A reference underflow is found in TLS handshake subsystem that causes a
direct use-after-free. Part of the crash log is like below:
[ 2.022114] ------------[ cut here ]------------
[ 2.022193] refcount_t: underflow; use-after-free.
[ 2.022288] WARNING: CPU: 0 PID: 60 at lib/refcount.c:28 refcount_warn_saturate+0xbe/0x110
[ 2.022432] Modules linked in:
[ 2.022848] RIP: 0010:refcount_warn_saturate+0xbe/0x110
[ 2.023231] RSP: 0018:ffffc900001bfe18 EFLAGS: 00000286
[ 2.023325] RAX: 0000000000000000 RBX: 0000000000000007 RCX: 00000000ffffdfff
[ 2.023438] RDX: 0000000000000000 RSI: 00000000ffffffea RDI: 0000000000000001
[ 2.023555] RBP: ffff888004c20098 R08: ffffffff82b392c8 R09: 00000000ffffdfff
[ 2.023693] R10: ffffffff82a592e0 R11: ffffffff82b092e0 R12: ffff888004c200d8
[ 2.023813] R13: 0000000000000000 R14: ffff888004c20000 R15: ffffc90000013ca8
[ 2.023930] FS: 0000000000000000(0000) GS:ffff88807dc00000(0000) knlGS:0000000000000000
[ 2.024062] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 2.024161] CR2: ffff888003601000 CR3: 0000000002a2e000 CR4: 00000000000006f0
[ 2.024275] Call Trace:
[ 2.024322] <TASK>
[ 2.024367] ? __warn+0x7f/0x130
[ 2.024430] ? refcount_warn_saturate+0xbe/0x110
[ 2.024513] ? report_bug+0x199/0x1b0
[ 2.024585] ? handle_bug+0x3c/0x70
[ 2.024676] ? exc_invalid_op+0x18/0x70
[ 2.024750] ? asm_exc_invalid_op+0x1a/0x20
[ 2.024830] ? refcount_warn_saturate+0xbe/0x110
[ 2.024916] ? refcount_warn_saturate+0xbe/0x110
[ 2.024998] __tcp_close+0x2f4/0x3d0
[ 2.025065] ? __pfx_kunit_generic_run_threadfn_adapter+0x10/0x10
[ 2.025168] tcp_close+0x1f/0x70
[ 2.025231] inet_release+0x33/0x60
[ 2.025297] sock_release+0x1f/0x80
[ 2.025361] handshake_req_cancel_test2+0x100/0x2d0
[ 2.025457] kunit_try_run_case+0x4c/0xa0
[ 2.025532] kunit_generic_run_threadfn_adapter+0x15/0x20
[ 2.025644] kthread+0xe1/0x110
[ 2.025708] ? __pfx_kthread+0x10/0x10
[ 2.025780] ret_from_fork+0x2c/0x50
One can enable CONFIG_NET_HANDSHAKE_KUNIT_TEST config to reproduce above
crash.
The root cause of this bug is that the commit 1ce77c998f
("net/handshake: Unpin sock->file if a handshake is cancelled") adds one
additional fput() function. That patch claims that the fput() is used to
enable sock->file to be freed even when user space never calls DONE.
However, it seems that the intended DONE routine will never give an
additional fput() of ths sock->file. The existing two of them are just
used to balance the reference added in sockfd_lookup().
This patch revert the mentioned commit to avoid the use-after-free. The
patched kernel could successfully pass the KUNIT test and boot to shell.
