Netdev List
 help / color / mirror / Atom feed
* [PATCH v6 08/10] ARM: mxs: add ocotp read function
From: Shawn Guo @ 2011-01-14 15:11 UTC (permalink / raw)
  To: davem, gerg, baruch, eric, bryan.wu, r64343, B32542,
	u.kleine-koenig
  Cc: Shawn Guo
In-Reply-To: <1294297998-26930-1-git-send-email-shawn.guo@freescale.com>

Signed-off-by: Shawn Guo <shawn.guo@freescale.com>
---
 arch/arm/mach-mxs/Kconfig               |    4 ++
 arch/arm/mach-mxs/Makefile              |    2 +
 arch/arm/mach-mxs/include/mach/common.h |    1 +
 arch/arm/mach-mxs/ocotp.c               |   90 +++++++++++++++++++++++++++++++
 4 files changed, 97 insertions(+), 0 deletions(-)
 create mode 100644 arch/arm/mach-mxs/ocotp.c

diff --git a/arch/arm/mach-mxs/Kconfig b/arch/arm/mach-mxs/Kconfig
index 8bfc8df..cd2fbdf 100644
--- a/arch/arm/mach-mxs/Kconfig
+++ b/arch/arm/mach-mxs/Kconfig
@@ -2,6 +2,9 @@ if ARCH_MXS
 
 source "arch/arm/mach-mxs/devices/Kconfig"
 
+config MXS_OCOTP
+	bool
+
 config SOC_IMX23
 	bool
 	select CPU_ARM926T
@@ -26,6 +29,7 @@ config MACH_MX28EVK
 	select SOC_IMX28
 	select MXS_HAVE_AMBA_DUART
 	select MXS_HAVE_PLATFORM_FEC
+	select MXS_OCOTP
 	default y
 	help
 	  Include support for MX28EVK platform. This includes specific
diff --git a/arch/arm/mach-mxs/Makefile b/arch/arm/mach-mxs/Makefile
index 39d3f9c..df501a8 100644
--- a/arch/arm/mach-mxs/Makefile
+++ b/arch/arm/mach-mxs/Makefile
@@ -1,6 +1,8 @@
 # Common support
 obj-y := clock.o devices.o gpio.o icoll.o iomux.o system.o timer.o
 
+obj-$(CONFIG_MXS_OCOTP) += ocotp.o
+
 obj-$(CONFIG_SOC_IMX23) += clock-mx23.o mm-mx23.o
 obj-$(CONFIG_SOC_IMX28) += clock-mx28.o mm-mx28.o
 
diff --git a/arch/arm/mach-mxs/include/mach/common.h b/arch/arm/mach-mxs/include/mach/common.h
index 59133eb..635bb5d 100644
--- a/arch/arm/mach-mxs/include/mach/common.h
+++ b/arch/arm/mach-mxs/include/mach/common.h
@@ -13,6 +13,7 @@
 
 struct clk;
 
+extern const u32 *mxs_get_ocotp(void);
 extern int mxs_reset_block(void __iomem *);
 extern void mxs_timer_init(struct clk *, int);
 
diff --git a/arch/arm/mach-mxs/ocotp.c b/arch/arm/mach-mxs/ocotp.c
new file mode 100644
index 0000000..65157a3
--- /dev/null
+++ b/arch/arm/mach-mxs/ocotp.c
@@ -0,0 +1,90 @@
+/*
+ * Copyright 2010 Freescale Semiconductor, Inc. All Rights Reserved.
+ *
+ * This program is free software; you can redistribute it and/or modify
+ * it under the terms of the GNU General Public License as published by
+ * the Free Software Foundation; either version 2 of the License, or
+ * (at your option) any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+ * GNU General Public License for more details.
+ */
+
+#include <linux/delay.h>
+#include <linux/err.h>
+#include <linux/mutex.h>
+
+#include <mach/mxs.h>
+
+#define OCOTP_WORD_OFFSET		0x20
+#define OCOTP_WORD_COUNT		0x20
+
+#define BM_OCOTP_CTRL_BUSY		(1 << 8)
+#define BM_OCOTP_CTRL_ERROR		(1 << 9)
+#define BM_OCOTP_CTRL_RD_BANK_OPEN	(1 << 12)
+
+static DEFINE_MUTEX(ocotp_mutex);
+static u32 ocotp_words[OCOTP_WORD_COUNT];
+
+const u32 *mxs_get_ocotp(void)
+{
+	void __iomem *ocotp_base = MXS_IO_ADDRESS(MXS_OCOTP_BASE_ADDR);
+	int timeout = 0x400;
+	size_t i;
+	static int once = 0;
+
+	if (once)
+		return ocotp_words;
+
+	mutex_lock(&ocotp_mutex);
+
+	/*
+	 * clk_enable(hbus_clk) for ocotp can be skipped
+	 * as it must be on when system is running.
+	 */
+
+	/* try to clear ERROR bit */
+	__mxs_clrl(BM_OCOTP_CTRL_ERROR, ocotp_base);
+
+	/* check both BUSY and ERROR cleared */
+	while ((__raw_readl(ocotp_base) &
+		(BM_OCOTP_CTRL_BUSY | BM_OCOTP_CTRL_ERROR)) && --timeout)
+		cpu_relax();
+
+	if (unlikely(!timeout))
+		goto error_unlock;
+
+	/* open OCOTP banks for read */
+	__mxs_setl(BM_OCOTP_CTRL_RD_BANK_OPEN, ocotp_base);
+
+	/* approximately wait 32 hclk cycles */
+	udelay(1);
+
+	/* poll BUSY bit becoming cleared */
+	timeout = 0x400;
+	while ((__raw_readl(ocotp_base) & BM_OCOTP_CTRL_BUSY) && --timeout)
+		cpu_relax();
+
+	if (unlikely(!timeout))
+		goto error_unlock;
+
+	for (i = 0; i < OCOTP_WORD_COUNT; i++)
+		ocotp_words[i] = __raw_readl(ocotp_base + OCOTP_WORD_OFFSET +
+						i * 0x10);
+
+	/* close banks for power saving */
+	__mxs_clrl(BM_OCOTP_CTRL_RD_BANK_OPEN, ocotp_base);
+
+	once = 1;
+
+	mutex_unlock(&ocotp_mutex);
+
+	return ocotp_words;
+
+error_unlock:
+	mutex_unlock(&ocotp_mutex);
+	pr_err("%s: timeout in reading OCOTP\n", __func__);
+	return NULL;
+}
-- 
1.7.1



