summaryrefslogtreecommitdiffstats
path: root/src/vlib
AgeCommit message (Collapse)AuthorFilesLines
2018-08-24Rename struct mallinfo -> struct dlmallinfoDave Barach1-1/+1
Also: don't #include /usr/include/malloc.h in dlmalloc.h Change-Id: Ic73ff8862cc8aba371488b912255e28dd96374ff Signed-off-by: Dave Barach <dave@barachs.net>
2018-08-17CMake as an alternative to autotools (experimental)Damjan Marion1-0/+98
Change-Id: Ibc59323e849810531dd0963e85493efad3b86857 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-08-14Run interior graph nodes before process nodesDave Barach1-14/+7
Reduces the chance of tripping over vectors in flight, especially in single-core cases. Change-Id: I132cdd3689f8e634f9a983af57219503817b8560 Signed-off-by: Dave Barach <dave@barachs.net>
2018-08-13Multiarch handling in different constructor macrosDamjan Marion3-0/+46
This significantly reduces need for ... in multiarch code. Simply constructor macros will jost create static unused entry if CLIB_MARCH_VARIANT is defined and that will be optimized out by compiler. Change-Id: I17d1c4ac0c903adcfadaa4a07de1b854c7ab14ac Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-08-03vlib: avoid double process dispatchFlorin Coras1-1/+9
Change-Id: I46467b1f149be9dfbd00e3ea6d60681d19acf235 Signed-off-by: Florin Coras <fcoras@cisco.com> Signed-off-by: Andrew Yourtchenko <ayourtch@gmail.com>
2018-08-01Improve cpu { coremask-% } configure optionYi He1-20/+4
Accept any sized hexadecimal bitmask specification to support platforms with hundreds of cores. Change-Id: Ib881db0cf60f78bdeffa13acfc2fc7fe7e128cc4 Signed-off-by: Yi He <yi.he@arm.com>
2018-07-27128-bit SIMD version of vlib_get_buffersDamjan Marion1-0/+8
Change-Id: I1a28ddf535c80ecf4ba4bf31659ff2fead1d8a64 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-07-23Buffer clone: copy opaque2Neale Ranns1-0/+2
Change-Id: I06e1dbfa0782bc03f2d0e77295b38b4fde7df35e Signed-off-by: Neale Ranns <nranns@cisco.com>
2018-07-20Add congestion drop in interface handoffDamjan Marion3-21/+57
This prevents deadlock in case when worker A sends to B and worker B sends to A Change-Id: Id9436960f932c58325fe4f5ef8ec67b50031aeda Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-07-20physmem: fix alloc when dlmaloc is usedDamjan Marion1-1/+5
Change-Id: I4b6cdf95ac694c8408eb6cf3259b335a8148b5aa Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-07-19Cleanup of handoff codeDamjan Marion1-0/+86
- removed handoff-dispatch node - removed some unused buffer metadata fields - enqueue to thread logic moved to inline function Change-Id: I7361e1d88f8cce74cd4fcec90d172eade1855cbd Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-07-18Add config option to use dlmalloc instead of mheapDave Barach6-18/+98
Configure w/ --enable-dlmalloc, see .../build-data/platforms/vpp.mk src/vppinfra/dlmalloc.[ch] are slightly modified versions of the well-known Doug Lea malloc. Main advantage: dlmalloc mspaces have no inherent size limit. Change-Id: I19b3f43f3c65bcfb82c1a265a97922d01912446e Signed-off-by: Dave Barach <dave@barachs.net>
2018-07-12Fix debug CLI node recycling bugsDave Barach1-5/+26
When creating a new - as opposed to recycled - debug CLI process node, perform a proper barrier sync and node runtime update. Otherwise, the graph replicas diverge for some period of time. That's not immediately fatal, but it's not a good idea, either. When renaming a debug cli process node, fix all of the name-vector replicas before freeing the [one-and-only] name vector. This fixes the so-called stats segment node runtime scraper crash, which tripped over a replicated dangling reference to the recently-freed debug CLI node name. Change-Id: Ieffabd9f003139e534b9d79b88370439907930e5 Signed-off-by: Dave Barach <dbarach@cisco.com>
2018-07-11threads: fix issue with setting main-core to 0Damjan Marion2-8/+15
Change-Id: Id3ccfcfa2a88cf7aa106f1202af7cd677de32575 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-07-04Do not override skip-cores if set in startup.confDamjan Marion1-4/+2
Change-Id: I97d75f58a46275fdad5add27dc59c731cc3ac64d Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-07-03Set main thread affinity before main heap is allocatedDamjan Marion1-7/+7
