- 29 Oct, 2022 21 commits
-
-
Willem de Bruijn authored
Extend packet socket option PACKET_IGNORE_OUTGOING to fanout groups. The socket option sets ptype.ignore_outgoing, which makes dev_queue_xmit_nit skip the socket. When the socket joins a fanout group, the option is not reflected in the struct ptype of the group. dev_queue_xmit_nit only tests the fanout ptype, so the flag is ignored once a socket joins a fanout group. Inheriting the option from a socket would change established behavior. Different sockets in the group can set different flags, and can also change them at runtime. Testing in packet_rcv_fanout defeats the purpose of the original patch, which is to avoid skb_clone in dev_queue_xmit_nit (esp. for MSG_ZEROCOPY packets). Instead, introduce a new fanout group flag with the same behavior. Tested with https://github.com/wdebruij/kerneltools/blob/master/tests/test_psock_fanout_ignore_outgoing.cSigned-off-by:
Willem de Bruijn <willemb@google.com> Reviewed-by:
Eric Dumazet <edumazet@google.com> Link: https://lore.kernel.org/r/20221027211014.3581513-1-willemdebruijn.kernel@gmail.comSigned-off-by:
Jakub Kicinski <kuba@kernel.org>
-
Lukasz Czapnik authored
In the event of a Tx hang it can be useful to read a variety of hardware registers to capture some state about why the transmit queue got stuck. Extend the ETHTOOL_GREGS dump provided by the ice driver with several CSR registers that provide such relevant information regarding the hardware Tx state. This enables capturing relevant data to enable debugging such a Tx hang. Signed-off-by:
Lukasz Czapnik <lukasz.czapnik@intel.com> Signed-off-by:
Mateusz Palczewski <mateusz.palczewski@intel.com> Tested-by: Gurucharan <gurucharanx.g@intel.com> (A Contingent worker at Intel) Link: https://lore.kernel.org/r/20221027104239.1691549-1-jacob.e.keller@intel.comSigned-off-by:
-
Jakub Kicinski authored
Russell King says: ==================== Clean up SFP register definitions This two-part patch series cleans up the SFP register definitions by 1. converting them from hex to decimal, as all the definitions in the documents use decimal, this makes it easier to cross-reference. 2. moving the bit definitions for each register along side their register address definition ==================== Link: https://lore.kernel.org/r/Y1qFvaDlLVM1fHdG@shell.armlinux.org.ukSigned-off-by:
-
Russell King (Oracle) authored
Just as we do for the A2h enum, arrange the A0h enum to have the field definitions next to their corresponding register index. Signed-off-by:
Russell King (Oracle) <rmk+kernel@armlinux.org.uk> Reviewed-by:
Andrew Lunn <andrew@lunn.ch> Signed-off-by:
-
Russell King (Oracle) authored
The register indexes in the standards are in decimal rather than hex, so lets specify them in decimal in the header file so we can easily cross-reference without converting between hex and decimal. Signed-off-by:
-
Jakub Kicinski authored
Russell King says: ==================== net: mtk_eth_soc: improve PCS implementation As a result of invesigations from Frank Wunderlich, we know a lot more about the Mediatek "SGMII" PCS block, and can implement the PCS support correctly. This series achieves that, and Frank has tested the final result and reports that it works for him. The series could do with further testing by others, but I suspect that is unlikely to happen until it is merged based on past performances with this driver. Briefly, the patches in order: 1. Add a new helper to get the link timer duration in nanoseconds 2. Add definitions for the newly discovered registers and updates to bit definitions, including bitmasks for the BMCR, BMSR and two advertisement registers. 3. Remove unnecessary/unused error handling (functions always returning zero.) 4. Adding the missing pcs_get_state() implementation. 5. Converting the code to use regmap_update_bits() rather than open-coding read-modify-write sequences. 