- 27 Mar, 2020 40 commits
-
-
Vladimir Oltean authored
Many switches don't have an explicit knob for configuring the MTU (maximum transmission unit per interface). Instead, they do the length-based packet admission checks on the ingress interface, for reasons that are easy to understand (why would you accept a packet in the queuing subsystem if you know you're going to drop it anyway). So it is actually the MRU that these switches permit configuring. In Linux there only exists the IFLA_MTU netlink attribute and the associated dev_set_mtu function. The comments like to play blind and say that it's changing the "maximum transfer unit", which is to say that there isn't any directionality in the meaning of the MTU word. So that is the interpretation that this patch is giving to things: MTU == MRU. When 2 interfaces having different MTUs are bridged, the bridge driver MTU auto-adjustment logic kicks in: what br_mtu_auto_adjust() does is it adjusts the MTU of the bridge net device itself (and not that of the slave net devices) to the minimum value of all slave interfaces, in order for forwarded packets to not exceed the MTU regardless of the interface they are received and send on. The idea behind this behavior, and why the slave MTUs are not adjusted, is that normal termination from Linux over the L2 forwarding domain should happen over the bridge net device, which _is_ properly limited by the minimum MTU. And termination over individual slave devices is possible even if those are bridged. But that is not "forwarding", so there's no reason to do normalization there, since only a single interface sees that packet. The problem with those switches that can only control the MRU is with the offloaded data path, where a packet received on an interface with MRU 9000 would still be forwarded to an interface with MRU 1500. And the br_mtu_auto_adjust() function does not really help, since the MTU configured on the bridge net device is ignored. In order to enforce the de-facto MTU == MRU rule for these switches, we need to do MTU normalization, which means: in order for no packet larger than the MTU configured on this port to be sent, then we need to limit the MRU on all ports that this packet could possibly come from. AKA since we are configuring the MRU via MTU, it means that all ports within a bridge forwarding domain should have the same MTU. And that is exactly what this patch is trying to do. >From an implementation perspective, we try to follow the intent of the user, otherwise there is a risk that we might livelock them (they try to change the MTU on an already-bridged interface, but we just keep changing it back in an attempt to keep the MTU normalized). So the MTU that the bridge is normalized to is either: - The most recently changed one: ip link set dev swp0 master br0 ip link set dev swp1 master br0 ip link set dev swp0 mtu 1400 This sequence will make swp1 inherit MTU 1400 from swp0. - The one of the most recently added interface to the bridge: ip link set dev swp0 master br0 ip link set dev swp1 mtu 1400 ip link set dev swp1 master br0 The above sequence will make swp0 inherit MTU 1400 as well. Suggested-by:
Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Vladimir Oltean authored
It is useful be able to configure port policers on a switch to accept frames of various sizes: - Increase the MTU for better throughput from the default of 1500 if it is known that there is no 10/100 Mbps device in the network. - Decrease the MTU to limit the latency of high-priority frames under congestion, or work around various network segments that add extra headers to packets which can't be fragmented. For DSA slave ports, this is mostly a pass-through callback, called through the regular ndo ops and at probe time (to ensure consistency across all supported switches). The CPU port is called with an MTU equal to the largest configured MTU of the slave ports. The assumption is that the user might want to sustain a bidirectional conversation with a partner over any switch port. The DSA master is configured the same as the CPU port, plus the tagger overhead. Since the MTU is by definition L2 payload (sans Ethernet header), it is up to each individual driver to figure out if it needs to do anything special for its frame tags on the