From: Lukas Wunner lukas@wunner.de

commit 23584c1ed3e15a6f4bfab8dc5a88d94ab929ee12 upstream.

The Power Fault Detected bit in the Slot Status register differs from all other hotplug events in that it is sticky: It can only be cleared after turning off slot power. Per PCIe r5.0, sec. 6.7.1.8:

If a power controller detects a main power fault on the hot-plug slot, it must automatically set its internal main power fault latch [...]. The main power fault latch is cleared when software turns off power to the hot-plug slot.

The stickiness used to cause interrupt storms and infinite loops which were fixed in 2009 by commits 5651c48cfafe ("PCI pciehp: fix power fault interrupt storm problem") and 99f0169c17f3 ("PCI: pciehp: enable software notification on empty slots").

Unfortunately in 2020 the infinite loop issue was inadvertently reintroduced by commit 8edf5332c393 ("PCI: pciehp: Fix MSI interrupt race"): The hardirq handler pciehp_isr() clears the PFD bit until pciehp's power_fault_detected flag is set. That happens in the IRQ thread pciehp_ist(), which never learns of the event because the hardirq handler is stuck in an infinite loop. Fix by setting the power_fault_detected flag already in the hardirq handler.

Link: https://bugzilla.kernel.org/show_bug.cgi?id=214989 Link: https://lore.kernel.org/linux-pci/DM8PR11MB5702255A6A92F735D90A4446868B9@DM8... Fixes: 8edf5332c393 ("PCI: pciehp: Fix MSI interrupt race") Link: https://lore.kernel.org/r/66eaeef31d4997ceea357ad93259f290ededecfd.163718722... Reported-by: Joseph Bao joseph.bao@intel.com Tested-by: Joseph Bao joseph.bao@intel.com Signed-off-by: Lukas Wunner lukas@wunner.de Signed-off-by: Bjorn Helgaas bhelgaas@google.com Cc: stable@vger.kernel.org # v4.19+ Cc: Stuart Hayes stuart.w.hayes@gmail.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/pci/hotplug/pciehp_hpc.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)

--- a/drivers/pci/hotplug/pciehp_hpc.c +++ b/drivers/pci/hotplug/pciehp_hpc.c @@ -642,6 +642,8 @@ read_status: */ if (ctrl->power_fault_detected) status &= ~PCI_EXP_SLTSTA_PFD; + else if (status & PCI_EXP_SLTSTA_PFD) + ctrl->power_fault_detected = true;

events |= status; if (!events) { @@ -651,7 +653,7 @@ read_status: }

if (status) { - pcie_capability_write_word(pdev, PCI_EXP_SLTSTA, events); + pcie_capability_write_word(pdev, PCI_EXP_SLTSTA, status);

/* * In MSI mode, all event bits must be zero before the port @@ -725,8 +727,7 @@ static irqreturn_t pciehp_ist(int irq, v }

/* Check Power Fault Detected */ - if ((events & PCI_EXP_SLTSTA_PFD) && !ctrl->power_fault_detected) { - ctrl->power_fault_detected = 1; + if (events & PCI_EXP_SLTSTA_PFD) { ctrl_err(ctrl, "Slot(%s): Power fault\n", slot_name(ctrl)); pciehp_set_indicators(ctrl, PCI_EXP_SLTCTL_PWR_IND_OFF, PCI_EXP_SLTCTL_ATTN_IND_ON);

From: Alex Elder elder@linaro.org

commit c1aaa01dbf4cef95af3e04a5a43986c290e06ea3 upstream.

Define a new replenish_flags bitmap to contain Boolean flags associated with an endpoint's replenishing state. Replace the replenish_enabled field with a flag in that bitmap. This is to prepare for the next patch, which adds another flag.

Signed-off-by: Alex Elder elder@linaro.org Signed-off-by: David S. Miller davem@davemloft.net Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ipa/ipa_endpoint.c | 8 ++++---- drivers/net/ipa/ipa_endpoint.h | 15 +++++++++++++-- 2 files changed, 17 insertions(+), 6 deletions(-)

--- a/drivers/net/ipa/ipa_endpoint.c +++ b/drivers/net/ipa/ipa_endpoint.c @@ -1069,7 +1069,7 @@ static void ipa_endpoint_replenish(struc u32 backlog; int delta;

- if (!endpoint->replenish_enabled) { + if (!test_bit(IPA_REPLENISH_ENABLED, endpoint->replenish_flags)) { if (add_one) atomic_inc(&endpoint->replenish_saved); return; @@ -1106,7 +1106,7 @@ static void ipa_endpoint_replenish_enabl u32 max_backlog; u32 saved;

- endpoint->replenish_enabled = true; + set_bit(IPA_REPLENISH_ENABLED, endpoint->replenish_flags); while ((saved = atomic_xchg(&endpoint->replenish_saved, 0))) atomic_add(saved, &endpoint->replenish_backlog);

@@ -1120,7 +1120,7 @@ static void ipa_endpoint_replenish_disab { u32 backlog;

- endpoint->replenish_enabled = false; + clear_bit(IPA_REPLENISH_ENABLED, endpoint->replenish_flags); while ((backlog = atomic_xchg(&endpoint->replenish_backlog, 0))) atomic_add(backlog, &endpoint->replenish_saved); } @@ -1665,7 +1665,7 @@ static void ipa_endpoint_setup_one(struc /* RX transactions require a single TRE, so the maximum * backlog is the same as the maximum outstanding TREs. */ - endpoint->replenish_enabled = false; + clear_bit(IPA_REPLENISH_ENABLED, endpoint->replenish_flags); atomic_set(&endpoint->replenish_saved, gsi_channel_tre_max(gsi, endpoint->channel_id)); atomic_set(&endpoint->replenish_backlog, 0); --- a/drivers/net/ipa/ipa_endpoint.h +++ b/drivers/net/ipa/ipa_endpoint.h @@ -41,6 +41,17 @@ enum ipa_endpoint_name { #define IPA_ENDPOINT_MAX 32 /* Max supported by driver */

/** + * enum ipa_replenish_flag: RX buffer replenish flags + * + * @IPA_REPLENISH_ENABLED: Whether receive buffer replenishing is enabled + * @IPA_REPLENISH_COUNT: Number of defined replenish flags + */ +enum ipa_replenish_flag { + IPA_REPLENISH_ENABLED, + IPA_REPLENISH_COUNT, /* Number of flags (must be last) */ +}; + +/** * struct ipa_endpoint - IPA endpoint information * @ipa: IPA pointer * @ee_id: Execution environmnent endpoint is associated with @@ -51,7 +62,7 @@ enum ipa_endpoint_name { * @trans_tre_max: Maximum number of TRE descriptors per transaction * @evt_ring_id: GSI event ring used by the endpoint * @netdev: Network device pointer, if endpoint uses one - * @replenish_enabled: Whether receive buffer replenishing is enabled + * @replenish_flags: Replenishing state flags * @replenish_ready: Number of replenish transactions without doorbell * @replenish_saved: Replenish requests held while disabled * @replenish_backlog: Number of buffers needed to fill hardware queue @@ -72,7 +83,7 @@ struct ipa_endpoint { struct net_device *netdev;

/* Receive buffer replenishing for RX endpoints */ - bool replenish_enabled; + DECLARE_BITMAP(replenish_flags, IPA_REPLENISH_COUNT); u32 replenish_ready; atomic_t replenish_saved; atomic_t replenish_backlog;

From: Maxime Ripard maxime@cerno.tech

Commit 20b0dfa86bef0e80b41b0e5ac38b92f23b6f27f9 upstream.

The original commit depended on a rework commit (724fc856c09e ("drm/vc4: hdmi: Split the CEC disable / enable functions in two")) that (rightfully) didn't reach stable.

However, probably because the context changed, when the patch was applied to stable the pm_runtime_put called got moved to the end of the vc4_hdmi_cec_adap_enable function (that would have become vc4_hdmi_cec_disable with the rework) to vc4_hdmi_cec_init.

This means that at probe time, we now drop our reference to the clocks and power domains and thus end up with a CPU hang when the CPU tries to access registers.

The call to pm_runtime_resume_and_get() is also problematic since the .adap_enable CEC hook is called both to enable and to disable the controller. That means that we'll now call pm_runtime_resume_and_get() at disable time as well, messing with the reference counting.

The behaviour we should have though would be to have pm_runtime_resume_and_get() called when the CEC controller is enabled, and pm_runtime_put when it's disabled.

We need to move things around a bit to behave that way, but it aligns stable with upstream.

