- 08 Dec, 2016 40 commits
-
-
Martin KaFai Lau authored
The XDP prog checks if the incoming packet matches any VIP:PORT combination in the BPF hashmap. If it is, it will encapsulate the packet with a IPv4/v6 header as instructed by the value of the BPF hashmap and then XDP_TX it out. The VIP:PORT -> IP-Encap-Info can be specified by the cmd args of the user prog. Acked-by:
Alexei Starovoitov <ast@kernel.org> Signed-off-by:
Martin KaFai Lau <kafai@fb.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Martin KaFai Lau authored
Reserve XDP_PACKET_HEADROOM for packet and enable bpf_xdp_adjust_head() support. This patch only affects the code path when XDP is active. After testing, the tx_dropped counter is incremented if the xdp_prog sends more than wire MTU. Signed-off-by:
Saeed Mahameed <saeedm@mellanox.com> Signed-off-by:
-
Martin KaFai Lau authored
When XDP is active in mlx4, mlx4 is using one page/pkt. At the same time (i.e. when XDP is active), it is currently limiting MTU to be FRAG_SZ0 - ETH_HLEN - (2 * VLAN_HLEN) which is 1514 in x86. AFAICT, we can at least raise the MTU limit up to PAGE_SIZE - ETH_HLEN - (2 * VLAN_HLEN) which this patch is doing. It will be useful in the next patch which allows XDP program to extend the packet by adding new header(s). Note: In the earlier XDP patches, there is already existing guard to ensure the page/pkt scheme only applies when XDP is active in mlx4. Signed-off-by:
-
Martin KaFai Lau authored
This patch allows XDP prog to extend/remove the packet data at the head (like adding or removing header). It is done by adding a new XDP helper bpf_xdp_adjust_head(). It also renames bpf_helper_changes_skb_data() to bpf_helper_changes_pkt_data() to better reflect that XDP prog does not work on skb. This patch adds one "xdp_adjust_head" bit to bpf_prog for the XDP-capable driver to check if the XDP prog requires bpf_xdp_adjust_head() support. The driver can then decide to error out during XDP_SETUP_PROG. Signed-off-by:
Daniel Borkmann <daniel@iogearbox.net> Acked-by:
John Fastabend <john.r.fastabend@intel.com> Signed-off-by:
-
David S. Miller authored
Woojung Huh says: ==================== phy: lan78xx: add phy fixup unregister functions & LAN7801 update V2 patch of adding phy fixup unregister function with use in LAN7801 update. ==================== Signed-off-by: -
Woojung Huh authored
Add LAN7801 MAC only support with phy fixup functions. Signed-off-by:
Woojung Huh <woojung.huh@microchip.com> Signed-off-by:
-
Woojung.Huh@microchip.com authored
>From : Woojung Huh <woojung.huh@microchip.com> Add functions to unregister phy fixup for modules. int phy_unregister_fixup(const char *bus_id, u32 phy_uid, u32 phy_uid_mask) Unregister phy fixup from phy_fixup_list per bus_id, phy_uid & phy_uid_mask int phy_unregister_fixup_for_uid(u32 phy_uid, u32 phy_uid_mask) Unregister phy fixup from phy_fixup_list. Use it for fixup registered by phy_register_fixup_for_uid() int phy_unregister_fixup_for_id(const char *bus_id) Unregister phy fixup from phy_fixup_list. Use it for fixup registered by phy_register_fixup_for_id() Signed-off-by:
-
Alexei Starovoitov authored
Commmits 57a09bf0 ("bpf: Detect identical PTR_TO_MAP_VALUE_OR_NULL registers") and 48461135 ("bpf: allow access into map value arrays") by themselves are correct, but in combination they make state equivalence ignore 'id' field of the register state which can lead to accepting invalid program. Fixes: 57a09bf0 ("bpf: Detect identical PTR_TO_MAP_VALUE_OR_NULL registers") Fixes: 48461135 ("bpf: allow access into map value arrays") Signed-off-by:
Thomas Graf <tgraf@suug.ch> Signed-off-by:
-
Eric Dumazet authored
sk_drops can be an often written field, do not read it unless application showed interest. Note that sk_drops can be read via inet_diag, so applications can avoid getting this info from every received packet. In the future, 'reading' sk_drops might require folding per node or per cpu fields, and thus become even more expensive than today. Signed-off-by:
Eric Dumazet <edumazet@google.com> Cc: Paolo Abeni <pabeni@redhat.com> Signed-off-by:
-
Eric Dumazet authored