^ permalink raw reply related

* Re: [PATCH v3 2/2] netlink: support setting devgroup parameters
From: jamal @ 2011-01-14 15:10 UTC (permalink / raw)
  To: Vlad Dogaru; +Cc: netdev, Octavian Purdila
In-Reply-To: <1294997911-13866-3-git-send-email-ddvlad@rosedu.org>

On Fri, 2011-01-14 at 11:38 +0200, Vlad Dogaru wrote:
> If a rtnetlink request specifies a negative or zero ifindex and has no
> interface name attribute, but has a group attribute, then the chenges
> are made to all the interfaces belonging to the specified group.
> 
> Signed-off-by: Vlad Dogaru <ddvlad@rosedu.org>

Acked-by: Jamal Hadi Salim <hadi@cyberus.ca>


cheers,
jamal


^ permalink raw reply

* Re: [PATCH v3 1/2] net_device: add support for network device groups
From: jamal @ 2011-01-14 15:10 UTC (permalink / raw)
  To: Vlad Dogaru; +Cc: netdev, Octavian Purdila
In-Reply-To: <1294997911-13866-2-git-send-email-ddvlad@rosedu.org>

On Fri, 2011-01-14 at 11:38 +0200, Vlad Dogaru wrote:
> Net devices can now be grouped, enabling simpler manipulation from
> userspace. This patch adds a group field to the net_device structure, as
> well as rtnetlink support to query and modify it.
> 
> Signed-off-by: Vlad Dogaru <ddvlad@rosedu.org>

Good stuff.

Acked-by: Jamal Hadi Salim <hadi@cyberus.ca>

cheers,
jamal


^ permalink raw reply

* Re: sch_sfb
From: Eric Dumazet @ 2011-01-14 15:09 UTC (permalink / raw)
  To: Juliusz Chroboczek; +Cc: Patrick McHardy, David Miller, netdev
In-Reply-To: <7ivd1rsj6n.fsf@lanthane.pps.jussieu.fr>

Le vendredi 14 janvier 2011 à 14:34 +0100, Juliusz Chroboczek a écrit :
> > I just looked at it out of interest after already having started my
> > own version.
> 
> >>   http://thread.gmane.org/gmane.linux.network/90225
> >>   http://thread.gmane.org/gmane.linux.network/90375
> 
> >> It was reviewed in particular by one Patrick McHardy.
> 
> > There's no reason to be pissed
> 
> Yes, there is.
> 
> First you object to my patch by making a bunch of unreasonable requests
> (notably that I use the in-kernel classifiers, which are not usable with
> Bloom filters).  Then it turns out you're implementing your own version
> "from scratch".  And then you claim that you never saw my version in the
> first place?
> 
> Patrick, what you're doing is not merely rude, it's actually unethical.
> 

Juliusz, I believe you overreact on this one.

Patrick reviews are really good and very reasonable. Yet, for a beginner
it might sounds difficult to understand and make the requested changes.

I sent you a private mail 2 weeks ago to offer my services to take SFB
and get it into kernel, because I read your previous attempts and
understood you had litle time to make the needed changes.

I got distracted by SFQ/CHOKe experiments but was about to work on SFB
when Patrick sent his mail. I really think we should work together, and
eventually offer SFB as a new qdisc to network admins :)

Thanks !