CPU port (it shouldn't except in special cases). So the MTU does not contain the tagger overhead on the CPU port. However the MTU of the DSA master, minus the tagger overhead, is used as a proxy for the MTU of the CPU port, which does not have a net device. This is to avoid uselessly calling the .change_mtu function on the CPU port when nothing should change. So it is safe to assume that the DSA master and the CPU port MTUs are apart by exactly the tagger's overhead in bytes. Some changes were made around dsa_master_set_mtu(), function which was now removed, for 2 reasons: - dev_set_mtu() already calls dev_validate_mtu(), so it's redundant to do the same thing in DSA - __dev_set_mtu() returns 0 if ops->ndo_change_mtu is an absent method That is to say, there's no need for this function in DSA, we can safely call dev_set_mtu() directly, take the rtnl lock when necessary, and just propagate whatever errors get reported (since the user probably wants to be informed). Some inspiration (mainly in the MTU DSA notifier) was taken from a vaguely similar patch from Murali and Florian, who are credited as co-developers down below. Co-developed-by:Murali Krishna Policharla <murali.policharla@broadcom.com> Signed-off-by:
-
Murali Krishna Policharla authored
Change DMA descriptor length to handle jumbo frames beyond 8192 bytes. Also update jumbo frame max size to include FCS, the DMA packet length received includes FCS. Signed-off-by:
Arun Parameswaran <arun.parameswaran@broadcom.com> Reviewed-by:
Ray Jui <ray.jui@broadcom.com> Signed-off-by:
-
Murali Krishna Policharla authored
The BCM7XX PHY family requires special configuration to pass jumbo frames. Do that during initial PHY setup. Signed-off-by:
Scott Branden <scott.branden@broadcom.com> Signed-off-by:
-
David S. Miller authored
Merge tag 'wireless-drivers-next-2020-03-27' of git://git.kernel.org/pub/scm/linux/kernel/git/kvalo/wireless-drivers-next Kalle Valo says: ==================== wireless-drivers-next patches for v5.7 Third set of patches for v5.7. Nothing really special this time, business as usual. When pulling this to net-next there's again a conflict in: drivers/net/wireless/intel/iwlwifi/pcie/drv.c To solve this drop these three lines from the conflict (the first hunk from "HEAD") as the whole AX200 block was moved above in the same file: IWL_DEV_INFO(0x2723, 0x1653, iwl_ax200_cfg_cc, iwl_ax200_killer_1650w_name), IWL_DEV_INFO(0x2723, 0x1654, iwl_ax200_cfg_cc, iwl_ax200_killer_1650x_name), IWL_DEV_INFO(0x2723, IWL_CFG_ANY, iwl_ax200_cfg_cc, iwl_ax200_name), And keep all the __IWL_DEV_INFO() entries (the second hunk). In other words, take everything from wireless-drivers-next. When running 'git diff' after the resolution the output should be empty. Major changes: brcmfmac * add USB autosuspend support ath11k * handle RX fragments * enable PN offload * add support for HE BSS color iwlwifi * support new FW API version * support for EDCA measurements * new scan API features * enable new firmware debugging code ==================== Kalle gave me directions on how to resolve the iwlwifi conflict as follows: ==================== When pulling this to net-next there's again a conflict in: drivers/net/wireless/intel/iwlwifi/pcie/drv.c To solve this drop these three lines from the conflict (the first hunk from "HEAD") as the whole AX200 block was moved above in the same file: IWL_DEV_INFO(0x2723, 0x1653, iwl_ax200_cfg_cc, iwl_ax200_killer_1650w_name), IWL_DEV_INFO(0x2723, 0x1654, iwl_ax200_cfg_cc, iwl_ax200_killer_1650x_name), IWL_DEV_INFO(0x2723, IWL_CFG_ANY, iwl_ax200_cfg_cc, iwl_ax200_name), And keep all the __IWL_DEV_INFO() entries (the second hunk). In other words, take everything from wireless-drivers-next. When running 'git diff' after the resolution the output should be empty. ==================== Signed-off-by:
-
David S. Miller authored
Julian Wiedmann says: ==================== s390/qeth: updates 2020-03-27 please apply the following patch series for qeth to netdev's net-next tree. Spring clean edition: - remove one sysfs attribute that was never put in use, - make support for OSN and OSX devices optional, and - probe for removal of the obsolete OSN support. ==================== Signed-off-by: -
Julian Wiedmann authored
