From patchwork Fri Nov 1 17:34:09 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223481 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 9D7C51515 for ; Fri, 1 Nov 2019 17:34:24 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 7B4DD217D9 for ; Fri, 1 Nov 2019 17:34:24 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="iEwwvhCE" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729542AbfKAReT (ORCPT ); Fri, 1 Nov 2019 13:34:19 -0400 Received: from us-smtp-1.mimecast.com ([207.211.31.81]:35065 "EHLO us-smtp-delivery-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729543AbfKAReT (ORCPT ); Fri, 1 Nov 2019 13:34:19 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629658; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=1ofVwMOE/ATx0q/xqI7IgeNPgnC6+tkVYzw7GkmuoBw=; b=iEwwvhCEEoULt6sefCHQjdii2rErNmYN5oEImcCII6djjFE4lCHRxRU14xlKPou+y7lHEs cLFyBSrKtohBt36jYJCmfnLMGO7kKT2hGGE2Top8W2a3eja2ENrpGUZHZUFfYSBB9Ung/J 6TNSzkmSJVa57Mve3zhlzZbws8dwJkI= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-251-p20JZCVjPPSL27BybuU1jg-1; Fri, 01 Nov 2019 13:34:14 -0400 Received: from smtp.corp.redhat.com (int-mx06.intmail.prod.int.phx2.redhat.com [10.5.11.16]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id BA84C800D49; Fri, 1 Nov 2019 17:34:12 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id BD54F5C290; Fri, 1 Nov 2019 17:34:09 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 01/11] pipe: Reduce #inclusion of pipe_fs_i.h [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:34:09 +0000 Message-ID: <157262964898.13142.10761929870592354381.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.16 X-MC-Unique: p20JZCVjPPSL27BybuU1jg-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Remove some #inclusions of linux/pipe_fs_i.h that don't seem to be necessary any more. Signed-off-by: David Howells --- fs/exec.c | 1 - fs/ocfs2/aops.c | 1 - security/smack/smack_lsm.c | 1 - 3 files changed, 3 deletions(-) diff --git a/fs/exec.c b/fs/exec.c index 555e93c7dec8..57bc7ef8d31b 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -59,7 +59,6 @@ #include #include #include -#include #include #include #include diff --git a/fs/ocfs2/aops.c b/fs/ocfs2/aops.c index 8de1c9d644f6..c50ac6b7415b 100644 --- a/fs/ocfs2/aops.c +++ b/fs/ocfs2/aops.c @@ -11,7 +11,6 @@ #include #include #include -#include #include #include #include diff --git a/security/smack/smack_lsm.c b/security/smack/smack_lsm.c index abeb09c30633..ecea41ce919b 100644 --- a/security/smack/smack_lsm.c +++ b/security/smack/smack_lsm.c @@ -28,7 +28,6 @@ #include #include #include -#include #include #include #include From patchwork Fri Nov 1 17:34:18 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223493 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id A1E1C1515 for ; Fri, 1 Nov 2019 17:34:35 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 77B8021928 for ; Fri, 1 Nov 2019 17:34:35 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="EEzFw/Uq" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729573AbfKARee (ORCPT ); Fri, 1 Nov 2019 13:34:34 -0400 Received: from us-smtp-2.mimecast.com ([207.211.31.81]:21959 "EHLO us-smtp-delivery-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729584AbfKARed (ORCPT ); Fri, 1 Nov 2019 13:34:33 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629671; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=OLwH1xhZtcs6oehBhP4aFawQY9wwncddHsWF2TUeyN4=; b=EEzFw/UqZ/rbuPTgxGLi7yLkEJLqZrWTak0Pej3M09dodsIG5sy7H6VWF29Uh1+ywpaot0 gh+lw/DvJTf+H7mav5KkgXSne1R4UptLuHIfXXMptF+tWwBZukJTB98qzbD7MjrmYIaqs3 5yDfOpEX8pE/76rsmWguKsWuaNXLHCQ= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-157-bpXnJyK9PXO9uxLy8SjMgQ-1; Fri, 01 Nov 2019 13:34:23 -0400 Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.phx2.redhat.com [10.5.11.12]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 512BB1800D7D; Fri, 1 Nov 2019 17:34:21 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id AA7C360BEC; Fri, 1 Nov 2019 17:34:18 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 02/11] Remove the nr_exclusive argument from __wake_up_sync_key() [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:34:18 +0000 Message-ID: <157262965796.13142.7304635748939123321.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.12 X-MC-Unique: bpXnJyK9PXO9uxLy8SjMgQ-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Remove the nr_exclusive argument from __wake_up_sync_key() and derived functions as everything seems to set it to 1. Note also that if it wasn't set to 1, it would clear WF_SYNC anyway. Signed-off-by: David Howells Acked-by: Peter Zijlstra (Intel) --- include/linux/wait.h | 8 ++++---- kernel/exit.c | 2 +- kernel/sched/wait.c | 14 ++++---------- 3 files changed, 9 insertions(+), 15 deletions(-) diff --git a/include/linux/wait.h b/include/linux/wait.h index 3eb7cae8206c..bb7676d396cd 100644 --- a/include/linux/wait.h +++ b/include/linux/wait.h @@ -201,9 +201,9 @@ void __wake_up(struct wait_queue_head *wq_head, unsigned int mode, int nr, void void __wake_up_locked_key(struct wait_queue_head *wq_head, unsigned int mode, void *key); void __wake_up_locked_key_bookmark(struct wait_queue_head *wq_head, unsigned int mode, void *key, wait_queue_entry_t *bookmark); -void __wake_up_sync_key(struct wait_queue_head *wq_head, unsigned int mode, int nr, void *key); +void __wake_up_sync_key(struct wait_queue_head *wq_head, unsigned int mode, void *key); void __wake_up_locked(struct wait_queue_head *wq_head, unsigned int mode, int nr); -void __wake_up_sync(struct wait_queue_head *wq_head, unsigned int mode, int nr); +void __wake_up_sync(struct wait_queue_head *wq_head, unsigned int mode); #define wake_up(x) __wake_up(x, TASK_NORMAL, 1, NULL) #define wake_up_nr(x, nr) __wake_up(x, TASK_NORMAL, nr, NULL) @@ -214,7 +214,7 @@ void __wake_up_sync(struct wait_queue_head *wq_head, unsigned int mode, int nr); #define wake_up_interruptible(x) __wake_up(x, TASK_INTERRUPTIBLE, 1, NULL) #define wake_up_interruptible_nr(x, nr) __wake_up(x, TASK_INTERRUPTIBLE, nr, NULL) #define wake_up_interruptible_all(x) __wake_up(x, TASK_INTERRUPTIBLE, 0, NULL) -#define wake_up_interruptible_sync(x) __wake_up_sync((x), TASK_INTERRUPTIBLE, 1) +#define wake_up_interruptible_sync(x) __wake_up_sync((x), TASK_INTERRUPTIBLE) /* * Wakeup macros to be used to report events to the targets. @@ -228,7 +228,7 @@ void __wake_up_sync(struct wait_queue_head *wq_head, unsigned int mode, int nr); #define wake_up_interruptible_poll(x, m) \ __wake_up(x, TASK_INTERRUPTIBLE, 1, poll_to_key(m)) #define wake_up_interruptible_sync_poll(x, m) \ - __wake_up_sync_key((x), TASK_INTERRUPTIBLE, 1, poll_to_key(m)) + __wake_up_sync_key((x), TASK_INTERRUPTIBLE, poll_to_key(m)) #define ___wait_cond_timeout(condition) \ ({ \ diff --git a/kernel/exit.c b/kernel/exit.c index a46a50d67002..a1ff25ef050e 100644 --- a/kernel/exit.c +++ b/kernel/exit.c @@ -1435,7 +1435,7 @@ static int child_wait_callback(wait_queue_entry_t *wait, unsigned mode, void __wake_up_parent(struct task_struct *p, struct task_struct *parent) { __wake_up_sync_key(&parent->signal->wait_chldexit, - TASK_INTERRUPTIBLE, 1, p); + TASK_INTERRUPTIBLE, p); } static long do_wait(struct wait_opts *wo) diff --git a/kernel/sched/wait.c b/kernel/sched/wait.c index c1e566a114ca..b4b52361dab7 100644 --- a/kernel/sched/wait.c +++ b/kernel/sched/wait.c @@ -169,7 +169,6 @@ EXPORT_SYMBOL_GPL(__wake_up_locked_key_bookmark); * __wake_up_sync_key - wake up threads blocked on a waitqueue. * @wq_head: the waitqueue * @mode: which threads - * @nr_exclusive: how many wake-one or wake-many threads to wake up * @key: opaque value to be passed to wakeup targets * * The sync wakeup differs that the waker knows that it will schedule @@ -183,26 +182,21 @@ EXPORT_SYMBOL_GPL(__wake_up_locked_key_bookmark); * accessing the task state. */ void __wake_up_sync_key(struct wait_queue_head *wq_head, unsigned int mode, - int nr_exclusive, void *key) + void *key) { - int wake_flags = 1; /* XXX WF_SYNC */ - if (unlikely(!wq_head)) return; - if (unlikely(nr_exclusive != 1)) - wake_flags = 0; - - __wake_up_common_lock(wq_head, mode, nr_exclusive, wake_flags, key); + __wake_up_common_lock(wq_head, mode, 1, WF_SYNC, key); } EXPORT_SYMBOL_GPL(__wake_up_sync_key); /* * __wake_up_sync - see __wake_up_sync_key() */ -void __wake_up_sync(struct wait_queue_head *wq_head, unsigned int mode, int nr_exclusive) +void __wake_up_sync(struct wait_queue_head *wq_head, unsigned int mode) { - __wake_up_sync_key(wq_head, mode, nr_exclusive, NULL); + __wake_up_sync_key(wq_head, mode, NULL); } EXPORT_SYMBOL_GPL(__wake_up_sync); /* For internal use only */ From patchwork Fri Nov 1 17:34:26 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223497 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 2F997913 for ; Fri, 1 Nov 2019 17:34:40 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 0EB8C21897 for ; Fri, 1 Nov 2019 17:34:40 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="J73b8rWV" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729623AbfKARej (ORCPT ); Fri, 1 Nov 2019 13:34:39 -0400 Received: from us-smtp-delivery-1.mimecast.com ([207.211.31.120]:51247 "EHLO us-smtp-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729443AbfKARei (ORCPT ); Fri, 1 Nov 2019 13:34:38 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629678; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=jLpXdslDrrhp+REdHr3l91hluU6shWm/KAlmhAfJqp8=; b=J73b8rWV2gk10JFe0x/QyokYOgxlB+NdAiMS3AvE1HRllTJAzUus8tr06vVKRMOY5LXdhQ Q/xTt7t4HZksbFnzBjFd4peMExCC69QUP/J0c7pSniPAMHYMRvyS2HfV4pFBDQuUVaPSxo ayR1dTmuVJQNznxLIKhAvPt9n2F5ujM= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-349-GwIvBTtLNaqHYo-nxOu7PA-1; Fri, 01 Nov 2019 13:34:34 -0400 Received: from smtp.corp.redhat.com (int-mx01.intmail.prod.int.phx2.redhat.com [10.5.11.11]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 499B6107ACC0; Fri, 1 Nov 2019 17:34:32 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id 690DB600D1; Fri, 1 Nov 2019 17:34:27 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 03/11] Add wake_up_interruptible_sync_poll_locked() [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:34:26 +0000 Message-ID: <157262966653.13142.15080563290216609050.