summaryrefslogtreecommitdiffstats
AgeCommit message (Collapse)AuthorFilesLines
2017-10-20Thread_index for thread0 in vlib_global_main not initialisedKeith Burns (alagalah)1-0/+1
Change-Id: Ie49ee865b197e8fe7bba170c115a4ccbf1013e5f Signed-off-by: Keith Burns (alagalah) <alagalah@gmail.com>
2017-10-20VCL: Fix converity warning CID 177864Dave Wallace1-2/+2
Change-Id: Id8578321381d14f9de827767ef0acf627f1535e4 Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-20VCL-LDPRELOAD: Refactor vcom_socket* and fix crash in vppcom_selectDave Wallace3-174/+190
- filter verbose debug output with VCOM_DEBUG > 2 - clean up nomenclature, renaming vppcom_*() functions to vcom_session_*() - fix vppcom_select crash with NULL maps. Change-Id: I6e416a096d6fd800aa26991c2439e24e8fc38cc5 Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-20Add reverse DNS (ip to name) resolutionDave Barach7-38/+402
Change-Id: Ic531d820b1846ff7363e5c396ac0b1176e87b401 Signed-off-by: Dave Barach <dave@barachs.net>
2017-10-20null-terminate some formatted stringGabriel Ganne4-0/+19
Any u8* variable created by format() is NOT null-terminated. Add the null terminating byte with vec_terminate_c_string(). If that variable is used by (at least) hash_get_mem(), then it needs to be null-terminated, as it will go through string_key_sum() which makes a call to strlen. Change-Id: I4e51e1b6668f557e53af3bb897cd281598eedbc0 Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
2017-10-20add cuckoo hashKlement Sekera8-1/+2483
Change-Id: I78215041588014e9e5c3599c60471ced610735bb Signed-off-by: Klement Sekera <ksekera@cisco.com>
2017-10-19LD_PRELOAD poll implementationshrinivasan ganapathy7-4/+748
- add poll implementation - implementation based on vppcom_poll - implementation based on vppcom_select - currently vppcom_select implementation is picked Change-Id: If6c2862ae72e9969335aca5b8085957c98287dc0 Signed-off-by: shrinivasan ganapathy <shrinivasanganapathy@gmail.com> Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-19VCL: cleanup misc. issues in vppcomDave Wallace1-6/+10
- Fix vppcom_select crash when n_bits == 0 - Enhance debug output - Remove port byte-swapping during accept Change-Id: I6ccd1040ceb82908d924220f558df803ab5eea30 Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-19vcl-ldpreload: test script enhancementsDave Wallace4-0/+59
Change-Id: Iff302a35cd5d35bf0671a63e8ba81a0c60a231eb Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-19make test: control sanity via SANITY=yes|no optionKlement Sekera1-2/+11
Allow skipping sanity when using DEBUG=gdb/gdbserver options to speed up development process by supplying SANITY=no option to make test. Change-Id: I6e619229f6866b22bdeb75654b4e750232ae22bd Signed-off-by: Klement Sekera <ksekera@cisco.com>
2017-10-19Add GENEVE tunnel maintainer.Marco Varlese1-0/+4
Change-Id: I30bbba4119accee2c157fb5b8202da3c87c59805 Signed-off-by: Marco Varlese <marco.varlese@suse.com>
2017-10-19Remove unused 'not_last' parameter from ip_add_del_routeMarek Gradzki4-17/+2
Vat supports setting value for the parameter, but 'not_last' is ignored by ip_add_del_route handler, so can be removed. This patch - updates ip.api, - removes vat handlers - updates vpp_papi_provider.py (also mpls_route_add_del with unused not_last) Change-Id: Ife15de123db4bc8247103a29b90bce1988e46534 Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
2017-10-19VPP-1024: rewrite buffer trajectory tracerDave Barach7-50/+161
Use a proper u16 * vector to capture node indices, since vpp w/ plugins now exceeds 255 graph nodes Change-Id: Ic48cad676fa3a6116413ddf08c083dd9660783f1 Signed-off-by: Dave Barach <dave@barachs.net>
2017-10-19VCL-LDPRELOAD: fix iperf3 socket_test.shDave Wallace3-50/+114
Change-Id: Ib6b52917af717d3341429163fb9ecc903cf717fb Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-18TEST,SPAN:add broadcast+QiniQ testsEyal Bari1-67/+137
Change-Id: I89c6923370af97bbd09db81a33b6d17f9424a049 Signed-off-by: Eyal Bari <ebari@cisco.com>