^ permalink raw reply

* Re: net 00/05: routing based send-to-self implementation
From: Kirill Smelkov @ 2011-01-14 15:02 UTC (permalink / raw)
  To: Jonathan Corbet; +Cc: Patrick McHardy, davem, netdev, Boris Kocherov
In-Reply-To: <20110114064048.5829220a@bike.lwn.net>

On Fri, Jan 14, 2011 at 06:40:48AM -0700, Jonathan Corbet wrote:
> On Fri, 14 Jan 2011 13:18:32 +0300
> Kirill Smelkov <kirr@mns.spb.ru> wrote:
> 
> >  ( Jonathan, I though something like this could be useful for LDD4 in
> >    revised snull not needing to play dirty tricks with IP addresses anymore )
> 
> Nice thought, thanks.  I don't know yet whether we'd want to do that or
> just run stuff on a virtual machine for LDD4 - lots of stuff to figure
> out still.

Just FYI: I've started with virtual machines, but figured out it is
(sometimes, maybe my fault) a bit of pain to coherently setup and also
that not all hardware have support for KVM - for example my netbook with
Atom N455, which I use half the time, does not support VT and they've
killed support for KQEMU in QEMU recently...

Anyway, good luck with LDD4!

Thanks,
Kirill

^ permalink raw reply

* [PATCH net-2.6 2/2] sfc: Restore the effect of the rss_cpus module parameter
From: Ben Hutchings @ 2011-01-14 14:25 UTC (permalink / raw)
  To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <1295014889.5386.1.camel@bwh-desktop>

Commit a4900ac ("sfc: Create multiple TX queues") accidentally
disabled the rss_cpus module parameter.

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
---
 drivers/net/sfc/efx.c |    3 +++
 1 files changed, 3 insertions(+), 0 deletions(-)

diff --git a/drivers/net/sfc/efx.c b/drivers/net/sfc/efx.c
index c2dc9a5..002bac7 100644
--- a/drivers/net/sfc/efx.c
+++ b/drivers/net/sfc/efx.c
@@ -1153,6 +1153,9 @@ static int efx_wanted_channels(void)
 	int count;
 	int cpu;
 
+	if (rss_cpus)
+		return rss_cpus;
+
 	if (unlikely(!zalloc_cpumask_var(&core_mask, GFP_KERNEL))) {
 		printk(KERN_WARNING
 		       "sfc: RSS disabled due to allocation failure\n");
-- 
1.7.3.4


-- 
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.


^ permalink raw reply related

* [PATCH net-2.6 1/2] sfc: Make efx_get_tx_queue() an inline function
From: Ben Hutchings @ 2011-01-14 14:25 UTC (permalink / raw)
  To: David Miller; +Cc: netdev, linux-net-drivers
In-Reply-To: <1295014889.5386.1.camel@bwh-desktop>

Signed-off-by: Ben Hutchings <bhutchings@solarflare.com>
---
 drivers/net/sfc/efx.c        |   15 +++------------
 drivers/net/sfc/net_driver.h |   10 ++++++++--
 2 files changed, 11 insertions(+), 14 deletions(-)

diff --git a/drivers/net/sfc/efx.c b/drivers/net/sfc/efx.c
index 711449c..c2dc9a5 100644
--- a/drivers/net/sfc/efx.c
+++ b/drivers/net/sfc/efx.c
@@ -1266,27 +1266,18 @@ static void efx_remove_interrupts(struct efx_nic *efx)
 	efx->legacy_irq = 0;
 }
 
-struct efx_tx_queue *
-efx_get_tx_queue(struct efx_nic *efx, unsigned index, unsigned type)
-{
-	unsigned tx_channel_offset =
-		separate_tx_channels ? efx->n_channels - efx->n_tx_channels : 0;
-	EFX_BUG_ON_PARANOID(index >= efx->n_tx_channels ||
-			    type >= EFX_TXQ_TYPES);
-	return &efx->channel[tx_channel_offset + index]->tx_queue[type];
-}
-
 static void efx_set_channels(struct efx_nic *efx)
 {
 	struct efx_channel *channel;
 	struct efx_tx_queue *tx_queue;
-	unsigned tx_channel_offset =
+
+	efx->tx_channel_offset =
 		separate_tx_channels ? efx->n_channels - efx->n_tx_channels : 0;
 
 	/* Channel pointers were set in efx_init_struct() but we now
 	 * need to clear them for TX queues in any RX-only channels. */
 	efx_for_each_channel(channel, efx) {
-		if (channel->channel - tx_channel_offset >=
+		if (channel->channel - efx->tx_channel_offset >=
 		    efx->n_tx_channels) {
 			efx_for_each_channel_tx_queue(tx_queue, channel)
 				tx_queue->channel = NULL;
diff --git a/drivers/net/sfc/net_driver.h b/drivers/net/sfc/net_driver.h
index bdce66d..28df866 100644
--- a/drivers/net/sfc/net_driver.h
+++ b/drivers/net/sfc/net_driver.h
@@ -735,6 +735,7 @@ struct efx_nic {
 	unsigned next_buffer_table;
 	unsigned n_channels;
 	unsigned n_rx_channels;
+	unsigned tx_channel_offset;
 	unsigned n_tx_channels;
 	unsigned int rx_buffer_len;
 	unsigned int rx_buffer_order;
@@ -929,8 +930,13 @@ efx_get_channel(struct efx_nic *efx, unsigned index)
 	     _channel = (_channel->channel + 1 < (_efx)->n_channels) ?	\
 		     (_efx)->channel[_channel->channel + 1] : NULL)
 
