Bryan O'Donoghue [Sat, 15 Jan 2022 00:16:46 +0000 (00:16 +0000)]
wcn36xx: Add SNR reporting via get_survey()
Add support for get_survey() reporting. Current channel and noise-floor are
reported, other parameters such as scan, busy, TX and RX time are not
immediately available.
Noise is a useful metric to report, so bring it out now.
Bryan O'Donoghue [Sat, 15 Jan 2022 00:16:45 +0000 (00:16 +0000)]
wcn36xx: Track SNR and RSSI for each RX frame
The BDs for each RX frame contain both the RSSI and SNR for the received
frame. If we track and store this information it can be useful to us in
get_survey() and potentially elsewhere.
Bryan O'Donoghue [Sat, 15 Jan 2022 00:16:44 +0000 (00:16 +0000)]
wcn36xx: Track the band and channel we are tuned to
Track the band and channel we are currently tuned to by way of pointers to
the standard structures that describe them both embedded within the driver.
Tracking of the pair makes it much easier when implementing
ieee80211_ops->get_survey to return quickly captured metrics for the
currently tuned channel.
Bryan O'Donoghue [Sat, 15 Jan 2022 00:16:43 +0000 (00:16 +0000)]
wcn36xx: Implement get_snr()
The wcn36xx BD phy descriptor returns both Received Signal Strength
Information (RSSI) and Signal To Noise Ratio (SNR) with each delivered BD.
The macro to extract this data is a simple-one liner, easily imported from
prima driver. This data will be useful to us when implementing
mac80211-ops->get_survey().
Francesco Magliocca [Wed, 12 Jan 2022 08:15:11 +0000 (10:15 +0200)]
ath10k: abstract htt_rx_desc structure
QCA6174 card often hangs with the current htt_rx_desc
memory layout in some circumstances, because its firmware
fails to handle length differences.
Therefore we must abstract the htt_rx_desc structure
and operations on it, to allow different wireless cards
to use different, unrelated rx descriptor structures.
Define a base htt_rx_desc structure and htt_rx_desc_v1
for use with the QCA family of ath10k supported cards
and htt_rx_desc_v2 for use with the WCN3990 card.
Define htt_rx_desc_ops which contains the abstract operations
to access the generic htt_rx_desc, give implementations
for each card and update htt_rx.c to use the defined
abstract interface to rx descriptors.
Lad Prabhakar [Fri, 24 Dec 2021 19:26:24 +0000 (19:26 +0000)]
wcn36xx: Use platform_get_irq_byname() to get the interrupt
platform_get_resource_byname(pdev, IORESOURCE_IRQ, ..) relies on static
allocation of IRQ resources in DT core code, this causes an issue
when using hierarchical interrupt domains using "interrupts" property
in the node as this bypasses the hierarchical setup and messes up the
irq chaining.
In preparation for removal of static setup of IRQ resource from DT core
code use platform_get_irq_byname().
Baochen Qiang [Wed, 12 Jan 2022 02:54:00 +0000 (10:54 +0800)]
ath11k: Reconfigure hardware rate for WCN6855 after vdev is started
There is an issue that WCN6855 tries to connect to an AP using
a hardware rate of 1Mb/s , even though the AP has announced
expected rates as [24, 36, 48, 54] in Probe Response frame.
The reason is that WCN6855 firmware clears hardware rate info
of management frames when vdev starts and uses 1Mb/s as default.
To solve it, reconfigure the rate after vdev is started.
Christophe JAILLET [Mon, 27 Dec 2021 16:58:10 +0000 (17:58 +0100)]
ath: dfs_pattern_detector: Avoid open coded arithmetic in memory allocation
kmalloc_array()/kcalloc() should be used to avoid potential overflow when
a multiplication is needed to compute the size of the requested memory.
kmalloc_array() can be used here instead of kcalloc() because the array is
fully initialized in the next 'for' loop.
Finally, 'cd->detectors' is defined as 'struct pri_detector **detectors;'.
So 'cd->detectors' and '*cd->detectors' are both some pointer.
So use a more logical 'sizeof(*cd->detectors)'.
Lad Prabhakar [Fri, 24 Dec 2021 19:26:25 +0000 (19:26 +0000)]
ath10k: Use platform_get_irq() to get the interrupt
platform_get_resource(pdev, IORESOURCE_IRQ, ..) relies on static
allocation of IRQ resources in DT core code, this causes an issue
when using hierarchical interrupt domains using "interrupts" property
in the node as this bypasses the hierarchical setup and messes up the
irq chaining.
In preparation for removal of static setup of IRQ resource from DT core
code use platform_get_irq().
