- 26 Aug, 2021 2 commits
-
-
Shaul Triebitz authored
If broadcast TWT is supported in the BSS, tell the firmware about it by setting the BROADCAST_TWT_SUPPORTED in the MAC context command. Signed-off-by:
Shaul Triebitz <shaul.triebitz@intel.com> Signed-off-by:
Luca Coelho <luciano.coelho@intel.com> Link: https://lore.kernel.org/r/iwlwifi.20210802170640.736c3b1bc915.I10583bb6f808aa60954da26106bbc8c26620cbe8@changeidSigned-off-by:
-
Mordechay Goodstein authored
Our HW supersets STBC for HE but never set it in assoc IE, fix it by setting it and enable using it. Signed-off-by:
Mordechay Goodstein <mordechay.goodstein@intel.com> Signed-off-by:
-
- 25 Aug, 2021 38 commits
-
-
David S. Miller authored
Juhee Kang says: ==================== samples: pktgen: enhance the ability to print the execution results of samples This patch series improves the ability to print the execution result of pktgen samples by adding a line which calls the function before termination and adding trap SIGINT. Also, this series documents the latest pktgen usage options. Currently, pktgen samples print the execution result when terminated usually. However, sample03 is not working properly. This is results of sample04 and sample03: # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample04_many_flows.sh -n 1 Running... ctrl^C to stop Device: eth0@0 Result: OK: 19(c5+d13) usec, 1 (60byte,0frags) 51762pps 24Mb/sec (24845760bps) errors: 0 # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample03_burst_single_flow.sh -n 1 Running... ctrl^C to stop Because sample03 doesn't call the function which prints the execution result when terminated normally, unlike other samples. So the first commit solves this issue by adding a line which calls the function before termination. Also, all pktgen samples are able to send infinite messages per thread by setting the count option to 0, and pktgen is stopped by Ctrl-C. However, the sample besides sample{3...5} don't work appropriately because Ctrl-C stops the script, not just pktgen. This is results of samples: # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample04_many_flows.sh -n 0 Running... ctrl^C to stop ^CDevice: eth0@0 Result: OK: 569657(c569538+d118) usec, 84650 (60byte,0frags) 148597pps 71Mb/sec (71326560bps) errors: 0 # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample01_simple.sh -n 0 Running... ctrl^C to stop ^C # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample02_multiqueue.sh -n 0 Running... ctrl^C to stop ^C # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample06_numa_awared_queue_irq_affinity.sh -n 0 Running... ctrl^C to stop ^C # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_bench_xmit_mode_netif_receive.sh -n 0 Running... ctrl^C to stop ^C # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_bench_xmit_mode_queue_xmit.sh -n 0 Running... ctrl^C to stop ^C So the second commit solves this issue by adding trap SIGINT. Also, changes control_c function to print_results to maintain consistency with other samples on the first commit and second commit. And current pktgen.rst documentation doesn't add the latest pktgen sample usage options such as count and IPv6, and so on. Also, the old pktgen sample scripts are still included in the document. The old scripts were removed by the commit a4b6ade8 ("samples/pktgen: remove remaining old pktgen sample scripts"). Thus, the last commit documents the latest pktgen sample usage and removes old sample scripts. And fixes a minor typo. ==================== Signed-off-by:David S. Miller <davem@davemloft.net>
-
Juhee Kang authored
Currently, the pktgen.rst documentation doesn't cover the latest pktgen sample usage options such as count and IPv6, and so on. Also, this documentation includes the old sample scripts which are no longer use because it was removed by the commit a4b6ade8 ("samples/pktgen : remove remaining old pktgen sample scripts") Thus, this commit documents pktgen sample usage using the latest options and removes old sample scripts, and fixes a minor typo. Signed-off-by:
Juhee Kang <claudiajkang@gmail.com> Signed-off-by:
