Some Bluetooth device will make up connection as PAN link, though the
connection is idle, it will still report the PAN link is active. The
coexistence mechanism will enable TDMA to protect the PAN, it makes
Wi-Fi throughput degrade at least 50%. But the link is idle, don't
need so much bandwidth. Add TDMA case to let Wi-Fi can do traffic 80%
bandwidth.
Ching-Te Ku [Thu, 3 Oct 2024 10:51:38 +0000 (18:51 +0800)]
wifi: rtw89: coex: Reorder Bluetooth info related logic
Reorder Bluetooth firmware related event index, it should be the same
with Wi-Fi firmware definition. To fix coexistence can not recognize
Bluetooth PAN(Personal area network) profile correctly, modified the
related logic.
Ching-Te Ku [Thu, 3 Oct 2024 10:51:37 +0000 (18:51 +0800)]
wifi: rtw89: coex: Update priority setting for Wi-Fi is scanning
Update coexistence priority setting for Wi-Fi scanning channel, the new
setting will allow Wi-Fi do RX while Bluetooth audio is not busy. Forced
to set new TDMA policy while RF calibration request come, to make sure
the calibration can do well, and switch to normal setting while the
calibration is done. Remove the code that no longer use.
Martin Kaistra [Mon, 30 Sep 2024 08:49:55 +0000 (10:49 +0200)]
wifi: rtl8xxxu: Perform update_beacon_work when beaconing is enabled
In STA+AP concurrent mode, performing a scan operation on one vif
temporarily stops beacons on the other. When the scan is completed,
beacons are enabled again with BSS_CHANGED_BEACON_ENABLED.
We can observe that no beacons are being sent when just
rtl8xxxu_start_tx_beacon() is being called.
Thus, also perform update_beacon_work in order to restore beaconing.
In range test, the RSSI is helpful to check attenuation of cable and align
difference between environments. Since data packets can be transmitted with
different rate and power, the RSSI of all packets can be variant.
Oppositely beacon is transmitted with the same rate and power, so beacon
RSSI will be relatively invariant, and more helpful to diagnose problems.
The output of beacon RSSI in unit of dBm looks like:
Zong-Zhe Yang [Wed, 25 Sep 2024 02:01:19 +0000 (10:01 +0800)]
wifi: rtw89: handle entity active flag per PHY
Originally, we have an active flag to record whether we have set PHY once.
After impending MLO support, there will be dual-PHY and they can be set
individually on Wi-Fi 7 chips. So, we now have active flag per PHY and
handle them individually.
Zong-Zhe Yang [Wed, 25 Sep 2024 02:01:18 +0000 (10:01 +0800)]
wifi: rtw89: initialize dual HW bands for MLO and control them by link
To support MLO, we initialize things on dual HW bands of Wi-Fi 7 chip.
And, each link will indicate which HW band it's bound to. So, in link
control flow, we control major things based on target link's HW band
instead of hardcode like RTW89_PHY_X or RTW89_MAC_X.
wifi: rtw89: 8852ce: fix gray code conversion for filter out EQ
To use manual mode to set value of filter out EQ, read the source value of
filter out EQ coded in gray code, and write to the target register in
binary number. The function to convert from gray code to binary number is
wrong originally, causing unexpected EQ, so fix it.
wifi: rtw89: 8922ae: disable PCI PHY EQ to improve compatibility
For adaption EQ circuit, this HW design and affected by EIEOS (Electrical
Idle Exit Order Set) amplitude from platform and process from IC, so
disable EQ to improve that.
wifi: rtw89: pci: consolidate PCI basic configurations for probe and resume
The PCI settings aren't always persistent after chip suspends, so
reconfigure the settings after chip resumes. Since most of these settings
are the same, consolidate them into a function to avoid missing somewhere.
Bitterblue Smith [Fri, 20 Sep 2024 19:27:30 +0000 (22:27 +0300)]
wifi: rtw88: Parse the RX descriptor with a single function
rtw8703b_query_rx_desc(), rtw8723d_query_rx_desc(),
rtw8821c_query_rx_desc(), rtw8822b_query_rx_desc(), and
rtw8822c_query_rx_desc() are almost identical, so replace them all with
a single function, rtw_rx_query_rx_desc().