Peter Seiderer [Wed, 12 Jan 2022 08:15:11 +0000 (10:15 +0200)]
ath5k: fix ah_txq_isr_txok_all setting
The struct ath5k_hw member ah_txq_isr_txok_all is never reset/assigned
outside of ath5k_hw_get_isr() and with the used bitwise-or in the
interrupt handling accumulates all ever set interrupt flags.
Fix this by clearing ah_txq_isr_txok_all before assigning.
Dan Carpenter [Wed, 12 Jan 2022 08:15:11 +0000 (10:15 +0200)]
ath11k: fix error code in ath11k_qmi_assign_target_mem_chunk()
The "ret" vairable is not set at this point. It could be uninitialized
or zero. The correct thing to return is -ENODEV.
Fixes: 6ac04bdc5edb ("ath11k: Use reserved host DDR addresses from DT for PCI devices") Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com> Link: https://lore.kernel.org/r/20220111071445.GA11243@kili
Miles Hu [Wed, 12 Jan 2022 08:15:10 +0000 (10:15 +0200)]
ath11k: enable RX PPDU stats in monitor co-exist mode
RX PPDU statistics collection is missing when monitor mode co-exists
with other modes. This commit combines the processing of the destination
ring with the status ring to fix the issue.
Tested-on: IPQ8074 hw2.0 AHB WLAN.HK.2.5.0.1-01179-QCAHKSWPL_SILICONZ-1 Signed-off-by: Miles Hu <milehu@codeaurora.org> Signed-off-by: Aloka Dixit <quic_alokad@quicinc.com> Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com> Link: https://lore.kernel.org/r/20220111032224.14093-1-quic_alokad@quicinc.com
Zekun Shen [Mon, 27 Dec 2021 03:12:13 +0000 (22:12 -0500)]
ath5k: fix OOB in ath5k_eeprom_read_pcal_info_5111
The bug was found during fuzzing. Stacktrace locates it in
ath5k_eeprom_convert_pcal_info_5111.
When none of the curve is selected in the loop, idx can go
up to AR5K_EEPROM_N_PD_CURVES. The line makes pd out of bound.
pd = &chinfo[pier].pd_curves[idx];
There are many OOB writes using pd later in the code. So I
added a sanity check for idx. Checks for other loops involving
AR5K_EEPROM_N_PD_CURVES are not needed as the loop index is not
used outside the loops.
The patch is NOT tested with real device.
The following is the fuzzing report
BUG: KASAN: slab-out-of-bounds in ath5k_eeprom_read_pcal_info_5111+0x126a/0x1390 [ath5k]
Write of size 1 at addr ffff8880174a4d60 by task modprobe/214
Wen Gong [Tue, 11 Jan 2022 14:42:52 +0000 (16:42 +0200)]
ath11k: free peer for station when disconnect from AP for QCA6390/WCN6855
Commit b4a0f54156ac ("ath11k: move peer delete after vdev stop of station
for QCA6390 and WCN6855") is to fix firmware crash by changing the WMI
command sequence, but actually skip all the peer delete operation, then
it lead commit 58595c9874c6 ("ath11k: Fixing dangling pointer issue upon
peer delete failure") not take effect, and then happened a use-after-free
warning from KASAN. because the peer->sta is not set to NULL and then used
later.
Change to only skip the WMI_PEER_DELETE_CMDID for QCA6390/WCN6855.
log of user-after-free:
[ 534.888665] BUG: KASAN: use-after-free in ath11k_dp_rx_update_peer_stats+0x912/0xc10 [ath11k]
[ 534.888696] Read of size 8 at addr ffff8881396bb1b8 by task rtcwake/2860
P Praneesh [Tue, 11 Jan 2022 14:42:52 +0000 (16:42 +0200)]
ath11k: add LDPC FEC type in 802.11 radiotap header
LDPC is one the FEC type advertised in msdu_start info2 for HT packet
type. Hence, add hardware specific callback for fetching LDPC
support from msdu start and enable RX_ENC_FLAG_LDPC flag while passing
rx status to mac80211.
Karthikeyan Periyasamy [Mon, 10 Jan 2022 14:24:14 +0000 (16:24 +0200)]
ath11k: Refactor the fallback routine when peer create fails
When there is an error in peer create process from
ath11k_peer_find(), the code attempts to handle a fallback
for peer create. When this fallback fails, the driver returns
the fallback return code rather than actual error code
(-ENOENT). So refactor the fallback routine to return
the actual error code.
ath11k: fix workqueue not getting destroyed after rmmod
Currently, ath11k_core_alloc() creates a single thread workqueue.
This workqueue is not detroyed during clean up when ath11k modules
are unloaded from the kernel and is left as it is.