2017-10-18LISP: fix crash when sending NSH map-request message, ONE-32Filip Tehlar1-0/+3
Change-Id: Ief8c3d3bec116e9f884981fb52af528f98b5f6ff Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
2017-10-18Plugin for IP-Address to Interface PuntingPierre Pfister7-0/+872
This plugin provides per-ip address to interface punting. When at least one rule is defined, the plugin receives all packets which destination is one of VPP's address but which was not processed by VPP (e.g., a TCP packet on a port that is not open, or a packet for a protocol which is not attached). Based on the set of configured rules, the destination address of each packet is used to send the packet on the associated interface. This plugin allows multiple containers to use VPP's TCP stack (or other features provided by VPP) while still being able to receive additional packets. Change-Id: I3e69bb7d98183bf5163cb9ecb564cb482de252ce Signed-off-by: Pierre Pfister <ppfister@cisco.com>
2017-10-18VPP-930: add missing ntohl to send_ip6_fib_details(...)Dave Barach1-1/+1
Change-Id: I0c1671f3eaf2dad084e3ac9fb124c9ed78273f50 Signed-off-by: Dave Barach <dave@barachs.net>
2017-10-18ipsec: use boolean or vs. bitwise or to avoid compiler errorAndrew Yourtchenko2-6/+6
Ubuntu 17.04, gcc version 6.3.0 20170406 (Ubuntu 6.3.0-12ubuntu2), "make build" fails with the few of the errors below: error: suggest parentheses around comparison in operand of ‘|’ [-Werror=parentheses] is_aead = (sa0->crypto_alg == IPSEC_CRYPTO_ALG_AES_GCM_128 | Solution: use the logical rather than the bitwise or. Change-Id: Iffcc1ed2e68b14b248159cb117593d32c623c553 Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
2017-10-18CSIT-844: fix binary api rx pthread heap push/popDave Barach4-7/+41
We need to push/pop the rx pthread's heap without affecting other thread(s). Search clib_per_cpu_mheaps, locate an unused slot. Duplicate the main thread heap pointer in that slot, and set __os_thread_index appropriately. Miscellaneous cleanups. Print exec_inband results as a vector, instead of as a format string. Don't bail out of vpp_api_test with results pending, e.g. at the end of a vpp_api_test script. Even though vpp will eventuallly garbage-collect them, We don't want to leave allocated reply messages lurking in the api message allocation rings... Change-Id: I0e8a25d1ff0d3700249dc330d079db16c2fcbc55 Signed-off-by: Dave Barach <dave@barachs.net>
2017-10-17session: fix connect corner case crash.Dave Wallace1-17/+20
Change-Id: I54ced42749432335183ee3085a9ccc5f95a87ae9 Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-17VPP-1032: fix coverity warningsDave Barach3-14/+25
Change-Id: Ic3a0c51e5408921051deaf2e50372d9654574b27 Signed-off-by: Dave Barach <dave@barachs.net>
2017-10-17Initial commit for phase 2, Add some simple validation.John DeNisco4-8/+235
Change-Id: I5b1d5600cdef4b05cc7c2f1cddb60aed2cc49ac2 Signed-off-by: John DeNisco <jdenisco@cisco.com>
2017-10-17TEST,SPAN: add rx+tx span testEyal Bari1-127/+78
also refactored verify function Change-Id: Ib93c99c833dd7c75e416374530c631f954957d98 Signed-off-by: Eyal Bari <ebari@cisco.com>
2017-10-17VCL_LDPRELOAD: Add unit test scripts.Dave Wallace5-0/+160
Change-Id: I5fbb3339a6dbd7b40fecf66bfda3fd3249b4a90c Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-17VCL: refactor VCL_LDPRELOAD env vars.Dave Wallace2-6/+4
- Update vcl-ldpreload references to use $WS_ROOT. Change-Id: I8e79f7c5e330eda6548a2baf84613b0a4d360811 Signed-off-by: Dave Wallace <dwallacelf@gmail.com>
2017-10-17Add API versioning to GENEVE tunnel implementation.Marco Varlese1-1/+3
Change-Id: I5974e5545d74af53c27938c8cdbae12745c38a54 Signed-off-by: Marco Varlese <marco.varlese@suse.com>
2017-10-17session: return local transport endpoint in connect replyFlorin Coras4-1/+20
Change-Id: I7794d5a0774017da4c1c15f45783a18754994ac8 Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-10-17Update CSIT tests 171002 -> 171017Jan Gelety1-1/+1
