From patchwork Tue Aug 27 18:02:30 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexander Aring X-Patchwork-Id: 13779965 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) (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 639141D278E for ; Tue, 27 Aug 2024 18:03:14 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=170.10.133.124 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781795; cv=none; b=GZ/Q8qNTKCjpUgW/n8Km4xl8YhpBR9PFGckWQcJ+dlKw1Dz1MV6FniiT4Yqc8J5fD+krRk+5Pkjkdw+P0JUHyUwzo5+ZghmPPag8EfTtVquRyTtUR9qNbMOK3K3B5F2oeuGg9mAqdRwjsnXErteigJIY0yERjNoziJKpA054ZmM= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781795; c=relaxed/simple; bh=CvsA/zhC/97Chjcmlq3gS0bR5udsf69Y4TjGQLAJrSE=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=BMDtwF0mV4PNO0PemdzxQE+YWhZZfMw838em0e2JeiYQYdclRnrg2VYE3vOvbY0EEw8zupXjQLF41lvLNR5Ib+aQJCzbIp8H5MI8yPsNMjWUP4nZJQDke5kl1V6i6Peu+XuCdGKQwiylbhWZzq1pt83ci1fCRFRjFx4hHdJ20Zw= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b=dINVc/k0; arc=none smtp.client-ip=170.10.133.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="dINVc/k0" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1724781793; h=from:from:reply-to:subject:subject: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=LUKN0kpZX28r3L3pwjaCmvRqCu3JTbKNbIZnDThKW3Q=; b=dINVc/k0oWr7YMr4jQTKJTvJwKEBYF8W/MQABuVdzcrfrTo7cRF6bpWXReWbfwfNPdIh58 hkwnEtJDL8jv7DYvH7h+M4jaGbOYa2GPGVV8WQ0GmMsKgr2EpmHJ4eV6Ybn2mhJnCSluVs KNRjoM4Wj550Z5qrHL8Qj3gxSm8TlIQ= Received: from mx-prod-mc-03.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-474-vtZYPNwkNImZ-BdRc3ea6A-1; Tue, 27 Aug 2024 14:03:10 -0400 X-MC-Unique: vtZYPNwkNImZ-BdRc3ea6A-1 Received: from mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.15]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-03.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id 18DA21955BEF; Tue, 27 Aug 2024 18:03:03 +0000 (UTC) Received: from fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com (fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com [10.6.24.150]) by mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id 0C9C61955DD6; Tue, 27 Aug 2024 18:02:58 +0000 (UTC) From: Alexander Aring To: teigland@redhat.com Cc: gfs2@lists.linux.dev, song@kernel.org, yukuai3@huawei.com, agruenba@redhat.com, mark@fasheh.com, jlbec@evilplan.org, joseph.qi@linux.alibaba.com, gregkh@linuxfoundation.org, rafael@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, ocfs2-devel@lists.linux.dev, netdev@vger.kernel.org, vvidic@valentin-vidic.from.hr, heming.zhao@suse.com, lucien.xin@gmail.com, paulmck@kernel.org, rcu@vger.kernel.org, juri.lelli@redhat.com, williams@redhat.com, aahringo@redhat.com Subject: [RFC 1/7] dlm: fix possible lkb_resource null dereference Date: Tue, 27 Aug 2024 14:02:30 -0400 Message-ID: <20240827180236.316946-2-aahringo@redhat.com> In-Reply-To: <20240827180236.316946-1-aahringo@redhat.com> References: <20240827180236.316946-1-aahringo@redhat.com> Precedence: bulk X-Mailing-List: rcu@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.0 on 10.30.177.15 This patch fixes a possible null pointer dereference when this function is called from request_lock() as lkb->lkb_resource is not assigned yet, only after validate_lock_args() by calling attach_lkb(). Another issue is that a resource name could be a non printable bytearray and we cannot assume to be ASCII coded. In this patch we just drop the printout of the resource name, the lkb id is enough to make a possible connection to a resource name if this exists. Signed-off-by: Alexander Aring --- fs/dlm/lock.c | 10 ++++------ 1 file changed, 4 insertions(+), 6 deletions(-) diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index 0e8d2b9bf908..121d2976986b 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -2861,16 +2861,14 @@ static int validate_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, case -EINVAL: /* annoy the user because dlm usage is wrong */ WARN_ON(1); - log_error(ls, "%s %d %x %x %x %d %d %s", __func__, + log_error(ls, "%s %d %x %x %x %d %d", __func__, rv, lkb->lkb_id, dlm_iflags_val(lkb), args->flags, - lkb->lkb_status, lkb->lkb_wait_type, - lkb->lkb_resource->res_name); + lkb->lkb_status, lkb->lkb_wait_type); break; default: - log_debug(ls, "%s %d %x %x %x %d %d %s", __func__, + log_debug(ls, "%s %d %x %x %x %d %d", __func__, rv, lkb->lkb_id, dlm_iflags_val(lkb), args->flags, - lkb->lkb_status, lkb->lkb_wait_type, - lkb->lkb_resource->res_name); + lkb->lkb_status, lkb->lkb_wait_type); break; } From patchwork Tue Aug 27 18:02:31 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexander Aring X-Patchwork-Id: 13779966 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) (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 92E041D2F5A for ; Tue, 27 Aug 2024 18:03:14 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=170.10.133.124 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781796; cv=none; b=J5vo2sRlBGJvEtEU87tKWMQF836DdOJA4HZCGQxCNwREG4/toZ66QfBnaJIDzYyhh/iMKSy6ePIWJiRgwoWLHwgzWrj0BXOIiL2gdwzpxFPsp89zkXSIbUNDC1k7zpjL4nq/yqYS91RjLYuwIodo6xbHzWHu0f7VNVlRDQ5lc6g= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781796; c=relaxed/simple; bh=FF9mY4DOg2BPKdXJQ5eiGZaB0qoS+e9hXmMBLwt7B7o=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=t0sTFEsBpEwe6hoXv3MzpfHP8uaRsQoL5/zJr32x5q6aSkygoL9n3wFOCASKa6neyE5CglVq8q1NyRkxWJ0wwkHg6y+W2KstEaMP1pwOUsIZbY7/tUITudN509jQ1NPx1Qb/Lt1KRU7Lqv7DESbZqYgIBuNAKQi1EbTNw4dxHMo= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b=A2RwYo9y; arc=none smtp.client-ip=170.10.133.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="A2RwYo9y" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1724781793; h=from:from:reply-to:subject:subject: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=NzGuFcOvuGGjMxt0Vp2I5r88XrXxRE/WxxnSQWtNRbA=; b=A2RwYo9ythmVnE2tLWrfx7YePL+iuKbc6KtYR9Z5YougOwC6F3sV4jRmKP1Cs+j0hTDOnz Z063EfszCPw1vVTbqEEwZbczTqtywF5NKjqrK+wQOGMaXjggn4e2O/3Z8UP0Ubl1Z0bzjz dheDXOsyY3SePqDtTxBmmQh6nrxCv8s= Received: from mx-prod-mc-04.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-364-hggUF37uM7-y7lXoaC8zdg-1; Tue, 27 Aug 2024 14:03:11 -0400 X-MC-Unique: hggUF37uM7-y7lXoaC8zdg-1 Received: from mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.15]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-04.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id 770E21955D58; Tue, 27 Aug 2024 18:03:07 +0000 (UTC) Received: from fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com (fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com [10.6.24.150]) by mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id B92D91955F1B; Tue, 27 Aug 2024 18:03:03 +0000 (UTC) From: Alexander Aring To: teigland@redhat.com Cc: gfs2@lists.linux.dev, song@kernel.org, yukuai3@huawei.com, agruenba@redhat.com, mark@fasheh.com, jlbec@evilplan.org, joseph.qi@linux.alibaba.com, gregkh@linuxfoundation.org, rafael@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, ocfs2-devel@lists.linux.dev, netdev@vger.kernel.org, vvidic@valentin-vidic.from.hr, heming.zhao@suse.com, lucien.xin@gmail.com, paulmck@kernel.org, rcu@vger.kernel.org, juri.lelli@redhat.com, williams@redhat.com, aahringo@redhat.com Subject: [RFC 2/7] dlm: fix swapped args sb_flags vs sb_status Date: Tue, 27 Aug 2024 14:02:31 -0400 Message-ID: <20240827180236.316946-3-aahringo@redhat.com> In-Reply-To: <20240827180236.316946-1-aahringo@redhat.com> References: <20240827180236.316946-1-aahringo@redhat.com> Precedence: bulk X-Mailing-List: rcu@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.0 on 10.30.177.15 The arguments got swapped by commit 986ae3c2a8df ("dlm: fix race between final callback and remove") fixing this now. Fixes: 986ae3c2a8df ("dlm: fix race between final callback and remove") Signed-off-by: Alexander Aring --- fs/dlm/ast.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/fs/dlm/ast.c b/fs/dlm/ast.c index 742b30b61c19..0fe8d80ce5e8 100644 --- a/fs/dlm/ast.c +++ b/fs/dlm/ast.c @@ -30,7 +30,7 @@ static void dlm_run_callback(uint32_t ls_id, uint32_t lkb_id, int8_t mode, trace_dlm_bast(ls_id, lkb_id, mode, res_name, res_length); bastfn(astparam, mode); } else if (flags & DLM_CB_CAST) { - trace_dlm_ast(ls_id, lkb_id, sb_status, sb_flags, res_name, + trace_dlm_ast(ls_id, lkb_id, sb_flags, sb_status, res_name, res_length); lksb->sb_status = sb_status; lksb->sb_flags = sb_flags; From patchwork Tue Aug 27 18:02:32 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexander Aring X-Patchwork-Id: 13779967 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.133.124]) (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 2B8711D363C for ; Tue, 27 Aug 2024 18:03:20 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=170.10.133.124 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781802; cv=none; b=YULgLaPh7IWbVYyfHfqVlzpKBMmMgVILWBDalVzvLJu8HxqQ3w7DquSj6xoESWXHlGoZ9ct/WWPZMcqmektMcTPN6S53fKW6Xdhit5mWuZ/asCU43lnlJpqeKIqShMENXDAN4lWzIpEoM99hRrRmTfRS8tsw1K/SLUWt+Zzc+HE= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781802; c=relaxed/simple; bh=EVl9XmBEGraFkDRmYKHcPjDfOyqLOV5AlkZhIsyCRFg=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=R6smSUJOfCTrtw2dPEoP5HyMvzLS/JEnO9BBWGKr1xFsssTUxig9rlhLrJEY2zdqy3HIbMW5x5ylQZ8LPdp4ybASduOZrEiQq/Swk+zCQ8sS2sSmTMu5Fn9ymEG30Gi79jymZ0EsHD5cyHIhcoukuAJtDJWSqu/7C3zE6dCSCNY= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b=HUj9xreH; arc=none smtp.client-ip=170.10.133.