Hardware is unable to realize all legal firmware interface state values
for hash_type. For example, if 4-tuple TCP_IPV4 hash is enabled,
4-tuple UDP_IPV4 hash must also be enabled. By providing the bits the
user intended to change instead of the possible illegal intermediate
states, the firmware is able to make better compromises when deciding
which bits to ignore.
With this new mechansim, we can now report the actual configured hash
back to the user. Add bnxt_hwrm_update_rss_hash_cfg() to report the
actual hash after user configuration.
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Extract common code into a new function. This will avoid duplication
in the next patch, which changes the update algorithm for both the P5
and legacy code paths.
No functional changes.
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com>
Signed-off-by: Michael Chan <michael.chan@broadcom.com>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Xin Long says:
====================
net: add helper support in tc act_ct for ovs offloading
Ilya reported an issue that FTP traffic would be broken when the OVS flow
with ct(commit,alg=ftp) installed in the OVS kernel module, and it was
caused by that TC didn't support the ftp helper offloaded from OVS.
This patchset is to add the helper support in act_ct for OVS offloading
in kernel net/sched.
The 1st and 2nd patches move some common code into nf_conntrack_helper from
openvswitch so that they could be used by net/sched in the 4th patch (Note
there are still some other common code used in both OVS and TC, and I will
extract it in other patches). The 3rd patch extracts another function in
net/sched to make the 4th patch easier to write. The 4th patch adds this
feature in net/sched.
The user space part will be added in another patch, and with it these OVS
flows (FTP over SNAT) can be used to test this feature:
table=0, in_port=veth1,tcp,tcp_dst=2121,ct_state=-trk \
actions=ct(table=1, nat), normal
table=0, in_port=veth2,tcp,ct_state=-trk actions=ct(table=1, nat)
table=0, in_port=veth1,tcp,ct_state=-trk actions=ct(table=0, nat)
table=0, in_port=veth1,tcp,ct_state=+trk+rel actions=ct(commit, nat),normal
table=0, in_port=veth1,tcp,ct_state=+trk+est actions=veth2"
table=1, in_port=veth1,tcp,tcp_dst=2121,ct_state=+trk+new \
actions=ct(commit, nat(src=7.7.16.1), alg=ftp),normal"
table=1, in_port=veth1,tcp,tcp_dst=2121,ct_state=+trk+est actions=veth2"
table=1, in_port=veth2,tcp,ct_state=+trk+est actions=veth1"
====================
Link: https://lore.kernel.org/r/cover.1667766782.git.lucien.xin@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
This patch is to add helper support in act_ct for OVS actions=ct(alg=xxx)
offloading, which is corresponding to Commit cae3a26275 ("openvswitch:
Allow attaching helpers to ct action") in OVS kernel part.
The difference is when adding TC actions family and proto cannot be got
from the filter/match, other than helper name in tb[TCA_CT_HELPER_NAME],
we also need to send the family in tb[TCA_CT_HELPER_FAMILY] and the
proto in tb[TCA_CT_HELPER_PROTO] to kernel.
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
This patch is to make the err path simple by calling tcf_ct_params_free(),
so that it won't cause problems when more members are added into param and
need freeing on the err path.
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Move ovs_ct_add_helper from openvswitch to nf_conntrack_helper and
rename as nf_ct_add_helper, so that it can be used in TC act_ct in
the next patch.
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Move ovs_ct_helper from openvswitch to nf_conntrack_helper and rename
as nf_ct_helper so that it can be used in TC act_ct in the next patch.
Note that it also adds the checks for the family and proto, as in TC
act_ct, the packets with correct family and proto are not guaranteed.
Acked-by: Marcelo Ricardo Leitner <marcelo.leitner@gmail.com>
Signed-off-by: Xin Long <lucien.xin@gmail.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Similar to what we do with the hash indirection table [1], when network
flow classification rules are forwarding traffic to channels greater
than the requested number of channels, fail the operation.
Without this, traffic could be directed to channels which no longer
exist (dropped) after changing number of channels.
[1] commit d4ab428627 ("ethtool: correctly ensure {GS}CHANNELS doesn't conflict with GS{RXFH}")
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Gal Pressman <gal@nvidia.com>
Link: https://lore.kernel.org/r/20221106123127.522985-1-gal@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
Currently after the hardware restart triggered from the driver, the
station interface connection remains intact, since a disconnect trigger
is not sent to userspace. This can lead to a problem in targets where
the wifi mac sequence is added by the firmware.