If workqueue is not destroyed, it could lead to kernel
memory scarcity in a longer run. This could affect self and
other drivers workability as well.
Add destroy workqueue in ath11k_core_free().
Tested on: IPQ8074 WLAN.HK.2.4.0.1-01746-QCAHKSWPL_SILICONZ-1
Wen Gong [Mon, 10 Jan 2022 14:24:13 +0000 (16:24 +0200)]
ath10k: fix memory overwrite of the WoWLAN wakeup packet pattern
In function ath10k_wow_convert_8023_to_80211(), it will do memcpy for
the new->pattern, and currently the new->pattern and new->mask is same
with the old, then the memcpy of new->pattern will also overwrite the
old->pattern, because the header format of new->pattern is 802.11,
its length is larger than the old->pattern which is 802.3. Then the
operation of "Copy frame body" will copy a mistake value because the
body memory has been overwrite when memcpy the new->pattern.
Assign another empty value to new_pattern to avoid the overwrite issue.
Yang Yingliang [Mon, 10 Jan 2022 14:24:13 +0000 (16:24 +0200)]
ath11k: add missing of_node_put() to avoid leak
The node pointer is returned by of_find_node_by_type()
or of_parse_phandle() with refcount incremented. Calling
of_node_put() to aovid the refcount leak.
Fixes: 6ac04bdc5edb ("ath11k: Use reserved host DDR addresses from DT for PCI devices") Reported-by: Hulk Robot <hulkci@huawei.com> Signed-off-by: Yang Yingliang <yangyingliang@huawei.com> Signed-off-by: Kalle Valo <quic_kvalo@quicinc.com> Link: https://lore.kernel.org/r/20211221114003.335557-1-yangyingliang@huawei.com
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
So, if dma_set_mask_and_coherent() succeeds, 'pci_using_dac' is known to be
1.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
So, if dma_set_mask_and_coherent() succeeds, 'pci_using_dac' is known to be
1.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
So, if dma_set_mask_and_coherent() succeeds, 'highdma' is known to be true.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
So, if dma_set_mask_and_coherent() succeeds, 'pci_using_dac' is known to be
1.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
Moreover, dma_set_mask_and_coherent() returns 0 or -EIO, so the return
code of the function can be used directly.
Finally, inline bnx2x_set_coherency_mask() because it is now only a wrapper
for a single dma_set_mask_and_coherent() call.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
Moreover, dma_set_mask_and_coherent() returns 0 or -EIO, so the return
code of the function can be used directly. There is no need to 'rc = -EIO'
explicitly.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
So if dma_set_mask_and_coherent() succeeds, 'netdev->features' will have
NETIF_F_HIGHDMA in all cases. Move the assignment of this feature in
be_netdev_init() instead be_probe() which is a much logical place.
Simplify code and remove some dead code accordingly.
As stated in [1], dma_set_mask() with a 64-bit mask never fails if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
So if dma_set_mask_and_coherent() succeeds, 'dma64' is know to be 'true'.
Simplify code and remove some dead code accordingly.
Christophe JAILLET [Sat, 8 Jan 2022 16:16:16 +0000 (17:16 +0100)]
bna: Simplify DMA setting
As stated in [1], dma_set_mask() with a 64-bit mask will never fail if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
So, if dma_set_mask_and_coherent() succeeds, 'using_dac' is known to be
'true'. This variable can be removed.
Simplify code and remove some dead code accordingly.
Christophe JAILLET [Sat, 8 Jan 2022 14:26:06 +0000 (15:26 +0100)]
net: alteon: Simplify DMA setting
As stated in [1], dma_set_mask() with a 64-bit mask will never fail if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
If dma_set_mask_and_coherent() succeeds, 'ap->pci_using_dac' is known to be
1. So 'pci_using_dac' can be removed from the 'struct ace_private'.
Simplify code and remove some dead code accordingly.
Christophe JAILLET [Sat, 8 Jan 2022 14:22:13 +0000 (15:22 +0100)]
myri10ge: Simplify DMA setting
As stated in [1], dma_set_mask() with a 64-bit mask will never fail if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
If dma_set_mask_and_coherent() succeeds, 'dac_enabled' is known to be 1.
Simplify code and remove some dead code accordingly.
Christophe JAILLET [Sat, 8 Jan 2022 13:48:59 +0000 (14:48 +0100)]
qlcnic: Simplify DMA setting
As stated in [1], dma_set_mask() with a 64-bit mask will never fail if
dev->dma_mask is non-NULL.
So, if it fails, the 32 bits case will also fail for the same reason.
So qlcnic_set_dma_mask(), (in qlcnic_main.c) can be simplified a lot and
inlined directly in its only caller.