Cc: stable@vger.kernel.org # 5.10.x Cc: stable@vger.kernel.org # 5.15.x Cc: stable@vger.kernel.org # 5.16.x Reported-by: Michael Stapelberg michael+drm@stapelberg.ch Signed-off-by: Maxime Ripard maxime@cerno.tech Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/gpu/drm/vc4/vc4_hdmi.c | 25 +++++++++++++------------ 1 file changed, 13 insertions(+), 12 deletions(-)

--- a/drivers/gpu/drm/vc4/vc4_hdmi.c +++ b/drivers/gpu/drm/vc4/vc4_hdmi.c @@ -1739,18 +1739,18 @@ static int vc4_hdmi_cec_adap_enable(stru u32 val; int ret;

- ret = pm_runtime_resume_and_get(&vc4_hdmi->pdev->dev); - if (ret) - return ret; + if (enable) { + ret = pm_runtime_resume_and_get(&vc4_hdmi->pdev->dev); + if (ret) + return ret;

- val = HDMI_READ(HDMI_CEC_CNTRL_5); - val &= ~(VC4_HDMI_CEC_TX_SW_RESET | VC4_HDMI_CEC_RX_SW_RESET | - VC4_HDMI_CEC_CNT_TO_4700_US_MASK | - VC4_HDMI_CEC_CNT_TO_4500_US_MASK); - val |= ((4700 / usecs) << VC4_HDMI_CEC_CNT_TO_4700_US_SHIFT) | - ((4500 / usecs) << VC4_HDMI_CEC_CNT_TO_4500_US_SHIFT); + val = HDMI_READ(HDMI_CEC_CNTRL_5); + val &= ~(VC4_HDMI_CEC_TX_SW_RESET | VC4_HDMI_CEC_RX_SW_RESET | + VC4_HDMI_CEC_CNT_TO_4700_US_MASK | + VC4_HDMI_CEC_CNT_TO_4500_US_MASK); + val |= ((4700 / usecs) << VC4_HDMI_CEC_CNT_TO_4700_US_SHIFT) | + ((4500 / usecs) << VC4_HDMI_CEC_CNT_TO_4500_US_SHIFT);

- if (enable) { HDMI_WRITE(HDMI_CEC_CNTRL_5, val | VC4_HDMI_CEC_TX_SW_RESET | VC4_HDMI_CEC_RX_SW_RESET); HDMI_WRITE(HDMI_CEC_CNTRL_5, val); @@ -1778,7 +1778,10 @@ static int vc4_hdmi_cec_adap_enable(stru HDMI_WRITE(HDMI_CEC_CPU_MASK_SET, VC4_HDMI_CPU_CEC); HDMI_WRITE(HDMI_CEC_CNTRL_5, val | VC4_HDMI_CEC_TX_SW_RESET | VC4_HDMI_CEC_RX_SW_RESET); + + pm_runtime_put(&vc4_hdmi->pdev->dev); } + return 0; }

@@ -1889,8 +1892,6 @@ static int vc4_hdmi_cec_init(struct vc4_ if (ret < 0) goto err_remove_handlers;

- pm_runtime_put(&vc4_hdmi->pdev->dev); - return 0;

err_remove_handlers:

From: John Hubbard jhubbard@nvidia.com

commit c36c04c2e132fc39f6b658bf607aed4425427fd7 upstream.

This reverts commit 54d516b1d62ff8f17cee2da06e5e4706a0d00b8a

That commit did a refactoring that effectively combined fast and slow gup paths (again). And that was again incorrect, for two reasons:

a) Fast gup and slow gup get reference counts on pages in different ways and with different goals: see Linus' writeup in commit cd1adf1b63a1 ("Revert "mm/gup: remove try_get_page(), call try_get_compound_head() directly""), and

b) try_grab_compound_head() also has a specific check for "FOLL_LONGTERM && !is_pinned(page)", that assumes that the caller can fall back to slow gup. This resulted in new failures, as recently report by Will McVicker [1].

But (a) has problems too, even though they may not have been reported yet. So just revert this.

Link: https://lore.kernel.org/r/20220131203504.3458775-1-willmcvicker@google.com [1] Fixes: 54d516b1d62f ("mm/gup: small refactoring: simplify try_grab_page()") Reported-and-tested-by: Will McVicker willmcvicker@google.com Cc: Christoph Hellwig hch@lst.de Cc: Minchan Kim minchan@google.com Cc: Matthew Wilcox willy@infradead.org Cc: Christian Borntraeger borntraeger@de.ibm.com Cc: Heiko Carstens hca@linux.ibm.com Cc: Vasily Gorbik gor@linux.ibm.com Cc: stable@vger.kernel.org # 5.15 Signed-off-by: John Hubbard jhubbard@nvidia.com Signed-off-by: Linus Torvalds torvalds@linux-foundation.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- mm/gup.c | 35 ++++++++++++++++++++++++++++++----- 1 file changed, 30 insertions(+), 5 deletions(-)

--- a/mm/gup.c +++ b/mm/gup.c @@ -124,8 +124,8 @@ static inline struct page *try_get_compo * considered failure, and furthermore, a likely bug in the caller, so a warning * is also emitted. */ -struct page *try_grab_compound_head(struct page *page, - int refs, unsigned int flags) +__maybe_unused struct page *try_grab_compound_head(struct page *page, + int refs, unsigned int flags) { if (flags & FOLL_GET) return try_get_compound_head(page, refs); @@ -208,10 +208,35 @@ static void put_compound_head(struct pag */ bool __must_check try_grab_page(struct page *page, unsigned int flags) { - if (!(flags & (FOLL_GET | FOLL_PIN))) - return true; + WARN_ON_ONCE((flags & (FOLL_GET | FOLL_PIN)) == (FOLL_GET | FOLL_PIN));

- return try_grab_compound_head(page, 1, flags); + if (flags & FOLL_GET) + return try_get_page(page); + else if (flags & FOLL_PIN) { + int refs = 1; + + page = compound_head(page); + + if (WARN_ON_ONCE(page_ref_count(page) <= 0)) + return false; + + if (hpage_pincount_available(page)) + hpage_pincount_add(page, 1); + else + refs = GUP_PIN_COUNTING_BIAS; + + /* + * Similar to try_grab_compound_head(): even if using the + * hpage_pincount_add/_sub() routines, be sure to + * *also* increment the normal page refcount field at least + * once, so that the page really is pinned. + */ + page_ref_add(page, refs); + + mod_node_page_state(page_pgdat(page), NR_FOLL_PIN_ACQUIRED, 1); + } + + return true; }

/**

From: Miklos Szeredi mszeredi@redhat.com

commit 94fd19752b28aa66c98e7991734af91dfc529f8f upstream.

Christoph Fritz is reporting that failure to copy up fileattr when upper doesn't support fileattr or xattr results in a regression.

Return success in these failure cases; this reverts overlayfs to the old behavior.

Add a pr_warn_once() in these cases to still let the user know about the copy up failures.

Reported-by: Christoph Fritz chf.fritz@googlemail.com Fixes: 72db82115d2b ("ovl: copy up sync/noatime fileattr flags") Cc: stable@vger.kernel.org # v5.15 Signed-off-by: Miklos Szeredi mszeredi@redhat.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- fs/overlayfs/copy_up.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)

--- a/fs/overlayfs/copy_up.c +++ b/fs/overlayfs/copy_up.c @@ -157,7 +157,9 @@ static int ovl_copy_fileattr(struct inod */ if (oldfa.flags & OVL_PROT_FS_FLAGS_MASK) { err = ovl_set_protattr(inode, new->dentry, &oldfa); - if (err) + if (err == -EPERM) + pr_warn_once("copying fileattr: no xattr on upper\n"); + else if (err) return err; }

@@ -167,6 +169,14 @@ static int ovl_copy_fileattr(struct inod

err = ovl_real_fileattr_get(new, &newfa); if (err) { + /* + * Returning an error if upper doesn't support fileattr will + * result in a regression, so revert to the old behavior. + */ + if (err == -ENOTTY || err == -EINVAL) { + pr_warn_once("copying fileattr: no support on upper\n"); + return 0; + } pr_warn("failed to retrieve upper fileattr (%pd2, err=%i)\n", new, err); return err;

From: J. Bruce Fields bfields@redhat.com

commit d19a7af73b5ecaac8168712d18be72b9db166768 upstream.

In my testing, we're sometimes hitting the request->fl_flags & FL_EXISTS case in posix_lock_inode, presumably just by random luck since we're not actually initializing fl_flags here.

This probably didn't matter before commit 7f024fcd5c97 ("Keep read and write fds with each nlm_file") since we wouldn't previously unlock unless we knew there were locks.

But now it causes lockd to give up on removing more locks.

We could just initialize fl_flags, but really it seems dubious to be calling vfs_lock_file with random values in some of the fields.

Fixes: 7f024fcd5c97 ("Keep read and write fds with each nlm_file") Signed-off-by: J. Bruce Fields bfields@redhat.com [ cel: fixed checkpatch.pl nit ] Signed-off-by: Chuck Lever chuck.lever@oracle.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- fs/lockd/svcsubs.c | 1 + 1 file changed, 1 insertion(+)

--- a/fs/lockd/svcsubs.c +++ b/fs/lockd/svcsubs.c @@ -180,6 +180,7 @@ static int nlm_unlock_files(struct nlm_f { struct file_lock lock;

+ locks_init_lock(&lock); lock.fl_type = F_UNLCK; lock.fl_start = 0; lock.fl_end = OFFSET_MAX;

From: Raed Salem raeds@nvidia.com

commit de47db0cf7f4a9c555ad204e06baa70b50a70d08 upstream.

IPsec Tunnel mode crypto offload software parser (SWP) setting in data path currently always set the inner L4 offset regardless of the encapsulated L4 header type and whether it exists in the first place, this breaks non TCP/UDP traffic as such.

Set the SWP inner L4 offset only when the IPsec tunnel encapsulated L4 header protocol is TCP/UDP.