After the target restart, its wifi mac sequence number gets reset to
zero. Hence AP to which our device is connected will receive frames with
a wifi mac sequence number jump to the past, thereby resulting in the
AP dropping all these frames, until the frame arrives with a wifi mac
sequence number which AP was expecting.
To avoid such frame drops, its better to trigger a station disconnect
upon target hardware restart which can be done with API
ieee80211_reconfig_disconnect exposed to mac80211.
The other targets are not affected by this change, since the hardware
params flag is not set.
Reported-by: kernel test robot <lkp@intel.com>
Tested-on: WCN6750 hw1.0 AHB WLAN.MSL.1.0.1-00887-QCAMSLSWPLZ-1
Signed-off-by: Youghandhar Chintala <quic_youghand@quicinc.com>
Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com>
Link: https://lore.kernel.org/r/20221104085403.11025-1-quic_youghand@quicinc.com
To catch missing SIGTRAP we employ a WARN in __perf_event_overflow(),
which fires if pending_sigtrap was already set: returning to user space
without consuming pending_sigtrap, and then having the event fire again
would re-enter the kernel and trigger the WARN.
This, however, seemed to miss the case where some events not associated
with progress in the user space task can fire and the interrupt handler
runs before the IRQ work meant to consume pending_sigtrap (and generate
the SIGTRAP).
syzbot gifted us this stack trace:
| WARNING: CPU: 0 PID: 3607 at kernel/events/core.c:9313 __perf_event_overflow
| Modules linked in:
| CPU: 0 PID: 3607 Comm: syz-executor100 Not tainted 6.1.0-rc2-syzkaller-00073-g88619e77b33d #0
| Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/11/2022
| RIP: 0010:__perf_event_overflow+0x498/0x540 kernel/events/core.c:9313
| <...>
| Call Trace:
| <TASK>
| perf_swevent_hrtimer+0x34f/0x3c0 kernel/events/core.c:10729
| __run_hrtimer kernel/time/hrtimer.c:1685 [inline]
| __hrtimer_run_queues+0x1c6/0xfb0 kernel/time/hrtimer.c:1749
| hrtimer_interrupt+0x31c/0x790 kernel/time/hrtimer.c:1811
| local_apic_timer_interrupt arch/x86/kernel/apic/apic.c:1096 [inline]
| __sysvec_apic_timer_interrupt+0x17c/0x640 arch/x86/kernel/apic/apic.c:1113
| sysvec_apic_timer_interrupt+0x40/0xc0 arch/x86/kernel/apic/apic.c:1107
| asm_sysvec_apic_timer_interrupt+0x16/0x20 arch/x86/include/asm/idtentry.h:649
| <...>
| </TASK>
In this case, syzbot produced a program with event type
PERF_TYPE_SOFTWARE and config PERF_COUNT_SW_CPU_CLOCK. The hrtimer
manages to fire again before the IRQ work got a chance to run, all while
never having returned to user space.
Improve the WARN to check for real progress in user space: approximate
this by storing a 32-bit hash of the current IP into pending_sigtrap,
and if an event fires while pending_sigtrap still matches the previous
IP, we assume no progress (false negatives are possible given we could
return to user space and trigger again on the same IP).
Fixes: ca6c21327c ("perf: Fix missing SIGTRAPs")
Reported-by: syzbot+b8ded3e2e2c6adde4990@syzkaller.appspotmail.com
Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20221031093513.3032814-1-elver@google.com
The previous attempt to augment carrier_down (see Link)
was not met with much enthusiasm so let's do the simple
thing of exposing what some devices already maintain.
Add a common ethtool statistic for link going down.
Currently users have to maintain per-driver mapping
to extract the right stat from the vendor-specific ethtool -S
stats. carrier_down does not fit the bill because it counts
a lot of software related false positives.
Add the statistic to the extended link state API to steer
vendors towards implementing all of it.
Implement for bnxt and all Linux-controlled PHYs. mlx5 and (possibly)
enic also have a counter for this but I leave the implementation
to their maintainers.
Link: https://lore.kernel.org/r/20220520004500.2250674-1-kuba@kernel.org
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Reviewed-by: Michael Chan <michael.chan@broadcom.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20221104190125.684910-1-kuba@kernel.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>
When using wpa_supplicant v2.10, this driver is no longer able to
associate with any AP and fails in the EAPOL 4-way handshake while
sending the 2/4 message to the AP. The problem is not present in
wpa_supplicant v2.9 or older. The problem stems from HostAP commit
144314eaa ("wpa_supplicant: Send EAPOL frames over nl80211 where available")
which changes the way EAPOL frames are sent, from them being send
at L2 frames to them being sent via nl80211 control port.
