From patchwork Mon Aug 15 11:04:08 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: =?utf-8?b?SsO8cmdlbiBHcm/Dnw==?= X-Patchwork-Id: 12943443 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id F1E68C48BE4 for ; Mon, 15 Aug 2022 11:04:38 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.387256.623425 (Exim 4.92) (envelope-from ) id 1oNXtJ-0001k4-TN; Mon, 15 Aug 2022 11:04:17 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 387256.623425; Mon, 15 Aug 2022 11:04:17 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1oNXtJ-0001jx-PX; Mon, 15 Aug 2022 11:04:17 +0000 Received: by outflank-mailman (input) for mailman id 387256; Mon, 15 Aug 2022 11:04:15 +0000 Received: from se1-gles-sth1-in.inumbo.com ([159.253.27.254] helo=se1-gles-sth1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1oNXtH-0001TU-Ij for xen-devel@lists.xenproject.org; Mon, 15 Aug 2022 11:04:15 +0000 Received: from smtp-out2.suse.de (smtp-out2.suse.de [195.135.220.29]) by se1-gles-sth1.inumbo.com (Halon) with ESMTPS id ff00fe8b-1c89-11ed-bd2e-47488cf2e6aa; Mon, 15 Aug 2022 13:04:13 +0200 (CEST) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by smtp-out2.suse.de (Postfix) with ESMTPS id 32B702012C; Mon, 15 Aug 2022 11:04:13 +0000 (UTC) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by imap2.suse-dmz.suse.de (Postfix) with ESMTPS id DB99813A93; Mon, 15 Aug 2022 11:04:12 +0000 (UTC) Received: from dovecot-director2.suse.de ([192.168.254.65]) by imap2.suse-dmz.suse.de with ESMTPSA id iDZONCwo+mLHBgAAMHmgww (envelope-from ); Mon, 15 Aug 2022 11:04:12 +0000 X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: ff00fe8b-1c89-11ed-bd2e-47488cf2e6aa DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=suse.com; s=susede1; t=1660561453; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=liuF5z62h6Zyt9FvDgpE6TztPWAGZMbvacA9LSHauHw=; b=bUSUrYi6OSB2OGNNu1WnuxpTee7e2w1RqAcmjTfcyU4aYRlxeo5DBO+dIVRN5CNx2PhxjR plFyRpvjklZyfKxHevIZ29fEk+ooCvS4Dl8H3PzzAy/yOvV6fhqJpdg3LI7KHipKKU+0U5 3K4Bd05jMVEq7FscgvS/a6LbFbEKkUw= From: Juergen Gross To: xen-devel@lists.xenproject.org Cc: Juergen Gross , George Dunlap , Dario Faggioli , Andrew Cooper , Jan Beulich , Julien Grall , Stefano Stabellini , Wei Liu Subject: [PATCH v2 1/3] xen/sched: introduce cpupool_update_node_affinity() Date: Mon, 15 Aug 2022 13:04:08 +0200 Message-Id: <20220815110410.19872-2-jgross@suse.com> X-Mailer: git-send-email 2.35.3 In-Reply-To: <20220815110410.19872-1-jgross@suse.com> References: <20220815110410.19872-1-jgross@suse.com> MIME-Version: 1.0 For updating the node affinities of all domains in a cpupool add a new function cpupool_update_node_affinity(). In order to avoid multiple allocations of cpumasks carve out memory allocation and freeing from domain_update_node_affinity() into new helpers, which can be used by cpupool_update_node_affinity(). Modify domain_update_node_affinity() to take an additional parameter for passing the allocated memory in and to allocate and free the memory via the new helpers in case NULL was passed. This will help later to pre-allocate the cpumasks in order to avoid allocations in stop-machine context. Signed-off-by: Juergen Gross Reviewed-by: Jan Beulich --- V2: - move helpers to core.c (Jan Beulich) - allocate/free memory in domain_update_node_aff() if NULL was passed in (Jan Beulich) --- xen/common/sched/core.c | 54 ++++++++++++++++++++++++++------------ xen/common/sched/cpupool.c | 39 +++++++++++++++------------ xen/common/sched/private.h | 7 +++++ xen/include/xen/sched.h | 9 ++++++- 4 files changed, 74 insertions(+), 35 deletions(-) diff --git a/xen/common/sched/core.c b/xen/common/sched/core.c index ff1ddc7624..085a9dd335 100644 --- a/xen/common/sched/core.c +++ b/xen/common/sched/core.c @@ -1824,9 +1824,28 @@ int vcpu_affinity_domctl(struct domain *d, uint32_t cmd, return ret; } -void domain_update_node_affinity(struct domain *d) +bool