While at it fix inner ip protocol read for setting MLX5_ETH_WQE_SWP_INNER_L4_UDP flag to address the case where the ip header protocol is IPv6.

Fixes: f1267798c980 ("net/mlx5: Fix checksum issue of VXLAN and IPsec crypto offload") Signed-off-by: Raed Salem raeds@nvidia.com Reviewed-by: Maor Dickman maord@nvidia.com Signed-off-by: Saeed Mahameed saeedm@nvidia.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.c | 13 ++++++++-- 1 file changed, 11 insertions(+), 2 deletions(-)

--- a/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_accel/ipsec_rxtx.c @@ -157,11 +157,20 @@ static void mlx5e_ipsec_set_swp(struct s /* Tunnel mode */ if (mode == XFRM_MODE_TUNNEL) { eseg->swp_inner_l3_offset = skb_inner_network_offset(skb) / 2; - eseg->swp_inner_l4_offset = skb_inner_transport_offset(skb) / 2; if (xo->proto == IPPROTO_IPV6) eseg->swp_flags |= MLX5_ETH_WQE_SWP_INNER_L3_IPV6; - if (inner_ip_hdr(skb)->protocol == IPPROTO_UDP) + + switch (xo->inner_ipproto) { + case IPPROTO_UDP: eseg->swp_flags |= MLX5_ETH_WQE_SWP_INNER_L4_UDP; + fallthrough; + case IPPROTO_TCP: + /* IP | ESP | IP | [TCP | UDP] */ + eseg->swp_inner_l4_offset = skb_inner_transport_offset(skb) / 2; + break; + default: + break; + } return; }

From: Vlad Buslov vladbu@nvidia.com

commit 04f8c12f031fcd0ffa0c72822eb665ceb2c872e7 upstream.

The mlx5_esw_bridge_cleanup() is expected to be called with rtnl lock taken, which is true for mlx5e_rep_bridge_cleanup() function but not for error handling code in mlx5e_rep_bridge_init(). Add missing rtnl lock/unlock calls and extend both mlx5_esw_bridge_cleanup() and its dual function mlx5_esw_bridge_init() with ASSERT_RTNL() to verify the invariant from now on.

Fixes: 7cd6a54a8285 ("net/mlx5: Bridge, handle FDB events") Fixes: 19e9bfa044f3 ("net/mlx5: Bridge, add offload infrastructure") Signed-off-by: Vlad Buslov vladbu@nvidia.com Reviewed-by: Roi Dayan roid@nvidia.com Signed-off-by: Saeed Mahameed saeedm@nvidia.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c | 2 ++ drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c | 4 ++++ 2 files changed, 6 insertions(+)

--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c @@ -509,7 +509,9 @@ err_register_swdev_blk: err_register_swdev: destroy_workqueue(br_offloads->wq); err_alloc_wq: + rtnl_lock(); mlx5_esw_bridge_cleanup(esw); + rtnl_unlock(); }

void mlx5e_rep_bridge_cleanup(struct mlx5e_priv *priv) --- a/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c @@ -1574,6 +1574,8 @@ struct mlx5_esw_bridge_offloads *mlx5_es { struct mlx5_esw_bridge_offloads *br_offloads;

+ ASSERT_RTNL(); + br_offloads = kvzalloc(sizeof(*br_offloads), GFP_KERNEL); if (!br_offloads) return ERR_PTR(-ENOMEM); @@ -1590,6 +1592,8 @@ void mlx5_esw_bridge_cleanup(struct mlx5 { struct mlx5_esw_bridge_offloads *br_offloads = esw->br_offloads;

+ ASSERT_RTNL(); + if (!br_offloads) return;

From: Maor Dickman maord@nvidia.com

commit ec41332e02bd0acf1f24206867bb6a02f5877a62 upstream.

Current implementation of bond netevent handler only check if the handled netdev is VF representor and it missing a check if the VF representor is on the same phys device of the bond handling the netevent.

Fix by adding the missing check and optimizing the check if the netdev is VF representor so it will not access uninitialized private data and crashes.

BUG: kernel NULL pointer dereference, address: 000000000000036c PGD 0 P4D 0 Oops: 0000 [#1] SMP NOPTI Workqueue: eth3bond0 bond_mii_monitor [bonding] RIP: 0010:mlx5e_is_uplink_rep+0xc/0x50 [mlx5_core] RSP: 0018:ffff88812d69fd60 EFLAGS: 00010282 RAX: 0000000000000000 RBX: ffff8881cf800000 RCX: 0000000000000000 RDX: ffff88812d69fe10 RSI: 000000000000001b RDI: ffff8881cf800880 RBP: ffff8881cf800000 R08: 00000445cabccf2b R09: 0000000000000008 R10: 0000000000000004 R11: 0000000000000008 R12: ffff88812d69fe10 R13: 00000000fffffffe R14: ffff88820c0f9000 R15: 0000000000000000 FS: 0000000000000000(0000) GS:ffff88846fb00000(0000) knlGS:0000000000000000 CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033 CR2: 000000000000036c CR3: 0000000103d80006 CR4: 0000000000370ea0 DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000 DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400 Call Trace: mlx5e_eswitch_uplink_rep+0x31/0x40 [mlx5_core] mlx5e_rep_is_lag_netdev+0x94/0xc0 [mlx5_core] mlx5e_rep_esw_bond_netevent+0xeb/0x3d0 [mlx5_core] raw_notifier_call_chain+0x41/0x60 call_netdevice_notifiers_info+0x34/0x80 netdev_lower_state_changed+0x4e/0xa0 bond_mii_monitor+0x56b/0x640 [bonding] process_one_work+0x1b9/0x390 worker_thread+0x4d/0x3d0 ? rescuer_thread+0x350/0x350 kthread+0x124/0x150 ? set_kthread_struct+0x40/0x40 ret_from_fork+0x1f/0x30

Fixes: 7e51891a237f ("net/mlx5e: Use netdev events to set/del egress acl forward-to-vport rule") Signed-off-by: Maor Dickman maord@nvidia.com Reviewed-by: Roi Dayan roid@nvidia.com Signed-off-by: Saeed Mahameed saeedm@nvidia.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx5/core/en/rep/bond.c | 32 +++++++----------- 1 file changed, 14 insertions(+), 18 deletions(-)

--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bond.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bond.c @@ -183,18 +183,7 @@ void mlx5e_rep_bond_unslave(struct mlx5_

static bool mlx5e_rep_is_lag_netdev(struct net_device *netdev) { - struct mlx5e_rep_priv *rpriv; - struct mlx5e_priv *priv; - - /* A given netdev is not a representor or not a slave of LAG configuration */ - if (!mlx5e_eswitch_rep(netdev) || !netif_is_lag_port(netdev)) - return false; - - priv = netdev_priv(netdev); - rpriv = priv->ppriv; - - /* Egress acl forward to vport is supported only non-uplink representor */ - return rpriv->rep->vport != MLX5_VPORT_UPLINK; + return netif_is_lag_port(netdev) && mlx5e_eswitch_vf_rep(netdev); }

static void mlx5e_rep_changelowerstate_event(struct net_device *netdev, void *ptr) @@ -210,9 +199,6 @@ static void mlx5e_rep_changelowerstate_e u16 fwd_vport_num; int err;

- if (!mlx5e_rep_is_lag_netdev(netdev)) - return; - info = ptr; lag_info = info->lower_state_info; /* This is not an event of a representor becoming active slave */ @@ -266,9 +252,6 @@ static void mlx5e_rep_changeupper_event( struct net_device *lag_dev; struct mlx5e_priv *priv;

- if (!mlx5e_rep_is_lag_netdev(netdev)) - return; - priv = netdev_priv(netdev); rpriv = priv->ppriv; lag_dev = info->upper_dev; @@ -293,6 +276,19 @@ static int mlx5e_rep_esw_bond_netevent(s unsigned long event, void *ptr) { struct net_device *netdev = netdev_notifier_info_to_dev(ptr); + struct mlx5e_rep_priv *rpriv; + struct mlx5e_rep_bond *bond; + struct mlx5e_priv *priv; + + if (!mlx5e_rep_is_lag_netdev(netdev)) + return NOTIFY_DONE; + + bond = container_of(nb, struct mlx5e_rep_bond, nb); + priv = netdev_priv(netdev); + rpriv = mlx5_eswitch_get_uplink_priv(priv->mdev->priv.eswitch, REP_ETH); + /* Verify VF representor is on the same device of the bond handling the netevent. */ + if (rpriv->uplink_priv.bond != bond) + return NOTIFY_DONE;

switch (event) { case NETDEV_CHANGELOWERSTATE:

From: Gal Pressman gal@nvidia.com

commit 4a08a131351e375a2969b98e46df260ed04dcba7 upstream.

When querying the module EEPROM, there was a misusage of the 'offset' variable vs the 'query.offset' field. Fix that by always using 'offset' and assigning its value to 'query.offset' right before the mcia register read call.

While at it, the cross-pages read size adjustment was changed to be more intuitive.

