- 20 Dec, 2017 35 commits
-
-
Julian Wiedmann authored
The L3 packet descriptor's 'dest_addr' field is used for a different purpose in RX descriptors. Clean up the hard-coded byte accesses and try to be more self-documenting. Signed-off-by:
Julian Wiedmann <jwi@linux.vnet.ibm.com> Signed-off-by:
David S. Miller <davem@davemloft.net>
-
Julian Wiedmann authored
When 1. an skb has no neighbour, and 2. skb->protocol is not IP[V6], we select the skb's cast type based on its destination MAC address. The multicast check is currently restricted to Multicast IP-mapped MACs. Extend it to also cover non-IP Multicast MACs. Signed-off-by:
-
Julian Wiedmann authored
Use the proper helpers to check for multicast IP addressing, and remove some ancient Token Ring code. Signed-off-by:
-
Julian Wiedmann authored
Instead of assuming that skb->data points to the Ethernet header, use the right helper and struct to access the Ethertype field. Signed-off-by:
-
Julian Wiedmann authored
Once all of qeth_l3_set_rx_mode()'s single-use helpers are folded back in, the two implementations actually look quite similar. So improve the readability by converting both set_rx_mode() routines to a common format. This also allows us to walk ip_mc_htable just once, instead of three times. Signed-off-by:
-
Julian Wiedmann authored
Be a little more self-documenting, and get rid of OSA_ADDR_LEN. Signed-off-by:
-
Julian Wiedmann authored
For adding/removing a MAC address, use just one helper each that handles both unicast and multicast. Saves one level of indirection for multicast addresses, while improving the error reporting for unicast addresses. Signed-off-by:
-
Julian Wiedmann authored
Instead of tracking the uc/mc state in each MAC address object, just check the multicast bit in the address itself. Signed-off-by:
-
Julian Wiedmann authored
commit "s390/qeth: use ip*_eth_mc_map helpers" removed the last occurrence of CONFIG_IPV6-dependent code. Signed-off-by:
-
Julian Wiedmann authored
Get rid of some wrapper indirection, and stop accessing the skb at hard-coded offsets. Signed-off-by:
-
Elena Reshetova authored
atomic_t variables are currently used to implement reference counters with the following properties: - counter is initialized to 1 using atomic_set() - a resource is freed upon counter reaching zero - once counter reaches zero, its further increments aren't allowed - counter schema uses basic atomic operations (set, inc, inc_not_zero, dec_and_test, etc.) Such atomic variables should be converted to a newly provided refcount_t type and API that prevents accidental counter overflows and underflows. This is important since overflows and underflows can lead to use-after-free situation and be exploitable. The variable qeth_reply.refcnt is used as pure reference counter. Convert it to refcount_t and fix up the operations. Suggested-by:
Kees Cook <keescook@chromium.org> Reviewed-by:
David Windsor <dwindsor@gmail.com> Reviewed-by:
Hans Liljestrand <ishkamiel@gmail.com> Signed-off-by:
Elena Reshetova <elena.reshetova@intel.com> [jwi: removed the WARN_ONs. Use CONFIG_REFCOUNT_FULL if you care.] Signed-off-by:
-
Elena Reshetova authored
atomic_t variables are currently used to implement reference counters with the following properties: - counter is initialized to 1 using atomic_set() - a resource is freed upon counter reaching zero - once counter reaches zero, its further increments aren't allowed - counter schema uses basic atomic operations (set, inc, inc_not_zero, dec_and_test, etc.) Such atomic variables should be converted to a newly provided refcount_t type and API that prevents accidental counter overflows and underflows. This is important since overflows and underflows can lead to use-after-free situation and be exploitable. The variable lcs_reply.refcnt is used as pure reference counter. Convert it to refcount_t and fix up the operations. Suggested-by:
-
git://git.open-mesh.org/linux-mergeDavid S. Miller authored
