David Epping says:
====================
net: phy: mscc: support VSC8501
this updated series of patches adds support for the VSC8501 Ethernet
PHY and fixes support for the VSC8502 PHY in cases where no other
software (like U-Boot) has initialized the PHY after power up.
The first patch simply adds the VSC8502 to the MODULE_DEVICE_TABLE,
where I guess it was unintentionally missing. I have no hardware to
test my change.
The second patch adds the VSC8501 PHY with exactly the same driver
implementation as the existing VSC8502.
The (new) third patch removes phydev locking from
vsc85xx_rgmii_set_skews(), as discussed for v2 of the patch set.
The (now) fourth patch fixes the initialization for VSC8501 and VSC8502.
I have tested this patch with VSC8501 on hardware in RGMII mode only.
https://ww1.microchip.com/downloads/aemDocuments/documents/UNG/ProductDocuments/DataSheets/VSC8501-03_Datasheet_60001741A.PDFhttps://ww1.microchip.com/downloads/aemDocuments/documents/UNG/ProductDocuments/DataSheets/VSC8502-03_Datasheet_60001742B.pdf
Table 4-42 "RGMII CONTROL, ADDRESS 20E2 (0X14)" Bit 11 for each of
them.
By default the RX_CLK is disabled for these PHYs. In cases where no
other software, like U-Boot, enabled the clock, this results in no
received packets being handed to the MAC.
The patch enables this clock output.
According to Microchip support (case number 01268776) this applies
to all modes (RGMII, GMII, and MII).
Other PHYs sharing the same register map and code, like
VSC8530/31/40/41 have the clock enabled and the relevant bit 11 is
reserved and read-only for them. As per previous discussion the
patch still clears the bit on these PHYs, too, possibly more easily
supporting other future PHYs implementing this functionality.
For the VSC8572 family of PHYs, having a different register map,
no such changes are applied.
====================
Link: https://lore.kernel.org/r/20230523153108.18548-1-david.epping@missinglinkelectronics.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
By default the VSC8501 and VSC8502 RGMII/GMII/MII RX_CLK output is
disabled. To allow packet forwarding towards the MAC it needs to be
enabled.
For other PHYs supported by this driver the clock output is enabled
by default.
Fixes: d316986331 ("net: phy: mscc: add support for VSC8502")
Signed-off-by: David Epping <david.epping@missinglinkelectronics.com>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Holding the struct phy_device (phydev) lock is unnecessary when
accessing phydev->interface in the PHY driver .config_init method,
which is the only place that vsc85xx_rgmii_set_skews() is called from.
The phy_modify_paged() function implements required MDIO bus level
locking, which can not be achieved by a phydev lock.
Signed-off-by: David Epping <david.epping@missinglinkelectronics.com>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The VSC8501 PHY can use the same driver implementation as the VSC8502.
Adding the PHY ID and copying the handler functions of VSC8502 is
sufficient to operate it.
Signed-off-by: David Epping <david.epping@missinglinkelectronics.com>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The mscc driver implements support for VSC8502, so its ID should be in
the MODULE_DEVICE_TABLE for automatic loading.
Signed-off-by: David Epping <david.epping@missinglinkelectronics.com>
Fixes: d316986331 ("net: phy: mscc: add support for VSC8502")
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Chuck Lever says:
====================
Bug fixes for net/handshake
Paolo observed that there is a possible leak of sock->file. I
haven't looked into that yet, but it seems to be separate from
the fixes in this series, so no need to hold these up.
====================
The submissions mentions net-next but it means netdev (perhaps
merge window left over when trees are converged). In any case,
it should have gone into net, but was instead applied to net-next
as commit deb2e484ba ("Merge branch 'net-handshake-fixes'").
These are fixes tho, and Chuck needs them to make progress with
the client so double-merging them into net... it is what it is :(
Link: https://lore.kernel.org/r/168381978252.84244.1933636428135211300.stgit@91.116.238.104.host.secureserver.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Enable the upper layer protocol to specify the SNI peername. This
avoids the need for tlshd to use a DNS lookup, which can return a
hostname that doesn't match the incoming certificate's SubjectName.
Fixes: 2fd5532044 ("net/handshake: Add a kernel API for requesting a TLSv1.3 handshake")
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
If user space never calls DONE, sock->file's reference count remains
elevated. Enable sock->file to be freed eventually in this case.
Reported-by: Jakub Kacinski <kuba@kernel.org>
Fixes: 3b3009ea8a ("net/handshake: Create a NETLINK service for handling handshake requests")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Fixes: 3b3009ea8a ("net/handshake: Create a NETLINK service for handling handshake requests")
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
trace_handshake_cmd_done_err() simply records the pointer in @req,
so initializing it to NULL is sufficient and safe.
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Fixes: 3b3009ea8a ("net/handshake: Create a NETLINK service for handling handshake requests")
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
If get_unused_fd_flags() fails, we ended up calling fput(sock->file)
twice.