OSN devices currently spend an awful long time in qeth_l2_set_online() until various unsupported HW cmds time out. This has been broken for over two years, ever since commit d22ffb5a ("s390/qeth: fix IPA command submission race") triggered a FW bug in cmd processing. Prior to commit 782e4a79 ("s390/qeth: don't poll for cmd IO completion"), this wait for timeout would have even been spent busy-polling. The offending patch was picked up by stable and all relevant distros, and yet noone noticed. OSN setups only ever worked in combination with an out-of-tree blob, and the last machine that even offered HW with OSN support was released back in 2015. Rather than attempting to work-around this FW issue for no actual gain, add a deprecation warning so anyone who still wants to maintain this part of the code can speak up. Else rip it all out in 2021. Signed-off-by:
Julian Wiedmann <jwi@linux.ibm.com> Signed-off-by:
-
Julian Wiedmann authored
The last machine generation that supports OSN is z13, and OSX is only supported up to z14. Allow users and distros to decide whether they still need support for these device types. Signed-off-by:
-
Julian Wiedmann authored
Ever since commit 4a71df50 ("qeth: new qeth device driver") introduced this attribute, it can be read & written but has no actual effect. Signed-off-by:
-
David S. Miller authored
Vasundhara Volam says: ==================== bnxt_en: Updates to devlink info_get cb This series adds support for a generic macro to devlink info_get cb. Adds support for fw.mgmt.api and board.id info to bnxt_en driver info_get cb. Also, updates the devlink-info.rst and bnxt.rst documentation accordingly. This series adds a patch to fix few macro names that maps to bnxt_en firmware versions. v1->v2: Remove ECN dev param, base_mh_addr and serial number info support in this series. Rename drv.spec macro to fw.api. --- v2->v3: Remove hw.addr info as it is per netdev but not per device info. --- v3->v4: Rename "fw.api" to "fw.mgmt.api". Also, add a patch that modifies few macro names in info_get command, to match the devlink documentation. ==================== Signed-off-by: -
Vasundhara Volam authored
Fix macro names to report fw.mgmt and fw.ncsi versions to match the devlink documentation. Example display after fixes: $ devlink dev info pci/0000:af:00.0 pci/0000:af:00.0: driver bnxt_en serial_number B0-26-28-FF-FE-25-84-20 versions: fixed: board.id BCM957454A4540 asic.id C454 asic.rev 1 running: fw 216.1.154.0 fw.psid 0.0.0 fw.mgmt 216.1.146.0 fw.mgmt.api 1.10.1 fw.ncsi 864.0.44.0 fw.roce 216.1.16.0 Fixes: 9599e036 ("bnxt_en: Add support for devlink info command") Signed-off-by:Vasundhara Volam <vasundhara-v.volam@broadcom.com> Signed-off-by:
Michael Chan <michael.chan@broadcom.com> Signed-off-by:
-
Vasundhara Volam authored
Add part number info from the vital product data to info_get command via devlink tool. Update bnxt.rst documentation as well. Cc: Jakub Kicinski <kuba@kernel.org> Signed-off-by:
-
Vasundhara Volam authored
Store the part number and serial number information from VPD in the bnxt structure. Follow up patch will add the support to display the information via devlink command. Signed-off-by:
-
Vasundhara Volam authored
This patch adds a new macro for serial number keyword. Acked-by:
Bjorn Helgaas <bhelgaas@google.com> Signed-off-by:
-
Vasundhara Volam authored
Display the minimum version of firmware interface spec supported between driver and firmware. Also update bnxt.rst documentation file. Signed-off-by:
-
Vasundhara Volam authored
Add definition and documentation for the new generic info "fw.mgmt.api". This macro specifies the version of the software interfaces between driver and firmware. Cc: Jakub Kicinski <kuba@kernel.org> Cc: Jacob Keller <jacob.e.keller@intel.com> Cc: Jiri Pirko <jiri@mellanox.com> Signed-off-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
-
David S. Miller authored
Ido Schimmel says: ==================== mlxsw: Various static checkers fixes Jakub told me he gets some warnings with W=1, so I decided to check with sparse, smatch and coccinelle as well. This patch set fixes all the issues found. None are actual bugs / regressions and therefore not targeted at net. Patches #1-#2 add missing kernel-doc comments. Patch #3 removes dead code. Patch #4 reworks the ACL code to avoid defining a static variable in a header file. Patch #5 removes unnecessary conversion to bool that coccinelle warns about. Patch #6 avoids false-positive uninitialized symbol errors emitted by smatch. ==================== Signed-off-by: -