Also, access the RX descriptor using a struct with __le32 members and
le32_get_bits().
wifi: rtw89: coex: initialize local .dbcc_2g_phy in _set_btg_ctrl()
For the case of DBCC enabled and fwrole version 0, the local variable
wl_rinfo.dbcc_2g_phy might not be set by following for-loop, leading
uninitialized variable before using.
The relation bewteen mac80211 link_id and our link instance is like below.
|\
(link_id) | \
0 -------- | |
1 -------- | | ------ instance-0 (link_id: X) -> work on HW band 0
2 -------- | |
... | | ------ instance-1 (link_id: Y) -> work on HW band 1
14 -------- | |
| /
|/
N.B. For cases of non-MLD connection, we set our link instance-0
active with link_id 0. So, our code flow can be compatible between
non-MLD connection and MLD connection.
Based on above, we tweak entire driver architecture first. But, we don't
dynamically enable multiple links here. That will be handled separately.
Most of the things changed here are changing flows to iterate all active
links and read bss_conf/link_sta data according to target link. And, for
cases of scan, ROC, WOW, we use instance-0 to deal with the request.
There are some things listed below, which work for now but need to extend
before multiple active links.
1. tx path
select suitable link instance among multiple active links
2. rx path
determine rx link by PPDU instead of always link instance-0
3. CAM
apply MLD pairwise key to any active links dynamically
Besides, PS code cannot easily work along with tweaking architecture. With
supporting MLO flag (currently false), we disable PS first and will fix it
by another commit in the following.
Zong-Zhe Yang [Mon, 16 Sep 2024 05:31:53 +0000 (13:31 +0800)]
wifi: rtw89: rename rtw89_sta to rtw89_sta_link ahead for MLO
This is an intermediate version that is separated from subsequent major
MLO changes, so some functions' namings are not really determined here.
e.g. struct rtw89_sta_link *sta_to_rtwsta_safe(struct ieee80211_sta *sta)
Zong-Zhe Yang [Mon, 16 Sep 2024 05:31:52 +0000 (13:31 +0800)]
wifi: rtw89: rename rtw89_vif to rtw89_vif_link ahead for MLO
This is an intermediate version that is separated from subsequent major
MLO changes, so some functions' namings are not really determined here.
e.g. struct rtw89_vif_link *vif_to_rtwvif_safe(struct ieee80211_vif *vif)
Driver uses skb_queue_purge() to purge TX skb, but not report tx status
causing "Have pending ack frames!" warning. Use ieee80211_purge_tx_queue()
to correct this.
Since ieee80211_purge_tx_queue() doesn't take locks, to prevent racing
between TX work and purge TX queue, flush and destroy TX work in advance.
wifi: rtw89: 8922a: rfk: support firmware command RX DCK v1 format
RX DCK stands for receiver DC calibration. The v1 format adds a field to
indicate the calibration is for operation channel or an average value for
all scanning channels. Update the format accordingly, and increase
firmware format sequence to 2.
Colin Ian King [Thu, 12 Sep 2024 13:53:35 +0000 (14:53 +0100)]
wifi: rtlwifi: make read-only arrays static const
Don't populate the read-only arrays params, toshiba_smid1, toshiba_smid2,
samsung_smid and lenovo_smid on the stack at run time, instead make them
static const.
wifi: wl1251: Use IRQF_NO_AUTOEN flag in request_irq()
As commit cbe16f35bee6 ("genirq: Add IRQF_NO_AUTOEN for request_irq/nmi()")
said, the code below is subobtimal. IRQF_NO_AUTOEN flag can be used by
drivers to request_irq(). It prevents the automatic enabling of the
requested interrupt in the same safe way. With that the usage can be
simplified and corrected.
wifi: mwifiex: Use IRQF_NO_AUTOEN flag in request_irq()
disable_irq() after request_irq() still has a time gap in which
interrupts can come. request_irq() with IRQF_NO_AUTOEN flag will
disable IRQ auto-enable when request IRQ.