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Suggested-by: Paolo Abeni <pabeni@redhat.com>
Fixes: 3b3009ea8a ("net/handshake: Create a NETLINK service for handling handshake requests")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
handshake_req_submit() now verifies that the socket has a file.
Fixes: 3b3009ea8a ("net/handshake: Create a NETLINK service for handling handshake requests")
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
-----BEGIN PGP SIGNATURE-----
iHUEABYIAB0WIQTFp0I1jqZrAX+hPRXbK58LschIgwUCZG4AiAAKCRDbK58LschI
g+xlAQCmefGbDuwPckZLnomvt6gl4bkIjs7kc1ySbG9QBnaInwD/WyrJaQIPijuD
qziHPAyx+MEgPseFU1b7Le35SZ66IwM=
=s4R1
-----END PGP SIGNATURE-----
Merge tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf
Daniel Borkmann says:
====================
pull-request: bpf 2023-05-24
We've added 19 non-merge commits during the last 10 day(s) which contain
a total of 20 files changed, 738 insertions(+), 448 deletions(-).
The main changes are:
1) Batch of BPF sockmap fixes found when running against NGINX TCP tests,
from John Fastabend.
2) Fix a memleak in the LRU{,_PERCPU} hash map when bucket locking fails,
from Anton Protopopov.
3) Init the BPF offload table earlier than just late_initcall,
from Jakub Kicinski.
4) Fix ctx access mask generation for 32-bit narrow loads of 64-bit fields,
from Will Deacon.
5) Remove a now unsupported __fallthrough in BPF samples,
from Andrii Nakryiko.
6) Fix a typo in pkg-config call for building sign-file,
from Jeremy Sowden.
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
bpf, sockmap: Test progs verifier error with latest clang
bpf, sockmap: Test FIONREAD returns correct bytes in rx buffer with drops
bpf, sockmap: Test FIONREAD returns correct bytes in rx buffer
bpf, sockmap: Test shutdown() correctly exits epoll and recv()=0
bpf, sockmap: Build helper to create connected socket pair
bpf, sockmap: Pull socket helpers out of listen test for general use
bpf, sockmap: Incorrectly handling copied_seq
bpf, sockmap: Wake up polling after data copy
bpf, sockmap: TCP data stall on recv before accept
bpf, sockmap: Handle fin correctly
bpf, sockmap: Improved check for empty queue
bpf, sockmap: Reschedule is now done through backlog
bpf, sockmap: Convert schedule_work into delayed_work
bpf, sockmap: Pass skb ownership through read_skb
bpf: fix a memory leak in the LRU and LRU_PERCPU hash maps
bpf: Fix mask generation for 32-bit narrow loads of 64-bit fields
samples/bpf: Drop unnecessary fallthrough
bpf: netdev: init the offload table earlier
selftests/bpf: Fix pkg-config call building sign-file
====================
Link: https://lore.kernel.org/r/20230524170839.13905-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
optlen is fetched without checking whether there is more than one byte to parse.
It can lead to out-of-bounds access.
Found by InfoTeCS on behalf of Linux Verification Center
(linuxtesting.org) with SVACE.
Fixes: c61a404325 ("[IPV6]: Find option offset by type.")
Signed-off-by: Gavrilov Ilia <Ilia.Gavrilov@infotecs.ru>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
-----BEGIN PGP SIGNATURE-----
iQEzBAABCAAdFiEEGhZs6bAKwk/OTgTpSD+KveBX+j4FAmRsUT8ACgkQSD+KveBX
+j61Zwf9GyvzrD29Lmu0/BTsLAnf7GAyJi/SMzXJ09Tp1dAYSWmF2DE3fzKvNoQ/
VT2udSKbZ96b2N9SGF396KZaV8gHxg23IAzILia1JDPd4Pn7YaNymAIWGU7vn+Tq
ErG7atPVnJV5R1H6SwO2KpOClG7jOjUPMF87uDCl2g+IpYNgjKa9hcnt5bguztC2
KBW/sV7BCYVWOUrmlSe1hH2Fn4djhga3i4JBIzjp55Dz1voIu5SHsT13Ou2/UuiC
1RDqBTJ9WvnviAxICbI96TLMJTFnDo9HFGHPIQRhZ6k25PIuWX6GLMKaceVlfCd+
BZvRG+PNOsDR9a9tFCjMBfx7KE0bMw==
=dwjN
-----END PGP SIGNATURE-----
Merge tag 'mlx5-fixes-2023-05-22' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5-fixes-2023-05-22
This series provides bug fixes for the mlx5 driver.