[ 0.733613] # Subtest: Handshake API tests
[ 0.734029] 1..11
[ 0.734255] KTAP version 1
[ 0.734542] # Subtest: req_alloc API fuzzing
[ 0.736104] ok 1 handshake_req_alloc NULL proto
[ 0.736114] ok 2 handshake_req_alloc CLASS_NONE
[ 0.736559] ok 3 handshake_req_alloc CLASS_MAX
[ 0.737020] ok 4 handshake_req_alloc no callbacks
[ 0.737488] ok 5 handshake_req_alloc no done callback
[ 0.737988] ok 6 handshake_req_alloc excessive privsize
[ 0.738529] ok 7 handshake_req_alloc all good
[ 0.739036] # req_alloc API fuzzing: pass:7 fail:0 skip:0 total:7
[ 0.739444] ok 1 req_alloc API fuzzing
[ 0.740065] ok 2 req_submit NULL req arg
[ 0.740436] ok 3 req_submit NULL sock arg
[ 0.740834] ok 4 req_submit NULL sock->file
[ 0.741236] ok 5 req_lookup works
[ 0.741621] ok 6 req_submit max pending
[ 0.741974] ok 7 req_submit multiple
[ 0.742382] ok 8 req_cancel before accept
[ 0.742764] ok 9 req_cancel after accept
[ 0.743151] ok 10 req_cancel after done
[ 0.743510] ok 11 req_destroy works
[ 0.743882] # Handshake API tests: pass:11 fail:0 skip:0 total:11
[ 0.744205] # Totals: pass:17 fail:0 skip:0 total:17
Acked-by: Chuck Lever <chuck.lever@oracle.com>
Fixes: 1ce77c998f ("net/handshake: Unpin sock->file if a handshake is cancelled")
Signed-off-by: Lin Ma <linma@zju.edu.cn>
Link: https://lore.kernel.org/r/20230613083204.633896-1-linma@zju.edu.cn
Link: https://lore.kernel.org/r/20230614015249.987448-1-linma@zju.edu.cn
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
* fix fragmentation for multi-link related elements
* fix callback copy/paste error
* fix multi-link locking
* remove double-locking of wiphy mutex
* transmit only on active links, not all
* activate links in the correct order
* don't remove links that weren't added
* disable soft-IRQs for LQ lock in iwlwifi
-----BEGIN PGP SIGNATURE-----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=IPqO
-----END PGP SIGNATURE-----
Merge tag 'wireless-2023-06-14' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless
Johannes Berg says:
====================
A couple of straggler fixes, mostly in the stack:
- fix fragmentation for multi-link related elements
- fix callback copy/paste error
- fix multi-link locking
- remove double-locking of wiphy mutex
- transmit only on active links, not all
- activate links in the correct order
- don't remove links that weren't added
- disable soft-IRQs for LQ lock in iwlwifi
* tag 'wireless-2023-06-14' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless:
wifi: iwlwifi: mvm: spin_lock_bh() to fix lockdep regression
wifi: mac80211: fragment per STA profile correctly
wifi: mac80211: Use active_links instead of valid_links in Tx
wifi: cfg80211: remove links only on AP
wifi: mac80211: take lock before setting vif links
wifi: cfg80211: fix link del callback to call correct handler
wifi: mac80211: fix link activation settings order
wifi: cfg80211: fix double lock bug in reg_wdev_chan_valid()
====================
Link: https://lore.kernel.org/r/20230614075502.11765-1-johannes@sipsolutions.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
A recent patch added a call to ext4_error() which is problematic since
some callers of the ext4_get_group_info() function may be holding a
spinlock, whereas ext4_error() must never be called in atomic context.
This triggered a report from Syzbot: "BUG: sleeping function called from
invalid context in ext4_update_super" (see the link below).
Therefore, drop the call to ext4_error() from ext4_get_group_info(). In
the meantime use eight characters tabs instead of nine characters ones.
Reported-by: syzbot+4acc7d910e617b360859@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/all/00000000000070575805fdc6cdb2@google.com/
Fixes: 5354b2af34 ("ext4: allow ext4_get_group_info() to fail")
Suggested-by: Theodore Ts'o <tytso@mit.edu>
Signed-off-by: Fabio M. De Francesco <fmdefrancesco@gmail.com>
Link: https://lore.kernel.org/r/20230614100446.14337-1-fmdefrancesco@gmail.com
This reverts commit ad3f09be6c.
The reverted commit was intended to simpfy the code to get group
descriptor block number in non-meta block group by assuming
s_gdb_count is block number used for all non-meta block group descriptors.
However s_gdb_count is block number used for all meta *and* non-meta
group descriptors. So s_gdb_group will be > actual group descriptor block
number used for all non-meta block group which should be "total non-meta
block group" / "group descriptors per block", e.g. s_first_meta_bg.
Signed-off-by: Kemeng Shi <shikemeng@huaweicloud.com>
Link: https://lore.kernel.org/r/20230613225025.3859522-1-shikemeng@huaweicloud.com
Fixes: ad3f09be6c ("ext4: remove unnecessary check in ext4_bg_num_gdb_nometa")
Cc: stable@kernel.org
Signed-off-by: Theodore Ts'o <tytso@mit.edu>
A recent patch replaced a tasklet execution of cq->comp_handler by a
direct call. While this made sense it let changes to cq->notify state be
unprotected and assumed that the cq completion machinery and the ulp done
callbacks were reentrant. The result is that in some cases completion
events can be lost. This patch moves the cq->comp_handler call inside of
the spinlock in rxe_cq_post which solves both issues. This is compatible
with the matching code in the request notify verb.