Fixes: e19b0a3474ab ("net/mlx5: Refactor module EEPROM query") Reported-by: Wang Yugui wangyugui@e16-tech.com Signed-off-by: Gal Pressman gal@nvidia.com Reviewed-by: Maxim Mikityanskiy maximmi@nvidia.com Signed-off-by: Saeed Mahameed saeedm@nvidia.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx5/core/port.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)

--- a/drivers/net/ethernet/mellanox/mlx5/core/port.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/port.c @@ -406,23 +406,24 @@ int mlx5_query_module_eeprom(struct mlx5

switch (module_id) { case MLX5_MODULE_ID_SFP: - mlx5_sfp_eeprom_params_set(&query.i2c_address, &query.page, &query.offset); + mlx5_sfp_eeprom_params_set(&query.i2c_address, &query.page, &offset); break; case MLX5_MODULE_ID_QSFP: case MLX5_MODULE_ID_QSFP_PLUS: case MLX5_MODULE_ID_QSFP28: - mlx5_qsfp_eeprom_params_set(&query.i2c_address, &query.page, &query.offset); + mlx5_qsfp_eeprom_params_set(&query.i2c_address, &query.page, &offset); break; default: mlx5_core_err(dev, "Module ID not recognized: 0x%x\n", module_id); return -EINVAL; }

- if (query.offset + size > MLX5_EEPROM_PAGE_LENGTH) + if (offset + size > MLX5_EEPROM_PAGE_LENGTH) /* Cross pages read, read until offset 256 in low page */ - size -= offset + size - MLX5_EEPROM_PAGE_LENGTH; + size = MLX5_EEPROM_PAGE_LENGTH - offset;

query.size = size; + query.offset = offset;

return mlx5_query_mcia(dev, &query, data); }

From: Dima Chumak dchumak@nvidia.com

commit 55b2ca702cfa744a9eb108915996a2294da47e71 upstream.

Only prio 1 is supported for nic mode when there is no ignore flow level support in firmware. But for switchdev mode, which supports fixed number of statically pre-allocated prios, this restriction is not relevant so it can be relaxed.

Fixes: d671e109bd85 ("net/mlx5: Fix tc max supported prio for nic mode") Signed-off-by: Dima Chumak dchumak@nvidia.com Reviewed-by: Roi Dayan roid@nvidia.com Signed-off-by: Saeed Mahameed saeedm@nvidia.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c | 7 ++++--- 1 file changed, 4 insertions(+), 3 deletions(-)

--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c @@ -121,12 +121,13 @@ u32 mlx5_chains_get_nf_ft_chain(struct m

u32 mlx5_chains_get_prio_range(struct mlx5_fs_chains *chains) { - if (!mlx5_chains_prios_supported(chains)) - return 1; - if (mlx5_chains_ignore_flow_level_supported(chains)) return UINT_MAX;

+ if (!chains->dev->priv.eswitch || + chains->dev->priv.eswitch->mode != MLX5_ESWITCH_OFFLOADS) + return 1; + /* We should get here only for eswitch case */ return FDB_TC_MAX_PRIO; }

From: Roi Dayan roid@nvidia.com

commit 880b517691908fb753019b9b27cd082e7617debd upstream.

When changing mode to switchdev, rep bridge init registered to netdevice notifier holds the devlink lock and then takes pernet_ops_rwsem. At that time deleting a netns holds pernet_ops_rwsem and then takes the devlink lock.

Example sequence is: $ ip netns add foo $ devlink dev eswitch set pci/0000:00:08.0 mode switchdev & $ ip netns del foo

deleting netns trace:

[ 1185.365555] ? devlink_pernet_pre_exit+0x74/0x1c0 [ 1185.368331] ? mutex_lock_io_nested+0x13f0/0x13f0 [ 1185.370984] ? xt_find_table+0x40/0x100 [ 1185.373244] ? __mutex_lock+0x24a/0x15a0 [ 1185.375494] ? net_generic+0xa0/0x1c0 [ 1185.376844] ? wait_for_completion_io+0x280/0x280 [ 1185.377767] ? devlink_pernet_pre_exit+0x74/0x1c0 [ 1185.378686] devlink_pernet_pre_exit+0x74/0x1c0 [ 1185.379579] ? devlink_nl_cmd_get_dumpit+0x3a0/0x3a0 [ 1185.380557] ? xt_find_table+0xda/0x100 [ 1185.381367] cleanup_net+0x372/0x8e0

changing mode to switchdev trace:

[ 1185.411267] down_write+0x13a/0x150 [ 1185.412029] ? down_write_killable+0x180/0x180 [ 1185.413005] register_netdevice_notifier+0x1e/0x210 [ 1185.414000] mlx5e_rep_bridge_init+0x181/0x360 [mlx5_core] [ 1185.415243] mlx5e_uplink_rep_enable+0x269/0x480 [mlx5_core] [ 1185.416464] ? mlx5e_uplink_rep_disable+0x210/0x210 [mlx5_core] [ 1185.417749] mlx5e_attach_netdev+0x232/0x400 [mlx5_core] [ 1185.418906] mlx5e_netdev_attach_profile+0x15b/0x1e0 [mlx5_core] [ 1185.420172] mlx5e_netdev_change_profile+0x15a/0x1d0 [mlx5_core] [ 1185.421459] mlx5e_vport_rep_load+0x557/0x780 [mlx5_core] [ 1185.422624] ? mlx5e_stats_grp_vport_rep_num_stats+0x10/0x10 [mlx5_core] [ 1185.424006] mlx5_esw_offloads_rep_load+0xdb/0x190 [mlx5_core] [ 1185.425277] esw_offloads_enable+0xd74/0x14a0 [mlx5_core]

Fix this by registering rep bridges for per net netdev notifier instead of global one, which operats on the net namespace without holding the pernet_ops_rwsem.

Fixes: 19e9bfa044f3 ("net/mlx5: Bridge, add offload infrastructure") Signed-off-by: Roi Dayan roid@nvidia.com Reviewed-by: Vlad Buslov vladbu@nvidia.com Signed-off-by: Saeed Mahameed saeedm@nvidia.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)

--- a/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/rep/bridge.c @@ -491,7 +491,7 @@ void mlx5e_rep_bridge_init(struct mlx5e_ }

br_offloads->netdev_nb.notifier_call = mlx5_esw_bridge_switchdev_port_event; - err = register_netdevice_notifier(&br_offloads->netdev_nb); + err = register_netdevice_notifier_net(&init_net, &br_offloads->netdev_nb); if (err) { esw_warn(mdev, "Failed to register bridge offloads netdevice notifier (err=%d)\n", err); @@ -526,7 +526,7 @@ void mlx5e_rep_bridge_cleanup(struct mlx return;

cancel_delayed_work_sync(&br_offloads->update_work); - unregister_netdevice_notifier(&br_offloads->netdev_nb); + unregister_netdevice_notifier_net(&init_net, &br_offloads->netdev_nb); unregister_switchdev_blocking_notifier(&br_offloads->nb_blk); unregister_switchdev_notifier(&br_offloads->nb); destroy_workqueue(br_offloads->wq);

From: Khalid Manaa khalidm@nvidia.com

commit b8d91145ed7cfa046cc07bcfb277465b9d45da73 upstream.

The HW doesn't wrap the CQE.shampo.header_index field according to the headers buffer size, instead it always increases it until reaching overflow of u16 size.

Thus the mlx5e_handle_rx_cqe_mpwrq_shampo handler should mask the CQE header_index field to find the actual header index in the headers buffer.

Fixes: f97d5c2a453e ("net/mlx5e: Add handle SHAMPO cqe support") Signed-off-by: Khalid Manaa khalidm@nvidia.com Reviewed-by: Tariq Toukan tariqt@nvidia.com Signed-off-by: Saeed Mahameed saeedm@nvidia.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h | 5 +++++ drivers/net/ethernet/mellanox/mlx5/core/en_rx.c | 4 ++-- 2 files changed, 7 insertions(+), 2 deletions(-)

--- a/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h +++ b/drivers/net/ethernet/mellanox/mlx5/core/en/txrx.h @@ -167,6 +167,11 @@ static inline u16 mlx5e_txqsq_get_next_p return pi; }

+static inline u16 mlx5e_shampo_get_cqe_header_index(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe) +{ + return be16_to_cpu(cqe->shampo.header_entry_index) & (rq->mpwqe.shampo->hd_per_wq - 1); +} + struct mlx5e_shampo_umr { u16 len; }; --- a/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/en_rx.c @@ -1116,7 +1116,7 @@ static void mlx5e_shampo_update_ipv6_udp static void mlx5e_shampo_update_fin_psh_flags(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe, struct tcphdr *skb_tcp_hd) { - u16 header_index = be16_to_cpu(cqe->shampo.header_entry_index); + u16 header_index = mlx5e_shampo_get_cqe_header_index(rq, cqe); struct tcphdr *last_tcp_hd; void *last_hd_addr;

@@ -1968,7 +1968,7 @@ mlx5e_free_rx_shampo_hd_entry(struct mlx static void mlx5e_handle_rx_cqe_mpwrq_shampo(struct mlx5e_rq *rq, struct mlx5_cqe64 *cqe) { u16 data_bcnt = mpwrq_get_cqe_byte_cnt(cqe) - cqe->shampo.header_size; - u16 header_index = be16_to_cpu(cqe->shampo.header_entry_index); + u16 header_index = mlx5e_shampo_get_cqe_header_index(rq, cqe); u32 wqe_offset = be32_to_cpu(cqe->shampo.data_offset); u16 cstrides = mpwrq_get_cqe_consumed_strides(cqe); u32 data_offset = wqe_offset & (PAGE_SIZE - 1);

From: Maor Dickman maord@nvidia.com

commit d8e5883d694bb053b19c4142a2d1f43a34f6fe2c upstream.