If dma_set_mask_and_coherent() succeeds, 'pci_using_dac' is known to be 1.
So it can be removed from all the calling chain.
qlcnic_setup_netdev() can finally be simplified as-well.
Yunsheng Lin [Fri, 7 Jan 2022 09:00:42 +0000 (17:00 +0800)]
page_pool: remove spinlock in page_pool_refill_alloc_cache()
As page_pool_refill_alloc_cache() is only called by
__page_pool_get_cached(), which assumes non-concurrent access
as suggested by the comment in __page_pool_get_cached(), and
ptr_ring allows concurrent access between consumer and producer,
so remove the spinlock in page_pool_refill_alloc_cache().
Taehee Yoo [Sun, 9 Jan 2022 16:37:02 +0000 (16:37 +0000)]
amt: fix wrong return type of amt_send_membership_update()
amt_send_membership_update() would return -1 but it's return type is bool.
So, it should be used TRUE instead of -1.
Fixes: cbc21dc1cfe9 ("amt: add data plane of amt interface") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Taehee Yoo <ap420073@gmail.com> Link: https://lore.kernel.org/r/20220109163702.6331-1-ap420073@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Pavel Skripkin [Thu, 6 Jan 2022 22:57:16 +0000 (01:57 +0300)]
net: mcs7830: handle usb read errors properly
Syzbot reported uninit value in mcs7830_bind(). The problem was in
missing validation check for bytes read via usbnet_read_cmd().
usbnet_read_cmd() internally calls usb_control_msg(), that returns
number of bytes read. Code should validate that requested number of bytes
was actually read.
So, this patch adds missing size validation check inside
mcs7830_get_reg() to prevent uninit value bugs
Reported-and-tested-by: syzbot+003c0a286b9af5412510@syzkaller.appspotmail.com Fixes: 2a36d7083438 ("USB: driver for mcs7830 (aka DeLOCK) USB ethernet adapter") Signed-off-by: Pavel Skripkin <paskripkin@gmail.com> Reviewed-by: Arnd Bergmann <arnd@arndb.de> Link: https://lore.kernel.org/r/20220106225716.7425-1-paskripkin@gmail.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
In this series patch, the interface kfree_skb_with_reason() is
introduced(), which is used to collect skb drop reason, and pass
it to 'kfree_skb' tracepoint. Therefor, 'drop_monitor' or eBPF is
able to monitor abnormal skb with detail reason.
In fact, this series patches are out of the intelligence of David
and Steve, I'm just a truck man :/
In the first patch, kfree_skb_with_reason() is introduced and
the 'reason' field is added to 'kfree_skb' tracepoint. In the
second patch, 'kfree_skb()' in replaced with 'kfree_skb_with_reason()'
in tcp_v4_rcv(). In the third patch, 'kfree_skb_with_reason()' is
used in __udp4_lib_rcv().
Changes since v3:
- fix some code style problems in skb.h
Changes since v2:
- rename kfree_skb_with_reason() to kfree_skb_reason()
- make kfree_skb() static inline, as Jakub suggested
Changes since v1:
- rename some drop reason, as David suggested
- add the third patch
====================
Jakub Kicinski [Sun, 9 Jan 2022 21:33:21 +0000 (13:33 -0800)]
net/mlx5e: Fix build error in fec_set_block_stats()
Build bot reports:
drivers/net/ethernet/mellanox/mlx5/core/en_stats.c: In function 'fec_set_block_stats':
drivers/net/ethernet/mellanox/mlx5/core/en_stats.c:1235:48: error: 'outl' undeclared (first use in this function); did you mean 'out'?
1235 | if (mlx5_core_access_reg(mdev, in, sz, outl, sz, MLX5_REG_PPCNT, 0, 0))
| ^~~~
| out
Jakub Kicinski [Mon, 10 Jan 2022 00:27:26 +0000 (16:27 -0800)]
Merge branch 'bnxt_en-update-for-net-next'
Michael Chan says:
====================
bnxt_en: Update for net-next
This series adds better error and debug logging for firmware messages.
We now also use the firmware provided timeout value for long running
commands instead of capping it to 40 seconds.
====================
Edwin Peer [Sun, 9 Jan 2022 23:54:45 +0000 (18:54 -0500)]
bnxt_en: improve firmware timeout messaging
While it has always been possible to infer that an HWRM command was
abandoned due to an unhealthy firmware status by the shortened timeout
reported, this change improves the log messaging to account for this
case explicitly. In the interests of further clarity, the firmware
status is now also reported in these new messages.
v2: Remove inline keyword for hwrm_wait_must_abort() in .c file.