Simon Wunderlich says: ==================== This feature/cleanup patchset includes the following patches: - bump version strings, by Simon Wunderlich - de-inline hash functions to save memory footprint, by Denys Vlasenko - Add License information to various files, by Sven Eckelmann (3 patches) - Change batman_adv.h from ISC to MIT, by Sven Eckelmann - Improve various includes, by Sven Eckelmann (5 patches) - Lots of kernel-doc work by Sven Eckelmann (8 patches) ==================== Signed-off-by: -
David S. Miller authored
Yafang Shao says: ==================== replace tcp_set_state tracepoint with inet_sock_set_state According to the discussion in the mail thread https://patchwork.kernel.org/patch/10099243/, tcp_set_state tracepoint is renamed to inet_sock_set_state tracepoint and is moved to include/trace/events/sock.h. With this new tracepoint, we can trace AF_INET/AF_INET6 sock state transitions. As there's only one single tracepoint for inet, so I didn't create a new trace file named trace/events/inet_sock.h, and just place it in include/trace/events/sock.h Currently TCP/DCCP/SCTP state transitions are traced with this tracepoint. - Why not more protocol ? If we really think that anonter protocol should be traced, I will modify the code to trace it. I just want to make the code easy and not output useless information. ==================== Signed-off-by:
-
Yafang Shao authored
With changes in inet_ files, SCTP state transitions are traced with inet_sock_set_state tracepoint. As SCTP state names, i.e. SCTP_SS_CLOSED, SCTP_SS_ESTABLISHED, have the same value with TCP state names. So the output info still print the TCP state names, that makes the code easy. Signed-off-by:
Yafang Shao <laoar.shao@gmail.com> Signed-off-by:
-
Yafang Shao authored
With changes in inet_ files, DCCP state transitions are traced with inet_sock_set_state tracepoint. Signed-off-by:
-
Yafang Shao authored
sk_state_load is only used by AF_INET/AF_INET6, so rename it to inet_sk_state_load and move it into inet_sock.h. sk_state_store is removed as it is not used any more. Signed-off-by:
-
Yafang Shao authored
As sk_state is a common field for struct sock, so the state transition tracepoint should not be a TCP specific feature. Currently it traces all AF_INET state transition, so I rename this tracepoint to inet_sock_set_state tracepoint with some minor changes and move it into trace/events/sock.h. We dont need to create a file named trace/events/inet_sock.h for this one single tracepoint. Two helpers are introduced to trace sk_state transition - void inet_sk_state_store(struct sock *sk, int newstate); - void inet_sk_set_state(struct sock *sk, int state); As trace header should not be included in other header files, so they are defined in sock.c. The protocol such as SCTP maybe compiled as a ko, hence export inet_sk_set_state(). Signed-off-by: -
Steven Rostedt (VMware) authored
The TCP trace events (specifically tcp_set_state), maps emums to symbol names via __print_symbolic(). But this only works for reading trace events from the tracefs trace files. If perf or trace-cmd were to record these events, the event format file does not convert the enum names into numbers, and you get something like: __print_symbolic(REC->oldstate, { TCP_ESTABLISHED, "TCP_ESTABLISHED" }, { TCP_SYN_SENT, "TCP_SYN_SENT" }, { TCP_SYN_RECV, "TCP_SYN_RECV" }, { TCP_FIN_WAIT1, "TCP_FIN_WAIT1" }, { TCP_FIN_WAIT2, "TCP_FIN_WAIT2" }, { TCP_TIME_WAIT, "TCP_TIME_WAIT" }, { TCP_CLOSE, "TCP_CLOSE" }, { TCP_CLOSE_WAIT, "TCP_CLOSE_WAIT" }, { TCP_LAST_ACK, "TCP_LAST_ACK" }, { TCP_LISTEN, "TCP_LISTEN" }, { TCP_CLOSING, "TCP_CLOSING" }, { TCP_NEW_SYN_RECV, "TCP_NEW_SYN_RECV" }) Where trace-cmd and perf do not know the values of those enums. Use the TRACE_DEFINE_ENUM() macros that will have the trace events convert the enum strings into their values at system boot. This will allow perf and trace-cmd to see actual numbers and not enums: __print_symbolic(REC->oldstate, { 1, "TCP_ESTABLISHED" }, { 2, "TCP_SYN_SENT" }, { 3, "TCP_SYN_RECV" }, { 4, "TCP_FIN_WAIT1" }, { 5, "TCP_FIN_WAIT2" }, { 6, "TCP_TIME_WAIT" }, { 7, "TCP_CLOSE" }, { 8, "TCP_CLOSE_WAIT" }, { 9, "TCP_LAST_ACK" }, { 10, "TCP_LISTEN" }, { 11, "TCP_CLOSING" }, { 12, "TCP_NEW_SYN_RECV" }) Signed-off-by:Steven Rostedt (VMware) <rostedt@goodmis.org> Acked-by:
Song Liu <songliubraving@fb.com> Signed-off-by:
-
Prashant Bhole authored
- Checking return value with IS_ERROR_OR_NULL - Added error handling where it was not handled Signed-off-by:
Prashant Bhole <bhole_prashant_q7@lab.ntt.co.jp> Acked-by:
Jakub Kicinski <jakub.kicinski@netronome.com> Signed-off-by:
-
Haishuang Yan authored
If md is NULL, tun_dst must be freed, otherwise it will cause memory leak. Fixes: ef7baf5e ("ip6_gre: add ip6 erspan collect_md mode") Cc: William Tu <u9012063@gmail.com> Signed-off-by:
Haishuang Yan <yanhaishuang@cmss.chinamobile.com> Signed-off-by:
-
Haishuang Yan authored
If md is NULL, tun_dst must be freed, otherwise it will cause memory leak. Fixes: 1a66a836 ("gre: add collect_md mode to ERSPAN tunnel") Cc: William Tu <u9012063@gmail.com> Signed-off-by:
-
Haishuang Yan authored
Same as ipv4 code, when ip6erspan_rcv call return PACKET_REJECT, we should call icmpv6_send to send icmp unreachable message in error path. Fixes: 5a963eb6 ("ip6_gre: Add ERSPAN native tunnel support") Acked-by:
William Tu <u9012063@gmail.com> Cc: William Tu <u9012063@gmail.com> Signed-off-by:
-
Haishuang Yan authored
When erspan_rcv call return PACKET_REJECT, we shoudn't call ipgre_rcv to process packets again, instead send icmp unreachable message in error path. Fixes: 84e54fe0 ("gre: introduce native tunnel support for ERSPAN") Acked-by:
-
Haishuang Yan authored
pskb_may_pull() can change skb->data, so we need to load ipv6h/ershdr at the right place. Fixes: 5a963eb6 ("ip6_gre: Add ERSPAN native tunnel support") Cc: William Tu <u9012063@gmail.com> Acked-by:
-
Andy Shevchenko authored
Get rid of yet another custom hex_dump_to_buffer(). The output is slightly changed, i.e. each byte followed by white space. Note, we don't use print_hex_dump() here since the original code uses nedev_dbg(). Acked-by:
Tom Lendacky <thomas.lendacky@amd.com> Signed-off-by:
Andy Shevchenko <andriy.shevchenko@linux.intel.com> Signed-off-by:
-
Michael Chan authored
Reported-by:
Marcelo Ricardo Leitner <marcelo.leitner@gmail.com> Signed-off-by:
Michael Chan <michael.chan@broadcom.com> Signed-off-by:
-
David S. Miller authored
Jakub Kicinski says: ==================== netdevsim: couple of build warning fixes This series fixes two harmless build warning about a symbol which should be static and an unused variable. ==================== Signed-off-by: -
Jakub Kicinski authored
skip_sw is set but no longer used. Signed-off-by:
Quentin Monnet <quentin.monnet@netronome.com> Signed-off-by:
-
Jakub Kicinski authored
struct device_type nsim_dev_type created for SR-IOV support should be static. Signed-off-by:
-
William Tu authored
Add test cases for gretap and ip6gretap, native mode and external (collect metadata) mode. Signed-off-by:
-
Andy Shevchenko authored
Replace sscanf() with mac_pton(). Signed-off-by:
-
Andy Shevchenko authored
Replace sscanf() with mac_pton(). Signed-off-by:
Andy Gospodarek <andy@greyhouse.net> Signed-off-by:
-
Andy Shevchenko authored
Use %pM to print MAC mac_pton() to convert it from ASCII to binary format, and ether_addr_copy() to copy. Signed-off-by:
Stephen Hemminger <stephen@networkplumber.org> Signed-off-by:
-
Alexey Kodanev authored