-
Juhee Kang authored
All pktgen samples can send indefinitely num messages per thread by setting the count option to 0(-n 0). If running sample with setting count 0 and press Ctrl-C to stop this program, the program prints the result of the execution so far. Currently, the samples besides sample{3...5} don't work properly. Because Ctrl-C stops the script, not just pktgen. This is results of samples: # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample04_many_flows.sh -n 0 Running... ctrl^C to stop ^CDevice: eth0@0 Result: OK: 569657(c569538+d118) usec, 84650 (60byte,0frags) 148597pps 71Mb/sec (71326560bps) errors: 0 # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample01_simple.sh -n 0 Running... ctrl^C to stop ^C In order to solve this, this commit adds trap SIGINT. Also, this commit changes control_c function to print_result to maintain consistency with other samples. Signed-off-by: -
Juhee Kang authored
Currently, most pktgen samples print the execution result when the program is terminated normally. However, sample03 doesn't work appropriately. This is results of samples: # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample04_many_flows.sh -n 1 Running... ctrl^C to stop Device: eth0@0 Result: OK: 19(c5+d13) usec, 1 (60byte,0frags) 51762pps 24Mb/sec (24845760bps) errors: 0 # DEV=eth0 DEST_IP=10.1.0.1 DST_MAC=00:11:22:33:44:55 ./pktgen_sample03_burst_single_flow.sh -n 1 Running... ctrl^C to stop The reason why it doesn't print the execution result when the program is terminated usually is that sample03 doesn't call the function which prints the result, unlike other samples. So, this commit solves this issue by calling the function before termination. Also, this commit changes control_c function to print_result to maintain consistency with other samples. Signed-off-by: -
David S. Miller authored
Sunil Goutham says: ==================== Octeontx2: Traffic shaping and SDP link config support This patch series adds support for traffic shaping configuration on all silicons available after 96xx C0. And also adds SDP link related configuration needed when Octeon is connected as an end-point and traffic needs to flow from end-point to host and vice versa. Series also has other changes like - New mbox messages in admin function driver for PF/VF drivers to retrieve available HW resource count. HW resources like block LFs, bandwidth profiles etc are covered. - Added PTP device ID for new CN10K and 95O silicons. - etc ==================== Signed-off-by: -
Sunil Goutham authored
Added mbox for PF/VF drivers to retrieve current ingress bandwidth profile free count. Also added current policer timeunit configuration info based on which ratelimiting decisions can be taken by PF/VF drivers. Signed-off-by:
Sunil Goutham <sgoutham@marvell.com> Signed-off-by:
-
Sunil Goutham authored
New usecases are popping up where in user wants to install common MCAM filters for all interfaces. Having channel verification will result in duplicating such MCAM filters for each of the ingress interface. Hence removed channel verification. Signed-off-by:
-
Subbaraya Sundeep authored
CN10K slicon has different device id for PTP device. Hence this patch updates the driver with new id. Though ptp driver being a separate driver AF manages configuring PTP block by all PFs. To manage ptp, AF driver checks in its probe whether 1. ptp hardware device found on silicon 2. A driver is bound to ptp device 3. The ptp driver probe is successful In failure of cases 1 and 3, AF proceeds with out ptp and for case 2 defers the probe. This patch refactors code also to check for all the PTP device ids given in ptp device ids table for case 1. Also added PTP device ID for 95O silicon Signed-off-by:
Subbaraya Sundeep <sbhatta@marvell.com> Signed-off-by:
-
George Cherian authored
Upon receiving the MBOX_MSG_FREE_RSRC_CNT, the AF will find out the current number of free resources and reply it back to the requester. No guarantee is given on the future state of the free resources yet. If another requester sends MBOX_MSG_ATTACH_RESOURCES after this call, the number of available resources might change. Signed-off-by:
George Cherian <george.cherian@marvell.com> Signed-off-by:
Stanislaw Kardach <skardach@marvell.com> Signed-off-by:
-
Radha Mohan Chintakuntla authored
Added support for packet IO via SDK links which is used when Octeon is connected as a end-point. Traffic host to end-point and vice versa flow through SDP links. This patch also support dual SDP blocks supported in 98xx silicon. Signed-off-by:
Radha Mohan Chintakuntla <radhac@marvell.com> Signed-off-by:
Nalla Pradeep <pnalla@marvell.com> Signed-off-by:
Subrahmanyam Nilla <snilla@marvell.com> Signed-off-by:
-
Harman Kalra authored
In 98xx, there are 2 NIX blocks and 4 LBK blocks present. The way these NIX-LBK should be configured depends on the use case. By default loopback functionality is supported in AF VF pairs which are attached to NIX0 and NIX1 LFs alternatively to ensure load balancing. NIX0 transmits a packet to LBK1 which will be received by NIX1 and packet transmitted by NIX1 will get received by NIX0 via LBK2. There are some requirements where only one AF VF is used and respective NIX is expected to operate in a mode where it can receive it own packet back. This can be achieved if NIX0 sends packet to LBK0 and not LBK1. Adding a flag in LF alloc request mailbox which can setup NIX0 to use LBK0 and NIX1 can use LBK3. Signed-off-by:
Harman Kalra <hkalra@marvell.com> Signed-off-by:
-
Subbaraya Sundeep authored
Unlike OcteonTx2, the channel numbers used by CGX/RPM and LBK on CN10K silicons aren't fixed in HW. They are SW programmable, hence we cannot derive transmit link from static channel numbers anymore. Get the same from admin function via mailbox. Signed-off-by:
-
Jerin Jacob authored
Before C0 HW revision, The RSS adder was computed based the following static formula. rss_adder<7:0> = flow_tag<7:0> ^ flow_tag<15:8> ^ flow_tag<23:16> ^ flow_tag<31:24> The above scheme has the following drawbacks: 1) It is not in line with other standard NIC behavior. 2) There can be an SW use case where SW can compute the hash upfront using Toeplitz function and predict the queue selection to optimize some packet lookup function. The nonstandard way of doing XOR makes the consumer to not predict the queue selection. C0 HW revision onwards, The HW can configure the rss_adder<7:0> as flow_tag<7:0> to align with standard NICs. This patch adds an option to select legacy RSS adder mode vs standard NIC behavior by setting NIX_LF_RSS_TAG_LSB_AS_ADDER flag. Since this bit field is used as reserved in old HW revisions, No need to have an additional HW version check. Signed-off-by:
Jerin Jacob <jerinj@marvell.com> Signed-off-by:
-
Nithin Dabilpuram authored
Starting from 96xx C0 onwards all silicons support traffic shaping. This patch enables that feature along with other changes - When PIR/CIR shaping config is modified, toggle SW_XOFF for config to take effect - Before SMQ flush, clear SW_XOFF at all parent schedulers - Support to read current transmit scheduler configuration via mbox Signed-off-by:
Nithin Dabilpuram <ndabilpuram@marvell.com> Signed-off-by:
Geetha sowjanya <gakula@marvell.com> Signed-off-by:
-
Cai Huoqing authored
it's helpful for complie test in other platform(e.g.X86) Signed-off-by:
Cai Huoqing <caihuoqing@baidu.com> Signed-off-by:
-
Nithin Dabilpuram authored
NIX_AF_TX_LINKX_NORM_CREDIT holds running counter of tx credits available per link. But, tx credits should be configured based on MTU config. So MTU change needs tx credit count update. An issue exists whereby when both PF & VF are enabled and PF traffic is flowing, if VF requests for MTU update, updating the NORM_CREDIT register will lead to corruption of credit count and subsequent deadlock of tx link as the NORM_CREDIT register holds running count. This patch provides workaround by pausing link traffic using NIX_AF_TL1X_SW_XOFF, waiting for existing packets to drain, and used credits be returned before updating new credit count. Signed-off-by:
-
Nithin Dabilpuram authored
