There is an issue with one-step timestamp based on UDP/IP. The peer will
discard the sync packet because of the wrong UDP checksum. For ENETC v1,
the software needs to update the UDP checksum when updating the
originTimestamp field, so that the hardware can correctly update the UDP
checksum when updating the correction field. Otherwise, the UDP checksum
in the sync packet will be wrong.
Fixes: 7294380c52 ("enetc: support PTP Sync packet one-step timestamping")
Cc: stable@vger.kernel.org
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20250224111251.1061098-6-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Actually ENETC VFs do not support HWTSTAMP_TX_ONESTEP_SYNC because only
ENETC PF can access PMa_SINGLE_STEP registers. And there will be a crash
if VFs are used to test one-step timestamp, the crash log as follows.
[ 129.110909] Unable to handle kernel paging request at virtual address 00000000000080c0
[ 129.287769] Call trace:
[ 129.290219] enetc_port_mac_wr+0x30/0xec (P)
[ 129.294504] enetc_start_xmit+0xda4/0xe74
[ 129.298525] enetc_xmit+0x70/0xec
[ 129.301848] dev_hard_start_xmit+0x98/0x118
Fixes: 41514737ec ("enetc: add get_ts_info interface for ethtool")
Cc: stable@vger.kernel.org
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20250224111251.1061098-5-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Set NETIF_F_GSO_UDP_L4 bit of hw_features and features because i.MX95
enetc and LS1028A driver implements UDP segmentation.
- i.MX95 ENETC supports UDP segmentation via LSO.
- LS1028A ENETC supports UDP segmentation since the commit 3d5b459ba0
("net: tso: add UDP segmentation support").
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Link: https://patch.msgid.link/20241219054755.1615626-5-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
ENETC rev 4.1 supports large send offload (LSO), segmenting large TCP
and UDP transmit units into multiple Ethernet frames. To support LSO,
software needs to fill some auxiliary information in Tx BD, such as LSO
header length, frame length, LSO maximum segment size, etc.
At 1Gbps link rate, TCP segmentation was tested using iperf3, and the
CPU performance before and after applying the patch was compared through
the top command. It can be seen that LSO saves a significant amount of
CPU cycles compared to software TSO.
Before applying the patch:
%Cpu(s): 0.1 us, 4.1 sy, 0.0 ni, 85.7 id, 0.0 wa, 0.5 hi, 9.7 si
After applying the patch:
%Cpu(s): 0.1 us, 2.3 sy, 0.0 ni, 94.5 id, 0.0 wa, 0.4 hi, 2.6 si
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Link: https://patch.msgid.link/20241219054755.1615626-4-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The max chained Tx BDs of latest ENETC (i.MX95 ENETC, rev 4.1) has been
increased to 63, but since the range of MAX_SKB_FRAGS is 17~45, so for
i.MX95 ENETC and later revision, it is better to set ENETC4_MAX_SKB_FRAGS
to MAX_SKB_FRAGS.
In addition, add max_frags in struct enetc_drvdata to indicate the max
chained BDs supported by device. Because the max number of chained BDs
supported by LS1028A and i.MX95 ENETC is different.
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20241219054755.1615626-3-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In addition to supporting Rx checksum offload, i.MX95 ENETC also supports
Tx checksum offload. The transmit checksum offload is implemented through
the Tx BD. To support Tx checksum offload, software needs to fill some
auxiliary information in Tx BD, such as IP version, IP header offset and
size, whether L4 is UDP or TCP, etc.
Same as Rx checksum offload, Tx checksum offload capability isn't defined
in register, so tx_csum bit is added to struct enetc_drvdata to indicate
whether the device supports Tx checksum offload.
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Link: https://patch.msgid.link/20241219054755.1615626-2-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Both ENETC PF and VF drivers share enetc_setup_tc_mqprio() to configure
MQPRIO. And enetc_setup_tc_mqprio() calls enetc_change_preemptible_tcs()
to configure preemptible TCs. However, only PF is able to configure
preemptible TCs. Because only PF has related registers, while VF does not
have these registers. So for VF, its hw->port pointer is NULL. Therefore,
VF will access an invalid pointer when accessing a non-existent register,
which will cause a crash issue. The simplified log is as follows.