LTP/udp6_ipsec_vti tests fail when sending large UDP datagrams over ip6_vti that require fragmentation and the underlying device has an MTU smaller than 1500 plus some extra space for headers. This happens because ip6_vti, by default, sets MTU to ETH_DATA_LEN and not updating it depending on a destination address or link parameter. Further attempts to send UDP packets may succeed because pmtu gets updated on ICMPV6_PKT_TOOBIG in vti6_err(). In case the lower device has larger MTU size, e.g. 9000, ip6_vti works but not using the possible maximum size, output packets have 1500 limit. The above cases require manual MTU setup after ip6_vti creation. However ip_vti already updates MTU based on lower device with ip_tunnel_bind_dev(). Here is the example when the lower device MTU is set to 9000: # ip a sh ltp_ns_veth2 ltp_ns_veth2@if7: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 9000 ... inet 10.0.0.2/24 scope global ltp_ns_veth2 inet6 fd00::2/64 scope global # ip li add vti6 type vti6 local fd00::2 remote fd00::1 # ip li show vti6 vti6@NONE: <POINTOPOINT,NOARP> mtu 1500 ... link/tunnel6 fd00::2 peer fd00::1 After the patch: # ip li add vti6 type vti6 local fd00::2 remote fd00::1 # ip li show vti6 vti6@NONE: <POINTOPOINT,NOARP> mtu 8832 ... link/tunnel6 fd00::2 peer fd00::1 Reported-by:Petr Vorel <pvorel@suse.cz> Signed-off-by:
Alexey Kodanev <alexey.kodanev@oracle.com> Signed-off-by:
-
- 19 Dec, 2017 5 commits
-
-
David S. Miller authored
Simon Horman says: ==================== nfp: flower: add Geneve tunnel support John Hurley says: This patchset adds support for offloading the encap and decap of Geneve tunnels to the NFP. In both cases, specifying well known port 6081 is a requirement for rule offload. Geneve firmware support has been recently added, so the patchset includes the reading of a fw symbol that defines a bitmap of newly supported features. Geneve will only be offloaded if the fw supports it. The new symbol is added in fw r5646. Geneve option fields are not supported as either a match or an action due there current exclussion from TC flower. Because Geneve (as both a match and action) behaves the same as other udp tunnels such as VXLAN, generic functions are created that handle both Geneve and VXLAN. It is anticapated that these functions will be modified to support options in future patches. The removal of an unused variable 'tun_dst_mask' is included as a separate patch here. This does not affect functionality. Also included are modifications to the test framework to check that the new encap and decap features are functioning correctly. ==================== Signed-off-by: -
John Hurley authored
Generate rules for the NFP to encapsulate packets in Geneve tunnels. Move the vxlan action code to generic udp tunnel actions and use core code for both vxlan and Geneve. Only support outputting to well known port 6081. Setting tunnel options is not supported yet. Only attempt to offload if the fw supports Geneve. Signed-off-by:
John Hurley <john.hurley@netronome.com> Reviewed-by:
Simon Horman <simon.horman@netronome.com> Signed-off-by:
-
John Hurley authored
Compile Geneve match fields for offloading to the NFP. The addition of Geneve overflows the 8 bit key_layer field, so apply extended metadata to the match cmsg allowing up to 32 more key_layer fields. Rather than adding new Geneve blocks, move the vxlan code to generic ipv4 udp tunnel structs and use these for both vxlan and Geneve. Matches are only supported when specifically mentioning well known port 6081. Geneve tunnel options are not yet included in the match. Only offload Geneve if the fw supports it - include check for this. Signed-off-by:
-
John Hurley authored
Extract the _abi_flower_extra_features symbol from the fw which gives a 64 bit bitmap of new features (on top of the flower base support) that the fw can offload. Store this bitmap in the priv data associated with each app. If the symbol does not exist, set the bitmap to 0. Signed-off-by:
-
John Hurley authored
The tunnel dest IP is required for separate offload to the NFP. It is already verified that a dest IP must be present and must be an exact match in the flower rule. Therefore, we can just extract the IP from the generated offload rule and remove the unused mask variable. The function is then no longer required to return the IP separately. Because tun_dst is localised to tunnel matches, move the declaration to the tunnel if branch. Signed-off-by:
-