Reviewed-by: Andy Gospodarek <gospo@broadcom.com> Signed-off-by: Edwin Peer <edwin.peer@broadcom.com> Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Edwin Peer [Sun, 9 Jan 2022 23:54:44 +0000 (18:54 -0500)]
bnxt_en: use firmware provided max timeout for messages
Some older devices cannot accommodate the 40 seconds timeout
cap for long running commands (such as NVRAM commands) due to
hardware limitations. Allow these devices to request more time for
these long running commands, but print a warning, since the longer
timeout may cause the hung task watchdog to trigger. In the case of a
firmware update operation, this is preferable to failing outright.
v2: Use bp->hwrm_cmd_max_timeout directly without the constants.
Fixes: 881d8353b05e ("bnxt_en: Add an upper bound for all firmware command timeouts.") Signed-off-by: Edwin Peer <edwin.peer@broadcom.com> Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Edwin Peer [Sun, 9 Jan 2022 23:54:43 +0000 (18:54 -0500)]
bnxt_en: improve VF error messages when PF is unavailable
The current driver design relies on the PF netdev being open in order
to intercept the following HWRM commands from a VF:
- HWRM_FUNC_VF_CFG
- HWRM_CFA_L2_FILTER_ALLOC
- HWRM_PORT_PHY_QCFG (only if FW_CAP_LINK_ADMIN is not supported)
If the PF is closed, then VFs are subjected to rather inscrutable error
messages in response to any configuration requests involving the above
command types. Recent firmware distinguishes this problem case from
other errors by returning HWRM_ERR_CODE_PF_UNAVAILABLE. In most cases,
the appropriate course of action is still to fail, but this can now be
accomplished with the aid of more user informative log messages. For L2
filter allocations that are already asynchronous, an automatic retry
seems more appropriate.
v2: Delete extra newline.
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com> Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Edwin Peer [Sun, 9 Jan 2022 23:54:42 +0000 (18:54 -0500)]
bnxt_en: add dynamic debug support for HWRM messages
Add logging of firmware messages. These can be useful for diagnosing
issues in the field, but due to their verbosity are only appropriate
at a debug message level.
Signed-off-by: Edwin Peer <edwin.peer@broadcom.com> Signed-off-by: Michael Chan <michael.chan@broadcom.com> Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Paul Blakey [Thu, 6 Jan 2022 15:38:04 +0000 (17:38 +0200)]
net: openvswitch: Fix ct_state nat flags for conns arriving from tc
Netfilter conntrack maintains NAT flags per connection indicating
whether NAT was configured for the connection. Openvswitch maintains
NAT flags on the per packet flow key ct_state field, indicating
whether NAT was actually executed on the packet.
When a packet misses from tc to ovs the conntrack NAT flags are set.
However, NAT was not necessarily executed on the packet because the
connection's state might still be in NEW state. As such, openvswitch
wrongly assumes that NAT was executed and sets an incorrect flow key
NAT flags.
Fix this, by flagging to openvswitch which NAT was actually done in
act_ct via tc_skb_ext and tc_skb_cb to the openvswitch module, so
the packet flow key NAT flags will be correctly set.
Fixes: b57dc7c13ea9 ("net/sched: Introduce action ct") Signed-off-by: Paul Blakey <paulb@nvidia.com> Acked-by: Jamal Hadi Salim <jhs@mojatatu.com> Link: https://lore.kernel.org/r/20220106153804.26451-1-paulb@nvidia.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
====================
Netfilter updates for net-next
The following patchset contains Netfilter updates for net-next. This
includes one patch to update ovs and act_ct to use nf_ct_put() instead
of nf_conntrack_put().
1) Add netns_tracker to nfnetlink_log and masquerade, from Eric Dumazet.
2) Remove redundant rcu read-size lock in nf_tables packet path.
3) Replace BUG() by WARN_ON_ONCE() in nft_payload.
4) Consolidate rule verdict tracing.
5) Replace WARN_ON() by WARN_ON_ONCE() in nf_tables core.
6) Make counter support built-in in nf_tables.
7) Add new field to conntrack object to identify locally generated
traffic, from Florian Westphal.
8) Prevent NAT from shadowing well-known ports, from Florian Westphal.
9) Merge nf_flow_table_{ipv4,ipv6} into nf_flow_table_inet, also from
Florian.
10) Remove redundant pointer in nft_pipapo AVX2 support, from Colin Ian King.
11) Replace opencoded max() in conntrack, from Jiapeng Chong.
12) Update conntrack to use refcount_t API, from Florian Westphal.
13) Move ip_ct_attach indirection into the nf_ct_hook structure.
14) Constify several pointer object in the netfilter codebase,
from Florian Westphal.