Please pull and let me know if there is any problem.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
The commit c6d96df9fa ("net: ethernet: mtk_eth_soc: drop generic vlan rx
offload, only use DSA untagging") makes VLAN RX offloading to be only used
on the SoCs without the MTK_NETSYS_V2 ability (which are not just MT7621
and MT7622). The commit disables the proper handling of special tagged
(DSA) frames, added with commit 87e3df4961 ("net-next: ethernet:
mediatek: add CDM able to recognize the tag for DSA"), for non
MTK_NETSYS_V2 SoCs when it finds a MAC that does not use DSA. So if the
other MAC uses DSA, the CDMQ component transmits DSA tagged frames to the
CPU improperly. This issue can be observed on frames with TCP, for example,
a TCP speed test using iperf3 won't work.
The commit disables the proper handling of special tagged (DSA) frames
because it assumes that these SoCs don't use more than one MAC, which is
wrong. Although I made Frank address this false assumption on the patch log
when they sent the patch on behalf of Felix, the code still made changes
with this assumption.
Therefore, the proper handling of special tagged (DSA) frames must be kept
enabled in all circumstances as it doesn't affect non DSA tagged frames.
Hardware DSA untagging, introduced with the commit 2d7605a729 ("net:
ethernet: mtk_eth_soc: enable hardware DSA untagging"), and VLAN RX
offloading are operations on the two CDM components of the frame engine,
CDMP and CDMQ, which connect to Packet DMA (PDMA) and QoS DMA (QDMA) and
are between the MACs and the CPU. These operations apply to all MACs of the
SoC so if one MAC uses DSA and the other doesn't, the hardware DSA
untagging operation will cause the CDMP component to transmit non DSA
tagged frames to the CPU improperly.
Since the VLAN RX offloading feature configuration was dropped, VLAN RX
offloading can only be used along with hardware DSA untagging. So, for the
case above, we need to disable both features and leave it to the CPU,
therefore software, to untag the DSA and VLAN tags.
So the correct way to handle this is:
For all SoCs:
Enable the proper handling of special tagged (DSA) frames
(MTK_CDMQ_IG_CTRL).
For non MTK_NETSYS_V2 SoCs:
Enable hardware DSA untagging (MTK_CDMP_IG_CTRL).
Enable VLAN RX offloading (MTK_CDMP_EG_CTRL).
When a non MTK_NETSYS_V2 SoC MAC does not use DSA:
Disable hardware DSA untagging (MTK_CDMP_IG_CTRL).
Disable VLAN RX offloading (MTK_CDMP_EG_CTRL).
Fixes: c6d96df9fa ("net: ethernet: mtk_eth_soc: drop generic vlan rx offload, only use DSA untagging")
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We had a good run, but after 4 weeks of use we heard someone
asking about pw-bot commands. Let's explain its existence
in the docs. It's not a complete documentation but hopefully
it's enough for the casual contributor. The project and scope
are in flux so the details would likely become out of date,
if we were to document more in depth.
Link: https://lore.kernel.org/all/20230522140057.GB18381@nucnuc.mle/
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230522230903.1853151-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Commit 50749f2dd6 ("tcp/udp: Fix memleaks of sk and zerocopy skbs with
TX timestamp.") added a call to skb_orphan_frags_rx() to fix leaks with
zerocopy skbs. But it ended up adding a leak of its own. When
skb_orphan_frags_rx() fails, the function just returns, leaking the skb
it just cloned. Free it before returning.
This bug was discovered and resolved using Coverity Static Analysis
Security Testing (SAST) by Synopsys, Inc.
Fixes: 50749f2dd6 ("tcp/udp: Fix memleaks of sk and zerocopy skbs with TX timestamp.")
Signed-off-by: Pratyush Yadav <ptyadav@amazon.de>
Reviewed-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://lore.kernel.org/r/20230522153020.32422-1-ptyadav@amazon.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The driver's interrupt service routine is requested with the
IRQF_NO_THREAD if MSI is available. This means that the routine is
invoked in hardirq context even on PREEMPT_RT. The routine itself is
relatively short and schedules a worker, performs register access and
schedules NAPI. On PREEMPT_RT, scheduling NAPI from hardirq results in
waking ksoftirqd for further processing so using NAPI threads with this
driver is highly recommended since it NULL routes the threaded-IRQ
efforts.
Adding rtl_hw_aspm_clkreq_enable() to the ISR is problematic on
PREEMPT_RT because the function uses spinlock_t locks which become
sleeping locks on PREEMPT_RT. The locks are only used to protect
register access and don't nest into other functions or locks. They are
also not used for unbounded period of time. Therefore it looks okay to
convert them to raw_spinlock_t.
Convert the three locks which are used from the interrupt service
routine to raw_spinlock_t.
Fixes: e1ed3e4d91 ("r8169: disable ASPM during NAPI poll")
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Reviewed-by: Heiner Kallweit <hkallweit1@gmail.com>
Link: https://lore.kernel.org/r/20230522134121.uxjax0F5@linutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
page_pool_ring_[un]lock() use in_softirq() to decide which
spin lock variant to use, and when they are called in the
context with in_softirq() being false, spin_lock_bh() is
called in page_pool_ring_lock() while spin_unlock() is
called in page_pool_ring_unlock(), because spin_lock_bh()
has disabled the softirq in page_pool_ring_lock(), which
causes inconsistency for spin lock pair calling.