An EAPOL frame sent as L2 frame is passed to the WiFi driver with
skb->protocol ETH_P_PAE, while EAPOL frame sent via nl80211 control
port has skb->protocol set to ETH_P_802_3 . The later happens in
ieee80211_tx_control_port(), where the EAPOL frame is encapsulated
into 802.3 frame.
The rsi_91x driver handles ETH_P_PAE EAPOL frames as high-priority
frames and sends them via highest-priority transmit queue, while
the ETH_P_802_3 frames are sent as regular frames. The EAPOL 4-way
handshake frames must be sent as highest-priority, otherwise the
4-way handshake times out.
Therefore, to fix this problem, inspect the skb control flags and
if flag IEEE80211_TX_CTRL_PORT_CTRL_PROTO is set, assume this is
an EAPOL frame and transmit the frame via high-priority queue just
like other ETH_P_PAE frames.
Fixes: 0eb42586cf ("rsi: data packet descriptor enhancements")
Signed-off-by: Marek Vasut <marex@denx.de>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20221104163339.227432-1-marex@denx.de
To detect TX or RX stuck, we implement SER (system error recovery) in
firmware to recover abnormal states of hardware, and report events to
driver. This kind of events could happen rarely per day.
SER might be true-positive or false-negative cases, and it could be failed
to recover true-positive case. We dump related registers to kernel message
at that moment and collect them from users, because they occur rarely,
randomly and hard to make sure we reproduce the same symptom. To address
problems accurately, add more registers by this patch.
It also might be false-positive cases that looks like TX or RX get stuck,
we need to dump registers from debugfs manually, so also add similar
things to debugfs as well.
Signed-off-by: Chia-Yuan Li <leo.li@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20221102014300.14091-3-pkshih@realtek.com
Dispatch is a component to decide packets forward to host, DMAC or
HAXIDMA. It contains CDT standing for CPU dispatcher, HDT standing
for host dispatcher, WDE standing for descriptor engine and PLE standing
for payload engine. STF is one kind of modes, it can be used if packet
send to hardware and doesn't need release report.
These debug port information can help to clarify the reason if
packets stuck in dispatch.
Signed-off-by: Chia-Yuan Li <leo.li@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20221102014300.14091-2-pkshih@realtek.com
Currently brcmfmac is expecting to be set for both
BRCMF_VIF_STATUS_EAP_SUCCESS and BRCMF_VIF_STATUS_EAP status bit based
on dongle event and those bits are cleared to complete connect request
successfully.
But when connect request is finished unsuccessfully, either
BRCMF_VIF_STATUS_EAP_SUCCESS / BRCMF_VIF_STATUS_EAP bits are not
cleared depending on how the connect fail event happens. These status
bits are carried over to following new connect request and this will lead
to generate below kernel warning for some case. Worst case status
mismatch happens between dongle and wpa_supplicant.
WARNING: ../net/wireless/sme.c:756 __cfg80211_connect_result+0x42c/0x4a0 [cfg80211]
The fix is to clear the BRCMF_VIF_STATUS_EAP_SUCCESS /
BRCMF_VIF_STATUS_EAP bits during the link down process and add to call
link down process when link down event received during
BRCMF_VIF_STATUS_CONNECTING as well as BRCMF_VIF_STATUS_CONNECTED
state.
Signed-off-by: Wataru Gohda <wataru.gohda@cypress.com>
Signed-off-by: Chi-hsien Lin <chi-hsien.lin@infineon.com>
Signed-off-by: Ian Lin <ian.lin@infineon.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>
Link: https://lore.kernel.org/r/20221024085215.27616-3-ian.lin@infineon.com
Although we tried to fix the regression for the recent changes with
the delayed card registration, it doesn't seem covering the all
cases; e.g. on Roland EDIROL M-100FX, where the generic quirk for
Roland devices is applied, it misses the card registration because the
detection of the last interface (apparently for MIDI) fails.
This patch is an attempt to recover from those failures by calling the
card register also at the error path for the secondary interfaces.
The card register condition is also extended to match with the old
check in the previous patch, too (i.e. the simple check of the
interface number) for catching the probe with errors.
Fixes: 39efc9c8a9 ("ALSA: usb-audio: Fix last interface check for registration")
Cc: <stable@vger.kernel.org>
Link: https://bugzilla.suse.com/show_bug.cgi?id=1205111
Link: https://lore.kernel.org/r/20221108065824.14418-1-tiwai@suse.de
Signed-off-by: Takashi Iwai <tiwai@suse.de>
For unmapped range, the returned map.m_llen is zero, and thus the
calculated count is unexpected zero.
