From 5f4fa1672d98fe99d2297b03add35346f1685d6b Mon Sep 17 00:00:00 2001 From: Ding Hui Date: Tue, 9 May 2023 19:11:47 +0800 Subject: [PATCH 1/8] iavf: Fix use-after-free in free_netdev We do netif_napi_add() for all allocated q_vectors[], but potentially do netif_napi_del() for part of them, then kfree q_vectors and leave invalid pointers at dev->napi_list. Reproducer: [root@host ~]# cat repro.sh #!/bin/bash pf_dbsf="0000:41:00.0" vf0_dbsf="0000:41:02.0" g_pids=() function do_set_numvf() { echo 2 >/sys/bus/pci/devices/${pf_dbsf}/sriov_numvfs sleep $((RANDOM%3+1)) echo 0 >/sys/bus/pci/devices/${pf_dbsf}/sriov_numvfs sleep $((RANDOM%3+1)) } function do_set_channel() { local nic=$(ls -1 --indicator-style=none /sys/bus/pci/devices/${vf0_dbsf}/net/) [ -z "$nic" ] && { sleep $((RANDOM%3)) ; return 1; } ifconfig $nic 192.168.18.5 netmask 255.255.255.0 ifconfig $nic up ethtool -L $nic combined 1 ethtool -L $nic combined 4 sleep $((RANDOM%3)) } function on_exit() { local pid for pid in "${g_pids[@]}"; do kill -0 "$pid" &>/dev/null && kill "$pid" &>/dev/null done g_pids=() } trap "on_exit; exit" EXIT while :; do do_set_numvf ; done & g_pids+=($!) while :; do do_set_channel ; done & g_pids+=($!) wait Result: [ 4093.900222] ================================================================== [ 4093.900230] BUG: KASAN: use-after-free in free_netdev+0x308/0x390 [ 4093.900232] Read of size 8 at addr ffff88b4dc145640 by task repro.sh/6699 [ 4093.900233] [ 4093.900236] CPU: 10 PID: 6699 Comm: repro.sh Kdump: loaded Tainted: G O --------- -t - 4.18.0 #1 [ 4093.900238] Hardware name: Powerleader PR2008AL/H12DSi-N6, BIOS 2.0 04/09/2021 [ 4093.900239] Call Trace: [ 4093.900244] dump_stack+0x71/0xab [ 4093.900249] print_address_description+0x6b/0x290 [ 4093.900251] ? free_netdev+0x308/0x390 [ 4093.900252] kasan_report+0x14a/0x2b0 [ 4093.900254] free_netdev+0x308/0x390 [ 4093.900261] iavf_remove+0x825/0xd20 [iavf] [ 4093.900265] pci_device_remove+0xa8/0x1f0 [ 4093.900268] device_release_driver_internal+0x1c6/0x460 [ 4093.900271] pci_stop_bus_device+0x101/0x150 [ 4093.900273] pci_stop_and_remove_bus_device+0xe/0x20 [ 4093.900275] pci_iov_remove_virtfn+0x187/0x420 [ 4093.900277] ? pci_iov_add_virtfn+0xe10/0xe10 [ 4093.900278] ? pci_get_subsys+0x90/0x90 [ 4093.900280] sriov_disable+0xed/0x3e0 [ 4093.900282] ? bus_find_device+0x12d/0x1a0 [ 4093.900290] i40e_free_vfs+0x754/0x1210 [i40e] [ 4093.900298] ? i40e_reset_all_vfs+0x880/0x880 [i40e] [ 4093.900299] ? pci_get_device+0x7c/0x90 [ 4093.900300] ? pci_get_subsys+0x90/0x90 [ 4093.900306] ? pci_vfs_assigned.part.7+0x144/0x210 [ 4093.900309] ? __mutex_lock_slowpath+0x10/0x10 [ 4093.900315] i40e_pci_sriov_configure+0x1fa/0x2e0 [i40e] [ 4093.900318] sriov_numvfs_store+0x214/0x290 [ 4093.900320] ? sriov_totalvfs_show+0x30/0x30 [ 4093.900321] ? __mutex_lock_slowpath+0x10/0x10 [ 4093.900323] ? __check_object_size+0x15a/0x350 [ 4093.900326] kernfs_fop_write+0x280/0x3f0 [ 4093.900329] vfs_write+0x145/0x440 [ 4093.900330] ksys_write+0xab/0x160 [ 4093.900332] ? __ia32_sys_read+0xb0/0xb0 [ 4093.900334] ? fput_many+0x1a/0x120 [ 4093.900335] ? filp_close+0xf0/0x130 [ 4093.900338] do_syscall_64+0xa0/0x370 [ 4093.900339] ? page_fault+0x8/0x30 [ 4093.900341] entry_SYSCALL_64_after_hwframe+0x65/0xca [ 4093.900357] RIP: 0033:0x7f16ad4d22c0 [ 4093.900359] Code: 73 01 c3 48 8b 0d d8 cb 2c 00 f7 d8 64 89 01 48 83 c8 ff c3 66 0f 1f 44 00 00 83 3d 89 24 2d 00 00 75 10 b8 01 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 31 c3 48 83 ec 08 e8 fe dd 01 00 48 89 04 24 [ 4093.900360] RSP: 002b:00007ffd6491b7f8 EFLAGS: 00000246 ORIG_RAX: 0000000000000001 [ 4093.900362] RAX: ffffffffffffffda RBX: 0000000000000002 RCX: 00007f16ad4d22c0 [ 4093.900363] RDX: 0000000000000002 RSI: 0000000001a41408 RDI: 0000000000000001 [ 4093.900364] RBP: 0000000001a41408 R08: 00007f16ad7a1780 R09: 00007f16ae1f2700 [ 4093.900364] R10: 0000000000000001 R11: 0000000000000246 R12: 0000000000000002 [ 4093.900365] R13: 0000000000000001 R14: 00007f16ad7a0620 R15: 0000000000000001 [ 4093.900367] [ 4093.900368] Allocated by task 820: [ 4093.900371] kasan_kmalloc+0xa6/0xd0 [ 4093.900373] __kmalloc+0xfb/0x200 [ 4093.900376] iavf_init_interrupt_scheme+0x63b/0x1320 [iavf] [ 4093.900380] iavf_watchdog_task+0x3d51/0x52c0 [iavf] [ 4093.900382] process_one_work+0x56a/0x11f0 [ 4093.900383] worker_thread+0x8f/0xf40 [ 4093.900384] kthread+0x2a0/0x390 [ 4093.900385] ret_from_fork+0x1f/0x40 [ 4093.900387] 0xffffffffffffffff [ 4093.900387] [ 4093.900388] Freed by task 6699: [ 4093.900390] __kasan_slab_free+0x137/0x190 [ 4093.900391] kfree+0x8b/0x1b0 [ 4093.900394] iavf_free_q_vectors+0x11d/0x1a0 [iavf] [ 4093.900397] iavf_remove+0x35a/0xd20 [iavf] [ 4093.900399] pci_device_remove+0xa8/0x1f0 [ 4093.900400] device_release_driver_internal+0x1c6/0x460 [ 4093.900401] pci_stop_bus_device+0x101/0x150 [ 4093.900402] pci_stop_and_remove_bus_device+0xe/0x20 [ 4093.900403] pci_iov_remove_virtfn+0x187/0x420 [ 4093.900404] sriov_disable+0xed/0x3e0 [ 4093.900409] i40e_free_vfs+0x754/0x1210 [i40e] [ 4093.900415] i40e_pci_sriov_configure+0x1fa/0x2e0 [i40e] [ 4093.900416] sriov_numvfs_store+0x214/0x290 [ 4093.900417] kernfs_fop_write+0x280/0x3f0 [ 4093.900418] vfs_write+0x145/0x440 [ 4093.900419] ksys_write+0xab/0x160 [ 4093.900420] do_syscall_64+0xa0/0x370 [ 4093.900421] entry_SYSCALL_64_after_hwframe+0x65/0xca [ 4093.900422] 0xffffffffffffffff [ 4093.900422] [ 4093.900424] The buggy address belongs to the object at ffff88b4dc144200 which belongs to the cache kmalloc-8k of size 8192 [ 4093.900425] The buggy address is located 5184 bytes inside of 8192-byte region [ffff88b4dc144200, ffff88b4dc146200) [ 4093.900425] The buggy address belongs to the page: [ 4093.900427] page:ffffea00d3705000 refcount:1 mapcount:0 mapping:ffff88bf04415c80 index:0x0 compound_mapcount: 0 [ 4093.900430] flags: 0x10000000008100(slab|head) [ 