-extern struct efx_tx_queue *
-efx_get_tx_queue(struct efx_nic *efx, unsigned index, unsigned type);
+static inline struct efx_tx_queue *
+efx_get_tx_queue(struct efx_nic *efx, unsigned index, unsigned type)
+{
+	EFX_BUG_ON_PARANOID(index >= efx->n_tx_channels ||
+			    type >= EFX_TXQ_TYPES);
+	return &efx->channel[efx->tx_channel_offset + index]->tx_queue[type];
+}
 
 static inline struct efx_tx_queue *
 efx_channel_get_tx_queue(struct efx_channel *channel, unsigned type)
-- 
1.7.3.4



-- 
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.


^ permalink raw reply related

* Re: [PATCH] CHOKe flow scheduler (0.8)
From: Eric Dumazet @ 2011-01-14 14:24 UTC (permalink / raw)
  To: Patrick McHardy; +Cc: Stephen Hemminger, David Miller, netdev
In-Reply-To: <4D3055C2.3060807@trash.net>

Le vendredi 14 janvier 2011 à 14:55 +0100, Patrick McHardy a écrit :
> On 14.01.2011 14:54, Patrick McHardy wrote:
> > On 14.01.2011 00:34, Stephen Hemminger wrote:
> >> +static int choke_enqueue(struct sk_buff *skb, struct Qdisc *sch)
> >> +{
> >> ...
> >> +	/* Is queue small? */
> >> +	if (p->qavg <= p->qth_min)
> >> +		p->qcount = -1;
> >> +	else {
> >> +		struct sk_buff *oskb;
> >> +		unsigned int idx;
> >> +
> >> +		/* Draw a packet at random from queue */
> >> +		oskb = choke_peek_random(sch, &idx);
> >> +
> >> +		/* Both packets from same flow ? */
> >> +		if (*(unsigned int *)(qdisc_skb_cb(oskb)->data) == hash) {
> >> +			/* Drop both packets */
> >> +			q->stats.matched++;
> >> +			choke_drop_by_idx(q, idx);
> >> +			sch->qstats.backlog -= qdisc_pkt_len(skb);
> >> +			--sch->q.qlen;
> >> +			qdisc_drop(oskb, sch);
> > 
> > You need to adjust the qlen values of parent qdiscs by calling
> > qdisc_tree_decrease_qlen(), they are not aware that a second
> > packet has been dropped.
> 
> I just saw that Eric already fixed this :)
> --

Thanks Patrick, I did that in choke_change() only, not on this part of
the code.


qdisc cbq 1: root refcnt 2 rate 1000Mbit cell 8b (bounded,isolated) prio no-transmit/8 weight 1000Mbit allot 1514b 
level 2 ewma 5 avpkt 1000b maxidle 0us 
 Sent 92016573 bytes 167811 pkt (dropped 615544, overlimits 982637 requeues 0) 
 rate 6248bit 9pps backlog 0b 191202p requeues 0 
  borrowed 0 overactions 0 avgidle 125 undertime 0
qdisc choke 11: parent 1:11 limit 130000b min 10833b max 32500b ewma 13 Plog 21 Scell_log 30
 Sent 91969929 bytes 167257 pkt (dropped 806746, overlimits 424342 requeues 0) 
 rate 56bit 0pps backlog 0b 0p requeues 0 
  marked 0 early 424342 pdrop 0 other 0 matched 191202


So yes : we see 191202 were CHOKed, and upper cbq leaks 191202 packets in 'backlog'

After following patch :

diff --git a/net/sched/sch_choke.c b/net/sched/sch_choke.c
index 29a91d8..5479f7e 100644
--- a/net/sched/sch_choke.c
+++ b/net/sched/sch_choke.c
@@ -223,6 +223,7 @@ static int choke_enqueue(struct sk_buff *skb, struct Qdisc *sch)
 			q->stats.matched++;
 			choke_drop_by_idx(q, idx);
 			sch->qstats.backlog -= qdisc_pkt_len(oskb);
+			qdisc_tree_decrease_qlen(sch, 1);
 			--sch->q.qlen;
 			qdisc_drop(oskb, sch);
 			goto congestion_drop;