Change-Id: I524909570fc1736f51fd437d6d30566c461139bd Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-07-02Add per-numa page allocation info to 'show memory'Damjan Marion1-2/+7
Change-Id: I64e4e3d68c0f3958323f30b12a26cfaafa8bad85 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-06-30cli: Validate cli_file_index on quit commandSteve Shin1-0/+4
VPP crash happens with 'vppctl quit quit' command. The 2nd quit command tries to access the file index which is already freed by the first quit. This can be avoided to validate cli_file_index. Change-Id: I880514c93523db2a727d7510c97950582cd6a6c8 Signed-off-by: Steve Shin <jonshin@cisco.com>
2018-06-27gcc8 and Wstringop-truncationMarco Varlese1-2/+4
gcc8 introduced a new warning (Wstringop-truncation) which in our case is being treated as error. Disabling the warning globally might introduce bugs related to string truncation which are not desired by the developer (e.g. bug). Instead, this patch disables the warning only for those occurences which have been verified to be non-bugs but the desired behaviour as per developer will. Change-Id: I0f04ff6b4fad44061e80a65af633fd7e0148a0c5 Signed-off-by: Marco Varlese <marco.varlese@suse.com>
2018-06-26add backtrace in unix_signal_handlerKingwel Xie1-9/+49
crash stack backtrace will be directed to syslog 1. make use of glic backtrace in execinfo.h. the old clib_backtrace is removed 2. install SIGABRT in signal handler, but have to remove it when backtrace is done. reason is to capture stack trace caused by SIGABRT. vPP ASSERT always call os_exit then abort(). we definitely want to know the trace of this situation. It is a little tricky to avoid SIGABRT infinite loop 3. always load symbols by calling clib_elf_main_init () in main(). Otherwise, PC addresses instead of symbols will be displayed. Change-Id: I150e15b94a4620b2ea4f08c73dc3e6ad1856de1e Signed-off-by: Kingwel Xie <kingwel.xie@ericsson.com>
2018-06-21configurable per-dispatch-cycle sleepDave Barach3-2/+25
Workaround for lack of driver interrupt support. Also quite handy for home gateway, laptop/vagrant, other use-cases not requiring maximum vectors/second for proper operation. Change-Id: Ifc4b98112450664beef67b89ab8a6940a3bf24b5 Signed-off-by: Dave Barach <dave@barachs.net>
2018-06-15STATS: Add more hierarchy to counters.Ole Troan1-1/+1
Put error counters under node_index. /err/<node-name>/<error-name> E.g: /err/ip4-lookup/Hash table collisions /err/ip6-rewrite/Buffer allocation error Error names are not globally unique, and this allows for walking all errors for a particular node. Put interface counters under the directory /if. E.g. /if/rx Put system counters under the directory /sys. E.g. /sys/vector_rate Change-Id: I5b794d16698f61bcb2063a8cd77a7c4ae36419b8 Signed-off-by: Ole Troan <ot@cisco.com>
2018-06-14Use __attribute__((weak)) references where necessaryDave Barach1-4/+17
It should be possible to use vlib without the vlibmemory library, etc. Change-Id: Ic2316b93d7dbb728fb4ff42a3ca8b0d747c9425e Signed-off-by: Dave Barach <dave@barachs.net>
2018-06-13Stat segment / client: show run" works nowDave Barach4-14/+44
Seems to have minimal-to-zero performance consequences. Data appears accurate: result match the debug CLI output. Checked at low rates, 27 MPPS sprayed across two worker threads. Change-Id: I09ede5150b88a91547feeee448a2854997613004 Signed-off-by: Dave Barach <dave@barachs.net>
2018-06-13add end of string for format name in unix_physmem_region_alloclollita liu1-1/+1
Change-Id: I8cb4d5d90fb321de6e5037a3d0440507db79ec75 Signed-off-by: lollita liu <lollita.liu@ericsson.com>
2018-06-13cli: 'restart' should close open files (VPP-1068)Chris Luke1-2/+17
- On 'restart' close all registered files (except stdio) so that the new process has a clean start; in particular, CLI sockets, API sockets, tun/af_packet etc descriptors all need to close so they're not left open but unused by the new VPP process. To do this we iterate all the files registered for the polling mechanism and close() them.[1] - While we're here, retain the original environment on 'restart'; several things make use of it. [1] An alternative mechanism would be to mark all files with CLOEXEC on opening; I think that is a little fragile for this fairly esoteric use case. Change-Id: I81b4e261c4d3c4e2948981231be899824dd4e69f Signed-off-by: Chris Luke <chrisy@flirble.org>
2018-06-10cli: Disable XON/XOFF in the ttyChris Luke1-0/+4
- CLI history forward-search is bound to ^S which is common, but that is also the tty's default control byte to pause output. So we disable XON/XOFF in the tty so that we can use ^S. Change-Id: I61717c77a11733d64eed7f8119677e7cd2e20029 Signed-off-by: Chris Luke <chrisy@flirble.org>
2018-06-10cli: Fix reverse-line-wrap in terminals (VPP-1295)Chris Luke1-86/+195
- Terminals do not reverse-line-wrap when the cursor is at the left edge and \b tries to make it go left. - Instead, we have to track the cursor position if we need to emit \b's and if we are at the left edge emit an ANSI sequence to relocate the cursor. Previously we usually simply calculated the new cursor position after a bunch of output had completed. - Further trickiness is required since most xterm-like terminals also defer moving the cursor to the next line when at the right edge[1], and then if they receive a \b move the cursor back one character too many. - This requires intricate reworking of everywhere that \b is emitted by the CLI code during command line editing. [1] Bash counters this issue by tracking the cursor position as output is generated and forcing the cursor to the next line (by emitting a space followed by \r) if it gets to this phantom cursor position); here we effectively do that but only if the user tries to go left when in this state. Change-Id: I7c1d7c0e24c53111a5810cebb504ccfdac743086 Signed-off-by: Chris Luke <chrisy@flirble.org>
2018-06-10cli: Fix off-by-one in the pagerChris Luke1-2/+2
- The last line in the pager buffer was sometimes missed when using space/pg-dn; simple off-by-one error. Change-Id: Id4e5f7cf0e5db4f719f87b9069d75427bc66d3f7 Signed-off-by: Chris Luke <chrisy@flirble.org>
2018-06-09Fix bug in vlib_buffer_free_from_ring_no_nextDamjan Marion1-1/+1
Change-Id: I332bb4578d1a3c79770985bf1f315d2ed823a3e5 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-06-08export counters in a memfd segmentDave Barach4-1/+82
also export per-node error counters directory entries implement object types Change-Id: I8ce8e0a754e1be9de895c44ed9be6533b4ecef0f Signed-off-by: Dave Barach <dave@barachs.net>
2018-06-05VPP API: Memory traceOle Troan1-9/+49
if you plan to put a hash into shared memory, the key sum and key equal functions MUST be set to constants such as KEY_FUNC_STRING, KEY_FUNC_MEM, etc. -lvppinfra is PIC, which means that the process which set up the hash won't have the same idea where the key sum and key compare functions live in other processes. Change-Id: Ib3b5963a0d2fb467b91e1f16274df66ac74009e9 Signed-off-by: Ole Troan <ot@cisco.com> Signed-off-by: Dave Barach <dave@barachs.net> Signed-off-by: Ole Troan <ot@cisco.com>
2018-06-03dpdk: improve buffer alloc perfomanceDamjan Marion2-3/+0
This is ~50% improvement in buffer alloc performance. For a 256 buffer allocation, it was ~10 clocks/buffer, now is < 5 clocks. Change-Id: I97590e240a79a42bcab5eb26587fc2d11e6eb163 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-31fix missing void pointer explicit type castZhiyong Yang1-1/+1
Change-Id: I5098e5d966abc44db045783883424b63083f08ee Signed-off-by: Zhiyong Yang <zhiyong.yang@intel.com>
2018-05-31Add src/vlib/config.h.in templateDamjan Marion1-0/+21
Change-Id: I9b81dcb9250cab09b88f161366da9475a7ee5dd6 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-29Add VLIB_NODE_FN() macro to simplify multiversioning of node functionsDamjan Marion2-0/+48
Change-Id: Ibab5e27277f618ceb2d543b9d6a1a5f191e7d1db Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-29Fixed vlib_buffer_clone with stale chained buffersYoann Desmouceaux1-2/+7
When calling vlib_buffer_clone() on a source vlib_buffer with no next buffer but whose total_length_not_including_first_buffer hadn't been properly zeroed out, the total_length_not_including_first_buffer of the clone was set to a wrong value. (see https://lists.fd.io/g/vpp-dev/topic/19869395) Change-Id: I4b503ece804e3933bb259be4c2148f84dafbea3e Signed-off-by: Yoann Desmouceaux <ydesmouc@cisco.com>
2018-05-19log: Validate the size of vec in vlib_logMohsin Kazmi1-0/+1
Change-Id: I6d1218c17ee055275596b9a49767f15994aa1b2b Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
2018-05-19Disable vector code in vlib_buffer_enqueue_to_next if no msb mask functionDamjan Marion1-1/+1
This fixes ARM64 build where we dont have defined u16x8_msb_mask(...) Change-Id: I864f5134a0d951601810c800f587d173b3b7ef41 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-18Add vlib_buffer_enqueue_to_next inline functionDamjan Marion1-0/+115
Change-Id: I1042c0fe179b57a00ce99c8d62cb1bdbe24d9184 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-18session: add session process nodeFlorin Coras1-1/+3
Add a session process node that handles main thread tx and retransmit in order to avoid having a polling input node. Change-Id: I3357e987c023a84b533b32793e37ab4204420f64 Signed-off-by: Florin Coras <fcoras@cisco.com>
2018-05-17Add buffer pointer-to-index and index-to-pointer array functionsDamjan Marion1-0/+143
Change-Id: Ib3fcc3ceb7f315389bcdecbb7d9632540a5dd6ba Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-05-17vlib: Fix WARN-ING macro in functionMohsin Kazmi1-1/+1
Change-Id: I238106c2afc46904fb0eb17164f30dbd1378892e Signed-off-by: Mohsin Kazmi <sykazmi@cisco.com>
2018-05-04Harmonize vec/pool_get_aligned object sizes and alignment requestsDave Barach1-0/+1
Object sizes must evenly divide alignment requests, or vice versa. Otherwise, only the first object will be aligned as requested. Three choices: add CLIB_CACHE_LINE_ALIGN_MARK(align_me) at the end of structures, manually pad to an even divisor or multiple of the alignment request, or use plain vectors/pools. static assert for enforcement. Change-Id: I41aa6ff1a58267301d32aaf4b9cd24678ac1c147 Signed-off-by: Dave Barach <dbarach@cisco.com>
2018-04-27A bit of buffer metadata reshuffling to accommodate flow_idDamjan Marion2-19/+12
Change-Id: I2794384557c6272fe217269b14a9db09eda19220 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-26vlib: set log tap level <level> does not work for some keywordsSteven1-2/+2
While some levels such as debug and emerg work, others don't. See below. DBGvpp# set log class tap level warn set log class tap level warn set logging class: unknown input `level warn' DBGvpp# set log class tap level debug set log class tap level debug DBGvpp# set log class tap level info set log class tap level info set logging class: unknown input `level info' DBGvpp# set log class tap level err set log class tap level err DBGvpp# set log class tap level crit set log class tap level crit set logging class: unknown input `level crit' DBGvpp# set log class tap level emerg set log class tap level emerg DBGvpp# Cause: The reason for the failure for the shorter keywords is level_str is unformatted with %v which is not null terminated. For example, the character after "info" could be anything in level_str. The memcmp with size of the macro keyword __##uc which includes the null character or 5 in this case and thus the comparison fails. Fix: Use %s which insure level_str is null terminated. Use strcmp to rule out false positve match like "debugxxx" against keyword "debug". Change-Id: I7a2d97a0f7f618df105da7eca791618dce04d21e Signed-off-by: Steven <sluong@cisco.com>
2018-04-18vlib: logging improvementsDamjan Marion2-32/+74
Change-Id: I1c72d9ddefdc83c7d5ab2564f89d83ffd3fa6110 Signed-off-by: Damjan Marion <damarion@cisco.com>
2018-04-17Add logging supportDamjan Marion3-0/+737
Change-Id: Ieac9cf50156dbbb4962411e900d59256441915ef Signed-off-by: Damjan Marion <damarion@cisco.com> Signed-off-by: Klement Sekera <ksekera@cisco.com>
2018-04-12Fixes for 'make UNATTENDED=yes CC=clang CXX=clang verify'Neale Ranns1-1/+2
Change-Id: I994649761fe2e66e12ae0e49a84fb1d0a966ddfb Signed-off-by: Neale Ranns <nranns@cisco.com>
2018-04-09features: don't break linked list, create separate one for arcDamjan Marion1-9/+11
We need to keep original linked list so destructire can remove entries. Change-Id: I5ff5ca0e1a417d88707255207725bba46433c943 Signed-off-by: Damjan Marion <damarion@cisco.com>