This patch fixes it by returning in_softirq state from
page_pool_producer_lock(), and use it to decide which
spin lock variant to use in page_pool_producer_unlock().
As pool->ring has both producer and consumer lock, so
rename it to page_pool_producer_[un]lock() to reflect
the actual usage. Also move them to page_pool.c as they
are only used there, and remove the 'inline' as the
compiler may have better idea to do inlining or not.
Fixes: 7886244736 ("net: page_pool: Add bulk support for ptr_ring")
Signed-off-by: Yunsheng Lin <linyunsheng@huawei.com>
Acked-by: Jesper Dangaard Brouer <brouer@redhat.com>
Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org>
Link: https://lore.kernel.org/r/20230522031714.5089-1-linyunsheng@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When BPF program drops pkts the sockmap logic 'eats' the packet and
updates copied_seq. In the PASS case where the sk_buff is accepted
we update copied_seq from recvmsg path so we need a new test to
handle the drop case.
Original patch series broke this resulting in
test_sockmap_skb_verdict_fionread:PASS:ioctl(FIONREAD) error 0 nsec
test_sockmap_skb_verdict_fionread:FAIL:ioctl(FIONREAD) unexpected ioctl(FIONREAD): actual 1503041772 != expected 256
After updated patch with fix.
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-14-john.fastabend@gmail.com
A bug was reported where ioctl(FIONREAD) returned zero even though the
socket with a SK_SKB verdict program attached had bytes in the msg
queue. The result is programs may hang or more likely try to recover,
but use suboptimal buffer sizes.
Add a test to check that ioctl(FIONREAD) returns the correct number of
bytes.
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-13-john.fastabend@gmail.com
When session gracefully shutdowns epoll needs to wake up and any recv()
readers should return 0 not the -EAGAIN they previously returned.
Note we use epoll instead of select to test the epoll wake on shutdown
event as well.
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-12-john.fastabend@gmail.com
A common operation for testing is to spin up a pair of sockets that are
connected. Then we can use these to run specific tests that need to
send data, check BPF programs and so on.
The sockmap_listen programs already have this logic lets move it into
the new sockmap_helpers header file for general use.
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-11-john.fastabend@gmail.com
No functional change here we merely pull the helpers in sockmap_listen.c
into a header file so we can use these in other programs. The tests we
are about to add aren't really _listen tests so doesn't make sense
to add them here.
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-10-john.fastabend@gmail.com
The read_skb() logic is incrementing the tcp->copied_seq which is used for
among other things calculating how many outstanding bytes can be read by
the application. This results in application errors, if the application
does an ioctl(FIONREAD) we return zero because this is calculated from
the copied_seq value.
To fix this we move tcp->copied_seq accounting into the recv handler so
that we update these when the recvmsg() hook is called and data is in
fact copied into user buffers. This gives an accurate FIONREAD value
as expected and improves ACK handling. Before we were calling the
tcp_rcv_space_adjust() which would update 'number of bytes copied to
user in last RTT' which is wrong for programs returning SK_PASS. The
bytes are only copied to the user when recvmsg is handled.
Doing the fix for recvmsg is straightforward, but fixing redirect and
SK_DROP pkts is a bit tricker. Build a tcp_psock_eat() helper and then
call this from skmsg handlers. This fixes another issue where a broken
socket with a BPF program doing a resubmit could hang the receiver. This
happened because although read_skb() consumed the skb through sock_drop()
it did not update the copied_seq. Now if a single reccv socket is
redirecting to many sockets (for example for lb) the receiver sk will be
hung even though we might expect it to continue. The hang comes from
not updating the copied_seq numbers and memory pressure resulting from
that.
We have a slight layer problem of calling tcp_eat_skb even if its not
a TCP socket. To fix we could refactor and create per type receiver
handlers. I decided this is more work than we want in the fix and we
already have some small tweaks depending on caller that use the
helper skb_bpf_strparser(). So we extend that a bit and always set
the strparser bit when it is in use and then we can gate the
seq_copied updates on this.
Fixes: 04919bed94 ("tcp: Introduce tcp_read_skb()")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-9-john.fastabend@gmail.com
When TCP stack has data ready to read sk_data_ready() is called. Sockmap
overwrites this with its own handler to call into BPF verdict program.
But, the original TCP socket had sock_def_readable that would additionally
wake up any user space waiters with sk_wake_async().
Sockmap saved the callback when the socket was created so call the saved
data ready callback and then we can wake up any epoll() logic waiting
on the read.
Note we call on 'copied >= 0' to account for returning 0 when a FIN is
received because we need to wake up user for this as well so they
can do the recvmsg() -> 0 and detect the shutdown.