update_node_aff_alloc(struct affinity_masks *affinity) { - cpumask_var_t dom_cpumask, dom_cpumask_soft; + if ( !alloc_cpumask_var(&affinity->hard) ) + return false; + if ( !alloc_cpumask_var(&affinity->soft) ) + { + free_cpumask_var(affinity->hard); + return false; + } + + return true; +} + +void update_node_aff_free(struct affinity_masks *affinity) +{ + free_cpumask_var(affinity->soft); + free_cpumask_var(affinity->hard); +} + +void domain_update_node_aff(struct domain *d, struct affinity_masks *affinity) +{ + struct affinity_masks masks = { }; cpumask_t *dom_affinity; const cpumask_t *online; struct sched_unit *unit; @@ -1836,14 +1855,16 @@ void domain_update_node_affinity(struct domain *d) if ( !d->vcpu || !d->vcpu[0] ) return; - if ( !zalloc_cpumask_var(&dom_cpumask) ) - return; - if ( !zalloc_cpumask_var(&dom_cpumask_soft) ) + if ( !affinity ) { - free_cpumask_var(dom_cpumask); - return; + affinity = &masks; + if ( !update_node_aff_alloc(affinity) ) + return; } + cpumask_clear(affinity->hard); + cpumask_clear(affinity->soft); + online = cpupool_domain_master_cpumask(d); spin_lock(&d->node_affinity_lock); @@ -1864,22 +1885,21 @@ void domain_update_node_affinity(struct domain *d) */ for_each_sched_unit ( d, unit ) { - cpumask_or(dom_cpumask, dom_cpumask, unit->cpu_hard_affinity); - cpumask_or(dom_cpumask_soft, dom_cpumask_soft, - unit->cpu_soft_affinity); + cpumask_or(affinity->hard, affinity->hard, unit->cpu_hard_affinity); + cpumask_or(affinity->soft, affinity->soft, unit->cpu_soft_affinity); } /* Filter out non-online cpus */ - cpumask_and(dom_cpumask, dom_cpumask, online); - ASSERT(!cpumask_empty(dom_cpumask)); + cpumask_and(affinity->hard, affinity->hard, online); + ASSERT(!cpumask_empty(affinity->hard)); /* And compute the intersection between hard, online and soft */ - cpumask_and(dom_cpumask_soft, dom_cpumask_soft, dom_cpumask); + cpumask_and(affinity->soft, affinity->soft, affinity->hard); /* * If not empty, the intersection of hard, soft and online is the * narrowest set we want. If empty, we fall back to hard&online. */ - dom_affinity = cpumask_empty(dom_cpumask_soft) ? - dom_cpumask : dom_cpumask_soft; + dom_affinity = cpumask_empty(affinity->soft) ? affinity->hard + : affinity->soft; nodes_clear(d->node_affinity); for_each_cpu ( cpu, dom_affinity ) @@ -1888,8 +1908,8 @@ void domain_update_node_affinity(struct domain *d) spin_unlock(&d->node_affinity_lock); - free_cpumask_var(dom_cpumask_soft); - free_cpumask_var(dom_cpumask); + if ( affinity == &masks ) + update_node_aff_free(affinity); } typedef long ret_t; diff --git a/xen/common/sched/cpupool.c b/xen/common/sched/cpupool.c index 2afe54f54d..58e082eb4c 100644 --- a/xen/common/sched/cpupool.c +++ b/xen/common/sched/cpupool.c @@ -410,6 +410,25 @@ int cpupool_move_domain(struct domain *d, struct cpupool *c) return ret; } +/* Update affinities of all domains in a cpupool. */ +static void cpupool_update_node_affinity(const struct cpupool *c) +{ + struct affinity_masks masks; + struct domain *d; + + if ( !update_node_aff_alloc(&masks) ) + return; + + rcu_read_lock(&domlist_read_lock); + + for_each_domain_in_cpupool(d, c) + domain_update_node_aff(d, &masks); + + rcu_read_unlock(&domlist_read_lock); + + update_node_aff_free(&masks); +} + /* * assign a specific cpu to a cpupool * cpupool_lock must be held @@ -417,7 +436,6 @@ int cpupool_move_domain(struct domain *d, struct cpupool *c) static int cpupool_assign_cpu_locked(struct cpupool *c, unsigned int cpu) { int ret; - struct domain *d; const cpumask_t *cpus; cpus = sched_get_opt_cpumask(c->gran, cpu); @@ -442,12 +460,7 @@ static int cpupool_assign_cpu_locked(struct cpupool *c, unsigned int cpu) rcu_read_unlock(&sched_res_rculock); - rcu_read_lock(&domlist_read_lock); - for_each_domain_in_cpupool(d, c) - { - domain_update_node_affinity(d); - } - rcu_read_unlock(&domlist_read_lock); + cpupool_update_node_affinity(c); return 