ice: Add netif_device_attach/detach into PF reset flow
Subsystems (3)
PATCH v4 1/1 ice: Add netif_device_attach/detach into PF reset flow
Ethtool callbacks can be executed while reset is in progress and try to access deleted resources, e.g. getting coalesce settings can result in a NULL pointer dereference seen below.
Reproduction steps:
Once the driver is fully initialized, trigger reset:
# echo 1 > /sys/class/net/<interface>/device/reset when reset is in progress try to get coalesce settings using ethtool: # ethtool -c <interface>
BUG: kernel NULL pointer dereference, address: 0000000000000020 PGD 0 P4D 0 Oops: Oops: 0000 [#1] PREEMPT SMP PTI CPU: 11 PID: 19713 Comm: ethtool Tainted: G S 6.10.0-rc7+ #7 RIP: 0010:ice_get_q_coalesce+0x2e/0xa0 [ice] RSP: 0018:ffffbab1e9bcf6a8 EFLAGS: 00010206 RAX: 000000000000000c RBX: ffff94512305b028 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffff9451c3f2e588 RDI: ffff9451c3f2e588 RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 R10: ffff9451c3f2e580 R11: 000000000000001f R12: ffff945121fa9000 R13: ffffbab1e9bcf760 R14: 0000000000000013 R15: ffffffff9e65dd40 FS: 00007faee5fbe740(0000) GS:ffff94546fd80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000020 CR3: 0000000106c2e005 CR4: 00000000001706f0 Call Trace: <TASK> ice_get_coalesce+0x17/0x30 [ice] coalesce_prepare_data+0x61/0x80 ethnl_default_doit+0xde/0x340 genl_family_rcv_msg_doit+0xf2/0x150 genl_rcv_msg+0x1b3/0x2c0 netlink_rcv_skb+0x5b/0x110 genl_rcv+0x28/0x40 netlink_unicast+0x19c/0x290 netlink_sendmsg+0x222/0x490 __sys_sendto+0x1df/0x1f0 __x64_sys_sendto+0x24/0x30 do_syscall_64+0x82/0x160 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7faee60d8e27
Calling netif_device_detach() before reset makes the net core not call the driver when ethtool command is issued, the attempt to execute an ethtool command during reset will result in the following message:
netlink error: No such device
instead of NULL pointer dereference. Once reset is done and ice_rebuild() is executing, the netif_device_attach() is called to allow for ethtool operations to occur again in a safe manner.
---
Changes since v1:
* Changed Fixes tag to point to another commit
* Minified the stacktrace
Changes since v2:
* Moved netif_device_attach() directly into ice_rebuild() and perform it only on main vsi
Changes since v3:
* Style changes requested by Przemek Kitszel
Suggestion from Kuba: https://lore.kernel.org/netdev/[email protected]/
Previous attempt (dropped because it introduced regression with link up): https://lore.kernel.org/netdev/[email protected]/
--- drivers/net/ethernet/intel/ice/ice_main.c | 7 +++++++ 1 file changed, 7 insertions(+)
drivers/net/ethernet/intel/ice/ice_main.c +7 -0
@@ -608,6 +608,9 @@ ice_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)memset (&vsi->mqprio_qopt,0 ,sizeof (vsi->mqprio_qopt)); } } + +if (vsi->netdev) +netif_device_detach (vsi->netdev); skip:/* clear SW filtering DB */
@@ -7591,6 +7594,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) {struct device *dev =ice_pf_to_dev (pf);struct ice_hw *hw = &pf->hw; +struct ice_vsi *vsi =ice_get_main_vsi (pf);bool dvm;int err;
@@ -7731,6 +7735,9 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)ice_rebuild_arfs (pf); } +if (vsi && vsi->netdev) +netif_device_attach (vsi->netdev); +ice_update_pf_netdev_link (pf);/* tell the firmware we are up */
Signature · 1 line
2.44.0
Quoted 88 lines · Dawid Osuchowski
Ethtool callbacks can be executed while reset is in progress and try to access deleted resources, e.g. getting coalesce settings can result in a NULL pointer dereference seen below.