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.11 X-MC-Unique: GwIvBTtLNaqHYo-nxOu7PA-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Add a wakeup call for a case whereby the caller already has the waitqueue spinlock held. This can be used by pipes to alter the ring buffer indices and issue a wakeup under the same spinlock. Signed-off-by: David Howells Acked-by: Peter Zijlstra (Intel) --- include/linux/wait.h | 3 +++ kernel/sched/wait.c | 23 +++++++++++++++++++++++ 2 files changed, 26 insertions(+) diff --git a/include/linux/wait.h b/include/linux/wait.h index bb7676d396cd..3283c8d02137 100644 --- a/include/linux/wait.h +++ b/include/linux/wait.h @@ -202,6 +202,7 @@ void __wake_up_locked_key(struct wait_queue_head *wq_head, unsigned int mode, vo void __wake_up_locked_key_bookmark(struct wait_queue_head *wq_head, unsigned int mode, void *key, wait_queue_entry_t *bookmark); void __wake_up_sync_key(struct wait_queue_head *wq_head, unsigned int mode, void *key); +void __wake_up_locked_sync_key(struct wait_queue_head *wq_head, unsigned int mode, void *key); void __wake_up_locked(struct wait_queue_head *wq_head, unsigned int mode, int nr); void __wake_up_sync(struct wait_queue_head *wq_head, unsigned int mode); @@ -229,6 +230,8 @@ void __wake_up_sync(struct wait_queue_head *wq_head, unsigned int mode); __wake_up(x, TASK_INTERRUPTIBLE, 1, poll_to_key(m)) #define wake_up_interruptible_sync_poll(x, m) \ __wake_up_sync_key((x), TASK_INTERRUPTIBLE, poll_to_key(m)) +#define wake_up_interruptible_sync_poll_locked(x, m) \ + __wake_up_locked_sync_key((x), TASK_INTERRUPTIBLE, poll_to_key(m)) #define ___wait_cond_timeout(condition) \ ({ \ diff --git a/kernel/sched/wait.c b/kernel/sched/wait.c index b4b52361dab7..ba059fbfc53a 100644 --- a/kernel/sched/wait.c +++ b/kernel/sched/wait.c @@ -191,6 +191,29 @@ void __wake_up_sync_key(struct wait_queue_head *wq_head, unsigned int mode, } EXPORT_SYMBOL_GPL(__wake_up_sync_key); +/** + * __wake_up_locked_sync_key - wake up a thread blocked on a locked waitqueue. + * @wq_head: the waitqueue + * @mode: which threads + * @key: opaque value to be passed to wakeup targets + * + * The sync wakeup differs in that the waker knows that it will schedule + * away soon, so while the target thread will be woken up, it will not + * be migrated to another CPU - ie. the two threads are 'synchronized' + * with each other. This can prevent needless bouncing between CPUs. + * + * On UP it can prevent extra preemption. + * + * If this function wakes up a task, it executes a full memory barrier before + * accessing the task state. + */ +void __wake_up_locked_sync_key(struct wait_queue_head *wq_head, + unsigned int mode, void *key) +{ + __wake_up_common(wq_head, mode, 1, WF_SYNC, key, NULL); +} +EXPORT_SYMBOL_GPL(__wake_up_locked_sync_key); + /* * __wake_up_sync - see __wake_up_sync_key() */ From patchwork Fri Nov 1 17:34:37 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223509 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 756BC913 for ; Fri, 1 Nov 2019 17:34:54 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 238DC218DE for ; Fri, 1 Nov 2019 17:34:54 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="FckLYBTu" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729689AbfKARew (ORCPT ); Fri, 1 Nov 2019 13:34:52 -0400 Received: from us-smtp-1.mimecast.com ([207.211.31.81]:59186 "EHLO us-smtp-delivery-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1728977AbfKARev (ORCPT ); Fri, 1 Nov 2019 13:34:51 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629687; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=YvdbzxCpv+fICSYb83023SgSKfIs5t/bfDI2DJc3eKg=; b=FckLYBTurtTVkFRlgwcjY6tlQsDRc2wpyakPvbVeHJ2wSKPMRNdYnLMMFkzARGpldP8Nua T3jZUWY66kdo0m3EfLhVJJVwGIW3m3Zy1pjXVT30gkgWrEFnPGa6fTbFsOpXA+gMa4YAVP E1jUsxBGQAW+8e6hcR9jZizfRww6URY= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-241-PAoLjEuMP7-FV82Z9sdXNA-1; Fri, 01 Nov 2019 13:34:43 -0400 Received: from smtp.corp.redhat.com (int-mx07.intmail.prod.int.phx2.redhat.com [10.5.11.22]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 97E99800A1A; Fri, 1 Nov 2019 17:34:41 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id 4FB691001920; Fri, 1 Nov 2019 17:34:38 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 04/11] pipe: Use head and tail pointers for the ring, not cursor and length [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:34:37 +0000 Message-ID: <157262967752.13142.696874122947836210.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.5.11.22 X-MC-Unique: PAoLjEuMP7-FV82Z9sdXNA-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Convert pipes to use head and tail pointers for the buffer ring rather than pointer and length as the latter requires two atomic ops to update (or a combined op) whereas the former only requires one. (1) The head pointer is the point at which production occurs and points to the slot in which the next buffer will be placed. This is equivalent to pipe->curbuf + pipe->nrbufs. The head pointer belongs to the write-side. (2) The tail pointer is the point at which consumption occurs. It points to the next slot to be consumed. This is equivalent to pipe->curbuf. The tail pointer belongs to the read-side. (3) head and tail are allowed to run to UINT_MAX and wrap naturally. They are only masked off when the array is being accessed, e.g.: pipe->bufs[head & mask] This means that it is not necessary to have a dead slot in the ring as head == tail isn't ambiguous. (4) The ring is empty if "head == tail". A helper, pipe_empty(), is provided for this. (5) The occupancy of the ring is "head - tail". A helper, pipe_occupancy(), is provided for this. (6) The number of free slots in the ring is "pipe->ring_size - occupancy". A helper, pipe_space_for_user() is provided to indicate how many slots userspace may use. (7) The ring is full if "head - tail >= pipe->ring_size". A helper, pipe_full(), is provided for this. Signed-off-by: David Howells Reported-by: kernel test robot --- drivers/char/virtio_console.c | 16 +- fs/fuse/dev.c | 31 +++-- fs/pipe.c | 170 +++++++++++++++----------- fs/splice.c | 190 ++++++++++++++++++----------- include/linux/pipe_fs_i.h | 60 +++++++++ include/linux/uio.h | 4 - lib/iov_iter.c | 269 +++++++++++++++++++++++------------------ 7 files changed, 448 insertions(+), 292 deletions(-) diff --git a/drivers/char/virtio_console.c b/drivers/char/virtio_console.c index 7270e7b69262..de35c56b3d96 100644 --- a/drivers/char/virtio_console.c +++ b/drivers/char/virtio_console.c @@ -919,6 +919,7 @@ static ssize_t port_fops_splice_write(struct pipe_inode_info *pipe, .pos = *ppos, .u.data = &sgl, }; + unsigned int occupancy; /* * Rproc_serial does not yet support splice. To support splice @@ -929,21 +930,18 @@ static ssize_t port_fops_splice_write(struct pipe_inode_info *pipe, if (is_rproc_serial(port->out_vq->vdev)) return -EINVAL; - /* - * pipe->nrbufs == 0 means there are no data to transfer, - * so this returns just 0 for no data. - */ pipe_lock(pipe); - if (!pipe->nrbufs) { - ret = 0; + ret = 0; + if (pipe_empty(pipe->head, pipe->tail)) goto error_out; - } ret = wait_port_writable(port, filp->f_flags & O_NONBLOCK); if (ret < 0) goto error_out; - buf = alloc_buf(port->portdev->vdev, 0, pipe->nrbufs); + occupancy = pipe_occupancy(pipe->head, pipe->tail); + buf = alloc_buf(port->portdev->vdev, 0, occupancy); + if (!buf) { ret = -ENOMEM; goto error_out; @@ -951,7 +949,7 @@ static ssize_t port_fops_splice_write(struct pipe_inode_info *pipe, sgl.n = 0; sgl.len = 0; - sgl.size = pipe->nrbufs; + sgl.size = occupancy; sgl.sg = buf->sg; sg_init_table(sgl.sg, sgl.size); ret = __splice_from_pipe(pipe, &sd, pipe_to_sg); diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index dadd617d826c..c56011f95a87 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -703,7 +703,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs) cs->pipebufs++; cs->nr_segs--; } else { - if (cs->nr_segs == cs->pipe->buffers) + if (cs->nr_segs >= cs->pipe->ring_size) return -EIO; page = alloc_page(GFP_HIGHUSER); @@ -879,7 +879,7 @@ static int fuse_ref_page(struct fuse_copy_state *cs, struct page *page, struct pipe_buffer *buf; int err; - if (cs->nr_segs == cs->pipe->buffers) + if (cs->nr_segs >= cs->pipe->ring_size) return -EIO; err = unlock_request(cs->req); @@ -1341,7 +1341,7 @@ static ssize_t fuse_dev_splice_read(struct file *in, loff_t *ppos, if (!fud) return -EPERM; - bufs = kvmalloc_array(pipe->buffers, sizeof(struct pipe_buffer), + bufs = kvmalloc_array(pipe->ring_size, sizeof(struct pipe_buffer), GFP_KERNEL); if (!bufs) return -ENOMEM; @@ -1353,7 +1353,7 @@ static ssize_t fuse_dev_splice_read(struct file *in, loff_t *ppos, if (ret < 0) goto out; - if (pipe->nrbufs + cs.nr_segs > pipe->buffers) { + if (pipe_occupancy(pipe->head, pipe->tail) + cs.nr_segs > pipe->ring_size) { ret = -EIO; goto out; } @@ -1935,6 +1935,7 @@ static ssize_t fuse_dev_splice_write(struct pipe_inode_info *pipe, struct file *out, loff_t *ppos, size_t len, unsigned int flags) { + unsigned int head, tail, mask, count; unsigned nbuf; unsigned idx; struct pipe_buffer *bufs; @@ -1949,8 +1950,12 @@ static ssize_t fuse_dev_splice_write(struct pipe_inode_info *pipe, pipe_lock(pipe); - bufs = kvmalloc_array(pipe->nrbufs, sizeof(struct pipe_buffer), - GFP_KERNEL); + head = pipe->head; + tail = pipe->tail; + mask = pipe->ring_size - 1; + count = head - tail; + + bufs = kvmalloc_array(count, sizeof(struct pipe_buffer), GFP_KERNEL); if (!bufs) { pipe_unlock(pipe); return -ENOMEM; @@ -1958,8 +1963,8 @@ static ssize_t fuse_dev_splice_write(struct pipe_inode_info *pipe, nbuf = 0; rem = 0; - for (idx = 0; idx < pipe->nrbufs && rem < len; idx++) - rem += pipe->bufs[(pipe->curbuf + idx) & (pipe->buffers - 1)].len; + for (idx = tail; idx < head && rem < len; idx++) + rem += pipe->bufs[idx & mask].len; ret = -EINVAL; if (rem < len) @@ -1970,16 +1975,16 @@ static ssize_t fuse_dev_splice_write(struct pipe_inode_info *pipe, struct pipe_buffer *ibuf; struct pipe_buffer *obuf; - BUG_ON(nbuf >= pipe->buffers); - BUG_ON(!pipe->nrbufs); - ibuf = &pipe->bufs[pipe->curbuf]; + BUG_ON(nbuf >= pipe->ring_size); + BUG_ON(tail == head); + ibuf = &pipe->bufs[tail & mask]; obuf = &bufs[nbuf]; if (rem >= ibuf->len) { *obuf = *ibuf; ibuf->ops = NULL; - pipe->curbuf = (pipe->curbuf + 1) & (pipe->buffers - 1); - pipe->nrbufs--; + tail++; + pipe->tail = tail; } else { if (!pipe_buf_get(pipe, ibuf)) goto out_free; diff --git a/fs/pipe.c b/fs/pipe.c index 8a2ab2f974bd..e9b361cb093e 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -43,10 +43,12 @@ unsigned long pipe_user_pages_hard; unsigned long pipe_user_pages_soft = PIPE_DEF_BUFFERS * INR_OPEN_CUR; /* - * We use a start+len construction, which provides full use of the - * allocated memory. - * -- Florian Coosmann (FGC) - * + * We use head and tail indices that aren't masked off, except at the point of + * dereference, but rather they're allowed to wrap naturally. This means there + * isn't a dead spot in the buffer, but the ring has to be a power of two and + * <= 2^31. + * -- David Howells 2019-09-23. + * * Reads with count = 0 should always return 0. * -- Julian Bradfield 1999-06-07. * @@ -285,10 +287,12 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to) ret = 0; __pipe_lock(pipe); for (;;) { - int bufs = pipe->nrbufs; - if (bufs) { - int curbuf = pipe->curbuf; - struct pipe_buffer *buf = pipe->bufs + curbuf; + unsigned int head = pipe->head; + unsigned int tail = pipe->tail; + unsigned int mask = pipe->ring_size - 1; + + if (!pipe_empty(head, tail)) { + struct pipe_buffer *buf = &pipe->bufs[tail & mask]; size_t chars = buf->len; size_t written; int error; @@ -321,17 +325,17 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to) if (!buf->len) { pipe_buf_release(pipe, buf); - curbuf = (curbuf + 1) & (pipe->buffers - 1); - pipe->curbuf = curbuf; - pipe->nrbufs = --bufs; + tail++; + pipe->tail = tail; do_wakeup = 1; } total_len -= chars; if (!total_len) break; /* common path: read succeeded */ + if (!pipe_empty(head, tail)) /* More to do? */ + continue; } - if (bufs) /* More to do? */ - continue; + if (!pipe->writers) break; if (!pipe->waiting_writers) { @@ -380,6 +384,7 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) { struct file *filp = iocb->ki_filp; struct pipe_inode_info *pipe = filp->private_data; + unsigned int head, tail, max_usage, mask; ssize_t ret = 0; int do_wakeup = 0; size_t total_len = iov_iter_count(from); @@ -397,12 +402,15 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) goto out; } + tail = pipe->tail; + head = pipe->head; + max_usage = pipe->ring_size; + mask = pipe->ring_size - 1; + /* We try to merge small writes */ chars = total_len & (PAGE_SIZE-1); /* size of the last buffer */ - if (pipe->nrbufs && chars != 0) { - int lastbuf = (pipe->curbuf + pipe->nrbufs - 1) & - (pipe->buffers - 1); - struct pipe_buffer *buf = pipe->bufs + lastbuf; + if (!pipe_empty(head, tail) && chars != 0) { + struct pipe_buffer *buf = &pipe->bufs[(head - 1) & mask]; int offset = buf->offset + buf->len; if (pipe_buf_can_merge(buf) && offset + chars <= PAGE_SIZE) { @@ -423,18 +431,16 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) } for (;;) { - int bufs; - if (!pipe->readers) { send_sig(SIGPIPE, current, 0); if (!ret) ret = -EPIPE; break; } - bufs = pipe->nrbufs; - if (bufs < pipe->buffers) { - int newbuf = (pipe->curbuf + bufs) & (pipe->buffers-1); - struct pipe_buffer *buf = pipe->bufs + newbuf; + + tail = pipe->tail; + if (!pipe_full(head, tail, max_usage)) { + struct pipe_buffer *buf = &pipe->bufs[head & mask]; struct page *page = pipe->tmp_page; int copied; @@ -470,14 +476,19 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) buf->ops = &packet_pipe_buf_ops; buf->flags = PIPE_BUF_FLAG_PACKET; } - pipe->nrbufs = ++bufs; + + head++; + pipe->head = head; pipe->tmp_page = NULL; if (!iov_iter_count(from)) break; } - if (bufs < pipe->buffers) + + if (!pipe_full(head, tail, max_usage)) continue; + + /* Wait for buffer space to become available. */ if (filp->f_flags & O_NONBLOCK) { if (!ret) ret = -EAGAIN; @@ -515,17 +526,19 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) static long pipe_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) { struct pipe_inode_info *pipe = filp->private_data; - int count, buf, nrbufs; + int count, head, tail, mask; switch (cmd) { case FIONREAD: __pipe_lock(pipe); count = 0; - buf = pipe->curbuf; - nrbufs = pipe->nrbufs; - while (--nrbufs >= 0) { - count += pipe->bufs[buf].len; - buf = (buf+1) & (pipe->buffers - 1); + head = pipe->head; + tail = pipe->tail; + mask = pipe->ring_size - 1; + + while (tail != head) { + count += pipe->bufs[tail & mask].len; + tail++; } __pipe_unlock(pipe); @@ -541,21 +554,25 @@ pipe_poll(struct file *filp, poll_table *wait) { __poll_t mask; struct pipe_inode_info *pipe = filp->private_data; - int nrbufs; + unsigned int head = READ_ONCE(pipe->head); + unsigned int tail = READ_ONCE(pipe->tail); poll_wait(filp, &pipe->wait, wait); + BUG_ON(pipe_occupancy(head, tail) > pipe->ring_size); + /* Reading only -- no need for acquiring the semaphore. */ - nrbufs = pipe->nrbufs; mask = 0; if (filp->f_mode & FMODE_READ) { - mask = (nrbufs > 0) ? EPOLLIN | EPOLLRDNORM : 0; + if (!pipe_empty(head, tail)) + mask |= EPOLLIN | EPOLLRDNORM; if (!pipe->writers && filp->f_version != pipe->w_counter) mask |= EPOLLHUP; } if (filp->f_mode & FMODE_WRITE) { - mask |= (nrbufs < pipe->buffers) ? EPOLLOUT | EPOLLWRNORM : 0; + if (!pipe_full(head, tail, pipe->ring_size)) + mask |= EPOLLOUT | EPOLLWRNORM; /* * Most Unices do not set EPOLLERR for FIFOs but on Linux they * behave exactly like pipes for poll(). @@ -679,7 +696,7 @@ struct pipe_inode_info *alloc_pipe_info(void) if (pipe->bufs) { init_waitqueue_head(&pipe->wait); pipe->r_counter = pipe->w_counter = 1; - pipe->buffers = pipe_bufs; + pipe->ring_size = pipe_bufs; pipe->user = user; mutex_init(&pipe->mutex); return pipe; @@ -697,9 +714,9 @@ void free_pipe_info(struct pipe_inode_info *pipe) { int i; - (void) account_pipe_buffers(pipe->user, pipe->buffers, 0); + (void) account_pipe_buffers(pipe->user, pipe->ring_size, 0); free_uid(pipe->user); - for (i = 0; i < pipe->buffers; i++) { + for (i = 0; i < pipe->ring_size; i++) { struct pipe_buffer *buf = pipe->bufs + i; if (buf->ops) pipe_buf_release(pipe, buf); @@ -880,7 +897,7 @@ SYSCALL_DEFINE1(pipe, int __user *, fildes) static int wait_for_partner(struct pipe_inode_info *pipe, unsigned int *cnt) { - int cur = *cnt; + int cur = *cnt; while (cur == *cnt) { pipe_wait(pipe); @@ -955,7 +972,7 @@ static int fifo_open(struct inode *inode, struct file *filp) } } break; - + case FMODE_WRITE: /* * O_WRONLY @@ -975,7 +992,7 @@ static int fifo_open(struct inode *inode, struct file *filp) goto err_wr; } break; - + case FMODE_READ | FMODE_WRITE: /* * O_RDWR @@ -1054,14 +1071,14 @@ unsigned int round_pipe_size(unsigned long size) static long pipe_set_size(struct pipe_inode_info *pipe, unsigned long arg) { struct pipe_buffer *bufs; - unsigned int size, nr_pages; + unsigned int size, nr_slots, head, tail, mask, n; unsigned long user_bufs; long ret = 0; size = round_pipe_size(arg); - nr_pages = size >> PAGE_SHIFT; + nr_slots = size >> PAGE_SHIFT; - if (!nr_pages) + if (!nr_slots) return -EINVAL; /* @@ -1071,13 +1088,13 @@ static long pipe_set_size(struct pipe_inode_info *pipe, unsigned long arg) * Decreasing the pipe capacity is always permitted, even * if the user is currently over a limit. */ - if (nr_pages > pipe->buffers && + if (nr_slots > pipe->ring_size && size > pipe_max_size && !capable(CAP_SYS_RESOURCE)) return -EPERM; - user_bufs = account_pipe_buffers(pipe->user, pipe->buffers, nr_pages); + user_bufs = account_pipe_buffers(pipe->user, pipe->ring_size, nr_slots); - if (nr_pages > pipe->buffers && + if (nr_slots > pipe->ring_size && (too_many_pipe_buffers_hard(user_bufs) || too_many_pipe_buffers_soft(user_bufs)) && is_unprivileged_user()) { @@ -1086,17 +1103,21 @@ static long pipe_set_size(struct pipe_inode_info *pipe, unsigned long arg) } /* - * We can shrink the pipe, if arg >= pipe->nrbufs. Since we don't - * expect a lot of shrink+grow operations, just free and allocate - * again like we would do for growing. If the pipe currently + * We can shrink the pipe, if arg is greater than the ring occupancy. + * Since we don't expect a lot of shrink+grow operations, just free and + * allocate again like we would do for growing. If the pipe currently * contains more buffers than arg, then return busy. */ - if (nr_pages < pipe->nrbufs) { + mask = pipe->ring_size - 1; + head = pipe->head; + tail = pipe->tail; + n = pipe_occupancy(pipe->head, pipe->tail); + if (nr_slots < n) { ret = -EBUSY; goto out_revert_acct; } - bufs = kcalloc(nr_pages, sizeof(*bufs), + bufs = kcalloc(nr_slots, sizeof(*bufs), GFP_KERNEL_ACCOUNT | __GFP_NOWARN); if (unlikely(!bufs)) { ret = -ENOMEM; @@ -1105,33 +1126,36 @@ static long pipe_set_size(struct pipe_inode_info *pipe, unsigned long arg) /* * The pipe array wraps around, so just start the new one at zero - * and adjust the indexes. + * and adjust the indices. */ - if (pipe->nrbufs) { - unsigned int tail; - unsigned int head; - - tail = pipe->curbuf + pipe->nrbufs; - if (tail < pipe->buffers) - tail = 0; - else - tail &= (pipe->buffers - 1); - - head = pipe->nrbufs - tail; - if (head) - memcpy(bufs, pipe->bufs + pipe->curbuf, head * sizeof(struct pipe_buffer)); - if (tail) - memcpy(bufs + head, pipe->bufs, tail * sizeof(struct pipe_buffer)); + if (n > 0) { + unsigned int h = head & mask; + unsigned int t = tail & mask; + if (h > t) { + memcpy(bufs, pipe->bufs + t, + n * sizeof(struct pipe_buffer)); + } else { + unsigned int tsize = pipe->ring_size - t; + if (h > 0) + memcpy(bufs + tsize, pipe->bufs, + h * sizeof(struct pipe_buffer)); + memcpy(bufs, pipe->bufs + t, + tsize * sizeof(struct pipe_buffer)); + } } - pipe->curbuf = 0; + head = n; + tail = 0; + kfree(pipe->bufs); pipe->bufs = bufs; - pipe->buffers = nr_pages; - return nr_pages * PAGE_SIZE; + pipe->ring_size = nr_slots; + pipe->tail = tail; + pipe->head = head; + return pipe->ring_size * PAGE_SIZE; out_revert_acct: - (void) account_pipe_buffers(pipe->user, nr_pages, pipe->buffers); + (void) account_pipe_buffers(pipe->user, nr_slots, pipe->ring_size); return ret; } @@ -1161,7 +1185,7 @@ long pipe_fcntl(struct file *file, unsigned int cmd, unsigned long arg) ret = pipe_set_size(pipe, arg); break; case F_GETPIPE_SZ: - ret = pipe->buffers * PAGE_SIZE; + ret = pipe->ring_size * PAGE_SIZE; break; default: ret = -EINVAL; diff --git a/fs/splice.c b/fs/splice.c index 98412721f056..22b0a47a35c0 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -185,6 +185,9 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe, struct splice_pipe_desc *spd) { unsigned int spd_pages = spd->nr_pages; + unsigned int tail = pipe->tail; + unsigned int head = pipe->head; + unsigned int mask = pipe->ring_size - 1; int ret = 0, page_nr = 0; if (!spd_pages) @@ -196,9 +199,8 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe, goto out; } - while (pipe->nrbufs < pipe->buffers) { - int newbuf = (pipe->curbuf + pipe->nrbufs) & (pipe->buffers - 1); - struct pipe_buffer *buf = pipe->bufs + newbuf; + while (!pipe_full(head, tail, pipe->ring_size)) { + struct pipe_buffer *buf = &pipe->bufs[head & mask]; buf->page = spd->pages[page_nr]; buf->offset = spd->partial[page_nr].offset; @@ -207,7 +209,8 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe, buf->ops = spd->ops; buf->flags = 0; - pipe->nrbufs++; + head++; + pipe->head = head; page_nr++; ret += buf->len; @@ -228,17 +231,19 @@ EXPORT_SYMBOL_GPL(splice_to_pipe); ssize_t add_to_pipe(struct pipe_inode_info *pipe, struct pipe_buffer *buf) { + unsigned int head = pipe->head; + unsigned int tail = pipe->tail; + unsigned int mask = pipe->ring_size - 1; int ret; if (unlikely(!pipe->readers)) { send_sig(SIGPIPE, current, 0); ret = -EPIPE; - } else if (pipe->nrbufs == pipe->buffers) { + } else if (pipe_full(head, tail, pipe->ring_size)) { ret = -EAGAIN; } else { - int newbuf = (pipe->curbuf + pipe->nrbufs) & (pipe->buffers - 1); - pipe->bufs[newbuf] = *buf; - pipe->nrbufs++; + pipe->bufs[head & mask] = *buf; + pipe->head = head + 1; return buf->len; } pipe_buf_release(pipe, buf); @@ -252,14 +257,14 @@ EXPORT_SYMBOL(add_to_pipe); */ int splice_grow_spd(const struct pipe_inode_info *pipe, struct splice_pipe_desc *spd) { - unsigned int buffers = READ_ONCE(pipe->buffers); + unsigned int max_usage = READ_ONCE(pipe->ring_size); - spd->nr_pages_max = buffers; - if (buffers <= PIPE_DEF_BUFFERS) + spd->nr_pages_max = max_usage; + if (max_usage <= PIPE_DEF_BUFFERS) return 0; - spd->pages = kmalloc_array(buffers, sizeof(struct page *), GFP_KERNEL); - spd->partial = kmalloc_array(buffers, sizeof(struct partial_page), + spd->pages = kmalloc_array(max_usage, sizeof(struct page *), GFP_KERNEL); + spd->partial = kmalloc_array(max_usage, sizeof(struct partial_page), GFP_KERNEL); if (spd->pages && spd->partial) @@ -298,10 +303,11 @@ ssize_t generic_file_splice_read(struct file *in, loff_t *ppos, { struct iov_iter to; struct kiocb kiocb; - int idx, ret; + unsigned int i_head; + int ret; iov_iter_pipe(&to, READ, pipe, len); - idx = to.idx; + i_head = to.head; init_sync_kiocb(&kiocb, in); kiocb.ki_pos = *ppos; ret = call_read_iter(in, &kiocb, &to); @@ -309,7 +315,7 @@ ssize_t generic_file_splice_read(struct file *in, loff_t *ppos, *ppos = kiocb.ki_pos; file_accessed(in); } else if (ret < 0) { - to.idx = idx; + to.head = i_head; to.iov_offset = 0; iov_iter_advance(&to, 0); /* to free what was emitted */ /* @@ -370,11 +376,12 @@ static ssize_t default_file_splice_read(struct file *in, loff_t *ppos, struct iov_iter to; struct page **pages; unsigned int nr_pages; + unsigned int mask; size_t offset, base, copied = 0; ssize_t res; int i; - if (pipe->nrbufs == pipe->buffers) + if (pipe_full(pipe->head, pipe->tail, pipe->ring_size)) return -EAGAIN; /* @@ -400,8 +407,9 @@ static ssize_t default_file_splice_read(struct file *in, loff_t *ppos, } } - pipe->bufs[to.idx].offset = offset; - pipe->bufs[to.idx].len -= offset; + mask = pipe->ring_size - 1; + pipe->bufs[to.head & mask].offset = offset; + pipe->bufs[to.head & mask].len -= offset; for (i = 0; i < nr_pages; i++) { size_t this_len = min_t(size_t, len, PAGE_SIZE - offset); @@ -443,7 +451,8 @@ static int pipe_to_sendpage(struct pipe_inode_info *pipe, more = (sd->flags & SPLICE_F_MORE) ? MSG_MORE : 0; - if (sd->len < sd->total_len && pipe->nrbufs > 1) + if (sd->len < sd->total_len && + pipe_occupancy(pipe->head, pipe->tail) > 1) more |= MSG_SENDPAGE_NOTLAST; return file->f_op->sendpage(file, buf->page, buf->offset, @@ -481,10 +490,13 @@ static void wakeup_pipe_writers(struct pipe_inode_info *pipe) static int splice_from_pipe_feed(struct pipe_inode_info *pipe, struct splice_desc *sd, splice_actor *actor) { + unsigned int head = pipe->head; + unsigned int tail = pipe->tail; + unsigned int mask = pipe->ring_size - 1; int ret; - while (pipe->nrbufs) { - struct pipe_buffer *buf = pipe->bufs + pipe->curbuf; + while (!pipe_empty(tail, head)) { + struct pipe_buffer *buf = &pipe->bufs[tail & mask]; sd->len = buf->len; if (sd->len > sd->total_len) @@ -511,8 +523,8 @@ static int splice_from_pipe_feed(struct