6. Adding out-of-band speed and duplex forcing for all non-inband modes not just the 802.3z link modes the code currently does. 7. Moving the release of the PHY power down to the main pcs_config() function. 8. Moving the interface speed selection to the main pcs_config() function. 9. Adding advertisement programming. 10. Adding correct link timer programming using the new helper in the first patch. 11. Adding support for 802.3z negotiation. There is one remaining issue - when configuring the PCS for in-band, for some reason the AN restart bit is always set. This should not be necessary, but requires further investigation with the hardware to find out whether it is really necessary. I suspect this was a work around for a previous poor implementation. ==================== Link: https://lore.kernel.org/r/Y1qDMw+DJLAJHT40@shell.armlinux.org.ukSigned-off-by: -
Russell King (Oracle) authored
As a result of help from Frank Wunderlich to investigate and test, we now know how to program this PCS for in-band 802.3z negotiation. Add support for this by moving the contents of the two functions into the common mtk_pcs_config() function and adding the register settings for 802.3z negotiation. Signed-off-by:
-
Russell King (Oracle) authored
Program the link timer appropriately for the interface mode being used, using the newly introduced phylink helper that provides the nanosecond link timer interval. The intervals are 1.6ms for SGMII based protocols and 10ms for 802.3z based protocols. Signed-off-by:
-
Russell King (Oracle) authored
Program the advertisement into the mtk PCS block. Signed-off-by:
-
Russell King (Oracle) authored
Move the selection of the underlying interface speed to the pcs_config function, so we always program the interface speed. Signed-off-by:
-
Russell King (Oracle) authored
The PHY power up is common to both configuration paths, so move it into the parent function. We need to do this for all serdes modes. Signed-off-by:
-
Russell King (Oracle) authored
Add support for forcing the link speed and duplex setting in the pcs_link_up() method for out of band modes, which will be useful when we finish converting the pcs_config() method. Until then, we still have to force duplex for 802.3z modes to work correctly. Signed-off-by:
-
Russell King (Oracle) authored
mtk_sgmii does a lot of read-modify-write operations, for which there is a specific regmap function. Use this function instead of open-coding the operations. Signed-off-by:
-
Russell King (Oracle) authored
Add a pcs_get_state() implementation which uses the advertisements to compute the resulting link modes, and BMSR contents to determine negotiation and link status. Signed-off-by:
-
Russell King (Oracle) authored
The functions called by the pcs_config() method always return zero, so there is no point trying to handle an error from these functions. Make these functions void, eliminate the "err" variable and simply return zero from the pcs_config() function itself. Signed-off-by:
-
Russell King (Oracle) authored
As a result of help from Frank Wunderlich to investigate and test, we know a bit more about the PCS on the Mediatek platforms. Update the definitions from this investigation. This PCS appears similar, but not identical to the Lynx PCS. Although not included in this patch, but for future reference, the PHY ID registers at offset 4 read as 0x4d544950 'MTIP'. Signed-off-by:
-
Russell King (Oracle) authored
Add a helper to convert the PHY interface mode to the required link timer setting as stated by the appropriate standard. Inappropriate interface modes return an error. Signed-off-by:
-
Jakub Kicinski authored
Sebastian Andrzej Siewior says: ==================== net: Remove the obsolte u64_stats_fetch_*_irq() This is the removal of u64_stats_fetch_*_irq() users in networking. The prerequisites are part of v6.1-rc1. The spi and bpf bits are not part of the series and have been routed directly. ==================== Link: https://lore.kernel.org/r/20221026132215.696950-1-bigeasy@linutronix.deSigned-off-by:
-
Thomas Gleixner authored