The variable modact is not initialized before used in command modify header allocation which can cause command to fail.

Fix by initializing modact with zeros.

Addresses-Coverity: ("Uninitialized scalar variable") Fixes: 8f1e0b97cc70 ("net/mlx5: E-Switch, Mark miss packets with new chain id mapping") Signed-off-by: Maor Dickman maord@nvidia.com Reviewed-by: Roi Dayan roid@nvidia.com Signed-off-by: Saeed Mahameed saeedm@nvidia.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)

--- a/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c +++ b/drivers/net/ethernet/mellanox/mlx5/core/lib/fs_chains.c @@ -212,7 +212,7 @@ static int create_chain_restore(struct fs_chain *chain) { struct mlx5_eswitch *esw = chain->chains->dev->priv.eswitch; - char modact[MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto)]; + u8 modact[MLX5_UN_SZ_BYTES(set_add_copy_action_in_auto)] = {}; struct mlx5_fs_chains *chains = chain->chains; enum mlx5e_tc_attr_to_reg chain_to_reg; struct mlx5_modify_hdr *mod_hdr;

From: Georgi Valkov gvalkov@abv.bg

commit 63e4b45c82ed1bde979da7052229a4229ce9cabf upstream.

When rx_buf is allocated we need to account for IPHETH_IP_ALIGN, which reduces the usable size by 2 bytes. Otherwise we have 1512 bytes usable instead of 1514, and if we receive more than 1512 bytes, ipheth_rcvbulk_callback is called with status -EOVERFLOW, after which the driver malfunctiones and all communication stops.

Resolves ipheth 2-1:4.2: ipheth_rcvbulk_callback: urb status: -75

Fixes: f33d9e2b48a3 ("usbnet: ipheth: fix connectivity with iOS 14") Signed-off-by: Georgi Valkov gvalkov@abv.bg Tested-by: Jan Kiszka jan.kiszka@siemens.com Link: https://lore.kernel.org/all/B60B8A4B-92A0-49B3-805D-809A2433B46C@abv.bg/ Link: https://lore.kernel.org/all/24851bd2769434a5fc24730dce8e8a984c5a4505.1643699... Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/usb/ipheth.c | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-)

--- a/drivers/net/usb/ipheth.c +++ b/drivers/net/usb/ipheth.c @@ -121,7 +121,7 @@ static int ipheth_alloc_urbs(struct iphe if (tx_buf == NULL) goto free_rx_urb;

- rx_buf = usb_alloc_coherent(iphone->udev, IPHETH_BUF_SIZE, + rx_buf = usb_alloc_coherent(iphone->udev, IPHETH_BUF_SIZE + IPHETH_IP_ALIGN, GFP_KERNEL, &rx_urb->transfer_dma); if (rx_buf == NULL) goto free_tx_buf; @@ -146,7 +146,7 @@ error_nomem:

static void ipheth_free_urbs(struct ipheth_device *iphone) { - usb_free_coherent(iphone->udev, IPHETH_BUF_SIZE, iphone->rx_buf, + usb_free_coherent(iphone->udev, IPHETH_BUF_SIZE + IPHETH_IP_ALIGN, iphone->rx_buf, iphone->rx_urb->transfer_dma); usb_free_coherent(iphone->udev, IPHETH_BUF_SIZE, iphone->tx_buf, iphone->tx_urb->transfer_dma); @@ -317,7 +317,7 @@ static int ipheth_rx_submit(struct iphet

usb_fill_bulk_urb(dev->rx_urb, udev, usb_rcvbulkpipe(udev, dev->bulk_in), - dev->rx_buf, IPHETH_BUF_SIZE, + dev->rx_buf, IPHETH_BUF_SIZE + IPHETH_IP_ALIGN, ipheth_rcvbulk_callback, dev); dev->rx_urb->transfer_flags |= URB_NO_TRANSFER_DMA_MAP;

From: Karen Sornek karen.sornek@intel.com

commit 6533e558c6505e94c3e0ed4281ed5e31ec985f4d upstream.

Fix the crash in kernel while dereferencing the NULL pointer, when the driver is unloaded and simultaneously the VSI rings are being stopped.

The hardware requires 50msec in order to finish RX queues disable. For this purpose the driver spins in mdelay function for the operation to be completed.

For example changing number of queues which requires reset would fail in the following call stack:

1) i40e_prep_for_reset 2) i40e_pf_quiesce_all_vsi 3) i40e_quiesce_vsi 4) i40e_vsi_close 5) i40e_down 6) i40e_vsi_stop_rings 7) i40e_vsi_control_rx -> disable requires the delay of 50msecs 8) continue back in i40e_down function where i40e_clean_tx_ring(vsi->tx_rings[i]) is going to crash

When the driver was spinning vsi_release called i40e_vsi_free_arrays where the vsi->tx_rings resources were freed and the pointer was set to NULL.

Fixes: 5b6d4a7f20b0 ("i40e: Fix crash during removing i40e driver") Signed-off-by: Slawomir Laba slawomirx.laba@intel.com Signed-off-by: Sylwester Dziedziuch sylwesterx.dziedziuch@intel.com Signed-off-by: Karen Sornek karen.sornek@intel.com Tested-by: Gurucharan G gurucharanx.g@intel.com Signed-off-by: Tony Nguyen anthony.l.nguyen@intel.com Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/intel/i40e/i40e.h | 1 + drivers/net/ethernet/intel/i40e/i40e_main.c | 19 ++++++++++++++++++- 2 files changed, 19 insertions(+), 1 deletion(-)

--- a/drivers/net/ethernet/intel/i40e/i40e.h +++ b/drivers/net/ethernet/intel/i40e/i40e.h @@ -144,6 +144,7 @@ enum i40e_state_t { __I40E_VIRTCHNL_OP_PENDING, __I40E_RECOVERY_MODE, __I40E_VF_RESETS_DISABLED, /* disable resets during i40e_remove */ + __I40E_IN_REMOVE, __I40E_VFS_RELEASING, /* This must be last as it determines the size of the BITMAP */ __I40E_STATE_SIZE__, --- a/drivers/net/ethernet/intel/i40e/i40e_main.c +++ b/drivers/net/ethernet/intel/i40e/i40e_main.c @@ -10863,6 +10863,9 @@ static void i40e_reset_and_rebuild(struc bool lock_acquired) { int ret; + + if (test_bit(__I40E_IN_REMOVE, pf->state)) + return; /* Now we wait for GRST to settle out. * We don't have to delete the VEBs or VSIs from the hw switch * because the reset will make them disappear. @@ -12222,6 +12225,8 @@ int i40e_reconfig_rss_queues(struct i40e

vsi->req_queue_pairs = queue_count; i40e_prep_for_reset(pf); + if (test_bit(__I40E_IN_REMOVE, pf->state)) + return pf->alloc_rss_size;

pf->alloc_rss_size = new_rss_size;

@@ -13048,6 +13053,10 @@ static int i40e_xdp_setup(struct i40e_vs if (need_reset) i40e_prep_for_reset(pf);

+ /* VSI shall be deleted in a moment, just return EINVAL */ + if (test_bit(__I40E_IN_REMOVE, pf->state)) + return -EINVAL; + old_prog = xchg(&vsi->xdp_prog, prog);

if (need_reset) { @@ -15938,8 +15947,13 @@ static void i40e_remove(struct pci_dev * i40e_write_rx_ctl(hw, I40E_PFQF_HENA(0), 0); i40e_write_rx_ctl(hw, I40E_PFQF_HENA(1), 0);

- while (test_bit(__I40E_RESET_RECOVERY_PENDING, pf->state)) + /* Grab __I40E_RESET_RECOVERY_PENDING and set __I40E_IN_REMOVE + * flags, once they are set, i40e_rebuild should not be called as + * i40e_prep_for_reset always returns early. + */ + while (test_and_set_bit(__I40E_RESET_RECOVERY_PENDING, pf->state)) usleep_range(1000, 2000); + set_bit(__I40E_IN_REMOVE, pf->state);

if (pf->flags & I40E_FLAG_SRIOV_ENABLED) { set_bit(__I40E_VF_RESETS_DISABLED, pf->state); @@ -16138,6 +16152,9 @@ static void i40e_pci_error_reset_done(st { struct i40e_pf *pf = pci_get_drvdata(pdev);

+ if (test_bit(__I40E_IN_REMOVE, pf->state)) + return; + i40e_reset_and_rebuild(pf, false, false); }

From: Shyam Sundar S K Shyam-sundar.S-k@amd.com

commit 5aac9108a180fc06e28d4e7fb00247ce603b72ee upstream.

There will be BUG_ON() triggered in include/linux/skbuff.h leading to intermittent kernel panic, when the skb length underflow is detected.

Fix this by dropping the packet if such length underflows are seen because of inconsistencies in the hardware descriptors.