root@ls1028ardb:~# tc qdisc add dev eno0vf0 parent root handle 100: \
mqprio num_tc 4 map 0 0 1 1 2 2 3 3 queues 1@0 1@1 1@2 1@3 hw 1
[ 187.290775] Unable to handle kernel paging request at virtual address 0000000000001f00
[ 187.424831] pc : enetc_mm_commit_preemptible_tcs+0x1c4/0x400
[ 187.430518] lr : enetc_mm_commit_preemptible_tcs+0x30c/0x400
[ 187.511140] Call trace:
[ 187.513588] enetc_mm_commit_preemptible_tcs+0x1c4/0x400
[ 187.518918] enetc_setup_tc_mqprio+0x180/0x214
[ 187.523374] enetc_vf_setup_tc+0x1c/0x30
[ 187.527306] mqprio_enable_offload+0x144/0x178
[ 187.531766] mqprio_init+0x3ec/0x668
[ 187.535351] qdisc_create+0x15c/0x488
[ 187.539023] tc_modify_qdisc+0x398/0x73c
[ 187.542958] rtnetlink_rcv_msg+0x128/0x378
[ 187.547064] netlink_rcv_skb+0x60/0x130
[ 187.550910] rtnetlink_rcv+0x18/0x24
[ 187.554492] netlink_unicast+0x300/0x36c
[ 187.558425] netlink_sendmsg+0x1a8/0x420
[ 187.606759] ---[ end trace 0000000000000000 ]---
In addition, some PFs also do not support configuring preemptible TCs,
such as eno1 and eno3 on LS1028A. It won't crash like it does for VFs,
but we should prevent these PFs from accessing these unimplemented
registers.
Fixes: 827145392a ("net: enetc: only commit preemptible TCs to hardware when MM TX is active")
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Suggested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Configuring TSN (Qbv, Qbu, PSFP) capabilities requires access to port
registers, which are available to the PSI but not the VSI.
Yet, the SI port capability register 0 (PSICAPR0), exposed to both PSIs
and VSIs, presents the same capabilities to the VF as to the PF, thus
leading the VF driver into thinking it can configure these features.
In the case of ENETC_SI_F_QBU, having it set in the VF leads to a crash:
root@ls1028ardb:~# tc qdisc add dev eno0vf0 parent root handle 100: \
mqprio num_tc 4 map 0 0 1 1 2 2 3 3 queues 1@0 1@1 1@2 1@3 hw 1
[ 187.290775] Unable to handle kernel paging request at virtual address 0000000000001f00
[ 187.424831] pc : enetc_mm_commit_preemptible_tcs+0x1c4/0x400
[ 187.430518] lr : enetc_mm_commit_preemptible_tcs+0x30c/0x400
[ 187.511140] Call trace:
[ 187.513588] enetc_mm_commit_preemptible_tcs+0x1c4/0x400
[ 187.518918] enetc_setup_tc_mqprio+0x180/0x214
[ 187.523374] enetc_vf_setup_tc+0x1c/0x30
[ 187.527306] mqprio_enable_offload+0x144/0x178
[ 187.531766] mqprio_init+0x3ec/0x668
[ 187.535351] qdisc_create+0x15c/0x488
[ 187.539023] tc_modify_qdisc+0x398/0x73c
[ 187.542958] rtnetlink_rcv_msg+0x128/0x378
[ 187.547064] netlink_rcv_skb+0x60/0x130
[ 187.550910] rtnetlink_rcv+0x18/0x24
[ 187.554492] netlink_unicast+0x300/0x36c
[ 187.558425] netlink_sendmsg+0x1a8/0x420
[ 187.606759] ---[ end trace 0000000000000000 ]---
while the other TSN features in the VF are harmless, because the
net_device_ops used for the VF driver do not expose entry points for
these other features.
These capability bits are in the process of being defeatured from the SI
registers. We should read them from the port capability register, where
they are also present, and which is naturally only exposed to the PF.
The change to blame (relevant for stable backports) is the one where
this started being a problem, aka when the kernel started to crash due
to the wrong capability seen by the VF driver.
Fixes: 827145392a ("net: enetc: only commit preemptible TCs to hardware when MM TX is active")
Reported-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The i.MX95 ENETC has been upgraded to revision 4.1, which is different
from the LS1028A ENETC (revision 1.0) except for the SI part. Therefore,
the fsl-enetc driver is incompatible with i.MX95 ENETC PF. So add new
nxp-enetc4 driver to support i.MX95 ENETC PF, and this driver will be
used to support the ENETC PF with major revision 4 for other SoCs in the
future.
Currently, the nxp-enetc4 driver only supports basic transmission feature
for i.MX95 ENETC PF, the more basic and advanced features will be added
in the subsequent patches. In addition, PCS support has not been added
yet, so 10G ENETC (ENETC instance 2) is not supported now.
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
There is a situation where num_tx_rings cannot be divided by bdr_int_num.
