aboutsummaryrefslogtreecommitdiffstats
path: root/src/plugins/perfmon
AgeCommit message (Collapse)AuthorFilesLines
2019-01-27Fix issue with cpu_id and numa_code captured too earlyDamjan Marion1-1/+1
Change-Id: I79b213b34c6071d14acf1922f89037a4a5a36c45 Signed-off-by: Damjan Marion <damarion@cisco.com>
2019-01-27perfmon: collect data on selected thread(s)Dave Barach3-18/+86
Add missing pre-input node runtime fork and refork code. unix-epoll-input runs on all threads; each instance needs its own runtime stats. Change-Id: I16b02e42d0c95f863161176c4bb9f9917bef809d Signed-off-by: Dave Barach <dave@barachs.net>
2019-01-24perfmon: enable pmc event before reading rdpmc indexDamjan Marion1-6/+6
Change-Id: I9b0a101e5d78c10257e3c5d8f5573c3eb29bfdef Signed-off-by: Damjan Marion <damarion@cisco.com>
2019-01-24perfmon plugin: 2-way parallel stat collectionDave Barach3-143/+228
As a FUD reduction measure, this patch implements 2-way parallel counter collection. Synthetic stat component counter pairs run at the same time. Running two counters (of any kind) at the same time naturally reduces the aggregate time required by an approximate factor-of-2, depending on whether an even or odd number of stats have been requested. I don't completely buy the argument that computing synthetic stats such as instructions-per-clock will be inaccurate if component counter values are collected sequentially. Given uniform traffic pattern, it must make no difference. As the collection interval increases, the difference between serial and parallel component counter collection will approach zero, see also the Central Limit theorem. Change-Id: I36ebdcf125e8882cca8a1929ec58f17fba1ad8f1 Signed-off-by: Dave Barach <dave@barachs.net>
2018-12-17Improve perfmon json table pickerDave Barach4-23/+341
Built a tool to chew up https://download.01.org/perfmon/mapfile.csv, and output a table in this format: typedef struct { u8 model; u8 stepping; u8 has_stepping; char *filename; } file_by_model_and_stepping_t; static const file_by_model_and_stepping_t fms_table [] = { /* model, stepping, stepping valid, file */ { 0x2E, 0x0, 0, "NehalemEX_core_V2.json" }, { 0x1E, 0x0, 0, "NehalemEP_core_V2.json" }, <snip> { 0x55, 0x5, 1, "cascadelakex_core_v1.00.json" }, { 0x55, 0x6, 1, "cascadelakex_core_v1.00.json" }, { 0x55, 0x7, 1, "cascadelakex_core_v1.00.json" }, <snip> Change-Id: Ie0e8a7e851799e9d060b966047745039c066ec7b Signed-off-by: Dave Barach <dave@barachs.net>
2018-11-14Remove c-11 memcpy checks from perf-critical codeDave Barach1-1/+1
Change-Id: Id4f37f5d4a03160572954a416efa1ef9b3d79ad1 Signed-off-by: Dave Barach <dave@barachs.net>
2018-10-25Add x86_64 perfmon tablesDave Barach3-20/+10
The license issue is resolved, so we can package the .json files. Added to the vpp-dev package in .tar.xz form, which saves a lot of space. Updated the perfmon error log entry: tell folks where to find the compressed tarball, and how to extract it. Change-Id: I3ed351fbf154cc3ba22d5f9c666acff77a2a14cf Signed-off-by: Dave Barach <dave@barachs.net>
2018-10-23VPP-1474: fix coverity warningDave Barach1-0/+3
Change-Id: I20f2fb14e00f3e7e96774959a4bf1a159ab9030f Signed-off-by: Dave Barach <dave@barachs.net>
2018-10-23perfmon.c: Register additional cpuids.Paul Vinciguerra1-4/+7
Added/tested additional cpuids from our testbed. Change-Id: Ifd3ea9e8e8231a8901966903bf5eceb635b82482 Signed-off-by: Paul Vinciguerra <pvinci@vinciconsulting.com>
2018-10-22X86_64 perf counter pluginDave Barach5-0/+1466
Change-Id: Ie5a00c15ee9536cc61afab57f6cadc1aa1972f3c Signed-off-by: Dave Barach <dave@barachs.net>