Received: from smtp.kernel.org (aws-us-west-2-korg-mail-alma10-1.taild15c8.ts.net [100.103.45.18]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 0E6FB4749CD for ; Thu, 20 Aug 2026 14:21:04 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=100.103.45.18 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787235667; cv=none; b=FEwQGOlKEi7gTVLSbh7TovvRiOwT2TOU956wu2dtX1PlRsPmVG8A3YFzL7mPntGrTvfrIDKdMGCRgBu+4bTBANoN3cqN0Dbpa9FMTasXkfWwdNpW+yqGywF8q3JJ139zZM6h3/ZIyGlHJCHz/gywuCScaHStw6e7paWwabwjAIg= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1787235667; c=relaxed/simple; bh=2fe6/7eiHicbbA/ix/MRAeBQ4apQGOmaLpxprrcuk/k=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=WCwzlQ3s+x6mWGJeZQ1xr0cyqorudxv/bKtQMCvT2Olq7qZero46vUvyzlb58n5NEL4DDZF6roZEjI/l4vuX9kT69ltWPAXk9vn45ChLefcl80RCF5GJmoUl5e4JNAg/SFa44zQbjQ5ES2J1r1L95/9ds3broiZb/vmvA2Cygs0= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b=DQjG+Lp9; arc=none smtp.client-ip=100.103.45.18 Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=kernel.org header.i=@kernel.org header.b="DQjG+Lp9" Received: by smtp.kernel.org (Postfix) with ESMTPSA id A48E91F000E9; Thu, 20 Aug 2026 14:21:03 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=kernel.org; s=k20260515; t=1787235664; bh=wKOpy0Ho8wpxP82OA8Sqjb0mfbSN++LnpuYjwxwt9ac=; h=From:To:Cc:Subject:Date:In-Reply-To:References; b=DQjG+Lp9cK4qGWmEaVUaXHHnCJuLrh726yQMUOturLogqCxuKWU6UozfbRMeoHg9Z ZqRXAwL9l+yncLigt+SAsRmUJ/CsJL2/oh6k3qVj0994cw4FqR/QSbxaO/7UXJUboV 353GjqdGcxNEFE0zRxL6FzHyp7CGHP3A+bZNHM+q0qSW+PAgGNeRfUgLHgQMKeUG4n 4f4idobT1meYpFxKxuW4lJp9W7aOMNqXPwNfDYXTnnP+lI/60Ek/sGOT5Uh2P760SW aTegzMsgVJYc1bAc4qdP5NCPod8j7M7m5PvMluf6qTH/XCvEwxHPc+zMpK/DA4hbGv ZfSVVWNEKIX7Q== From: Sasha Levin To: stable@vger.kernel.org Cc: Peter Zijlstra , Ravi Bangoria , Sasha Levin Subject: [PATCH 6.12.y 1/2] perf: Unify perf_event_free_task() / perf_event_exit_task_context() Date: Thu, 20 Aug 2026 10:21:01 -0400 Message-ID: <20260820142102.1884457-1-sashal@kernel.org> X-Mailer: git-send-email 2.53.0 In-Reply-To: <2026081729-cruelty-pouch-23ea@gregkh> References: <2026081729-cruelty-pouch-23ea@gregkh> Precedence: bulk X-Mailing-List: stable@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Transfer-Encoding: 8bit From: Peter Zijlstra [ Upstream commit 90661365021a6d0d7f3a2c5046ebe33e4df53b92 ] Both perf_event_free_task() and perf_event_exit_task_context() are very similar, except perf_event_exit_task_context() is a little more generic / makes less assumptions. Signed-off-by: Peter Zijlstra (Intel) Reviewed-by: Ravi Bangoria Link: https://lkml.kernel.org/r/20250307193723.274039710@infradead.org Stable-dep-of: 42c5ca1f0a28 ("perf/core: Fix group leader use-after-free after sibling detach") Signed-off-by: Sasha Levin --- kernel/events/core.c | 93 ++++++++++++-------------------------------- 1 file changed, 25 insertions(+), 68 deletions(-) diff --git a/kernel/events/core.c b/kernel/events/core.c index da77f856e1c83..8e2629d5305e6 100644 --- a/kernel/events/core.c +++ b/kernel/events/core.c @@ -13522,13 +13522,11 @@ perf_event_exit_event(struct perf_event *event, perf_event_wakeup(event); } -static void perf_event_exit_task_context(struct task_struct *child) +static void perf_event_exit_task_context(struct task_struct *child, bool exit) { struct perf_event_context *child_ctx, *clone_ctx = NULL; struct perf_event *child_event, *next; - WARN_ON_ONCE(child != current); - child_ctx = perf_pin_task_context(child); if (!child_ctx) return; @@ -13551,7 +13549,8 @@ static void perf_event_exit_task_context(struct task_struct *child) * in. */ raw_spin_lock_irq(&child_ctx->lock); - task_ctx_sched_out(child_ctx, NULL, EVENT_ALL); + if (exit) + task_ctx_sched_out(child_ctx, NULL, EVENT_ALL); /* * Now that the context is inactive, destroy the task <-> ctx relation @@ -13560,7 +13559,7 @@ static void perf_event_exit_task_context(struct task_struct *child) RCU_INIT_POINTER(child->perf_event_ctxp, NULL); put_ctx(child_ctx); /* cannot be last */ WRITE_ONCE(child_ctx->task, TASK_TOMBSTONE); - put_task_struct(current); /* cannot be last */ + put_task_struct(child); /* cannot be last */ clone_ctx = unclone_ctx(child_ctx); raw_spin_unlock_irq(&child_ctx->lock); @@ -13573,13 +13572,31 @@ static void perf_event_exit_task_context(struct task_struct *child) * won't get any samples after PERF_RECORD_EXIT. We can however still * get a few PERF_RECORD_READ events. */ - perf_event_task(child, child_ctx, 0); + if (exit) + perf_event_task(child, child_ctx, 0); list_for_each_entry_safe(child_event, next, &child_ctx->event_list, event_entry) perf_event_exit_event(child_event, child_ctx, 0); mutex_unlock(&child_ctx->mutex); + if (!exit) { + /* + * perf_event_release_kernel() could still have a reference on + * this context. In that case we must wait for these events to + * have been freed (in particular all their references to this + * task must've been dropped). + * + * Without this copy_process() will unconditionally free this + * task (irrespective of its reference count) and + * _free_event()'s put_task_struct(event->hw.target) will be a + * use-after-free. + * + * Wait for all events to drop their context reference. + */ + wait_var_event(&child_ctx->refcount, + refcount_read(&child_ctx->refcount) == 1); + } put_ctx(child_ctx); } @@ -13607,7 +13624,7 @@ void perf_event_exit_task(struct task_struct *child) } mutex_unlock(&child->perf_event_mutex); - perf_event_exit_task_context(child); + perf_event_exit_task_context(child, true); /* * The perf_event_exit_task_context calls perf_event_task @@ -13618,25 +13635,6 @@ void perf_event_exit_task(struct task_struct *child) perf_event_task(child, NULL, 0); } -static void perf_free_event(struct perf_event *event, - struct perf_event_context *ctx) -{ - struct perf_event *parent = event->parent; - - if (WARN_ON_ONCE(!parent)) - return; - - mutex_lock(&parent->child_mutex); - list_del_init(&event->child_list); - mutex_unlock(&parent->child_mutex); - - raw_spin_lock_irq(&ctx->lock); - perf_group_detach(event); - list_del_event(event, ctx); - raw_spin_unlock_irq(&ctx->lock); - put_event(event); -} - /* * Free a context as created by inheritance by perf_event_init_task() below, * used by fork() in case of fail. @@ -13646,48 +13644,7 @@ static void perf_free_event(struct perf_event *event, */ void perf_event_free_task(struct task_struct *task) { - struct perf_event_context *ctx; - struct perf_event *event, *tmp; - - ctx = rcu_access_pointer(task->perf_event_ctxp); - if (!ctx) - return; - - mutex_lock(&ctx->mutex); - raw_spin_lock_irq(&ctx->lock); - /* - * Destroy the task <-> ctx relation and mark the context dead. - * - * This is important because even though the task hasn't been - * exposed yet the context has been (through child_list). - */ - RCU_INIT_POINTER(task->perf_event_ctxp, NULL); - WRITE_ONCE(ctx->task, TASK_TOMBSTONE); - put_task_struct(task); /* cannot be last */ - raw_spin_unlock_irq(&ctx->lock); - - - list_for_each_entry_safe(event, tmp, &ctx->event_list, event_entry) - perf_free_event(event, ctx); - - mutex_unlock(&ctx->mutex); - - /* - * perf_event_release_kernel() could've stolen some of our - * child events and still have them on its free_list. In that - * case we must wait for these events to have been freed (in - * particular all their references to this task must've been - * dropped). - * - * Without this copy_process() will unconditionally free this - * task (irrespective of its reference count) and - * _free_event()'s put_task_struct(event->hw.target) will be a - * use-after-free. - * - * Wait for all events to drop their context reference. - */ - wait_var_event(&ctx->refcount, refcount_read(&ctx->refcount) == 1); - put_ctx(ctx); /* must be last */ + perf_event_exit_task_context(task, false); } void perf_event_delayed_put(struct task_struct *task) -- 2.53.0