Clear and disable interrupt before queueing work as there might be a chance that work gets completed on other core faster and interrupt enable as a part of the work completes before interrupt disable in the interrupt context. This leads to permanent disable of interrupt. Signed-off-by:
-
Geetha sowjanya authored
Set NPA batch allocation engine to process 35 cache lines per turn on CN10k platform. Signed-off-by:
-
Shaokun Zhang authored
Function 'mctp_dev_get_rtnl' is declared twice, so remove the repeated declaration. Cc: Jeremy Kerr <jk@codeconstruct.com.au> Cc: Matt Johnston <matt@codeconstruct.com.au> Cc: "David S. Miller" <davem@davemloft.net> Cc: Jakub Kicinski <kuba@kernel.org> Signed-off-by:
Shaokun Zhang <zhangshaokun@hisilicon.com> Signed-off-by:
-
David S. Miller authored
Merge tag 'linux-can-next-for-5.15-20210825' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next Marc Kleine says: ==================== pull-request: can-next 2021-08-25 this is a pull request of 4 patches for net-next/master. The first patch is by Cai Huoqing, and enables COMPILE_TEST for the rcar CAN drivers. Lad Prabhakar contributes a patch for the rcar_canfd driver, fixing a redundant assignment. The last 2 patches are by Tang Bin, target the mscan driver, and clean up the driver by converting it to of_device_get_match_data() and removing a useless BUG_ON. ==================== Signed-off-by:
-
David S. Miller authored
Biju Das says: ==================== Add Factorisation code to support Gigabit Ethernet driver The DMAC and EMAC blocks of Gigabit Ethernet IP found on RZ/G2L SoC are similar to the R-Car Ethernet AVB IP. The Gigabit Ethernet IP consists of Ethernet controller (E-MAC), Internal TCP/IP Offload Engine (TOE) and Dedicated Direct memory access controller (DMAC). With a few changes in the driver we can support both IPs. This patch series aims to add factorisation code to support RZ/G2L SoC, hardware feature bits for gPTP feature, Multiple irq feature and optional reset support. Ref:- * https://lore.kernel.org/linux-renesas-soc/TYCPR01MB59334319695607A2683C1A5E86E59@TYCPR01MB5933.jpnprd01.prod.outlook.com/T/#t ==================== Signed-off-by:
-
Biju Das authored
Reset support is present on R-Car. Let's support it, if it is available. Signed-off-by:
Biju Das <biju.das.jz@bp.renesas.com> Reviewed-by:
Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> Signed-off-by:
-
Biju Das authored
The E-MAC IP on the R-Car AVB module has different initialization parameters for RX frame size, duplex settings, different offset for transfer speed setting and has magic packet detection support compared to E-MAC on RZ/G2L Gigabit Ethernet module. Factorise the ravb_emac_init function to support the later SoC. Signed-off-by:
-
Biju Das authored
The DMAC IP on the R-Car AVB module has different initialization parameters for RCR, TGC, TCCR, RIC0, RIC2, and TIC compared to DMAC IP on the RZ/G2L Gigabit Ethernet module. Factorise the ravb_dmac_init function to support the later SoC. Signed-off-by:
-
Biju Das authored
RZ/G2L supports HW checksum on RX and TX whereas R-Car supports on RX. Factorise ravb_set_features to support this feature. Signed-off-by:
-
Biju Das authored
R-Car supports 100 and 1000 Mbps transfer speed whereas RZ/G2L in addition support 10Mbps. Factorise ravb_adjust_link function in order to support 10Mbps speed. Signed-off-by:
-
Biju Das authored
R-Car uses an extended descriptor in RX whereas, RZ/G2L uses normal descriptor in RX. Factorise the ravb_rx function to support the later SoC. Signed-off-by:
-
Biju Das authored
The ravb_ring_init function uses an extended descriptor in RX for R-Car and normal descriptor for RZ/G2L. Add a helper function for RX ring buffer allocation to support later SoC. Signed-off-by:
-
Biju Das authored
The ravb_ring_format function uses an extended descriptor in RX for R-Car compared to the normal descriptor for RZ/G2L. Factorise RX ring buffer buildup to extend the support for later SoC. Signed-off-by:
-
Biju Das authored
R-Car uses extended descriptor in RX, whereas RZ/G2L uses normal descriptor. Factorise ravb_ring_free function so that it can support later SoC. Signed-off-by:
-
Biju Das authored
There are some H/W differences for the gPTP feature between R-Car Gen3, R-Car Gen2, and RZ/G2L as below. 1) On R-Car Gen3, gPTP support is active in config mode. 2) On R-Car Gen2, gPTP support is not active in config mode. 3) RZ/G2L does not support the gPTP feature. Add a ptp_cfg_active hw feature bit to struct ravb_hw_info for supporting gPTP active in config mode for R-Car Gen3. This patch also removes enum ravb_chip_id, chip_id from both struct ravb_hw_info and struct ravb_private, as it is unused. Signed-off-by:
-
Biju Das authored
There are some H/W differences for the gPTP feature between R-Car Gen3, R-Car Gen2, and RZ/G2L as below. 1) On R-Car Gen2, gPTP support is not active in config mode. 2) On R-Car Gen3, gPTP support is active in config mode. 3) RZ/G2L does not support the gPTP feature. Add a no_ptp_cfg_active hw feature bit to struct ravb_hw_info for handling gPTP for R-Car Gen2. Signed-off-by:
-
Biju Das authored
R-Car Gen3 supports separate interrupts for E-MAC and DMA queues, whereas R-Car Gen2 and RZ/G2L have a single interrupt instead. Add a multi_irq hw feature bit to struct ravb_hw_info to enable this only for R-Car Gen3. Signed-off-by:
-
Biju Das authored
For addressing 4 bytes alignment restriction on transmission buffer for R-Car Gen2 we use 2 descriptors whereas it is a single descriptor for other cases. Replace the macros NUM_TX_DESC_GEN[23] with magic number and add a comment to explain it. Signed-off-by:
Geert Uytterhoeven <geert+renesas@glider.be> Reviewed-by:
-
David S. Miller authored
Vladimir Oltean says: ==================== Make sja1105 treat tag_8021q VLANs more like real DSA tags This series solves a nuisance with the sja1105 driver, which is that non-DSA tagged packets sent directly by the DSA master would still exit the switch just fine. We also had an issue for packets coming from the outside world with a crafted DSA tag, the switch would not reject that tag but think it was valid. ====================
-
Vladimir Oltean authored
Introduced in commit 38b5beea ("net: dsa: sja1105: prepare tagger for handling DSA tags and VLAN simultaneously"), the sja1105_xmit_tpid function solved quite a different problem than our needs are now. Then, we used best-effort VLAN filtering and we were using the xmit_tpid to tunnel packets coming from an 8021q upper through the TX VLAN allocated by tag_8021q to that egress port. The need for a different VLAN protocol depending on switch revision came from the fact that this in itself was more of a hack to trick the hardware into accepting tunneled VLANs in the first place. Right now, we deny 8021q uppers (see sja1105_prechangeupper). Even if we supported them again, we would not do that using the same method of {tunneling the VLAN on egress, retagging the VLAN on ingress} that we had in the best-effort VLAN filtering mode. It seems rather simpler that we just allocate a VLAN in the VLAN table that is simply not used by the bridge at all, or by any other port. Anyway, I have 2 gripes with the current sja1105_xmit_tpid: 1. When sending packets on behalf of a VLAN-aware bridge (with the new TX forwarding offload framework) plus untagged (with the tag_8021q VLAN added by the tagger) packets, we can see that on SJA1105P/Q/R/S and later (which have a qinq_tpid of ETH_P_8021AD), some packets sent through the DSA master have a VLAN protocol of 0x8100 and others of 0x88a8. This is strange and there is no reason for it now. If we have a bridge and are therefore forced to send using that bridge's TPID, we can as well blend with that bridge's VLAN protocol for all packets. 2. The sja1105_xmit_tpid introduces a dependency on the sja1105 driver, because it looks inside dp->priv. It is desirable to keep as much separation between taggers and switch drivers as possible. Now it doesn't do that anymore. Signed-off-by:
Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by:
-
Vladimir Oltean authored