Under UDP flood, many softirq producers try to add packets to UDP receive queue, and one user thread is burning one cpu trying to dequeue packets as fast as possible. Two parts of the per packet cost are : - copying payload from kernel space to user space, - freeing memory pieces associated with skb. If socket is under pressure, softirq handler(s) can try to pull in skb->head the payload of the packet if it fits. Meaning the softirq handler(s) can free/reuse the page fragment immediately, instead of letting udp_recvmsg() do this hundreds of usec later, possibly from another node. Additional gains : - We reduce skb->truesize and thus can store more packets per SO_RCVBUF - We avoid cache line misses at copyout() time and consume_skb() time, and avoid one put_page() with potential alien freeing on NUMA hosts. This comes at the cost of a copy, bounded to available tail room, which is usually small. (We might have to fix GRO_MAX_HEAD which looks bigger than necessary) This patch gave me about 5 % increase in throughput in my tests. skb_condense() helper could probably used in other contexts. Signed-off-by:
-
David S. Miller authored
Raghu Vatsavayi says: ==================== liquidio VF data path Following is V3 patch series that adds support for VF data path related features. It also has following changes related to previous comments: 1) Remove unnecessary "void *" casting. 2) Remove inline for functions and let gcc decide. Please apply patches in following order as some of them depend on earlier patches. ==================== Signed-off-by: -
Raghu Vatsavayi authored
Adds support for VF receive data control path. Signed-off-by:
Raghu Vatsavayi <raghu.vatsavayi@caviumnetworks.com> Signed-off-by:
Derek Chickles <derek.chickles@caviumnetworks.com> Signed-off-by:
Satanand Burla <satananda.burla@caviumnetworks.com> Signed-off-by:
Felix Manlunas <felix.manlunas@caviumnetworks.com> Signed-off-by:
-
Raghu Vatsavayi authored
Adds support for freeing VF xmit buffers. Signed-off-by:
-
Raghu Vatsavayi authored
Adds support for transmit functionality in VF. Signed-off-by:
-
Raghu Vatsavayi authored
Adds support for VF scatter gather lists. Signed-off-by:
-
Raghu Vatsavayi authored
Adds support for configuring mtu, multicast and mac address. Signed-off-by:
-
Raghu Vatsavayi authored
Adds support for VF link status related changes. Signed-off-by:
-
Raghu Vatsavayi authored
Adds support for VF link initialization and offload features. Signed-off-by:
-
Eric Dumazet authored
RFS is not commonly used, so add a jump label to avoid some conditionals in fast path. Signed-off-by:
-
Zhang Shengju authored
Remove first zero for add, and use full attribute name in comments. Signed-off-by:
Zhang Shengju <zhangshengju@cmss.chinamobile.com> Signed-off-by:
-
David S. Miller authored
Niklas Cassel says: ==================== net: stmmac: make DMA programmable burst length more configurable Make DMA programmable burst length more configurable in the stmmac driver. This is done by adding support for independent pbl for tx/rx through DT. More fine grained tuning of pbl is possible thanks to a DT property saying that we should NOT multiply pbl values by x8/x4 in hardware. All new DT properties are optional, and created in a way that it will not affect any existing DT configurations. Changes since V1: Created cover-letter. Rebased patch set against next-20161205, since conflicting patches to stmmac_platform.c has been merged since V1. Changes since V2: Moved default value initialization of pbl to stmmac_platform.c and added a check for pbl != 0 in stmmac_main.c, to catch a possble pbl == 0 from pci glue. ==================== Signed-off-by: -
Niklas Cassel authored
The driver currently always sets the PBLx8/PBLx4 bit, which means that the pbl values configured via the pbl/txpbl/rxpbl DT properties are always multiplied by 8/4 in the hardware. In order to allow the DT to configure lower pbl values, while at the same time not changing behavior of any existing device trees using the pbl/txpbl/rxpbl settings, add a property to disable the multiplication of the pbl by 8/4 in the hardware. Suggested-by:
Rabin Vincent <rabinv@axis.com> Signed-off-by:
Niklas Cassel <niklas.cassel@axis.com> Acked-by:
Alexandre Torgue <alexandre.torgue@st.com> Signed-off-by:
-
Niklas Cassel authored
GMAC and newer supports independent programmable burst lengths for DMA tx/rx. Add new optional devicetree properties representing this. To be backwards compatible, snps,pbl will still be valid, but snps,txpbl/snps,rxpbl will override the value in snps,pbl if set. If the IP is synthesized to use the AXI interface, there is a register and a matching DT property inside the optional stmmac-axi-config DT node for controlling burst lengths, named snps,blen. However, using this register, it is not possible to control tx and rx independently. Also, this register is not available if the IP was synthesized with, e.g., the AHB interface. Signed-off-by:
-
Niklas Cassel authored
DMA_BUS_MODE_RPBL_MASK is really 6 bits, just like DMA_BUS_MODE_PBL_MASK. Signed-off-by:
-
Niklas Cassel authored
commit 64c3b252 ("net: stmmac: fixed the pbl setting with DT") changed the parsing of the DT binding. Before 64c3b252, snps,fixed-burst and snps,mixed-burst were parsed regardless if the property snps,pbl existed or not. After the commit, fixed burst and mixed burst are only parsed if snps,pbl exists. Now when snps,aal has been added, it too is only parsed if snps,pbl exists. Since the DT binding does not specify that fixed burst, mixed burst or aal depend on snps,pbl being specified, undo changes introduced by 64c3b252. The issue commit 64c3b252 ("net: stmmac: fixed the pbl setting with DT") tries to address is solved in another way: The databook specifies that all values other than 1, 2, 4, 8, 16, or 32 results in undefined behavior, so snps,pbl = <0> is invalid. If pbl is 0 after parsing, set pbl to DEFAULT_DMA_PBL. This handles the case where the property is omitted, and also handles the case where the property is specified without any data. Signed-off-by:
-
Niklas Cassel authored
Use struct stmmac_dma_cfg *dma_cfg as an argument rather than using all the struct members as individual arguments. Signed-off-by:
-
Niklas Cassel authored
All drivers except pci glue layer calls stmmac_probe_config_dt. stmmac_probe_config_dt does a kzalloc dma_cfg. pci glue layer does kzalloc dma_cfg explicitly, so all current drivers does a kzalloc dma_cfg. Return an error if no DMA configuration is found, that way we can assume that the DMA configuration always exists. Signed-off-by:
-
Tobias Klauser authored
Use module_pci_driver() to get rid of some boilerplate code. Signed-off-by:
Tobias Klauser <tklauser@distanz.ch> Acked-by:
Lino Sanfilippo <LinoSanfilippo@gmx.de> Signed-off-by:
-
David S. Miller authored
Simon Horman says: ==================== net/sched: cls_flower: Support matching on ICMP this series adds support for matching on ICMP type and code to cls_flower. Changes v5->v6: * Restore missing signed-off-by Changes v4->v5: * Drop all helpers Changes v3->v4: * Do not add icmp to struct flow_keys, it is not needed * Do not test for ICMP protocols in packet in __skb_flow_dissect, this is also not needed Changes v2->v3: * Add FLOW_DISSECTOR_KEY_ICMP and use separate structure for ICMP Changes v1->v2: * Include all dissector helpers in first patch ==================== Signed-off-by: -
Simon Horman authored
Support matching on ICMP type and code. Example usage: tc qdisc add dev eth0 ingress tc filter add dev eth0 protocol ip parent ffff: flower \ indev eth0 ip_proto icmp type 8 code 0 action drop tc filter add dev eth0 protocol ipv6 parent ffff: flower \ indev eth0 ip_proto icmpv6 type 128 code 0 action drop Signed-off-by:
Simon Horman <simon.horman@netronome.com> Acked-by:
Jiri Pirko <jiri@mellanox.com> Signed-off-by:
-
Simon Horman authored
Allow dissection of ICMP(V6) type and code. This should only occur if a packet is ICMP(V6) and the dissector has FLOW_DISSECTOR_KEY_ICMP set. There are currently no users of FLOW_DISSECTOR_KEY_ICMP. A follow-up patch will allow FLOW_DISSECTOR_KEY_ICMP to be used by the flower classifier. Signed-off-by:
-
Niklas Cassel authored