Fixes: 622c36f143fc ("amd-xgbe: Fix jumbo MTU processing on newer hardware") Suggested-by: Tom Lendacky thomas.lendacky@amd.com Signed-off-by: Shyam Sundar S K Shyam-sundar.S-k@amd.com Acked-by: Tom Lendacky thomas.lendacky@amd.com Link: https://lore.kernel.org/r/20220127092003.2812745-1-Shyam-sundar.S-k@amd.com Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ethernet/amd/xgbe/xgbe-drv.c | 12 +++++++++++- 1 file changed, 11 insertions(+), 1 deletion(-)

--- a/drivers/net/ethernet/amd/xgbe/xgbe-drv.c +++ b/drivers/net/ethernet/amd/xgbe/xgbe-drv.c @@ -2555,6 +2555,14 @@ read_again: buf2_len = xgbe_rx_buf2_len(rdata, packet, len); len += buf2_len;

+ if (buf2_len > rdata->rx.buf.dma_len) { + /* Hardware inconsistency within the descriptors + * that has resulted in a length underflow. + */ + error = 1; + goto skip_data; + } + if (!skb) { skb = xgbe_create_skb(pdata, napi, rdata, buf1_len); @@ -2584,8 +2592,10 @@ skip_data: if (!last || context_next) goto read_again;

- if (!skb) + if (!skb || error) { + dev_kfree_skb(skb); goto next_packet; + }

/* Be sure we don't exceed the configured MTU */ max_len = netdev->mtu + ETH_HLEN;

From: Eric Dumazet edumazet@google.com

commit 04c2a47ffb13c29778e2a14e414ad4cb5a5db4b5 upstream.

Whenever tc_new_tfilter() jumps back to replay: label, we need to make sure @q and @chain local variables are cleared again, or risk use-after-free as in [1]

For consistency, apply the same fix in tc_ctl_chain()

BUG: KASAN: use-after-free in mini_qdisc_pair_swap+0x1b9/0x1f0 net/sched/sch_generic.c:1581 Write of size 8 at addr ffff8880985c4b08 by task syz-executor.4/1945

CPU: 0 PID: 1945 Comm: syz-executor.4 Not tainted 5.17.0-rc1-syzkaller-00495-gff58831fa02d #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011 Call Trace: <TASK> __dump_stack lib/dump_stack.c:88 [inline] dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106 print_address_description.constprop.0.cold+0x8d/0x336 mm/kasan/report.c:255 __kasan_report mm/kasan/report.c:442 [inline] kasan_report.cold+0x83/0xdf mm/kasan/report.c:459 mini_qdisc_pair_swap+0x1b9/0x1f0 net/sched/sch_generic.c:1581 tcf_chain_head_change_item net/sched/cls_api.c:372 [inline] tcf_chain0_head_change.isra.0+0xb9/0x120 net/sched/cls_api.c:386 tcf_chain_tp_insert net/sched/cls_api.c:1657 [inline] tcf_chain_tp_insert_unique net/sched/cls_api.c:1707 [inline] tc_new_tfilter+0x1e67/0x2350 net/sched/cls_api.c:2086 rtnetlink_rcv_msg+0x80d/0xb80 net/core/rtnetlink.c:5583 netlink_rcv_skb+0x153/0x420 net/netlink/af_netlink.c:2494 netlink_unicast_kernel net/netlink/af_netlink.c:1317 [inline] netlink_unicast+0x539/0x7e0 net/netlink/af_netlink.c:1343 netlink_sendmsg+0x904/0xe00 net/netlink/af_netlink.c:1919 sock_sendmsg_nosec net/socket.c:705 [inline] sock_sendmsg+0xcf/0x120 net/socket.c:725 ____sys_sendmsg+0x331/0x810 net/socket.c:2413 ___sys_sendmsg+0xf3/0x170 net/socket.c:2467 __sys_sendmmsg+0x195/0x470 net/socket.c:2553 __do_sys_sendmmsg net/socket.c:2582 [inline] __se_sys_sendmmsg net/socket.c:2579 [inline] __x64_sys_sendmmsg+0x99/0x100 net/socket.c:2579 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae RIP: 0033:0x7f2647172059 Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48 RSP: 002b:00007f2645aa5168 EFLAGS: 00000246 ORIG_RAX: 0000000000000133 RAX: ffffffffffffffda RBX: 00007f2647285100 RCX: 00007f2647172059 RDX: 040000000000009f RSI: 00000000200002c0 RDI: 0000000000000006 RBP: 00007f26471cc08d R08: 0000000000000000 R09: 0000000000000000 R10: 9e00000000000000 R11: 0000000000000246 R12: 0000000000000000 R13: 00007fffb3f7f02f R14: 00007f2645aa5300 R15: 0000000000022000 </TASK>

Allocated by task 1944: kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38 kasan_set_track mm/kasan/common.c:45 [inline] set_alloc_info mm/kasan/common.c:436 [inline] ____kasan_kmalloc mm/kasan/common.c:515 [inline] ____kasan_kmalloc mm/kasan/common.c:474 [inline] __kasan_kmalloc+0xa9/0xd0 mm/kasan/common.c:524 kmalloc_node include/linux/slab.h:604 [inline] kzalloc_node include/linux/slab.h:726 [inline] qdisc_alloc+0xac/0xa10 net/sched/sch_generic.c:941 qdisc_create.constprop.0+0xce/0x10f0 net/sched/sch_api.c:1211 tc_modify_qdisc+0x4c5/0x1980 net/sched/sch_api.c:1660 rtnetlink_rcv_msg+0x413/0xb80 net/core/rtnetlink.c:5592 netlink_rcv_skb+0x153/0x420 net/netlink/af_netlink.c:2494 netlink_unicast_kernel net/netlink/af_netlink.c:1317 [inline] netlink_unicast+0x539/0x7e0 net/netlink/af_netlink.c:1343 netlink_sendmsg+0x904/0xe00 net/netlink/af_netlink.c:1919 sock_sendmsg_nosec net/socket.c:705 [inline] sock_sendmsg+0xcf/0x120 net/socket.c:725 ____sys_sendmsg+0x331/0x810 net/socket.c:2413 ___sys_sendmsg+0xf3/0x170 net/socket.c:2467 __sys_sendmmsg+0x195/0x470 net/socket.c:2553 __do_sys_sendmmsg net/socket.c:2582 [inline] __se_sys_sendmmsg net/socket.c:2579 [inline] __x64_sys_sendmmsg+0x99/0x100 net/socket.c:2579 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae

Freed by task 3609: kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38 kasan_set_track+0x21/0x30 mm/kasan/common.c:45 kasan_set_free_info+0x20/0x30 mm/kasan/generic.c:370 ____kasan_slab_free mm/kasan/common.c:366 [inline] ____kasan_slab_free+0x130/0x160 mm/kasan/common.c:328 kasan_slab_free include/linux/kasan.h:236 [inline] slab_free_hook mm/slub.c:1728 [inline] slab_free_freelist_hook+0x8b/0x1c0 mm/slub.c:1754 slab_free mm/slub.c:3509 [inline] kfree+0xcb/0x280 mm/slub.c:4562 rcu_do_batch kernel/rcu/tree.c:2527 [inline] rcu_core+0x7b8/0x1540 kernel/rcu/tree.c:2778 __do_softirq+0x29b/0x9c2 kernel/softirq.c:558

Last potentially related work creation: kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38 __kasan_record_aux_stack+0xbe/0xd0 mm/kasan/generic.c:348 __call_rcu kernel/rcu/tree.c:3026 [inline] call_rcu+0xb1/0x740 kernel/rcu/tree.c:3106 qdisc_put_unlocked+0x6f/0x90 net/sched/sch_generic.c:1109 tcf_block_release+0x86/0x90 net/sched/cls_api.c:1238 tc_new_tfilter+0xc0d/0x2350 net/sched/cls_api.c:2148 rtnetlink_rcv_msg+0x80d/0xb80 net/core/rtnetlink.c:5583 netlink_rcv_skb+0x153/0x420 net/netlink/af_netlink.c:2494 netlink_unicast_kernel net/netlink/af_netlink.c:1317 [inline] netlink_unicast+0x539/0x7e0 net/netlink/af_netlink.c:1343 netlink_sendmsg+0x904/0xe00 net/netlink/af_netlink.c:1919 sock_sendmsg_nosec net/socket.c:705 [inline] sock_sendmsg+0xcf/0x120 net/socket.c:725 ____sys_sendmsg+0x331/0x810 net/socket.c:2413 ___sys_sendmsg+0xf3/0x170 net/socket.c:2467 __sys_sendmmsg+0x195/0x470 net/socket.c:2553 __do_sys_sendmmsg net/socket.c:2582 [inline] __se_sys_sendmmsg net/socket.c:2579 [inline] __x64_sys_sendmmsg+0x99/0x100 net/socket.c:2579 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae

The buggy address belongs to the object at ffff8880985c4800 which belongs to the cache kmalloc-1k of size 1024 The buggy address is located 776 bytes inside of 1024-byte region [ffff8880985c4800, ffff8880985c4c00) The buggy address belongs to the page: page:ffffea0002617000 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x985c0 head:ffffea0002617000 order:3 compound_mapcount:0 compound_pincount:0 flags: 0xfff00000010200(slab|head|node=0|zone=1|lastcpupid=0x7ff) raw: 00fff00000010200 0000000000000000 dead000000000122 ffff888010c41dc0 raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000 page dumped because: kasan: bad access detected page_owner tracks the page as allocated page last allocated via order 3, migratetype Unmovable, gfp_mask 0x1d20c0(__GFP_IO|__GFP_FS|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP|__GFP_NOMEMALLOC|__GFP_HARDWALL), pid 1941, ts 1038999441284, free_ts 1033444432829 prep_new_page mm/page_alloc.c:2434 [inline] get_page_from_freelist+0xa72/0x2f50 mm/page_alloc.c:4165 __alloc_pages+0x1b2/0x500 mm/page_alloc.c:5389 alloc_pages+0x1aa/0x310 mm/mempolicy.c:2271 alloc_slab_page mm/slub.c:1799 [inline] allocate_slab mm/slub.c:1944 [inline] new_slab+0x28a/0x3b0 mm/slub.c:2004 ___slab_alloc+0x87c/0xe90 mm/slub.c:3018 __slab_alloc.constprop.0+0x4d/0xa0 mm/slub.c:3105 slab_alloc_node mm/slub.c:3196 [inline] slab_alloc mm/slub.c:3238 [inline] __kmalloc+0x2fb/0x340 mm/slub.c:4420 kmalloc include/linux/slab.h:586 [inline] kzalloc include/linux/slab.h:715 [inline] __register_sysctl_table+0x112/0x1090 fs/proc/proc_sysctl.c:1335 neigh_sysctl_register+0x2c8/0x5e0 net/core/neighbour.c:3787 devinet_sysctl_register+0xb1/0x230 net/ipv4/devinet.c:2618 inetdev_init+0x286/0x580 net/ipv4/devinet.c:278 inetdev_event+0xa8a/0x15d0 net/ipv4/devinet.c:1532 notifier_call_chain+0xb5/0x200 kernel/notifier.c:84 call_netdevice_notifiers_info+0xb5/0x130 net/core/dev.c:1919 call_netdevice_notifiers_extack net/core/dev.c:1931 [inline] call_netdevice_notifiers net/core/dev.c:1945 [inline] register_netdevice+0x1073/0x1500 net/core/dev.c:9698 veth_newlink+0x59c/0xa90 drivers/net/veth.c:1722 page last free stack trace: reset_page_owner include/linux/page_owner.h:24 [inline] free_pages_prepare mm/page_alloc.c:1352 [inline] free_pcp_prepare+0x374/0x870 mm/page_alloc.c:1404 free_unref_page_prepare mm/page_alloc.c:3325 [inline] free_unref_page+0x19/0x690 mm/page_alloc.c:3404 release_pages+0x748/0x1220 mm/swap.c:956 tlb_batch_pages_flush mm/mmu_gather.c:50 [inline] tlb_flush_mmu_free mm/mmu_gather.c:243 [inline] tlb_flush_mmu+0xe9/0x6b0 mm/mmu_gather.c:250 zap_pte_range mm/memory.c:1441 [inline] zap_pmd_range mm/memory.c:1490 [inline] zap_pud_range mm/memory.c:1519 [inline] zap_p4d_range mm/memory.c:1540 [inline] unmap_page_range+0x1d1d/0x2a30 mm/memory.c:1561 unmap_single_vma+0x198/0x310 mm/memory.c:1606 unmap_vmas+0x16b/0x2f0 mm/memory.c:1638 exit_mmap+0x201/0x670 mm/mmap.c:3178 __mmput+0x122/0x4b0 kernel/fork.c:1114 mmput+0x56/0x60 kernel/fork.c:1135 exit_mm kernel/exit.c:507 [inline] do_exit+0xa3c/0x2a30 kernel/exit.c:793 do_group_exit+0xd2/0x2f0 kernel/exit.c:935 __do_sys_exit_group kernel/exit.c:946 [inline] __se_sys_exit_group kernel/exit.c:944 [inline] __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:944 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae

Memory state around the buggy address: ffff8880985c4a00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8880985c4a80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb

ffff8880985c4b00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb

^ ffff8880985c4b80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb ffff8880985c4c00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc

Fixes: 470502de5bdb ("net: sched: unlock rules update API") Signed-off-by: Eric Dumazet edumazet@google.com Cc: Vlad Buslov vladbu@mellanox.com Cc: Jiri Pirko jiri@mellanox.com Cc: Cong Wang xiyou.wangcong@gmail.com Reported-by: syzbot syzkaller@googlegroups.com Link: https://lore.kernel.org/r/20220131172018.3704490-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- net/sched/cls_api.c | 11 +++++++---- 1 file changed, 7 insertions(+), 4 deletions(-)

--- a/net/sched/cls_api.c +++ b/net/sched/cls_api.c @@ -1945,9 +1945,9 @@ static int tc_new_tfilter(struct sk_buff bool prio_allocate; u32 parent; u32 chain_index; - struct Qdisc *q = NULL; + struct Qdisc *q; struct tcf_chain_info chain_info; - struct tcf_chain *chain = NULL; + struct tcf_chain *chain; struct tcf_block *block; struct tcf_proto *tp; unsigned long cl; @@ -1976,6 +1976,8 @@ replay: tp = NULL; cl = 0; block = NULL; + q = NULL; + chain = NULL; flags = 0;

if (prio == 0) { @@ -2798,8 +2800,8 @@ static int tc_ctl_chain(struct sk_buff * struct tcmsg *t; u32 parent; u32 chain_index; - struct Qdisc *q = NULL; - struct tcf_chain *chain = NULL; + struct Qdisc *q; + struct tcf_chain *chain; struct tcf_block *block; unsigned long cl; int err; @@ -2809,6 +2811,7 @@ static int tc_ctl_chain(struct sk_buff * return -EPERM;

replay: + q = NULL; err = nlmsg_parse_deprecated(n, sizeof(*t), tca, TCA_MAX, rtm_tca_policy, extack); if (err < 0)

From: Alex Elder elder@linaro.org

commit 34a081761e4e3c35381cbfad609ebae2962fe2f8 upstream.

In some cases, the IPA hardware needs to request the always-on subsystem (AOSS) to coordinate with the IPA microcontroller to retain IPA register values at power collapse. This is done by issuing a QMP request to the AOSS microcontroller. A similar request ondoes that request.

We must get and hold the "QMP" handle early, because we might get back EPROBE_DEFER for that. But the actual request should be sent while we know the IPA clock is active, and when we know the microcontroller is operational.

Fixes: 1aac309d3207 ("net: ipa: use autosuspend") Signed-off-by: Alex Elder elder@linaro.org Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- drivers/net/ipa/ipa_power.c | 52 ++++++++++++++++++++++++++++++++++++++++++++ drivers/net/ipa/ipa_power.h | 7 +++++ drivers/net/ipa/ipa_uc.c | 5 ++++ 3 files changed, 64 insertions(+)

--- a/drivers/net/ipa/ipa_power.c +++ b/drivers/net/ipa/ipa_power.c @@ -11,6 +11,8 @@ #include <linux/pm_runtime.h> #include <linux/bitops.h>

+#include "linux/soc/qcom/qcom_aoss.h" + #include "ipa.h" #include "ipa_power.h" #include "ipa_endpoint.h" @@ -64,6 +66,7 @@ enum ipa_power_flag { * struct ipa_power - IPA power management information * @dev: IPA device pointer * @core: IPA core clock + * @qmp: QMP handle for AOSS communication * @spinlock: Protects modem TX queue enable/disable * @flags: Boolean state flags * @interconnect_count: Number of elements in interconnect[] @@ -72,6 +75,7 @@ enum ipa_power_flag { struct ipa_power { struct device *dev; struct clk *core; + struct qmp *qmp; spinlock_t spinlock; /* used with STOPPED/STARTED power flags */ DECLARE_BITMAP(flags, IPA_POWER_FLAG_COUNT); u32 interconnect_count; @@ -382,6 +386,47 @@ void ipa_power_modem_queue_active(struct clear_bit(IPA_POWER_FLAG_STARTED, ipa->power->flags); }

+static int ipa_power_retention_init(struct ipa_power *power) +{ + struct qmp *qmp = qmp_get(power->dev); + + if (IS_ERR(qmp)) { + if (PTR_ERR(qmp) == -EPROBE_DEFER) + return -EPROBE_DEFER; + + /* We assume any other error means it's not defined/needed */ + qmp = NULL; + } + power->qmp = qmp; + + return 0; +} + +static void ipa_power_retention_exit(struct ipa_power *power) +{ + qmp_put(power->qmp); + power->qmp = NULL; +} + +/* Control register retention on power collapse */ +void ipa_power_retention(struct ipa *ipa, bool enable) +{ + static const char fmt[] = "{ class: bcm, res: ipa_pc, val: %c }"; + struct ipa_power *power = ipa->power; + char buf[36]; /* Exactly enough for fmt[]; size a multiple of 4 */ + int ret; + + if (!power->qmp) + return; /* Not needed on this platform */ + + (void)snprintf(buf, sizeof(buf), fmt, enable ? '1' : '0'); + + ret = qmp_send(power->qmp, buf, sizeof(buf)); + if (ret) + dev_err(power->dev, "error %d sending QMP %sable request\n", + ret, enable ? "en" : "dis"); +} + int ipa_power_setup(struct ipa *ipa) { int ret; @@ -438,12 +483,18 @@ ipa_power_init(struct device *dev, const if (ret) goto err_kfree;

