From nobody Thu Dec 09 02:39:29 2021 X-Original-To: dev-commits-src-all@mlmmj.nyi.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2610:1c1:1:606c::19:1]) by mlmmj.nyi.freebsd.org (Postfix) with ESMTP id 2608C18DE990; Thu, 9 Dec 2021 02:39:30 +0000 (UTC) (envelope-from git@FreeBSD.org) Received: from mxrelay.nyi.freebsd.org (mxrelay.nyi.freebsd.org [IPv6:2610:1c1:1:606c::19:3]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256 client-signature RSA-PSS (4096 bits) client-digest SHA256) (Client CN "mxrelay.nyi.freebsd.org", Issuer "R3" (verified OK)) by mx1.freebsd.org (Postfix) with ESMTPS id 4J8dXd6LW8z3l7F; Thu, 9 Dec 2021 02:39:29 +0000 (UTC) (envelope-from git@FreeBSD.org) Received: from gitrepo.freebsd.org (gitrepo.freebsd.org [IPv6:2610:1c1:1:6068::e6a:5]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (4096 bits) server-digest SHA256) (Client did not present a certificate) by mxrelay.nyi.freebsd.org (Postfix) with ESMTPS id B5FDB1E9F2; Thu, 9 Dec 2021 02:39:29 +0000 (UTC) (envelope-from git@FreeBSD.org) Received: from gitrepo.freebsd.org ([127.0.1.44]) by gitrepo.freebsd.org (8.16.1/8.16.1) with ESMTP id 1B92dTWn040320; Thu, 9 Dec 2021 02:39:29 GMT (envelope-from git@gitrepo.freebsd.org) Received: (from git@localhost) by gitrepo.freebsd.org (8.16.1/8.16.1/Submit) id 1B92dTRL040319; Thu, 9 Dec 2021 02:39:29 GMT (envelope-from git) Date: Thu, 9 Dec 2021 02:39:29 GMT Message-Id: <202112090239.1B92dTRL040319@gitrepo.freebsd.org> To: src-committers@FreeBSD.org, dev-commits-src-all@FreeBSD.org, dev-commits-src-main@FreeBSD.org From: Alexander Motin Subject: git: 63346fef3382 - main - mca: Some error handling logic improvements. List-Id: Commit messages for all branches of the src repository List-Archive: https://lists.freebsd.org/archives/dev-commits-src-all List-Help: List-Post: List-Subscribe: List-Unsubscribe: Sender: owner-dev-commits-src-all@freebsd.org X-BeenThere: dev-commits-src-all@freebsd.org MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: 8bit X-Git-Committer: mav X-Git-Repository: src X-Git-Refname: refs/heads/main X-Git-Reftype: branch X-Git-Commit: 63346fef3382b0ee02366cd9799930f8dbc15a52 Auto-Submitted: auto-generated ARC-Message-Signature: i=1; a=rsa-sha256; c=relaxed/relaxed; d=freebsd.org; s=dkim; t=1639017569; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding; bh=4JfMCRqUThI87DGCvrbhCwYM9EmgmqNRF6XAQ7dgvsI=; b=ZGTzA2rOq5uKashBH+hkXAWM5bf8JOoyZIahyCPdFfQjApHMT1o1GbP1zhVcMkoKMFCCKs OCPt3yXTJZ6gHRaX1TdDuAeZ5nzq6Sk9HLpZRva5pw98Q677ZMwdeHY17BVOSRZ2G1+eir ZPg0NoGSCBg9lt0pvEXDLgE8iQLBhDS/M2Wd253H+0UyrxB0Yj8QW+UlIRYuDXJGWbh2Ju tEADit3YA+NjnYdrU8RCBqr2NDHFhWwdPNTxl0Wspl11Llk9qhRv8zdZIDY0KdFI+YVtCE J6WsvSTt8NaYE0tl3dcLBvxmiZqq15syFmFA9qu4jTb7yYrpHeQz9JR/NwDqcQ== ARC-Seal: i=1; s=dkim; d=freebsd.org; t=1639017569; a=rsa-sha256; cv=none; b=lbbYg0SVYgtye6wmUzEBhyCr4o03LjF8qznIUwqojY9EVHBr6bQe07CDxaZIXaxQ7vV/ID xi59gOrzBEKmkzUWLVQ0cEZyUR5ZfwG7jR0PpqhROwN6+J6l+U7aBOb34+T9K+KNhTIgzm gV4gbOa9eY/fcRruPxpV6eMBGME1nLaxQ95NE7Q4L2kKsqGkWrm+kWvNP5sjxhPmtQTeR6 naTbC47c4kPp7Pz/x0GPK2M1Jus4TQr8Vvd7SIJik0QFwbYbYwtCXz34nmqVVyYak0bpxh ONm6ELeBMWbpTBJn4IQmmgFiCx8NVVIV8FakVAW6OP8cIG4Wh+O9lwnDohCsXQ== ARC-Authentication-Results: i=1; mx1.freebsd.org; none X-ThisMailContainsUnwantedMimeParts: N The branch main has been updated by mav: URL: https://cgit.FreeBSD.org/src/commit/?id=63346fef3382b0ee02366cd9799930f8dbc15a52 commit 63346fef3382b0ee02366cd9799930f8dbc15a52 Author: Alexander Motin AuthorDate: 2021-12-09 02:06:11 +0000 Commit: Alexander Motin CommitDate: 2021-12-09 02:39:24 +0000 mca: Some error handling logic improvements. - Enable local MCEs on capable Intel CPUs. It delivers exceptions