0; } @@ -456,18 +469,14 @@ static int cpupool_unassign_cpu_finish(struct cpupool *c) { int cpu = cpupool_moving_cpu; const cpumask_t *cpus; - struct domain *d; int ret; if ( c != cpupool_cpu_moving ) return -EADDRNOTAVAIL; - /* - * We need this for scanning the domain list, both in - * cpu_disable_scheduler(), and at the bottom of this function. - */ rcu_read_lock(&domlist_read_lock); ret = cpu_disable_scheduler(cpu); + rcu_read_unlock(&domlist_read_lock); rcu_read_lock(&sched_res_rculock); cpus = get_sched_res(cpu)->cpus; @@ -494,11 +503,7 @@ static int cpupool_unassign_cpu_finish(struct cpupool *c) } rcu_read_unlock(&sched_res_rculock); - for_each_domain_in_cpupool(d, c) - { - domain_update_node_affinity(d); - } - rcu_read_unlock(&domlist_read_lock); + cpupool_update_node_affinity(c); return ret; } diff --git a/xen/common/sched/private.h b/xen/common/sched/private.h index a870320146..38251b1f7b 100644 --- a/xen/common/sched/private.h +++ b/xen/common/sched/private.h @@ -593,6 +593,13 @@ affinity_balance_cpumask(const struct sched_unit *unit, int step, cpumask_copy(mask, unit->cpu_hard_affinity); } +struct affinity_masks { + cpumask_var_t hard; + cpumask_var_t soft; +}; + +bool update_node_aff_alloc(struct affinity_masks *affinity); +void update_node_aff_free(struct affinity_masks *affinity); void sched_rm_cpu(unsigned int cpu); const cpumask_t *sched_get_opt_cpumask(enum sched_gran opt, unsigned int cpu); void schedule_dump(struct cpupool *c); diff --git a/xen/include/xen/sched.h b/xen/include/xen/sched.h index e2b3b6daa3..666264b8c3 100644 --- a/xen/include/xen/sched.h +++ b/xen/include/xen/sched.h @@ -663,8 +663,15 @@ static inline void get_knownalive_domain(struct domain *d) ASSERT(!(atomic_read(&d->refcnt) & DOMAIN_DESTROYED)); } +struct affinity_masks; + int domain_set_node_affinity(struct domain *d, const nodemask_t *affinity); -void domain_update_node_affinity(struct domain *d); +void domain_update_node_aff(struct domain *d, struct affinity_masks *affinity); + +static inline void domain_update_node_affinity(struct domain *d) +{ + domain_update_node_aff(d, NULL); +} /* * To be implemented by each architecture, sanity checking the configuration From patchwork Mon Aug 15 11:04:09 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: =?utf-8?b?SsO8cmdlbiBHcm/Dnw==?= X-Patchwork-Id: 12943441 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id DC1D2C00140 for ; Mon, 15 Aug 2022 11:04:37 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.387258.623447 (Exim 4.92) (envelope-from ) id 1oNXtL-0002FK-JB; Mon, 15 Aug 2022 11:04:19 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 387258.623447; Mon, 15 Aug 2022 11:04:19 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1oNXtL-0002FB-FX; Mon, 15 Aug 2022 11:04:19 +0000 Received: by outflank-mailman (input) for mailman id 387258; Mon, 15 Aug 2022 11:04:18 +0000 Received: from se1-gles-flk1-in.inumbo.com ([94.247.172.50] helo=se1-gles-flk1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1oNXtK-0001af-6J for xen-devel@lists.xenproject.org; Mon, 15 Aug 2022 11:04:18 +0000 Received: from smtp-out2.suse.de (smtp-out2.suse.de [195.135.220.29]) by se1-gles-flk1.inumbo.com (Halon) with ESMTPS id ff2758f8-1c89-11ed-924f-1f966e50362f; Mon, 15 Aug 2022 13:04:16 +0200 (CEST) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by smtp-out2.suse.de (Postfix) with ESMTPS id 6676F20781; Mon, 15 Aug 2022 11:04:13 +0000 (UTC) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by imap2.suse-dmz.suse.de (Postfix) with ESMTPS id 3A11813A93; Mon, 15 Aug 2022 11:04:13 +0000 (UTC) Received: from dovecot-director2.suse.de ([192.168.254.65]) by imap2.suse-dmz.suse.de with ESMTPSA id WFy8DC0o+mLHBgAAMHmgww (envelope-from ); Mon, 15 Aug 2022 11:04:13 +0000 X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: ff2758f8-1c89-11ed-924f-1f966e50362f DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=suse.com; s=susede1; t=1660561453; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=cPpZ0KnROqbgy44j3t9FMKxKlKGwxhvHykExU9F22/A=; b=p4EDn98+z9gItWm5S+kJQSA0laXVnu7hAZ+hAXW0QdImMNChYyGIFKhEHFjoZnCaq0bVgR OkmAJJzbS2tqSkaWl/D7/C5WDRd6u9ueo3UL8bzzseH6vzH+uipTzSl9vgNS7DJDPyJ+ym qTTfgs+PzLHHP/cWR4nOtpCk/blvIqw= From: Juergen Gross To: xen-devel@lists.xenproject.org Cc: Juergen Gross , George Dunlap , Dario Faggioli Subject: [PATCH v2 2/3] xen/sched: carve out memory allocation and freeing from schedule_cpu_rm() Date: Mon, 15 Aug 2022 13:04:09 +0200 Message-Id: <20220815110410.19872-3-jgross@suse.com> X-Mailer: git-send-email 2.35.3 In-Reply-To: <20220815110410.19872-1-jgross@suse.com> References: <20220815110410.19872-1-jgross@suse.com> MIME-Version: 1.0 In order to prepare not allocating or freeing memory from schedule_cpu_rm(), move this functionality to dedicated functions. For now call those functions from schedule_cpu_rm(). No change of behavior expected. Signed-off-by: Juergen Gross --- V2: - add const (Jan Beulich) - use "unsigned int" for loop index (Jan Beulich) - use xmalloc_flex_struct() (Jan Beulich) - use XFREE() (Jan Beulich) - hold rcu lock longer (Jan Beulich) - add ASSERT() (Jan Beulich) Signed-off-by: Juergen Gross --- xen/common/sched/core.c | 133 +++++++++++++++++++++---------------- xen/common/sched/private.h | 9 +++ 2 files changed, 86 insertions(+), 56 deletions(-) diff --git a/xen/common/sched/core.c b/xen/common/sched/core.c index 085a9dd335..d0b6513b6f 100644 --- a/xen/common/sched/core.c +++ b/xen/common/sched/core.c @@ -3237,6 +3237,65 @@ out: return ret; } +static struct cpu_rm_data *schedule_cpu_rm_alloc(unsigned int cpu) +{ + struct cpu_rm_data *data; + const struct sched_resource *sr; + unsigned int idx; + + rcu_read_lock(&sched_res_rculock); + + sr = get_sched_res(cpu); + data = xmalloc_flex_struct(struct cpu_rm_data, sr, sr->granularity - 1); + if ( !data ) + goto out; + + data->old_ops = sr->scheduler; + data->vpriv_old = idle_vcpu[cpu]->sched_unit->priv; + data->ppriv_old = sr->sched_priv; + + for ( idx = 0; idx < sr->granularity - 1; idx++ ) + { + data->sr[idx] = sched_alloc_res(); + if ( data->sr[idx] ) + { + data->sr[idx]->sched_unit_idle = sched_alloc_unit_mem(); + if ( !data->sr[idx]->sched_unit_idle ) + { + sched_res_free(&data->sr[idx]->rcu); + data->sr[idx] = NULL; + } + } + if ( !data->sr[idx] ) + { + while ( idx > 0 ) + sched_res_free(&data->sr[--idx]->rcu); + XFREE(data); + goto out; + } + + data->sr[idx]->curr = data->sr[idx]->sched_unit_idle; + data->sr[idx]->scheduler = &sched_idle_ops; + data->sr[idx]->granularity = 1; + + /* We want the lock not to change when replacing the resource. */ + data->sr[idx]->schedule_lock = sr->schedule_lock; + } + + out: + rcu_read_unlock(&sched_res_rculock); + + return data; +} + +static void schedule_cpu_rm_free(struct cpu_rm_data *mem, unsigned int cpu) +{ + sched_free_udata(mem->old_ops, mem->vpriv_old); + sched_free_pdata(mem->old_ops, mem->ppriv_old, cpu); + + xfree(mem); +} + /* * Remove a pCPU from its cpupool. Its scheduler becomes &sched_idle_ops * (the idle scheduler). @@ -3245,53 +3304,23 @@ out: */ int schedule_cpu_rm(unsigned int cpu) { - void *ppriv_old, *vpriv_old; - struct sched_resource *sr, **sr_new = NULL; + struct sched_resource *sr; + struct cpu_rm_data *data; struct sched_unit *unit; - struct scheduler *old_ops; spinlock_t *old_lock; unsigned long flags; - int idx, ret = -ENOMEM; + int idx = 0; unsigned int cpu_iter; + data = schedule_cpu_rm_alloc(cpu); + if ( !data ) + return -ENOMEM; + rcu_read_lock(&sched_res_rculock); sr = get_sched_res(cpu); - old_ops = sr->scheduler; - if ( sr->granularity > 1 ) - { - sr_new = xmalloc_array(struct sched_resource *, sr->granularity - 1); - if ( !sr_new ) - goto out; - for ( idx = 0; idx < sr->granularity - 1; idx++ ) - { - sr_new[idx] = sched_alloc_res(); - if ( sr_new[idx] ) - { - sr_new[idx]->sched_unit_idle = sched_alloc_unit_mem(); - if ( !sr_new[idx]->sched_unit_idle ) - { - sched_res_free(&sr_new[idx]->rcu); - sr_new[idx] = NULL; - } - } - if ( !sr_new[idx] ) - { - for ( idx--; idx >= 0; idx-- ) - sched_res_free(&sr_new[idx]->rcu); - goto out; - } - sr_new[idx]->curr = sr_new[idx]->sched_unit_idle; - sr_new[idx]->scheduler = &sched_idle_ops; - sr_new[idx]->granularity = 1; - - /* We want the lock not to change when replacing the resource. */ - sr_new[idx]->schedule_lock = sr->schedule_lock; - } - } - - ret = 0; + ASSERT(sr->granularity); ASSERT(sr->cpupool != NULL); ASSERT(cpumask_test_cpu(cpu, &cpupool_free_cpus)); ASSERT(!cpumask_test_cpu(cpu, sr->cpupool->cpu_valid)); @@ -3299,10 +3328,6 @@ int schedule_cpu_rm(unsigned int cpu) /* See comment in schedule_cpu_add() regarding lock switching. */ old_lock = pcpu_schedule_lock_irqsave(cpu, &flags); - vpriv_old = idle_vcpu[cpu]->sched_unit->priv; - ppriv_old = sr->sched_priv; - - idx = 0; for_each_cpu ( cpu_iter, sr->cpus ) { per_cpu(sched_res_idx, cpu_iter) = 0; @@ -3316,27 +3341,27 @@ int schedule_cpu_rm(unsigned int cpu) else { /* Initialize unit. */ - unit = sr_new[idx]->sched_unit_idle; - unit->res = sr_new[idx]; + unit = data->sr[idx]->sched_unit_idle; + unit->res = data->sr[idx]; unit->is_running = true; sched_unit_add_vcpu(unit, idle_vcpu[cpu_iter]); sched_domain_insert_unit(unit, idle_vcpu[cpu_iter]->domain); /* Adjust cpu masks of resources (old and new). */ cpumask_clear_cpu(cpu_iter, sr->cpus); - cpumask_set_cpu(cpu_iter, sr_new[idx]->cpus); + cpumask_set_cpu(cpu_iter, data->sr[idx]->cpus); cpumask_set_cpu(cpu_iter, &sched_res_mask); /* Init timer. */ - init_timer(&sr_new[idx]->s_timer, s_timer_fn, NULL, cpu_iter); + init_timer(&data->sr[idx]->s_timer, s_timer_fn, NULL, cpu_iter); /* Last resource initializations and insert resource pointer. */ - sr_new[idx]->master_cpu = cpu_iter; - set_sched_res(cpu_iter, sr_new[idx]); + data->sr[idx]->master_cpu = cpu_iter; + set_sched_res(cpu_iter, data->sr[idx]); /* Last action: set the new lock pointer. */ smp_mb(); - sr_new[idx]->schedule_lock = &sched_free_cpu_lock; + data->sr[idx]->schedule_lock = &sched_free_cpu_lock; idx++; } @@ -3352,16 +3377,12 @@ int schedule_cpu_rm(unsigned int cpu) /* _Not_ pcpu_schedule_unlock(): schedule_lock may have changed! */ spin_unlock_irqrestore(old_lock, flags); - sched_deinit_pdata(old_ops, ppriv_old, cpu); - - sched_free_udata(old_ops, vpriv_old); - sched_free_pdata(old_ops, ppriv_old, cpu); + sched_deinit_pdata(data->old_ops, data->ppriv_old, cpu); -out: rcu_read_unlock(&sched_res_rculock); - xfree(sr_new); + schedule_cpu_rm_free(data, cpu); - return ret; + return 0; } struct scheduler *scheduler_get_default(void) diff --git a/xen/common/sched/private.h b/xen/common/sched/private.h index 38251b1f7b..601d639699 100644 --- a/xen/common/sched/private.h +++ b/xen/common/sched/private.h @@ -600,6 +600,15 @@ struct affinity_masks { bool update_node_aff_alloc(struct affinity_masks *affinity); void update_node_aff_free(struct affinity_masks *affinity); + +/* Memory allocation related data for schedule_cpu_rm(). */ +struct cpu_rm_data { + const struct scheduler *old_ops; + void *ppriv_old; + void *vpriv_old; + struct sched_resource *sr[]; +}; + void sched_rm_cpu(unsigned int cpu); const cpumask_t *sched_get_opt_cpumask(enum sched_gran opt, unsigned int cpu); void schedule_dump(struct cpupool *c); From patchwork Mon Aug 15 11:04:10 2022 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: =?utf-8?b?SsO8cmdlbiBHcm/Dnw==?