15) Tree-wide replacement of nf_conntrack_put() by nf_ct_put(), also
from Florian.
16) Fix egress splat due to incorrect rcu notation, from Florian.
17) Move stateful fields of connlimit, last, quota, numgen and limit
out of the expression data area.
18) Build a blob to represent the ruleset in nf_tables, this is a
requirement of the new register tracking infrastructure.
19) Add NFT_REG32_NUM to define the maximum number of 32-bit registers.
20) Add register tracking infrastructure to skip redundant
store-to-register operations, this includes support for payload,
meta and bitwise expresssions.
* git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf-next: (32 commits)
netfilter: nft_meta: cancel register tracking after meta update
netfilter: nft_payload: cancel register tracking after payload update
netfilter: nft_bitwise: track register operations
netfilter: nft_meta: track register operations
netfilter: nft_payload: track register operations
netfilter: nf_tables: add register tracking infrastructure
netfilter: nf_tables: add NFT_REG32_NUM
netfilter: nf_tables: add rule blob layout
netfilter: nft_limit: move stateful fields out of expression data
netfilter: nft_limit: rename stateful structure
netfilter: nft_numgen: move stateful fields out of expression data
netfilter: nft_quota: move stateful fields out of expression data
netfilter: nft_last: move stateful fields out of expression data
netfilter: nft_connlimit: move stateful fields out of expression data
netfilter: egress: avoid a lockdep splat
net: prefer nf_ct_put instead of nf_conntrack_put
netfilter: conntrack: avoid useless indirection during conntrack destruction
netfilter: make function op structures const
netfilter: core: move ip_ct_attach indirection to struct nf_ct_hook
netfilter: conntrack: convert to refcount_t api
...
====================
Pablo Neira Ayuso [Sun, 9 Jan 2022 16:11:24 +0000 (17:11 +0100)]
netfilter: nft_bitwise: track register operations
Check if the destination register already contains the data that this
bitwise expression performs. This allows to skip this redundant
operation.
If the destination contains a different bitwise operation, cancel the
register tracking information. If the destination contains no bitwise
operation, update the register tracking information.
Update the payload and meta expression to check if this bitwise
operation has been already performed on the register. Hence, both the
payload/meta and the bitwise expressions are reduced.
There is also a special case: If source register != destination register
and source register is not updated by a previous bitwise operation, then
transfer selector from the source register to the destination register.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pablo Neira Ayuso [Sun, 9 Jan 2022 16:11:23 +0000 (17:11 +0100)]
netfilter: nft_meta: track register operations
Check if the destination register already contains the data that this
meta store expression performs. This allows to skip this redundant
operation. If the destination contains a different selector, update
the register tracking information.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Pablo Neira Ayuso [Sun, 9 Jan 2022 16:11:22 +0000 (17:11 +0100)]
netfilter: nft_payload: track register operations
Check if the destination register already contains the data that this
payload store expression performs. This allows to skip this redundant
operation. If the destination contains a different selector, update
the register tracking information.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
This patch adds new infrastructure to skip redundant selector store
operations on the same register to achieve a performance boost from
the packet path.
This is particularly noticeable in pure linear rulesets but it also
helps in rulesets which are already heaving relying in maps to avoid
ruleset linear inspection.
The idea is to keep data of the most recurrent store operations on
register to reuse them with cmp and lookup expressions.
This infrastructure allows for dynamic ruleset updates since the ruleset
blob reduction happens from the kernel.
Userspace still needs to be updated to maximize register utilization to
cooperate to improve register data reuse / reduce number of store on
register operations.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
The new structure nft_rule_dp represents the rule in a more compact way
(smaller memory footprint) compared to the control-plane nft_rule
structure.
The ruleset blob is a read-only data structure. The first field contains
the blob size, then the rules containing expressions. There is a trailing
rule which is used by the tracing infrastructure which is equivalent to
the NULL rule marker in the previous representation. The blob size field
does not include the size of this trailing rule marker.
The ruleset blob is generated from the commit path.
This patch reuses the infrastructure available since 0cbc06b3faba
("netfilter: nf_tables: remove synchronize_rcu in commit phase") to
build the array of rules per chain.
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Florian Westphal [Fri, 7 Jan 2022 04:03:26 +0000 (05:03 +0100)]
net: prefer nf_ct_put instead of nf_conntrack_put
Its the same as nf_conntrack_put(), but without the
need for an indirect call. The downside is a module dependency on
nf_conntrack, but all of these already depend on conntrack anyway.
Cc: Paul Blakey <paulb@mellanox.com> Cc: dev@openvswitch.org Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
There are two _put helpers:
nf_ct_put and nf_conntrack_put. The latter is what should be used in
code that MUST NOT cause a linker dependency on the conntrack module
(e.g. calls from core network stack).