pipe_inode_info *pipe, struct splice_des if (!buf->len) { pipe_buf_release(pipe, buf); - pipe->curbuf = (pipe->curbuf + 1) & (pipe->buffers - 1); - pipe->nrbufs--; + tail++; + pipe->tail = tail; if (pipe->files) sd->need_wakeup = true; } @@ -543,7 +555,7 @@ static int splice_from_pipe_next(struct pipe_inode_info *pipe, struct splice_des if (signal_pending(current)) return -ERESTARTSYS; - while (!pipe->nrbufs) { + while (pipe_empty(pipe->head, pipe->tail)) { if (!pipe->writers) return 0; @@ -686,7 +698,7 @@ iter_file_splice_write(struct pipe_inode_info *pipe, struct file *out, .pos = *ppos, .u.file = out, }; - int nbufs = pipe->buffers; + int nbufs = pipe->ring_size; struct bio_vec *array = kcalloc(nbufs, sizeof(struct bio_vec), GFP_KERNEL); ssize_t ret; @@ -699,16 +711,19 @@ iter_file_splice_write(struct pipe_inode_info *pipe, struct file *out, splice_from_pipe_begin(&sd); while (sd.total_len) { struct iov_iter from; + unsigned int head = pipe->head; + unsigned int tail = pipe->tail; + unsigned int mask = pipe->ring_size - 1; size_t left; - int n, idx; + int n; ret = splice_from_pipe_next(pipe, &sd); if (ret <= 0) break; - if (unlikely(nbufs < pipe->buffers)) { + if (unlikely(nbufs < pipe->ring_size)) { kfree(array); - nbufs = pipe->buffers; + nbufs = pipe->ring_size; array = kcalloc(nbufs, sizeof(struct bio_vec), GFP_KERNEL); if (!array) { @@ -719,16 +734,13 @@ iter_file_splice_write(struct pipe_inode_info *pipe, struct file *out, /* build the vector */ left = sd.total_len; - for (n = 0, idx = pipe->curbuf; left && n < pipe->nrbufs; n++, idx++) { - struct pipe_buffer *buf = pipe->bufs + idx; + for (n = 0; !pipe_empty(head, tail) && left && n < nbufs; tail++, n++) { + struct pipe_buffer *buf = &pipe->bufs[tail & mask]; size_t this_len = buf->len; if (this_len > left) this_len = left; - if (idx == pipe->buffers - 1) - idx = -1; - ret = pipe_buf_confirm(pipe, buf); if (unlikely(ret)) { if (ret == -ENODATA) @@ -752,14 +764,15 @@ iter_file_splice_write(struct pipe_inode_info *pipe, struct file *out, *ppos = sd.pos; /* dismiss the fully eaten buffers, adjust the partial one */ + tail = pipe->tail; while (ret) { - struct pipe_buffer *buf = pipe->bufs + pipe->curbuf; + struct pipe_buffer *buf = &pipe->bufs[tail & mask]; if (ret >= buf->len) { ret -= buf->len; buf->len = 0; pipe_buf_release(pipe, buf); - pipe->curbuf = (pipe->curbuf + 1) & (pipe->buffers - 1); - pipe->nrbufs--; + tail++; + pipe->tail = tail; if (pipe->files) sd.need_wakeup = true; } else { @@ -942,15 +955,17 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, sd->flags &= ~SPLICE_F_NONBLOCK; more = sd->flags & SPLICE_F_MORE; - WARN_ON_ONCE(pipe->nrbufs != 0); + WARN_ON_ONCE(!pipe_empty(pipe->head, pipe->tail)); while (len) { + unsigned int p_space; size_t read_len; loff_t pos = sd->pos, prev_pos = pos; /* Don't try to read more the pipe has space for. */ - read_len = min_t(size_t, len, - (pipe->buffers - pipe->nrbufs) << PAGE_SHIFT); + p_space = pipe->ring_size - + pipe_occupancy(pipe->head, pipe->tail); + read_len = min_t(size_t, len, p_space << PAGE_SHIFT); ret = do_splice_to(in, &pos, pipe, read_len, flags); if (unlikely(ret <= 0)) goto out_release; @@ -989,7 +1004,7 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, } done: - pipe->nrbufs = pipe->curbuf = 0; + pipe->tail = pipe->head = 0; file_accessed(in); return bytes; @@ -998,8 +1013,8 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, * If we did an incomplete transfer we must release * the pipe buffers in question: */ - for (i = 0; i < pipe->buffers; i++) { - struct pipe_buffer *buf = pipe->bufs + i; + for (i = 0; i < pipe->ring_size; i++) { + struct pipe_buffer *buf = &pipe->bufs[i]; if (buf->ops) pipe_buf_release(pipe, buf); @@ -1075,7 +1090,7 @@ static int wait_for_space(struct pipe_inode_info *pipe, unsigned flags) send_sig(SIGPIPE, current, 0); return -EPIPE; } - if (pipe->nrbufs != pipe->buffers) + if (!pipe_full(pipe->head, pipe->tail, pipe->ring_size)) return 0; if (flags & SPLICE_F_NONBLOCK) return -EAGAIN; @@ -1442,16 +1457,16 @@ static int ipipe_prep(struct pipe_inode_info *pipe, unsigned int flags) int ret; /* - * Check ->nrbufs without the inode lock first. This function + * Check the pipe occupancy without the inode lock first. This function * is speculative anyways, so missing one is ok. */ - if (pipe->nrbufs) + if (!pipe_empty(pipe->head, pipe->tail)) return 0; ret = 0; pipe_lock(pipe); - while (!pipe->nrbufs) { + while (pipe_empty(pipe->head, pipe->tail)) { if (signal_pending(current)) { ret = -ERESTARTSYS; break; @@ -1480,16 +1495,16 @@ static int opipe_prep(struct pipe_inode_info *pipe, unsigned int flags) int ret; /* - * Check ->nrbufs without the inode lock first. This function + * Check pipe occupancy without the inode lock first. This function * is speculative anyways, so missing one is ok. */ - if (pipe->nrbufs < pipe->buffers) + if (pipe_full(pipe->head, pipe->tail, pipe->ring_size)) return 0; ret = 0; pipe_lock(pipe); - while (pipe->nrbufs >= pipe->buffers) { + while (pipe_full(pipe->head, pipe->tail, pipe->ring_size)) { if (!pipe->readers) { send_sig(SIGPIPE, current, 0); ret = -EPIPE; @@ -1520,7 +1535,10 @@ static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, size_t len, unsigned int flags) { struct pipe_buffer *ibuf, *obuf; - int ret = 0, nbuf; + unsigned int i_head, o_head; + unsigned int i_tail, o_tail; + unsigned int i_mask, o_mask; + int ret = 0; bool input_wakeup = false; @@ -1540,7 +1558,14 @@ static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, */ pipe_double_lock(ipipe, opipe); + i_tail = ipipe->tail; + i_mask = ipipe->ring_size - 1; + o_head = opipe->head; + o_mask = opipe->ring_size - 1; + do { + size_t o_len; + if (!opipe->readers) { send_sig(SIGPIPE, current, 0); if (!ret) @@ -1548,14 +1573,18 @@ static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, break; } - if (!ipipe->nrbufs && !ipipe->writers) + i_head = ipipe->head; + o_tail = opipe->tail; + + if (pipe_empty(i_head, i_tail) && !ipipe->writers) break; /* * Cannot make any progress, because either the input * pipe is empty or the output pipe is full. */ - if (!ipipe->nrbufs || opipe->nrbufs >= opipe->buffers) { + if (pipe_empty(i_head, i_tail) || + pipe_full(o_head, o_tail, opipe->ring_size)) { /* Already processed some buffers, break */ if (ret) break; @@ -1575,9 +1604,8 @@ static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, goto retry; } - ibuf = ipipe->bufs + ipipe->curbuf; - nbuf = (opipe->curbuf + opipe->nrbufs) & (opipe->buffers - 1); - obuf = opipe->bufs + nbuf; + ibuf = &ipipe->bufs[i_tail & i_mask]; + obuf = &opipe->bufs[o_head & o_mask]; if (len >= ibuf->len) { /* @@ -1585,10 +1613,12 @@ static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, */ *obuf = *ibuf; ibuf->ops = NULL; - opipe->nrbufs++; - ipipe->curbuf = (ipipe->curbuf + 1) & (ipipe->buffers - 1); - ipipe->nrbufs--; + i_tail++; + ipipe->tail = i_tail; input_wakeup = true; + o_len = obuf->len; + o_head++; + opipe->head = o_head; } else { /* * Get a reference to this pipe buffer, @@ -1610,12 +1640,14 @@ static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, pipe_buf_mark_unmergeable(obuf); obuf->len = len; - opipe->nrbufs++; - ibuf->offset += obuf->len; - ibuf->len -= obuf->len; + ibuf->offset += len; + ibuf->len -= len; + o_len = len; + o_head++; + opipe->head = o_head; } - ret += obuf->len; - len -= obuf->len; + ret += o_len; + len -= o_len; } while (len); pipe_unlock(ipipe); @@ -1641,7 +1673,10 @@ static int link_pipe(struct pipe_inode_info *ipipe, size_t len, unsigned int flags) { struct pipe_buffer *ibuf, *obuf; - int ret = 0, i = 0, nbuf; + unsigned int i_head, o_head; + unsigned int i_tail, o_tail; + unsigned int i_mask, o_mask; + int ret = 0; /* * Potential ABBA deadlock, work around it by ordering lock @@ -1650,6 +1685,11 @@ static int link_pipe(struct pipe_inode_info *ipipe, */ pipe_double_lock(ipipe, opipe); + i_tail = ipipe->tail; + i_mask = ipipe->ring_size - 1; + o_head = opipe->head; + o_mask = opipe->ring_size - 1; + do { if (!opipe->readers) { send_sig(SIGPIPE, current, 0); @@ -1658,15 +1698,19 @@ static int link_pipe(struct pipe_inode_info *ipipe, break; } + i_head = ipipe->head; + o_tail = opipe->tail; + /* - * If we have iterated all input buffers or ran out of + * If we have iterated all input buffers or run out of * output room, break. */ - if (i >= ipipe->nrbufs || opipe->nrbufs >= opipe->buffers) + if (pipe_empty(i_head, i_tail) || + pipe_full(o_head, o_tail, opipe->ring_size)) break; - ibuf = ipipe->bufs + ((ipipe->curbuf + i) & (ipipe->buffers-1)); - nbuf = (opipe->curbuf + opipe->nrbufs) & (opipe->buffers - 1); + ibuf = &ipipe->bufs[i_tail & i_mask]; + obuf = &opipe->bufs[o_head & o_mask]; /* * Get a reference to this pipe buffer, @@ -1678,7 +1722,6 @@ static int link_pipe(struct pipe_inode_info *ipipe, break; } - obuf = opipe->bufs + nbuf; *obuf = *ibuf; /* @@ -1691,11 +1734,12 @@ static int link_pipe(struct pipe_inode_info *ipipe, if (obuf->len > len) obuf->len = len; - - opipe->nrbufs++; ret += obuf->len; len -= obuf->len; - i++; + + o_head++; + opipe->head = o_head; + i_tail++; } while (len); /* diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index 5c626fdc10db..96158ca80456 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -30,9 +30,9 @@ struct pipe_buffer { * struct pipe_inode_info - a linux kernel pipe * @mutex: mutex protecting the whole thing * @wait: reader/writer wait point in case of empty/full pipe - * @nrbufs: the number of non-empty pipe buffers in this pipe - * @buffers: total number of buffers (should be a power of 2) - * @curbuf: the current pipe buffer entry + * @head: The point of buffer production + * @tail: The point of buffer consumption + * @ring_size: total number of buffers (should be a power of 2) * @tmp_page: cached released page * @readers: number of current readers of this pipe * @writers: number of current writers of this pipe @@ -48,7 +48,9 @@ struct pipe_buffer { struct pipe_inode_info { struct mutex mutex; wait_queue_head_t wait; - unsigned int nrbufs, curbuf, buffers; + unsigned int head; + unsigned int tail; + unsigned int ring_size; unsigned int readers; unsigned int writers; unsigned int files; @@ -104,6 +106,56 @@ struct pipe_buf_operations { bool (*get)(struct pipe_inode_info *, struct pipe_buffer *); }; +/** + * pipe_empty - Return true if the pipe is empty + * @head: The pipe ring head pointer + * @tail: The pipe ring tail pointer + */ +static inline bool pipe_empty(unsigned int head, unsigned int tail) +{ + return head == tail; +} + +/** + * pipe_occupancy - Return number of slots used in the pipe + * @head: The pipe ring head pointer + * @tail: The pipe ring tail pointer + */ +static inline unsigned int pipe_occupancy(unsigned int head, unsigned int tail) +{ + return head - tail; +} + +/** + * pipe_full - Return true if the pipe is full + * @head: The pipe ring head pointer + * @tail: The pipe ring tail pointer + * @limit: The maximum amount of slots available. + */ +static inline bool pipe_full(unsigned int head, unsigned int tail, + unsigned int limit) +{ + return pipe_occupancy(head, tail) >= limit; +} + +/** + * pipe_space_for_user - Return number of slots available to userspace + * @head: The pipe ring head pointer + * @tail: The pipe ring tail pointer + * @pipe: The pipe info structure + */ +static inline unsigned int pipe_space_for_user(unsigned int head, unsigned int tail, + struct pipe_inode_info *pipe) +{ + unsigned int p_occupancy, p_space; + + p_occupancy = pipe_occupancy(head, tail); + if (p_occupancy >= pipe->ring_size) + return 0; + p_space = pipe->ring_size - p_occupancy; + return p_space; +} + /** * pipe_buf_get - get a reference to a pipe_buffer * @pipe: the pipe that the buffer belongs to diff --git a/include/linux/uio.h b/include/linux/uio.h index ab5f523bc0df..9576fd8158d7 