Now that the 32bit UP oddity is gone and 32bit uses always a sequence count, there is no need for the fetch_irq() variants anymore. Convert to the regular interface. Signed-off-by:
Thomas Gleixner <tglx@linutronix.de> Signed-off-by:
Sebastian Andrzej Siewior <bigeasy@linutronix.de> Acked-by:
Peter Zijlstra (Intel) <peterz@infradead.org> Signed-off-by:
-
Thomas Gleixner authored
Now that the 32bit UP oddity is gone and 32bit uses always a sequence count, there is no need for the fetch_irq() variants anymore. Convert to the regular interface. Signed-off-by:
-
Jakub Kicinski authored
==================== pull-request: wireless-next-2022-10-28 First set of patches v6.2. mac80211 refactoring continues for Wi-Fi 7. All mac80211 driver are now converted to use internal TX queues, this might cause some regressions so we wanted to do this early in the cycle. Note: wireless tree was merged[1] to wireless-next to avoid some conflicts with mac80211 patches between the trees. Unfortunately there are still two smaller conflicts in net/mac80211/util.c which Stephen also reported[2]. In the first conflict initialise scratch_len to "params->scratch_len ?: 3 * params->len" (note number 3, not 2!) and in the second conflict take the version which uses elems->scratch_pos. [1] https://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next.git/commit/?id=dfd2d876b3fda1790bc0239ba4c6967e25d16e91 [2] https://lore.kernel.org/all/20221020032340.5cf101c0@canb.auug.org.au/ mac80211 - preparation for Wi-Fi 7 Multi-Link Operation (MLO) continues - add API to show the link STAs in debugfs - all mac80211 drivers are now using mac80211 internal TX queues (iTXQs) rtw89 - support 8852BE rtl8xxxu - support RTL8188FU brmfmac - support two station interfaces concurrently bcma - support SPROM rev 11 ==================== Link: https://lore.kernel.org/r/20221028132943.304ECC433B5@smtp.kernel.orgSigned-off-by:
-
- 28 Oct, 2022 19 commits
-
-
Dmitry Torokhov authored
Because we enable the clock immediately after acquiring it in probe, we can combine the 2 operations and use devm_clk_get_optional_enabled() helper. Signed-off-by:
Dmitry Torokhov <dmitry.torokhov@gmail.com> Reviewed-by:
Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
David S. Miller authored
Jiawen Wu says: ==================== net: WangXun txgbe ethernet driver This patch series adds support for WangXun 10 gigabit NIC, to initialize hardware, set mac address, and register netdev. Change log: v6: address comments: Jakub Kicinski: check with scripts/kernel-doc v5: address comments: Jakub Kicinski: clean build with W=1 C=1 v4: address comments: Andrew Lunn: https://lore.kernel.org/all/YzXROBtztWopeeaA@lunn.ch/ v3: address comments: Andrew Lunn: remove hw function ops, reorder functions, use BIT(n) for register bit offset, move the same code of txgbe and ngbe to libwx v2: address comments: Andrew Lunn: https://lore.kernel.org/netdev/YvRhld5rD%2FxgITEg@lunn.ch/ ==================== Signed-off-by: -
Jiawen Wu authored
Add MAC address related operations, and register netdev. Signed-off-by:
Jiawen Wu <jiawenwu@trustnetic.com> Signed-off-by:
-
Jiawen Wu authored
Reset and initialize the hardware by configuring the MAC layer. Signed-off-by:
-
Jiawen Wu authored
Get PCI config space info, set LAN id and check flash status. Signed-off-by:
-
David S. Miller authored
Mubashir Adnan Qureshi says: ==================== net: Add PLB functionality to TCP This patch series adds PLB (Protective Load Balancing) to TCP and hooks it up to DCTCP. PLB is disabled by default and can be enabled using relevant sysctls and support from underlying CC. PLB (Protective Load Balancing) is a host based mechanism for load balancing across switch links. It leverages congestion signals(e.g. ECN) from transport layer to randomly change the path of the connection experiencing congestion. PLB changes the path of the connection by changing the outgoing IPv6 flow label for IPv6 connections (implemented in Linux by calling sk_rethink_txhash()). Because of this implementation mechanism, PLB can currently only work for IPv6 traffic. For more information, see the SIGCOMM 2022 paper: https://doi.org/10.1145/3544216.3544226 ==================== Signed-off-by:
-
Mubashir Adnan Qureshi authored
rcv_wnd can be useful to diagnose TCP performance where receiver window becomes the bottleneck. rehash reports the PLB and timeout triggered rehash attempts by the TCP connection. Signed-off-by:
Mubashir Adnan Qureshi <mubashirq@google.com> Signed-off-by:
Yuchung Cheng <ycheng@google.com> Signed-off-by:
Neal Cardwell <ncardwell@google.com> Reviewed-by:
-
Mubashir Adnan Qureshi authored
A u32 counter is added to tcp_sock for counting the number of PLB triggered rehashes for a TCP connection. An SNMP counter is also added to count overall PLB triggered rehash events for a host. These counters are hooked up to PLB implementation for DCTCP. TCP_NLA_REHASH is added to SCM_TIMESTAMPING_OPT_STATS that reports the rehash attempts triggered due to PLB or timeouts. This gives a historical view of sustained congestion or timeouts experienced by the TCP connection. Signed-off-by:
-
Mubashir Adnan Qureshi authored
PLB support is added to TCP DCTCP code. As DCTCP uses ECN as the congestion signal, PLB also uses ECN to make decisions whether to change the path or not upon sustained congestion. Signed-off-by:
-
Mubashir Adnan Qureshi authored
Congestion control algorithms track PLB state and cause the connection to trigger a path change when either of the 2 conditions is satisfied: - No packets are in flight and (# consecutive congested rounds >= sysctl_tcp_plb_idle_rehash_rounds) - (# consecutive congested rounds >= sysctl_tcp_plb_rehash_rounds) A round (RTT) is marked as congested when congestion signal (ECN ce_ratio) over an RTT is greater than sysctl_tcp_plb_cong_thresh. In the event of RTO, PLB (via tcp_write_timeout()) triggers a path change and disables congestion-triggered path changes for random time between (sysctl_tcp_plb_suspend_rto_sec, 2*sysctl_tcp_plb_suspend_rto_sec) to avoid hopping onto the "connectivity blackhole". RTO-triggered path changes can still happen during this cool-off period. Signed-off-by:
-
Mubashir Adnan Qureshi authored
PLB (Protective Load Balancing) is a host based mechanism for load balancing across switch links. It leverages congestion signals(e.g. ECN) from transport layer to randomly change the path of the connection experiencing congestion. PLB changes the path of the connection by changing the outgoing IPv6 flow label for IPv6 connections (implemented in Linux by calling sk_rethink_txhash()). Because of this implementation mechanism, PLB can currently only work for IPv6 traffic. For more information, see the SIGCOMM 2022 paper: https://doi.org/10.1145/3544216.3544226 This commit adds new sysctl knobs and sets their default values for TCP PLB. Signed-off-by:
-
David S. Miller authored
Raju Lakkaraju says: ==================== net: phy: mxl-gpy: Add MDI-X This patch series add the MDI-X feature to GPY211 PHYs and Also Change return type to gpy_update_interface() function ==================== Signed-off-by: -
Raju Lakkaraju authored
Add support for MDI-X status and configuration for GPY211 chips Signed-off-by:
Raju Lakkaraju <Raju.Lakkaraju@microchip.com> Signed-off-by:
-
Raju Lakkaraju authored
gpy_update_interface() is called from gpy_read_status() which does return error codes. gpy_read_status() would benefit from returning -EINVAL, etc. Signed-off-by:
-
git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-nextJakub Kicinski authored