only to the affected CPU instead of global broadcast, requiring a lot of synchronization between CPUs. AMD always deliver MCEs locally. - Make MCE handler process only uncorrected errors, while CMCI and polling only corrected. It reduces synchronization problems between them and is explicitly recommended by the documentation. - Add minimal support for uncorrected software recoverable errors on Intel CPUs. It allows to avoid kernel panics in case uncorrected errors do not affect current operation, like ones found during scrub or write. Such errors are only logged, postponing the panic until the corrupted data will actually be needed (that may never happen). - Reduce polling period from 1 hour to 5 minutes. MFC after: 2 weeks --- sys/x86/x86/mca.c | 114 ++++++++++++++++++++++++++++++++++++------------------ 1 file changed, 77 insertions(+), 37 deletions(-) diff --git a/sys/x86/x86/mca.c b/sys/x86/x86/mca.c index c3befa46615b..e3386b5b7459 100644 --- a/sys/x86/x86/mca.c +++ b/sys/x86/x86/mca.c @@ -126,7 +126,7 @@ static STAILQ_HEAD(, mca_internal) mca_freelist; static int mca_freecount; static STAILQ_HEAD(, mca_internal) mca_records; static STAILQ_HEAD(, mca_internal) mca_pending; -static int mca_ticks = 3600; /* Check hourly by default. */ +static int mca_ticks = 300; static struct taskqueue *mca_tq; static struct task mca_resize_task; static struct timeout_task mca_scan_task; @@ -608,16 +608,60 @@ mca_log(const struct mca_record *rec) printf("MCA: Misc 0x%llx\n", (long long)rec->mr_misc); } +static bool +mca_is_mce(uint64_t mcg_cap, uint64_t status, bool *recoverablep) +{ + + /* Corrected error. */ + if ((status & MC_STATUS_UC) == 0) + return (0); + + /* Spurious MCA error. */ + if ((status & MC_STATUS_EN) == 0) + return (0); + + /* The processor does not support software error recovery. */ + if (!ser_supported(mcg_cap)) { + *recoverablep = false; + return (1); + } + + /* Context might have been corrupted. */ + if (status & MC_STATUS_PCC) { + *recoverablep = false; + return (1); + } + + /* Uncorrected software recoverable. */ + if (status & MC_STATUS_S) { + /* Action required vs optional. */ + if (status & MC_STATUS_AR) + *recoverablep = false; + return (1); + } + + /* Uncorrected no action required. */ + return (0); +} + static int -mca_check_status(int bank, struct mca_record *rec) +mca_check_status(enum scan_mode mode, uint64_t mcg_cap, int bank, + struct mca_record *rec, bool *recoverablep) { uint64_t status; u_int p[4]; + bool mce, recover; status = rdmsr(mca_msr_ops.status(bank)); if (!(status & MC_STATUS_VAL)) return (0); + recover = *recoverablep; + mce = mca_is_mce(mcg_cap, status, &recover); + if (mce != (mode == MCE)) + return (0); + *recoverablep = recover; + /* Save exception information. */ rec->mr_status = status; rec->mr_bank = bank; @@ -639,7 +683,7 @@ mca_check_status(int bank, struct mca_record *rec) * Clear machine check. Don't do this for uncorrectable * errors so that the BIOS can see them. */ - if (!(rec->mr_status & (MC_STATUS_PCC | MC_STATUS_UC))) { + if (!mce || recover) { wrmsr(mca_msr_ops.status(bank), 0); do_cpuid(0, p); } @@ -837,24 +881,15 @@ amd_thresholding_update(enum scan_mode mode, int bank, int valid) * reported immediately via mca_log(). The current thread must be * pinned when this is called. The 'mode' parameter indicates if we * are being called from the MC exception handler, the CMCI handler, - * or the periodic poller. In the MC exception case this function - * returns true if the system is restartable. Otherwise, it returns a - * count of the number of valid MC records found. + * or the periodic poller. */ static int -mca_scan(enum scan_mode mode, int *recoverablep) +mca_scan(enum scan_mode mode, bool *recoverablep) { struct mca_record rec; - uint64_t mcg_cap, ucmask; - int count, i, recoverable, valid; - - count = 0; - recoverable = 1; - ucmask = MC_STATUS_UC | MC_STATUS_PCC; + uint64_t mcg_cap; + int count = 0, i, valid; - /* When handling a MCE#, treat the OVER flag as non-restartable. */ - if (mode == MCE) - ucmask |= MC_STATUS_OVER; mcg_cap = rdmsr(MSR_MCG_CAP); for (i = 0; i < (mcg_cap & MCG_CAP_COUNT); i++) { #ifdef DEV_APIC @@ -866,16 +901,13 @@ mca_scan(enum scan_mode mode, int *recoverablep) continue; #endif - valid = mca_check_status(i, &rec); + valid = mca_check_status(mode, mcg_cap, i, &rec, recoverablep); if (valid) { count++; - if (rec.mr_status & ucmask) { - recoverable = 0; - mtx_lock_spin(&mca_lock); + if (*recoverablep) + mca_record_entry(mode, &rec); + else mca_log(&rec); - mtx_unlock_spin(&mca_lock); - } - mca_record_entry(mode, &rec); } #ifdef DEV_APIC @@ -891,8 +923,6 @@ mca_scan(enum scan_mode mode, int *recoverablep) } #endif } - if (recoverablep != NULL) - *recoverablep = recoverable; return (count); } @@ -962,21 +992,21 @@ static void mca_scan_cpus(void *context, int pending) { struct thread *td; - int count, cpu; + int cpu; + bool recoverable = true; mca_resize_freelist(); td = curthread; - count = 0; thread_lock(td); CPU_FOREACH(cpu) { sched_bind(td, cpu); thread_unlock(td); - count += mca_scan(POLLED, NULL); + mca_scan(POLLED, &recoverable); thread_lock(td); sched_unbind(td); } thread_unlock(td); - if (count != 0) + if (!STAILQ_EMPTY(&mca_pending)) mca_process_records(POLLED); taskqueue_enqueue_timeout_sbt(mca_tq, &mca_scan_task, mca_ticks * SBT_1S, 0, C_PREL(1)); @@ -1371,6 +1401,13 @@ _mca_init(int boot) mca_msr_ops.misc = mca_smca_misc_reg; } + /* Enable local MCE if supported. */ + if (cpu_vendor_id == CPU_VENDOR_INTEL && + (mcg_cap & MCG_CAP_LMCE_P) && + (rdmsr(MSR_IA32_FEATURE_CONTROL) & + IA32_FEATURE_CONTROL_LMCE_EN)) + wrmsr(MSR_MCG_EXT_CTL, rdmsr(MSR_MCG_EXT_CTL) | 1); + /* * The cmci_monitor() must not be executed * simultaneously by several CPUs. @@ -1422,7 +1459,7 @@ _mca_init(int boot) mtx_unlock_spin(&mca_lock); #ifdef DEV_APIC - if (!amd_thresholding_supported() && + if (cmci_supported(mcg_cap) && PCPU_GET(cmci_mask) != 0 && boot) lapic_enable_cmc(); #endif @@ -1465,7 +1502,8 @@ void mca_intr(void) { uint64_t mcg_status; - int recoverable, count; + int count; + bool lmcs, recoverable; if (!(cpu_feature & CPUID_MCA)) { /* @@ -1475,14 +1513,15 @@ mca_intr(void) printf("MC Type: 0x%jx Address: 0x%jx\n", (uintmax_t)rdmsr(MSR_P5_MC_TYPE), (uintmax_t)rdmsr(MSR_P5_MC_ADDR)); - panic("Machine check"); + panic("Machine check exception"); } /* Scan the banks and check for any non-recoverable errors. */ - count = mca_scan(MCE, &recoverable); mcg_status = rdmsr(MSR_MCG_STATUS); - if (!(mcg_status & MCG_STATUS_RIPV)) - recoverable = 0; + recoverable = (mcg_status & MCG_STATUS_RIPV) != 0; + lmcs = (cpu_vendor_id != CPU_VENDOR_INTEL || + (mcg_status & MCG_STATUS_LMCS)); + count = mca_scan(MCE, &recoverable); if (!recoverable) { /* @@ -1490,7 +1529,7 @@ mca_intr(void) * Some errors will assert a machine check on all CPUs, but * only certain CPUs will find a valid bank to log. */ - while (count == 0) + while (!lmcs && count == 0) cpu_spinwait(); panic("Unrecoverable machine check exception"); @@ -1505,6 +1544,7 @@ mca_intr(void) void cmc_intr(void) { + bool recoverable = true; /* * Serialize MCA bank scanning to prevent collisions from @@ -1512,7 +1552,7 @@ cmc_intr(void) * * If we found anything, log them to the console. */ - if (mca_scan(CMCI, NULL) != 0) + if (mca_scan(CMCI, &recoverable) != 0) mca_process_records(CMCI); } #endif