Fixes: 853402a00823 ("mwifiex: Enable WoWLAN for both sdio and pcie") Signed-off-by: Jinjie Ruan <ruanjinjie@huawei.com> Acked-by: Brian Norris <briannorris@chromium.org> Signed-off-by: Kalle Valo <kvalo@kernel.org> Link: https://patch.msgid.link/20240910124314.698896-3-ruanjinjie@huawei.com
wifi: p54: Use IRQF_NO_AUTOEN flag in request_irq()
disable_irq() after request_irq() still has a time gap in which
interrupts can come. request_irq() with IRQF_NO_AUTOEN flag will
disable IRQ auto-enable when request IRQ.
wifi: brcmfmac: add flag for random seed during firmware download
Providing the random seed to firmware was tied to the fact that the
device has a valid OTP, which worked for some Apple chips. However,
it turns out the BCM43752 device also needs the random seed in order
to get firmware running. Suspect it is simply tied to the firmware
branch used for the device. Introducing a mechanism to allow setting
it for a device through the device table.
Co-developed-by: Ondrej Jirman <megi@xff.cz> Signed-off-by: Ondrej Jirman <megi@xff.cz> Co-developed-by: Arend van Spriel <arend.vanspriel@broadcom.com> Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com> Signed-off-by: Jacobe Zang <jacobe.zang@wesion.com> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.com> Tested-by: Sebastian Reichel <sebastian.reichel@collabora.com> # On RK3588 EVB1 Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com> Signed-off-by: Kalle Valo <kvalo@kernel.org> Link: https://patch.msgid.link/20240910-wireless-mainline-v14-4-9d80fea5326d@wesion.com
wifi: brcmfmac: Add optional lpo clock enable support
WiFi modules often require 32kHz clock to function. Add support to
enable the clock to PCIe driver and move "brcm,bcm4329-fmac" check
to the top of brcmf_of_probe. Change function prototypes from void
to int and add appropriate errno's for return values that will be
send to bus when error occurred.
Co-developed-by: Ondrej Jirman <megi@xff.cz> Signed-off-by: Ondrej Jirman <megi@xff.cz> Co-developed-by: Arend van Spriel <arend.vanspriel@broadcom.com> Signed-off-by: Arend van Spriel <arend.vanspriel@broadcom.com> Reviewed-by: Sai Krishna <saikrishnag@marvell.com> Signed-off-by: Jacobe Zang <jacobe.zang@wesion.com> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.com> Tested-by: Sebastian Reichel <sebastian.reichel@collabora.com> # On RK3588 EVB1 Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com> Signed-off-by: Kalle Valo <kvalo@kernel.org> Link: https://patch.msgid.link/20240910-wireless-mainline-v14-3-9d80fea5326d@wesion.com
dt-bindings: net: wireless: brcm4329-fmac: add clock description for AP6275P
Not only AP6275P Wi-Fi device but also all Broadcom wireless devices allow
external low power clock input. In DTS the clock as an optional choice in
the absence of an internal clock.
Reviewed-by: Arend van Spriel <arend.vanspriel@broadcom.com> Reviewed-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Jacobe Zang <jacobe.zang@wesion.com> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.com> Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com> Signed-off-by: Kalle Valo <kvalo@kernel.org> Link: https://patch.msgid.link/20240910-wireless-mainline-v14-2-9d80fea5326d@wesion.com
It's the device id used by AP6275P which is the Wi-Fi module
used by Rockchip's RK3588 evaluation board and also used in
some other RK3588 boards.
Acked-by: Arend van Spriel <arend.vanspriel@broadcom.com> Acked-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by: Jacobe Zang <jacobe.zang@wesion.com> Reviewed-by: Sebastian Reichel <sebastian.reichel@collabora.com> Signed-off-by: Kalle Valo <kvalo@kernel.org> Link: https://patch.msgid.link/20240910-wireless-mainline-v14-1-9d80fea5326d@wesion.com
Maxime Chevallier [Fri, 13 Sep 2024 10:05:14 +0000 (12:05 +0200)]
net: ethtool: phy: Don't set the context dev pointer for unfiltered DUMP
The context info allows continuing DUMP requests, shall they fill the
netlink buffer.
In the case of filtered DUMP requests, a reference on the netdev is
grabbed in the .start() callback and release in .done().
Unfiltered DUMP request don't need the dev pointer to be set in the context
info, doing so will trigger an unwanted netdev_put() in .done().