- update of CSIT operational branch to be used for VPP-patch test Change-Id: Ib098170bbcdbbb01c25ef197d9181cfd826d2854 Signed-off-by: Jan Gelety <jgelety@cisco.com>
2017-10-16session: fix build when debugging is onFlorin Coras3-3/+3
Change-Id: I193832a0cd0557fffc034a1223f67fa64a4d45ae Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-10-16udp: refactor udp codeFlorin Coras40-1243/+1941
Change-Id: I44d5c9df7c49b8d4d5677c6d319033b2da3e6b80 Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-10-16DNS name resolver improvementsDave Barach4-31/+217
- Cache intermediate CNAME records - Bug fixes Change-Id: I06dcb558212fc5e9434281493c872577cf9b83e1 Signed-off-by: Dave Barach <dave@barachs.net>
2017-10-16LDPRELOAD: Remove hack code in getpeernameSteven1-14/+0
Now the real implementation of getpeername is there, we remove the hack version of getpeername in LDPRELOAD. Change-Id: I8fa2ba93d346f78d04129a87ee17a834cba537da Signed-off-by: Steven <sluong@cisco.com>
2017-10-16LISP: fix map-request counters, ONE-25Filip Tehlar1-6/+5
Change-Id: I198f58a84c4692408f9205052af24ee22df7aeaa Signed-off-by: Filip Tehlar <ftehlar@cisco.com>
2017-10-16jvpp: stop plugin intialization on first mismatchMarek Gradzki9-12/+15
Java bindings use get_message_id from jvpp-common to detect if messages known at compile time are avaliable at runtime. In case of missing entry, Java exception is propagated via JNI using (*env)->ThrowNew. But this function does not end code execution so, in order to prevent unexpected behaviour (e.g. calling vl_msg_api_set_handlers with id == 0), get_message_id caller should do it manually. Change-Id: I2edb5013fd3658dcdd77a867b5cdf62e559ee071 Signed-off-by: Marek Gradzki <mgradzki@cisco.com>
2017-10-16lb plugin - fix format() type mismatchesGabriel Ganne1-2/+3
* (vip - lbm->vips) is u64; change format from [%u] to [%lu] * vip->plen is u8, but format looks for u32; add exlicit cast (this cast was done implicitely) On ARM platforms, these prevent a loop in the second call to format_white_space() which would get an invalid (huge) indent value; the result *looked like* an infinite loop. Change-Id: I675ef2f98e4ba3d9e8aef12022d38b1d22981da8 Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
2017-10-15VCL/LDPRELOAD: getpeername and getsocknameSteven2-42/+39
Add support for getpeername and getsockname system calls Change-Id: Ie22787b967bb2a5fead0f5fcffd779e4f39b3302 Signed-off-by: Steven <sluong@cisco.com>
2017-10-15Fix minor issues in clib_spinlock_unlock()Dave Barach1-1/+3
Change-Id: I20ce799c9dd57332c06003b466ee7c36169bce98 Signed-off-by: Dave Barach <dave@barachs.net>
2017-10-15ip: avoid arp assert if out of buffers (VPP-1030)Florin Coras1-0/+4
Change-Id: Ia31b978c6c1619c3e0075a84fcbbb6ccbf1c0076 Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-10-15Revert "Enforce FIB table creation before use"Florin Coras8-31/+101
This reverts commit f9342023c19887da656133e2688a90d70383b0c5. Reverting to unblock master. No idea why jjb +1ed this patch! On closer inspection it looks like it -1ed it and subsequently changed opinion. CSIT tests should be fixed before re-merging. Change-Id: I26608912a962c52083073e16c7c9d2cc44a3cc8d Signed-off-by: Florin Coras <fcoras@cisco.com>
2017-10-14NAT: fix delete of sessions for 1:1 NAT if 1 worker (VPP-1023)Matus Fabian1-1/+1
Change-Id: I2446c646de7f227f9438dd7ef93a455ba5af0102 Signed-off-by: Matus Fabian <matfabia@cisco.com>
2017-10-14change format_get_indent() to use u32 instead of uwordGabriel Ganne6-7/+8
This follows commit d3c008d108aa2187d1a2afe2833b4de25ca2c2ab by Christophe Fontaine. Change-Id: I0c4df40df44be2ac0ab25817fa050a1f619eca4d Signed-off-by: Gabriel Ganne <gabriel.ganne@enea.com>
2017-10-14Fix bug with temporary directory when building rpms.Thomas F Herbert1-1/+1
Fixes bug introduced in commit 5349f94d. JIRA: VPP-1014 Change-Id: Ia18f4c6f5f1124306cce790a36f6de970d186687 Signed-off-by: Thomas F Herbert <therbert@redhat.com> (cherry picked from commit 8c5bfbcc09f44f3ed2b87b0b5c257fbb7b735f97)