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="HUj9xreH" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1724781800; h=from:from:reply-to:subject:subject: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=WdJBi9yY/9Wqn0S5jG8EeDv8lkEIJtCuMq5NHG08cG8=; b=HUj9xreHBVhTu8VR9BJrT+cgPIy9P/5u10Z3Go73AiDqgHep/cDorfc8s1W5w+FCwcCgkG MY863EgYIUwUKoOTDnhqh36c30TtFAhSatUj/Ac6Dgq/qRZ2dPY19z6MZMHo7ONROqOe0L p1zgokRFmQwn3rebLxUzhYLStqT8bhU= Received: from mx-prod-mc-03.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-358-qBmchWQzOR6rjT6349ZJ2A-1; Tue, 27 Aug 2024 14:03:14 -0400 X-MC-Unique: qBmchWQzOR6rjT6349ZJ2A-1 Received: from mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.15]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-03.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id BD06C19560B1; Tue, 27 Aug 2024 18:03:10 +0000 (UTC) Received: from fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com (fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com [10.6.24.150]) by mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id 6B81A1955DD6; Tue, 27 Aug 2024 18:03:07 +0000 (UTC) From: Alexander Aring To: teigland@redhat.com Cc: gfs2@lists.linux.dev, song@kernel.org, yukuai3@huawei.com, agruenba@redhat.com, mark@fasheh.com, jlbec@evilplan.org, joseph.qi@linux.alibaba.com, gregkh@linuxfoundation.org, rafael@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, ocfs2-devel@lists.linux.dev, netdev@vger.kernel.org, vvidic@valentin-vidic.from.hr, heming.zhao@suse.com, lucien.xin@gmail.com, paulmck@kernel.org, rcu@vger.kernel.org, juri.lelli@redhat.com, williams@redhat.com, aahringo@redhat.com Subject: [RFC 3/7] dlm: make add_to_waiters() that is can't fail Date: Tue, 27 Aug 2024 14:02:32 -0400 Message-ID: <20240827180236.316946-4-aahringo@redhat.com> In-Reply-To: <20240827180236.316946-1-aahringo@redhat.com> References: <20240827180236.316946-1-aahringo@redhat.com> Precedence: bulk X-Mailing-List: rcu@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.0 on 10.30.177.15 If add_to_waiters() fails we have a problem because the previous called functions such as validate_lock_args() or validate_unlock_args() sets specific lkb values that are set for a request, there exists no way back to revert those changes. When there is a pending lock request the original request arguments will be overwritten with unknown consequences. The good news are that I believe those cases that we fail in add_to_waiters() can't happen or very unlikely to happen (only if the DLM user does stupid API things), but if so we have the above mentioned problem. There are two conditions that will be removed here. The first one is the -EINVAL case which contains is_overlap_unlock() or (is_overlap_cancel() and mstype == DLM_MSG_CANCEL). The is_overlap_unlock() is missing for the normal UNLOCK case which is moved to validate_unlock_args(). The is_overlap_cancel() already happens in validate_unlock_args() when DLM_LKF_CANCEL is set. In case of validate_lock_args() we check on is_overlap() when it is not a new request, on a new request the lkb is always new and does not have those values set. The -EBUSY check can't happen in case as for non new lock requests (when DLM_LKF_CONVERT is set) we already check in validate_lock_args() for lkb_wait_type and is_overlap(). Then there is only validate_unlock_args() that will never hit the default case because dlm_unlock() will produce DLM_MSG_UNLOCK and DLM_MSG_CANCEL messages. Signed-off-by: Alexander Aring --- fs/dlm/lock.c | 43 ++++++++++++++----------------------------- 1 file changed, 14 insertions(+), 29 deletions(-) diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index 121d2976986b..8cb5a537bfd3 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -1703,19 +1703,11 @@ static int msg_reply_type(int mstype) /* add/remove lkb from global waiters list of lkb's waiting for a reply from a remote node */ -static int add_to_waiters(struct dlm_lkb *lkb, int mstype, int to_nodeid) +static void add_to_waiters(struct dlm_lkb *lkb, int mstype, int to_nodeid) { struct dlm_ls *ls = lkb->lkb_resource->res_ls; - int error = 0; spin_lock_bh(&ls->ls_waiters_lock); - - if (is_overlap_unlock(lkb) || - (is_overlap_cancel(lkb) && (mstype == DLM_MSG_CANCEL))) { - error = -EINVAL; - goto out; - } - if (lkb->lkb_wait_type || is_overlap_cancel(lkb)) { switch (mstype) { case DLM_MSG_UNLOCK: @@ -1725,7 +1717,11 @@ static int add_to_waiters(struct dlm_lkb *lkb, int mstype, int to_nodeid) set_bit(DLM_IFL_OVERLAP_CANCEL_BIT, &lkb->lkb_iflags); break; default: - error = -EBUSY; + /* should never happen as validate_lock_args() checks + * on lkb_wait_type and validate_unlock_args() only + * creates UNLOCK or CANCEL messages. + */ + WARN_ON_ONCE(1); goto out; } lkb->lkb_wait_count++; @@ -1747,12 +1743,7 @@ static int add_to_waiters(struct dlm_lkb *lkb, int mstype, int to_nodeid) hold_lkb(lkb); list_add(&lkb->lkb_wait_reply, &ls->ls_waiters); out: - if (error) - log_error(ls, "addwait error %x %d flags %x %d %d %s", - lkb->lkb_id, error, dlm_iflags_val(lkb), mstype, - lkb->lkb_wait_type, lkb->lkb_resource->res_name); spin_unlock_bh(&ls->ls_waiters_lock); - return error; } /* We clear the RESEND flag because we might be taking an lkb off the waiters @@ -2926,13 +2917,16 @@ static int validate_unlock_args(struct dlm_lkb *lkb, struct dlm_args *args) goto out; } + if (is_overlap_unlock(lkb)) + goto out; + /* cancel not allowed with another cancel/unlock in progress */ if (args->flags & DLM_LKF_CANCEL) { if (lkb->lkb_exflags & DLM_LKF_CANCEL) goto out; - if (is_overlap(lkb)) + if (is_overlap_cancel(lkb)) goto out; if (test_bit(DLM_IFL_RESEND_BIT, &lkb->lkb_iflags)) { @@ -2970,9 +2964,6 @@ static int validate_unlock_args(struct dlm_lkb *lkb, struct dlm_args *args) if (lkb->lkb_exflags & DLM_LKF_FORCEUNLOCK) goto out; - if (is_overlap_unlock(lkb)) - goto out; - if (test_bit(DLM_IFL_RESEND_BIT, &lkb->lkb_iflags)) { set_bit(DLM_IFL_OVERLAP_UNLOCK_BIT, &lkb->lkb_iflags); rv = -EBUSY; @@ -3608,10 +3599,7 @@ static int send_common(struct dlm_rsb *r, struct dlm_lkb *lkb, int mstype) to_nodeid = r->res_nodeid; - error = add_to_waiters(lkb, mstype, to_nodeid); - if (error) - return error; - + add_to_waiters(lkb, mstype, to_nodeid); error = create_message(r, lkb, to_nodeid, mstype, &ms, &mh); if (error) goto fail; @@ -3714,10 +3702,7 @@ static int send_lookup(struct dlm_rsb *r, struct dlm_lkb *lkb) to_nodeid = dlm_dir_nodeid(r); - error = add_to_waiters(lkb, DLM_MSG_LOOKUP, to_nodeid); - if (error) - return error; - + add_to_waiters(lkb, DLM_MSG_LOOKUP, to_nodeid); error = create_message(r, NULL, to_nodeid, DLM_MSG_LOOKUP, &ms, &mh); if (error) goto fail; @@ -6342,8 +6327,8 @@ int dlm_debug_add_lkb_to_waiters(struct dlm_ls *ls, uint32_t lkb_id, if (error) return error; - error = add_to_waiters(lkb, mstype, to_nodeid); + add_to_waiters(lkb, mstype, to_nodeid); dlm_put_lkb(lkb); - return error; + return 0; } From patchwork Tue Aug 27 18:02:33 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexander Aring X-Patchwork-Id: 13779968 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) (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 CD0071D414C for ; Tue, 27 Aug 2024 18:03:24 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=170.10.129.124 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781806; cv=none; b=cNbHR7XZrd+AbFo8k9mbL3XpTcIdre0Jad4zTkSmqH0wr+vq6DhoSF/sDk6oY+oU0zZsvLQmyDqznbBJfe0dZkcInzPfoYKlx4NIr7phhWVBMjgJY9EVPxj/3bhIFTDxeO0uPy9yADEIVq1g0znLNd+UhuNUAWbttn3qfBnisdM= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781806; c=relaxed/simple; bh=LrvL7Ph/QlOSx7zujU2LmzLvPYnl+n/ew0XfqKHb0TI=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=NHbkYRxg0zOK8CwKVEE64Z9zaO2kMw4Kd447upkz3m6kJC9Y3dEjgeZPBA73U7TykjAsQfPj1GCk78PwqO2ScAPA17M4eG1mvt7O6WRvyv4EOizZWuxLrEiJrHeZmagG7o+ksGmVY8MSXpBj4qXSF8Yq3Ib61eh3VTZhwD451zQ= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b=O2KeA5O2; arc=none smtp.client-ip=170.10.129.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="O2KeA5O2" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1724781803; h=from:from:reply-to:subject:subject: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=SCDbnalavEPDZ6X8+FuzaIDIjdKt3WdCbgJCmYzM6cw=; b=O2KeA5O2Z+J3Gt7gYnt266JcpxL44N04h1mFJm9/TiJ+orUpLqyIkp5owLDHlGSF6QSB6L YNnbVc67UFjG2FeJg9XV1HKud3YI3ZwZO5cPaieU3lcxEmgXYRzNTAjbludJ1hjsF2G5sa qybN8D1sge8MF+iSdao5Ulwy6929JaE= Received: from mx-prod-mc-02.