4093.900433] raw: 0010000000008100 dead000000000100 dead000000000200 ffff88bf04415c80 [ 4093.900434] raw: 0000000000000000 0000000000030003 00000001ffffffff 0000000000000000 [ 4093.900434] page dumped because: kasan: bad access detected [ 4093.900435] [ 4093.900435] Memory state around the buggy address: [ 4093.900436] ffff88b4dc145500: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 4093.900437] ffff88b4dc145580: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 4093.900438] >ffff88b4dc145600: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 4093.900438] ^ [ 4093.900439] ffff88b4dc145680: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 4093.900440] ffff88b4dc145700: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb [ 4093.900440] ================================================================== Although the patch #2 (of 2) can avoid the issue triggered by this repro.sh, there still are other potential risks that if num_active_queues is changed to less than allocated q_vectors[] by unexpected, the mismatched netif_napi_add/del() can also cause UAF. Since we actually call netif_napi_add() for all allocated q_vectors unconditionally in iavf_alloc_q_vectors(), so we should fix it by letting netif_napi_del() match to netif_napi_add(). Fixes: 5eae00c57f5e ("i40evf: main driver core") Signed-off-by: Ding Hui Cc: Donglin Peng Cc: Huang Cun Reviewed-by: Simon Horman Reviewed-by: Madhu Chittim Reviewed-by: Leon Romanovsky Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf_main.c | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index a483eb185c99..89d88c3a7add 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -1828,19 +1828,16 @@ static int iavf_alloc_q_vectors(struct iavf_adapter *adapter) static void iavf_free_q_vectors(struct iavf_adapter *adapter) { int q_idx, num_q_vectors; - int napi_vectors; if (!adapter->q_vectors) return; num_q_vectors = adapter->num_msix_vectors - NONQ_VECS; - napi_vectors = adapter->num_active_queues; for (q_idx = 0; q_idx < num_q_vectors; q_idx++) { struct iavf_q_vector *q_vector = &adapter->q_vectors[q_idx]; - if (q_idx < napi_vectors) - netif_napi_del(&q_vector->napi); + netif_napi_del(&q_vector->napi); } kfree(adapter->q_vectors); adapter->q_vectors = NULL; From 7c4bced3caa749ce468b0c5de711c98476b23a52 Mon Sep 17 00:00:00 2001 From: Ding Hui Date: Tue, 9 May 2023 19:11:48 +0800 Subject: [PATCH 2/8] iavf: Fix out-of-bounds when setting channels on remove If we set channels greater during iavf_remove(), and waiting reset done would be timeout, then returned with error but changed num_active_queues directly, that will lead to OOB like the following logs. Because the num_active_queues is greater than tx/rx_rings[] allocated actually. Reproducer: [root@host ~]# cat repro.sh #!/bin/bash pf_dbsf="0000:41:00.0" vf0_dbsf="0000:41:02.0" g_pids=() function do_set_numvf() { echo 2 >/sys/bus/pci/devices/${pf_dbsf}/sriov_numvfs sleep $((RANDOM%3+1)) echo 0 >/sys/bus/pci/devices/${pf_dbsf}/sriov_numvfs sleep $((RANDOM%3+1)) } function do_set_channel() { local nic=$(ls -1 --indicator-style=none /sys/bus/pci/devices/${vf0_dbsf}/net/) [ -z "$nic" ] && { sleep $((RANDOM%3)) ; return 1; } ifconfig $nic 192.168.18.5 netmask 255.255.255.0 ifconfig $nic up ethtool -L $nic combined 1 ethtool -L $nic combined 4 sleep $((RANDOM%3)) } function on_exit() { local pid for pid in "${g_pids[@]}"; do kill -0 "$pid" &>/dev/null && kill "$pid" &>/dev/null done g_pids=() } trap "on_exit; exit" EXIT while :; do do_set_numvf ; done & g_pids+=($!) while :; do do_set_channel ; done & g_pids+=($!) wait Result: [ 3506.152887] iavf 0000:41:02.0: Removing device [ 3510.400799] ================================================================== [ 3510.400820] BUG: KASAN: slab-out-of-bounds in iavf_free_all_tx_resources+0x156/0x160 [iavf] [ 3510.400823] Read of size 8 at addr ffff88b6f9311008 by task repro.sh/55536 [ 3510.400823] [ 3510.400830] CPU: 101 PID: 55536 Comm: repro.sh Kdump: loaded Tainted: G O --------- -t - 4.18.0 #1 [ 3510.400832] Hardware name: Powerleader PR2008AL/H12DSi-N6, BIOS 2.0 04/09/2021 [ 3510.400835] Call Trace: [ 3510.400851] dump_stack+0x71/0xab [ 3510.400860] print_address_description+0x6b/0x290 [ 3510.400865] ? iavf_free_all_tx_resources+0x156/0x160 [iavf] [ 3510.400868] kasan_report+0x14a/0x2b0 [ 3510.400873] iavf_free_all_tx_resources+0x156/0x160 [iavf] [ 3510.400880] iavf_remove+0x2b6/0xc70 [iavf] [ 3510.400884] ? iavf_free_all_rx_resources+0x160/0x160 [iavf] [ 3510.400891] ? wait_woken+0x1d0/0x1d0 [ 3510.400895] ? notifier_call_chain+0xc1/0x130 [ 3510.400903] pci_device_remove+0xa8/0x1f0 [ 3510.400910] device_release_driver_internal+0x1c6/0x460 [ 3510.400916] pci_stop_bus_device+0x101/0x150 [ 3510.400919] pci_stop_and_remove_bus_device+0xe/0x20 [ 3510.400924] pci_iov_remove_virtfn+0x187/0x420 [ 3510.400927] ? pci_iov_add_virtfn+0xe10/0xe10 [ 3510.400929] ? pci_get_subsys+0x90/0x90 [ 3510.400932] sriov_disable+0xed/0x3e0 [ 3510.400936] ? bus_find_device+0x12d/0x1a0 [ 3510.400953] i40e_free_vfs+0x754/0x1210 [i40e] [ 3510.400966] ? i40e_reset_all_vfs+0x880/0x880 [i40e] [ 3510.400968] ? pci_get_device+0x7c/0x90 [ 3510.400970] ? pci_get_subsys+0x90/0x90 [ 3510.400982] ? pci_vfs_assigned.part.7+0x144/0x210 [ 3510.400987] ? __mutex_lock_slowpath+0x10/0x10 [ 3510.400996] i40e_pci_sriov_configure+0x1fa/0x2e0 [i40e] [ 3510.401001] sriov_numvfs_store+0x214/0x290 [ 3510.401005] ? sriov_totalvfs_show+0x30/0x30 [ 3510.401007] ? __mutex_lock_slowpath+0x10/0x10 [ 3510.401011] ? __check_object_size+0x15a/0x350 [ 3510.401018] kernfs_fop_write+0x280/0x3f0 [ 3510.401022] vfs_write+0x145/0x440 [ 3510.401025] ksys_write+0xab/0x160 [ 3510.401028] ? __ia32_sys_read+0xb0/0xb0 [ 3510.401031] ? fput_many+0x1a/0x120 [ 3510.401032] ? filp_close+0xf0/0x130 [ 3510.401038] do_syscall_64+0xa0/0x370 [ 3510.401041] ? page_fault+0x8/0x30 [ 