Fixes: 04919bed94 ("tcp: Introduce tcp_read_skb()")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-8-john.fastabend@gmail.com
A common mechanism to put a TCP socket into the sockmap is to hook the
BPF_SOCK_OPS_{ACTIVE_PASSIVE}_ESTABLISHED_CB event with a BPF program
that can map the socket info to the correct BPF verdict parser. When
the user adds the socket to the map the psock is created and the new
ops are assigned to ensure the verdict program will 'see' the sk_buffs
as they arrive.
Part of this process hooks the sk_data_ready op with a BPF specific
handler to wake up the BPF verdict program when data is ready to read.
The logic is simple enough (posted here for easy reading)
static void sk_psock_verdict_data_ready(struct sock *sk)
{
struct socket *sock = sk->sk_socket;
if (unlikely(!sock || !sock->ops || !sock->ops->read_skb))
return;
sock->ops->read_skb(sk, sk_psock_verdict_recv);
}
The oversight here is sk->sk_socket is not assigned until the application
accepts() the new socket. However, its entirely ok for the peer application
to do a connect() followed immediately by sends. The socket on the receiver
is sitting on the backlog queue of the listening socket until its accepted
and the data is queued up. If the peer never accepts the socket or is slow
it will eventually hit data limits and rate limit the session. But,
important for BPF sockmap hooks when this data is received TCP stack does
the sk_data_ready() call but the read_skb() for this data is never called
because sk_socket is missing. The data sits on the sk_receive_queue.
Then once the socket is accepted if we never receive more data from the
peer there will be no further sk_data_ready calls and all the data
is still on the sk_receive_queue(). Then user calls recvmsg after accept()
and for TCP sockets in sockmap we use the tcp_bpf_recvmsg_parser() handler.
The handler checks for data in the sk_msg ingress queue expecting that
the BPF program has already run from the sk_data_ready hook and enqueued
the data as needed. So we are stuck.
To fix do an unlikely check in recvmsg handler for data on the
sk_receive_queue and if it exists wake up data_ready. We have the sock
locked in both read_skb and recvmsg so should avoid having multiple
runners.
Fixes: 04919bed94 ("tcp: Introduce tcp_read_skb()")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-7-john.fastabend@gmail.com
The sockmap code is returning EAGAIN after a FIN packet is received and no
more data is on the receive queue. Correct behavior is to return 0 to the
user and the user can then close the socket. The EAGAIN causes many apps
to retry which masks the problem. Eventually the socket is evicted from
the sockmap because its released from sockmap sock free handling. The
issue creates a delay and can cause some errors on application side.
To fix this check on sk_msg_recvmsg side if length is zero and FIN flag
is set then set return to zero. A selftest will be added to check this
condition.
Fixes: 04919bed94 ("tcp: Introduce tcp_read_skb()")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: William Findlay <will@isovalent.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-6-john.fastabend@gmail.com
We noticed some rare sk_buffs were stepping past the queue when system was
under memory pressure. The general theory is to skip enqueueing
sk_buffs when its not necessary which is the normal case with a system
that is properly provisioned for the task, no memory pressure and enough
cpu assigned.
But, if we can't allocate memory due to an ENOMEM error when enqueueing
the sk_buff into the sockmap receive queue we push it onto a delayed
workqueue to retry later. When a new sk_buff is received we then check
if that queue is empty. However, there is a problem with simply checking
the queue length. When a sk_buff is being processed from the ingress queue
but not yet on the sockmap msg receive queue its possible to also recv
a sk_buff through normal path. It will check the ingress queue which is
zero and then skip ahead of the pkt being processed.
Previously we used sock lock from both contexts which made the problem
harder to hit, but not impossible.
To fix instead of popping the skb from the queue entirely we peek the
skb from the queue and do the copy there. This ensures checks to the
queue length are non-zero while skb is being processed. Then finally
when the entire skb has been copied to user space queue or another
socket we pop it off the queue. This way the queue length check allows
bypassing the queue only after the list has been completely processed.
To reproduce issue we run NGINX compliance test with sockmap running and
observe some flakes in our testing that we attributed to this issue.
Fixes: 04919bed94 ("tcp: Introduce tcp_read_skb()")
Suggested-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: William Findlay <will@isovalent.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-5-john.fastabend@gmail.com
Now that the backlog manages the reschedule() logic correctly we can drop
the partial fix to reschedule from recvmsg hook.
Rescheduling on recvmsg hook was added to address a corner case where we
still had data in the backlog state but had nothing to kick it and
reschedule the backlog worker to run and finish copying data out of the
state. This had a couple limitations, first it required user space to
kick it introducing an unnecessary EBUSY and retry. Second it only
handled the ingress case and egress redirects would still be hung.
With the correct fix, pushing the reschedule logic down to where the
enomem error occurs we can drop this fix.