Fixes: 78b26a3353 ("RDMA/rxe: Remove tasklet call from rxe_cq.c")
Link: https://lore.kernel.org/r/20230612155032.17036-1-rpearsonhpe@gmail.com
Signed-off-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Jason Gunthorpe <jgg@nvidia.com>
We've seen the in-flight count go into negative with some
internal stress testing in Microsoft.
Adding a WARN when this happens, in hope of understanding
why this happens when it happens.
Signed-off-by: Shyam Prasad N <sprasad@microsoft.com>
Reviewed-by: Bharath SM <bharathsm@microsoft.com>
Signed-off-by: Steve French <stfrench@microsoft.com>
umount can race with lease break so need to check if
tcon->ses->server is still valid to send the lease
break response.
Reviewed-by: Bharath SM <bharathsm@microsoft.com>
Reviewed-by: Shyam Prasad N <sprasad@microsoft.com>
Fixes: 59a556aebc ("SMB3: drop reference to cfile before sending oplock break")
Signed-off-by: Steve French <stfrench@microsoft.com>
This filter already exists for excluding IPv6 SNMP stats. Extend its
definition to also exclude IFLA_VF_INFO stats in RTM_GETLINK.
This patch constitutes a partial fix for a netlink attribute nesting
overflow bug in IFLA_VFINFO_LIST. By excluding the stats when the
requester doesn't need them, the truncation of the VF list is avoided.
While it was technically only the stats added in commit c5a9f6f0ab
("net/core: Add drop counters to VF statistics") breaking the camel's
back, the appreciable size of the stats data should never have been
included without due consideration for the maximum number of VFs
supported by PCI.
Fixes: 3b766cd832 ("net/core: Add reading VF statistics through the PF netdevice")
Fixes: c5a9f6f0ab ("net/core: Add drop counters to VF statistics")
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Cc: Edwin Peer <espeer@gmail.com>
Signed-off-by: Gal Pressman <gal@nvidia.com>
Link: https://lore.kernel.org/r/20230611105108.122586-1-gal@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Petr Machata says:
====================
mlxsw: Preparations for out-of-order-operations patches
The mlxsw driver currently makes the assumption that the user applies
configuration in a bottom-up manner. Thus netdevices need to be added to
the bridge before IP addresses are configured on that bridge or SVI added
on top of it. Enslaving a netdevice to another netdevice that already has
uppers is in fact forbidden by mlxsw for this reason. Despite this safety,
it is rather easy to get into situations where the offloaded configuration
is just plain wrong.
As an example, take a front panel port, configure an IP address: it gets a
RIF. Now enslave the port to a bridge, and the RIF is gone. Remove the
port from the bridge again, but the RIF never comes back. There is a number
of similar situations, where changing the configuration there and back
utterly breaks the offload.
Over the course of the following several patchsets, mlxsw code is going to
be adjusted to diminish the space of wrongly offloaded configurations.
Ideally the offload state will reflect the actual state, regardless of the
sequence of operation used to construct that state.
No functional changes are intended in this patchset yet. Rather the patches
prepare the codebase for easier introduction of functional changes in later
patchsets.
- In patch #1, extract a helper to join a RIF of a given port, if there is
one. In patch #2, use it in a newly-added helper to join a LAG interface.
- In patches #3, #4 and #5, add helpers that abstract away the rif->dev
access. This will make it simpler in the future to change the way the
deduction is done. In patch #6, do this for deduction from nexthop group
info to RIF.
- In patch #7, add a helper to destroy a RIF. So far RIF was destroyed
simply by kfree'ing it.
- In patch #8, add a helper to check if any IP addresses are configured on
a netdevice. This helper will be useful later.
- In patch #9, add a helper to migrate a RIF. This will be a convenient
place to put extensions later on.
- Patch #10 move IPIP initialization up to make ipip_ops_arr available
earlier.
====================
Link: https://lore.kernel.org/r/cover.1686581444.git.petrm@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
mlxsw will need to keep track of certain devices that are not related to
any of its front panel ports. This includes IPIP netdevices. To be able to
query the list of supported IPIP types, router->ipip_ops_arr needs to be
initialized.
To that end, move the IPIP initialization up (and finalization
correspondingly down).