Reported-by: syzbot+e9ed4e4368d450c8f9db@syzkaller.appspotmail.com Closes: https://lore.kernel.org/netdev/000000000000d3bf150621d361a7@google.com/ Fixes: 17194be4c8e1 ("net: ethtool: Introduce a command to list PHYs on an interface") Signed-off-by: Maxime Chevallier <maxime.chevallier@bootlin.com> Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev> Link: https://patch.msgid.link/20240913100515.167341-1-maxime.chevallier@bootlin.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The first patch is by Uwe Kleine-König targets all CAN platform driver
and switches back the remove function to struct
platform_driver::remove().
A patch by Stefan Mätje fixes the help text of the ESD USB driver.
Jake Hamby's patch masks an unneeded interrupt in the m_can driver.
The last 2 patches target the rockchip_canfd driver. Arnd Bergmann's
patch reworks the delay calculation for the timekeeping worker, a
patch by me fixes the decoding of the error code register.
* tag 'linux-can-next-for-6.12-20240911' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next:
can: rockchip_canfd: rkcanfd_handle_error_int_reg_ec(): fix decoding of error code register
can: rockchip_canfd: rkcanfd_timestamp_init(): rework delay calculation
can: m_can: m_can_chip_config(): mask timestamp wraparound IRQ
can: usb: Kconfig: Fix list of devices for esd_usb driver
can: Switch back to struct platform_driver::remove()
====================
Gaosheng Cui [Thu, 12 Sep 2024 01:19:49 +0000 (09:19 +0800)]
MIPS: Remove the obsoleted code for include/linux/mv643xx.h
Most of the drivers which used this header have been deleted, most
of these code is obsoleted, move the only defines that are actually
used into arch/powerpc/platforms/chrp/pegasos_eth.c and delete the
file completely.
Jakub Kicinski [Sat, 14 Sep 2024 04:17:14 +0000 (21:17 -0700)]
Merge branch 'enic-report-per-queue-stats'
Nelson Escobar says:
====================
enic: Report per queue stats
Patch #1: Use a macro instead of static const variables for array sizes. I
didn't want to add more static const variables in the next patch
so clean up the existing ones first.
Patch #2: Collect per queue statistics
Patch #3: Report per queue stats in netdev qstats
Patch #4: Report some per queue stats in ethtool
# NETIF="eno6" tools/testing/selftests/drivers/net/stats.py
KTAP version 1
1..5
ok 1 stats.check_pause # XFAIL pause not supported by the device
ok 2 stats.check_fec # XFAIL FEC not supported by the device
ok 3 stats.pkt_byte_sum
ok 4 stats.qstat_by_ifindex
ok 5 stats.check_down
====================
net: fib_rules: Add DSCP selector support
Currently, the kernel rejects IPv4 FIB rules that try to match on the
upper three DSCP bits:
# ip -4 rule add tos 0x1c table 100
# ip -4 rule add tos 0x3c table 100
Error: Invalid tos.
The reason for that is that historically users of the FIB lookup API
only populated the lower three DSCP bits in the TOS field of the IPv4
flow key ('flowi4_tos'), which fits the TOS definition from the initial
IPv4 specification (RFC 791).
This is not very useful nowadays and instead some users want to be able
to match on the six bits DSCP field, which replaced the TOS and IP
precedence fields over 25 years ago (RFC 2474). In addition, the current
behavior differs between IPv4 and IPv6 which does allow users to match
on the entire DSCP field using the TOS selector.
Recent patchsets made sure that callers of the FIB lookup API now
populate the entire DSCP field in the IPv4 flow key. Therefore, it is
now possible to extend FIB rules to match on DSCP.
This is done by adding a new DSCP attribute which is implemented for
both IPv4 and IPv6 to provide user space programs a consistent behavior
between both address families.
The behavior of the old TOS selector is unchanged and IPv4 FIB rules
using it will only match on the lower three DSCP bits. The kernel will
reject rules that try to use both selectors.
Patch #1 adds the new DSCP attribute but rejects its usage.
Patches #2-#3 implement IPv4 and IPv6 support.
Patch #4 allows user space to use the new attribute.