Everything seems fine :

qdisc cbq 1: root refcnt 2 rate 1000Mbit cell 8b (bounded,isolated) prio no-transmit/8 weight 1000Mbit allot 1514b 
level 2 ewma 5 avpkt 1000b maxidle 0us 
 Sent 93962148 bytes 170940 pkt (dropped 633053, overlimits 1010490 requeues 0) 
 rate 568bit 1pps backlog 0b 0p requeues 0 
  borrowed 0 overactions 0 avgidle 125 undertime 0
qdisc choke 11: parent 1:11 limit 130000b min 10833b max 32500b ewma 13 Plog 21 Scell_log 30
 Sent 93957621 bytes 170888 pkt (dropped 829160, overlimits 436946 requeues 0) 
 rate 104bit 0pps backlog 0b 0p requeues 0 
  marked 0 early 436946 pdrop 0 other 0 matched 196107



^ permalink raw reply related

* pull request: sfc-2.6 2011-01-14
From: Ben Hutchings @ 2011-01-14 14:21 UTC (permalink / raw)
  To: David Miller; +Cc: netdev, sf-linux-drivers

The following changes since commit 5b919f833d9d60588d026ad82d17f17e8872c7a9:

  net: ax25: fix information leak to userland harder (2011-01-12 00:34:49 -0800)

are available in the git repository at:
  git://git.kernel.org/pub/scm/linux/kernel/git/bwh/sfc-2.6.git master

A minor optimisation and a regression fix.

Ben.

Ben Hutchings (2):
      sfc: Make efx_get_tx_queue() an inline function
      sfc: Restore the effect of the rss_cpus module parameter

 drivers/net/sfc/efx.c        |   18 ++++++------------
 drivers/net/sfc/net_driver.h |   10 ++++++++--
 2 files changed, 14 insertions(+), 14 deletions(-)

-- 
Ben Hutchings, Senior Software Engineer, Solarflare Communications
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.




^ permalink raw reply

* Re: net 00/05: routing based send-to-self implementation
From: Jonathan Corbet @ 2011-01-14 13:40 UTC (permalink / raw)
  To: Kirill Smelkov; +Cc: Patrick McHardy, davem, netdev, Boris Kocherov
In-Reply-To: <20110114101832.GA3170@tugrik.mns.mnsspb.ru>

On Fri, 14 Jan 2011 13:18:32 +0300
Kirill Smelkov <kirr@mns.spb.ru> wrote:

>  ( Jonathan, I though something like this could be useful for LDD4 in
>    revised snull not needing to play dirty tricks with IP addresses anymore )

Nice thought, thanks.  I don't know yet whether we'd want to do that or
just run stuff on a virtual machine for LDD4 - lots of stuff to figure
out still.

Thanks,

jon

^ permalink raw reply

* [PATCH resend] Bluetooth: l2cap: fix misuse of logical operation in place of bitop
From: David Sterba @ 2011-01-14 13:59 UTC (permalink / raw)
  To: akpm
  Cc: marcel, linux-bluetooth, netdev, linux-kernel, David Sterba,
	Gustavo F. Padovan, João Paulo Rechi Vita

CC: Marcel Holtmann <marcel@holtmann.org>
CC: "Gustavo F. Padovan" <padovan@profusion.mobi>
CC: João Paulo Rechi Vita <jprvita@profusion.mobi>
Signed-off-by: David Sterba <dsterba@suse.cz>
---

Andrew, this has not been picked up by maintainers since 27.12., please consider it for -mm.

 net/bluetooth/l2cap.c |    4 ++--
 1 files changed, 2 insertions(+), 2 deletions(-)

diff --git a/net/bluetooth/l2cap.c b/net/bluetooth/l2cap.c
index cd8f6ea..bdfdfdc 100644
--- a/net/bluetooth/l2cap.c
+++ b/net/bluetooth/l2cap.c
@@ -1893,8 +1893,8 @@ static int l2cap_sock_sendmsg(struct kiocb *iocb, struct socket *sock, struct ms
 		if (pi->mode == L2CAP_MODE_STREAMING) {
 			l2cap_streaming_send(sk);
 		} else {
-			if (pi->conn_state & L2CAP_CONN_REMOTE_BUSY &&
-					pi->conn_state && L2CAP_CONN_WAIT_F) {
+			if ((pi->conn_state & L2CAP_CONN_REMOTE_BUSY) &&
+					(pi->conn_state & L2CAP_CONN_WAIT_F)) {
 				err = len;
 				break;
 			}
-- 
1.7.3.4.626.g73e7b


^ permalink raw reply related

* Re: [PATCH] CHOKe flow scheduler (0.8)
From: Patrick McHardy @ 2011-01-14 13:55 UTC (permalink / raw)
  To: Stephen Hemminger; +Cc: Eric Dumazet, David Miller, netdev
In-Reply-To: <4D305598.1010207@trash.net>