Ido Schimmel authored
Suppress the following smatch errors. None of these are actually possible with current code paths. drivers/net/ethernet/mellanox/mlxsw//spectrum_router.c:1220 mlxsw_sp_ipip_entry_find_decap() error: uninitialized symbol 'saddrp'. drivers/net/ethernet/mellanox/mlxsw//spectrum_router.c:1220 mlxsw_sp_ipip_entry_find_decap() error: uninitialized symbol 'saddr_len'. drivers/net/ethernet/mellanox/mlxsw//spectrum_router.c:1221 mlxsw_sp_ipip_entry_find_decap() error: uninitialized symbol 'saddr_prefix_len'. drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c:1390 mlxsw_sp_netdevice_ipip_ol_reg_event() error: uninitialized symbol 'ipipt'. drivers/net/ethernet/mellanox/mlxsw/spectrum_router.c:3255 mlxsw_sp_nexthop_group_update() error: uninitialized symbol 'err'. Signed-off-by:
Ido Schimmel <idosch@mellanox.com> Reviewed-by:
-
Ido Schimmel authored
Suppress following warning from coccinelle: drivers/net/ethernet/mellanox/mlxsw//switchx2.c:183:63-68: WARNING: conversion to bool not needed here Signed-off-by:
-
Ido Schimmel authored
The static array 'mlxsw_afk_element_infos' in 'core_acl_flex_keys.h' is copied to each file that includes the header, but not all use it. This results in the following warnings when compiling with W=1: drivers/net/ethernet/mellanox/mlxsw//core_acl_flex_keys.h:76:44: warning: ‘mlxsw_afk_element_infos’ defined but not used [-Wunused-const-variable=] One way to suppress the warning is to mark the array with '__maybe_unused', but another option is to remove it from the header file entirely. Change 'struct mlxsw_afk_element_inst' to store the key to the array ('element') instead of the array value keyed by 'element'. Adjust the different users accordingly. Signed-off-by: -
Ido Schimmel authored
In merge commit 50853808 ("Merge branch 'mlxsw-Prepare-for-VLAN-aware-bridge-w-VxLAN'") I flipped mlxsw to use emulated 802.1Q FIDs and correspondingly emulated VLAN RIFs. This means that the non-emulated variants are no longer used. Remove them and suppress the following warnings when compiling with W=1: drivers/net/ethernet/mellanox/mlxsw//spectrum_router.c:7572:38: warning: ‘mlxsw_sp_rif_vlan_ops’ defined but not used [-Wunused-const-variable=] drivers/net/ethernet/mellanox/mlxsw//spectrum_fid.c:584:41: warning: ‘mlxsw_sp_fid_8021q_family’ defined but not used [-Wunused-const-variable=] Signed-off-by:
-
Ido Schimmel authored
Suppress following warnings when compiling with W=1: drivers/net/ethernet/mellanox/mlxsw//spectrum_router.c:1552: warning: Function parameter or member 'mlxsw_sp' not described in '__mlxsw_sp_ipip_entry_update_tunnel' drivers/net/ethernet/mellanox/mlxsw//spectrum_router.c:1552: warning: Function parameter or member 'ipip_entry' not described in '__mlxsw_sp_ipip_entry_update_tunnel' drivers/net/ethernet/mellanox/mlxsw//spectrum_router.c:1552: warning: Function parameter or member 'extack' not described in '__mlxsw_sp_ipip_entry_update_tunnel' Signed-off-by:
-
Ido Schimmel authored
Suppress following warning when compiling with W=1: drivers/net/ethernet/mellanox/mlxsw//i2c.c:78: warning: Function parameter or member 'cmd' not described in 'mlxsw_i2c' Signed-off-by:
-
Kalle Valo authored
Merge tag 'iwlwifi-next-for-kalle-2020-03-27' of git://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/iwlwifi-next Second set of iwlwifi patches intended for v5.7 * Refactoring of the device selection algorithms continues; * Improvement in the initialization fo SoC-based devices; * Support for FW scan API; * Some additions to our FW debuggging capabilities; * More refactoring of the device selection algorithms; * Support new FW API version; * Support for EDCA measurements; * New scan API features; * Enable new debugging code; * Some other small fixes and clean-ups;
-
Johannes Berg authored
Some of the places using this want the newline, but not all, so remove the newline from it and generate it in the debugfs files where it's desired. The effect of this is not printing double newlines in debug log messages. Change-Id: Ia59b0abbef16b6783fcabc095c5fde16bd047a26 Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