devicetree binding for stmmac states: - compatible: Should be "snps,dwmac-<ip_version>", "snps,dwmac" For backwards compatibility: "st,spear600-gmac" is also supported. Previously, when specifying "snps,dwmac-4.10a", "snps,dwmac" as your compatible string, plat_stmmacenet_data would have both has_gmac and has_gmac4 set. This would lead to stmmac_hw_init calling dwmac1000_setup rather than dwmac4_setup, resulting in a non-functional driver. This happened since the check for has_gmac is done before the check for has_gmac4. However, the order should not matter, so it does not make sense to have both set. If something is valid for both, you should do as the stmmac_interrupt does: if (priv->plat->has_gmac || priv->plat->has_gmac4) ... The places where it was obvious that the author actually meant if (has_gmac || has_gmac4) rather than if (has_gmac) has been updated. Signed-off-by:
-
Niklas Cassel authored
devicetree binding for stmmac states: - compatible: Should be "snps,dwmac-<ip_version>", "snps,dwmac" For backwards compatibility: "st,spear600-gmac" is also supported. Since dwmac-generic.c calls stmmac_probe_config_dt explicitly, another alternative would have been to remove all compatible strings other than "snps,dwmac" and "st,spear600-gmac" from dwmac-generic.c. However, that would probably do more good than harm, since when trying to figure out what hardware a certain driver supports, you usually look at the compatible strings in the struct of_device_id, and not in some function defined in a completely different file. No functional change intended. Signed-off-by:
Giuseppe Cavallaro <peppe.cavallaro@st.com> Signed-off-by:
-
Niklas Cassel authored
snps,tso was previously placed under AXI BUS Mode parameters, suggesting that the property should be in the stmmac-axi-config node. TSO (TCP Segmentation Offloading) has nothing to do with AXI BUS Mode parameters, and the parser actually expects it to be in the root node, not in the stmmac-axi-config. Also added a note about snps,tso only being available on GMAC4 and newer. Signed-off-by:
Rob Herring <robh@kernel.org> Acked-by:
-
Tobias Klauser authored
Do not open code getting the MAC address exclusively from the "local-mac-address" property, but instead use of_get_mac_address() which looks up the MAC address using the 3 typical property names. Also avoid casting away the const qualifier of the return value by making temac_init_mac_address() take a const void* address. Follows commit b34296a9 ("net: ethoc: Utilize of_get_mac_address()"). Cc: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
-
Tobias Klauser authored
Do not open code getting the MAC address exclusively from the "local-mac-address" property, but instead use of_get_mac_address() which looks up the MAC address using the 3 typical property names. Also avoid casting away the const qualifier of the return value by making axienet_set_mac_address() take a const void* address. Follows commit b34296a9 ("net: ethoc: Utilize of_get_mac_address()"). Cc: Florian Fainelli <f.fainelli@gmail.com> Signed-off-by:
-
David S. Miller authored
Or Gerlitz says: ==================== net/sched: cls_flower: Add support for matching on dissection flags This series add the UAPI to provide set of flags for matching, where the flags provided from user-space are mapped to flow-dissector flags. The 1st flag allows to match on whether the packet is an IP fragment and corresponds to the FLOW_DIS_IS_FRAGMENT flag. v2->v3: - replace BIT() with << (kbuild test robot) v1->v2: - dropped the flow dissector patch (#1) as no changes are needed there (Jiri) - applied code review comments from Jiri to the flower patch ==================== Signed-off-by: -
Or Gerlitz authored
Enable offloading of matching on packets being fragments. Signed-off-by:
Or Gerlitz <ogerlitz@mellanox.com> Reviewed-by:
Paul Blakey <paulb@mellanox.com> Reviewed-by:
-
Or Gerlitz authored
Add UAPI to provide set of flags for matching, where the flags provided from user-space are mapped to flow-dissector flags. The 1st flag allows to match on whether the packet is an IP fragment and corresponds to the FLOW_DIS_IS_FRAGMENT flag. Signed-off-by:
-
Dan Carpenter authored
These two statements were not indented correctly so it's sort of confusing. Signed-off-by:
Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by:
-