The sja1105 driver is a bit special in its use of VLAN headers as DSA tags. This is because in VLAN-aware mode, the VLAN headers use an actual TPID of 0x8100, which is understood even by the DSA master as an actual VLAN header. Furthermore, control packets such as PTP and STP are transmitted with no VLAN header as a DSA tag, because, depending on switch generation, there are ways to steer these control packets towards a precise egress port other than VLAN tags. Transmitting control packets as untagged means leaving a door open for traffic in general to be transmitted as untagged from the DSA master, and for it to traverse the switch and exit a random switch port according to the FDB lookup. This behavior is a bit out of line with other DSA drivers which have native support for DSA tagging. There, it is to be expected that the switch only accepts DSA-tagged packets on its CPU port, dropping everything that does not match this pattern. We perhaps rely a bit too much on the switches' hardware dropping on the CPU port, and place no other restrictions in the kernel data path to avoid that. For example, sja1105 is also a bit special in that STP/PTP packets are transmitted using "management routes" (sja1105_port_deferred_xmit): when sending a link-local packet from the CPU, we must first write a SPI message to the switch to tell it to expect a packet towards multicast MAC DA 01-80-c2-00-00-0e, and to route it towards port 3 when it gets it. This entry expires as soon as it matches a packet received by the switch, and it needs to be reinstalled for the next packet etc. All in all quite a ghetto mechanism, but it is all that the sja1105 switches offer for injecting a control packet. The driver takes a mutex for serializing control packets and making the pairs of SPI writes of a management route and its associated skb atomic, but to be honest, a mutex is only relevant as long as all parties agree to take it. With the DSA design, it is possible to open an AF_PACKET socket on the DSA master net device, and blast packets towards 01-80-c2-00-00-0e, and whatever locking the DSA switch driver might use, it all goes kaput because management routes installed by the driver will match skbs sent by the DSA master, and not skbs generated by the driver itself. So they will end up being routed on the wrong port. So through the lens of that, maybe it would make sense to avoid that from happening by doing something in the network stack, like: introduce a new bit in struct sk_buff, like xmit_from_dsa. Then, somewhere around dev_hard_start_xmit(), introduce the following check: if (netdev_uses_dsa(dev) && !skb->xmit_from_dsa) kfree_skb(skb); Ok, maybe that is a bit drastic, but that would at least prevent a bunch of problems. For example, right now, even though the majority of DSA switches drop packets without DSA tags sent by the DSA master (and therefore the majority of garbage that user space daemons like avahi and udhcpcd and friends create), it is still conceivable that an aggressive user space program can open an AF_PACKET socket and inject a spoofed DSA tag directly on the DSA master. We have no protection against that; the packet will be understood by the switch and be routed wherever user space says. Furthermore: there are some DSA switches where we even have register access over Ethernet, using DSA tags. So even user space drivers are possible in this way. This is a huge hole. However, the biggest thing that bothers me is that udhcpcd attempts to ask for an IP address on all interfaces by default, and with sja1105, it will attempt to get a valid IP address on both the DSA master as well as on sja1105 switch ports themselves. So with IP addresses in the same subnet on multiple interfaces, the routing table will be messed up and the system will be unusable for traffic until it is configured manually to not ask for an IP address on the DSA master itself. It turns out that it is possible to avoid that in the sja1105 driver, at least very superficially, by requesting the switch to drop VLAN-untagged packets on the CPU port. With the exception of control packets, all traffic originated from tag_sja1105.c is already