FreeBSD source tree https://cgit-dev.freebsd.org/src/atom?h=releng%2F12.2 2020-08-19T13:07:04Z 2020-08-19T13:07:04Z Hans Petter Selasky hselasky@FreeBSD.org 2020-08-19T13:07:04Z urn:sha1:898a95874a5bde996e5e3455d62aa3e5f927b7c1 Remove the "config" taskqgroup and its KPIs. Equivalent functionality is already provided by taskqueue(9), just use that instead. Interfaces may be detached from a taskqueue_thread task, for example by prison_complete(), so after r359438, when draining the queue we may end up deadlocking. 2020-08-04T12:04:47Z Ryan Moeller freqlabs@FreeBSD.org 2020-08-04T12:04:47Z urn:sha1:83fbe66ef7255fa4d1223ea0bd82fc54e1e8122d libpmc: Use known pmc_cpuid buffer size Use the existing PMC_CPUID_LEN to size pmc_cpuid in the kernel and various buffers for reading it in libpmc. This avoids some extra syscalls and malloc/frees. While in here, use strlcpy to copy a user-provided cpuid string instead of memcpy, to make sure we terminate the buffer. Reviewed by: mav Sponsored by: iXsystems, Inc. Differential Revision: https://reviews.freebsd.org/D25679 2020-07-28T02:53:44Z Ryan Moeller freqlabs@FreeBSD.org 2020-07-28T02:53:44Z urn:sha1:898b2ca1291b8959e9959ca861fe30f457594353 hwpmc: Always set pmc_cpuid to something pmc_cpuid was uninitialized for most AMD processor families. We can still populate this string for unimplemented families. Also added a CPUID_TO_STEPPING macro and converted existing code to use it. Reviewed by: mav Sponsored by: iXsystems, Inc. Differential Revision: https://reviews.freebsd.org/D25673 2020-07-20T13:50:10Z Alexander Motin mav@FreeBSD.org 2020-07-20T13:50:10Z urn:sha1:c4896d816f6081b194d261e03f5beade4a3730b2 It follows the equivalent Linux change to be able to differentiate skylakex and cascadelakex, sharing the same model but not stepping. 2020-05-01T00:40:18Z Alexander Motin mav@FreeBSD.org 2020-05-01T00:40:18Z urn:sha1:2b136caaf356a9aaa9fdafa0b81a4f7cfba0730d According to the 325462-071US document, they should be the same. 2020-04-24T15:02:05Z Alexander Motin mav@FreeBSD.org 2020-04-24T15:02:05Z urn:sha1:5bddeab3424b5190e40894a76b50fc650728496a Recognize new micro-architecture in hwpmc_intel driver. Based on Intel document 325462-071US. Tested with tools/test/hwpmc/pmctest.py on Atom E3930 SoC. 2020-04-24T14:47:55Z Alexander Motin mav@FreeBSD.org 2020-04-24T14:47:55Z urn:sha1:84e82cf5f5e2559ab73b47a6bba291eb4c673517 - amd_intr() does not account for the offset (0x200) in the counter MSR address and ends up accessing invalid regions while reading counter value after the 4th counter (0xC001000[8,9,..]) and erroneously updates the counter values for counters [1-4]. - amd_intr() should only check core pmcs for interrupts since other types of pmcs (L3,DF) cannot generate interrupts. - fix pmc NMI's being ignored due to NMI latency on newer AMD processors Note that this fixes a kernel panic due to GPFs accessing MSRs on higher core count AMD cpus (seen on both Rome 7502P, and Threadripper 2990WX 32-core CPUs) 2020-04-06T14:58:24Z Mark Johnston markj@FreeBSD.org 2020-04-06T14:58:24Z urn:sha1:3bc47a0786de1cb45a75a3b5e5b09fd8e5831fe4 Fix accounting of hwpmc's thread descriptor freelist. 2020-03-12T20:59:00Z Matt Macy mmacy@FreeBSD.org 2020-03-12T20:59:00Z urn:sha1:5fc79c36bfb40bbf7263a186838ce0a898454c91 Don't drop samples with callchain pending Tested by: mjg@ MFC of r353483 2020-02-14T13:26:45Z Konstantin Belousov kib@FreeBSD.org 2020-02-14T13:26:45Z urn:sha1:d4081c82eda3f34ff8ca91b0ff14b0acbadf33eb pmc: Add Hygon Dhyana support.