Prior to the refactoring introduced by commit 1ae9470c3e ("erofs:
clean up .read_folio() and .readahead() in fscache mode"), only the
readahead routine suffers from this. With the refactoring of making
.read_folio() and .readahead() calling one common routine, both
read_folio and readahead have this issue now.
Fix this by calculating count separately in unmapped condition.
Fixes: c665b394b9 ("erofs: implement fscache-based data readahead")
Fixes: 1ae9470c3e ("erofs: clean up .read_folio() and .readahead() in fscache mode")
Signed-off-by: Jingbo Xu <jefflexu@linux.alibaba.com>
Reviewed-by: Gao Xiang <hsiangkao@linux.alibaba.com>
Reviewed-by: Chao Yu <chao@kernel.org>
Link: https://lore.kernel.org/r/20221104054028.52208-3-jefflexu@linux.alibaba.com
Signed-off-by: Gao Xiang <hsiangkao@linux.alibaba.com>
SAT SCSI/ATA Translation specification requires SCSI SYNCHRONIZE CACHE
(10) and (16) commands both shall be translated to ATA flush command.
Also, ZBC Zoned Block Commands specification mandates SYNCHRONIZE CACHE
(16) command support. However, libata translates only SYNCHRONIZE CACHE
(10). This results in SYNCHRONIZE CACHE (16) command failures on SATA
drives and then libata translation does not conform to ZBC. To avoid the
failure, add support for SYNCHRONIZE CACHE (16).
Signed-off-by: Shin'ichiro Kawasaki <shinichiro.kawasaki@wdc.com>
Cc: stable@vger.kernel.org
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Martin K. Petersen <martin.petersen@oracle.com>
Signed-off-by: Damien Le Moal <damien.lemoal@opensource.wdc.com>
When disabling dma channel, a TCF flag is set and as TCIE is enabled, an
interrupt is raised.
On a busy system, the interrupt may have latency and the user can ask for
dmaengine_resume while stm32-dma driver has not yet managed the complete
pause (backup of registers to restore state in resume).
To avoid such a case, instead of waiting the interrupt to backup the
registers, do it just after disabling the channel and discard Transfer
Complete interrupt in case the channel is paused.
Fixes: 099a9a94be ("dmaengine: stm32-dma: add device_pause/device_resume support")
Signed-off-by: Amelie Delaunay <amelie.delaunay@foss.st.com>
Link: https://lore.kernel.org/r/20221024083611.132588-1-amelie.delaunay@foss.st.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
>From Intel IAA spec [1], Intel IAA does not support batch processing.
Two batch related default values for IAA are incorrect in current code:
(1) The max batch size of device is set during device initialization,
that indicates batch is supported. It should be always 0 on IAA.
(2) The max batch size of work queue is set to WQ_DEFAULT_MAX_BATCH (32)
as the default value regardless of Intel DSA or IAA device during
work queue setup and cleanup. It should be always 0 on IAA.
Fix the issues by setting the max batch size of device and max batch
size of work queue to 0 on IAA device, that means batch is not
supported.
[1]: https://cdrdv2.intel.com/v1/dl/getContent/721858
Fixes: 23084545db ("dmaengine: idxd: set max_xfer and max_batch for RO device")
Fixes: 92452a72eb ("dmaengine: idxd: set defaults for wq configs")
Fixes: bfe1d56091 ("dmaengine: idxd: Init and probe for Intel data accelerators")
Signed-off-by: Xiaochen Shen <xiaochen.shen@intel.com>
Reviewed-by: Dave Jiang <dave.jiang@intel.com>
Reviewed-by: Fenghua Yu <fenghua.yu@intel.com>
Link: https://lore.kernel.org/r/20220930201528.18621-2-xiaochen.shen@intel.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
The first IRQ is required, but IRQs 1 through (nb_phy_chans - 1) are
optional, because on some platforms (e.g. PXA168) there is a single IRQ
shared between all channels.
This change inhibits a flood of "IRQ index # not found" messages at
startup. Tested on a PXA168-based device.
Fixes: 7723f4c5ec ("driver core: platform: Add an error message to platform_get_irq*()")
Signed-off-by: Doug Brown <doug@schmorgal.com>
Link: https://lore.kernel.org/r/20220906000709.52705-1-doug@schmorgal.com
Signed-off-by: Vinod Koul <vkoul@kernel.org>