Fixes: bec217197b ("skmsg: Schedule psock work if the cached skb exists on the psock")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-4-john.fastabend@gmail.com
Sk_buffs are fed into sockmap verdict programs either from a strparser
(when the user might want to decide how framing of skb is done by attaching
another parser program) or directly through tcp_read_sock. The
tcp_read_sock is the preferred method for performance when the BPF logic is
a stream parser.
The flow for Cilium's common use case with a stream parser is,
tcp_read_sock()
sk_psock_verdict_recv
ret = bpf_prog_run_pin_on_cpu()
sk_psock_verdict_apply(sock, skb, ret)
// if system is under memory pressure or app is slow we may
// need to queue skb. Do this queuing through ingress_skb and
// then kick timer to wake up handler
skb_queue_tail(ingress_skb, skb)
schedule_work(work);
The work queue is wired up to sk_psock_backlog(). This will then walk the
ingress_skb skb list that holds our sk_buffs that could not be handled,
but should be OK to run at some later point. However, its possible that
the workqueue doing this work still hits an error when sending the skb.
When this happens the skbuff is requeued on a temporary 'state' struct
kept with the workqueue. This is necessary because its possible to
partially send an skbuff before hitting an error and we need to know how
and where to restart when the workqueue runs next.
Now for the trouble, we don't rekick the workqueue. This can cause a
stall where the skbuff we just cached on the state variable might never
be sent. This happens when its the last packet in a flow and no further
packets come along that would cause the system to kick the workqueue from
that side.
To fix we could do simple schedule_work(), but while under memory pressure
it makes sense to back off some instead of continue to retry repeatedly. So
instead to fix convert schedule_work to schedule_delayed_work and add
backoff logic to reschedule from backlog queue on errors. Its not obvious
though what a good backoff is so use '1'.
To test we observed some flakes whil running NGINX compliance test with
sockmap we attributed these failed test to this bug and subsequent issue.
>From on list discussion. This commit
bec217197b41("skmsg: Schedule psock work if the cached skb exists on the psock")
was intended to address similar race, but had a couple cases it missed.
Most obvious it only accounted for receiving traffic on the local socket
so if redirecting into another socket we could still get an sk_buff stuck
here. Next it missed the case where copied=0 in the recv() handler and
then we wouldn't kick the scheduler. Also its sub-optimal to require
userspace to kick the internal mechanisms of sockmap to wake it up and
copy data to user. It results in an extra syscall and requires the app
to actual handle the EAGAIN correctly.
Fixes: 04919bed94 ("tcp: Introduce tcp_read_skb()")
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: William Findlay <will@isovalent.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-3-john.fastabend@gmail.com
The read_skb hook calls consume_skb() now, but this means that if the
recv_actor program wants to use the skb it needs to inc the ref cnt
so that the consume_skb() doesn't kfree the sk_buff.
This is problematic because in some error cases under memory pressure
we may need to linearize the sk_buff from sk_psock_skb_ingress_enqueue().
Then we get this,
skb_linearize()
__pskb_pull_tail()
pskb_expand_head()
BUG_ON(skb_shared(skb))
Because we incremented users refcnt from sk_psock_verdict_recv() we
hit the bug on with refcnt > 1 and trip it.
To fix lets simply pass ownership of the sk_buff through the skb_read
call. Then we can drop the consume from read_skb handlers and assume
the verdict recv does any required kfree.
Bug found while testing in our CI which runs in VMs that hit memory
constraints rather regularly. William tested TCP read_skb handlers.
[ 106.536188] ------------[ cut here ]------------
[ 106.536197] kernel BUG at net/core/skbuff.c:1693!
[ 106.536479] invalid opcode: 0000 [#1] PREEMPT SMP PTI
[ 106.536726] CPU: 3 PID: 1495 Comm: curl Not tainted 5.19.0-rc5 #1
[ 106.537023] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS ArchLinux 1.16.0-1 04/01/2014
[ 106.537467] RIP: 0010:pskb_expand_head+0x269/0x330
[ 106.538585] RSP: 0018:ffffc90000138b68 EFLAGS: 00010202
[ 106.538839] RAX: 000000000000003f RBX: ffff8881048940e8 RCX: 0000000000000a20
[ 106.539186] RDX: 0000000000000002 RSI: 0000000000000000 RDI: ffff8881048940e8
[ 106.539529] RBP: ffffc90000138be8 R08: 00000000e161fd1a R09: 0000000000000000
[ 106.539877] R10: 0000000000000018 R11: 0000000000000000 R12: ffff8881048940e8
[ 106.540222] R13: 0000000000000003 R14: 0000000000000000 R15: ffff8881048940e8
[ 106.540568] FS: 00007f277dde9f00(0000) GS:ffff88813bd80000(0000) knlGS:0000000000000000
[ 106.540954] CS: 0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[ 106.541227] CR2: 00007f277eeede64 CR3: 000000000ad3e000 CR4: 00000000000006e0
[ 106.541569] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[ 106.541915] DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
[ 106.542255] Call Trace:
[ 106.542383] <IRQ>
[ 106.542487] __pskb_pull_tail+0x4b/0x3e0
[ 106.542681] skb_ensure_writable+0x85/0xa0
[ 106.542882] sk_skb_pull_data+0x18/0x20
[ 106.543084] bpf_prog_b517a65a242018b0_bpf_skskb_http_verdict+0x3a9/0x4aa9
[ 106.543536] ? migrate_disable+0x66/0x80
[ 106.543871] sk_psock_verdict_recv+0xe2/0x310
[ 106.544258] ? sk_psock_write_space+0x1f0/0x1f0
[ 106.544561] tcp_read_skb+0x7b/0x120
[ 106.544740] tcp_data_queue+0x904/0xee0
[ 106.544931] tcp_rcv_established+0x212/0x7c0
[ 106.545142] tcp_v4_do_rcv+0x174/0x2a0
[ 106.545326] tcp_v4_rcv+0xe70/0xf60
[ 106.545500] ip_protocol_deliver_rcu+0x48/0x290
[ 106.545744] ip_local_deliver_finish+0xa7/0x150
Fixes: 04919bed94 ("tcp: Introduce tcp_read_skb()")
Reported-by: William Findlay <will@isovalent.com>
Signed-off-by: John Fastabend <john.fastabend@gmail.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: William Findlay <will@isovalent.com>
Reviewed-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20230523025618.113937-2-john.fastabend@gmail.com
With a raw socket bound to IPPROTO_RAW (ie with hdrincl enabled), the
protocol field of the flow structure, build by raw_sendmsg() /
rawv6_sendmsg()), is set to IPPROTO_RAW. This breaks the ipsec policy
lookup when some policies are defined with a protocol in the selector.
For ipv6, the sin6_port field from 'struct sockaddr_in6' could be used to
specify the protocol. Just accept all values for IPPROTO_RAW socket.
For ipv4, the sin_port field of 'struct sockaddr_in' could not be used
without breaking backward compatibility (the value of this field was never
checked). Let's add a new kind of control message, so that the userland
could specify which protocol is used.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
CC: stable@vger.kernel.org
Signed-off-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Link: https://lore.kernel.org/r/20230522120820.1319391-1-nicolas.dichtel@6wind.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
It was noticing that after a while when unloading/loading the driver and
sending traffic through the switch, it would stop working. It would stop
forwarding any traffic and the only way to get out of this was to do a
power cycle of the board. The root cause seems to be that the switch
core is initialized twice. Apparently initializing twice the switch core
disturbs the pointers in the queue systems in the HW, so after a while
it would stop sending the traffic.
Unfortunetly, it is not possible to use a reset of the switch here,
because the reset line is connected to multiple devices like MDIO,
SGPIO, FAN, etc. So then all the devices will get reseted when the
network driver will be loaded.
So the fix is to check if the core is initialized already and if that is
the case don't initialize it again.
Fixes: db8bcaad53 ("net: lan966x: add the basic lan966x driver")
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Reviewed-by: Simon Horman <simon.horman@corigine.com>
Link: https://lore.kernel.org/r/20230522120038.3749026-1-horatiu.vultur@microchip.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
After the cited patch, mlx5_irq xarray index can be different then
mlx5_irq MSIX table index.
Fix it by storing both mlx5_irq xarray index and MSIX table index.
Fixes: 3354822cde ("net/mlx5: Use dynamic msix vectors allocation")
Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Eli Cohen <elic@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
The cited patch deny the user of changing the affinity of mlx5 irqs,
which break backward compatibility.
Hence, allow the user to change the affinity of mlx5 irqs.
Fixes: bbac70c741 ("net/mlx5: Use newer affinity descriptor")
Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Eli Cohen <elic@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
From one hand, mlx5 driver is allowing to probe PFs in parallel.
From the other hand, devcom, which is a share resource between PFs, is
registered without any lock. This might resulted in memory problems.
Hence, use the global mlx5_dev_list_lock in order to serialize devcom
registration.
Fixes: fadd59fc50 ("net/mlx5: Introduce inter-device communication mechanism")
Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
In case devcom allocation is failed, mlx5 is always freeing the priv.
However, this priv might have been allocated by a different thread,
and freeing it might lead to use-after-free bugs.
Fix it by freeing the priv only in case it was allocated by the
running thread.
Fixes: fadd59fc50 ("net/mlx5: Introduce inter-device communication mechanism")
Signed-off-by: Shay Drory <shayd@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
devcom events are sent to all registered component. Following the
cited patch, it is possible for two components, e.g.: two eswitches,
to send devcom events, while both components are registered. This
means eswitch layer will do double un/pairing, which is double
allocation and free of resources, even though only one un/pairing is
needed. flow example:
cpu0 cpu1
---- ----
mlx5_devlink_eswitch_mode_set(dev0)
esw_offloads_devcom_init()
mlx5_devcom_register_component(esw0)
mlx5_devlink_eswitch_mode_set(dev1)
esw_offloads_devcom_init()
mlx5_devcom_register_component(esw1)
mlx5_devcom_send_event()
mlx5_devcom_send_event()
Hence, check whether the eswitches are already un/paired before
free/allocation of resources.