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
RIF configuration contains a number of parameters that cannot be changed
after the RIF is created. For the IPIP loopbacks, this is currently worked
around by creating a new RIF with the desired configuration changes
applied, and updating next hops to the new RIF, and then destroying the old
RIF. This operation will be useful as a reusable atom, so extract a helper
to that effect.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
This function will be useful later as the driver will need to retroactively
create RIFs for new uppers with addresses.
Add another helper that assumes RCU lock, and restructure the code to
skip the IPv6 branch not through conditioning on the addr_list_empty
variable, but by directly returning the result value. This makes the skip
more obvious than it previously was.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Right now freeing the object that mlxsw uses to keep track of a RIF is as
simple as calling a kfree. But later on as CRIF abstraction is brought in,
it will involve severing the link between CRIF and its RIF as well. Better
to have the logic encapsulated in a helper.
Since a helper is being introduced, make it a full-fledged destructor and
have it validate that the objects tracked at the RIF have been released.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
To abstract away deduction of RIF from the corresponding next hop group
info (NHGI), mlxsw currently uses a macro. In its current form, that macro
is impossible to extend to more general computation. Therefore introduce a
helper, mlxsw_sp_nhgi_rif(), and use it throughout. This will make it
possible to change the deduction path easily later on.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
In order to abstract away deduction of netdevice from the corresponding
next hop, introduce a helper, mlxsw_sp_nexthop_dev(), and use it
throughout. This will make it possible to change the deduction path easily
later on.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
The previous patch added a helper to access a netdevice given a RIF. Using
this helper in mlxsw_sp_rif_create() is unreasonable: the netdevice was
given in RIF creation parameters. Just take it there.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
In order to abstract away deduction of netdevice from the corresponding
RIF, introduce a helper, mlxsw_sp_rif_dev(), and use it throughout. This
will make it possible to change the deduction path easily later on.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Currently, joining a LAG very simply means that the LAG RIF should be
joined by the subport representing untagged traffic. If the RIF does not
exist, it does not have to be created: if the user wants there to be RIF
for the LAG device, they are supposed to add an IP address, and they are
supposed to do it after tha LAG becomes mlxsw upper.
We can also assume that the LAG has no uppers, otherwise the enslavement is
not allowed.
In the future, these ordering dependencies should be removed. That means
that joining LAG will be more complex operation, possibly involving a lazy
RIF creation, and possibly joining / lazily creating RIFs for VLAN uppers
of the LAG. It will be handy to have a dedicated function that handles all
this. The new function mlxsw_sp_router_port_join_lag() is that.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Split out of mlxsw_sp_port_vlan_router_join() the part that checks for RIF
and dispatches to __mlxsw_sp_port_vlan_router_join(), leaving it as wrapper
that just manages the router lock.
The new function, mlxsw_sp_port_vlan_router_join_existing(), will be useful
as an atom in later patches.
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Amit Cohen <amcohen@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Setting the IPv6 address generation mode of a net device during its
creation never worked, but after commit b0ad3c1790 ("rtnetlink: call
validate_linkmsg in rtnl_create_link") it explicitly fails [1]. The
failure is caused by the fact that validate_linkmsg() is called before
the net device is registered, when it still does not have an 'inet6_dev'.
Likewise, raising the net device before setting the address generation
mode is meaningless, because by the time the mode is set, the address
has already been generated.
Therefore, fix the test to first create the net device, then set its
IPv6 address generation mode and finally bring it up.
[1]
# ip link add name mydev addrgenmode eui64 type dummy
RTNETLINK answers: Address family not supported by protocol
Fixes: ba95e79309 ("selftests: forwarding: hw_stats_l3: Add a new test")
Signed-off-by: Danielle Ratson <danieller@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Link: https://lore.kernel.org/r/f3b05d85b2bc0c3d6168fe8f7207c6c8365703db.1686580046.git.petrm@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
mini_Qdisc_pair::p_miniq is a double pointer to mini_Qdisc, initialized
in ingress_init() to point to net_device::miniq_ingress. ingress Qdiscs
access this per-net_device pointer in mini_qdisc_pair_swap(). Similar
for clsact Qdiscs and miniq_egress.
Unfortunately, after introducing RTNL-unlocked RTM_{NEW,DEL,GET}TFILTER
requests (thanks Hillf Danton for the hint), when replacing ingress or
clsact Qdiscs, for example, the old Qdisc ("@old") could access the same
miniq_{in,e}gress pointer(s) concurrently with the new Qdisc ("@new"),
causing race conditions [1] including a use-after-free bug in
mini_qdisc_pair_swap() reported by syzbot:
BUG: KASAN: slab-use-after-free in mini_qdisc_pair_swap+0x1c2/0x1f0 net/sched/sch_generic.c:1573
Write of size 8 at addr ffff888045b31308 by task syz-executor690/14901
...