Pablo Neira Ayuso says: ==================== Netfilter updates for net-next 1) Move struct nft_payload_set definition to .c file where it is only used. 2) Shrink transport and inner header offset fields in the nft_pktinfo structure to 16-bits, from Florian Westphal. 3) Get rid of nft_objref Kbuild toggle, make it built-in into nf_tables. This expression is used to instantiate conntrack helpers in nftables. After removing the conntrack helper auto-assignment toggle it this feature became more important so move it to the nf_tables core module. Also from Florian. 4) Extend the existing function to calculate payload inner header offset to deal with the GRE and IPIP transport protocols. 6) Add inner expression support for nf_tables. This new expression provides a packet parser for tunneled packets which uses a userspace description of the expected inner headers. The inner expression invokes the payload expression (via direct call) to match on the inner header protocol fields using the inner link, network and transport header offsets. An example of the bytecode generated from userspace to match on IP source encapsulated in a VxLAN packet: # nft --debug=netlink add rule netdev x y udp dport 4789 vxlan ip saddr 1.2.3.4 netdev x y [ meta load l4proto => reg 1 ] [ cmp eq reg 1 0x00000011 ] [ payload load 2b @ transport header + 2 => reg 1 ] [ cmp eq reg 1 0x0000b512 ] [ inner type vxlan hdrsize 8 flags f [ meta load protocol => reg 1 ] ] [ cmp eq reg 1 0x00000008 ] [ inner type vxlan hdrsize 8 flags f [ payload load 4b @ network header + 12 => reg 1 ] ] [ cmp eq reg 1 0x04030201 ] 7) Store inner link, network and transport header offsets in percpu area to parse inner packet header once only. Matching on a different tunnel type invalidates existing offsets in the percpu area and it invokes the inner tunnel parser again. 8) Add support for inner meta matching. This support for NFTA_META_PROTOCOL, which specifies the inner ethertype, and NFT_META_L4PROTO, which specifies the inner transport protocol. 9) Extend nft_inner to parse GENEVE optional fields to calculate the link layer offset. 10) Update inner expression so tunnel offset points to GRE header to normalize tunnel header handling. This also allows to perform different interpretations of the GRE header from userspace. * git://git.kernel.org/pub/scm/linux/kernel/git/netfilter/nf-next: netfilter: nft_inner: set tunnel offset to GRE header offset netfilter: nft_inner: add geneve support netfilter: nft_meta: add inner match support netfilter: nft_inner: add percpu inner context netfilter: nft_inner: support for inner tunnel header matching netfilter: nft_payload: access ipip payload for inner offset netfilter: nft_payload: access GRE payload via inner offset netfilter: nft_objref: make it builtin netfilter: nf_tables: reduce nft_pktinfo by 8 bytes netfilter: nft_payload: move struct nft_payload_set definition where it belongs ==================== Link: https://lore.kernel.org/r/20221026132227.3287-1-pablo@netfilter.orgSigned-off-by: -
Yang Li authored
./drivers/net/ethernet/freescale/dpaa2/dpaa2-xsk.c:453:42-47: WARNING: conversion to bool not needed here Link: https://bugzilla.openanolis.cn/show_bug.cgi?id=2577Reported-by:
Abaci Robot <abaci@linux.alibaba.com> Signed-off-by:
Yang Li <yang.lee@linux.alibaba.com> Link: https://lore.kernel.org/r/20221026051824.38730-1-yang.lee@linux.alibaba.comSigned-off-by:
-
Jakub Kicinski authored
Shannon Nelson says: ==================== ionic: VF attr replay and other updates For better VF management when a FW update restart or a FW crash recover is detected, the PF now will replay any user specified VF attributes to be sure the FW hasn't lost them in the restart. Newer FW offers more packet processing offloads, so we now support them in the driver. A small refactor of the Rx buffer fill cleans a bit of code and will help future work on buffer caching. ==================== Link: https://lore.kernel.org/r/20221026143744.11598-1-snelson@pensando.ioSigned-off-by:
-
Neel Patel authored
The same pre-work code is used before each call to ionic_rx_fill(), so bring it in and make it a part of the routine. Signed-off-by:
Neel Patel <neel@pensando.io> Signed-off-by:
Shannon Nelson <snelson@pensando.io> Signed-off-by:
-
Neel Patel authored
Support stateless offloads for GRE, VXLAN, GENEVE, IPXIP4 and IPXIP6 when the FW supports them. Signed-off-by:
-