= X-Patchwork-Id: 12943444 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from lists.xenproject.org (lists.xenproject.org [192.237.175.120]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.lore.kernel.org (Postfix) with ESMTPS id E9024C25B0E for ; Mon, 15 Aug 2022 11:04:38 +0000 (UTC) Received: from list by lists.xenproject.org with outflank-mailman.387257.623430 (Exim 4.92) (envelope-from ) id 1oNXtK-0001nZ-8T; Mon, 15 Aug 2022 11:04:18 +0000 X-Outflank-Mailman: Message body and most headers restored to incoming version Received: by outflank-mailman (output) from mailman id 387257.623430; Mon, 15 Aug 2022 11:04:18 +0000 Received: from localhost ([127.0.0.1] helo=lists.xenproject.org) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1oNXtK-0001mp-1o; Mon, 15 Aug 2022 11:04:18 +0000 Received: by outflank-mailman (input) for mailman id 387257; Mon, 15 Aug 2022 11:04:16 +0000 Received: from se1-gles-sth1-in.inumbo.com ([159.253.27.254] helo=se1-gles-sth1.inumbo.com) by lists.xenproject.org with esmtp (Exim 4.92) (envelope-from ) id 1oNXtI-0001TU-BY for xen-devel@lists.xenproject.org; Mon, 15 Aug 2022 11:04:16 +0000 Received: from smtp-out1.suse.de (smtp-out1.suse.de [195.135.220.28]) by se1-gles-sth1.inumbo.com (Halon) with ESMTPS id ff473844-1c89-11ed-bd2e-47488cf2e6aa; Mon, 15 Aug 2022 13:04:13 +0200 (CEST) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by smtp-out1.suse.de (Postfix) with ESMTPS id A12EA33B41; Mon, 15 Aug 2022 11:04:13 +0000 (UTC) Received: from imap2.suse-dmz.suse.de (imap2.suse-dmz.suse.de [192.168.254.74]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature ECDSA (P-521) server-digest SHA512) (No client certificate requested) by imap2.suse-dmz.suse.de (Postfix) with ESMTPS id 6E85E13A93; Mon, 15 Aug 2022 11:04:13 +0000 (UTC) Received: from dovecot-director2.suse.de ([192.168.254.65]) by imap2.suse-dmz.suse.de with ESMTPSA id 0FeqGS0o+mLHBgAAMHmgww (envelope-from ); Mon, 15 Aug 2022 11:04:13 +0000 X-BeenThere: xen-devel@lists.xenproject.org List-Id: Xen developer discussion List-Unsubscribe: , List-Post: List-Help: List-Subscribe: , Errors-To: xen-devel-bounces@lists.xenproject.org Precedence: list Sender: "Xen-devel" X-Inumbo-ID: ff473844-1c89-11ed-bd2e-47488cf2e6aa DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=suse.com; s=susede1; t=1660561453; h=from:from:reply-to:date:date:message-id:message-id:to:to:cc:cc: mime-version:mime-version: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=oK+J2+KdcgfEZPELpjligxuZYoDpVDFh6cULumqb2ak=; b=K2nxOL7bI3qzmuvr5d9pRJ55fBdrPURYd8EC0wm902mSbcA0W+eYhcaKIl61C+RMyPVRZx 3WCcmrxcmxFlTQYGbYVK7Ljrn0q2Q83Wj068qf7jW5hY9UYUYG7Rg0yrKBs5ztfXZpQeEl ZXfxuWgqchLrp+Ie4DAXHC2vID/GUeE= From: Juergen Gross To: xen-devel@lists.xenproject.org Cc: Juergen Gross , George Dunlap , Dario Faggioli , Gao Ruifeng Subject: [PATCH v2 3/3] xen/sched: fix cpu hotplug Date: Mon, 15 Aug 2022 13:04:10 +0200 Message-Id: <20220815110410.19872-4-jgross@suse.com> X-Mailer: git-send-email 2.35.3 In-Reply-To: <20220815110410.19872-1-jgross@suse.com> References: <20220815110410.19872-1-jgross@suse.com> MIME-Version: 1.0 Cpu cpu unplugging is calling schedule_cpu_rm() via stop_machine_run() with interrupts disabled, thus any memory allocation or freeing must be avoided. Since commit 5047cd1d5dea ("xen/common: Use enhanced ASSERT_ALLOC_CONTEXT in xmalloc()") this restriction is being enforced via an assertion, which will now fail. Before that commit cpu unplugging in normal configurations was working just by chance as only the cpu performing schedule_cpu_rm() was doing active work. With core scheduling enabled, however, failures could result from memory allocations not being properly propagated to other cpus' TLBs. Fix this mess by allocating needed memory before entering stop_machine_run() and freeing any memory only after having finished stop_machine_run(). Fixes: 1ec410112cdd ("xen/sched: support differing granularity in schedule_cpu_[add/rm]()") Reported-by: Gao Ruifeng Signed-off-by: Juergen Gross Reviewed-by: Jan Beulich --- V2: - move affinity mask allocation into schedule_cpu_rm_alloc() (Jan Beulich) --- xen/common/sched/core.c | 27 +++++++++++---- xen/common/sched/cpupool.c | 68 +++++++++++++++++++++++++++++--------- xen/common/sched/private.h | 5 ++- 3 files changed, 78 insertions(+), 22 deletions(-) diff --git a/xen/common/sched/core.c b/xen/common/sched/core.c index d0b6513b6f..ec09c8e5a9 100644 --- a/xen/common/sched/core.c +++ b/xen/common/sched/core.c @@ -3237,7 +3237,7 @@ out: return ret; } -static struct cpu_rm_data *schedule_cpu_rm_alloc(unsigned int cpu) +struct cpu_rm_data *schedule_cpu_rm_alloc(unsigned int cpu, bool aff_alloc) { struct cpu_rm_data *data; const struct sched_resource *sr; @@ -3250,6 +3250,17 @@ static struct cpu_rm_data *schedule_cpu_rm_alloc(unsigned int cpu) if ( !data ) goto out; + if ( aff_alloc ) + { + if ( !update_node_aff_alloc(&data->affinity) ) + { + XFREE(data); + goto out; + } + } + else + memset(&data->affinity, 0, sizeof(data->affinity)); + data->old_ops = sr->scheduler; data->vpriv_old = idle_vcpu[cpu]->sched_unit->priv; data->ppriv_old = sr->sched_priv; @@ -3270,6 +3281,7 @@ static struct cpu_rm_data *schedule_cpu_rm_alloc(unsigned int cpu) { while ( idx > 0 ) sched_res_free(&data->sr[--idx]->rcu); + update_node_aff_free(&data->affinity); XFREE(data); goto out; } @@ -3288,10 +3300,11 @@ static struct cpu_rm_data *schedule_cpu_rm_alloc(unsigned int cpu) return data; } -static void schedule_cpu_rm_free(struct cpu_rm_data *mem, unsigned int cpu) +void schedule_cpu_rm_free(struct cpu_rm_data *mem, unsigned int cpu) { sched_free_udata(mem->old_ops, mem->vpriv_old); sched_free_pdata(mem->old_ops, mem->ppriv_old, cpu); + update_node_aff_free(&mem->affinity); xfree(mem); } @@ -3302,17 +3315,18 @@ static void schedule_cpu_rm_free(struct cpu_rm_data *mem, unsigned int cpu) * The cpu is already marked as "free" and not valid any longer for its * cpupool. */ -int schedule_cpu_rm(unsigned int cpu) +int schedule_cpu_rm(unsigned int cpu, struct cpu_rm_data *data) { struct sched_resource *sr; - struct cpu_rm_data *data; struct sched_unit *unit; spinlock_t *old_lock; unsigned long flags; int idx = 0; unsigned int cpu_iter; + bool freemem = !data; - data = schedule_cpu_rm_alloc(cpu); + if ( !data ) + data = schedule_cpu_rm_alloc(cpu, false); if ( !data ) return -ENOMEM; @@ -3380,7 +3394,8 @@ int schedule_cpu_rm(unsigned int cpu) sched_deinit_pdata(data->old_ops, data->ppriv_old, cpu); rcu_read_unlock(&sched_res_rculock); - schedule_cpu_rm_free(data, cpu); + if ( freemem ) + schedule_cpu_rm_free(data, cpu); return 0; } diff --git a/xen/common/sched/cpupool.c b/xen/common/sched/cpupool.c index 58e082eb4c..2506861e4f 100644 --- a/xen/common/sched/cpupool.c +++ b/xen/common/sched/cpupool.c @@ -411,22 +411,28 @@ int cpupool_move_domain(struct domain *d, struct cpupool *c) } /* Update affinities of all domains in a cpupool. */ -static void cpupool_update_node_affinity(const struct cpupool *c) +static void cpupool_update_node_affinity(const struct cpupool *c, + struct affinity_masks *masks) { - struct affinity_masks masks; + struct affinity_masks local_masks; struct domain *d; - if ( !update_node_aff_alloc(&masks) ) - return; + if ( !masks ) + { + if ( !update_node_aff_alloc(&local_masks) ) + return; + masks = &local_masks; + } rcu_read_lock(&domlist_read_lock); for_each_domain_in_cpupool(d, c) - domain_update_node_aff(d, &masks); + domain_update_node_aff(d, masks); rcu_read_unlock(&domlist_read_lock); - update_node_aff_free(&masks); + if ( masks == &local_masks ) + update_node_aff_free(masks); } /* @@ -460,15 +466,17 @@ static int cpupool_assign_cpu_locked(struct cpupool *c, unsigned int cpu) rcu_read_unlock(&sched_res_rculock); - cpupool_update_node_affinity(c); + cpupool_update_node_affinity(c, NULL); return 0; } -static int cpupool_unassign_cpu_finish(struct cpupool *c) +static int cpupool_unassign_cpu_finish(struct cpupool *c, + struct cpu_rm_data *mem) { int cpu = cpupool_moving_cpu; const cpumask_t *cpus; + struct affinity_masks *masks = mem ? &mem->affinity : NULL; int ret; if ( c != cpupool_cpu_moving ) @@ -491,7 +499,7 @@ static int cpupool_unassign_cpu_finish(struct cpupool *c) */ if ( !ret ) { - ret = schedule_cpu_rm(cpu); + ret = schedule_cpu_rm(cpu, mem); if ( ret ) cpumask_andnot(&cpupool_free_cpus, &cpupool_free_cpus, cpus); else @@ -503,7 +511,7 @@ static int cpupool_unassign_cpu_finish(struct cpupool *c) } rcu_read_unlock(&sched_res_rculock); - cpupool_update_node_affinity(c); + cpupool_update_node_affinity(c, masks); return ret; } @@ -567,7 +575,7 @@ static long cf_check cpupool_unassign_cpu_helper(void *info) cpupool_cpu_moving->cpupool_id, cpupool_moving_cpu); spin_lock(&cpupool_lock); - ret = cpupool_unassign_cpu_finish(c); + ret = cpupool_unassign_cpu_finish(c, NULL); spin_unlock(&cpupool_lock); debugtrace_printk("cpupool_unassign_cpu ret=%ld\n", ret); @@ -714,7 +722,7 @@ static int cpupool_cpu_add(unsigned int cpu) * This function is called in stop_machine context, so we can be sure no * non-idle vcpu is active on the system. */ -static void cpupool_cpu_remove(unsigned int cpu) +static void cpupool_cpu_remove(unsigned int cpu, struct cpu_rm_data *mem) { int ret; @@ -722,7 +730,7 @@ static void cpupool_cpu_remove(unsigned int cpu) if ( !cpumask_test_cpu(cpu, &cpupool_free_cpus) ) { - ret = cpupool_unassign_cpu_finish(cpupool0); + ret = cpupool_unassign_cpu_finish(cpupool0, mem); BUG_ON(ret); } cpumask_clear_cpu(cpu, &cpupool_free_cpus); @@ -788,7 +796,7 @@ static void cpupool_cpu_remove_forced(unsigned int cpu) { ret = cpupool_unassign_cpu_start(c, master_cpu); BUG_ON(ret); - ret = cpupool_unassign_cpu_finish(c); + ret = cpupool_unassign_cpu_finish(c, NULL); BUG_ON(ret); } } @@ -1008,10 +1016,21 @@ static int cf_check cpu_callback( { unsigned int cpu = (unsigned long)hcpu; int rc = 0; + static struct cpu_rm_data *mem; switch ( action ) { case CPU_DOWN_FAILED: + if ( system_state <= SYS_STATE_active ) + { + if ( mem ) + { + schedule_cpu_rm_free(mem, cpu); + mem = NULL; + } + rc = cpupool_cpu_add(cpu); + } + break; case CPU_ONLINE: if ( system_state <= SYS_STATE_active ) rc = cpupool_cpu_add(cpu); @@ -1019,12 +1038,31 @@ static int cf_check cpu_callback( case CPU_DOWN_PREPARE: /* Suspend/Resume don't change assignments of cpus to cpupools. */ if ( system_state <= SYS_STATE_active ) + { rc = cpupool_cpu_remove_prologue(cpu); + if ( !rc ) + { + ASSERT(!mem); + mem = schedule_cpu_rm_alloc(cpu, true); + rc = mem ? 0 : -ENOMEM; + } + } break; case CPU_DYING: /* Suspend/Resume don't change assignments of cpus to cpupools. */ if ( system_state <= SYS_STATE_active ) - cpupool_cpu_remove(cpu); + { + ASSERT(mem); + cpupool_cpu_remove(cpu, mem); + } + break; + case CPU_DEAD: + if ( system_state <= SYS_STATE_active ) + { + ASSERT(mem); + schedule_cpu_rm_free(mem, cpu); + mem = NULL; + } break; case CPU_RESUME_FAILED: cpupool_cpu_remove_forced(cpu); diff --git a/xen/common/sched/private.h b/xen/common/sched/private.h index 601d639699..cc7a6cb571 100644 --- a/xen/common/sched/private.h +++ b/xen/common/sched/private.h @@ -603,6 +603,7 @@ void update_node_aff_free(struct affinity_masks *affinity); /* Memory allocation related data for schedule_cpu_rm(). */ struct cpu_rm_data { + struct affinity_masks affinity; const struct scheduler *old_ops; void *ppriv_old; void *vpriv_old; @@ -617,7 +618,9 @@ struct scheduler *scheduler_alloc(unsigned int sched_id); void scheduler_free(struct scheduler *sched); int cpu_disable_scheduler(unsigned int cpu); int schedule_cpu_add(unsigned int cpu, struct cpupool *c); -int schedule_cpu_rm(unsigned int cpu); +struct cpu_rm_data *schedule_cpu_rm_alloc(unsigned int cpu, bool aff_alloc); +void schedule_cpu_rm_free(struct cpu_rm_data *mem, unsigned int cpu); +int schedule_cpu_rm(unsigned int cpu, struct cpu_rm_data *mem); int sched_move_domain(struct domain *d, struct cpupool *c); struct cpupool *cpupool_get_by_id(unsigned int poolid); void cpupool_put(struct cpupool *pool);