Luca Coelho <luciano.coelho@intel.com>
-
Johannes Berg authored
There's no need for this to be exposed outside of the tx.c file, make it static. Change-Id: I41d40008311b108d0578bd2ec73c5477e700a839 Signed-off-by:
-
Luca Coelho authored
Start supporting API version 53 for 22000 series. Change-Id: I5725e46394f3f53c3069723fd513cc53c7df383d Signed-off-by: -
Johannes Berg authored
This is used to initialize the default value, but refers back to the value itself, essentially leading to a val = val assignment at init time ... that's useless, remove it and use _NODEF. Change-Id: I725923016563c34ce2fa057bf7c12984e1041c49 Signed-off-by:
-
Luca Coelho authored
Now that YoYo is more mature, we can enable it by default, so we collect data in the new way. Change-Id: Ic1a147f935286b085ca8bdb248a7493b7c6341ea Signed-off-by: -
Tova Mussai authored
The FW changed the maximum number of scan offload profiles to 8 in new APIs. Support it by changing the scan_offload_profile_cfg struct to be more dynamic, so we can reuse most of the code and only change size of the profiles array. Change-Id: I65210cf541af46e2675a8c764d5871f7f5b070d6 Signed-off-by:
Tova Mussai <tova.mussai@intel.com> Signed-off-by:
-
rotem saado authored
When we have an assert during D3 we want to dump the internal buffer memory even if are we working in DRAM debug mode. We should not block dumping it. Change-Id: I69aad2d4904c4f8bb653c61e8781a2e07780054f Signed-off-by:
Rotem Saado <rotem.saado@intel.com> Signed-off-by:
-
Luca Coelho authored
We don't support the FPGA versions of this card combination anymore. Remove the cfg mangling that tries to load it and all the relevant structures. Change-Id: I190652101afcab682cfba873d062992f11efca32 Signed-off-by: -
Luca Coelho authored
We don't support QnJ HR STEP A anymore. Remove the structures we used for it. Change-Id: I0dfd88232bdc8ff2dd9c4368b8ed9a0c40c86bc8 Signed-off-by: -
Luca Coelho authored
A couple of SoCs, which can be recognized by PCI device IDs 0xA0F0 and 0x43F0, need a longer wait for the xtal to stabilize. To handle this, add a new trans_cfg structure for Qu devices with a larger xtal_latency value and apply them to the devices recognized by these IDs. Also add a flag that allows us to inform the FW that the low latency xtal should be used. Change-Id: I8a14c6af45ea14d8e7f1ef38a589158f38d0c0ea Signed-off-by: -
Luca Coelho authored
This new command is mostly backwards compatible, with the exception that the device_type element was changed into a bitmask. The device type bit remains the same (because we only had 0 and 1 anyway), but when using v1 we can't set any other bits, because that would change the integer. Other than that, the struct remains the same and the driver can set the device_type bit in both cases, but it can only set the low_latency bit if VER_2 is used. Change-Id: Ib68d4c821ebcce253b42ed0ea15881fb4e3e01da Signed-off-by: -
Avraham Stern authored
Add support for requesting trigger-based / non trigger-based measurements. Change-Id: Ib4d0c471da9c50d9981541a7f5926db384a0f7ce Signed-off-by:
Avraham Stern <avraham.stern@intel.com> Signed-off-by:
-
Luca Coelho authored
Now that we identify the correct cfgs with the new tables for Qu step C and QuZ with Jf, we can remove the mangling we do later on. Change-Id: Ic01ce67db147e897ad2424f0e05a70a00d2c620e Signed-off-by: -
Luca Coelho authored
All the QnJ devices have a similar matching to the other Qu devices, but needs a different configuration. Convert the QnJ devices to the new table accordingly. Change-Id: If236ef3d0da3e605a3379922818f5897e0affd7e Signed-off-by: -
Luca Coelho authored
Add new generic iwl_trans structures for these devices and apply the correct cfg depending on the device characteristics. Since we have to match Qu with IWL_CONFIG_ANY, we also need to move the Hr devices to the new table, but for now we keep matching on PCI device and subsystem device IDs. Change-Id: I14e9146a99621ff11ce50bc746a4b88af508fee0 Signed-off-by: -
Luca Coelho authored
We need to use different firmware versions for different HW steps with certain devices. Prepare for this differentiation by adding HW step to the new device table. Change-Id: Ib1afb7b0c89e9dc2d26e6d32ea19e978c17ba1dd Signed-off-by:
-