Call Trace:
<TASK>
__dump_stack lib/dump_stack.c:88 [inline]
dump_stack_lvl+0xd9/0x150 lib/dump_stack.c:106
print_address_description.constprop.0+0x2c/0x3c0 mm/kasan/report.c:319
print_report mm/kasan/report.c:430 [inline]
kasan_report+0x11c/0x130 mm/kasan/report.c:536
mini_qdisc_pair_swap+0x1c2/0x1f0 net/sched/sch_generic.c:1573
tcf_chain_head_change_item net/sched/cls_api.c:495 [inline]
tcf_chain0_head_change.isra.0+0xb9/0x120 net/sched/cls_api.c:509
tcf_chain_tp_insert net/sched/cls_api.c:1826 [inline]
tcf_chain_tp_insert_unique net/sched/cls_api.c:1875 [inline]
tc_new_tfilter+0x1de6/0x2290 net/sched/cls_api.c:2266
...
@old and @new should not affect each other. In other words, @old should
never modify miniq_{in,e}gress after @new, and @new should not update
@old's RCU state.
Fixing without changing sch_api.c turned out to be difficult (please
refer to Closes: for discussions). Instead, make sure @new's first call
always happen after @old's last call (in {ingress,clsact}_destroy()) has
finished:
In qdisc_graft(), return -EBUSY if @old has any ongoing filter requests,
and call qdisc_destroy() for @old before grafting @new.
Introduce qdisc_refcount_dec_if_one() as the counterpart of
qdisc_refcount_inc_nz() used for filter requests. Introduce a
non-static version of qdisc_destroy() that does a TCQ_F_BUILTIN check,
just like qdisc_put() etc.
Depends on patch "net/sched: Refactor qdisc_graft() for ingress and
clsact Qdiscs".
[1] To illustrate, the syzkaller reproducer adds ingress Qdiscs under
TC_H_ROOT (no longer possible after commit c7cfbd1150 ("net/sched:
sch_ingress: Only create under TC_H_INGRESS")) on eth0 that has 8
transmission queues:
Thread 1 creates ingress Qdisc A (containing mini Qdisc a1 and a2),
then adds a flower filter X to A.
Thread 2 creates another ingress Qdisc B (containing mini Qdisc b1 and
b2) to replace A, then adds a flower filter Y to B.
Thread 1 A's refcnt Thread 2
RTM_NEWQDISC (A, RTNL-locked)
qdisc_create(A) 1
qdisc_graft(A) 9
RTM_NEWTFILTER (X, RTNL-unlocked)
__tcf_qdisc_find(A) 10
tcf_chain0_head_change(A)
mini_qdisc_pair_swap(A) (1st)
|
| RTM_NEWQDISC (B, RTNL-locked)
RCU sync 2 qdisc_graft(B)
| 1 notify_and_destroy(A)
|
tcf_block_release(A) 0 RTM_NEWTFILTER (Y, RTNL-unlocked)
qdisc_destroy(A) tcf_chain0_head_change(B)
tcf_chain0_head_change_cb_del(A) mini_qdisc_pair_swap(B) (2nd)
mini_qdisc_pair_swap(A) (3rd) |
... ...
Here, B calls mini_qdisc_pair_swap(), pointing eth0->miniq_ingress to
its mini Qdisc, b1. Then, A calls mini_qdisc_pair_swap() again during
ingress_destroy(), setting eth0->miniq_ingress to NULL, so ingress
packets on eth0 will not find filter Y in sch_handle_ingress().
This is just one of the possible consequences of concurrently accessing
miniq_{in,e}gress pointers.
Fixes: 7a096d579e ("net: sched: ingress: set 'unlocked' flag for Qdisc ops")
Fixes: 87f373921c ("net: sched: ingress: set 'unlocked' flag for clsact Qdisc ops")
Reported-by: syzbot+b53a9c0d1ea4ad62da8b@syzkaller.appspotmail.com
Closes: https://lore.kernel.org/r/0000000000006cf87705f79acf1a@google.com/
Cc: Hillf Danton <hdanton@sina.com>
Cc: Vlad Buslov <vladbu@mellanox.com>
Signed-off-by: Peilin Ye <peilin.ye@bytedance.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Grafting ingress and clsact Qdiscs does not need a for-loop in
qdisc_graft(). Refactor it. No functional changes intended.