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-629-3JmjOdStOcu7MPESNnoFGg-1; Tue, 27 Aug 2024 14:03:18 -0400 X-MC-Unique: 3JmjOdStOcu7MPESNnoFGg-1 Received: from mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.15]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-02.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id DE35C1955D55; Tue, 27 Aug 2024 18:03:14 +0000 (UTC) Received: from fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com (fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com [10.6.24.150]) by mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id D03CA1955DD8; Tue, 27 Aug 2024 18:03:10 +0000 (UTC) From: Alexander Aring To: teigland@redhat.com Cc: gfs2@lists.linux.dev, song@kernel.org, yukuai3@huawei.com, agruenba@redhat.com, mark@fasheh.com, jlbec@evilplan.org, joseph.qi@linux.alibaba.com, gregkh@linuxfoundation.org, rafael@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, ocfs2-devel@lists.linux.dev, netdev@vger.kernel.org, vvidic@valentin-vidic.from.hr, heming.zhao@suse.com, lucien.xin@gmail.com, paulmck@kernel.org, rcu@vger.kernel.org, juri.lelli@redhat.com, williams@redhat.com, aahringo@redhat.com Subject: [RFC 4/7] dlm: add our_nodeid to tracepoints Date: Tue, 27 Aug 2024 14:02:33 -0400 Message-ID: <20240827180236.316946-5-aahringo@redhat.com> In-Reply-To: <20240827180236.316946-1-aahringo@redhat.com> References: <20240827180236.316946-1-aahringo@redhat.com> Precedence: bulk X-Mailing-List: rcu@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.0 on 10.30.177.15 This patch adds our_nodeid to some DLM tracepoints that are necessary for the DLM kernel verifier to know from which nodeid the traceevent comes from. This is useful when using DLM in net-namespaces to get a whole cluster-view of DLM in traces. Signed-off-by: Alexander Aring --- fs/dlm/ast.c | 23 +++++++++++++---------- fs/dlm/dlm_internal.h | 1 + fs/dlm/user.c | 9 +++++---- include/trace/events/dlm.h | 36 +++++++++++++++++++++++------------- 4 files changed, 42 insertions(+), 27 deletions(-) diff --git a/fs/dlm/ast.c b/fs/dlm/ast.c index 0fe8d80ce5e8..01de0d4b9450 100644 --- a/fs/dlm/ast.c +++ b/fs/dlm/ast.c @@ -18,20 +18,21 @@ #include "user.h" #include "ast.h" -static void dlm_run_callback(uint32_t ls_id, uint32_t lkb_id, int8_t mode, - uint32_t flags, uint8_t sb_flags, int sb_status, - struct dlm_lksb *lksb, +static void dlm_run_callback(int our_nodeid, uint32_t ls_id, uint32_t lkb_id, + int8_t mode, uint32_t flags, uint8_t sb_flags, + int sb_status, struct dlm_lksb *lksb, void (*astfn)(void *astparam), void (*bastfn)(void *astparam, int mode), void *astparam, const char *res_name, size_t res_length) { if (flags & DLM_CB_BAST) { - trace_dlm_bast(ls_id, lkb_id, mode, res_name, res_length); + trace_dlm_bast(our_nodeid, ls_id, lkb_id, mode, res_name, + res_length); bastfn(astparam, mode); } else if (flags & DLM_CB_CAST) { - trace_dlm_ast(ls_id, lkb_id, sb_flags, sb_status, res_name, - res_length); + trace_dlm_ast(our_nodeid, ls_id, lkb_id, sb_flags, sb_status, + res_name, res_length); lksb->sb_status = sb_status; lksb->sb_flags = sb_flags; astfn(astparam); @@ -40,8 +41,8 @@ static void dlm_run_callback(uint32_t ls_id, uint32_t lkb_id, int8_t mode, static void dlm_do_callback(struct dlm_callback *cb) { - dlm_run_callback(cb->ls_id, cb->lkb_id, cb->mode, cb->flags, - cb->sb_flags, cb->sb_status, cb->lkb_lksb, + dlm_run_callback(cb->our_nodeid, cb->ls_id, cb->lkb_id, cb->mode, + cb->flags, cb->sb_flags, cb->sb_status, cb->lkb_lksb, cb->astfn, cb->bastfn, cb->astparam, cb->res_name, cb->res_length); dlm_free_cb(cb); @@ -130,6 +131,7 @@ int dlm_get_cb(struct dlm_lkb *lkb, uint32_t flags, int mode, return -ENOMEM; /* for tracing */ + (*cb)->our_nodeid = ls->ls_dn->our_node->id; (*cb)->lkb_id = lkb->lkb_id; (*cb)->ls_id = ls->ls_global_id; memcpy((*cb)->res_name, rsb->res_name, rsb->res_length); @@ -185,8 +187,9 @@ void dlm_add_cb(struct dlm_lkb *lkb, uint32_t flags, int mode, int status, list_add(&cb->list, &ls->ls_cb_delay); } else { if (test_bit(LSFL_SOFTIRQ, &ls->ls_flags)) { - dlm_run_callback(ls->ls_global_id, lkb->lkb_id, mode, flags, - sbflags, status, lkb->lkb_lksb, + dlm_run_callback(ls->ls_dn->our_node->id, + ls->ls_global_id, lkb->lkb_id, mode, + flags, sbflags, status, lkb->lkb_lksb, lkb->lkb_astfn, lkb->lkb_bastfn, lkb->lkb_astparam, rsb->res_name, rsb->res_length); diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h index 2de5ef2653cd..bc3ff1b64e0c 100644 --- a/fs/dlm/dlm_internal.h +++ b/fs/dlm/dlm_internal.h @@ -234,6 +234,7 @@ struct dlm_callback { bool copy_lvb; struct dlm_lksb *lkb_lksb; unsigned char lvbptr[DLM_USER_LVB_LEN]; + int our_nodeid; union { void *astparam; /* caller's ast arg */ diff --git a/fs/dlm/user.c b/fs/dlm/user.c index 1b682f8f95b6..c4d6e67ff63e 100644 --- a/fs/dlm/user.c +++ b/fs/dlm/user.c @@ -868,13 +868,14 @@ static ssize_t device_read(struct file *file, char __user *buf, size_t count, spin_unlock_bh(&proc->asts_spin); if (cb->flags & DLM_CB_BAST) { - trace_dlm_bast(cb->ls_id, cb->lkb_id, cb->mode, cb->res_name, - cb->res_length); + trace_dlm_bast(cb->our_nodeid, cb->ls_id, cb->lkb_id, + cb->mode, cb->res_name, cb->res_length); } else if (cb->flags & DLM_CB_CAST) { cb->lkb_lksb->sb_status = cb->sb_status; cb->lkb_lksb->sb_flags = cb->sb_flags; - trace_dlm_ast(cb->ls_id, cb->lkb_id, cb->sb_status, - cb->sb_flags, cb->res_name, cb->res_length); + trace_dlm_ast(cb->our_nodeid, cb->ls_id, cb->lkb_id, + cb->sb_status, cb->sb_flags, cb->res_name, + cb->res_length); } ret = copy_result_to_user(&cb->ua, diff --git a/include/trace/events/dlm.h b/include/trace/events/dlm.h index af160082c9e3..2621bb7ac3a8 100644 --- a/include/trace/events/dlm.h +++ b/include/trace/events/dlm.h @@ -98,6 +98,7 @@ TRACE_EVENT(dlm_lock_start, TP_ARGS(ls, lkb, name, namelen, mode, flags), TP_STRUCT__entry( + __field(unsigned int, our_nodeid) __field(__u32, ls_id) __field(__u32, lkb_id) __field(int, mode) @@ -109,6 +110,7 @@ TRACE_EVENT(dlm_lock_start, TP_fast_assign( struct dlm_rsb *r; + __entry->our_nodeid = ls->ls_dn->our_node->id; __entry->ls_id = ls->ls_global_id; __entry->lkb_id = lkb->lkb_id; __entry->mode = mode; @@ -123,8 +125,8 @@ TRACE_EVENT(dlm_lock_start, __get_dynamic_array_len(res_name)); ), - TP_printk("ls_id=%u lkb_id=%x mode=%s flags=%s res_name=%s", - __entry->ls_id, __entry->lkb_id, + TP_printk("our_nodeid=%u ls_id=%u lkb_id=%x mode=%s flags=%s res_name=%s", + __entry->our_nodeid, __entry->ls_id, __entry->lkb_id, show_lock_mode(__entry->mode), show_lock_flags(__entry->flags), __print_hex_str(__get_dynamic_array(res_name), @@ -141,6 +143,7 @@ TRACE_EVENT(dlm_lock_end, TP_ARGS(ls, lkb, name, namelen, mode, flags, error, kernel_lock), TP_STRUCT__entry( + __field(unsigned int, our_nodeid) __field(__u32, ls_id) __field(__u32, lkb_id) __field(int, mode) @@ -153,6 +156,7 @@ TRACE_EVENT(dlm_lock_end, TP_fast_assign( struct dlm_rsb *r; + __entry->our_nodeid = ls->ls_dn->our_node->id; __entry->ls_id = ls->ls_global_id; __entry->lkb_id = lkb->lkb_id; __entry->mode = mode; @@ -178,8 +182,8 @@ TRACE_EVENT(dlm_lock_end, ), - TP_printk("ls_id=%u lkb_id=%x mode=%s flags=%s error=%d res_name=%s", - __entry->ls_id, __entry->lkb_id, + TP_printk("our_nodeid=%u ls_id=%u lkb_id=%x mode=%s flags=%s error=%d res_name=%s", + __entry->our_nodeid, __entry->ls_id, __entry->lkb_id, show_lock_mode(__entry->mode), show_lock_flags(__entry->flags), __entry->error, __print_hex_str(__get_dynamic_array(res_name), @@ -189,12 +193,13 @@ TRACE_EVENT(dlm_lock_end, TRACE_EVENT(dlm_bast, - TP_PROTO(__u32 ls_id, __u32 lkb_id, int mode, + TP_PROTO(unsigned int our_nodeid, __u32 ls_id, __u32 lkb_id, int mode, const char *res_name, size_t res_length), - TP_ARGS(ls_id, lkb_id, mode, res_name, res_length), + TP_ARGS(our_nodeid, ls_id, lkb_id, mode, res_name, res_length), TP_STRUCT__entry( + __field(unsigned int, our_nodeid) __field(__u32, ls_id) __field(__u32, lkb_id) __field(int, mode) @@ -202,6 +207,7 @@ TRACE_EVENT(dlm_bast, ), TP_fast_assign( + __entry->our_nodeid = our_nodeid; __entry->ls_id = ls_id; __entry->lkb_id = lkb_id; __entry->mode = mode; @@ -210,8 +216,8 @@ TRACE_EVENT(dlm_bast, __get_dynamic_array_len(res_name)); ), - TP_printk("ls_id=%u lkb_id=%x mode=%s res_name=%s", - __entry->ls_id, __entry->lkb_id, + TP_printk("our_nodeid=%u ls_id=%u lkb_id=%x mode=%s res_name=%s", + __entry->our_nodeid, __entry->ls_id, __entry->lkb_id, show_lock_mode(__entry->mode), __print_hex_str(__get_dynamic_array(res_name), __get_dynamic_array_len(res_name))) @@ -220,12 +226,15 @@ TRACE_EVENT(dlm_bast, TRACE_EVENT(dlm_ast, - TP_PROTO(__u32 ls_id, __u32 lkb_id, __u8 sb_flags, int sb_status, - const char *res_name, size_t res_length), + TP_PROTO(unsigned int our_nodeid, __u32 ls_id, __u32 lkb_id, + __u8 sb_flags, int sb_status, const char *res_name, + size_t res_length), - TP_ARGS(ls_id, lkb_id, sb_flags, sb_status, res_name, res_length), + TP_ARGS(our_nodeid, ls_id, lkb_id, sb_flags, sb_status, res_name, + res_length), TP_STRUCT__entry( + __field(unsigned int, our_nodeid) __field(__u32, ls_id) __field(__u32, lkb_id) __field(__u8, sb_flags) @@ -234,6 +243,7 @@ TRACE_EVENT(dlm_ast, ), TP_fast_assign( + __entry->our_nodeid = our_nodeid; __entry->ls_id = ls_id; __entry->lkb_id = lkb_id; __entry->sb_flags = sb_flags; @@ -243,8 +253,8 @@ TRACE_EVENT(dlm_ast, __get_dynamic_array_len(res_name)); ), - TP_printk("ls_id=%u lkb_id=%x sb_flags=%s sb_status=%d res_name=%s", - __entry->ls_id, __entry->lkb_id, + TP_printk("our_nodeid=%u ls_id=%u lkb_id=%x sb_flags=%s sb_status=%d res_name=%s", + __entry->our_nodeid, __entry->ls_id, __entry->lkb_id, show_dlm_sb_flags(__entry->sb_flags), __entry->sb_status, __print_hex_str(__get_dynamic_array(res_name), __get_dynamic_array_len(res_name))) From patchwork Tue Aug 27 18:02:34 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexander Aring X-Patchwork-Id: 13779969 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) (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 732E61D2F43 for ; Tue, 27 Aug 2024 18:03:27 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=170.10.129.124 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781809; cv=none; b=jqsyiKqWfLTe9u2Alua+3q0oNkoSd0nvPImD02ITC1rLsVXfwEVjwZl63kbO4cSOsD103qFuECC6/v/91/Zw/ue6/7GFj6SjL9PnUlSxsBg+bXEKBvRzGmTf36U9Lo2ANJnzf6gkWwftze0KeahOLlDfEauJ7sx91HeoZsOSZyw= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781809; c=relaxed/simple; bh=OJls2vsbsH6uelH/T/HgOzTBW8h0QC3AGr0F+CJvFJs=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=jOKVuk49zu5x9WeTQhYteRBqE4w9OzHdVvvj1kD7zxg3AaJPSt4opUP0BmDfyxOIC20w1mbsX6t9Ao9fHrucbXr151d50JbSTe4I3uMLTBQDTTwfhEaghp6ydaglmUpy1ZTbqiiLyVACZbI2/emJNj45E24t9eGXTfmk+Ydm/Pk= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b=UjETfX3Y; arc=none smtp.client-ip=170.10.129.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="UjETfX3Y" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1724781806; h=from:from:reply-to:subject:subject: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=0n3QoftpE0QdZNyZ3i3U02OdoDsTNz9pYwwpvpU4g2w=; b=UjETfX3Y8+vcQGapPnxSIbHcbtdzn8nZYR53nLrgMv5IMHduvpE/pTO0yD4/eM1tPYYH8k CykAQaX+cPaFdxv7WsDHCjbonw/BgQvV9/heVjRw+A1lRcErLsA4m/cVCTrDh8pHsLykpw 8aiRxZmxU1HvFQ6K16gCctEl0tqJ0kc= Received: from mx-prod-mc-04.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-121-IpDebxFUMKGY-2feeLPpiw-1; Tue, 27 Aug 2024 14:03:22 -0400 X-MC-Unique: IpDebxFUMKGY-2feeLPpiw-1 Received: from mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.15]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-04.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id 692F91955D4C; Tue, 27 Aug 2024 18:03:18 +0000 (UTC) Received: from fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com (fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com [10.6.24.150]) by mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id 1B1741955BE1; Tue, 27 Aug 2024 18:03:14 +0000 (UTC) From: Alexander Aring To: teigland@redhat.com Cc: gfs2@lists.linux.dev, song@kernel.org, yukuai3@huawei.com, agruenba@redhat.com, mark@fasheh.com, jlbec@evilplan.org, joseph.qi@linux.alibaba.com, gregkh@linuxfoundation.org, rafael@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, ocfs2-devel@lists.linux.dev, netdev@vger.kernel.org, vvidic@valentin-vidic.from.hr, heming.zhao@suse.com, lucien.xin@gmail.com, paulmck@kernel.org, rcu@vger.kernel.org, juri.lelli@redhat.com, williams@redhat.com, aahringo@redhat.com Subject: [RFC 5/7] dlm: add lkb rv mode to ast tracepoint Date: Tue, 27 Aug 2024 14:02:34 -0400 Message-ID: <20240827180236.316946-6-aahringo@redhat.com> In-Reply-To: <20240827180236.316946-1-aahringo@redhat.com> References: <20240827180236.316946-1-aahringo@redhat.com> Precedence: bulk X-Mailing-List: rcu@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.0 on 10.30.177.15 This patch adds the lkb_rv_mode to the ast tracepoint. The lkb_rv_mode is the requested mode by dlm_lock() requests. We cannot use lkb_mode as dlm internally sometimes changes this value and for cases like the dlm kernel verifier we want to check on DLM correctness from what the user is seeing. This new tracepoint events tells us what the requested mode was. Signed-off-by: Alexander Aring --- fs/dlm/ast.c | 9 +++++---- fs/dlm/dlm_internal.h | 2 ++ fs/dlm/lock.c | 1 + fs/dlm/user.c | 2 +- include/trace/events/dlm.h | 11 +++++++---- 5 files changed, 16 insertions(+), 9 deletions(-) diff --git a/fs/dlm/ast.c b/fs/dlm/ast.c index 01de0d4b9450..fc2ca37011d8 100644 --- a/fs/dlm/ast.c +++ b/fs/dlm/ast.c @@ -24,7 +24,7 @@ static void dlm_run_callback(int our_nodeid, uint32_t ls_id, uint32_t lkb_id, void (*astfn)(void *astparam), void (*bastfn)(void *astparam, int mode), void *astparam, const char *res_name, - size_t res_length) + size_t res_length, int rv_mode) { if (flags & DLM_CB_BAST) { trace_dlm_bast(our_nodeid, ls_id, lkb_id, mode, res_name, @@ -32,7 +32,7 @@ static void dlm_run_callback(int our_nodeid, uint32_t ls_id, uint32_t lkb_id, bastfn(astparam, mode); } else if (flags & DLM_CB_CAST) { trace_dlm_ast(our_nodeid, ls_id, lkb_id, sb_flags, sb_status, - res_name, res_length); + res_name, res_length, rv_mode); lksb->sb_status = sb_status; lksb->sb_flags = sb_flags; astfn(astparam); @@ -44,7 +44,7 @@ static void dlm_do_callback(struct dlm_callback *cb) dlm_run_callback(cb->our_nodeid, cb->ls_id, cb->lkb_id, cb->mode, cb->flags, cb->sb_flags, cb->sb_status, cb->lkb_lksb, cb->astfn, cb->bastfn, cb->astparam, - cb->res_name, cb->res_length); + cb->res_name, cb->res_length, cb->rv_mode); dlm_free_cb(cb); } @@ -134,6 +134,7 @@ int dlm_get_cb(struct dlm_lkb *lkb, uint32_t flags, int mode, (*cb)->our_nodeid = ls->ls_dn->our_node->id; (*cb)->lkb_id = lkb->lkb_id; (*cb)->ls_id = ls->ls_global_id; + (*cb)->rv_mode = lkb->lkb_rv_mode; memcpy((*cb)->res_name, rsb->res_name, rsb->res_length); (*cb)->res_length = rsb->res_length; @@ -192,7 +193,7 @@ void dlm_add_cb(struct dlm_lkb *lkb, uint32_t flags, int mode, int status, flags, sbflags, status, lkb->lkb_lksb, lkb->lkb_astfn, lkb->lkb_bastfn, lkb->lkb_astparam, rsb->res_name, - rsb->res_length); + rsb->res_length, lkb->lkb_rv_mode); } else { rv = dlm_get_queue_cb(lkb, flags, mode, status, sbflags, &cb); if (!rv) diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h index bc3ff1b64e0c..3f630696f7ab 100644 --- a/fs/dlm/dlm_internal.h +++ b/fs/dlm/dlm_internal.h @@ -247,6 +247,7 @@ struct dlm_callback { size_t res_length; uint32_t ls_id; uint32_t lkb_id; + int rv_mode; struct list_head list; }; @@ -296,6 +297,7 @@ struct dlm_lkb { void *lkb_astparam; /* caller's ast arg */ struct dlm_user_args *lkb_ua; }; + int lkb_rv_mode; struct rcu_head rcu; }; diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index 8cb5a537bfd3..21bb9603a0df 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -2844,6 +2844,7 @@ static int validate_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, lkb->lkb_lksb = args->lksb; lkb->lkb_lvbptr = args->lksb->sb_lvbptr; lkb->lkb_ownpid = (int) current->pid; + lkb->lkb_rv_mode = args->mode; rv = 0; out: switch (rv) { diff --git a/fs/dlm/user.c b/fs/dlm/user.c index c4d6e67ff63e..75fb85676e90 100644 --- a/fs/dlm/user.c +++ b/fs/dlm/user.c @@ -875,7 +875,7 @@ static ssize_t device_read(struct file *file, char __user *buf, size_t count, cb->lkb_lksb->sb_flags = cb->sb_flags; trace_dlm_ast(cb->our_nodeid, cb->ls_id, cb->lkb_id, cb->sb_status, cb->sb_flags, cb->res_name, - cb->res_length); + cb->res_length, cb->rv_mode); } ret = copy_result_to_user(&cb->ua, diff --git a/include/trace/events/dlm.h b/include/trace/events/dlm.h index 2621bb7ac3a8..f8d7ca451760 100644 --- a/include/trace/events/dlm.h +++ b/include/trace/events/dlm.h @@ -228,10 +228,10 @@ TRACE_EVENT(dlm_ast, TP_PROTO(unsigned int our_nodeid, __u32 ls_id, __u32 lkb_id, __u8 sb_flags, int sb_status, const char *res_name, - size_t res_length), + size_t res_length, int rv_mode), TP_ARGS(our_nodeid, ls_id, lkb_id, sb_flags, sb_status, res_name, - res_length), + res_length, rv_mode), TP_STRUCT__entry( __field(unsigned int, our_nodeid) @@ -239,6 +239,7 @@ TRACE_EVENT(dlm_ast, __field(__u32, lkb_id) __field(__u8, sb_flags) __field(int, sb_status) + __field(int, rv_mode) __dynamic_array(unsigned char, res_name, res_length) ), @@ -248,16 +249,18 @@ TRACE_EVENT(dlm_ast, __entry->lkb_id = lkb_id; __entry->sb_flags = sb_flags; __entry->sb_status = sb_status; + __entry->rv_mode = rv_mode; memcpy(__get_dynamic_array(res_name), res_name, __get_dynamic_array_len(res_name)); ), - TP_printk("our_nodeid=%u ls_id=%u lkb_id=%x sb_flags=%s sb_status=%d res_name=%s", + TP_printk("our_nodeid=%u ls_id=%u lkb_id=%x sb_flags=%s sb_status=%d res_name=%s rv_mode=%d", __entry->our_nodeid, __entry->ls_id, __entry->lkb_id, show_dlm_sb_flags(__entry->sb_flags), __entry->sb_status, __print_hex_str(__get_dynamic_array(res_name), - __get_dynamic_array_len(res_name))) + __get_dynamic_array_len(res_name)), + __entry->rv_mode) ); From patchwork Tue Aug 27 18:02:35 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexander Aring X-Patchwork-Id: 13779970 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) (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 6BE971D415F for ; Tue, 27 Aug 2024 18:03:28 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=170.10.129.124 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781810; cv=none; b=UJWeGS9XBYyvuEwtmBZi1FxUhwIaYtnBUVxsCSQk7RU7a34IT9ArxR5UXhcw0++H/qqesZBxUBp1tZYmWNTlWMx2FoE99bG/0F2H4HKyLc8Ts5jXlbZjdelTF1i6qh3MO6gULLkL8JF11ivfgJ3iaMq563qAA/YeNCKwVlWx9O8= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781810; c=relaxed/simple; bh=vRZxlA4NhcmLd0q6bUaelJnaLNQfejDBlwC1tsp+b8s=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=reaZhDPvXc+XxQa7ag5ufkjJzF7XxZRKp48mMFo0H92O94log5oYfef79vQ1sAWw/a7euB1/HSwQGZK1ZKxnpj3a8bUjE2e7cjTVm59ThE4LmxMcp0OgAUgJdLyZPNsgfOCW3muuHSpNr13BEUzn6KjOH815IVGpmjNeaIzq36E= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b=Oj7N8snw; arc=none smtp.client-ip=170.10.129.