3510.401043] entry_SYSCALL_64_after_hwframe+0x65/0xca [ 3510.401073] RIP: 0033:0x7f3a9bb842c0 [ 3510.401079] Code: 73 01 c3 48 8b 0d d8 cb 2c 00 f7 d8 64 89 01 48 83 c8 ff c3 66 0f 1f 44 00 00 83 3d 89 24 2d 00 00 75 10 b8 01 00 00 00 0f 05 <48> 3d 01 f0 ff ff 73 31 c3 48 83 ec 08 e8 fe dd 01 00 48 89 04 24 [ 3510.401080] RSP: 002b:00007ffc05f1fe18 EFLAGS: 00000246 ORIG_RAX: 0000000000000001 [ 3510.401083] RAX: ffffffffffffffda RBX: 0000000000000002 RCX: 00007f3a9bb842c0 [ 3510.401085] RDX: 0000000000000002 RSI: 0000000002327408 RDI: 0000000000000001 [ 3510.401086] RBP: 0000000002327408 R08: 00007f3a9be53780 R09: 00007f3a9c8a4700 [ 3510.401086] R10: 0000000000000001 R11: 0000000000000246 R12: 0000000000000002 [ 3510.401087] R13: 0000000000000001 R14: 00007f3a9be52620 R15: 0000000000000001 [ 3510.401090] [ 3510.401093] Allocated by task 76795: [ 3510.401098] kasan_kmalloc+0xa6/0xd0 [ 3510.401099] __kmalloc+0xfb/0x200 [ 3510.401104] iavf_init_interrupt_scheme+0x26f/0x1310 [iavf] [ 3510.401108] iavf_watchdog_task+0x1d58/0x4050 [iavf] [ 3510.401114] process_one_work+0x56a/0x11f0 [ 3510.401115] worker_thread+0x8f/0xf40 [ 3510.401117] kthread+0x2a0/0x390 [ 3510.401119] ret_from_fork+0x1f/0x40 [ 3510.401122] 0xffffffffffffffff [ 3510.401123] In timeout handling, we should keep the original num_active_queues and reset num_req_queues to 0. Fixes: 4e5e6b5d9d13 ("iavf: Fix return of set the new channel count") Signed-off-by: Ding Hui Cc: Donglin Peng Cc: Huang Cun Reviewed-by: Leon Romanovsky Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf_ethtool.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/drivers/net/ethernet/intel/iavf/iavf_ethtool.c b/drivers/net/ethernet/intel/iavf/iavf_ethtool.c index 6f171d1d85b7..92443f8e9fbd 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_ethtool.c +++ b/drivers/net/ethernet/intel/iavf/iavf_ethtool.c @@ -1863,7 +1863,7 @@ static int iavf_set_channels(struct net_device *netdev, } if (i == IAVF_RESET_WAIT_COMPLETE_COUNT) { adapter->flags &= ~IAVF_FLAG_REINIT_ITR_NEEDED; - adapter->num_active_queues = num_req; + adapter->num_req_queues = 0; return -EOPNOTSUPP; } From a77ed5c5b768e9649be240a2d864e5cd9c6a2015 Mon Sep 17 00:00:00 2001 From: Ahmed Zaki Date: Fri, 19 May 2023 15:46:02 -0600 Subject: [PATCH 3/8] iavf: use internal state to free traffic IRQs If the system tries to close the netdev while iavf_reset_task() is running, __LINK_STATE_START will be cleared and netif_running() will return false in iavf_reinit_interrupt_scheme(). This will result in iavf_free_traffic_irqs() not being called and a leak as follows: [7632.489326] remove_proc_entry: removing non-empty directory 'irq/999', leaking at least 'iavf-enp24s0f0v0-TxRx-0' [7632.490214] WARNING: CPU: 0 PID: 10 at fs/proc/generic.c:718 remove_proc_entry+0x19b/0x1b0 is shown when pci_disable_msix() is later called. Fix by using the internal adapter state. The traffic IRQs will always exist if state == __IAVF_RUNNING. Fixes: 5b36e8d04b44 ("i40evf: Enable VF to request an alternate queue allocation") Signed-off-by: Ahmed Zaki Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf_main.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-) diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index 89d88c3a7add..058e19c6f94a 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -1929,15 +1929,16 @@ static void iavf_free_rss(struct iavf_adapter *adapter) /** * iavf_reinit_interrupt_scheme - Reallocate queues and vectors * @adapter: board private structure + * @running: true if adapter->state == __IAVF_RUNNING * * Returns 0 on success, negative on failure **/ -static int iavf_reinit_interrupt_scheme(struct iavf_adapter *adapter) +static int iavf_reinit_interrupt_scheme(struct iavf_adapter *adapter, bool running) { struct net_device *netdev = adapter->netdev; int err; - if (netif_running(netdev)) + if (running) iavf_free_traffic_irqs(adapter); iavf_free_misc_irq(adapter); iavf_reset_interrupt_capability(adapter); @@ -3053,7 +3054,7 @@ static void iavf_reset_task(struct work_struct *work) if ((adapter->flags & IAVF_FLAG_REINIT_MSIX_NEEDED) || (adapter->flags & IAVF_FLAG_REINIT_ITR_NEEDED)) { - err = iavf_reinit_interrupt_scheme(adapter); + err = iavf_reinit_interrupt_scheme(adapter, running); if (err) goto reset_err; } From c2ed2403f12c74a74a0091ed5d830e72c58406e8 Mon Sep 17 00:00:00 2001 From: Marcin Szycik Date: Mon, 5 Jun 2023 10:52:22 -0400 Subject: [PATCH 4/8] iavf: Wait for reset in callbacks which trigger it There was a fail when trying to add the interface to bonding right after changing the MTU on the interface. It was caused by bonding interface unable to open the interface due to interface being in __RESETTING state because of MTU change. Add new reset_waitqueue to indicate that reset has finished. Add waiting for reset to finish in callbacks which trigger hw reset: iavf_set_priv_flags(), iavf_change_mtu() and iavf_set_ringparam(). We use a 5000ms timeout period because on Hyper-V based systems, this operation takes around 3000-4000ms. In normal circumstances, it doesn't take more than 500ms to complete. Add a function iavf_wait_for_reset() to reuse waiting for reset code and use it also in iavf_set_channels(), which already waits for reset. We don't use error handling in iavf_set_channels() as this could cause the device to be in incorrect state if the reset was scheduled but hit timeout or the waitng function was interrupted by a signal. Fixes: 4e5e6b5d9d13 ("iavf: Fix return of set the new channel count") Signed-off-by: Marcin Szycik Co-developed-by: Dawid Wesierski Signed-off-by: Dawid Wesierski Signed-off-by: Sylwester Dziedziuch Signed-off-by: Kamil Maziarz Signed-off-by: Mateusz Palczewski Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf.h | 2 + .../net/ethernet/intel/iavf/iavf_ethtool.c | 31 ++++++----- drivers/net/ethernet/intel/iavf/iavf_main.c | 51 ++++++++++++++++++- .../net/ethernet/intel/iavf/iavf_virtchnl.c | 1 + 4 files changed, 68 insertions(+), 17 deletions(-) diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h index f80f2735e688..a5cab19eb6a8 100644 --- a/drivers/net/ethernet/intel/iavf/iavf.h +++ b/drivers/net/ethernet/intel/iavf/iavf.h @@ -257,6 +257,7 @@ struct iavf_adapter { struct work_struct adminq_task; struct delayed_work client_task; wait_queue_head_t down_waitqueue; + wait_queue_head_t reset_waitqueue; wait_queue_head_t vc_waitqueue; struct iavf_q_vector *q_vectors; struct list_head vlan_filter_list; @@ -582,4 +583,5 @@ void iavf_add_adv_rss_cfg(struct iavf_adapter *adapter); void iavf_del_adv_rss_cfg(struct iavf_adapter *adapter); struct iavf_mac_filter *iavf_add_filter(struct iavf_adapter *adapter, const u8 *macaddr); +int iavf_wait_for_reset(struct iavf_adapter *adapter); #endif /* _IAVF_H_ */ diff --git a/drivers/net/ethernet/intel/iavf/iavf_ethtool.c b/drivers/net/ethernet/intel/iavf/iavf_ethtool.c index 92443f8e9fbd..b7141c2a941d 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_ethtool.c +++ b/drivers/net/ethernet/intel/iavf/iavf_ethtool.c @@ -484,6 +484,7 @@ static int iavf_set_priv_flags(struct net_device *netdev, u32 flags) { struct iavf_adapter *adapter = netdev_priv(netdev); u32 orig_flags, new_flags, changed_flags; + int ret = 0; u32 i; orig_flags = READ_ONCE(adapter->flags); @@ -533,10 +534,13 @@ static int iavf_set_priv_flags(struct net_device *netdev, u32 flags) if (netif_running(netdev)) { adapter->flags |= IAVF_FLAG_RESET_NEEDED; queue_work(adapter->wq, &adapter->reset_task); + ret = iavf_wait_for_reset(adapter); + if (ret) + netdev_warn(netdev, "Changing private flags timeout or interrupted waiting for reset"); } } - return 0; + return ret; } /** @@ -627,6 +631,7 @@ static int iavf_set_ringparam(struct net_device *netdev, { struct iavf_adapter *adapter = netdev_priv(netdev); u32 new_rx_count, new_tx_count; + int ret = 0; if ((ring->rx_mini_pending) || (ring->rx_jumbo_pending)) return -EINVAL; @@ -673,9 +678,12 @@ static int iavf_set_ringparam(struct net_device *netdev, if (netif_running(netdev)) { adapter->flags |= IAVF_FLAG_RESET_NEEDED; queue_work(adapter->wq, &adapter->reset_task); + ret = iavf_wait_for_reset(adapter); + if (ret) + netdev_warn(netdev, "Changing ring parameters timeout or interrupted waiting for reset"); } - return 0; + return ret; } /** @@ -1830,7 +1838,7 @@ static int iavf_set_channels(struct net_device *netdev, { struct iavf_adapter *adapter = netdev_priv(netdev); u32 num_req = ch->combined_count; - int i; + int ret = 0; if ((adapter->vf_res->vf_cap_flags & VIRTCHNL_VF_OFFLOAD_ADQ) && adapter->num_tc) { @@ -1854,20 +1862,11 @@ static int iavf_set_channels(struct net_device *netdev, adapter->flags |= IAVF_FLAG_REINIT_ITR_NEEDED; iavf_schedule_reset(adapter); - /* wait for the reset is done */ - for (i = 0; i < IAVF_RESET_WAIT_COMPLETE_COUNT; i++) { - msleep(IAVF_RESET_WAIT_MS); - if (adapter->flags & IAVF_FLAG_RESET_PENDING) - continue; - break; - } - if (i == IAVF_RESET_WAIT_COMPLETE_COUNT) { - adapter->flags &= ~IAVF_FLAG_REINIT_ITR_NEEDED; - adapter->num_req_queues = 0; - return -EOPNOTSUPP; - } + ret = iavf_wait_for_reset(adapter); + if (ret) + netdev_warn(netdev, "Changing channel count timeout or interrupted waiting for reset"); - return 0; + return ret; } /** diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index 058e19c6f94a..b89933aa5bfe 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -166,6 +166,45 @@ static struct iavf_adapter *iavf_pdev_to_adapter(struct pci_dev *pdev) return netdev_priv(pci_get_drvdata(pdev)); } +/** + * iavf_is_reset_in_progress - Check if a reset is in progress + * @adapter: board private structure + */ +static bool iavf_is_reset_in_progress(struct iavf_adapter *adapter) +{ + if (adapter->state == __IAVF_RESETTING || + adapter->flags & (IAVF_FLAG_RESET_PENDING | + IAVF_FLAG_RESET_NEEDED)) + return true; + + return false; +} + +/** + * iavf_wait_for_reset - Wait for reset to finish. + * @adapter: board private structure + * + * Returns 0 if reset finished successfully, negative on timeout or interrupt. + */ +int iavf_wait_for_reset(struct iavf_adapter *adapter) +{ + int ret = wait_event_interruptible_timeout(adapter->reset_waitqueue, + !iavf_is_reset_in_progress(adapter), + msecs_to_jiffies(5000)); + + /* If ret < 0 then it means wait was interrupted. + * If ret == 0 then it means we got a timeout while waiting + * for reset to finish. + * If ret > 0 it means reset has finished. + */ + if (ret > 0) + return 0; + else if (ret < 0) + return -EINTR; + else + return -EBUSY; +} + /** * iavf_allocate_dma_mem_d - OS specific memory alloc for shared code * @hw: pointer to the HW structure @@ -3149,6 +3188,7 @@ static void iavf_reset_task(struct work_struct *work) adapter->flags &= ~IAVF_FLAG_REINIT_ITR_NEEDED; + wake_up(&adapter->reset_waitqueue); mutex_unlock(&adapter->client_lock); mutex_unlock(&adapter->crit_lock); @@ -4313,6 +4353,7 @@ static int iavf_close(struct net_device *netdev) static int iavf_change_mtu(struct net_device *netdev, int new_mtu) { struct iavf_adapter *adapter = netdev_priv(netdev); + int ret = 0; netdev_dbg(netdev, "changing MTU from %d to %d\n", netdev->mtu, new_mtu); @@ -4325,9 +4366,14 @@ static int iavf_change_mtu(struct net_device *netdev, int new_mtu) if (netif_running(netdev)) { adapter->flags |= IAVF_FLAG_RESET_NEEDED; queue_work(adapter->wq, &adapter->reset_task); + ret = iavf_wait_for_reset(adapter); + if (ret < 0) + netdev_warn(netdev, "MTU change interrupted waiting for reset"); + else if (ret) + netdev_warn(netdev, "MTU change timed out waiting for reset"); } - return 0; + return ret; } #define NETIF_VLAN_OFFLOAD_FEATURES (NETIF_F_HW_VLAN_CTAG_RX | \ @@ -4928,6 +4974,9 @@ static int iavf_probe(struct pci_dev *pdev, const struct pci_device_id *ent) /* Setup the wait queue for indicating transition to down status */ init_waitqueue_head(&adapter->down_waitqueue); + /* Setup the wait queue for indicating transition to running state */ + init_waitqueue_head(&adapter->reset_waitqueue); + /* Setup the wait queue for indicating virtchannel events */ init_waitqueue_head(&adapter->vc_waitqueue); diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c index 7c0578b5457b..1bab896aaf40 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c +++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c @@ -2285,6 +2285,7 @@ void iavf_virtchnl_completion(struct iavf_adapter *adapter, case VIRTCHNL_OP_ENABLE_QUEUES: /* enable transmits */ iavf_irq_enable(adapter, true); + wake_up(&adapter->reset_waitqueue); adapter->flags &= ~IAVF_FLAG_QUEUES_DISABLED; break; case VIRTCHNL_OP_DISABLE_QUEUES: From d2806d960e8387945b53edf7ed9d71ab3ab5f073 Mon Sep 17 00:00:00 2001 From: Marcin Szycik Date: Mon, 5 Jun 2023 10:52:23 -0400 Subject: [PATCH 5/8] Revert "iavf: Detach device during reset task" This reverts commit aa626da947e9cd30c4cf727493903e1adbb2c0a0. Detaching device during reset was not fully fixing the rtnl locking issue, as there could be a situation where callback was already in progress before detaching netdev. Furthermore, detaching netdevice causes TX timeouts if traffic is running. To reproduce: ip netns exec ns1 iperf3 -c $PEER_IP -t 600 --logfile /dev/null & while :; do for i in 200 7000 400 5000 300 3000 ; do ip netns exec ns1 ip link set $VF1 mtu $i sleep 2 done sleep 10 done Currently, callbacks such as iavf_change_mtu() wait for the reset. If the reset fails to acquire the rtnl_lock, they schedule the netdev update for later while continuing the reset flow. Operations like MTU changes are performed under the rtnl_lock. Therefore, when the operation finishes, another callback that uses rtnl_lock can start. Signed-off-by: Dawid Wesierski Signed-off-by: Marcin Szycik Signed-off-by: Mateusz Palczewski Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf_main.c | 13 ++----------- 1 file changed, 2 insertions(+), 11 deletions(-) diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index b89933aa5bfe..957ad6e63f6f 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -2977,11 +2977,6 @@ static void iavf_reset_task(struct work_struct *work) int i = 0, err; bool running; - /* Detach interface to avoid subsequent NDO callbacks */ - rtnl_lock(); - netif_device_detach(netdev); - rtnl_unlock(); - /* When device is being removed it doesn't make sense to run the reset * task, just return in such a case. */ @@ -2989,7 +2984,7 @@ static void iavf_reset_task(struct work_struct *work) if (adapter->state != __IAVF_REMOVE) queue_work(adapter->wq, &adapter->reset_task); - goto reset_finish; + return; } while (!mutex_trylock(&adapter->client_lock)) @@ -3192,7 +3187,7 @@ static void iavf_reset_task(struct work_struct *work) mutex_unlock(&adapter->client_lock); mutex_unlock(&adapter->crit_lock); - goto reset_finish; + return; reset_err: if (running) { set_bit(__IAVF_VSI_DOWN, adapter->vsi.state); @@ -3213,10 +3208,6 @@ static void iavf_reset_task(struct work_struct *work) } dev_err(&adapter->pdev->dev, "failed to allocate resources during reinit\n"); -reset_finish: - rtnl_lock(); - netif_device_attach(netdev); - rtnl_unlock(); } /** From d916d273041b0b5652434ff27aec716ff90992ac Mon Sep 17 00:00:00 2001 From: Marcin Szycik Date: Mon, 5 Jun 2023 10:52:24 -0400 Subject: [PATCH 6/8] Revert "iavf: Do not restart Tx queues after reset task failure" This reverts commit 08f1c147b7265245d67321585c68a27e990e0c4b. Netdev is no longer being detached during reset, so this fix can be reverted. We leave the removal of "hacky" IFF_UP flag update. Signed-off-by: Marcin Szycik Signed-off-by: Mateusz Palczewski Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf_main.c | 15 --------------- 1 file changed, 15 deletions(-) diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index 957ad6e63f6f..0c12a752429c 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -3042,11 +3042,6 @@ static void iavf_reset_task(struct work_struct *work) iavf_disable_vf(adapter); mutex_unlock(&adapter->client_lock); mutex_unlock(&adapter->crit_lock); - if (netif_running(netdev)) { - rtnl_lock(); - dev_close(netdev); - rtnl_unlock(); - } return; /* Do not attempt to reinit. It's dead, Jim. */ } @@ -3197,16 +3192,6 @@ static void iavf_reset_task(struct work_struct *work) mutex_unlock(&adapter->client_lock); mutex_unlock(&adapter->crit_lock); - - if (netif_running(netdev)) { - /* Close device to ensure that Tx queues will not be started - * during netif_device_attach() at the end of the reset task. - */ - rtnl_lock(); - dev_close(netdev); - rtnl_unlock(); - } - dev_err(&adapter->pdev->dev, "failed to allocate resources during reinit\n"); } From d1639a17319ba78a018280cd2df6577a7e5d9fab Mon Sep 17 00:00:00 2001 From: Ahmed Zaki Date: Mon, 5 Jun 2023 10:52:25 -0400 Subject: [PATCH 7/8] iavf: fix a deadlock caused by rtnl and driver's lock circular dependencies A driver's lock (crit_lock) is used to serialize all the driver's tasks. Lockdep, however, shows a circular dependency between rtnl and crit_lock. This happens when an ndo that already holds the rtnl requests the driver to reset, since the reset task (in some paths) tries to grab rtnl to either change real number of queues of update netdev features. [566.241851] ====================================================== [566.241893] WARNING: possible circular locking dependency detected [566.241936] 6.2.14-100.fc36.x86_64+debug #1 Tainted: G OE [566.241984] ------------------------------------------------------ [566.242025] repro.sh/2604 is trying to acquire lock: [566.242061] ffff9280fc5ceee8 (&adapter->crit_lock){+.+.}-{3:3}, at: iavf_close+0x3c/0x240 [iavf] [566.242167] but task is already holding lock: [566.242209] ffffffff9976d350 (rtnl_mutex){+.