Tested-by: Pedro Tammela <pctammela@mojatatu.com>
Acked-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Jamal Hadi Salim <jhs@mojatatu.com>
Reviewed-by: Vlad Buslov <vladbu@nvidia.com>
Signed-off-by: Peilin Ye <peilin.ye@bytedance.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Currently UNREPLIED and UNASSURED connections are added to the nf flow
table. This causes the following connection packets to be processed
by the flow table which then skips conntrack_in(), and thus such the
connections will remain UNREPLIED and UNASSURED even if reply traffic
is then seen. Even still, the unoffloaded reply packets are the ones
triggering hardware update from new to established state, and if
there aren't any to triger an update and/or previous update was
missed, hardware can get out of sync with sw and still mark
packets as new.
Fix the above by:
1) Not skipping conntrack_in() for UNASSURED packets, but still
refresh for hardware, as before the cited patch.
2) Try and force a refresh by reply-direction packets that update
the hardware rules from new to established state.
3) Remove any bidirectional flows that didn't failed to update in
hardware for re-insertion as bidrectional once any new packet
arrives.
Fixes: 6a9bad0069 ("net/sched: act_ct: offload UDP NEW connections")
Co-developed-by: Vlad Buslov <vladbu@nvidia.com>
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Signed-off-by: Paul Blakey <paulb@nvidia.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Link: https://lore.kernel.org/r/1686313379-117663-1-git-send-email-paulb@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Lockdep on 6.4-rc on ThinkPad X1 Carbon 5th says
=====================================================
WARNING: SOFTIRQ-safe -> SOFTIRQ-unsafe lock order detected
6.4.0-rc5 #1 Not tainted
-----------------------------------------------------
kworker/3:1/49 [HC0[0]:SC0[4]:HE1:SE0] is trying to acquire:
ffff8881066fa368 (&mvm_sta->deflink.lq_sta.rs_drv.pers.lock){+.+.}-{2:2}, at: rs_drv_get_rate+0x46/0xe7
and this task is already holding:
ffff8881066f80a8 (&sta->rate_ctrl_lock){+.-.}-{2:2}, at: rate_control_get_rate+0xbd/0x126
which would create a new lock dependency:
(&sta->rate_ctrl_lock){+.-.}-{2:2} -> (&mvm_sta->deflink.lq_sta.rs_drv.pers.lock){+.+.}-{2:2}
but this new dependency connects a SOFTIRQ-irq-safe lock:
(&sta->rate_ctrl_lock){+.-.}-{2:2}
etc. etc. etc.
Changing the spin_lock() in rs_drv_get_rate() to spin_lock_bh() was not
enough to pacify lockdep, but changing them all on pers.lock has worked.
Fixes: a8938bc881 ("wifi: iwlwifi: mvm: Add locking to the rate read flow")
Signed-off-by: Hugh Dickins <hughd@google.com>
Link: https://lore.kernel.org/r/79ffcc22-9775-cb6d-3ffd-1a517c40beef@google.com
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
sopass won't be set if wolopt doesn't change. This means the following
will fail to set the correct sopass.
ethtool -s eth0 wol s sopass 11:22:33:44:55:66
ethtool -s eth0 wol s sopass 22:44:55:66:77:88
Make sure we call into the driver layer set_wol if sopass is different.
Fixes: 55b24334c0 ("ethtool: ioctl: improve error checking for set_wol")
Signed-off-by: Justin Chen <justin.chen@broadcom.com>
Link: https://lore.kernel.org/r/1686605822-34544-1-git-send-email-justin.chen@broadcom.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Some qdiscs and classifiers have recently been retired from kernel.
However, tc-testing config is still cluttered with them which causes noise
when using merge_config.sh script to update existing config for tc-testing
compatibility. Remove the config settings for affected qdiscs and
classifiers.
Fixes: fb38306ceb ("net/sched: Retire ATM qdisc")
Fixes: 051d442098 ("net/sched: Retire CBQ qdisc")
Fixes: bbe77c14ee ("net/sched: Retire dsmark qdisc")
Fixes: 265b4da82d ("net/sched: Retire rsvp classifier")
Fixes: 8c710f7525 ("net/sched: Retire tcindex classifier")
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Pedro Tammela <pctammela@mojatatu.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Setting very small value of db like 10ms introduces rounding errors when
converting to/from jiffies on some kernel configs. For example, on 250hz
the actual value will be set to 12ms which causes the test to fail:
# $ sudo ./tdc.py -d eth2 -e 3410
# -- ns/SubPlugin.__init__
# Test 3410: Create SFB with db setting
#
# All test results:
#
# 1..1
# not ok 1 3410 - Create SFB with db setting
# Could not match regex pattern. Verify command output:
# qdisc sfb 1: root refcnt 2 rehash 600s db 12ms limit 1000p max 25p target 20p increment 0.000503548 decrement 4.57771e-05 penalty_rate 10pps penalty_burst 20p
Set the value to 100ms instead which currently seem to work on 100hz,
250hz, 300hz and 1000hz kernel configs.
