Date: Mon, 1 Jun 2015 18:08:56 +0000 (UTC) From: John Baldwin <jhb@FreeBSD.org> To: src-committers@freebsd.org, svn-src-all@freebsd.org, svn-src-stable@freebsd.org, svn-src-stable-10@freebsd.org Subject: svn commit: r283886 - in stable: 10/sys/dev/hwpmc 9/sys/dev/hwpmc Message-ID: <201506011808.t51I8uQ1022510@svn.freebsd.org>
next in thread | raw e-mail | index | archive | help
Author: jhb Date: Mon Jun 1 18:08:56 2015 New Revision: 283886 URL: https://svnweb.freebsd.org/changeset/base/283886 Log: MFC 283123: Fix two bugs that could result in PMC sampling effectively stopping. In both cases, the the effect of the bug was that a very small positive number was written to the counter. This means that a large number of events needed to occur before the next sampling interrupt would trigger. Even with very frequently occurring events like clock cycles wrapping all the way around could take a long time. Both bugs occurred when updating the saved reload count for an outgoing thread on a context switch. First, the counter-independent code compares the current reload count against the count set when the thread switched in and generates a delta to apply to the saved count. If this delta causes the reload counter to go negative, it would add a full reload interval to wrap it around to a positive value. The fix is to add the full reload interval if the resulting counter is zero. Second, occasionally the raw counter value read during a context switch has actually wrapped, but an interrupt has not yet triggered. In this case the existing logic would return a very large reload count (e.g. 2^48 - 2 if the counter had overflowed by a count of 2). This was seen both for fixed-function and programmable counters on an E5-2643. Workaround this case by returning a reload count of zero. PR: 198149 Sponsored by: Norse Corp, Inc. Modified: stable/10/sys/dev/hwpmc/hwpmc_core.c stable/10/sys/dev/hwpmc/hwpmc_mod.c Directory Properties: stable/10/ (props changed) Changes in other areas also in this revision: Modified: stable/9/sys/dev/hwpmc/hwpmc_core.c stable/9/sys/dev/hwpmc/hwpmc_mod.c Directory Properties: stable/9/sys/ (props changed) stable/9/sys/dev/ (props changed) Modified: stable/10/sys/dev/hwpmc/hwpmc_core.c ============================================================================== --- stable/10/sys/dev/hwpmc/hwpmc_core.c Mon Jun 1 18:05:30 2015 (r283885) +++ stable/10/sys/dev/hwpmc/hwpmc_core.c Mon Jun 1 18:08:56 2015 (r283886) @@ -203,6 +203,10 @@ core_pcpu_fini(struct pmc_mdep *md, int static pmc_value_t iaf_perfctr_value_to_reload_count(pmc_value_t v) { + + /* If the PMC has overflowed, return a reload count of zero. */ + if ((v & (1ULL << (core_iaf_width - 1))) == 0) + return (0); v &= (1ULL << core_iaf_width) - 1; return (1ULL << core_iaf_width) - v; } @@ -1786,6 +1790,10 @@ static const int niap_events = sizeof(ia static pmc_value_t iap_perfctr_value_to_reload_count(pmc_value_t v) { + + /* If the PMC has overflowed, return a reload count of zero. */ + if ((v & (1ULL << (core_iap_width - 1))) == 0) + return (0); v &= (1ULL << core_iap_width) - 1; return (1ULL << core_iap_width) - v; } Modified: stable/10/sys/dev/hwpmc/hwpmc_mod.c ============================================================================== --- stable/10/sys/dev/hwpmc/hwpmc_mod.c Mon Jun 1 18:05:30 2015 (r283885) +++ stable/10/sys/dev/hwpmc/hwpmc_mod.c Mon Jun 1 18:08:56 2015 (r283886) @@ -1440,7 +1440,7 @@ pmc_process_csw_out(struct thread *td) tmp += pm->pm_sc.pm_reloadcount; mtx_pool_lock_spin(pmc_mtxpool, pm); pp->pp_pmcs[ri].pp_pmcval -= tmp; - if ((int64_t) pp->pp_pmcs[ri].pp_pmcval < 0) + if ((int64_t) pp->pp_pmcs[ri].pp_pmcval <= 0) pp->pp_pmcs[ri].pp_pmcval += pm->pm_sc.pm_reloadcount; mtx_pool_unlock_spin(pmc_mtxpool, pm);
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?201506011808.t51I8uQ1022510>