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="Oj7N8snw" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1724781807; h=from:from:reply-to:subject:subject: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=LykJQUiuG045SO6zmRnu8bef9DxxtsLHq0x3Z+ewaXo=; b=Oj7N8snw5BXRgDtK0eAjMOEKZOH8GhCj3Y2udMGFWDN4Ixn3qcOzgz4IRbEX+qVPTt2fo9 Dt3s38b6r1xqtPCUafQztZ/BUX+7H92nvnlT+wrc4EOAjPLDPavDRS7CnhJL1Pe9flI4n5 vx+xJVvymSDfT44r4YL2JUoJu/6z55o= Received: from mx-prod-mc-04.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-433--ayIjcuCPze9ryjYvQt7vA-1; Tue, 27 Aug 2024 14:03:24 -0400 X-MC-Unique: -ayIjcuCPze9ryjYvQt7vA-1 Received: from mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.15]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-04.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id C75D11955D4F; Tue, 27 Aug 2024 18:03:21 +0000 (UTC) Received: from fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com (fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com [10.6.24.150]) by mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id 7C8861955F1B; Tue, 27 Aug 2024 18:03:18 +0000 (UTC) From: Alexander Aring To: teigland@redhat.com Cc: gfs2@lists.linux.dev, song@kernel.org, yukuai3@huawei.com, agruenba@redhat.com, mark@fasheh.com, jlbec@evilplan.org, joseph.qi@linux.alibaba.com, gregkh@linuxfoundation.org, rafael@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, ocfs2-devel@lists.linux.dev, netdev@vger.kernel.org, vvidic@valentin-vidic.from.hr, heming.zhao@suse.com, lucien.xin@gmail.com, paulmck@kernel.org, rcu@vger.kernel.org, juri.lelli@redhat.com, williams@redhat.com, aahringo@redhat.com Subject: [RFC 6/7] dlm: add more tracepoints for DLM kernel verifier Date: Tue, 27 Aug 2024 14:02:35 -0400 Message-ID: <20240827180236.316946-7-aahringo@redhat.com> In-Reply-To: <20240827180236.316946-1-aahringo@redhat.com> References: <20240827180236.316946-1-aahringo@redhat.com> Precedence: bulk X-Mailing-List: rcu@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.0 on 10.30.177.15 This patch adds more useful tracepoints for the kernel verifier. The lock/unlock validation tracepoints are useful to store the request state mode that the lock must change to, at this time the lock request cannot fail anymore. (It can fail but there might be upcoming changes because we can't deal with that e.g. -ENOMEM cases). Another tracepoint is dlm_release_lockspace() that signals us that a node is leaving the lockspace and all locks holding should be dropped. Signed-off-by: Alexander Aring --- fs/dlm/lock.c | 10 +++-- fs/dlm/lockspace.c | 4 ++ include/trace/events/dlm.h | 80 ++++++++++++++++++++++++++++++++++++++ 3 files changed, 91 insertions(+), 3 deletions(-) diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index 21bb9603a0df..597418cba76a 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -2811,7 +2811,8 @@ static int set_unlock_args(uint32_t flags, void *astarg, struct dlm_args *args) } static int validate_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, - struct dlm_args *args) + struct dlm_args *args, const char *res_name, + size_t res_length) { int rv = -EBUSY; @@ -2845,6 +2846,7 @@ static int validate_lock_args(struct dlm_ls *ls, struct dlm_lkb *lkb, lkb->lkb_lvbptr = args->lksb->sb_lvbptr; lkb->lkb_ownpid = (int) current->pid; lkb->lkb_rv_mode = args->mode; + trace_dlm_lock_validated(ls, lkb, args, res_name, res_length); rv = 0; out: switch (rv) { @@ -2988,6 +2990,7 @@ static int validate_unlock_args(struct dlm_lkb *lkb, struct dlm_args *args) lkb->lkb_exflags |= args->flags; dlm_set_sbflags_val(lkb, 0); lkb->lkb_astparam = args->astparam; + trace_dlm_unlock_validated(ls, lkb, args); rv = 0; out: switch (rv) { @@ -3264,7 +3267,7 @@ static int request_lock(struct dlm_ls *ls, struct dlm_lkb *lkb, struct dlm_rsb *r; int error; - error = validate_lock_args(ls, lkb, args); + error = validate_lock_args(ls, lkb, args, name, len); if (error) return error; @@ -3295,7 +3298,8 @@ static int convert_lock(struct dlm_ls *ls, struct dlm_lkb *lkb, hold_rsb(r); lock_rsb(r); - error = validate_lock_args(ls, lkb, args); + error = validate_lock_args(ls, lkb, args, r->res_name, + r->res_length); if (error) goto out; diff --git a/fs/dlm/lockspace.c b/fs/dlm/lockspace.c index 092f7017b896..c19d797264c5 100644 --- a/fs/dlm/lockspace.c +++ b/fs/dlm/lockspace.c @@ -9,6 +9,8 @@ ******************************************************************************* ******************************************************************************/ +#include + #include #include @@ -781,6 +783,8 @@ static int release_lockspace(struct dlm_net *dn, struct dlm_ls *ls, int force) return rv; } + trace_dlm_release_lockspace(dn->our_node->id, ls->ls_global_id); + if (dn->ls_count == 1) dlm_midcomms_version_wait(ls->ls_dn); diff --git a/include/trace/events/dlm.h b/include/trace/events/dlm.h index f8d7ca451760..facad6251e43 100644 --- a/include/trace/events/dlm.h +++ b/include/trace/events/dlm.h @@ -338,6 +338,86 @@ TRACE_EVENT(dlm_unlock_end, ); +TRACE_EVENT(dlm_release_lockspace, + + TP_PROTO(unsigned int our_nodeid, __u32 ls_id), + + TP_ARGS(our_nodeid, ls_id), + + TP_STRUCT__entry( + __field(unsigned int, our_nodeid) + __field(__u32, ls_id) + ), + + TP_fast_assign( + __entry->our_nodeid = our_nodeid; + __entry->ls_id = ls_id; + ), + + TP_printk("our_nodeid=%u ls_id=%u", + __entry->our_nodeid, __entry->ls_id) + +); + +TRACE_EVENT(dlm_lock_validated, + + TP_PROTO(struct dlm_ls *ls, struct dlm_lkb *lkb, struct dlm_args *args, + const char *res_name, size_t res_length), + + TP_ARGS(ls, lkb, args, res_name, res_length), + + TP_STRUCT__entry( + __field(uint32_t, our_nodeid) + __field(uint32_t, ls_id) + __dynamic_array(unsigned char, res_name, res_length) + __field(int, mode) + ), + + TP_fast_assign( + __entry->our_nodeid = ls->ls_dn->our_node->id; + __entry->ls_id = ls->ls_global_id; + memcpy(__get_dynamic_array(res_name), res_name, + __get_dynamic_array_len(res_name)); + __entry->mode = args->mode; + ), + + TP_printk("our_nodeid=%u ls_id=%u res_name=%s mode=%d", + __entry->our_nodeid, __entry->ls_id, + __print_hex_str(__get_dynamic_array(res_name), + __get_dynamic_array_len(res_name)), + __entry->mode) + +); + +TRACE_EVENT(dlm_unlock_validated, + + TP_PROTO(struct dlm_ls *ls, struct dlm_lkb *lkb, struct dlm_args *args), + + TP_ARGS(ls, lkb, args), + + TP_STRUCT__entry( + __field(uint32_t, our_nodeid) + __field(uint32_t, ls_id) + __dynamic_array(unsigned char, res_name, + lkb->lkb_resource->res_length) + ), + + TP_fast_assign( + struct dlm_rsb *r = lkb->lkb_resource; + + __entry->our_nodeid = ls->ls_dn->our_node->id; + __entry->ls_id = ls->ls_global_id; + memcpy(__get_dynamic_array(res_name), r->res_name, + __get_dynamic_array_len(res_name)); + ), + + TP_printk("our_nodeid=%u ls_id=%u res_name=%s", + __entry->our_nodeid, __entry->ls_id, + __print_hex_str(__get_dynamic_array(res_name), + __get_dynamic_array_len(res_name))) + +); + DECLARE_EVENT_CLASS(dlm_rcom_template, TP_PROTO(uint32_t dst, uint32_t h_seq, const struct dlm_rcom *rc), From patchwork Tue Aug 27 18:02:36 2024 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Alexander Aring X-Patchwork-Id: 13779971 Received: from us-smtp-delivery-124.mimecast.com (us-smtp-delivery-124.mimecast.com [170.10.129.124]) (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 9020F1D4613 for ; Tue, 27 Aug 2024 18:03:36 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=170.10.129.124 ARC-Seal: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781819; cv=none; b=Bdq8wdZy4zmK7jIXwMXv8ktr2ZzpnILbba+MFZeudSfCMu/1xE7mPvfxQqq/u0AcXDn09heDkD/BJurXYN1Fj8Bwh5DL4plJb34jv6wc4nBfmFJex20LhDhaMwVHsD8+4Ve3ZQDvMNhaRdyUFkhHR1V72a5VEvyH8lZN9BYV7hw= ARC-Message-Signature: i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1724781819; c=relaxed/simple; bh=2EYGC1vMs0Hpw0whCd3bNyo1/02HGjJXSqnQRxHrocY=; h=From:To:Cc:Subject:Date:Message-ID:In-Reply-To:References: MIME-Version; b=At+DwcwIczf3nJMT4xA0J5/57rF3fezJEimF9dGPbOGjBFwVmJ4KnbDXVtk8JlnQ8/dezT0weDAXmfgK5Se9o52NTqRvnXOq9ZnJuGv/Xg1j1q/kM1r55MvqEYu7MmIAFCAPk6LuMYsH3zb2CePQuQNfclLzjhdO2mRfTFCpfAM= ARC-Authentication-Results: i=1; smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com; spf=pass smtp.mailfrom=redhat.com; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b=bSZAD5vD; arc=none smtp.client-ip=170.10.129.124 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=none dis=none) header.from=redhat.