Fixes: 6ad92dc56f ("selftests/tc-testing: add selftests for sfb qdisc")
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Pedro Tammela <pctammela@mojatatu.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
This code returns directly but it should instead call of_node_put()
to drop some reference counts.
Fixes: dab2b265dd ("net: ethernet: ti: am65-cpsw: Add support for SERDES configuration")
Signed-off-by: Dan Carpenter <dan.carpenter@linaro.org>
Reviewed-by: Roger Quadros <rogerq@kernel.org>
Link: https://lore.kernel.org/r/e3012f0c-1621-40e6-bf7d-03c276f6e07f@kili.mountain
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
After commit b8a1a4cd5a ("i2c: Provide a temporary .probe_new()
call-back type"), all drivers being converted to .probe_new() and then
commit 03c835f498 ("i2c: Switch .probe() to not take an id parameter")
convert back to (the new) .probe() to be able to eventually drop
.probe_new() from struct i2c_driver.
Signed-off-by: Uwe Kleine-König <u.kleine-koenig@pengutronix.de>
Link: https://lore.kernel.org/r/20230612071641.836976-1-u.kleine-koenig@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The given value of 1518 seems to refer to the layer 2 ethernet frame
size without 802.1Q tag. Actual use of the "max-frame-size" including in
the consumer of the "altr,tse-1.0" compatible is the MTU.
Fixes: 95acd4c7b6 ("nios2: Device tree support")
Fixes: 61c610ec61 ("nios2: Add Max10 device tree")
Cc: <stable@vger.kernel.org>
Signed-off-by: Janne Grunau <j@jannau.net>
Signed-off-by: Dinh Nguyen <dinguyen@kernel.org>
Commit 445164e8c1 ("spi: dw: Replace spi->chip_select references with
function calls") replaced direct access to spi.chip_select with
spi_*_chipselect calls but incorrectly replaced a set instance with a
get instance, replace the incorrect instance.
Fixes: 445164e8c1 ("spi: dw: Replace spi->chip_select references with function calls")
Signed-off-by: Abe Kohandel <abe.kohandel@intel.com>
Reviewed-by: Andy Shevchenko <andriy.shevchenko@linux.intel.com>
Acked-by: Serge Semin <fancer.lancer@gmail.com>
Link: https://lore.kernel.org/r/20230613162103.569812-1-abe.kohandel@intel.com
Signed-off-by: Mark Brown <broonie@kernel.org>
Jakub Kicinski says:
====================
tools: ynl-gen: improvements for DPLL
A set of improvements needed to generate the kernel code for DPLL.
====================
Link: https://lore.kernel.org/r/20230612155920.1787579-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Instead of reimplementing policies in MutliAttr for every
underlying type forward the calls to the base type.
This will be needed for DPLL which uses a multi-attr nest,
and currently gets an invalid NLA_NEST policy generated.
Reviewed-by: Jacob Keller <Jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Scalar range validation assumes enums start at 0.
Teach it to properly calculate the value range.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The "qcom,paired" schema is all wrong. First, it's a list rather than an
object(dictionary). Second, it is missing a required type. The meta-schema
normally catches this, but schemas under "$defs" was not getting checked.
A fix for that is pending.
Fixes: f9a06b8109 ("dt-bindings: pinctrl: qcom,pmic-mpp: Convert qcom pmic mpp bindings to YAML")
Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Link: https://lore.kernel.org/r/20230418150606.1528107-1-robh@kernel.org
Signed-off-by: Rob Herring <robh@kernel.org>
regcache_maple_sync() tries to sync all cached values no matter
whether it's writable or not. OTOH, regache_sync_val() does care the
wrtability and returns -EIO for a read-only register. This results in
an error message like:
snd_hda_codec_realtek hdaudioC0D0: Unable to sync register 0x2f0009. -5
and the sync loop is aborted incompletely.
This patch adds the writable register check to regcache_sync_val() for
addressing the bug above.
Note that, although we may add the check in the caller side
(regcache_maple_sync()), here we put in regcache_sync_val(), so that a
similar case like this can be avoided in future.