For example, num_tx_rings is 8 and bdr_int_num is 3. According to the
previous logic, this results in two tx_bdr corresponding memories not
being allocated, so when sending packets to tx ring 6 or 7, wild pointers
will be accessed. Of course, this issue doesn't exist on LS1028A, because
its num_tx_rings is 8, and bdr_int_num is either 1 or 2. However, there
is a risk for the upcoming i.MX95. Therefore, it is necessary to ensure
that each tx_bdr can be allocated to the corresponding memory.
Signed-off-by: Clark Wang <xiaoning.wang@nxp.com>
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Extract enetc_int_vector_init() and enetc_int_vector_destroy() from
enetc_alloc_msix() so that the code is more concise and readable.
Signed-off-by: Clark Wang <xiaoning.wang@nxp.com>
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The verdor ID and device ID of i.MX95 EMDIO are different from LS1028A
EMDIO, so add new vendor ID and device ID to pci_device_id table to
support i.MX95 EMDIO.
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ERR050089 workaround causes performance degradation and potential
functional issues (e.g., RCU stalls) under certain workloads. Since
new SoCs like i.MX95 do not require this workaround, use a static key
to compile out enetc_lock_mdio() and enetc_unlock_mdio() at runtime,
improving performance and avoiding unnecessary logic.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Compile enetc_pf_common.c as a standalone module to allow shared usage
between ENETC v1 and v4 PF drivers. Add struct enetc_pf_ops to register
different hardware operation interfaces for both ENETC v1 and v4 PF
drivers.
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The ENETC PF driver of LS1028A (rev 1.0) is incompatible with the version
used on the i.MX95 platform (rev 4.1), except for the station interface
(SI) part. To reduce code redundancy and prepare for a new driver for rev
4.1 and later, extract shared interfaces from enetc_pf.c and move them to
enetc_pf_common.c. This refactoring lays the groundwork for compiling
enetc_pf_common.c into a shared driver for both platforms' PF drivers.
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
The netc-blk-ctrl driver is used to configure Integrated Endpoint
Register Block (IERB) and Privileged Register Block (PRB) of NETC.
For i.MX platforms, it is also used to configure the NETCMIX block.
The IERB contains registers that are used for pre-boot initialization,
debug, and non-customer configuration. The PRB controls global reset
and global error handling for NETC. The NETCMIX block is mainly used
to set MII protocol and PCS protocol of the links, it also contains
settings for some other functions.
Note the IERB configuration registers can only be written after being
unlocked by PRB, otherwise, all write operations are inhibited. A warm
reset is performed when the IERB is unlocked, and it results in an FLR
to all NETC devices. Therefore, all NETC device drivers must be probed
or initialized after the warm reset is finished.
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Frank Li <Frank.Li@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
In the previous implementation, vf_state is allocated memory only when VF
is enabled. However, net_device_ops::ndo_set_vf_mac() may be called before
VF is enabled to configure the MAC address of VF. If this is the case,
enetc_pf_set_vf_mac() will access vf_state, resulting in access to a null
pointer. The simplified error log is as follows.
root@ls1028ardb:~# ip link set eno0 vf 1 mac 00:0c:e7:66:77:89
[ 173.543315] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000004
[ 173.637254] pc : enetc_pf_set_vf_mac+0x3c/0x80 Message from sy
[ 173.641973] lr : do_setlink+0x4a8/0xec8
[ 173.732292] Call trace:
[ 173.734740] enetc_pf_set_vf_mac+0x3c/0x80
[ 173.738847] __rtnl_newlink+0x530/0x89c
[ 173.742692] rtnl_newlink+0x50/0x7c
[ 173.746189] rtnetlink_rcv_msg+0x128/0x390
[ 173.750298] netlink_rcv_skb+0x60/0x130
[ 173.754145] rtnetlink_rcv+0x18/0x24
[ 173.757731] netlink_unicast+0x318/0x380
[ 173.761665] netlink_sendmsg+0x17c/0x3c8
Fixes: d4fd0404c1 ("enetc: Introduce basic PF and VF ENETC ethernet drivers")
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20241031060247.1290941-2-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net_dim() is currently passed a struct dim_sample argument by value.
struct dim_sample is 24 bytes. Since this is greater 16 bytes, x86-64
passes it on the stack. All callers have already initialized dim_sample
on the stack, so passing it by value requires pushing a duplicated copy
to the stack. Either witing to the stack and immediately reading it, or
perhaps dereferencing addresses relative to the stack pointer in a chain
of push instructions, seems to perform quite poorly.