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=redhat.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="bSZAD5vD" DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1724781814; h=from:from:reply-to:subject:subject: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=OpnAX1fOM2MzxfYtvNDSARUGYAHYe8Rt2hfnrOKJSbk=; b=bSZAD5vDox0Aw/dFtPfDP/YwWvqjuMLQAe0WO0Fb92HPBJbglSoVaEpUwlgx7lIbYMP+zi qCBvgHL4YCkc1gu4Xre0p3h/MD3iai1AmlpwnsFCl65O9QOyjYzPDFaPHpbRcR3oRNzJNx lTRNFiroxO9FbFhLrKYe+f1QNaH2rec= Received: from mx-prod-mc-05.mail-002.prod.us-west-2.aws.redhat.com (ec2-54-186-198-63.us-west-2.compute.amazonaws.com [54.186.198.63]) by relay.mimecast.com with ESMTP with STARTTLS (version=TLSv1.3, cipher=TLS_AES_256_GCM_SHA384) id us-mta-361-h3AoxaQyPYG3P68gVUJ3tw-1; Tue, 27 Aug 2024 14:03:28 -0400 X-MC-Unique: h3AoxaQyPYG3P68gVUJ3tw-1 Received: from mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com [10.30.177.15]) (using TLSv1.3 with cipher TLS_AES_256_GCM_SHA384 (256/256 bits) key-exchange X25519 server-signature RSA-PSS (2048 bits) server-digest SHA256) (No client certificate requested) by mx-prod-mc-05.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTPS id 56C0419560A3; Tue, 27 Aug 2024 18:03:25 +0000 (UTC) Received: from fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com (fs-i40c-03.mgmt.fast.eng.rdu2.dc.redhat.com [10.6.24.150]) by mx-prod-int-02.mail-002.prod.us-west-2.aws.redhat.com (Postfix) with ESMTP id DE0FD1955F1B; Tue, 27 Aug 2024 18:03:21 +0000 (UTC) From: Alexander Aring To: teigland@redhat.com Cc: gfs2@lists.linux.dev, song@kernel.org, yukuai3@huawei.com, agruenba@redhat.com, mark@fasheh.com, jlbec@evilplan.org, joseph.qi@linux.alibaba.com, gregkh@linuxfoundation.org, rafael@kernel.org, akpm@linux-foundation.org, linux-kernel@vger.kernel.org, linux-raid@vger.kernel.org, ocfs2-devel@lists.linux.dev, netdev@vger.kernel.org, vvidic@valentin-vidic.from.hr, heming.zhao@suse.com, lucien.xin@gmail.com, paulmck@kernel.org, rcu@vger.kernel.org, juri.lelli@redhat.com, williams@redhat.com, aahringo@redhat.com Subject: [RFC 7/7] rv: add dlm compatible lock state kernel verifier Date: Tue, 27 Aug 2024 14:02:36 -0400 Message-ID: <20240827180236.316946-8-aahringo@redhat.com> In-Reply-To: <20240827180236.316946-1-aahringo@redhat.com> References: <20240827180236.316946-1-aahringo@redhat.com> Precedence: bulk X-Mailing-List: rcu@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 X-Scanned-By: MIMEDefang 3.0 on 10.30.177.15 This patch adds the DLM kernel lock state verifier. It can be simply activated by: echo "dlm" > /sys/kernel/tracing/rv/enabled_monitors echo "printk" > /sys/kernel/tracing/rv/monitors/dlm/reactors then run any kind of workload on DLM to check on DLM correctness in sense of compatible lock states and their current holders. For example there cannot be two lock holders or more for a specific lock holding the exclusive lock. IMPORTANT: This kernel verifier for DLM only makes sense to use it with combination of DLM's net-namespace feature to run a DLM cluster on one Linux kernel instance. This offers us a whole cluster view in the Linux tracing subsystem that this verifier takes advantage of! The above note is the reason why this verifier works kinda different than other verifiers. We build a layer to have a cluster view and check if a lock state is still compatible with all current lock holder states. That's why we have a rhashtable in the verifier to keep track of all current cluster wide locks. However we use da events to check if our model is violating or not. As example gfs2 can be used to produce some kind of filesystem benchmark that produces DLM lock handling. This verifier will then check if DLM is working as expected. Most useful feature is checking DLM recovery handling if nodes leave the lockspace and DLM is still correct. Signed-off-by: Alexander Aring --- Documentation/trace/rv/monitor_dlm.rst | 77 +++++ include/trace/events/rv.h | 9 + kernel/trace/rv/Kconfig | 18 + kernel/trace/rv/Makefile | 1 + kernel/trace/rv/monitors/dlm/dlm.c | 445 +++++++++++++++++++++++++ kernel/trace/rv/monitors/dlm/dlm.h | 38 +++ kernel/trace/rv/monitors/dlm/dlm_da.h | 143 ++++++++ tools/verification/models/dlm.dot | 14 + 8 files changed, 745 insertions(+) create mode 100644 Documentation/trace/rv/monitor_dlm.rst create mode 100644 kernel/trace/rv/monitors/dlm/dlm.c create mode 100644 kernel/trace/rv/monitors/dlm/dlm.h create mode 100644 kernel/trace/rv/monitors/dlm/dlm_da.h create mode 100644 tools/verification/models/dlm.dot diff --git a/Documentation/trace/rv/monitor_dlm.rst b/Documentation/trace/rv/monitor_dlm.rst new file mode 100644 index 000000000000..95cdf3d1a904 --- /dev/null +++ b/Documentation/trace/rv/monitor_dlm.rst @@ -0,0 +1,77 @@ +Monitor dlm +============ + +- Name: dlm - dlm runtime lock compatibility verifier. + Only makes sense with DLM net-names paces because we + need the whole traced cluster view. +- Type: per-dlm_lock deterministic automaton +- Author: Alexander Aring + +Description +----------- + +This is a per-dlm lock compatibility monitor, with the following +definition:: + + | + with | + others v + compatible +-------------+ + +--------- | | + | | valid | + +--------> | | + +-------------+ + | + | all_unlock + v + #=============# + H free H + #=============# + +This model is on a per cluster wide DLM lock basis. Each cluster +node can hold a specific lock resource in a certain lock mode. +This lock mode is either compatible or not compatible with all +other nodes holding the particular lock resource in a different +lock mode. A simple lock state is the Exclusive Lock state. Two +nodes can never held the exclusive lock for a specific lock +resource at the same time. This is what "with others compatible" +edge means when a node changes the lock state and checks if the +lock state is still compatible with other holders of the lock, we +are still in the valid state. + +If all holders for a specific lock resource that we track switch +to unlock state, we free the monitoring resource as we don't track +the lock correctness anymore. The lock can be monitored again if +the same lock resource switches to a valid lock state. + +This monitor introduce also another lock state to signal that a +lock state is in transition. The user signals a lock state change +and we waiting for a lock state completion (ast) callback. At this +time the user cannot assume to hold the state in a certain state +until completion and we need to ignore lock holders they are in +transition. + +IMPORTANT NOTE: + +This monitor makes only sense when having a cluster wide view in +the local Linux tracing subsystem. For now this means a DLM user +should construct a cluster with several nodes by using +net-namespaces. This will allow the DLM monitor to track cluster +wide lock changes. The monitor also works on a real cluster with +several machines as nodes, but it will not make any sense as we +don't check on any cluster-wide DLM correctness. Only for per-node +local DLM correctness, which is unlikely to break. + +There might be ideas to use time synchronized tracing to get a +cluster wide Linux tracing view and run the kernel verifier on +a real cluster, however this isn't supported yet and only an idea +to how to might handle it. + +This monitor is different than other current monitors. It builds +an nonexitent layer that represents the current cluster state that +we don't track in such a way in DLM. That's why it only works +with DLM and net-namespaces together. + +Specification +------------- +Grapviz Dot file in tools/verification/models/dlm.dot diff --git a/include/trace/events/rv.h b/include/trace/events/rv.h index 56592da9301c..cd031a4d994d 100644 --- a/include/trace/events/rv.h +++ b/include/trace/events/rv.h @@ -66,6 +66,15 @@ DEFINE_EVENT(error_da_monitor, error_wip, TP_PROTO(char *state, char *event), TP_ARGS(state, event)); #endif /* CONFIG_RV_MON_WIP */ +#ifdef CONFIG_RV_MON_DLM +DEFINE_EVENT(event_da_monitor, event_dlm, + TP_PROTO(char *state, char *event, char *next_state, bool final_state), + TP_ARGS(state, event, next_state, final_state)); + +DEFINE_EVENT(error_da_monitor, error_dlm, + TP_PROTO(char *state, char *event), + TP_ARGS(state, event)); +#endif /* CONFIG_RV_MON_DLM */ #endif /* CONFIG_DA_MON_EVENTS_IMPLICIT */ #ifdef CONFIG_DA_MON_EVENTS_ID diff --git a/kernel/trace/rv/Kconfig b/kernel/trace/rv/Kconfig index 831779607e84..bc12b72088c4 100644 --- a/kernel/trace/rv/Kconfig +++ b/kernel/trace/rv/Kconfig @@ -50,6 +50,24 @@ config RV_MON_WWNR For further information, see: Documentation/trace/rv/monitor_wwnr.rst +config RV_MON_DLM + depends on RV + depends on DLM + depends on NET_NS + bool "dlm monitor" + help + Enable dlm (runtime lock