Reproduction steps:
Once the driver is fully initialized, trigger reset:# echo 1 > /sys/class/net/<interface>/device/reset when reset is in progress try to get coalesce settings using ethtool: # ethtool -c <interface>BUG: kernel NULL pointer dereference, address: 0000000000000020 PGD 0 P4D 0 Oops: Oops: 0000 [#1] PREEMPT SMP PTI CPU: 11 PID: 19713 Comm: ethtool Tainted: G S 6.10.0-rc7+ #7 RIP: 0010:ice_get_q_coalesce+0x2e/0xa0 [ice] RSP: 0018:ffffbab1e9bcf6a8 EFLAGS: 00010206 RAX: 000000000000000c RBX: ffff94512305b028 RCX: 0000000000000000 RDX: 0000000000000000 RSI: ffff9451c3f2e588 RDI: ffff9451c3f2e588 RBP: 0000000000000000 R08: 0000000000000000 R09: 0000000000000000 R10: ffff9451c3f2e580 R11: 000000000000001f R12: ffff945121fa9000 R13: ffffbab1e9bcf760 R14: 0000000000000013 R15: ffffffff9e65dd40 FS: 00007faee5fbe740(0000) GS:ffff94546fd80000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 0000000000000020 CR3: 0000000106c2e005 CR4: 00000000001706f0 Call Trace: <TASK> ice_get_coalesce+0x17/0x30 [ice] coalesce_prepare_data+0x61/0x80 ethnl_default_doit+0xde/0x340 genl_family_rcv_msg_doit+0xf2/0x150 genl_rcv_msg+0x1b3/0x2c0 netlink_rcv_skb+0x5b/0x110 genl_rcv+0x28/0x40 netlink_unicast+0x19c/0x290 netlink_sendmsg+0x222/0x490 __sys_sendto+0x1df/0x1f0 __x64_sys_sendto+0x24/0x30 do_syscall_64+0x82/0x160 entry_SYSCALL_64_after_hwframe+0x76/0x7e RIP: 0033:0x7faee60d8e27Calling
netif_device_detach()before reset makes the net core not call the driver when ethtool command is issued, the attempt to execute an ethtool command during reset will result in the following message:netlink error: No such device
instead of NULL pointer dereference. Once reset is done and
ice_rebuild()is executing, thenetif_device_attach()is called to allow for ethtool operations to occur again in a safe manner.Fixes fcea6f3da546 ("ice: Add stats and ethtool support")Suggested-by Jakub Kicinski <[email protected]>Reviewed-by Igor Bagnucki <[email protected]>Signed-off-by Dawid Osuchowski <[email protected]>---
Changes since v1:
* Changed Fixes tag to point to another commit
* Minified the stacktraceChanges since v2:
* Movednetif_device_attach()directly intoice_rebuild()and perform it only on main vsiChanges since v3:
* Style changes requested by Przemek KitszelSuggestion from Kuba: https://lore.kernel.org/netdev/[email protected]/
Previous attempt (dropped because it introduced regression with link up): https://lore.kernel.org/netdev/[email protected]/--- drivers/net/ethernet/intel/ice/ice_main.c | 7 +++++++ 1 file changed, 7 insertions(+)drivers/net/ethernet/intel/ice/ice_main.c +3
-0@@ -608,6 +608,9 @@ ice_prepare_for_reset(struct ice_pf *pf, enum ice_reset_req reset_type)memset (&vsi->mqprio_qopt,0 ,sizeof (vsi->mqprio_qopt)); } } + +if (vsi->netdev) +netif_device_detach (vsi->netdev); skip:/* clear SW filtering DB */
drivers/net/ethernet/intel/ice/ice_main.c static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) · line 7597
Quoted 4 lines
drivers/net/ethernet/intel/ice/ice_main.c +1
-0@@ -7591,6 +7594,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) {struct device *dev =ice_pf_to_dev (pf);struct ice_hw *hw = &pf->hw; +struct ice_vsi *vsi =ice_get_main_vsi (pf);
we have an unwritten rule that is called 'reverse christmas tree' which requires us to have declarations of variables sorted from longest to shortest.
struct ice_vsi *vsi = ice_get_main_vsi(pf); struct device *dev = ice_pf_to_dev(pf); struct ice_hw *hw = &pf->hw; bool dvm; int err;
Quoted 16 lines
drivers/net/ethernet/intel/ice/ice_main.c +3
-0bool dvm;int err;@@ -7731,6 +7735,9 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type)ice_rebuild_arfs (pf); } +if (vsi && vsi->netdev) +netif_device_attach (vsi->netdev); +ice_update_pf_netdev_link (pf);/* tell the firmware we are up */ Signature · 1 line
2.44.0
Quoted 11 lines · Maciej Fijalkowski
drivers/net/ethernet/intel/ice/ice_main.c static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) · line 7597
Quoted 5 lines · Dawid Osuchowski
drivers/net/ethernet/intel/ice/ice_main.c +1
-0@@ -7591,6 +7594,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) {struct device *dev =ice_pf_to_dev (pf);struct ice_hw *hw = &pf->hw; +struct ice_vsi *vsi =ice_get_main_vsi (pf);we have an unwritten rule that is called 'reverse christmas tree' which requires us to have declarations of variables sorted from longest to shortest.
Quoted 6 lines
struct ice_vsi *vsi = ice_get_main_vsi(pf); struct device *dev = ice_pf_to_dev(pf); struct ice_hw *hw = &pf->hw; bool dvm; int err;
My apologies, I was not aware of that unwritten rule. Will fix in next revision, thanks. Other than that, does the rest of the changes look okay to you?
--Dawid
Przemek Kitszel
maintainerQuoted 11 lines · Maciej Fijalkowski
drivers/net/ethernet/intel/ice/ice_main.c static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) · line 7597
Quoted 5 lines · Dawid Osuchowski
drivers/net/ethernet/intel/ice/ice_main.c +1
-0@@ -7591,6 +7594,7 @@ static void ice_rebuild(struct ice_pf *pf, enum ice_reset_req reset_type) {struct device *dev =ice_pf_to_dev (pf);struct ice_hw *hw = &pf->hw; +struct ice_vsi *vsi =ice_get_main_vsi (pf);we have an unwritten rule that is called 'reverse christmas tree' which requires us to have declarations of variables sorted from longest to shortest.
Hah, the rule is written down actually :P
https://docs.kernel.org/process/maintainer-netdev.html