In a heavy TCP workload, mlx5e_handle_rx_dim() consumes 3% of CPU time,
94% of which is attributed to the first push instruction to copy
dim_sample on the stack for the call to net_dim():
// Call ktime_get()
0.26 |4ead2: call 4ead7 <mlx5e_handle_rx_dim+0x47>
// Pass the address of struct dim in %rdi
|4ead7: lea 0x3d0(%rbx),%rdi
// Set dim_sample.pkt_ctr
|4eade: mov %r13d,0x8(%rsp)
// Set dim_sample.byte_ctr
|4eae3: mov %r12d,0xc(%rsp)
// Set dim_sample.event_ctr
0.15 |4eae8: mov %bp,0x10(%rsp)
// Duplicate dim_sample on the stack
94.16 |4eaed: push 0x10(%rsp)
2.79 |4eaf1: push 0x10(%rsp)
0.07 |4eaf5: push %rax
// Call net_dim()
0.21 |4eaf6: call 4eafb <mlx5e_handle_rx_dim+0x6b>
To allow the caller to reuse the struct dim_sample already on the stack,
pass the struct dim_sample by reference to net_dim().
Signed-off-by: Caleb Sander Mateos <csander@purestorage.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Shannon Nelson <shannon.nelson@amd.com>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Reviewed-by: Arthur Kiyanovski <akiyano@amazon.com>
Reviewed-by: Louis Peens <louis.peens@corigine.com>
Link: https://patch.msgid.link/20241031002326.3426181-2-csander@purestorage.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The MAC address of VF can be configured through the mailbox mechanism of
ENETC, but the previous implementation forgot to set the MAC address in
net_device, resulting in the SMAC of the sent frames still being the old
MAC address. Since the MAC address in the hardware has been changed, Rx
cannot receive frames with the DMAC address as the new MAC address. The
most obvious phenomenon is that after changing the MAC address, we can
see that the MAC address of eno0vf0 has not changed through the "ifconfig
eno0vf0" command and the IP address cannot be obtained .
root@ls1028ardb:~# ifconfig eno0vf0 down
root@ls1028ardb:~# ifconfig eno0vf0 hw ether 00:04:9f:3a:4d:56 up
root@ls1028ardb:~# ifconfig eno0vf0
eno0vf0: flags=4163<UP,BROADCAST,RUNNING,MULTICAST> mtu 1500
ether 66:36:2c:3b:87:76 txqueuelen 1000 (Ethernet)
RX packets 794 bytes 69239 (69.2 KB)
RX errors 0 dropped 0 overruns 0 frame 0
TX packets 11 bytes 2226 (2.2 KB)
TX errors 0 dropped 0 overruns 0 carrier 0 collisions 0
Fixes: beb74ac878 ("enetc: Add vf to pf messaging support")
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Reviewed-by: Claudiu Manoil <claudiu.manoil@nxp.com>
Link: https://patch.msgid.link/20241029090406.841836-1-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When running "xdp-bench tx eno0" to test the XDP_TX feature of ENETC
on LS1028A, it was found that if the command was re-run multiple times,
Rx could not receive the frames, and the result of xdp-bench showed
that the rx rate was 0.
root@ls1028ardb:~# ./xdp-bench tx eno0
Hairpinning (XDP_TX) packets on eno0 (ifindex 3; driver fsl_enetc)
Summary 2046 rx/s 0 err,drop/s
Summary 0 rx/s 0 err,drop/s
Summary 0 rx/s 0 err,drop/s
Summary 0 rx/s 0 err,drop/s
By observing the Rx PIR and CIR registers, CIR is always 0x7FF and
PIR is always 0x7FE, which means that the Rx ring is full and can no
longer accommodate other Rx frames. Therefore, the problem is caused
by the Rx BD ring not being cleaned up.
Further analysis of the code revealed that the Rx BD ring will only
be cleaned if the "cleaned_cnt > xdp_tx_in_flight" condition is met.
Therefore, some debug logs were added to the driver and the current
values of cleaned_cnt and xdp_tx_in_flight were printed when the Rx
BD ring was full. The logs are as follows.
[ 178.762419] [XDP TX] >> cleaned_cnt:1728, xdp_tx_in_flight:2140
[ 178.771387] [XDP TX] >> cleaned_cnt:1941, xdp_tx_in_flight:2110
[ 178.776058] [XDP TX] >> cleaned_cnt:1792, xdp_tx_in_flight:2110
From the results, the max value of xdp_tx_in_flight has reached 2140.
However, the size of the Rx BD ring is only 2048. So xdp_tx_in_flight
did not drop to 0 after enetc_stop() is called and the driver does not
clear it. The root cause is that NAPI is disabled too aggressively,
without having waited for the pending XDP_TX frames to be transmitted,
and their buffers recycled, so that xdp_tx_in_flight cannot naturally
drop to 0. Later, enetc_free_tx_ring() does free those stale, unsent
XDP_TX packets, but it is not coded up to also reset xdp_tx_in_flight,
hence the manifestation of the bug.