Everyone else should call nf_ct_put() instead.
A followup patch will convert a few nf_conntrack_put() calls to
nf_ct_put(), in particular from modules that already have a conntrack
dependency such as act_ct or even nf_conntrack itself.
Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Florian Westphal [Fri, 7 Jan 2022 04:03:22 +0000 (05:03 +0100)]
netfilter: conntrack: convert to refcount_t api
Convert nf_conn reference counting from atomic_t to refcount_t based api.
refcount_t api provides more runtime sanity checks and will warn on
certain constructs, e.g. refcount_inc() on a zero reference count, which
usually indicates use-after-free.
For this reason template allocation is changed to init the refcount to
1, the subsequenct add operations are removed.
Likewise, init_conntrack() is changed to set the initial refcount to 1
instead refcount_inc().
This is safe because the new entry is not (yet) visible to other cpus.
Signed-off-by: Florian Westphal <fw@strlen.de> Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
Jakub Kicinski [Sun, 9 Jan 2022 22:14:08 +0000 (14:14 -0800)]
Merge tag 'for-net-next-2022-01-07' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next
Luiz Augusto von Dentz says:
====================
bluetooth-next pull request for net-next:
- Add support for Foxconn QCA 0xe0d0
- Fix HCI init sequence on MacBook Air 8,1 and 8,2
- Fix Intel firmware loading on legacy ROM devices
* tag 'for-net-next-2022-01-07' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next:
Bluetooth: hci_sock: fix endian bug in hci_sock_setsockopt()
Bluetooth: L2CAP: uninitialized variables in l2cap_sock_setsockopt()
Bluetooth: btqca: sequential validation
Bluetooth: btusb: Add support for Foxconn QCA 0xe0d0
Bluetooth: btintel: Fix broken LED quirk for legacy ROM devices
Bluetooth: hci_event: Rework hci_inquiry_result_with_rssi_evt
Bluetooth: btbcm: disable read tx power for MacBook Air 8,1 and 8,2
Bluetooth: hci_qca: Fix NULL vs IS_ERR_OR_NULL check in qca_serdev_probe
Bluetooth: hci_bcm: Check for error irq
====================
Jakub Kicinski [Sun, 9 Jan 2022 22:00:49 +0000 (14:00 -0800)]
Merge tag 'linux-can-fixes-for-5.16-20220109' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can
Marc Kleine-Budde says:
====================
pull-request: can 2022-01-09
The first patch is by Johan Hovold and fixes a mem leak in the error
path of the softing_cs driver.
The next patch is by me and fixes a set but not used variable warning
in the softing driver.
Jiasheng Jiang's patch for the xilinx_can driver adds the missing
error checking when getting the IRQ.
Lad Prabhakar contributes a patch for the rcar_canfd driver to fix a
mem leak in the error path.
The last patch is by Brian Silverman and properly initializes the send
USB messages to avoid spurious CAN error frames.
* tag 'linux-can-fixes-for-5.16-20220109' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can:
can: gs_usb: gs_can_start_xmit(): zero-initialize hf->{flags,reserved}
can: rcar_canfd: rcar_canfd_channel_probe(): make sure we free CAN network device
can: xilinx_can: xcan_probe(): check for error irq
can: softing: softing_startstop(): fix set but not used variable warning
can: softing_cs: softingcs_probe(): fix memleak on registration failure
====================
this is a pull request of 22 patches for net-next/master.
The first patch is by Tom Rix and fixes an uninitialized variable in
the janz-ican3 driver (introduced in linux-can-next-for-5.17-20220105).
The next 13 patches are by my and target the mcp251xfd driver. First
several cleanup patches, then the driver is prepared for the upcoming
ethtool ring parameter and IRQ coalescing support, which is added in a
later pull request.
The remaining 8 patches are by Dario Binacchi and me and enhance the
flexcan driver. The driver is moved into a sub directory. An ethtool
private flag is added to optionally disable CAN RTR frame reception,
to make use of more RX buffers. The resulting RX buffer configuration
can be read by ethtool ring parameter support. Finally documentation
for the ethtool private flag is added to the
Documentation/networking/device_drivers/can directory.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
No information is deliberately sent in hf->flags in host -> device
communications, but the open-source candleLight firmware echoes it
back, which can result in the GS_CAN_FLAG_OVERFLOW flag being set and
generating spurious ERRORFRAMEs.
While there also initialize the reserved member with 0.