+.}-{3:3}, at: iavf_remove+0x6b5/0x730 [iavf] [566.242300] which lock already depends on the new lock. [566.242353] the existing dependency chain (in reverse order) is: [566.242401] -> #1 (rtnl_mutex){+.+.}-{3:3}: [566.242451] __mutex_lock+0xc1/0xbb0 [566.242489] iavf_init_interrupt_scheme+0x179/0x440 [iavf] [566.242560] iavf_watchdog_task+0x80b/0x1400 [iavf] [566.242627] process_one_work+0x2b3/0x560 [566.242663] worker_thread+0x4f/0x3a0 [566.242696] kthread+0xf2/0x120 [566.242730] ret_from_fork+0x29/0x50 [566.242763] -> #0 (&adapter->crit_lock){+.+.}-{3:3}: [566.242815] __lock_acquire+0x15ff/0x22b0 [566.242869] lock_acquire+0xd2/0x2c0 [566.242901] __mutex_lock+0xc1/0xbb0 [566.242934] iavf_close+0x3c/0x240 [iavf] [566.242997] __dev_close_many+0xac/0x120 [566.243036] dev_close_many+0x8b/0x140 [566.243071] unregister_netdevice_many_notify+0x165/0x7c0 [566.243116] unregister_netdevice_queue+0xd3/0x110 [566.243157] iavf_remove+0x6c1/0x730 [iavf] [566.243217] pci_device_remove+0x33/0xa0 [566.243257] device_release_driver_internal+0x1bc/0x240 [566.243299] pci_stop_bus_device+0x6c/0x90 [566.243338] pci_stop_and_remove_bus_device+0xe/0x20 [566.243380] pci_iov_remove_virtfn+0xd1/0x130 [566.243417] sriov_disable+0x34/0xe0 [566.243448] ice_free_vfs+0x2da/0x330 [ice] [566.244383] ice_sriov_configure+0x88/0xad0 [ice] [566.245353] sriov_numvfs_store+0xde/0x1d0 [566.246156] kernfs_fop_write_iter+0x15e/0x210 [566.246921] vfs_write+0x288/0x530 [566.247671] ksys_write+0x74/0xf0 [566.248408] do_syscall_64+0x58/0x80 [566.249145] entry_SYSCALL_64_after_hwframe+0x72/0xdc [566.249886] other info that might help us debug this: [566.252014] Possible unsafe locking scenario: [566.253432] CPU0 CPU1 [566.254118] ---- ---- [566.254800] lock(rtnl_mutex); [566.255514] lock(&adapter->crit_lock); [566.256233] lock(rtnl_mutex); [566.256897] lock(&adapter->crit_lock); [566.257388] *** DEADLOCK *** The deadlock can be triggered by a script that is continuously resetting the VF adapter while doing other operations requiring RTNL, e.g: while :; do ip link set $VF up ethtool --set-channels $VF combined 2 ip link set $VF down ip link set $VF up ethtool --set-channels $VF combined 4 ip link set $VF down done Any operation that triggers a reset can substitute "ethtool --set-channles" As a fix, add a new task "finish_config" that do all the work which needs rtnl lock. With the exception of iavf_remove(), all work that require rtnl should be called from this task. As for iavf_remove(), at the point where we need to call unregister_netdevice() (and grab rtnl_lock), we make sure the finish_config task is not running (cancel_work_sync()) to safely grab rtnl. Subsequent finish_config work cannot restart after that since the task is guarded by the __IAVF_IN_REMOVE_TASK bit in iavf_schedule_finish_config(). Fixes: 5ac49f3c2702 ("iavf: use mutexes for locking of critical sections") Signed-off-by: Ahmed Zaki Signed-off-by: Mateusz Palczewski Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf.h | 2 + drivers/net/ethernet/intel/iavf/iavf_main.c | 114 +++++++++++++----- .../net/ethernet/intel/iavf/iavf_virtchnl.c | 1 + 3 files changed, 85 insertions(+), 32 deletions(-) diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h index a5cab19eb6a8..bf5e3c8e97e0 100644 --- a/drivers/net/ethernet/intel/iavf/iavf.h +++ b/drivers/net/ethernet/intel/iavf/iavf.h @@ -255,6 +255,7 @@ struct iavf_adapter { struct workqueue_struct *wq; struct work_struct reset_task; struct work_struct adminq_task; + struct work_struct finish_config; struct delayed_work client_task; wait_queue_head_t down_waitqueue; wait_queue_head_t reset_waitqueue; @@ -521,6 +522,7 @@ int iavf_process_config(struct iavf_adapter *adapter); int iavf_parse_vf_resource_msg(struct iavf_adapter *adapter); void iavf_schedule_reset(struct iavf_adapter *adapter); void iavf_schedule_request_stats(struct iavf_adapter *adapter); +void iavf_schedule_finish_config(struct iavf_adapter *adapter); void iavf_reset(struct iavf_adapter *adapter); void iavf_set_ethtool_ops(struct net_device *netdev); void iavf_update_stats(struct iavf_adapter *adapter); diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index 0c12a752429c..2a2ae3c4337b 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -1690,10 +1690,10 @@ static int iavf_set_interrupt_capability(struct iavf_adapter *adapter) adapter->msix_entries[vector].entry = vector; err = iavf_acquire_msix_vectors(adapter, v_budget); + if (!err) + iavf_schedule_finish_config(adapter); out: - netif_set_real_num_rx_queues(adapter->netdev, pairs); - netif_set_real_num_tx_queues(adapter->netdev, pairs); return err; } @@ -1913,9 +1913,7 @@ static int iavf_init_interrupt_scheme(struct iavf_adapter *adapter) goto err_alloc_queues; } - rtnl_lock(); err = iavf_set_interrupt_capability(adapter); - rtnl_unlock(); if (err) { dev_err(&adapter->pdev->dev, "Unable to setup interrupt capabilities\n"); @@ -2001,6 +1999,78 @@ static int iavf_reinit_interrupt_scheme(struct iavf_adapter *adapter, bool runni return err; } +/** + * iavf_finish_config - do all netdev work that needs RTNL + * @work: our work_struct + * + * Do work that needs both RTNL and crit_lock. + **/ +static void iavf_finish_config(struct work_struct *work) +{ + struct iavf_adapter *adapter; + int pairs, err; + + adapter = container_of(work, struct iavf_adapter, finish_config); + + /* Always take RTNL first to prevent circular lock dependency */ + rtnl_lock(); + mutex_lock(&adapter->crit_lock); + + if ((adapter->flags & IAVF_FLAG_SETUP_NETDEV_FEATURES) && + adapter->netdev_registered && + !test_bit(__IAVF_IN_REMOVE_TASK, &adapter->crit_section)) { + netdev_update_features(adapter->netdev); + adapter->flags &= ~IAVF_FLAG_SETUP_NETDEV_FEATURES; + } + + switch (adapter->state) { + case __IAVF_DOWN: + if (!adapter->netdev_registered) { + err = register_netdevice(adapter->netdev); + if (err) { + dev_err(&adapter->pdev->dev, "Unable to register netdev (%d)\n", + err); + + /* go back and try again.