One option would be to cover this extra condition in enetc_free_tx_ring(),
but now that the ENETC_TX_DOWN exists, we have created a window at
the beginning of enetc_stop() where NAPI can still be scheduled, but
any concurrent enqueue will be blocked. Therefore, enetc_wait_bdrs()
and enetc_disable_tx_bdrs() can be called with NAPI still scheduled,
and it is guaranteed that this will not wait indefinitely, but instead
give us an indication that the pending TX frames have orderly dropped
to zero. Only then should we call napi_disable().
This way, enetc_free_tx_ring() becomes entirely redundant and can be
dropped as part of subsequent cleanup.
The change also refactors enetc_start() so that it looks like the
mirror opposite procedure of enetc_stop().
Fixes: ff58fda090 ("net: enetc: prioritize ability to go down over packet processing")
Cc: stable@vger.kernel.org
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20241010092056.298128-5-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The Tx BD rings are disabled first in enetc_stop() and the driver
waits for them to become empty. This operation is not safe while
the ring is actively transmitting frames, and will cause the ring
to not be empty and hardware exception. As described in the NETC
block guide, software should only disable an active Tx ring after
all pending ring entries have been consumed (i.e. when PI = CI).
Disabling a transmit ring that is actively processing BDs risks
a HW-SW race hazard whereby a hardware resource becomes assigned
to work on one or more ring entries only to have those entries be
removed due to the ring becoming disabled.
When testing XDP_REDIRECT feautre, although all frames were blocked
from being put into Tx rings during ring reconfiguration, the similar
warning log was still encountered:
fsl_enetc 0000:00:00.2 eno2: timeout for tx ring #6 clear
fsl_enetc 0000:00:00.2 eno2: timeout for tx ring #7 clear
The reason is that when there are still unsent frames in the Tx ring,
disabling the Tx ring causes the remaining frames to be unable to be
sent out. And the Tx ring cannot be restored, which means that even
if the xdp program is uninstalled, the Tx frames cannot be sent out
anymore. Therefore, correct the operation order in enect_start() and
enect_stop().
Fixes: ff58fda090 ("net: enetc: prioritize ability to go down over packet processing")
Cc: stable@vger.kernel.org
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20241010092056.298128-4-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When testing the XDP_REDIRECT function on the LS1028A platform, we
found a very reproducible issue that the Tx frames can no longer be
sent out even if XDP_REDIRECT is turned off. Specifically, if there
is a lot of traffic on Rx direction, when XDP_REDIRECT is turned on,
the console may display some warnings like "timeout for tx ring #6
clear", and all redirected frames will be dropped, the detailed log
is as follows.
root@ls1028ardb:~# ./xdp-bench redirect eno0 eno2
Redirecting from eno0 (ifindex 3; driver fsl_enetc) to eno2 (ifindex 4; driver fsl_enetc)
[203.849809] fsl_enetc 0000:00:00.2 eno2: timeout for tx ring #5 clear
[204.006051] fsl_enetc 0000:00:00.2 eno2: timeout for tx ring #6 clear
[204.161944] fsl_enetc 0000:00:00.2 eno2: timeout for tx ring #7 clear
eno0->eno2 1420505 rx/s 1420590 err,drop/s 0 xmit/s
xmit eno0->eno2 0 xmit/s 1420590 drop/s 0 drv_err/s 15.71 bulk-avg
eno0->eno2 1420484 rx/s 1420485 err,drop/s 0 xmit/s
xmit eno0->eno2 0 xmit/s 1420485 drop/s 0 drv_err/s 15.71 bulk-avg
By analyzing the XDP_REDIRECT implementation of enetc driver, the
driver will reconfigure Tx and Rx BD rings when a bpf program is
installed or uninstalled, but there is no mechanisms to block the
redirected frames when enetc driver reconfigures rings. Similarly,
XDP_TX verdicts on received frames can also lead to frames being
enqueued in the Tx rings. Because XDP ignores the state set by the
netif_tx_wake_queue() API, so introduce the ENETC_TX_DOWN flag to
suppress transmission of XDP frames.
Fixes: c33bfaf91c ("net: enetc: set up XDP program under enetc_reconfigure()")
Cc: stable@vger.kernel.org
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20241010092056.298128-3-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The xdp_drops statistic indicates the number of XDP frames dropped in
the Rx direction. However, enetc_xdp_drop() is also used in XDP_TX and
XDP_REDIRECT actions. If frame loss occurs in these two actions, the
frames loss count should not be included in xdp_drops, because there
are already xdp_tx_drops and xdp_redirect_failures to count the frame
loss of these two actions, so it's better to remove xdp_drops statistic
from enetc_xdp_drop() and increase xdp_drops in XDP_DROP action.