On 14.01.2011 14:54, Patrick McHardy wrote:
> On 14.01.2011 00:34, Stephen Hemminger wrote:
>> +static int choke_enqueue(struct sk_buff *skb, struct Qdisc *sch)
>> +{
>> ...
>> +	/* Is queue small? */
>> +	if (p->qavg <= p->qth_min)
>> +		p->qcount = -1;
>> +	else {
>> +		struct sk_buff *oskb;
>> +		unsigned int idx;
>> +
>> +		/* Draw a packet at random from queue */
>> +		oskb = choke_peek_random(sch, &idx);
>> +
>> +		/* Both packets from same flow ? */
>> +		if (*(unsigned int *)(qdisc_skb_cb(oskb)->data) == hash) {
>> +			/* Drop both packets */
>> +			q->stats.matched++;
>> +			choke_drop_by_idx(q, idx);
>> +			sch->qstats.backlog -= qdisc_pkt_len(skb);
>> +			--sch->q.qlen;
>> +			qdisc_drop(oskb, sch);
> 
> You need to adjust the qlen values of parent qdiscs by calling
> qdisc_tree_decrease_qlen(), they are not aware that a second
> packet has been dropped.

I just saw that Eric already fixed this :)

^ permalink raw reply

* Re: [PATCH] CHOKe flow scheduler (0.8)
From: Patrick McHardy @ 2011-01-14 13:54 UTC (permalink / raw)
  To: Stephen Hemminger; +Cc: Eric Dumazet, David Miller, netdev
In-Reply-To: <20110113153436.70d3c0a3@s6510>

On 14.01.2011 00:34, Stephen Hemminger wrote:
> +static int choke_enqueue(struct sk_buff *skb, struct Qdisc *sch)
> +{
> ...
> +	/* Is queue small? */
> +	if (p->qavg <= p->qth_min)
> +		p->qcount = -1;
> +	else {
> +		struct sk_buff *oskb;
> +		unsigned int idx;
> +
> +		/* Draw a packet at random from queue */
> +		oskb = choke_peek_random(sch, &idx);
> +
> +		/* Both packets from same flow ? */
> +		if (*(unsigned int *)(qdisc_skb_cb(oskb)->data) == hash) {
> +			/* Drop both packets */
> +			q->stats.matched++;
> +			choke_drop_by_idx(q, idx);
> +			sch->qstats.backlog -= qdisc_pkt_len(skb);
> +			--sch->q.qlen;
> +			qdisc_drop(oskb, sch);

You need to adjust the qlen values of parent qdiscs by calling
qdisc_tree_decrease_qlen(), they are not aware that a second
packet has been dropped.

> +			goto congestion_drop;
> +		}
> +

^ permalink raw reply

* [PATCH v3 13/16] nfs: introduce mount option 'rpcmount'
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs, David S. Miller, Rob Landley, Al Viro,
	containers, netdev, linux-kernel, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas@openvz.org>

It specifies rpc_pipefs to use. /var/lib/nfs/rpc_pipefs, by default.

Signed-off-by: Kirill A. Shutemov <kas@openvz.org>
---
 fs/nfs/callback.c         |    5 ++-
 fs/nfs/callback.h         |    3 +-
 fs/nfs/client.c           |   46 ++++++++++++++++++++++++++++++++++++--------
 fs/nfs/internal.h         |   10 +++++++-
 fs/nfs/mount_clnt.c       |    3 +-
 fs/nfs/namespace.c        |    3 +-
 fs/nfs/nfs4namespace.c    |   22 +++++++++++---------
 fs/nfs/super.c            |   20 +++++++++++++++++++
 include/linux/nfs_fs_sb.h |    1 +
 9 files changed, 86 insertions(+), 27 deletions(-)

diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index 7a535c8..c9814fb 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -276,7 +276,8 @@ int nfs4_set_callback_sessionid(struct nfs_client *clp)
 /*
  * Bring up the callback thread if it is not already up.
  */
-int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt)
+int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt,
+		struct vfsmount *rpcmount)
 {
 	struct svc_serv *serv = NULL;
 	struct svc_rqst *rqstp;
@@ -291,7 +292,7 @@ int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt)
 		nfs_callback_bc_serv(minorversion, xprt, cb_info);
 		goto out;
 	}
-	serv = svc_create(&nfs4_callback_program, init_rpc_pipefs,
+	serv = svc_create(&nfs4_callback_program, rpcmount,
 			NFS4_CALLBACK_BUFSIZE, NULL);
 	if (!serv) {
 		ret = -ENOMEM;
diff --git a/fs/nfs/callback.h b/fs/nfs/callback.h
index d3b44f9..9496e0f 100644
--- a/fs/nfs/callback.h
+++ b/fs/nfs/callback.h
@@ -175,7 +175,8 @@ extern __be32 nfs4_callback_getattr(struct cb_getattrargs *args,
 extern __be32 nfs4_callback_recall(struct cb_recallargs *args, void *dummy,
 				   struct cb_process_state *cps);
 #ifdef CONFIG_NFS_V4
-extern int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt);
+extern int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt,
+		struct vfsmount *rpcmount);
 extern void nfs_callback_down(int minorversion);
 extern int nfs4_validate_delegation_stateid(struct nfs_delegation *delegation,
 					    const nfs4_stateid *stateid);
diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index ad3b5e8..8dd0e99 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -131,6 +131,7 @@ struct nfs_client_initdata {
 	const struct nfs_rpc_ops *rpc_ops;
 	int proto;
 	u32 minorversion;
+	struct vfsmount *rpcmount;
 };
 
