drivers/net: fix segfault in secondary process
Checks
Commit Message
Calling rte_eth_dev_info_get() on secondary process cause a crash
because eth_dev->device is not set properly.
Fixes: ee27edbe0c10 ("drivers/net: share vdev data to secondary process")
Cc: stable@dpdk.org
Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
---
Cc: Vipin Varghese <vipin.varghese@intel.com>
---
drivers/net/af_packet/rte_eth_af_packet.c | 1 +
drivers/net/bonding/rte_eth_bond_pmd.c | 1 +
drivers/net/failsafe/failsafe.c | 1 +
drivers/net/kni/rte_eth_kni.c | 1 +
drivers/net/null/rte_eth_null.c | 1 +
drivers/net/octeontx/octeontx_ethdev.c | 1 +
drivers/net/pcap/rte_eth_pcap.c | 1 +
drivers/net/tap/rte_eth_tap.c | 2 ++
drivers/net/vhost/rte_eth_vhost.c | 1 +
9 files changed, 10 insertions(+)
Comments
On Thu, 19 Jul 2018 17:45:56 +0100
Ferruh Yigit <ferruh.yigit@intel.com> wrote:
> Calling rte_eth_dev_info_get() on secondary process cause a crash
> because eth_dev->device is not set properly.
>
> Fixes: ee27edbe0c10 ("drivers/net: share vdev data to secondary process")
> Cc: stable@dpdk.org
>
> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Good catch.
Maybe eth_dev should do this for the device drivers?
Better to make device drivers as simple and safe as possible.
There seem to be a lot of bugs related to secondary process model.
Do we have a test suite for that.
> -----Original Message-----
> From: dev [mailto:dev-bounces@dpdk.org] On Behalf Of Ferruh Yigit
> Sent: Friday, July 20, 2018 12:46 AM
> To: John W. Linville <linville@tuxdriver.com>; Doherty, Declan
> <declan.doherty@intel.com>; Chas Williams <chas3@att.com>; Gaetan Rivet
> <gaetan.rivet@6wind.com>; Tetsuya Mukawa <mtetsuyah@gmail.com>;
> Santosh Shukla <santosh.shukla@caviumnetworks.com>; Jerin Jacob
> <jerin.jacob@caviumnetworks.com>; Wiles, Keith <keith.wiles@intel.com>;
> Maxime Coquelin <maxime.coquelin@redhat.com>; Bie, Tiwei
> <tiwei.bie@intel.com>; Wang, Zhihong <zhihong.wang@intel.com>
> Cc: dev@dpdk.org; Yigit, Ferruh <ferruh.yigit@intel.com>; stable@dpdk.org;
> Varghese, Vipin <vipin.varghese@intel.com>
> Subject: [dpdk-dev] [PATCH] drivers/net: fix segfault in secondary process
>
> Calling rte_eth_dev_info_get() on secondary process cause a crash because
> eth_dev->device is not set properly.
>
> Fixes: ee27edbe0c10 ("drivers/net: share vdev data to secondary process")
> Cc: stable@dpdk.org
>
> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
Yes, this is the same issue I met when detaching the device from a secondary process.
The backend device should be assigned on secondary.
Reviewed-by: Qi Zhang <qi.z.zhang@intel.com>
On 7/19/2018 5:32 PM, Stephen Hemminger wrote:
> On Thu, 19 Jul 2018 17:45:56 +0100
> Ferruh Yigit <ferruh.yigit@intel.com> wrote:
>
>> Calling rte_eth_dev_info_get() on secondary process cause a crash
>> because eth_dev->device is not set properly.
>>
>> Fixes: ee27edbe0c10 ("drivers/net: share vdev data to secondary process")
>> Cc: stable@dpdk.org
>>
>> Signed-off-by: Ferruh Yigit <ferruh.yigit@intel.com>
>
> Good catch.
>
> Maybe eth_dev should do this for the device drivers?
> Better to make device drivers as simple and safe as possible.
Agreed, let me try to create a helper function for this.
>
> There seem to be a lot of bugs related to secondary process model.
> Do we have a test suite for that.
>
@@ -936,6 +936,7 @@ rte_pmd_af_packet_probe(struct rte_vdev_device *dev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = &ops;
+ eth_dev->device = &dev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}
@@ -3184,6 +3184,7 @@ bond_probe(struct rte_vdev_device *dev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = &default_dev_ops;
+ eth_dev->device = &dev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}
@@ -328,6 +328,7 @@ rte_pmd_failsafe_probe(struct rte_vdev_device *vdev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = &failsafe_ops;
+ eth_dev->device = &vdev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}
@@ -420,6 +420,7 @@ eth_kni_probe(struct rte_vdev_device *vdev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = ð_kni_ops;
+ eth_dev->device = &vdev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}
@@ -624,6 +624,7 @@ rte_pmd_null_probe(struct rte_vdev_device *dev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = &ops;
+ eth_dev->device = &dev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}
@@ -1183,6 +1183,7 @@ octeontx_probe(struct rte_vdev_device *dev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = &octeontx_dev_ops;
+ eth_dev->device = &dev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}
@@ -1017,6 +1017,7 @@ pmd_pcap_probe(struct rte_vdev_device *dev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = &ops;
+ eth_dev->device = &dev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}
@@ -1931,6 +1931,7 @@ rte_pmd_tun_probe(struct rte_vdev_device *dev)
return -1;
}
eth_dev->dev_ops = &ops;
+ eth_dev->device = &dev->device;
return 0;
}
@@ -2000,6 +2001,7 @@ rte_pmd_tap_probe(struct rte_vdev_device *dev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = &ops;
+ eth_dev->device = &dev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}
@@ -1354,6 +1354,7 @@ rte_pmd_vhost_probe(struct rte_vdev_device *dev)
}
/* TODO: request info from primary to set up Rx and Tx */
eth_dev->dev_ops = &ops;
+ eth_dev->device = &dev->device;
rte_eth_dev_probing_finish(eth_dev);
return 0;
}