2017-10-14Deps added to Makefile must also be in spec file.Thomas F Herbert2-5/+7
Deps are required for Fedora too. This patch is merged in version stable/17.10. It should also be merged in master because it fixes breaking rpm builds in some circumstances. JIRA: VPP-1015 Cherry-picked from stable/1710 Change-Id: I10807069742cdd6b09a0f34d9d05e9cae4146ec3 Signed-off-by: Thomas F Herbert <therbert@redhat.com>
2017-10-14Add API call to set keys on IPsec tunnel intfMatthew Smith3-0/+152
There was already a CLI command and a libvnet function to set the keys on an existing IPsec tunnel interface. Expose this via the API. Change-Id: I226a9616f680fc022f04447177a2e8232690657f Signed-off-by: Matthew Smith <mgsmith@netgate.com>
2017-10-14Source Lookup progammable via APINeale Ranns12-13/+201
Change-Id: I5d5d4f22b6369d504455a644f73076d772fbcfb4 Signed-off-by: Neale Ranns <nranns@cisco.com>
2017-10-14vhost: crash under heavy traffic condition due to memory corruption (VPP-1016)Steven1-2/+33
With heavy traffic, tx code path may crash due to memory corruption Thread 5 "vpp_wk_2" received signal SIGSEGV, Segmentation fault. [Switching to Thread 0x7fff3995c700 (LWP 2505)] 0x00007ffff73675e8 in vhost_user_if_input (vm=0x7fffb5f5bf9c, vum=0x7ffff7882a40 <vhost_user_main>, vui=0x7fffb65570c4, qid=0, node=0x7fffb6577dac, mode=VNET_HW_INTERFACE_RX_MODE_POLLING) at /home/sluong/vpp-master/vpp/build-data/../src/vnet/devices/virtio/vhost-user.c:1610 1610 bi_current = (vum->cpus[thread_index].rx_buffers) [vum->cpus[thread_index].rx_buffers_len]; (gdb) p vum->cpus[thread_index].rx_buffers_len $2 = 793212607 (gdb) Apparently, some code accidentally wrote the bad value in rx_buffers_len. rx_buffers_len should never be greater than 1024 since that is how many buffers we request each time. After debugging many hours, I discovered that the memory corruption happens in the tx code path right here on line 2176. { vhost_copy_t *cpy = &vum->cpus[thread_index].copy[copy_len]; copy_len++; cpy->len = bytes_left; cpy->len = (cpy->len > buffer_len) ? buffer_len : cpy->len; cpy->dst = buffer_map_addr; cpy->src = (uword) vlib_buffer_get_current (current_b0) + current_b0->current_length - bytes_left; (gdb) p cpy $3 = (vhost_copy_t *) 0x7fffb554077c (gdb) p copy_len $4 = 1025 (gdb) p &vum->cpus[3].rx_buffers_len $8 = (u32 *) 0x7fffb5540784 copy_len is picking up the index entry 1024 before it was incremented. copy array has only 1024 members (0 - 1023 are valid). The assignment here in cpy surely causes memory corruption. It is only discovered later when the memory location that it corrupted is used. The condition for the crash is to transmit jumbo frames under heavy volume. Since ring size is 1024, with one packet taking up one index for frame size (less 2048), it does not cause overflow. With jumbo frames, it requires multiple indices for one packet, it can cause the overflow under heavy traffic. The fix is to do copy out when we have 1000 entries in the array to avoid overflow. Change-Id: Iefbc739b8e80470f1cf13123113f8331ffcd0eb2 Signed-off-by: Steven <sluong@cisco.com>
2017-10-14plugins/dpdk: align memory to avoid potential segfault and false sharingGeorgina Sheehan1-1/+1
Made Update to src/plugins/dpdk/buffer.c Change-Id: I87bb8f38974a7be274c1b1d205f5513e7d068e48 Signed-off-by: Georgina <georgina.sheehan@intel.com>
2017-10-14Use AESNI=y only on x86_64 machinesBrian Brooks1-0/+6
This fixes the native build on arm64 machines. Change-Id: I89bff01beedb4c8e26ac55fab2dd1ed39754bf26 Signed-off-by: Brian Brooks <brian.brooks@arm.com> Reviewed-by: Ola Liljedahl <ola.liljedahl@arm.com>
2017-10-14VPP-1029: Don't call clib_longjmp(...) directly from the SIGTERM handlerDave Barach2-4/+8
It's way too easy to imagine leaving a mutex or a spin-lock held in the /vpe-api shared-memory segment, or elsewhere. Set a volatile variable and check it in a safe place... Change-Id: I9d91c38cffeb921143c272162d055c9c24a6c312 Signed-off-by: Dave Barach <dave@barachs.net>