 /*
@@ -167,6 +168,7 @@ static struct nfs_client *nfs_alloc_client(const struct nfs_client_initdata *cl_
 	clp->cl_rpcclient = ERR_PTR(-EINVAL);
 
 	clp->cl_proto = cl_init->proto;
+	clp->cl_rpcmount = mntget(cl_init->rpcmount);
 
 #ifdef CONFIG_NFS_V4
 	err = nfs_get_cb_ident_idr(clp, cl_init->minorversion);
@@ -286,6 +288,7 @@ static void nfs_free_client(struct nfs_client *clp)
 	if (clp->cl_machine_cred != NULL)
 		put_rpccred(clp->cl_machine_cred);
 
+	mntput(clp->cl_rpcmount);
 	kfree(clp->cl_hostname);
 	kfree(clp);
 
@@ -471,6 +474,9 @@ static struct nfs_client *nfs_match_client(const struct nfs_client_initdata *dat
 		/* Match the full socket address */
 		if (!nfs_sockaddr_cmp(sap, clap))
 			continue;
+		/* Match rpc_pipefs mount point */
+		if (clp->cl_rpcmount->mnt_sb != data->rpcmount->mnt_sb)
+			continue;
 
 		atomic_inc(&clp->cl_count);
 		return clp;
@@ -629,7 +635,7 @@ static int nfs_create_rpc_client(struct nfs_client *clp,
 		.program	= &nfs_program,
 		.version	= clp->rpc_ops->version,
 		.authflavor	= flavor,
-		.rpcmount	= init_rpc_pipefs,
+		.rpcmount	= clp->cl_rpcmount,
 	};
 
 	if (discrtry)
@@ -664,7 +670,7 @@ static void nfs_destroy_server(struct nfs_server *server)
 /*
  * Version 2 or 3 lockd setup
  */
-static int nfs_start_lockd(struct nfs_server *server)
+static int nfs_start_lockd(struct nfs_server *server, struct vfsmount *rpcmount)
 {
 	struct nlm_host *host;
 	struct nfs_client *clp = server->nfs_client;
@@ -675,7 +681,7 @@ static int nfs_start_lockd(struct nfs_server *server)
 		.nfs_version	= clp->rpc_ops->version,
 		.noresvport	= server->flags & NFS_MOUNT_NORESVPORT ?
 					1 : 0,
-		.rpcmount	= init_rpc_pipefs,
+		.rpcmount	= rpcmount,
 	};
 
 	if (nlm_init.nfs_version > 3)
@@ -823,8 +829,16 @@ static int nfs_init_server(struct nfs_server *server,
 		cl_init.rpc_ops = &nfs_v3_clientops;
 #endif
 
+	cl_init.rpcmount = get_rpc_pipefs(data->rpcmount);
+	if (IS_ERR(cl_init.rpcmount)) {
+		dprintk("<-- nfs_init_server() = error %ld\n",
+				PTR_ERR(cl_init.rpcmount));
+		return PTR_ERR(cl_init.rpcmount);
+	}
+
 	/* Allocate or find a client reference we can use */
 	clp = nfs_get_client(&cl_init);
+	mntput(cl_init.rpcmount);
 	if (IS_ERR(clp)) {
 		dprintk("<-- nfs_init_server() = error %ld\n", PTR_ERR(clp));
 		return PTR_ERR(clp);
@@ -856,7 +870,7 @@ static int nfs_init_server(struct nfs_server *server,
 	server->acdirmax = data->acdirmax * HZ;
 
 	/* Start lockd here, before we might error out */
-	error = nfs_start_lockd(server);
+	error = nfs_start_lockd(server, clp->cl_rpcmount);
 	if (error < 0)
 		goto error;
 
@@ -1270,7 +1284,8 @@ static int nfs4_init_callback(struct nfs_client *clp)
 		}
 
 		error = nfs_callback_up(clp->cl_mvops->minor_version,
-					clp->cl_rpcclient->cl_xprt);
+					clp->cl_rpcclient->cl_xprt,
+					clp->cl_rpcmount);
 		if (error < 0) {
 			dprintk("%s: failed to start callback. Error = %d\n",
 				__func__, error);
@@ -1370,7 +1385,8 @@ static int nfs4_set_client(struct nfs_server *server,
 		const char *ip_addr,
 		rpc_authflavor_t authflavour,
 		int proto, const struct rpc_timeout *timeparms,
-		u32 minorversion)
+		u32 minorversion,
+		struct vfsmount *rpcmount)
 {
 	struct nfs_client_initdata cl_init = {
 		.hostname = hostname,
@@ -1379,6 +1395,7 @@ static int nfs4_set_client(struct nfs_server *server,
 		.rpc_ops = &nfs_v4_clientops,
 		.proto = proto,
 		.minorversion = minorversion,
+		.rpcmount = rpcmount,
 	};
 	struct nfs_client *clp;
 	int error;
@@ -1485,6 +1502,7 @@ static int nfs4_init_server(struct nfs_server *server,
 		const struct nfs_parsed_mount_data *data)
 {
 	struct rpc_timeout timeparms;
+	struct vfsmount *rpcmount;
 	int error;
 
 	dprintk("--> nfs4_init_server()\n");
@@ -1499,6 +1517,11 @@ static int nfs4_init_server(struct nfs_server *server,
 			server->caps |= NFS_CAP_READDIRPLUS;
 	server->options = data->options;
 