*/ + iavf_free_rss(adapter); + iavf_free_misc_irq(adapter); + iavf_reset_interrupt_capability(adapter); + iavf_change_state(adapter, + __IAVF_INIT_CONFIG_ADAPTER); + goto out; + } + adapter->netdev_registered = true; + } + + /* Set the real number of queues when reset occurs while + * state == __IAVF_DOWN + */ + fallthrough; + case __IAVF_RUNNING: + pairs = adapter->num_active_queues; + netif_set_real_num_rx_queues(adapter->netdev, pairs); + netif_set_real_num_tx_queues(adapter->netdev, pairs); + break; + + default: + break; + } + +out: + mutex_unlock(&adapter->crit_lock); + rtnl_unlock(); +} + +/** + * iavf_schedule_finish_config - Set the flags and schedule a reset event + * @adapter: board private structure + **/ +void iavf_schedule_finish_config(struct iavf_adapter *adapter) +{ + if (!test_bit(__IAVF_IN_REMOVE_TASK, &adapter->crit_section)) + queue_work(adapter->wq, &adapter->finish_config); +} + /** * iavf_process_aq_command - process aq_required flags * and sends aq command @@ -2638,22 +2708,8 @@ static void iavf_init_config_adapter(struct iavf_adapter *adapter) netif_carrier_off(netdev); adapter->link_up = false; - - /* set the semaphore to prevent any callbacks after device registration - * up to time when state of driver will be set to __IAVF_DOWN - */ - rtnl_lock(); - if (!adapter->netdev_registered) { - err = register_netdevice(netdev); - if (err) { - rtnl_unlock(); - goto err_register; - } - } - - adapter->netdev_registered = true; - netif_tx_stop_all_queues(netdev); + if (CLIENT_ALLOWED(adapter)) { err = iavf_lan_add_device(adapter); if (err) @@ -2666,7 +2722,6 @@ static void iavf_init_config_adapter(struct iavf_adapter *adapter) iavf_change_state(adapter, __IAVF_DOWN); set_bit(__IAVF_VSI_DOWN, adapter->vsi.state); - rtnl_unlock(); iavf_misc_irq_enable(adapter); wake_up(&adapter->down_waitqueue); @@ -2686,10 +2741,11 @@ static void iavf_init_config_adapter(struct iavf_adapter *adapter) /* request initial VLAN offload settings */ iavf_set_vlan_offload_features(adapter, 0, netdev->features); + iavf_schedule_finish_config(adapter); return; + err_mem: iavf_free_rss(adapter); -err_register: iavf_free_misc_irq(adapter); err_sw_init: iavf_reset_interrupt_capability(adapter); @@ -2716,15 +2772,6 @@ static void iavf_watchdog_task(struct work_struct *work) goto restart_watchdog; } - if ((adapter->flags & IAVF_FLAG_SETUP_NETDEV_FEATURES) && - adapter->netdev_registered && - !test_bit(__IAVF_IN_REMOVE_TASK, &adapter->crit_section) && - rtnl_trylock()) { - netdev_update_features(adapter->netdev); - rtnl_unlock(); - adapter->flags &= ~IAVF_FLAG_SETUP_NETDEV_FEATURES; - } - if (adapter->flags & IAVF_FLAG_PF_COMMS_FAILED) iavf_change_state(adapter, __IAVF_COMM_FAILED); @@ -4942,6 +4989,7 @@ static int iavf_probe(struct pci_dev *pdev, const struct pci_device_id *ent) INIT_WORK(&adapter->reset_task, iavf_reset_task); INIT_WORK(&adapter->adminq_task, iavf_adminq_task); + INIT_WORK(&adapter->finish_config, iavf_finish_config); INIT_DELAYED_WORK(&adapter->watchdog_task, iavf_watchdog_task); INIT_DELAYED_WORK(&adapter->client_task, iavf_client_task); queue_delayed_work(adapter->wq, &adapter->watchdog_task, @@ -5084,13 +5132,15 @@ static void iavf_remove(struct pci_dev *pdev) usleep_range(500, 1000); } cancel_delayed_work_sync(&adapter->watchdog_task); + cancel_work_sync(&adapter->finish_config); + rtnl_lock(); if (adapter->netdev_registered) { - rtnl_lock(); unregister_netdevice(netdev); adapter->netdev_registered = false; - rtnl_unlock(); } + rtnl_unlock(); + if (CLIENT_ALLOWED(adapter)) { err = iavf_lan_del_device(adapter); if (err) diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c index 1bab896aaf40..073ac29ed84c 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c +++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c @@ -2237,6 +2237,7 @@ void iavf_virtchnl_completion(struct iavf_adapter *adapter, iavf_process_config(adapter); adapter->flags |= IAVF_FLAG_SETUP_NETDEV_FEATURES; + iavf_schedule_finish_config(adapter); iavf_set_queue_vlan_tag_loc(adapter); From c34743daca0eb1dc855831a5210f0800a850088e Mon Sep 17 00:00:00 2001 From: Ahmed Zaki Date: Mon, 5 Jun 2023 10:52:26 -0400 Subject: [PATCH 8/8] iavf: fix reset task race with iavf_remove() The reset task is currently scheduled from the watchdog or adminq tasks. First, all direct calls to schedule the reset task are replaced with the iavf_schedule_reset(), which is modified to accept the flag showing the type of reset. To prevent the reset task from starting once iavf_remove() starts, we need to check the __IAVF_IN_REMOVE_TASK bit before we schedule it. This is now easily added to iavf_schedule_reset(). Finally, remove the check for IAVF_FLAG_RESET_NEEDED in the watchdog task. It is redundant since all callers who set the flag immediately schedules the reset task. Fixes: 3ccd54ef44eb ("iavf: Fix init state closure on remove") Fixes: 14756b2ae265 ("iavf: Fix __IAVF_RESETTING state usage") Signed-off-by: Ahmed Zaki Signed-off-by: Mateusz Palczewski Tested-by: Rafal Romanowski Signed-off-by: Tony Nguyen --- drivers/net/ethernet/intel/iavf/iavf.h | 2 +- .../net/ethernet/intel/iavf/iavf_ethtool.c | 8 ++--- drivers/net/ethernet/intel/iavf/iavf_main.c | 32 +++++++------------ .../net/ethernet/intel/iavf/iavf_virtchnl.c | 3 +- 4 files changed, 16 insertions(+), 29 deletions(-) diff --git a/drivers/net/ethernet/intel/iavf/iavf.h b/drivers/net/ethernet/intel/iavf/iavf.h index bf5e3c8e97e0..8cbdebc5b698 100644 --- a/drivers/net/ethernet/intel/iavf/iavf.h +++ b/drivers/net/ethernet/intel/iavf/iavf.h @@ -520,7 +520,7 @@ int iavf_up(struct iavf_adapter *adapter); void iavf_down(struct iavf_adapter *adapter); int iavf_process_config(struct iavf_adapter *adapter); int iavf_parse_vf_resource_msg(struct iavf_adapter *adapter); -void iavf_schedule_reset(struct iavf_adapter *adapter); +void iavf_schedule_reset(struct iavf_adapter *adapter, u64 flags); void iavf_schedule_request_stats(struct iavf_adapter *adapter); void iavf_schedule_finish_config(struct iavf_adapter *adapter); void iavf_reset(struct iavf_adapter *adapter); diff --git a/drivers/net/ethernet/intel/iavf/iavf_ethtool.c b/drivers/net/ethernet/intel/iavf/iavf_ethtool.c index b7141c2a941d..2f47cfa7f06e 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_ethtool.c +++ b/drivers/net/ethernet/intel/iavf/iavf_ethtool.c @@ -532,8 +532,7 @@ static int iavf_set_priv_flags(struct net_device *netdev, u32 flags) /* issue a reset to force legacy-rx change to take effect */ if (changed_flags & IAVF_FLAG_LEGACY_RX) { if (netif_running(netdev)) { - adapter->flags |= IAVF_FLAG_RESET_NEEDED; - queue_work(adapter->wq, &adapter->reset_task); + iavf_schedule_reset(adapter, IAVF_FLAG_RESET_NEEDED); ret = iavf_wait_for_reset(adapter); if (ret) netdev_warn(netdev, "Changing private flags timeout or interrupted waiting for reset"); @@ -676,8 +675,7 @@ static int iavf_set_ringparam(struct net_device *netdev, } if (netif_running(netdev)) { - adapter->flags |= IAVF_FLAG_RESET_NEEDED; - queue_work(adapter->wq, &adapter->reset_task); + iavf_schedule_reset(adapter, IAVF_FLAG_RESET_NEEDED); ret = iavf_wait_for_reset(adapter); if (ret) netdev_warn(netdev, "Changing ring parameters timeout or interrupted waiting for reset"); @@ -1860,7 +1858,7 @@ static int iavf_set_channels(struct net_device *netdev, adapter->num_req_queues = num_req; adapter->flags |= IAVF_FLAG_REINIT_ITR_NEEDED; - iavf_schedule_reset(adapter); + iavf_schedule_reset(adapter, IAVF_FLAG_RESET_NEEDED); ret = iavf_wait_for_reset(adapter); if (ret) diff --git a/drivers/net/ethernet/intel/iavf/iavf_main.c b/drivers/net/ethernet/intel/iavf/iavf_main.c index 2a2ae3c4337b..3a88d413ddee 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_main.c +++ b/drivers/net/ethernet/intel/iavf/iavf_main.c @@ -301,12 +301,14 @@ static int iavf_lock_timeout(struct mutex *lock, unsigned int msecs) /** * iavf_schedule_reset - Set the flags and schedule a reset event * @adapter: board private structure + * @flags: IAVF_FLAG_RESET_PENDING or IAVF_FLAG_RESET_NEEDED **/ -void iavf_schedule_reset(struct iavf_adapter *adapter) +void iavf_schedule_reset(struct iavf_adapter *adapter, u64 flags) { - if (!(adapter->flags & - (IAVF_FLAG_RESET_PENDING | IAVF_FLAG_RESET_NEEDED))) { - adapter->flags |= IAVF_FLAG_RESET_NEEDED; + if (!test_bit(__IAVF_IN_REMOVE_TASK, &adapter->crit_section) && + !(adapter->flags & + (IAVF_FLAG_RESET_PENDING | IAVF_FLAG_RESET_NEEDED))) { + adapter->flags |= flags; queue_work(adapter->wq, &adapter->reset_task); } } @@ -334,7 +336,7 @@ static void iavf_tx_timeout(struct net_device *netdev, unsigned int txqueue) struct iavf_adapter *adapter = netdev_priv(netdev); adapter->tx_timeout_count++; - iavf_schedule_reset(adapter); + iavf_schedule_reset(adapter, IAVF_FLAG_RESET_NEEDED); } /** @@ -2478,7 +2480,7 @@ int iavf_parse_vf_resource_msg(struct iavf_adapter *adapter) adapter->vsi_res->num_queue_pairs); adapter->flags |= IAVF_FLAG_REINIT_MSIX_NEEDED; adapter->num_req_queues = adapter->vsi_res->num_queue_pairs; - iavf_schedule_reset(adapter); + iavf_schedule_reset(adapter, IAVF_FLAG_RESET_NEEDED); return -EAGAIN; } @@ -2775,14 +2777,6 @@ static void iavf_watchdog_task(struct work_struct *work) if (adapter->flags & IAVF_FLAG_PF_COMMS_FAILED) iavf_change_state(adapter, __IAVF_COMM_FAILED); - if (adapter->flags & IAVF_FLAG_RESET_NEEDED) { - adapter->aq_required = 0; - adapter->current_op = VIRTCHNL_OP_UNKNOWN; - mutex_unlock(&adapter->crit_lock); - queue_work(adapter->wq, &adapter->reset_task); - return; - } - switch (adapter->state) { case __IAVF_STARTUP: iavf_startup(adapter); @@ -2910,11 +2904,10 @@ static void iavf_watchdog_task(struct work_struct *work) /* check for hw reset */ reg_val = rd32(hw, IAVF_VF_ARQLEN1) & IAVF_VF_ARQLEN1_ARQENABLE_MASK; if (!reg_val) { - adapter->flags |= IAVF_FLAG_RESET_PENDING; adapter->aq_required = 0; adapter->current_op = VIRTCHNL_OP_UNKNOWN; dev_err(&adapter->pdev->dev, "Hardware reset detected\n"); - queue_work(adapter->wq, &adapter->reset_task); + iavf_schedule_reset(adapter, IAVF_FLAG_RESET_PENDING); mutex_unlock(&adapter->crit_lock); queue_delayed_work(adapter->wq, &adapter->watchdog_task, HZ * 2); @@ -3288,9 +3281,7 @@ static void iavf_adminq_task(struct work_struct *work) } while (pending); mutex_unlock(&adapter->crit_lock); - if ((adapter->flags & - (IAVF_FLAG_RESET_PENDING | IAVF_FLAG_RESET_NEEDED)) || - adapter->state == __IAVF_RESETTING) + if (iavf_is_reset_in_progress(adapter)) goto freedom; /* check for error indications */ @@ -4387,8 +4378,7 @@ static int iavf_change_mtu(struct net_device *netdev, int new_mtu) } if (netif_running(netdev)) { - adapter->flags |= IAVF_FLAG_RESET_NEEDED; - queue_work(adapter->wq, &adapter->reset_task); + iavf_schedule_reset(adapter, IAVF_FLAG_RESET_NEEDED); ret = iavf_wait_for_reset(adapter); if (ret < 0) netdev_warn(netdev, "MTU change interrupted waiting for reset"); diff --git a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c index 073ac29ed84c..be3c007ce90a 100644 --- a/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c +++ b/drivers/net/ethernet/intel/iavf/iavf_virtchnl.c @@ -1961,9 +1961,8 @@ void iavf_virtchnl_completion(struct iavf_adapter *adapter, case VIRTCHNL_EVENT_RESET_IMPENDING: dev_info(&adapter->pdev->dev, "Reset indication received from the PF\n"); if (!(adapter->flags & IAVF_FLAG_RESET_PENDING)) { - adapter->flags |= IAVF_FLAG_RESET_PENDING; dev_info(&adapter->pdev->dev, "Scheduling reset task\n"); - queue_work(adapter->wq, &adapter->reset_task); + iavf_schedule_reset(adapter, IAVF_FLAG_RESET_PENDING); } break; default: