From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S965304AbeCHJ2T (ORCPT ); Thu, 8 Mar 2018 04:28:19 -0500 Received: from terminus.zytor.com ([198.137.202.136]:58169 "EHLO terminus.zytor.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S935057AbeCHJ2Q (ORCPT ); Thu, 8 Mar 2018 04:28:16 -0500 Date: Thu, 8 Mar 2018 01:28:02 -0800 From: tip-bot for Ashok Raj Message-ID: Cc: arjan.van.de.ven@intel.com, tglx@linutronix.de, linux-kernel@vger.kernel.org, mingo@kernel.org, hpa@zytor.com, bp@suse.de, ashok.raj@intel.com, thomas.lendacky@amd.com Reply-To: ashok.raj@intel.com, bp@suse.de, thomas.lendacky@amd.com, tglx@linutronix.de, linux-kernel@vger.kernel.org, arjan.van.de.ven@intel.com, hpa@zytor.com, mingo@kernel.org In-Reply-To: <20180228102846.13447-8-bp@alien8.de> References: <20180228102846.13447-8-bp@alien8.de> To: linux-tip-commits@vger.kernel.org Subject: [tip:x86/pti] x86/microcode: Synchronize late microcode loading Git-Commit-ID: a5321aec6412b20b5ad15db2d6b916c05349dbff X-Mailer: tip-git-log-daemon Robot-ID: Robot-Unsubscribe: Contact to get blacklisted from these emails MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Content-Type: text/plain; charset=UTF-8 Content-Disposition: inline Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Commit-ID: a5321aec6412b20b5ad15db2d6b916c05349dbff Gitweb: https://git.kernel.org/tip/a5321aec6412b20b5ad15db2d6b916c05349dbff Author: Ashok Raj AuthorDate: Wed, 28 Feb 2018 11:28:46 +0100 Committer: Thomas Gleixner CommitDate: Thu, 8 Mar 2018 10:19:26 +0100 x86/microcode: Synchronize late microcode loading Original idea by Ashok, completely rewritten by Borislav. Before you read any further: the early loading method is still the preferred one and you should always do that. The following patch is improving the late loading mechanism for long running jobs and cloud use cases. Gather all cores and serialize the microcode update on them by doing it one-by-one to make the late update process as reliable as possible and avoid potential issues caused by the microcode update. [ Borislav: Rewrite completely. ] Co-developed-by: Borislav Petkov Signed-off-by: Ashok Raj Signed-off-by: Borislav Petkov Signed-off-by: Thomas Gleixner Tested-by: Tom Lendacky Tested-by: Ashok Raj Reviewed-by: Tom Lendacky Cc: Arjan Van De Ven Link: https://lkml.kernel.org/r/20180228102846.13447-8-bp@alien8.de --- arch/x86/kernel/cpu/microcode/core.c | 118 +++++++++++++++++++++++++++-------- 1 file changed, 92 insertions(+), 26 deletions(-) diff --git a/arch/x86/kernel/cpu/microcode/core.c b/arch/x86/kernel/cpu/microcode/core.c index 5dd157d48606..70ecbc8099c9 100644 --- a/arch/x86/kernel/cpu/microcode/core.c +++ b/arch/x86/kernel/cpu/microcode/core.c @@ -22,13 +22,16 @@ #define pr_fmt(fmt) "microcode: " fmt #include +#include #include #include #include #include #include +#include #include #include +#include #include #include @@ -64,6 +67,11 @@ LIST_HEAD(microcode_cache); */ static DEFINE_MUTEX(microcode_mutex); +/* + * Serialize late loading so that CPUs get updated one-by-one. + */ +static DEFINE_SPINLOCK(update_lock); + struct ucode_cpu_info ucode_cpu_info[NR_CPUS]; struct cpu_info_ctx { @@ -486,6 +494,19 @@ static void __exit microcode_dev_exit(void) /* fake device for request_firmware */ static struct platform_device *microcode_pdev; +/* + * Late loading dance. Why the heavy-handed stomp_machine effort? + * + * - HT siblings must be idle and not execute other code while the other sibling + * is loading microcode in order to avoid any negative interactions caused by + * the loading. + * + * - In addition, microcode update on the cores must be serialized until this + * requirement can be relaxed in the future. Right now, this is conservative + * and good. + */ +#define SPINUNIT 100 /* 100 nsec */ + static int check_online_cpus(void) { if (num_online_cpus() == num_present_cpus()) @@ -496,23 +517,85 @@ static int check_online_cpus(void) return -EINVAL; } -static enum ucode_state reload_for_cpu(int cpu) +static atomic_t late_cpus; + +/* + * Returns: + * < 0 - on error + * 0 - no update done + * 1 - microcode was updated + */ +static int __reload_late(void *info) { - struct ucode_cpu_info *uci = ucode_cpu_info + cpu; + unsigned int timeout = NSEC_PER_SEC; + int all_cpus = num_online_cpus(); + int cpu = smp_processor_id(); + enum ucode_state err; + int ret = 0; - if (!uci->valid) - return UCODE_OK; + atomic_dec(&late_cpus); + + /* + * Wait for all CPUs to arrive. A load will not be attempted unless all + * CPUs show up. + * */ + while (atomic_read(&late_cpus)) { + if (timeout < SPINUNIT) { + pr_err("Timeout while waiting for CPUs rendezvous, remaining: %d\n", + atomic_read(&late_cpus)); + return -1; + } + + ndelay(SPINUNIT); + timeout -= SPINUNIT; + + touch_nmi_watchdog(); + } + + spin_lock(&update_lock); + apply_microcode_local(&err); + spin_unlock(&update_lock); + + if (err > UCODE_NFOUND) { + pr_warn("Error reloading microcode on CPU %d\n", cpu); + ret = -1; + } else if (err == UCODE_UPDATED) { + ret = 1; + } - return apply_microcode_on_target(cpu); + atomic_inc(&late_cpus); + + while (atomic_read(&late_cpus) != all_cpus) + cpu_relax(); + + return ret; +} + +/* + * Reload microcode late on all CPUs. Wait for a sec until they + * all gather together. + */ +static int microcode_reload_late(void) +{ + int ret; + + atomic_set(&late_cpus, num_online_cpus()); + + ret = stop_machine_cpuslocked(__reload_late, NULL, cpu_online_mask); + if (ret < 0) + return ret; + else if (ret > 0) + microcode_check(); + + return ret; } static ssize_t reload_store(struct device *dev, struct device_attribute *attr, const char *buf, size_t size) { - int cpu, bsp = boot_cpu_data.cpu_index; enum ucode_state tmp_ret = UCODE_OK; - bool do_callback = false; + int bsp = boot_cpu_data.cpu_index; unsigned long val; ssize_t ret = 0; @@ -534,30 +617,13 @@ static ssize_t reload_store(struct device *dev, goto put; mutex_lock(µcode_mutex); - - for_each_online_cpu(cpu) { - tmp_ret = reload_for_cpu(cpu); - if (tmp_ret > UCODE_NFOUND) { - pr_warn("Error reloading microcode on CPU %d\n", cpu); - - /* set retval for the first encountered reload error */ - if (!ret) - ret = -EINVAL; - } - - if (tmp_ret == UCODE_UPDATED) - do_callback = true; - } - - if (!ret && do_callback) - microcode_check(); - + ret = microcode_reload_late(); mutex_unlock(µcode_mutex); put: put_online_cpus(); - if (!ret) + if (ret >= 0) ret = size; return ret;