Fixes: 7ed2bc8007 ("net: enetc: add support for XDP_TX")
Cc: stable@vger.kernel.org
Signed-off-by: Wei Fang <wei.fang@nxp.com>
Reviewed-by: Maciej Fijalkowski <maciej.fijalkowski@intel.com>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://patch.msgid.link/20241010092056.298128-2-wei.fang@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
asm/unaligned.h is always an include of asm-generic/unaligned.h;
might as well move that thing to linux/unaligned.h and include
that - there's nothing arch-specific in that header.
auto-generated by the following:
for i in `git grep -l -w asm/unaligned.h`; do
sed -i -e "s/asm\/unaligned.h/linux\/unaligned.h/" $i
done
for i in `git grep -l -w asm-generic/unaligned.h`; do
sed -i -e "s/asm-generic\/unaligned.h/linux\/unaligned.h/" $i
done
git mv include/asm-generic/unaligned.h include/linux/unaligned.h
git mv tools/include/asm-generic/unaligned.h tools/include/linux/unaligned.h
sed -i -e "/unaligned.h/d" include/asm-generic/Kbuild
sed -i -e "s/__ASM_GENERIC/__LINUX/" include/linux/unaligned.h tools/include/linux/unaligned.h
The enetc driver uses ifdefs when checking whether
CONFIG_FSL_ENETC_PTP_CLOCK is enabled in a number of places. This works
if the driver is built-in but fails if the driver is available as a
kernel module. Replace the instances of ifdef with use of the IS_ENABLED
macro, that will evaluate as true when this feature is built as a kernel
module and follows the kernel's coding style.
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Signed-off-by: Martyn Welch <martyn.welch@collabora.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://patch.msgid.link/20240912173742.484549-1-martyn.welch@collabora.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In prevision to add new UAPI for hwtstamp we will be limited to the struct
ethtool_ts_info that is currently passed in fixed binary format through the
ETHTOOL_GET_TS_INFO ethtool ioctl. It would be good if new kernel code
already started operating on an extensible kernel variant of that
structure, similar in concept to struct kernel_hwtstamp_config vs struct
hwtstamp_config.
Since struct ethtool_ts_info is in include/uapi/linux/ethtool.h, here
we introduce the kernel-only structure in include/linux/ethtool.h.
The manual copy is then made in the function called by ETHTOOL_GET_TS_INFO.
Acked-by: Shannon Nelson <shannon.nelson@amd.com>
Acked-by: Alexandra Winter <wintera@linux.ibm.com>
Signed-off-by: Kory Maincent <kory.maincent@bootlin.com>
Link: https://patch.msgid.link/20240709-feature_ptp_netnext-v17-6-b5317f50df2a@bootlin.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
As clang points out, the error message in enetc_setup_xdp_prog()
still does not fit in the buffer and will be truncated:
drivers/net/ethernet/freescale/enetc/enetc.c:2771:3: error: 'snprintf' will always be truncated; specified size is 80, but format string expands to at least 87 [-Werror,-Wformat-truncation]
Replace it with an even shorter message that should fit.
Fixes: f968c56417 ("net: enetc: shorten enetc_setup_xdp_prog() error message to fit NETLINK_MAX_FMTMSG_LEN")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20240326223825.4084412-3-arnd@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In order to pass EEE link modes beyond bit 32 to userspace we have to
complement the 32 bit bitmaps in struct ethtool_eee with linkmode
bitmaps. Therefore, similar to ethtool_link_settings and
ethtool_link_ksettings, add a struct ethtool_keee. In a first step
it's an identical copy of ethtool_eee. This patch simply does a
s/ethtool_eee/ethtool_keee/g for all users.
No functional change intended.
Suggested-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
The driver code proper is handled by the lynx_pcs. The enetc just needs
to populate phylink's supported_interfaces array, and return true for
this phy-mode in enetc_port_has_pcs(), such that it creates an internal
MDIO bus through which the Lynx PCS registers are accessed.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20240103113445.3892971-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The get/set_rxfh ethtool ops currently takes the rxfh (RSS) parameters
as direct function arguments. This will force us to change the API (and
all drivers' functions) every time some new parameters are added.
This is part 1/2 of the fix, as suggested in [1]:
- First simplify the code by always providing a pointer to all params
(indir, key and func); the fact that some of them may be NULL seems
like a weird historic thing or a premature optimization.