100644 --- a/include/linux/uio.h +++ b/include/linux/uio.h @@ -45,8 +45,8 @@ struct iov_iter { union { unsigned long nr_segs; struct { - int idx; - int start_idx; + unsigned int head; + unsigned int start_head; }; }; }; diff --git a/lib/iov_iter.c b/lib/iov_iter.c index 639d5e7014c1..6607e9c875ce 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -325,28 +325,33 @@ static size_t copy_page_from_iter_iovec(struct page *page, size_t offset, size_t static bool sanity(const struct iov_iter *i) { struct pipe_inode_info *pipe = i->pipe; - int idx = i->idx; - int next = pipe->curbuf + pipe->nrbufs; + unsigned int p_head = pipe->head; + unsigned int p_tail = pipe->tail; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int p_occupancy = pipe_occupancy(p_head, p_tail); + unsigned int i_head = i->head; + unsigned int idx; + if (i->iov_offset) { struct pipe_buffer *p; - if (unlikely(!pipe->nrbufs)) + if (unlikely(p_occupancy == 0)) goto Bad; // pipe must be non-empty - if (unlikely(idx != ((next - 1) & (pipe->buffers - 1)))) + if (unlikely(i_head != p_head - 1)) goto Bad; // must be at the last buffer... - p = &pipe->bufs[idx]; + p = &pipe->bufs[i_head & p_mask]; if (unlikely(p->offset + p->len != i->iov_offset)) goto Bad; // ... at the end of segment } else { - if (idx != (next & (pipe->buffers - 1))) + if (i_head != p_head) goto Bad; // must be right after the last buffer } return true; Bad: - printk(KERN_ERR "idx = %d, offset = %zd\n", i->idx, i->iov_offset); - printk(KERN_ERR "curbuf = %d, nrbufs = %d, buffers = %d\n", - pipe->curbuf, pipe->nrbufs, pipe->buffers); - for (idx = 0; idx < pipe->buffers; idx++) + printk(KERN_ERR "idx = %d, offset = %zd\n", i_head, i->iov_offset); + printk(KERN_ERR "head = %d, tail = %d, buffers = %d\n", + p_head, p_tail, pipe->ring_size); + for (idx = 0; idx < pipe->ring_size; idx++) printk(KERN_ERR "[%p %p %d %d]\n", pipe->bufs[idx].ops, pipe->bufs[idx].page, @@ -359,18 +364,15 @@ static bool sanity(const struct iov_iter *i) #define sanity(i) true #endif -static inline int next_idx(int idx, struct pipe_inode_info *pipe) -{ - return (idx + 1) & (pipe->buffers - 1); -} - static size_t copy_page_to_iter_pipe(struct page *page, size_t offset, size_t bytes, struct iov_iter *i) { struct pipe_inode_info *pipe = i->pipe; struct pipe_buffer *buf; + unsigned int p_tail = pipe->tail; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int i_head = i->head; size_t off; - int idx; if (unlikely(bytes > i->count)) bytes = i->count; @@ -382,8 +384,7 @@ static size_t copy_page_to_iter_pipe(struct page *page, size_t offset, size_t by return 0; off = i->iov_offset; - idx = i->idx; - buf = &pipe->bufs[idx]; + buf = &pipe->bufs[i_head & p_mask]; if (off) { if (offset == off && buf->page == page) { /* merge with the last one */ @@ -391,18 +392,21 @@ static size_t copy_page_to_iter_pipe(struct page *page, size_t offset, size_t by i->iov_offset += bytes; goto out; } - idx = next_idx(idx, pipe); - buf = &pipe->bufs[idx]; + i_head++; + buf = &pipe->bufs[i_head & p_mask]; } - if (idx == pipe->curbuf && pipe->nrbufs) + if (pipe_full(i_head, p_tail, pipe->ring_size)) return 0; - pipe->nrbufs++; + buf->ops = &page_cache_pipe_buf_ops; - get_page(buf->page = page); + get_page(page); + buf->page = page; buf->offset = offset; buf->len = bytes; + + pipe->head = i_head; i->iov_offset = offset + bytes; - i->idx = idx; + i->head = i_head; out: i->count -= bytes; return bytes; @@ -480,24 +484,30 @@ static inline bool allocated(struct pipe_buffer *buf) return buf->ops == &default_pipe_buf_ops; } -static inline void data_start(const struct iov_iter *i, int *idxp, size_t *offp) +static inline void data_start(const struct iov_iter *i, + unsigned int *iter_headp, size_t *offp) { + unsigned int p_mask = i->pipe->ring_size - 1; + unsigned int iter_head = i->head; size_t off = i->iov_offset; - int idx = i->idx; - if (off && (!allocated(&i->pipe->bufs[idx]) || off == PAGE_SIZE)) { - idx = next_idx(idx, i->pipe); + + if (off && (!allocated(&i->pipe->bufs[iter_head & p_mask]) || + off == PAGE_SIZE)) { + iter_head++; off = 0; } - *idxp = idx; + *iter_headp = iter_head; *offp = off; } static size_t push_pipe(struct iov_iter *i, size_t size, - int *idxp, size_t *offp) + int *iter_headp, size_t *offp) { struct pipe_inode_info *pipe = i->pipe; + unsigned int p_tail = pipe->tail; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int iter_head; size_t off; - int idx; ssize_t left; if (unlikely(size > i->count)) @@ -506,33 +516,34 @@ static size_t push_pipe(struct iov_iter *i, size_t size, return 0; left = size; - data_start(i, &idx, &off); - *idxp = idx; + data_start(i, &iter_head, &off); + *iter_headp = iter_head; *offp = off; if (off) { left -= PAGE_SIZE - off; if (left <= 0) { - pipe->bufs[idx].len += size; + pipe->bufs[iter_head & p_mask].len += size; return size; } - pipe->bufs[idx].len = PAGE_SIZE; - idx = next_idx(idx, pipe); + pipe->bufs[iter_head & p_mask].len = PAGE_SIZE; + iter_head++; } - while (idx != pipe->curbuf || !pipe->nrbufs) { + while (!pipe_full(iter_head, p_tail, pipe->ring_size)) { + struct pipe_buffer *buf = &pipe->bufs[iter_head & p_mask]; struct page *page = alloc_page(GFP_USER); if (!page) break; - pipe->nrbufs++; - pipe->bufs[idx].ops = &default_pipe_buf_ops; - pipe->bufs[idx].page = page; - pipe->bufs[idx].offset = 0; - if (left <= PAGE_SIZE) { - pipe->bufs[idx].len = left; + + buf->ops = &default_pipe_buf_ops; + buf->page = page; + buf->offset = 0; + buf->len = max_t(ssize_t, left, PAGE_SIZE); + left -= buf->len; + iter_head++; + pipe->head = iter_head; + + if (left == 0) return size; - } - pipe->bufs[idx].len = PAGE_SIZE; - left -= PAGE_SIZE; - idx = next_idx(idx, pipe); } return size - left; } @@ -541,23 +552,26 @@ static size_t copy_pipe_to_iter(const void *addr, size_t bytes, struct iov_iter *i) { struct pipe_inode_info *pipe = i->pipe; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int i_head; size_t n, off; - int idx; if (!sanity(i)) return 0; - bytes = n = push_pipe(i, bytes, &idx, &off); + bytes = n = push_pipe(i, bytes, &i_head, &off); if (unlikely(!n)) return 0; - for ( ; n; idx = next_idx(idx, pipe), off = 0) { + do { size_t chunk = min_t(size_t, n, PAGE_SIZE - off); - memcpy_to_page(pipe->bufs[idx].page, off, addr, chunk); - i->idx = idx; + memcpy_to_page(pipe->bufs[i_head & p_mask].page, off, addr, chunk); + i->head = i_head; i->iov_offset = off + chunk; n -= chunk; addr += chunk; - } + off = 0; + i_head++; + } while (n); i->count -= bytes; return bytes; } @@ -573,28 +587,31 @@ static size_t csum_and_copy_to_pipe_iter(const void *addr, size_t bytes, __wsum *csum, struct iov_iter *i) { struct pipe_inode_info *pipe = i->pipe; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int i_head; size_t n, r; size_t off = 0; __wsum sum = *csum; - int idx; if (!sanity(i)) return 0; - bytes = n = push_pipe(i, bytes, &idx, &r); + bytes = n = push_pipe(i, bytes, &i_head, &r); if (unlikely(!n)) return 0; - for ( ; n; idx = next_idx(idx, pipe), r = 0) { + do { size_t chunk = min_t(size_t, n, PAGE_SIZE - r); - char *p = kmap_atomic(pipe->bufs[idx].page); + char *p = kmap_atomic(pipe->bufs[i_head & p_mask].page); sum = csum_and_memcpy(p + r, addr, chunk, sum, off); kunmap_atomic(p); - i->idx = idx; + i->head = i_head; i->iov_offset = r + chunk; n -= chunk; off += chunk; addr += chunk; - } + r = 0; + i_head++; + } while (n); i->count -= bytes; *csum = sum; return bytes; @@ -645,29 +662,32 @@ static size_t copy_pipe_to_iter_mcsafe(const void *addr, size_t bytes, struct iov_iter *i) { struct pipe_inode_info *pipe = i->pipe; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int i_head; size_t n, off, xfer = 0; - int idx; if (!sanity(i)) return 0; - bytes = n = push_pipe(i, bytes, &idx, &off); + bytes = n = push_pipe(i, bytes, &i_head, &off); if (unlikely(!n)) return 0; - for ( ; n; idx = next_idx(idx, pipe), off = 0) { + do { size_t chunk = min_t(size_t, n, PAGE_SIZE - off); unsigned long rem; - rem = memcpy_mcsafe_to_page(pipe->bufs[idx].page, off, addr, - chunk); - i->idx = idx; + rem = memcpy_mcsafe_to_page(pipe->bufs[i_head & p_mask].page, + off, addr, chunk); + i->head = i_head; i->iov_offset = off + chunk - rem; xfer += chunk - rem; if (rem) break; n -= chunk; addr += chunk; - } + off = 0; + i_head++; + } while (n); i->count -= xfer; return xfer; } @@ -925,23 +945,26 @@ EXPORT_SYMBOL(copy_page_from_iter); static size_t pipe_zero(size_t bytes, struct iov_iter *i) { struct pipe_inode_info *pipe = i->pipe; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int i_head; size_t n, off; - int idx; if (!sanity(i)) return 0; - bytes = n = push_pipe(i, bytes, &idx, &off); + bytes = n = push_pipe(i, bytes, &i_head, &off); if (unlikely(!n)) return 0; - for ( ; n; idx = next_idx(idx, pipe), off = 0) { + do { size_t chunk = min_t(size_t, n, PAGE_SIZE - off); - memzero_page(pipe->bufs[idx].page, off, chunk); - i->idx = idx; + memzero_page(pipe->bufs[i_head & p_mask].page, off, chunk); + i->head = i_head; i->iov_offset = off + chunk; n -= chunk; - } + off = 0; + i_head++; + } while (n); i->count -= bytes; return bytes; } @@ -987,20 +1010,26 @@ EXPORT_SYMBOL(iov_iter_copy_from_user_atomic); static inline void pipe_truncate(struct iov_iter *i) { struct pipe_inode_info *pipe = i->pipe; - if (pipe->nrbufs) { + unsigned int p_tail = pipe->tail; + unsigned int p_head = pipe->head; + unsigned int p_mask = pipe->ring_size - 1; + + if (!pipe_empty(p_head, p_tail)) { + struct pipe_buffer *buf; + unsigned int i_head = i->head; size_t off = i->iov_offset; - int idx = i->idx; - int nrbufs = (idx - pipe->curbuf) & (pipe->buffers - 1); + if (off) { - pipe->bufs[idx].len = off - pipe->bufs[idx].offset; - idx = next_idx(idx, pipe); - nrbufs++; + buf = &pipe->bufs[i_head & p_mask]; + buf->len = off - buf->offset; + i_head++; } - while (pipe->nrbufs > nrbufs) { - pipe_buf_release(pipe, &pipe->bufs[idx]); - idx = next_idx(idx, pipe); - pipe->nrbufs--; + while (p_head != i_head) { + p_head--; + pipe_buf_release(pipe, &pipe->bufs[p_head & p_mask]); } + + pipe->head = p_head; } } @@ -1011,18 +1040,20 @@ static void pipe_advance(struct iov_iter *i, size_t size) size = i->count; if (size) { struct pipe_buffer *buf; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int i_head = i->head; size_t off = i->iov_offset, left = size; - int idx = i->idx; + if (off) /* make it relative to the beginning of buffer */ - left += off - pipe->bufs[idx].offset; + left += off - pipe->bufs[i_head & p_mask].offset; while (1) { - buf = &pipe->bufs[idx]; + buf = &pipe->bufs[i_head & p_mask]; if (left <= buf->len) break; left -= buf->len; - idx = next_idx(idx, pipe); + i_head++; } - i->idx = idx; + i->head = i_head; i->iov_offset = buf->offset + left; } i->count -= size; @@ -1053,25 +1084,27 @@ void iov_iter_revert(struct iov_iter *i, size_t unroll) i->count += unroll; if (unlikely(iov_iter_is_pipe(i))) { struct pipe_inode_info *pipe = i->pipe; - int idx = i->idx; + unsigned int p_mask = pipe->ring_size - 1; + unsigned int i_head = i->head; size_t off = i->iov_offset; while (1) { - size_t n = off - pipe->bufs[idx].offset; + struct pipe_buffer *b = &pipe->bufs[i_head & p_mask]; + size_t n = off - b->offset; if (unroll < n) { off -= unroll; break; } unroll -= n; - if (!unroll && idx == i->start_idx) { + if (!unroll && i_head == i->start_head) { off = 0; break; } - if (!idx--) - idx = pipe->buffers - 1; - off = pipe->bufs[idx].offset + pipe->bufs[idx].len; + i_head--; + b = &pipe->bufs[i_head & p_mask]; + off = b->offset + b->len; } i->iov_offset = off; - i->idx = idx; + i->head = i_head; pipe_truncate(i); return; } @@ -1159,13 +1192,13 @@ void iov_iter_pipe(struct iov_iter *i, unsigned int direction, size_t count) { BUG_ON(direction != READ); - WARN_ON(pipe->nrbufs == pipe->buffers); + WARN_ON(pipe_full(pipe->head, pipe->tail, pipe->ring_size)); i->type = ITER_PIPE | READ; i->pipe = pipe; - i->idx = (pipe->curbuf + pipe->nrbufs) & (pipe->buffers - 1); + i->head = pipe->head; i->iov_offset = 0; i->count = count; - i->start_idx = i->idx; + i->start_head = i->head; } EXPORT_SYMBOL(iov_iter_pipe); @@ -1189,11 +1222,12 @@ EXPORT_SYMBOL(iov_iter_discard); unsigned long iov_iter_alignment(const struct iov_iter *i) { + unsigned int p_mask = i->pipe->ring_size - 1; unsigned long res = 0; size_t size = i->count; if (unlikely(iov_iter_is_pipe(i))) { - if (size && i->iov_offset && allocated(&i->pipe->bufs[i->idx])) + if (size && i->iov_offset && allocated(&i->pipe->bufs[i->head & p_mask])) return size | i->iov_offset; return size; } @@ -1231,19 +1265,20 @@ EXPORT_SYMBOL(iov_iter_gap_alignment); static inline ssize_t __pipe_get_pages(struct iov_iter *i, size_t maxsize, struct page **pages, - int idx, + int iter_head, size_t *start) { struct pipe_inode_info *pipe = i->pipe; - ssize_t n = push_pipe(i, maxsize, &idx, start); + unsigned int p_mask = pipe->ring_size - 1; + ssize_t n = push_pipe(i, maxsize, &iter_head, start); if (!n) return -EFAULT; maxsize = n; n += *start; while (n > 0) { - get_page(*pages++ = pipe->bufs[idx].page); - idx = next_idx(idx, pipe); + get_page(*pages++ = pipe->bufs[iter_head & p_mask].page); + iter_head++; n -= PAGE_SIZE; } @@ -1254,9 +1289,8 @@ static ssize_t pipe_get_pages(struct iov_iter *i, struct page **pages, size_t maxsize, unsigned maxpages, size_t *start) { - unsigned npages; + unsigned int iter_head, npages; size_t capacity; - int idx; if (!maxsize) return 0; @@ -1264,12 +1298,12 @@ static ssize_t pipe_get_pages(struct iov_iter *i, if (!sanity(i)) return -EFAULT; - data_start(i, &idx, start); - /* some of this one + all after this one */ - npages = ((i->pipe->curbuf - idx - 1) & (i->pipe->buffers - 1)) + 1; - capacity = min(npages,maxpages) * PAGE_SIZE - *start; + data_start(i, &iter_head, start); + /* Amount of free space: some of this one + all after this one */ + npages = pipe_space_for_user(iter_head, i->pipe->tail, i->pipe); + capacity = min(npages, maxpages) * PAGE_SIZE - *start; - return __pipe_get_pages(i, min(maxsize, capacity), pages, idx, start); + return __pipe_get_pages(i, min(maxsize, capacity), pages, iter_head, start); } ssize_t iov_iter_get_pages(struct iov_iter *i, @@ -1323,9 +1357,8 @@ static ssize_t pipe_get_pages_alloc(struct iov_iter *i, size_t *start) { struct page **p; + unsigned int iter_head, npages; ssize_t n; - int idx; - int npages; if (!maxsize) return 0; @@ -1333,9 +1366,9 @@ static ssize_t pipe_get_pages_alloc(struct iov_iter *i, if (!sanity(i)) return -EFAULT; - data_start(i, &idx, start); - /* some of this one + all after this one */ - npages = ((i->pipe->curbuf - idx - 1) & (i->pipe->buffers - 1)) + 1; + data_start(i, &iter_head, start); + /* Amount of free space: some of this one + all after this one */ + npages = pipe_space_for_user(iter_head, i->pipe->tail, i->pipe); n = npages * PAGE_SIZE - *start; if (maxsize > n) maxsize = n; @@ -1344,7 +1377,7 @@ static ssize_t pipe_get_pages_alloc(struct iov_iter *i, p = get_pages_array(npages); if (!p) return -ENOMEM; - n = __pipe_get_pages(i, maxsize, p, idx, start); + n = __pipe_get_pages(i, maxsize, p, iter_head, start); if (n > 0) *pages = p; else @@ -1560,15 +1593,15 @@ int iov_iter_npages(const struct iov_iter *i, int maxpages) if (unlikely(iov_iter_is_pipe(i))) { struct pipe_inode_info *pipe = i->pipe; + unsigned int iter_head; size_t off; - int idx; if (!sanity(i)) return 0; - data_start(i, &idx, &off); + data_start(i, &iter_head, &off); /* some of this one + all after this one */ - npages = ((pipe->curbuf - idx - 1) & (pipe->buffers - 1)) + 1; + npages = pipe_space_for_user(iter_head, pipe->tail, pipe); if (npages >= maxpages) return maxpages; } else iterate_all_kinds(i, size, v, ({ From patchwork Fri Nov 1 17:34:46 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223513 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id AB7851515 for ; Fri, 1 Nov 2019 17:34:59 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 7F529217F9 for ; Fri, 1 Nov 2019 17:34:59 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="BGS7yE9z" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729711AbfKARe6 (ORCPT ); Fri, 1 Nov 2019 13:34:58 -0400 Received: from us-smtp-1.mimecast.com ([207.211.31.81]:44211 "EHLO us-smtp-delivery-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729565AbfKARe6 (ORCPT ); Fri, 1 Nov 2019 13:34:58 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629696; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=n3+Jnro9bGTxUGjNWG4PT+hlbnJz5YjGU3Jvtd/eEGU=; b=BGS7yE9za3JadxWsUX05Yx4e83ppS34SB0fVl01EwmJ379aTt4Wjm2ctJm1OEDoIwj5AHf 5viQ2ZrmLt5ZTERsRO0mnpid0fCfEXqH6+BAyi6tmJVOn/XfEjlWgMffsxZ5gJY5X+DyQR gPo6ivCXsfFbooERN4zyhAhgcAl9NNo= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-59-p6rvO6F6NI2v9viUzMq_WQ-1; Fri, 01 Nov 2019 13:34:54 -0400 Received: from smtp.corp.redhat.com (int-mx06.intmail.prod.int.phx2.redhat.com [10.5.11.16]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id F40AF2EDC; Fri, 1 Nov 2019 17:34:51 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id 95A685C28E; Fri, 1 Nov 2019 17:34:47 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 05/11] pipe: Allow pipes to have kernel-reserved slots [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:34:46 +0000 Message-ID: <157262968684.13142.910827322486465736.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.16 X-MC-Unique: p6rvO6F6NI2v9viUzMq_WQ-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Split pipe->ring_size into two numbers: (1) pipe->ring_size - indicates the hard size of the pipe ring. (2) pipe->max_usage - indicates the maximum number of pipe ring slots that userspace orchestrated events can fill. This allows for a pipe that is both writable by the general kernel notification facility and by userspace, allowing plenty of ring space for notifications to be added whilst preventing userspace from being able to pin too much unswappable kernel space. Signed-off-by: David Howells --- fs/fuse/dev.c | 8 ++++---- fs/pipe.c | 10 ++++++---- fs/splice.c | 26 +++++++++++++------------- include/linux/pipe_fs_i.h | 6 +++++- lib/iov_iter.c | 4 ++-- 5 files changed, 30 insertions(+), 24 deletions(-) diff --git a/fs/fuse/dev.c b/fs/fuse/dev.c index c56011f95a87..423b6c657bf0 100644 --- a/fs/fuse/dev.c +++ b/fs/fuse/dev.c @@ -703,7 +703,7 @@ static int fuse_copy_fill(struct fuse_copy_state *cs) cs->pipebufs++; cs->nr_segs--; } else { - if (cs->nr_segs >= cs->pipe->ring_size) + if (cs->nr_segs >= cs->pipe->max_usage) return -EIO; page = alloc_page(GFP_HIGHUSER); @@ -879,7 +879,7 @@ static int fuse_ref_page(struct fuse_copy_state *cs, struct page *page, struct pipe_buffer *buf; int err; - if (cs->nr_segs >= cs->pipe->ring_size) + if (cs->nr_segs >= cs->pipe->max_usage) return -EIO; err = unlock_request(cs->req); @@ -1341,7 +1341,7 @@ static ssize_t fuse_dev_splice_read(struct file *in, loff_t *ppos, if (!fud) return -EPERM; - bufs = kvmalloc_array(pipe->ring_size, sizeof(struct pipe_buffer), + bufs = kvmalloc_array(pipe->max_usage, sizeof(struct pipe_buffer), GFP_KERNEL); if (!bufs) return -ENOMEM; @@ -1353,7 +1353,7 @@ static ssize_t fuse_dev_splice_read(struct file *in, loff_t *ppos, if (ret < 0) goto out; - if (pipe_occupancy(pipe->head, pipe->tail) + cs.nr_segs > pipe->ring_size) { + if (pipe_occupancy(pipe->head, pipe->tail) + cs.nr_segs > pipe->max_usage) { ret = -EIO; goto out; } diff --git a/fs/pipe.c b/fs/pipe.c index e9b361cb093e..69afeab8a73a 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -404,7 +404,7 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) tail = pipe->tail; head = pipe->head; - max_usage = pipe->ring_size; + max_usage = pipe->max_usage; mask = pipe->ring_size - 1; /* We try to merge small writes */ @@ -571,7 +571,7 @@ pipe_poll(struct file *filp, poll_table *wait) } if (filp->f_mode & FMODE_WRITE) { - if (!pipe_full(head, tail, pipe->ring_size)) + if (!pipe_full(head, tail, pipe->max_usage)) mask |= EPOLLOUT | EPOLLWRNORM; /* * Most Unices do not set EPOLLERR for FIFOs but on Linux they @@ -696,6 +696,7 @@ struct pipe_inode_info *alloc_pipe_info(void) if (pipe->bufs) { init_waitqueue_head(&pipe->wait); pipe->r_counter = pipe->w_counter = 1; + pipe->max_usage = pipe_bufs; pipe->ring_size = pipe_bufs; pipe->user = user; mutex_init(&pipe->mutex); @@ -1150,9 +1151,10 @@ static long pipe_set_size(struct pipe_inode_info *pipe, unsigned long arg) kfree(pipe->bufs); pipe->bufs = bufs; pipe->ring_size = nr_slots; + pipe->max_usage = nr_slots; pipe->tail = tail; pipe->head = head; - return pipe->ring_size * PAGE_SIZE; + return pipe->max_usage * PAGE_SIZE; out_revert_acct: (void) account_pipe_buffers(pipe->user, nr_slots, pipe->ring_size); @@ -1185,7 +1187,7 @@ long pipe_fcntl(struct file *file, unsigned int cmd, unsigned long arg) ret = pipe_set_size(pipe, arg); break; case F_GETPIPE_SZ: - ret = pipe->ring_size * PAGE_SIZE; + ret = pipe->max_usage * PAGE_SIZE; break; default: ret = -EINVAL; diff --git a/fs/splice.c b/fs/splice.c index 22b0a47a35c0..c521090a0469 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -199,7 +199,7 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe, goto out; } - while (!pipe_full(head, tail, pipe->ring_size)) { + while (!pipe_full(head, tail, pipe->max_usage)) { struct pipe_buffer *buf = &pipe->bufs[head & mask]; buf->page = spd->pages[page_nr]; @@ -239,7 +239,7 @@ ssize_t add_to_pipe(struct pipe_inode_info *pipe, struct pipe_buffer *buf) if (unlikely(!pipe->readers)) { send_sig(SIGPIPE, current, 0); ret = -EPIPE; - } else if (pipe_full(head, tail, pipe->ring_size)) { + } else if (pipe_full(head, tail, pipe->max_usage)) { ret = -EAGAIN; } else { pipe->bufs[head & mask] = *buf; @@ -257,7 +257,7 @@ EXPORT_SYMBOL(add_to_pipe); */ int splice_grow_spd(const struct pipe_inode_info *pipe, struct splice_pipe_desc *spd) { - unsigned int max_usage = READ_ONCE(pipe->ring_size); + unsigned int max_usage = READ_ONCE(pipe->max_usage); spd->nr_pages_max = max_usage; if (max_usage <= PIPE_DEF_BUFFERS) @@ -381,7 +381,7 @@ static ssize_t default_file_splice_read(struct file *in, loff_t *ppos, ssize_t res; int i; - if (pipe_full(pipe->head, pipe->tail, pipe->ring_size)) + if (pipe_full(pipe->head, pipe->tail, pipe->max_usage)) return -EAGAIN; /* @@ -698,7 +698,7 @@ iter_file_splice_write(struct pipe_inode_info *pipe, struct file *out, .pos = *ppos, .u.file = out, }; - int nbufs = pipe->ring_size; + int nbufs = pipe->max_usage; struct bio_vec *array = kcalloc(nbufs, sizeof(struct bio_vec), GFP_KERNEL); ssize_t ret; @@ -721,9 +721,9 @@ iter_file_splice_write(struct pipe_inode_info *pipe, struct file *out, if (ret <= 0) break; - if (unlikely(nbufs < pipe->ring_size)) { + if (unlikely(nbufs < pipe->max_usage)) { kfree(array); - nbufs = pipe->ring_size; + nbufs = pipe->max_usage; array = kcalloc(nbufs, sizeof(struct bio_vec), GFP_KERNEL); if (!array) { @@ -963,7 +963,7 @@ ssize_t splice_direct_to_actor(struct file *in, struct splice_desc *sd, loff_t pos = sd->pos, prev_pos = pos; /* Don't try to read more the pipe has space for. */ - p_space = pipe->ring_size - + p_space = pipe->max_usage - pipe_occupancy(pipe->head, pipe->tail); read_len = min_t(size_t, len, p_space << PAGE_SHIFT); ret = do_splice_to(in, &pos, pipe, read_len, flags); @@ -1090,7 +1090,7 @@ static int wait_for_space(struct pipe_inode_info *pipe, unsigned flags) send_sig(SIGPIPE, current, 0); return -EPIPE; } - if (!pipe_full(pipe->head, pipe->tail, pipe->ring_size)) + if (!pipe_full(pipe->head, pipe->tail, pipe->max_usage)) return 0; if (flags & SPLICE_F_NONBLOCK) return -EAGAIN; @@ -1498,13 +1498,13 @@ static int opipe_prep(struct pipe_inode_info *pipe, unsigned int flags) * Check pipe occupancy without the inode lock first. This function * is speculative anyways, so missing one is ok. */ - if (pipe_full(pipe->head, pipe->tail, pipe->ring_size)) + if (pipe_full(pipe->head, pipe->tail, pipe->max_usage)) return 0; ret = 0; pipe_lock(pipe); - while (pipe_full(pipe->head, pipe->tail, pipe->ring_size)) { + while (pipe_full(pipe->head, pipe->tail, pipe->max_usage)) { if (!pipe->readers) { send_sig(SIGPIPE, current, 0); ret = -EPIPE; @@ -1584,7 +1584,7 @@ static int splice_pipe_to_pipe(struct pipe_inode_info *ipipe, * pipe is empty or the output pipe is full. */ if (pipe_empty(i_head, i_tail) || - pipe_full(o_head, o_tail, opipe->ring_size)) { + pipe_full(o_head, o_tail, opipe->max_usage)) { /* Already processed some buffers, break */ if (ret) break; @@ -1706,7 +1706,7 @@ static int link_pipe(struct pipe_inode_info *ipipe, * output room, break. */ if (pipe_empty(i_head, i_tail) || - pipe_full(o_head, o_tail, opipe->ring_size)) + pipe_full(o_head, o_tail, opipe->max_usage)) break; ibuf = &ipipe->bufs[i_tail & i_mask]; diff --git a/include/linux/pipe_fs_i.h b/include/linux/pipe_fs_i.h index 96158ca80456..44f2245debda 100644 --- a/include/linux/pipe_fs_i.h +++ b/include/linux/pipe_fs_i.h @@ -32,6 +32,7 @@ struct pipe_buffer { * @wait: reader/writer wait point in case of empty/full pipe * @head: The point of buffer production * @tail: The point of buffer consumption + * @max_usage: The maximum number of slots that may be used in the ring * @ring_size: total number of buffers (should be a power of 2) * @tmp_page: cached released page * @readers: number of current readers of this pipe @@ -50,6 +51,7 @@ struct pipe_inode_info { wait_queue_head_t wait; unsigned int head; unsigned int tail; + unsigned int max_usage; unsigned int ring_size; unsigned int readers; unsigned int writers; @@ -150,9 +152,11 @@ static inline unsigned int pipe_space_for_user(unsigned int head, unsigned int t unsigned int p_occupancy, p_space; p_occupancy = pipe_occupancy(head, tail); - if (p_occupancy >= pipe->ring_size) + if (p_occupancy >= pipe->max_usage) return 0; p_space = pipe->ring_size - p_occupancy; + if (p_space > pipe->max_usage) + p_space = pipe->max_usage; return p_space; } diff --git a/lib/iov_iter.c b/lib/iov_iter.c index 6607e9c875ce..8225279b96e3 100644 --- a/lib/iov_iter.c +++ b/lib/iov_iter.c @@ -395,7 +395,7 @@ static size_t copy_page_to_iter_pipe(struct page *page, size_t offset, size_t by i_head++; buf = &pipe->bufs[i_head & p_mask]; } - if (pipe_full(i_head, p_tail, pipe->ring_size)) + if (pipe_full(i_head, p_tail, pipe->max_usage)) return 0; buf->ops = &page_cache_pipe_buf_ops; @@ -528,7 +528,7 @@ static size_t push_pipe(struct iov_iter *i, size_t size, pipe->bufs[iter_head & p_mask].len = PAGE_SIZE; iter_head++; } - while (!pipe_full(iter_head, p_tail, pipe->ring_size)) { + while (!pipe_full(iter_head, p_tail, pipe->max_usage)) { struct pipe_buffer *buf = &pipe->bufs[iter_head & p_mask]; struct page *page = alloc_page(GFP_USER); if (!page) From patchwork Fri Nov 1 17:34:57 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223529 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 8EE81913 for ; Fri, 1 Nov 2019 17:35:14 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 6BE14217F9 for ; Fri, 1 Nov 2019 17:35:14 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="MUd1OMSp" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729759AbfKARfN (ORCPT ); Fri, 1 Nov 2019 13:35:13 -0400 Received: from us-smtp-1.mimecast.com ([207.211.31.81]:23652 "EHLO us-smtp-1.mimecast.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1729741AbfKARfH (ORCPT ); Fri, 1 Nov 2019 13:35:07 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629705; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=Qoe4+68SxhMAUVnSP6Av5bTGeZl7pUYmSw55Mh+o64M=; b=MUd1OMSp97oaKyHEOD8SLUfBByCJmokllkYGa3v8qyS895A98rrHcxuPmZHzx2fRFfowwM N38DrPQRT/SEQhbHwO+tkLN9UZZO5zr7dp0ToL9TvdZa4vNsfi4dmTc7v1QgjiyTB5u5eu m12A49G73QREE1A4gGCe7SiOrkE7wRg= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-13-L0312JSnMXO9VBv0kwfEmw-1; Fri, 01 Nov 2019 13:35:02 -0400 Received: from smtp.corp.redhat.com (int-mx03.intmail.prod.int.phx2.redhat.com [10.5.11.13]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id A4D81800A1A; Fri, 1 Nov 2019 17:35:00 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id DFDA860876; Fri, 1 Nov 2019 17:34:57 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 06/11] pipe: Advance tail pointer inside of wait spinlock in pipe_read() [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:34:57 +0000 Message-ID: <157262969721.13142.294641621235161592.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.13 X-MC-Unique: L0312JSnMXO9VBv0kwfEmw-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Advance the pipe ring tail pointer inside of wait spinlock in pipe_read() so that the pipe can be written into with kernel notifications from contexts where pipe->mutex cannot be taken. Signed-off-by: David Howells --- fs/pipe.c | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-) diff --git a/fs/pipe.c b/fs/pipe.c index 69afeab8a73a..ea134f69a292 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -325,9 +325,14 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to) if (!buf->len) { pipe_buf_release(pipe, buf); + spin_lock_irq(&pipe->wait.lock); tail++; pipe->tail = tail; - do_wakeup = 1; + do_wakeup = 0; + wake_up_interruptible_sync_poll_locked( + &pipe->wait, EPOLLOUT | EPOLLWRNORM); + spin_unlock_irq(&pipe->wait.lock); + kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); } total_len -= chars; if (!total_len) @@ -359,6 +364,7 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to) if (do_wakeup) { wake_up_interruptible_sync_poll(&pipe->wait, EPOLLOUT | EPOLLWRNORM); kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); + do_wakeup = 0; } pipe_wait(pipe); } From patchwork Fri Nov 1 17:35:05 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223545 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id AC258913 for ; Fri, 1 Nov 2019 17:35:28 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 8C4C021855 for ; Fri, 1 Nov 2019 17:35:28 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="FjspXGU2" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729707AbfKARfU (ORCPT ); Fri, 1 Nov 2019 13:35:20 -0400 Received: from us-smtp-1.mimecast.com ([207.211.31.81]:56885 "EHLO us-smtp-delivery-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729741AbfKARfS (ORCPT ); Fri, 1 Nov 2019 13:35:18 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629717; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=f31ZclG/RJxENvGCAs10O3ojjOrA9kpGCTfILvXe/Gk=; b=FjspXGU2PeyaFJLGdB1SmRC3gIlTz/m5daYv6ZZLcIfCGGKqlIxVfnhh4SChMMRgA3iGUr abrT8jtk6dDOb0WDqc0KEG8mtQ42U6JpT97KjeWOROK2MDcvE0dmgucZ+uvzzqoYe/ldfm 0poOjs1JtvhnADsZVIuK8SKGxSczF44= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-273-596x9dcRPvmD5UotO2FJKA-1; Fri, 01 Nov 2019 13:35:14 -0400 Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.phx2.redhat.com [10.5.11.12]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id AFCBD800686; Fri, 1 Nov 2019 17:35:10 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id A78ED60BEC; Fri, 1 Nov 2019 17:35:06 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 07/11] pipe: Conditionalise wakeup in pipe_read() [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:35:05 +0000 Message-ID: <157262970587.13142.9476858866693533437.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.12 X-MC-Unique: 596x9dcRPvmD5UotO2FJKA-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Only do a wakeup in pipe_read() if we made space in a completely full buffer. The producer shouldn't be waiting on pipe->wait otherwise. Signed-off-by: David Howells --- fs/pipe.c | 15 ++++++--------- 1 file changed, 6 insertions(+), 9 deletions(-) diff --git a/fs/pipe.c b/fs/pipe.c index ea134f69a292..c16950e36ded 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -328,11 +328,13 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to) spin_lock_irq(&pipe->wait.lock); tail++; pipe->tail = tail; - do_wakeup = 0; - wake_up_interruptible_sync_poll_locked( - &pipe->wait, EPOLLOUT | EPOLLWRNORM); + do_wakeup = 1; + if (head - (tail - 1) == pipe->max_usage) + wake_up_interruptible_sync_poll_locked( + &pipe->wait, EPOLLOUT | EPOLLWRNORM); spin_unlock_irq(&pipe->wait.lock); - kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); + if (head - (tail - 1) == pipe->max_usage) + kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); } total_len -= chars; if (!total_len) @@ -361,11 +363,6 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to) ret = -ERESTARTSYS; break; } - if (do_wakeup) { - wake_up_interruptible_sync_poll(&pipe->wait, EPOLLOUT | EPOLLWRNORM); - kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); - do_wakeup = 0; - } pipe_wait(pipe); } __pipe_unlock(pipe); From patchwork Fri Nov 1 17:35:15 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223541 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 42D421747 for ; Fri, 1 Nov 2019 17:35:27 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 239B121855 for ; Fri, 1 Nov 2019 17:35:27 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="Dm1aJdKj" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729812AbfKARfZ (ORCPT ); Fri, 1 Nov 2019 13:35:25 -0400 Received: from us-smtp-delivery-1.mimecast.com ([207.211.31.120]:52181 "EHLO us-smtp-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1727988AbfKARfZ (ORCPT ); Fri, 1 Nov 2019 13:35:25 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629724; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=bQVPVip8ff0LcfPaTnPD2rpsZ9HBejI/JIzJ0jq//Ps=; b=Dm1aJdKjRmY8+wc4naer/Kd0Acg5LKpbt+yDsjsZdYkMNYo+j+zAMr2DaiDXbT4TRZU038 zuMBX03CNnm3uv0wxV1MYdC9rERpW3cUsoYmE6OZH+RVVaGYxwQSOcvaEB8ZJvRvUrV2tQ 8qivf5Cvx9KgdnljbkXvNTThq1PfSY8= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-287-pqc9grtDOSGwWyyMavPlOA-1; Fri, 01 Nov 2019 13:35:21 -0400 Received: from smtp.corp.redhat.com (int-mx08.intmail.prod.int.phx2.redhat.com [10.5.11.23]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 35FA82EDC; Fri, 1 Nov 2019 17:35:19 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id 965F619C58; Fri, 1 Nov 2019 17:35:16 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 08/11] pipe: Rearrange sequence in pipe_write() to preallocate slot [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:35:15 +0000 Message-ID: <157262971590.13142.1426435250123334464.