tx_timer_active = 1; > + mod_timer(&channel->tx_timer, > + jiffies + usecs_to_jiffies(pdata->tx_usecs)); > + } > } > > static void xgbe_init_timers(struct xgbe_prv_data *pdata)
diff --git a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c index 614c0278419b..6b73648b3779 100644 --- a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c +++ b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c @@ -682,10 +682,24 @@ static void xgbe_service(struct work_struct *work) static void xgbe_service_timer(struct timer_list *t) { struct xgbe_prv_data *pdata = from_timer(pdata, t, service_timer); + struct xgbe_channel *channel; + unsigned int i; queue_work(pdata->dev_workqueue, &pdata->service_work); mod_timer(&pdata->service_timer, jiffies + HZ); + + if (!pdata->tx_usecs) + return; + + for (i = 0; i < pdata->channel_count; i++) { + channel = pdata->channel[i]; + if (!channel->tx_ring || channel->tx_timer_active) + break; + channel->tx_timer_active = 1; + mod_timer(&channel->tx_timer, + jiffies + usecs_to_jiffies(pdata->tx_usecs)); + } } static void xgbe_init_timers(struct xgbe_prv_data *pdata)
The existing implementation uses software logic to accumulate tx completions until the specified time (1ms) is met and then poll them. However, there exists a tiny gap which leads to a race between resetting and checking the tx_activate flag. Due to this the tx completions are not reported to upper layer and tx queue timeout kicks-in restarting the device. To address this, introduce a tx cleanup mechanism as part of the periodic maintenance process. Fixes: c5aa9e3b8156 ("amd-xgbe: Initial AMD 10GbE platform driver") Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com> Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com> --- drivers/net/ethernet/amd/xgbe/xgbe-drv.c | 14 ++++++++++++++ 1 file changed, 14 insertions(+)