VLAN-tagged, so only STP and PTP packets need to be converted. For that, we need to uphold the equivalence between an untagged and a pvid-tagged packet, and to remember that the CPU port of sja1105 uses a pvid of 4095. Now that we drop untagged traffic on the CPU port, non-aggressive user space applications like udhcpcd stop bothering us, and sja1105 effectively becomes just as vulnerable to the aggressive kind of user space programs as other DSA switches are (ok, users can also create 8021q uppers on top of the DSA master in the case of sja1105, but in future patches we can easily deny that, but it still doesn't change the fact that VLAN-tagged packets can still be injected over raw sockets). Signed-off-by:
-
Vladimir Oltean authored
Currently it is possible for an attacker to craft packets with a fake DSA tag and send them to us, and our user ports will accept them and preserve that VLAN when transmitting towards the CPU. Then the tagger will be misled into thinking that the packets came on a different port than they really came on. Up until recently there wasn't a good option to prevent this from happening. In SJA1105P and later, the MAC Configuration Table introduced two options called: - DRPSITAG: Drop Single Inner Tagged Frames - DRPSOTAG: Drop Single Outer Tagged Frames Because the sja1105 driver classifies all VLANs as "outer VLANs" (S-Tags), it would be in principle possible to enable the DRPSOTAG bit on ports using tag_8021q, and drop on ingress all packets which have a VLAN tag. When the switch is VLAN-unaware, this works, because it uses a custom TPID of 0xdadb, so any "tagged" packets received on a user port are probably a spoofing attempt. But when the switch overall is VLAN-aware, and some ports are standalone (therefore they use tag_8021q), the TPID is 0x8100, and the port can receive a mix of untagged and VLAN-tagged packets. The untagged ones will be classified to the tag_8021q pvid, and the tagged ones to the VLAN ID from the packet header. Yes, it is true that since commit 4fbc08bd ("net: dsa: sja1105: deny 8021q uppers on ports") we no longer support this mixed mode, but that is a temporary limitation which will eventually be lifted. It would be nice to not introduce one more restriction via DRPSOTAG, which would make the standalone ports of a VLAN-aware switch drop genuinely VLAN-tagged packets. Also, the DRPSOTAG bit is not available on the first generation of switches (SJA1105E, SJA1105T). So since one of the key features of this driver is compatibility across switch generations, this makes it an even less desirable approach. The breakthrough comes from commit bef0746c ("net: dsa: sja1105: make sure untagged packets are dropped on ingress ports with no pvid"), where it became obvious that untagged packets are not dropped even if the ingress port is not in the VMEMB_PORT vector of that port's pvid. However, VLAN-tagged packets are subject to VLAN ingress checking/dropping. This means that instead of using the catch-all DRPSOTAG bit introduced in SJA1105P, we can drop tagged packets on a per-VLAN basis, and this is already compatible with SJA1105E/T. This patch adds an "allowed_ingress" argument to sja1105_vlan_add(), and we call it with "false" for tag_8021q VLANs on user ports. The tag_8021q VLANs still need to be allowed, of course, on ingress to DSA ports and CPU ports. We also need to refine the drop_untagged check in sja1105_commit_pvid to make it not freak out about this new configuration. Currently it will try to keep the configuration consistent between untagged and pvid-tagged packets, so if the pvid of a port is 1 but VLAN 1 is not in VMEMB_PORT, packets tagged with VID 1 will behave the same as untagged packets, and be dropped. This behavior is what we want for ports under a VLAN-aware bridge, but for the ports with a tag_8021q pvid, we want untagged packets to be accepted, but packets tagged with a header recognized by the switch as a tag_8021q VLAN to be dropped. So only restrict the drop_untagged check to apply to the bridge_pvid, not to the tag_8021q_pvid. Signed-off-by:
-