+ ret = ipa_power_retention_init(power); + if (ret) + goto err_interconnect_exit; + pm_runtime_set_autosuspend_delay(dev, IPA_AUTOSUSPEND_DELAY); pm_runtime_use_autosuspend(dev); pm_runtime_enable(dev);

return power;

+err_interconnect_exit: + ipa_interconnect_exit(power); err_kfree: kfree(power); err_clk_put: @@ -460,6 +511,7 @@ void ipa_power_exit(struct ipa_power *po

pm_runtime_disable(dev); pm_runtime_dont_use_autosuspend(dev); + ipa_power_retention_exit(power); ipa_interconnect_exit(power); kfree(power); clk_put(clk); --- a/drivers/net/ipa/ipa_power.h +++ b/drivers/net/ipa/ipa_power.h @@ -41,6 +41,13 @@ void ipa_power_modem_queue_wake(struct i void ipa_power_modem_queue_active(struct ipa *ipa);

/** + * ipa_power_retention() - Control register retention on power collapse + * @ipa: IPA pointer + * @enable: Whether retention should be enabled or disabled + */ +void ipa_power_retention(struct ipa *ipa, bool enable); + +/** * ipa_power_setup() - Set up IPA power management * @ipa: IPA pointer * --- a/drivers/net/ipa/ipa_uc.c +++ b/drivers/net/ipa/ipa_uc.c @@ -11,6 +11,7 @@

#include "ipa.h" #include "ipa_uc.h" +#include "ipa_power.h"

/** * DOC: The IPA embedded microcontroller @@ -154,6 +155,7 @@ static void ipa_uc_response_hdlr(struct case IPA_UC_RESPONSE_INIT_COMPLETED: if (ipa->uc_powered) { ipa->uc_loaded = true; + ipa_power_retention(ipa, true); pm_runtime_mark_last_busy(dev); (void)pm_runtime_put_autosuspend(dev); ipa->uc_powered = false; @@ -184,6 +186,9 @@ void ipa_uc_deconfig(struct ipa *ipa)

ipa_interrupt_remove(ipa->interrupt, IPA_IRQ_UC_1); ipa_interrupt_remove(ipa->interrupt, IPA_IRQ_UC_0); + if (ipa->uc_loaded) + ipa_power_retention(ipa, false); + if (!ipa->uc_powered) return;

From: Tianchen Ding dtcccc@linux.alibaba.com

commit c80d401c52a2d1baf2a5afeb06f0ffe678e56d23 upstream.

subparts_cpus should be limited as a subset of cpus_allowed, but it is updated wrongly by using cpumask_andnot(). Use cpumask_and() instead to fix it.

Fixes: ee8dde0cd2ce ("cpuset: Add new v2 cpuset.sched.partition flag") Signed-off-by: Tianchen Ding dtcccc@linux.alibaba.com Reviewed-by: Waiman Long longman@redhat.com Signed-off-by: Tejun Heo tj@kernel.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- kernel/cgroup/cpuset.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)

--- a/kernel/cgroup/cpuset.c +++ b/kernel/cgroup/cpuset.c @@ -1615,8 +1615,7 @@ static int update_cpumask(struct cpuset * Make sure that subparts_cpus is a subset of cpus_allowed. */ if (cs->nr_subparts_cpus) { - cpumask_andnot(cs->subparts_cpus, cs->subparts_cpus, - cs->cpus_allowed); + cpumask_and(cs->subparts_cpus, cs->subparts_cpus, cs->cpus_allowed); cs->nr_subparts_cpus = cpumask_weight(cs->subparts_cpus); } spin_unlock_irq(&callback_lock);

From: Eric Dumazet edumazet@google.com

commit e42e70ad6ae2ae511a6143d2e8da929366e58bd9 upstream.

When packet_setsockopt( PACKET_FANOUT_DATA ) reads po->fanout, no lock is held, meaning that another thread can change po->fanout.

Given that po->fanout can only be set once during the socket lifetime (it is only cleared from fanout_release()), we can use READ_ONCE()/WRITE_ONCE() to document the race.

BUG: KCSAN: data-race in packet_setsockopt / packet_setsockopt

write to 0xffff88813ae8e300 of 8 bytes by task 14653 on cpu 0: fanout_add net/packet/af_packet.c:1791 [inline] packet_setsockopt+0x22fe/0x24a0 net/packet/af_packet.c:3931 __sys_setsockopt+0x209/0x2a0 net/socket.c:2180 __do_sys_setsockopt net/socket.c:2191 [inline] __se_sys_setsockopt net/socket.c:2188 [inline] __x64_sys_setsockopt+0x62/0x70 net/socket.c:2188 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae

read to 0xffff88813ae8e300 of 8 bytes by task 14654 on cpu 1: packet_setsockopt+0x691/0x24a0 net/packet/af_packet.c:3935 __sys_setsockopt+0x209/0x2a0 net/socket.c:2180 __do_sys_setsockopt net/socket.c:2191 [inline] __se_sys_setsockopt net/socket.c:2188 [inline] __x64_sys_setsockopt+0x62/0x70 net/socket.c:2188 do_syscall_x64 arch/x86/entry/common.c:50 [inline] do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80 entry_SYSCALL_64_after_hwframe+0x44/0xae

value changed: 0x0000000000000000 -> 0xffff888106f8c000

Reported by Kernel Concurrency Sanitizer on: CPU: 1 PID: 14654 Comm: syz-executor.3 Not tainted 5.16.0-syzkaller #0 Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011

Fixes: 47dceb8ecdc1 ("packet: add classic BPF fanout mode") Signed-off-by: Eric Dumazet edumazet@google.com Cc: Willem de Bruijn willemb@google.com Reported-by: syzbot syzkaller@googlegroups.com Link: https://lore.kernel.org/r/20220201022358.330621-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- net/packet/af_packet.c | 8 ++++++-- 1 file changed, 6 insertions(+), 2 deletions(-)

--- a/net/packet/af_packet.c +++ b/net/packet/af_packet.c @@ -1788,7 +1788,10 @@ static int fanout_add(struct sock *sk, s err = -ENOSPC; if (refcount_read(&match->sk_ref) < match->max_num_members) { __dev_remove_pack(&po->prot_hook); - po->fanout = match; + + /* Paired with packet_setsockopt(PACKET_FANOUT_DATA) */ + WRITE_ONCE(po->fanout, match); + po->rollover = rollover; rollover = NULL; refcount_set(&match->sk_ref, refcount_read(&match->sk_ref) + 1); @@ -3941,7 +3944,8 @@ packet_setsockopt(struct socket *sock, i } case PACKET_FANOUT_DATA: { - if (!po->fanout) + /* Paired with the WRITE_ONCE() in fanout_add() */ + if (!READ_ONCE(po->fanout)) return -EINVAL;

return fanout_set_data(po, optval, optlen);

From: Eric Dumazet edumazet@google.com

commit b67985be400969578d4d4b17299714c0e5d2c07b upstream.

tcp_shift_skb_data() might collapse three packets into a larger one.

P_A, P_B, P_C -> P_ABC

Historically, it used a single tcp_skb_can_collapse_to(P_A) call, because it was enough.

In commit 85712484110d ("tcp: coalesce/collapse must respect MPTCP extensions"), this call was replaced by a call to tcp_skb_can_collapse(P_A, P_B)

But the now needed test over P_C has been missed.

This probably broke MPTCP.

Then later, commit 9b65b17db723 ("net: avoid double accounting for pure zerocopy skbs") added an extra condition to tcp_skb_can_collapse(), but the missing call from tcp_shift_skb_data() is also breaking TCP zerocopy, because P_A and P_C might have different skb_zcopy_pure() status.

Fixes: 85712484110d ("tcp: coalesce/collapse must respect MPTCP extensions") Fixes: 9b65b17db723 ("net: avoid double accounting for pure zerocopy skbs") Signed-off-by: Eric Dumazet edumazet@google.com Cc: Mat Martineau mathew.j.martineau@linux.intel.com Cc: Talal Ahmad talalahmad@google.com Cc: Arjun Roy arjunroy@google.com Cc: Willem de Bruijn willemb@google.com Acked-by: Soheil Hassas Yeganeh soheil@google.com Acked-by: Paolo Abeni pabeni@redhat.com Link: https://lore.kernel.org/r/20220201184640.756716-1-eric.dumazet@gmail.com Signed-off-by: Jakub Kicinski kuba@kernel.org Signed-off-by: Greg Kroah-Hartman gregkh@linuxfoundation.org --- net/ipv4/tcp_input.c | 2 ++ 1 file changed, 2 insertions(+)

--- a/net/ipv4/tcp_input.c +++ b/net/ipv4/tcp_input.c @@ -1660,6 +1660,8 @@ static struct sk_buff *tcp_shift_skb_dat (mss != tcp_skb_seglen(skb))) goto out;

+ if (!tcp_skb_can_collapse(prev, skb)) + goto out; len = skb->len; pcount = tcp_skb_pcount(skb); if (tcp_skb_shift(prev, skb, pcount, len))