Fixes: f033c26de5 ("regmap: Add maple tree based register cache")
Link: https://lore.kernel.org/r/877cs7g6f1.wl-tiwai@suse.de
Signed-off-by: Takashi Iwai <tiwai@suse.de>
Link: https://lore.kernel.org/r/20230613112240.3361-1-tiwai@suse.de
Signed-off-by: Mark Brown <broonie@kernel.org>
MAC version 21H supports the 10Mbps speed. So, extend support to
platforms that support it.
Acked-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Reviewed-by: Sridhar Samudrala <sridhar.samudrala@intel.com>
Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Naveen Mamindlapalli says:
====================
RVU NIX AF driver updates
This patch series includes a few enhancements and other updates to the
RVU NIX AF driver.
The first patch adds devlink option to configure NPC MCAM high priority
zone entries reservation. This is useful when the requester needs more
high priority entries than default reserved entries.
The second patch adds support for RSS hash computation using L3 SRC or
DST only, or L4 SRC or DST only.
The third patch updates DWRR MTU configuration for CN10KB silicon. HW uses
the DWRR MTU to compute DWRR weight.
Patch 4 configures the LBK link in TL3_TL2 configuration only when switch
mode is enabled.
Patch 5 adds an option in the mailbox request to enable/disable DROP_RE bit
which drops packets with L2 errors when set.
Patch 6 updates SMQ flush mechanism to stop other child nodes from
enqueuing any packets while SMQ flush is active. Otherwise SMQ flush may
timeout.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
When multiple transmit scheduler queues feed a TL1 transmit link, the
SMQ flush initiated on a low priority queue might get stuck when a high
priority queue fully subscribes the transmit link. This inturn effects
interface teardown. To avoid this, temporarily XOFF all TL1's other
immediate child transmit scheduler queues and also clear any rate limit
configuration on all the scheduler queues in SMQ(flush) hierarchy.
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add option to toggle DROP_RE bit in rx cfg mbox. This helps in
modifying the config runtime as opposed to setting available via
nix_lf_alloc() mbox at NIX LF init time.
Signed-off-by: Nithin Dabilpuram <ndabilpuram@marvell.com>
Signed-off-by: Jerin Jacob Kollanukkaran <jerinj@marvell.com>
Signed-off-by: Sunil Kovvuri Goutham <sgoutham@marvell.com>
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Currently, all the TL3_TL2 nodes are being configured to enable
switch LBK channel 63 in them. Instead enable them only when switch
mode is enabled.
Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com>
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The DWRR MTU config added for SDP and RPM/LBK links on CN10K
silicon is further extended on CK10KB silicon variant and made
it configurable. Now there are 4 DWRR MTU config to choose while
setting transmit scheduler's RR_WEIGHT.
Here we are reserving one config for each of RPM, SDP and LBK.
NIXX_AF_DWRR_MTUX(0) ---> RPM
NIXX_AF_DWRR_MTUX(1) ---> SDP
NIXX_AF_DWRR_MTUX(2) ---> LBK
PF/VF drivers can choose the DWRR_MTU to be used by setting
SMQX_CFG[pkt_link_type] to one of above. TLx_SCHEDULE[RR_WEIGHT]
is to be as configured 'quantum / 2^DWRR_MTUX[MTU]'. DWRR_MTU
of each link is exposed to PF/VF drivers via mailbox for
RR_WEIGHT calculation.
Signed-off-by: Sunil Goutham <sgoutham@marvell.com>
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Add support to select L3 SRC or DST only, L4 SRC or DST only for RSS
calculation.
AF consumer may have requirement as we can select only SRC or DST data for
RSS calculation in L3, L4 layers. With this requirement there will be
following combinations, IPV[4,6]_SRC_ONLY, IPV[4,6]_DST_ONLY,
[TCP,UDP,SCTP]_SRC_ONLY, [TCP,UDP,SCTP]_DST_ONLY. So, instead of creating
a bit for each combination, we are using upper 4 bits (31:28) in the
flow_key_cfg to represent the SRC, DST selection. 31 => L3_SRC,
30 => L3_DST, 29 => L4_SRC, 28 => L4_DST. These won't be part of flow_cfg,
so that we don't need to change the existing ABI.
Signed-off-by: Kiran Kumar K <kirankumark@marvell.com>
Signed-off-by: Geetha sowjanya <gakula@marvell.com>
Signed-off-by: Naveen Mamindlapalli <naveenm@marvell.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