It will simplify the drivers if all pointers are always present.
- Then make the functions take a dev pointer, and a pointer to a
single struct wrapping all arguments. The set_* should also take
an extack.
Link: https://lore.kernel.org/netdev/20231121152906.2dd5f487@kernel.org/ [1]
Suggested-by: Jakub Kicinski <kuba@kernel.org>
Suggested-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Ahmed Zaki <ahmed.zaki@intel.com>
Link: https://lore.kernel.org/r/20231213003321.605376-2-ahmed.zaki@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
NETLINK_MAX_FMTMSG_LEN is currently hardcoded to 80, and we provide an
error printf-formatted string having 96 characters including the
terminating \0. Assuming each %d (representing a queue) gets replaced by
a number having at most 2 digits (a reasonable assumption), the final
string is also 96 characters wide, which is too much.
Reduce the verbiage a bit by removing some (partially) redundant words,
which makes the new printf-formatted string be 73 characters wide with
the trailing newline.
Fixes: 800db2d125 ("net: enetc: ensure we always have a minimum number of TXQs for stack")
Reported-by: kernel test robot <lkp@intel.com>
Closes: https://lore.kernel.org/lkml/202311061336.4dsWMT1h-lkp@intel.com/
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20231106160311.616118-1-vladimir.oltean@nxp.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pull networking updates from Jakub Kicinski:
"Including fixes from netfilter and bpf.
Current release - regressions:
- eth: stmmac: fix failure to probe without MAC interface specified
Current release - new code bugs:
- docs: netlink: fix missing classic_netlink doc reference
Previous releases - regressions:
- deal with integer overflows in kmalloc_reserve()
- use sk_forward_alloc_get() in sk_get_meminfo()
- bpf_sk_storage: fix the missing uncharge in sk_omem_alloc
- fib: avoid warn splat in flow dissector after packet mangling
- skb_segment: call zero copy functions before using skbuff frags
- eth: sfc: check for zero length in EF10 RX prefix
Previous releases - always broken:
- af_unix: fix msg_controllen test in scm_pidfd_recv() for
MSG_CMSG_COMPAT
- xsk: fix xsk_build_skb() dereferencing possible ERR_PTR()
- netfilter:
- nft_exthdr: fix non-linear header modification
- xt_u32, xt_sctp: validate user space input
- nftables: exthdr: fix 4-byte stack OOB write
- nfnetlink_osf: avoid OOB read
- one more fix for the garbage collection work from last release
- igmp: limit igmpv3_newpack() packet size to IP_MAX_MTU
- bpf, sockmap: fix preempt_rt splat when using raw_spin_lock_t
- handshake: fix null-deref in handshake_nl_done_doit()
- ip: ignore dst hint for multipath routes to ensure packets are
hashed across the nexthops
- phy: micrel:
- correct bit assignments for cable test errata
- disable EEE according to the KSZ9477 errata
Misc:
- docs/bpf: document compile-once-run-everywhere (CO-RE) relocations
- Revert "net: macsec: preserve ingress frame ordering", it appears
to have been developed against an older kernel, problem doesn't
exist upstream"
* tag 'net-6.6-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (95 commits)
net: enetc: distinguish error from valid pointers in enetc_fixup_clear_rss_rfs()
Revert "net: team: do not use dynamic lockdep key"
net: hns3: remove GSO partial feature bit
net: hns3: fix the port information display when sfp is absent
net: hns3: fix invalid mutex between tc qdisc and dcb ets command issue
net: hns3: fix debugfs concurrency issue between kfree buffer and read
net: hns3: fix byte order conversion issue in hclge_dbg_fd_tcam_read()
net: hns3: Support query tx timeout threshold by debugfs
net: hns3: fix tx timeout issue
net: phy: Provide Module 4 KSZ9477 errata (DS80000754C)
netfilter: nf_tables: Unbreak audit log reset
netfilter: ipset: add the missing IP_SET_HASH_WITH_NET0 macro for ip_set_hash_netportnet.c
netfilter: nft_set_rbtree: skip sync GC for new elements in this transaction
netfilter: nf_tables: uapi: Describe NFTA_RULE_CHAIN_ID
netfilter: nfnetlink_osf: avoid OOB read
netfilter: nftables: exthdr: fix 4-byte stack OOB write
selftests/bpf: Check bpf_sk_storage has uncharged sk_omem_alloc
bpf: bpf_sk_storage: Fix the missing uncharge in sk_omem_alloc
bpf: bpf_sk_storage: Fix invalid wait context lockdep report
s390/bpf: Pass through tail call counter in trampolines
...