Lad Prabhakar [Thu, 6 Jan 2022 11:48:00 +0000 (11:48 +0000)]
can: rcar_canfd: rcar_canfd_channel_probe(): make sure we free CAN network device
Make sure we free CAN network device in the error path. There are
several jumps to fail label after allocating the CAN network device
successfully. This patch places the free_candev() under fail label so
that in failure path a jump to fail label frees the CAN network
device.
Fixes: 76e9353a80e9 ("can: rcar_canfd: Add support for RZ/G2L family") Link: https://lore.kernel.org/all/20220106114801.20563-1-prabhakar.mahadev-lad.rj@bp.renesas.com Reported-by: Pavel Machek <pavel@denx.de> Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> Reviewed-by: Kieran Bingham <kieran.bingham+renesas@ideasonboard.com> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Jiasheng Jiang [Fri, 24 Dec 2021 02:13:24 +0000 (10:13 +0800)]
can: xilinx_can: xcan_probe(): check for error irq
For the possible failure of the platform_get_irq(), the returned irq
could be error number and will finally cause the failure of the
request_irq().
Consider that platform_get_irq() can now in certain cases return
-EPROBE_DEFER, and the consequences of letting request_irq()
effectively convert that into -EINVAL, even at probe time rather than
later on. So it might be better to check just now.
Marc Kleine-Budde [Sat, 8 Jan 2022 20:57:51 +0000 (21:57 +0100)]
can: softing: softing_startstop(): fix set but not used variable warning
In the function softing_startstop() the variable error_reporting is
assigned but not used. The code that uses this variable is commented
out. Its stated that the functionality is not finally verified.
To fix the warning:
| drivers/net/can/softing/softing_fw.c:424:9: error: variable 'error_reporting' set but not used [-Werror,-Wunused-but-set-variable]
remove the comment, activate the code, but add a "0 &&" to the if
expression and rely on the optimizer rather than the preprocessor to
remove the code.
Johan Hovold [Wed, 22 Dec 2021 10:48:43 +0000 (11:48 +0100)]
can: softing_cs: softingcs_probe(): fix memleak on registration failure
In case device registration fails during probe, the driver state and
the embedded platform device structure needs to be freed using
platform_device_put() to properly free all resources (e.g. the device
name).
Fixes: 0a0b7a5f7a04 ("can: add driver for Softing card") Link: https://lore.kernel.org/all/20211222104843.6105-1-johan@kernel.org Cc: stable@vger.kernel.org # 2.6.38 Signed-off-by: Johan Hovold <johan@kernel.org> Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org> Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>
Dario Binacchi [Sat, 8 Jan 2022 18:16:33 +0000 (19:16 +0100)]
can: flexcan: add ethtool support to get rx/tx ring parameters
This patch adds ethtool support to get the number of message buffers
configured for reception/transmission, which may also depends on
runtime configurations such as the 'rx-rtr' flag state.
Marc Kleine-Budde [Tue, 4 Jan 2022 15:28:00 +0000 (16:28 +0100)]
can: flexcan: add ethtool support to change rx-rtr setting during runtime
This patch adds a private flag to the flexcan driver to switch the
"rx-rtr" setting on and off.
"rx-rtr" on - Receive RTR frames. (default)
The CAN controller can and will receive RTR frames.
On some IP cores the controller cannot receive RTR
frames in the more performant "RX mailbox" mode and
will use "RX FIFO" mode instead.
"rx-rtr" off - Waive ability to receive RTR frames. (not supported on all IP cores)
This mode activates the "RX mailbox mode" for better
performance, on some IP cores RTR frames cannot be
received anymore.
The "RX FIFO" mode uses a FIFO with a depth of 6 CAN frames.
The "RX mailbox" mode uses up to 62 mailboxes.
Marc Kleine-Budde [Fri, 7 Jan 2022 13:29:15 +0000 (14:29 +0100)]
can: flexcan: add more quirks to describe RX path capabilities
Most flexcan IP cores support 2 RX modes:
- FIFO
- mailbox
Some IP core versions cannot receive CAN RTR messages via mailboxes.
This patch adds quirks to document this.
This information will be used in a later patch to switch from FIFO to
more performant mailbox mode at the expense of losing the ability to
receive RTR messages. This trade off is beneficial in certain use
cases.
Marc Kleine-Budde [Thu, 6 Jan 2022 11:45:46 +0000 (12:45 +0100)]
can: flexcan: rename RX modes
Most flexcan IP cores support 2 RX modes:
- FIFO
- mailbox
The names for these modes were chosen to reflect the name of the
rx-offload mode they are using.
The name of the RX modes should better reflect their difference with
regards the flexcan IP core. So this patch renames the various
occurrences of OFF_FIFO to RX_FIFO and OFF_TIMESTAMP to RX_MAILBOX:
projects / users / dwmw2 / linux.git / log