Test that locally generated traffic from a socket that specifies a DS
Field using the IP_TOS / IPV6_TCLASS socket options is correctly
redirected using a FIB rule that matches on DSCP. Add negative tests to
verify that the rule is not it when it should not. Test with both IPv4
and IPv6 and with both TCP and UDP sockets.
Now that both IPv4 and IPv6 support the new DSCP selector, enable user
space to configure FIB rules that make use of it by changing the policy
of the new DSCP attribute so that it accepts values in the range of [0,
63].
Use NLA_U8 rather than NLA_UINT as the field is of fixed size.
Implement support for the new DSCP selector that allows IPv6 FIB rules
to match on the entire DSCP field. This is done despite the fact that
the above can be achieved using the existing TOS selector, so that user
space program will be able to work with IPv4 and IPv6 rules in the same
way.
Differentiate between both selectors by adding a new bit in the IPv6 FIB
rule structure that is only set when the 'FRA_DSCP' attribute is
specified by user space. Reject rules that use both selectors.
Implement support for the new DSCP selector that allows IPv4 FIB rules
to match on the entire DSCP field, unlike the existing TOS selector that
only matches on the three lower DSCP bits.
Differentiate between both selectors by adding a new bit in the IPv4 FIB
rule structure (in an existing one byte hole) that is only set when the
'FRA_DSCP' attribute is specified by user space. Reject rules that use
both selectors.
The FIB rule TOS selector is implemented differently between IPv4 and
IPv6. In IPv4 it is used to match on the three "Type of Services" bits
specified in RFC 791, while in IPv6 is it is used to match on the six
DSCP bits specified in RFC 2474.
Add a new FIB rule attribute to allow matching on DSCP. The attribute
will be used to implement a 'dscp' selector in ip-rule with a consistent
behavior between IPv4 and IPv6.
For now, set the type of the attribute to 'NLA_REJECT' so that user
space will not be able to configure it. This restriction will be lifted
once both IPv4 and IPv6 support the new attribute.
net: ethtool: Enhance error messages sent to user space
During the firmware flashing process, notifications are sent to user
space to provide progress updates. When an error occurs, an error
message is sent to indicate what went wrong.
In some cases, appropriate error messages are missing.
Add relevant error messages where applicable, allowing user space to better
understand the issues encountered.
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.
Qianqiang Liu [Fri, 13 Sep 2024 01:47:32 +0000 (09:47 +0800)]
net: ag71xx: remove dead code path
The "err" is always zero, so the following branch can never be executed:
if (err) {
ndev->stats.rx_dropped++;
kfree_skb(skb);
}
Therefore, the "if" statement can be removed.
Use "ndev->stats.rx_errors" to count "napi_build_skb()" failure
Jakub Kicinski [Sat, 14 Sep 2024 02:50:25 +0000 (19:50 -0700)]
Merge tag 'for-net-next-2024-09-12' 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:
- btusb: Add MediaTek MT7925-B22M support ID 0x13d3:0x3604
- btusb: Add Realtek RTL8852C support ID 0x0489:0xe122
- btrtl: Add the support for RTL8922A
- btusb: Add 2 USB HW IDs for MT7925 (0xe118/e)
- btnxpuart: Add support for ISO packets
- btusb: Add Mediatek MT7925 support ID 0x13d3:0x3608
- btsdio: Do not bind to non-removable CYW4373
- hci_uart: Add support for Amlogic HCI UART
* tag 'for-net-next-2024-09-12' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth-next: (27 commits)
Bluetooth: btintel_pcie: Allocate memory for driver private data
Bluetooth: btusb: Fix not handling ZPL/short-transfer
Bluetooth: btusb: Add 2 USB HW IDs for MT7925 (0xe118/e)
Bluetooth: btsdio: Do not bind to non-removable CYW4373
Bluetooth: hci_sync: Ignore errors from HCI_OP_REMOTE_NAME_REQ_CANCEL
Bluetooth: CMTP: Mark BT_CMTP as DEPRECATED
Bluetooth: replace deprecated strncpy with strscpy_pad
Bluetooth: hci_core: Fix sending MGMT_EV_CONNECT_FAILED
Bluetooth: btrtl: Set msft ext address filter quirk for RTL8852B
Bluetooth: Use led_set_brightness() in LED trigger activate() callback
Bluetooth: btrtl: Use kvmemdup to simplify the code
Bluetooth: btusb: Add Mediatek MT7925 support ID 0x13d3:0x3608
Bluetooth: btrtl: Add the support for RTL8922A
Bluetooth: hci_ldisc: Use speed set by btattach as oper_speed
Bluetooth: hci_conn: Remove redundant memset after kzalloc
Bluetooth: L2CAP: Remove unused declarations
dt-bindings: bluetooth: bring the HW description closer to reality for wcn6855
Bluetooth: btnxpuart: Add support for ISO packets
Bluetooth: hci_h4: Add support for ISO packets in h4_recv.h
Bluetooth: btusb: Add Realtek RTL8852C support ID 0x0489:0xe122
...