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.84 on 10.5.11.23 X-MC-Unique: pqc9grtDOSGwWyyMavPlOA-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Rearrange the sequence in pipe_write() so that the allocation of the new buffer, the allocation of a ring slot and the attachment to the ring is done under the pipe wait spinlock and then the lock is dropped and the buffer can be filled. The data copy needs to be done with the spinlock unheld and irqs enabled, so the lock needs to be dropped first. However, the reader can't progress as we're holding pipe->mutex. We also need to drop the lock as that would impact others looking at the pipe waitqueue, such as poll(), the consumer and a future kernel message writer. We just abandon the preallocated slot if we get a copy error. Future writes may continue it and a future read will eventually recycle it. Signed-off-by: David Howells --- fs/pipe.c | 51 +++++++++++++++++++++++++++++++++------------------ 1 file changed, 33 insertions(+), 18 deletions(-) diff --git a/fs/pipe.c b/fs/pipe.c index c16950e36ded..ce77ac0d8901 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -387,7 +387,7 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) { struct file *filp = iocb->ki_filp; struct pipe_inode_info *pipe = filp->private_data; - unsigned int head, tail, max_usage, mask; + unsigned int head, max_usage, mask; ssize_t ret = 0; int do_wakeup = 0; size_t total_len = iov_iter_count(from); @@ -405,14 +405,13 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) goto out; } - tail = pipe->tail; head = pipe->head; max_usage = pipe->max_usage; mask = pipe->ring_size - 1; /* We try to merge small writes */ chars = total_len & (PAGE_SIZE-1); /* size of the last buffer */ - if (!pipe_empty(head, tail) && chars != 0) { + if (!pipe_empty(head, pipe->tail) && chars != 0) { struct pipe_buffer *buf = &pipe->bufs[(head - 1) & mask]; int offset = buf->offset + buf->len; @@ -441,8 +440,8 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) break; } - tail = pipe->tail; - if (!pipe_full(head, tail, max_usage)) { + head = pipe->head; + if (!pipe_full(head, pipe->tail, max_usage)) { struct pipe_buffer *buf = &pipe->bufs[head & mask]; struct page *page = pipe->tmp_page; int copied; @@ -455,40 +454,56 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) } pipe->tmp_page = page; } + + /* Allocate a slot in the ring in advance and attach an + * empty buffer. If we fault or otherwise fail to use + * it, either the reader will consume it or it'll still + * be there for the next write. + */ + spin_lock_irq(&pipe->wait.lock); + + head = pipe->head; + pipe->head = head + 1; + /* Always wake up, even if the copy fails. Otherwise * we lock up (O_NONBLOCK-)readers that sleep due to * syscall merging. * FIXME! Is this really true? */ - do_wakeup = 1; - copied = copy_page_from_iter(page, 0, PAGE_SIZE, from); - if (unlikely(copied < PAGE_SIZE && iov_iter_count(from))) { - if (!ret) - ret = -EFAULT; - break; - } - ret += copied; + wake_up_interruptible_sync_poll_locked( + &pipe->wait, EPOLLIN | EPOLLRDNORM); + + spin_unlock_irq(&pipe->wait.lock); + kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); /* Insert it into the buffer array */ + buf = &pipe->bufs[head & mask]; buf->page = page; buf->ops = &anon_pipe_buf_ops; buf->offset = 0; - buf->len = copied; + buf->len = 0; buf->flags = 0; if (is_packetized(filp)) { buf->ops = &packet_pipe_buf_ops; buf->flags = PIPE_BUF_FLAG_PACKET; } - - head++; - pipe->head = head; pipe->tmp_page = NULL; + copied = copy_page_from_iter(page, 0, PAGE_SIZE, from); + if (unlikely(copied < PAGE_SIZE && iov_iter_count(from))) { + if (!ret) + ret = -EFAULT; + break; + } + ret += copied; + buf->offset = 0; + buf->len = copied; + if (!iov_iter_count(from)) break; } - if (!pipe_full(head, tail, max_usage)) + if (!pipe_full(head, pipe->tail, max_usage)) continue; /* Wait for buffer space to become available. */ From patchwork Fri Nov 1 17:35:24 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223553 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id CBEE01747 for ; Fri, 1 Nov 2019 17:35:40 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id ABF8121855 for ; Fri, 1 Nov 2019 17:35:40 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="e8kS7un1" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728362AbfKARfj (ORCPT ); Fri, 1 Nov 2019 13:35:39 -0400 Received: from us-smtp-1.mimecast.com ([205.139.110.61]:40879 "EHLO us-smtp-delivery-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729803AbfKARfd (ORCPT ); Fri, 1 Nov 2019 13:35:33 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629732; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=G/3IjFBlrcu9yWMCYn5372hkbcEcVSpB5vL0sJqAX3o=; b=e8kS7un1jUu9yv5WMzhOdDPLizxERroGi1aiDCbdM5A5s4PDZpvYv+1sMwjFAw1JDB25HR GRLlS5rN4nVJ2rys5uxtSd/Jp8o4lHFnh+rpDIsOEcYfe0O/zPRQUyBb3roUqEuBGuy7OP oFMdksJhdA0d5ngnRP5qd8n+SXFMMc0= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-268-vc67HLdpPwahTj8KtFndSw-1; Fri, 01 Nov 2019 13:35:29 -0400 Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.phx2.redhat.com [10.5.11.12]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 8C0271800D6B; Fri, 1 Nov 2019 17:35:27 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id 1F0F860BEC; Fri, 1 Nov 2019 17:35:24 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 09/11] pipe: Remove redundant wakeup from pipe_write() [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:35:24 +0000 Message-ID: <157262972442.13142.11480420475866936493.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.12 X-MC-Unique: vc67HLdpPwahTj8KtFndSw-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Remove a redundant wakeup from pipe_write(). Signed-off-by: David Howells --- fs/pipe.c | 5 ----- 1 file changed, 5 deletions(-) diff --git a/fs/pipe.c b/fs/pipe.c index ce77ac0d8901..d7b8d3f22987 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -517,11 +517,6 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) ret = -ERESTARTSYS; break; } - if (do_wakeup) { - wake_up_interruptible_sync_poll(&pipe->wait, EPOLLIN | EPOLLRDNORM); - kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); - do_wakeup = 0; - } pipe->waiting_writers++; pipe_wait(pipe); pipe->waiting_writers--; From patchwork Fri Nov 1 17:35:32 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223559 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id 1BF051599 for ; Fri, 1 Nov 2019 17:35:47 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id EF01021855 for ; Fri, 1 Nov 2019 17:35:46 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="LWU3p0ik" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1729895AbfKARfm (ORCPT ); Fri, 1 Nov 2019 13:35:42 -0400 Received: from us-smtp-1.mimecast.com ([207.211.31.81]:38716 "EHLO us-smtp-delivery-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729439AbfKARfm (ORCPT ); Fri, 1 Nov 2019 13:35:42 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629741; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=N2gLZQTmOtfe94PwgELEjIRO8IuV3gWS7pjAAw1ld4Q=; b=LWU3p0ikQbApW8GBZRyNBYPVF1420EKSoUdelk9Miu/iTy9ZYUbluIznJOFlfw/1ci/zIU 0ENhXrAnb8bacE/dvrEckrGiWvV+2bosOVF5lZAxgD1W6iIj4TCfbFZbFrJXDTOPRMmE4x 2tcEzX2ofjsMExD0jwAVvz43RrOg50c= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-31-c0mbb6TfPoiAsyuG_xveUQ-1; Fri, 01 Nov 2019 13:35:38 -0400 Received: from smtp.corp.redhat.com (int-mx02.intmail.prod.int.phx2.redhat.com [10.5.11.12]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 3B4502EDC; Fri, 1 Nov 2019 17:35:36 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id 9C9B160C63; Fri, 1 Nov 2019 17:35:33 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 10/11] pipe: Check for ring full inside of the spinlock in pipe_write() [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:35:32 +0000 Message-ID: <157262973277.13142.11996381016183033698.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.12 X-MC-Unique: c0mbb6TfPoiAsyuG_xveUQ-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Make pipe_write() check to see if the ring has become full between it taking the pipe mutex, checking the ring status and then taking the spinlock. This can happen if a notification is written into the pipe as that happens without the pipe mutex. Signed-off-by: David Howells --- fs/pipe.c | 5 +++++ 1 file changed, 5 insertions(+) diff --git a/fs/pipe.c b/fs/pipe.c index d7b8d3f22987..aba2455caabe 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -463,6 +463,11 @@ pipe_write(struct kiocb *iocb, struct iov_iter *from) spin_lock_irq(&pipe->wait.lock); head = pipe->head; + if (pipe_full(head, pipe->tail, max_usage)) { + spin_unlock_irq(&pipe->wait.lock); + continue; + } + pipe->head = head + 1; /* Always wake up, even if the copy fails. Otherwise From patchwork Fri Nov 1 17:35:41 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: David Howells X-Patchwork-Id: 11223569 Return-Path: Received: from mail.kernel.org (pdx-korg-mail-1.web.codeaurora.org [172.30.200.123]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id AD72D913 for ; Fri, 1 Nov 2019 17:35:55 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.kernel.org (Postfix) with ESMTP id 8D26F20862 for ; Fri, 1 Nov 2019 17:35:55 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=redhat.com header.i=@redhat.com header.b="ATN+ZACa" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1728821AbfKARfy (ORCPT ); Fri, 1 Nov 2019 13:35:54 -0400 Received: from us-smtp-2.mimecast.com ([205.139.110.61]:30551 "EHLO us-smtp-delivery-1.mimecast.com" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1729932AbfKARfv (ORCPT ); Fri, 1 Nov 2019 13:35:51 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1572629750; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=Vi11bsC86DIjvR3JukBeRWivqFrcNqVuRmOTzZJ3x6E=; b=ATN+ZACaQxhHrOMEJVgoBf/W41sr40yN4v97GH3Rg+f+WQpJQ3GBSyVpmV7SPsN2oGJUu1 rMQKwXsB2O5IZuD/AS71e84Pj5W4ULhJvHt0QRudHeJG55oDwnzAo9nU5LZs5MfnKE0YYY V3/A2wvJnA3RpXYpr7yOyVXwgan8iGQ= Received: from mimecast-mx01.redhat.com (mimecast-mx01.redhat.com [209.132.183.4]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-280-jObNGFxmMiqYnw6I3pTN-A-1; Fri, 01 Nov 2019 13:35:46 -0400 Received: from smtp.corp.redhat.com (int-mx05.intmail.prod.int.phx2.redhat.com [10.5.11.15]) (using TLSv1.2 with cipher AECDH-AES256-SHA (256/256 bits)) (No client certificate requested) by mimecast-mx01.redhat.com (Postfix) with ESMTPS id 03A9A800D49; Fri, 1 Nov 2019 17:35:45 +0000 (UTC) Received: from warthog.procyon.org.uk (ovpn-121-40.rdu2.redhat.com [10.10.121.40]) by smtp.corp.redhat.com (Postfix) with ESMTP id 3C6B45D6B7; Fri, 1 Nov 2019 17:35:42 +0000 (UTC) Organization: Red Hat UK Ltd. Registered Address: Red Hat UK Ltd, Amberley Place, 107-111 Peascod Street, Windsor, Berkshire, SI4 1TE, United Kingdom. Registered in England and Wales under Company Registration No. 3798903 Subject: [RFC PATCH 11/11] pipe: Increase the writer-wakeup threshold to reduce context-switch count [ver #3] From: David Howells To: torvalds@linux-foundation.org Cc: dhowells@redhat.com, Rasmus Villemoes , Greg Kroah-Hartman , Peter Zijlstra , nicolas.dichtel@6wind.com, raven@themaw.net, Christian Brauner , dhowells@redhat.com, keyrings@vger.kernel.org, linux-usb@vger.kernel.org, linux-block@vger.kernel.org, linux-security-module@vger.kernel.org, linux-fsdevel@vger.kernel.org, linux-api@vger.kernel.org, linux-security-module@vger.kernel.org, linux-kernel@vger.kernel.org Date: Fri, 01 Nov 2019 17:35:41 +0000 Message-ID: <157262974145.13142.5910403713310791673.stgit@warthog.procyon.org.uk> In-Reply-To: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> References: <157262963995.13142.5568934007158044624.stgit@warthog.procyon.org.uk> User-Agent: StGit/unknown-version MIME-Version: 1.0 X-Scanned-By: MIMEDefang 2.79 on 10.5.11.15 X-MC-Unique: jObNGFxmMiqYnw6I3pTN-A-1 X-Mimecast-Spam-Score: 0 Sender: linux-usb-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-usb@vger.kernel.org Increase the threshold at which the reader sends a wake event to the writers in the queue such that the queue must be half empty before the wake is issued rather than the wake being issued when just a single slot available. This reduces the number of context switches in the tests significantly, without altering the amount of work achieved. With my pipe-bench program, there's a 20% reduction versus an unpatched kernel. Suggested-by: Konstantin Khlebnikov Signed-off-by: David Howells --- fs/pipe.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-) diff --git a/fs/pipe.c b/fs/pipe.c index aba2455caabe..9cd5cbef9552 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -324,16 +324,18 @@ pipe_read(struct kiocb *iocb, struct iov_iter *to) } if (!buf->len) { + bool wake; pipe_buf_release(pipe, buf); spin_lock_irq(&pipe->wait.lock); tail++; pipe->tail = tail; do_wakeup = 1; - if (head - (tail - 1) == pipe->max_usage) + wake = head - (tail - 1) == pipe->max_usage / 2; + if (wake) wake_up_interruptible_sync_poll_locked( &pipe->wait, EPOLLOUT | EPOLLWRNORM); spin_unlock_irq(&pipe->wait.lock); - if (head - (tail - 1) == pipe->max_usage) + if (wake) kill_fasync(&pipe->fasync_writers, SIGIO, POLL_OUT); } total_len -= chars;