Pull modules updates from Luis Chamberlain:
"Summary of the changes worth highlighting from most interesting to
boring below:
- Christoph Hellwig's symbol_get() fix to Nvidia's efforts to
circumvent the protection he put in place in year 2020 to prevent
proprietary modules from using GPL only symbols, and also ensuring
proprietary modules which export symbols grandfather their taint.
That was done through year 2020 commit 262e6ae708 ("modules:
inherit TAINT_PROPRIETARY_MODULE"). Christoph's new fix is done by
clarifing __symbol_get() was only ever intended to prevent module
reference loops by Linux kernel modules and so making it only find
symbols exported via EXPORT_SYMBOL_GPL(). The circumvention tactic
used by Nvidia was to use symbol_get() to purposely swift through
proprietary module symbols and completely bypass our traditional
EXPORT_SYMBOL*() annotations and community agreed upon
restrictions.
A small set of preamble patches fix up a few symbols which just
needed adjusting for this on two modules, the rtc ds1685 and the
networking enetc module. Two other modules just needed some build
fixing and removal of use of __symbol_get() as they can't ever be
modular, as was done by Arnd on the ARM pxa module and Christoph
did on the mmc au1xmmc driver.
This is a good reminder to us that symbol_get() is just a hack to
address things which should be fixed through Kconfig at build time
as was done in the later patches, and so ultimately it should just
go.
- Extremely late minor fix for old module layout 055f23b74b
("module: check for exit sections in layout_sections() instead of
module_init_section()") by James Morse for arm64. Note that this
layout thing is old, it is *not* Song Liu's commit ac3b432839
("module: replace module_layout with module_memory"). The issue
however is very odd to run into and so there was no hurry to get
this in fast.
- Although the fix did not go through the modules tree I'd like to
highlight the fix by Peter Zijlstra in commit 5409730962
("x86/static_call: Fix __static_call_fixup()") now merged in your
tree which came out of what was originally suspected to be a
fallout of the the newer module layout changes by Song Liu commit
ac3b432839 ("module: replace module_layout with module_memory")
instead of module_init_section()"). Thanks to the report by
Christian Bricart and the debugging by Song Liu & Peter that turned
to be noted as a kernel regression in place since v5.19 through
commit ee88d363d1 ("x86,static_call: Use alternative RET
encoding").
I highlight this to reflect and clarify that we haven't seen more
fallout from ac3b432839 ("module: replace module_layout with
module_memory").
- RISC-V toolchain got mapping symbol support which prefix symbols
with "$" to help with alignment considerations for disassembly.
This is used to differentiate between incompatible instruction
encodings when disassembling. RISC-V just matches what ARM/AARCH64
did for alignment considerations and Palmer Dabbelt extended
is_mapping_symbol() to accept these symbols for RISC-V. We already
had support for this for all architectures but it also checked for
the second character, the RISC-V check Dabbelt added was just for
the "$". After a bit of testing and fallout on linux-next and based
on feedback from Masahiro Yamada it was decided to simplify the
check and treat the first char "$" as unique for all architectures,
and so we no make is_mapping_symbol() for all archs if the symbol
starts with "$".
The most relevant commit for this for RISC-V on binutils was:
https://sourceware.org/pipermail/binutils/2021-July/117350.html
- A late fix by Andrea Righi (today) to make module zstd
decompression use vmalloc() instead of kmalloc() to account for
large compressed modules. I suspect we'll see similar things for
other decompression algorithms soon.
- samples/hw_breakpoint minor fixes by Rong Tao, Arnd Bergmann and
Chen Jiahao"
* tag 'modules-6.6-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/mcgrof/linux:
module/decompress: use vmalloc() for zstd decompression workspace
kallsyms: Add more debug output for selftest
ARM: module: Use module_init_layout_section() to spot init sections
arm64: module: Use module_init_layout_section() to spot init sections
module: Expose module_init_layout_section()
modules: only allow symbol_get of EXPORT_SYMBOL_GPL modules
rtc: ds1685: use EXPORT_SYMBOL_GPL for ds1685_rtc_poweroff
net: enetc: use EXPORT_SYMBOL_GPL for enetc_phc_index
mmc: au1xmmc: force non-modular build and remove symbol_get usage
ARM: pxa: remove use of symbol_get()
samples/hw_breakpoint: mark sample_hbp as static
samples/hw_breakpoint: fix building without module unloading
samples/hw_breakpoint: Fix kernel BUG 'invalid opcode: 0000'
modpost, kallsyms: Treat add '$'-prefixed symbols as mapping symbols
kernel: params: Remove unnecessary ‘0’ values from err
module: Ignore RISC-V mapping symbols too