====================
Mina Almasry [Fri, 13 Sep 2024 06:07:45 +0000 (06:07 +0000)]
memory-provider: disable building dmabuf mp on !CONFIG_PAGE_POOL
When CONFIG_TRACEPOINTS=y but CONFIG_PAGE_POOL=n, we end up with this
build failure that is reported by the 0-day bot:
ld: vmlinux.o: in function `mp_dmabuf_devmem_alloc_netmems':
>> (.text+0xc37286): undefined reference to `__tracepoint_page_pool_state_hold'
>> ld: (.text+0xc3729a): undefined reference to `__SCT__tp_func_page_pool_state_hold'
>> ld: vmlinux.o:(__jump_table+0x10c48): undefined reference to `__tracepoint_page_pool_state_hold'
>> ld: vmlinux.o:(.static_call_sites+0xb824): undefined reference to `__SCK__tp_func_page_pool_state_hold'
The root cause is that in this configuration, traces are enabled but the
page_pool specific trace_page_pool_state_hold is not registered.
There is no reason to build the dmabuf memory provider when
CONFIG_PAGE_POOL is not present, as it's really a provider to the
page_pool.
In fact the whole NET_DEVMEM is RX path-only at the moment, so we can
make the entire config dependent on the PAGE_POOL.
Note that this may need to be revisited after/while devmem TX is
added, as devmem TX likely does not need CONFIG_PAGE_POOL. For now this
build fix is sufficient.
Reported-by: kernel test robot <lkp@intel.com> Closes: https://lore.kernel.org/oe-kbuild-all/202409131239.ysHQh4Tv-lkp@intel.com/ Signed-off-by: Mina Almasry <almasrymina@google.com> Reviewed-by: Simon Horman <horms@kernel.org> Tested-by: Simon Horman <horms@kernel.org> # build-tested Link: https://patch.msgid.link/20240913060746.2574191-1-almasrymina@google.com Signed-off-by: Jakub Kicinski <kuba@kernel.org>
David S. Miller [Fri, 13 Sep 2024 09:49:01 +0000 (10:49 +0100)]
Merge branch 'am65-cpsw-rx-mq'
Roger Quadros says:
====================
net: ethernet: ti: am65-cpsw: Add multi queue RX support
am65-cpsw can support up to 8 queues at Rx. So far we have
been using only one queue (i.e. default flow) for all RX traffic.
This series adds multi-queue support. The driver starts with
1 RX queue by default. User can increase the RX queues via ethtool,
e.g. 'ethtool -L ethx rx <N>'
The series also adds regmap and regfield support to some of the
ALE registers. It adds Policer/Classifier registers and fields.
Converting the existing ALE control APIs to regfields can be a separate
exercise.
Some helper functions are added to read/write to the Policer/Classifier
registers and a default Classifier setup function is added that
routes packets based on their PCP/DSCP priority to different RX queues.
Signed-off-by: Roger Quadros <rogerq@kernel.org>
---
Changes in v4:
- Use single macro AM65_CPSW_MAX_QUEUES for both TX and RX queues
to simplify code
- reuse am65_cpsw_get/set_per_queue_coalesce for am65_cpsw_get/set_coalesce.
- return -EINVAL if unsupported tx/rx_coalesce_usecs in
am65_cpsw_set_coalesce.