Fixes: 09b278462f ("net: devlink: enable parallel ops on netlink interface")
Signed-off-by: Shay Drory <shayd@nvidia.com>
Reviewed-by: Mark Bloch <mbloch@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Check in the mlx5e_ptp_poll_ts_cq context if the ptp tx sq should be woken
up. Before change, the ptp tx sq may never wake up if the ptp tx ts skb
fifo is full when mlx5e_poll_tx_cq checks if the queue should be woken up.
Fixes: 1880bc4e4a ("net/mlx5e: Add TX port timestamp support")
Signed-off-by: Rahul Rameshbabu <rrameshbabu@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Fix spacing for the error and also the correct error code pointer.
Fixes: c9b9dcb430 ("net/mlx5: Move device memory management to mlx5_core")
Signed-off-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
With introduction of post action infrastructure most of the users of encap
attribute had been modified in order to obtain the correct attribute by
calling mlx5e_tc_get_encap_attr() helper instead of assuming encap action
is always on default attribute. However, the cited commit didn't modify
mlx5e_invalidate_encap() which prevents it from destroying correct modify
header action which leads to a warning [0]. Fix the issue by using correct
attribute.
[0]:
Feb 21 09:47:35 c-237-177-40-045 kernel: WARNING: CPU: 17 PID: 654 at drivers/net/ethernet/mellanox/mlx5/core/en_tc.c:684 mlx5e_tc_attach_mod_hdr+0x1cc/0x230 [mlx5_core]
Feb 21 09:47:35 c-237-177-40-045 kernel: RIP: 0010:mlx5e_tc_attach_mod_hdr+0x1cc/0x230 [mlx5_core]
Feb 21 09:47:35 c-237-177-40-045 kernel: Call Trace:
Feb 21 09:47:35 c-237-177-40-045 kernel: <TASK>
Feb 21 09:47:35 c-237-177-40-045 kernel: mlx5e_tc_fib_event_work+0x8e3/0x1f60 [mlx5_core]
Feb 21 09:47:35 c-237-177-40-045 kernel: ? mlx5e_take_all_encap_flows+0xe0/0xe0 [mlx5_core]
Feb 21 09:47:35 c-237-177-40-045 kernel: ? lock_downgrade+0x6d0/0x6d0
Feb 21 09:47:35 c-237-177-40-045 kernel: ? lockdep_hardirqs_on_prepare+0x273/0x3f0
Feb 21 09:47:35 c-237-177-40-045 kernel: ? lockdep_hardirqs_on_prepare+0x273/0x3f0
Feb 21 09:47:35 c-237-177-40-045 kernel: process_one_work+0x7c2/0x1310
Feb 21 09:47:35 c-237-177-40-045 kernel: ? lockdep_hardirqs_on_prepare+0x3f0/0x3f0
Feb 21 09:47:35 c-237-177-40-045 kernel: ? pwq_dec_nr_in_flight+0x230/0x230
Feb 21 09:47:35 c-237-177-40-045 kernel: ? rwlock_bug.part.0+0x90/0x90
Feb 21 09:47:35 c-237-177-40-045 kernel: worker_thread+0x59d/0xec0
Feb 21 09:47:35 c-237-177-40-045 kernel: ? __kthread_parkme+0xd9/0x1d0
Fixes: 8300f22526 ("net/mlx5e: Create new flow attr for multi table actions")
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
SW Steering uses RC QP for writing STEs to ICM. This writingis done in LB
(loopback), and FL (force-loopback) QP is preferred for performance. FL is
available when RoCE is enabled or disabled based on RoCE caps.
This patch adds reading of FL capability from HCA caps in addition to the
existing reading from RoCE caps, thus fixing the case where we didn't
have loopback enabled when RoCE was disabled.
Fixes: 7304d603a5 ("net/mlx5: DR, Add support for force-loopback QP")
Signed-off-by: Itamar Gozlan <igozlan@nvidia.com>
Signed-off-by: Yevgeny Kliteynik <kliteyn@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
When calculating crc for hash index we use the function crc32 that
calculates for little-endian (LE) arch.
Then we convert it to network endianness using htonl(), but it's wrong
to do the conversion in BE archs since the crc32 value is already LE.
The solution is to switch the bytes from the crc result for all types
of arc.
Fixes: 40416d8ede ("net/mlx5: DR, Replace CRC32 implementation to use kernel lib")
Signed-off-by: Erez Shitrit <erezsh@nvidia.com>
Reviewed-by: Alex Vesker <valex@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