compatibility verifier) sample monitor, + this monitor will check on DLM lock correctness in sense of + checking on compatible lock modes during DLM runtime. E.g. two + cluster wide lock holders that holding the exclusive lock state + for a specific lock. + + IMPORTANT: the verifier only works on DLMs net-namespace feature + that is e.g. supported by gfs2. + + For further information, see: + Documentation/trace/rv/monitor_dlm.rst + config RV_REACTORS bool "Runtime verification reactors" default y diff --git a/kernel/trace/rv/Makefile b/kernel/trace/rv/Makefile index 963d14875b45..b1ac0d69ebef 100644 --- a/kernel/trace/rv/Makefile +++ b/kernel/trace/rv/Makefile @@ -3,6 +3,7 @@ obj-$(CONFIG_RV) += rv.o obj-$(CONFIG_RV_MON_WIP) += monitors/wip/wip.o obj-$(CONFIG_RV_MON_WWNR) += monitors/wwnr/wwnr.o +obj-$(CONFIG_RV_MON_DLM) += monitors/dlm/dlm.o obj-$(CONFIG_RV_REACTORS) += rv_reactors.o obj-$(CONFIG_RV_REACT_PRINTK) += reactor_printk.o obj-$(CONFIG_RV_REACT_PANIC) += reactor_panic.o diff --git a/kernel/trace/rv/monitors/dlm/dlm.c b/kernel/trace/rv/monitors/dlm/dlm.c new file mode 100644 index 000000000000..2f384b5b08b6 --- /dev/null +++ b/kernel/trace/rv/monitors/dlm/dlm.c @@ -0,0 +1,445 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#define MODULE_NAME "dlm_rv" +#include +#include + +#include "dlm.h" +#include "dlm_da.h" +#include "../../../fs/dlm/lock.h" + +/* out of DLM DLM API mode values */ +#define STATE_MODE_UNLOCK -1 +#define STATE_MODE_IN_TRANSITION -2 + +struct dlm_rv_lock_key { + uint32_t ls_id; + char res_name[DLM_RESNAME_MAXLEN]; +}; + +struct dlm_rv_lock { + union rv_dlm_lock_monitor rv; + + struct dlm_rv_lock_key key; + struct rhash_head node; + + struct list_head holders; + + /* holder for lock list */ + struct list_head list; +}; + +/* protected by dlm_rv_hash_lock */ +struct dlm_rv_holder { + unsigned int nodeid; + int mode; + + struct list_head list; +}; + +static const struct rhashtable_params dlm_rv_hash_params = { + .nelem_hint = 3, /* start small */ + .key_len = sizeof(struct dlm_rv_lock_key), + .key_offset = offsetof(struct dlm_rv_lock, key), + .head_offset = offsetof(struct dlm_rv_lock, node), + .automatic_shrinking = true, +}; + +static LIST_HEAD(dlm_rv_locks); +static struct rhashtable dlm_rv_hash; +static DEFINE_SPINLOCK(dlm_rv_hash_lock); +static struct kmem_cache *lk_cache; +static struct kmem_cache *hl_cache; + +/* + * Entry point for the per-dlm_lock monitor. + */ +#define DECLARE_DA_MON_PER_DLM_LOCK(name, type) \ + \ +DECLARE_AUTOMATA_HELPERS(name, type) \ +DECLARE_DA_MON_GENERIC_HELPERS(name, type) \ +DECLARE_DA_MON_MODEL_HANDLER_PER_DLM_LOCK(name, type) \ +DECLARE_DA_MON_INIT_PER_DLM_LOCK(name, type) \ +DECLARE_DA_MON_MONITOR_HANDLER_PER_DLM_LOCK(name, type) + +static struct rv_monitor rv_dlm; +DECLARE_DA_MON_PER_DLM_LOCK(dlm, unsigned char); + +static struct dlm_rv_lock * +lookup_lock(uint32_t ls_id, const char *res_name, size_t res_length) +{ + struct dlm_rv_lock_key key = { + .ls_id = ls_id, + }; + + WARN_ON(res_length > DLM_RESNAME_MAXLEN); + + /* the key.res_name is DLM_RESNAME_MAXLEN */ + memcpy(&key.res_name, res_name, res_length); + return rhashtable_lookup_fast(&dlm_rv_hash, &key, dlm_rv_hash_params); +} + +static struct dlm_rv_holder * +lookup_holder(struct dlm_rv_lock *lk, unsigned int nodeid) +{ + struct dlm_rv_holder *iter, *hl = NULL; + + list_for_each_entry(iter, &lk->holders, list) { + if (iter->nodeid == nodeid) { + hl = iter; + break; + } + } + + return hl; +} + +/* set a specific mode to a lock node holder identified by the nodeid */ +static void set_holder_state(struct dlm_rv_lock *lk, unsigned int nodeid, + int mode) +{ + struct dlm_rv_holder *hl; + + hl = lookup_holder(lk, nodeid); + if (hl) { + hl->mode = mode; + return; + } + + /* we only create holders they are not start with UNLOCK */ + if (mode == STATE_MODE_UNLOCK) + return; + + hl = kmem_cache_zalloc(hl_cache, GFP_ATOMIC); + if (WARN_ON_ONCE(!hl)) + return; + + hl->nodeid = nodeid; + hl->mode = mode; + + list_add(&hl->list, &lk->holders); +} + +/* check if all lock holders except the one from nodeid is still + * compatible with the mode given by mode. Usually the nodeid which + * is skipped has the applied mode as parameter to check if the + * state change is valid. + */ +static int check_valid_lock_holders(struct dlm_rv_lock *lk, int mode, + unsigned int nodeid) +{ + struct dlm_rv_holder *hl; + + list_for_each_entry(hl, &lk->holders, list) { + /* ignore ourself + * ignore pending lock states + */ + if (hl->nodeid == nodeid || + hl->mode == STATE_MODE_IN_TRANSITION) + continue; + + if (!dlm_modes_compat(mode, hl->mode)) + return 0; + } + + return 1; +} + +/* check if all holders for a lock are in unlock state */ +static int check_all_unlock_holders(struct dlm_rv_lock *lk) +{ + struct dlm_rv_holder *hl; + + /* should never happen but when we delete the lk */ + if (WARN_ON(list_empty(&lk->holders))) + return 1; + + list_for_each_entry(hl, &lk->holders, list) { + if (hl->mode != STATE_MODE_UNLOCK) + return 0; + } + + return 1; +} + +/* drop all lock holders for a specific lock */ +static void drop_all_lock_holders(struct dlm_rv_lock *lk) +{ + struct dlm_rv_holder *hl, *tmp; + + list_for_each_entry_safe(hl, tmp, &lk->holders, list) { + list_del(&hl->list); + kmem_cache_free(hl_cache, hl); + } +} + +/* unlock specific lock holder if available and if all lock holders + * are in unlock state, we remove and free the lock. + */ +static void unlock_lock_holder(struct dlm_rv_lock *lk, unsigned int nodeid) +{ + int rv; + + set_holder_state(lk, nodeid, STATE_MODE_UNLOCK); + rv = check_all_unlock_holders(lk); + if (rv) { + drop_all_lock_holders(lk); + + list_del(&lk->list); + rhashtable_remove_fast(&dlm_rv_hash, &lk->node, + dlm_rv_hash_params); + /* move into final state */ + da_handle_event_dlm(lk, all_unlock_dlm); + kmem_cache_free(lk_cache, lk); + } +} + +static void handle_dlm_ast(void *data, unsigned int our_nodeid, __u32 ls_id, + __u32 lkb_id, __u8 sb_flags, int sb_status, + const char *res_name, size_t res_length, int mode) +{ + struct dlm_rv_holder *hl; + struct dlm_rv_lock *lk; + int rv; + + switch (sb_status) { + case -DLM_EUNLOCK: + /* handle an unlock of an lock we saw before */ + spin_lock_bh(&dlm_rv_hash_lock); + /* switch to unlock state if there is a lock available + * and check if all locks are in unlock mode, see + * unlock_lock_holder(). + */ + lk = lookup_lock(ls_id, res_name, res_length); + if (lk) + unlock_lock_holder(lk, our_nodeid); + spin_unlock_bh(&dlm_rv_hash_lock); + return; + case 0: + /* successful lock state change */ + break; + default: + /* ignored */ + return; + } + + spin_lock_bh(&dlm_rv_hash_lock); + lk = lookup_lock(ls_id, res_name, res_length); + if (!lk) { + /* start to begin tracking DLM cluster lock */ + lk = kmem_cache_zalloc(lk_cache, GFP_ATOMIC); + if (WARN_ON_ONCE(!lk)) { + spin_unlock_bh(&dlm_rv_hash_lock); + return; + } + + lk->key.ls_id = ls_id; + memcpy(lk->key.res_name, res_name, res_length); + INIT_LIST_HEAD(&lk->holders); + + da_monitor_reset_dlm(da_get_monitor_dlm(lk)); + da_handle_start_event_dlm(lk, with_others_compatible_dlm); + set_holder_state(lk, our_nodeid, mode); + + list_add_tail(&lk->list, &dlm_rv_locks); + rv = rhashtable_insert_fast(&dlm_rv_hash, &lk->node, + dlm_rv_hash_params); + spin_unlock_bh(&dlm_rv_hash_lock); + + WARN_ON(rv); + return; + } + + /* lock is known, change it's state and check if it doesn't + * violate the DLM cluster wide compatible lock modes + */ + set_holder_state(lk, our_nodeid, mode); + rv = check_valid_lock_holders(lk, mode, our_nodeid); + if (rv) { + /* the whole validation process, this event signals + * everything is fine and DLM works correctly there + * are no cluster-wide locks that violates DLM locking. + */ + da_handle_event_dlm(lk, with_others_compatible_dlm); + } else { + /* print all holders of the lock when a invalid lock state is entered */ + console_lock(); + pr_info("---\n"); + pr_info("ls_id %u lkb_id: 0x%08x\n", ls_id, lkb_id); + pr_info("holders:\n"); + list_for_each_entry(hl, &lk->holders, list) { + pr_info("\tnodeid: %u mode: %d\n", hl->nodeid, + hl->mode); + } + pr_info("---\n"); + console_unlock(); + + /* move into an invalid state change, we don't have a edge for that + * so we just use event_max_dlm. + */ + da_handle_event_dlm(lk, event_max_dlm); + } + spin_unlock_bh(&dlm_rv_hash_lock); +} + +/* set the holder to transition state as lock downgrades can issue + * grant messages to other nodes we need to ignore if a lock on a + * specific node is in state transition. From point of DLM API + * the user cannot assume to still hold the lock at this point + * anyway. + */ +static void set_holder_transition(uint32_t ls_id, const char *res_name, + size_t res_length, uint32_t our_nodeid) +{ + struct dlm_rv_holder *hl; + struct dlm_rv_lock *lk; + + spin_lock_bh(&dlm_rv_hash_lock); + lk = lookup_lock(ls_id, res_name, res_length); + if (lk) { + hl = lookup_holder(lk, our_nodeid); + if (hl) + hl->mode = STATE_MODE_IN_TRANSITION; + } + spin_unlock_bh(&dlm_rv_hash_lock); +} + +/* after a lock request got validated it cannot fail */ +static void handle_dlm_lock_validated(void *data, struct dlm_ls *ls, + struct dlm_lkb *lkb, + struct dlm_args *args, + const char *res_name, size_t res_length) +{ + set_holder_transition(ls->ls_global_id, res_name, + res_length, ls->ls_dn->our_node->id); +} + +static void handle_dlm_unlock_validated(void *data, struct dlm_ls *ls, + struct dlm_lkb *lkb, + struct dlm_args *args) +{ + set_holder_transition(ls->ls_global_id, + lkb->lkb_resource->res_name, + lkb->lkb_resource->res_length, + ls->ls_dn->our_node->id); +} + +/* remove all holders, recovery will fast this up and we need to drop them */ +static void handle_dlm_release_lockspace(void *data, unsigned int our_nodeid, + __u32 ls_id) +{ + struct dlm_rv_lock *lk, *lk_tmp; + + spin_lock_bh(&dlm_rv_hash_lock); + list_for_each_entry_safe(lk, lk_tmp, &dlm_rv_locks, list) { + if (lk->key.ls_id != ls_id) + continue; + + /* unlock all locks for the node that calls + * dlm_release_lockspace(). It's not necessary + * from the DLM API that a node need to unlock + * all locks before calling dlm_release_lockspace() + * there is even an optimization because each recovery + * will deal with that locally. However we handle a + * dlm_release_lockspace() on a specific node as + * unlock all locks. + */ + unlock_lock_holder(lk, our_nodeid); + } + spin_unlock_bh(&dlm_rv_hash_lock); +} + +static void rhash_lock_free(void *ptr, void *arg) +{ + struct dlm_rv_lock *lk = ptr; + + list_del(&lk->list); + drop_all_lock_holders(lk); + kmem_cache_free(lk_cache, lk); +} + +static int enable_dlm(void) +{ + int retval; + + retval = rhashtable_init(&dlm_rv_hash, &dlm_rv_hash_params); + if (retval) + return retval; + + retval = da_monitor_init_dlm(); + if (retval) { + rhashtable_destroy(&dlm_rv_hash); + return retval; + } + + rv_attach_trace_probe("dlm", dlm_ast, handle_dlm_ast); + rv_attach_trace_probe("dlm", dlm_lock_validated, handle_dlm_lock_validated); + rv_attach_trace_probe("dlm", dlm_unlock_validated, handle_dlm_unlock_validated); + rv_attach_trace_probe("dlm", dlm_release_lockspace, handle_dlm_release_lockspace); + + return 0; +} + +static void disable_dlm(void) +{ + rv_dlm.enabled = 0; + + rv_detach_trace_probe("dlm", dlm_ast, handle_dlm_ast); + rv_detach_trace_probe("dlm", dlm_lock_validated, handle_dlm_lock_validated); + rv_detach_trace_probe("dlm", dlm_unlock_validated, handle_dlm_unlock_validated); + rv_detach_trace_probe("dlm", dlm_release_lockspace, handle_dlm_release_lockspace); + + da_monitor_destroy_dlm(); + + rhashtable_free_and_destroy(&dlm_rv_hash, rhash_lock_free, NULL); +} + +static struct rv_monitor rv_dlm = { + .name = "dlm", + .description = "dlm runtime lock compatibility verifier", + .enable = enable_dlm, + .disable = disable_dlm, + .reset = da_monitor_reset_all_dlm, + .enabled = 0, +}; + +static int __init register_dlm(void) +{ + lk_cache = KMEM_CACHE(dlm_rv_lock, 0); + if (!lk_cache) + return -ENOMEM; + + hl_cache = KMEM_CACHE(dlm_rv_holder, 0); + if (!hl_cache) { + kmem_cache_destroy(lk_cache); + return -ENOMEM; + } + + rv_register_monitor(&rv_dlm); + return 0; +} + +static void __exit unregister_dlm(void) +{ + rv_unregister_monitor(&rv_dlm); + + kmem_cache_destroy(hl_cache); + kmem_cache_destroy(lk_cache); +} + +module_init(register_dlm); +module_exit(unregister_dlm); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Alexander Aring "); +MODULE_DESCRIPTION("dlm: runtime lock compatibility verifier"); diff --git a/kernel/trace/rv/monitors/dlm/dlm.h b/kernel/trace/rv/monitors/dlm/dlm.h new file mode 100644 index 000000000000..514614be2ca9 --- /dev/null +++ b/kernel/trace/rv/monitors/dlm/dlm.h @@ -0,0 +1,38 @@ +enum states_dlm { + valid_dlm = 0, + free_dlm, + state_max_dlm +}; + +#define INVALID_STATE state_max_dlm + +enum events_dlm { + all_unlock_dlm = 0, + with_others_compatible_dlm, + event_max_dlm +}; + +struct automaton_dlm { + char *state_names[state_max_dlm]; + char *event_names[event_max_dlm]; + unsigned char function[state_max_dlm][event_max_dlm]; + unsigned char initial_state; + bool final_states[state_max_dlm]; +}; + +static const struct automaton_dlm automaton_dlm = { + .state_names = { + "valid", + "free" + }, + .event_names = { + "all_unlock", + "with_others_compatible" + }, + .function = { + { free_dlm, valid_dlm }, + { INVALID_STATE, INVALID_STATE }, + }, + .initial_state = valid_dlm, + .final_states = { 0, 1 }, +}; diff --git a/kernel/trace/rv/monitors/dlm/dlm_da.h b/kernel/trace/rv/monitors/dlm/dlm_da.h new file mode 100644 index 000000000000..064ed5085b30 --- /dev/null +++ b/kernel/trace/rv/monitors/dlm/dlm_da.h @@ -0,0 +1,143 @@ +// SPDX-License-Identifier: GPL-2.0-only + +#ifndef __DLM_DA_RV__ +#define __DLM_DA_RV__ + +/* + * Event handler for per_dlm_lock monitors. + */ +#define DECLARE_DA_MON_MODEL_HANDLER_PER_DLM_LOCK(name, type) \ + \ +static inline bool da_event_##name(struct da_monitor *da_mon, struct dlm_rv_lock *lk, \ + enum events_##name event) \ +{ \ + type curr_state = da_monitor_curr_state_##name(da_mon); \ + type next_state = model_get_next_state_##name(curr_state, event); \ + \ + if (next_state != INVALID_STATE) { \ + da_monitor_set_state_##name(da_mon, next_state); \ + \ + trace_event_##name(model_get_state_name_##name(curr_state), \ + model_get_event_name_##name(event), \ + model_get_state_name_##name(next_state), \ + model_is_final_state_##name(next_state)); \ + \ + return true; \ + } \ + \ + if (rv_reacting_on_##name()) \ + cond_react_##name(format_react_msg_##name(curr_state, event)); \ + \ + trace_error_##name(model_get_state_name_##name(curr_state), \ + model_get_event_name_##name(event)); \ + \ + return false; \ +} + +/* + * Functions to define, init and get a per-dlm-lock monitor. + */ +#define DECLARE_DA_MON_INIT_PER_DLM_LOCK(name, type) \ + \ +/* \ + * da_get_monitor_##name - return the monitor in the allocated slot for tsk \ + */ \ +static inline struct da_monitor *da_get_monitor_##name(struct dlm_rv_lock *lk) \ +{ \ + return &lk->rv.da_mon; \ +} \ + \ +static void da_monitor_reset_all_##name(void) \ +{ \ +} \ + \ +/* \ + * da_monitor_init_##name - initialize the per-task monitor \ + * \ + * Try to allocate a slot in the task's vector of monitors. If there \ + * is an available slot, use it and reset all task's monitor. \ + */ \ +static int da_monitor_init_##name(void) \ +{ \ + da_monitor_reset_all_##name(); \ + return 0; \ +} \ + \ +/* \ + * da_monitor_destroy_##name - return the allocated slot \ + */ \ +static inline void da_monitor_destroy_##name(void) \ +{ \ + return; \ +} + +/* + * Handle event for per task. + */ +#define DECLARE_DA_MON_MONITOR_HANDLER_PER_DLM_LOCK(name, type) \ + \ +static inline void \ +__da_handle_event_##name(struct da_monitor *da_mon, struct dlm_rv_lock *lk, \ + enum events_##name event) \ +{ \ + bool retval; \ + \ + retval = da_event_##name(da_mon, lk, event); \ + if (!retval) \ + da_monitor_reset_##name(da_mon); \ +} \ + \ +/* \ + * da_handle_event_##name - handle an event \ + */ \ +static inline void \ +da_handle_event_##name(struct dlm_rv_lock *lk, enum events_##name event) \ +{ \ + struct da_monitor *da_mon = da_get_monitor_##name(lk); \ + bool retval; \ + \ + retval = da_monitor_handling_event_##name(da_mon); \ + if (!retval) \ + return; \ + \ + __da_handle_event_##name(da_mon, lk, event); \ +} \ + \ +/* \ + * da_handle_start_event_##name - start monitoring or handle event \ + * \ + * This function is used to notify the monitor that the system is returning \ + * to the initial state, so the monitor can start monitoring in the next event. \ + * Thus: \ + * \ + * If the monitor already started, handle the event. \ + * If the monitor did not start yet, start the monitor but skip the event. \ + */ \ +static inline bool \ +da_handle_start_event_##name(struct dlm_rv_lock *lk, enum events_##name event) \ +{ \ + struct da_monitor *da_mon; \ + \ + if (!da_monitor_enabled_##name()) \ + return 0; \ + \ + da_mon = da_get_monitor_##name(lk); \ + \ + if (unlikely(!da_monitoring_##name(da_mon))) { \ + da_monitor_start_##name(da_mon); \ + return 0; \ + } \ + \ + __da_handle_event_##name(da_mon, lk, event); \ + \ + return 1; \ +} + +/* + * Futher monitor types are expected, so make this a union. + */ +union rv_dlm_lock_monitor { + struct da_monitor da_mon; +}; + +#endif /* __DLM_DA_RV__ */ diff --git a/tools/verification/models/dlm.dot b/tools/verification/models/dlm.dot new file mode 100644 index 000000000000..43092c865e3b --- /dev/null +++ b/tools/verification/models/dlm.dot @@ -0,0 +1,14 @@ +digraph state_automaton { + {node [shape = circle] "valid"}; + {node [shape = plaintext, style=invis, label=""] "__init_valid"}; + {node [shape = doublecircle] "free"}; + "__init_valid" -> "valid"; + "valid" [label = "valid", color = green3] + "valid" -> "valid" [ label = "with_others_compatible" ]; + "free" [label = "free"] + "valid" -> "free" [ label = "all_unlock" ]; + { rank = min ; + "__init_valid"; + "valid"; + } +}