- reverse Xmas tree declaration order fixes in cpsw_ale
- Link to v3: https://lore.kernel.org/r/20240703-am65-cpsw-multi-rx-v3-0-f11cd860fd72@kernel.org
Changes in v3:
- code style fixes
- squashed patches 5 and 6
- added comment about priority to thread mapping table.
- Added Reviewed-by Simon Horman.
- Link to v2: https://lore.kernel.org/r/20240628-am65-cpsw-multi-rx-v2-0-c399cb77db56@kernel.org
Changes in v2:
- rebase to net/next
- fixed RX stall issue during iperf
- Link to v1: https://lore.kernel.org/r/20240606-am65-cpsw-multi-rx-v1-0-0704b0cb6fdc@kernel.org
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Roger Quadros [Tue, 10 Sep 2024 09:24:02 +0000 (12:24 +0300)]
net: ethernet: ti: cpsw_ale: add policer/classifier helpers and setup defaults
The Policer registers in the ALE register space are just shadow registers
and use an index field in the policer table control register to read/write
to the actual Polier registers.
Add helper functions to Read and Write to Policer registers.
Also add a helper function to set the thread value to classifier/policer
mapping. Any packet that first matches the classifier will be sent to the
thread (flow) that is set in the classifier to thread mapping table.
If not set then it goes to the default flow.
Default behaviour is to have 8 classifiers to map 8 DSCP/PCP
priorities to N receive threads (flows). N depends on number of
RX channels enabled for the port.
As per the standard [1] User prioritie 1 (Background) and 2 (Spare) have
lower priority than the user priority 0 (default). User priority 1 being
of the lowest priority.
[1] IEEE802.1D-2004, IEEE Standard for Local and metropolitan area networks
Table G-2 - Traffic type acronyms
Table G-3 - Defining traffic types
Signed-off-by: Roger Quadros <rogerq@kernel.org> Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Roger Quadros [Tue, 10 Sep 2024 09:23:58 +0000 (12:23 +0300)]
net: ethernet: ti: am65-cpsw: Introduce multi queue Rx
am65-cpsw can support up to 8 queues at Rx.
Use a macro AM65_CPSW_MAX_RX_QUEUES to indicate that.
As there is only one DMA channel for RX traffic, the
8 queues come as 8 flows in that channel.
By default, we will start with 1 flow as defined by the
macro AM65_CPSW_DEFAULT_RX_CHN_FLOWS.
User can change the number of flows by ethtool like so
'ethtool -L ethx rx <N>'
All traffic will still come on flow 0. To get traffic on
different flows the Classifiers will need to be set up.
Signed-off-by: Roger Quadros <rogerq@kernel.org> Reviewed-by: Simon Horman <horms@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Mina Almasry [Fri, 13 Sep 2024 03:28:24 +0000 (03:28 +0000)]
memory-provider: fix compilation issue without SYSFS
When CONFIG_SYSFS is not set, the kernel fails to compile:
net/core/page_pool_user.c:368:45: error: implicit declaration of function 'get_netdev_rx_queue_index' [-Werror=implicit-function-declaration]
368 | if (pool->slow.queue_idx == get_netdev_rx_queue_index(rxq)) {
| ^~~~~~~~~~~~~~~~~~~~~~~~~
When CONFIG_SYSFS is not set, get_netdev_rx_queue_index() is not defined
as well.
Fix by removing the ifdef around get_netdev_rx_queue_index(). It is not
needed anymore after commit e817f85652c1 ("xdp: generic XDP handling of
xdp_rxq_info") removed most of the CONFIG_SYSFS ifdefs.
Jakub Kicinski [Fri, 13 Sep 2024 03:48:33 +0000 (20:48 -0700)]
Merge branch 'mlx5-updates-2024-09-11'
Saeed Mahameed says:
====================
Misc updates to mlx5 driver:
1) Fix HW steering ret value and align with kdoc
2) Flow steering cleanups and add support for no append at software level
3) Support for sync reset using hot reset
4) RX SW counter to cover no-split events in header/data split mode
5) Make affinity of SFs configurable
====================
net/mlx5e: SHAMPO, Add no-split ethtool counters for header/data split
When SHAMPO can't identify the protocol/header of a packet, it will
yield a packet that is not split - all the packet is in the data part.
Count this value in packets and bytes.
net/mlx5: Skip HotPlug check on sync reset using hot reset
Sync reset request is nacked by the driver when PCIe bridge connected to
mlx5 device has HotPlug interrupt enabled. However, when using reset
method of hot reset this check can be skipped as Hotplug is supported on
this reset method.
Signed-off-by: Moshe Shemesh <moshe@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Link: https://patch.msgid.link/20240911201757.1505453-12-saeed@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net/mlx5: Add support for sync reset using hot reset
On device that supports sync reset for firmware activate using hot
reset, the driver queries the required reset method while handling the
sync reset request. If the required reset method is hot reset, the
driver will use pci_reset_bus() to reset the PCI link instead of the
link toggle.
Signed-off-by: Moshe Shemesh <moshe@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Link: https://patch.msgid.link/20240911201757.1505453-11-saeed@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net/mlx5: Add device cap for supporting hot reset in sync reset flow
New devices with new FW can support sync reset for firmware activate
using hot reset. Add capability for supporting it and add MFRL field to
query from FW which type of PCI reset method to use while handling sync
reset events.
Signed-off-by: Moshe Shemesh <moshe@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Link: https://patch.msgid.link/20240911201757.1505453-10-saeed@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Mark Bloch [Wed, 11 Sep 2024 20:17:50 +0000 (13:17 -0700)]
net/mlx5: fs, add support for no append at software level
Native capability for some steering engines lacks support for adding an
additional match with the same value to the same flow group. To accommodate
the NO APPEND flag in these scenarios, we include the new rule in the
existing flow table entry (fte) without immediate hardware commitment. When
a request is made to delete the corresponding hardware rule, we then commit
the pending rule to hardware.
Only one pending rule is supported because NO_APPEND is primarily used
during replacement operations. In this scenario, a rule is initially added.
When it needs replacement, the new rule is added with NO_APPEND set. Only
after the insertion of the new rule is the original rule deleted.
Signed-off-by: Mark Bloch <mbloch@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Link: https://patch.msgid.link/20240911201757.1505453-9-saeed@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Mark Bloch [Wed, 11 Sep 2024 20:17:49 +0000 (13:17 -0700)]
net/mlx5: fs, separate action and destination into distinct struct
Introduce a dedicated structure to encapsulate flow context, actions,
destination count, and modification mask. This refactoring lays the
groundwork for forthcoming patches that will integrate the NO APPEND
software logic. Future modifications should focus solely on these
specific fields.
Signed-off-by: Mark Bloch <mbloch@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Link: https://patch.msgid.link/20240911201757.1505453-8-saeed@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net/mlx5: fs, make get_root_namespace API function
As preparation for HW Steering support, where the function
get_root_namespace() is needed to get root FDB, make it an API function
and rename it to mlx5_get_root_namespace().
Reviewed-by: Yevgeny Kliteynik <kliteyn@nvidia.com> Signed-off-by: Moshe Shemesh <moshe@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com> Reviewed-by: Jacob Keller <jacob.e.keller@intel.com> Reviewed-by: Kalesh AP <kalesh-anakkur.purayil@broadcom.com> Link: https://patch.msgid.link/20240911201757.1505453-5-saeed@kernel.org Signed-off-by: Jakub Kicinski <kuba@kernel.org>
net/mlx5: fs, move steering common function to fs_cmd.h
As preparation for HW steering support in fs core level, move SW
steering helper function that can be reused by HW steering to fs_cmd.h.
The function mlx5_fs_cmd_is_fw_term_table() checks if a flow table is a
flow steering termination table and so should be handled by FW steering.
====================
net: Use IRQF_NO_AUTOEN flag in request_irq()
As commit cbe16f35bee6 ("genirq: Add IRQF_NO_AUTOEN for request_irq/nmi()")
said, reqeust_irq() and then disable_irq() is unsafe.
IRQF_NO_AUTOEN flag can be used by drivers to request_irq(). It prevents
the automatic enabling of the requested interrupt in the same safe way.
With that the usage can be simplified and corrected.
====================
disable_irq() after request_irq() still has a time gap in which
interrupts can come. request_irq() with IRQF_NO_AUTOEN flag will
disable IRQ auto-enable when request IRQ.