+	rpcmount = get_rpc_pipefs(data->rpcmount);
+	if (IS_ERR(rpcmount)) {
+		error = PTR_ERR(rpcmount);
+		goto error;
+	}
 	/* Get a client record */
 	error = nfs4_set_client(server,
 			data->nfs_server.hostname,
@@ -1508,7 +1531,9 @@ static int nfs4_init_server(struct nfs_server *server,
 			data->auth_flavors[0],
 			data->nfs_server.protocol,
 			&timeparms,
-			data->minorversion);
+			data->minorversion,
+			rpcmount);
+	mntput(rpcmount);
 	if (error < 0)
 		goto error;
 
@@ -1598,7 +1623,10 @@ struct nfs_server *nfs4_create_referral_server(struct nfs_clone_mount *data,
 				data->authflavor,
 				parent_server->client->cl_xprt->prot,
 				parent_server->client->cl_timeout,
-				parent_client->cl_mvops->minor_version);
+				parent_client->cl_mvops->minor_version,
+				parent_client->cl_rpcmount);
+
+
 	if (error < 0)
 		goto error;
 
@@ -1672,7 +1700,7 @@ struct nfs_server *nfs_clone_server(struct nfs_server *source,
 		(unsigned long long) server->fsid.major,
 		(unsigned long long) server->fsid.minor);
 
-	error = nfs_start_lockd(server);
+	error = nfs_start_lockd(server, server->nfs_client->cl_rpcmount);
 	if (error < 0)
 		goto out_free_server;
 
diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h
index bfa3a34..b9554d7 100644
--- a/fs/nfs/internal.h
+++ b/fs/nfs/internal.h
@@ -86,6 +86,7 @@ struct nfs_parsed_mount_data {
 	unsigned int		version;
 	unsigned int		minorversion;
 	char			*fscache_uniq;
+	char			*rpcmount;
 
 	struct {
 		struct sockaddr_storage	address;
@@ -120,6 +121,7 @@ struct nfs_mount_request {
 	int			noresvport;
 	unsigned int		*auth_flav_len;
 	rpc_authflavor_t	*auth_flavs;
+	struct vfsmount		*rpcmount;
 };
 
 extern int nfs_mount(struct nfs_mount_request *info);
@@ -164,10 +166,14 @@ static inline void nfs_fs_proc_exit(void)
 
 /* nfs4namespace.c */
 #ifdef CONFIG_NFS_V4
-extern struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry);
+extern struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent,
+		struct dentry *dentry,
+		struct vfsmount *rpcmount);
 #else
 static inline
-struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry)
+struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent,
+		struct dentry *dentry,
+		struct vfsmount *rpcmount)
 {
 	return ERR_PTR(-ENOENT);
 }
diff --git a/fs/nfs/mount_clnt.c b/fs/nfs/mount_clnt.c
index 6227875..205cc02 100644
--- a/fs/nfs/mount_clnt.c
+++ b/fs/nfs/mount_clnt.c
@@ -13,7 +13,6 @@
 #include <linux/in.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/sched.h>
-#include <linux/sunrpc/rpc_pipe_fs.h>
 #include <linux/nfs_fs.h>
 #include "internal.h"
 
@@ -162,7 +161,7 @@ int nfs_mount(struct nfs_mount_request *info)
 		.program	= &mnt_program,
 		.version	= info->version,
 		.authflavor	= RPC_AUTH_UNIX,
-		.rpcmount	= init_rpc_pipefs,
+		.rpcmount	= info->rpcmount,
 	};
 	struct rpc_clnt		*mnt_clnt;
 	int			status;
diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c
index 74aaf39..dad0129 100644
--- a/fs/nfs/namespace.c
+++ b/fs/nfs/namespace.c
@@ -144,7 +144,8 @@ static void * nfs_follow_mountpoint(struct dentry *dentry, struct nameidata *nd)
 		goto out_err;
 
 	if (fattr->valid & NFS_ATTR_FATTR_V4_REFERRAL)
-		mnt = nfs_do_refmount(nd->path.mnt, nd->path.dentry);
+		mnt = nfs_do_refmount(nd->path.mnt, nd->path.dentry,
+				server->nfs_client->cl_rpcmount);
 	else
 		mnt = nfs_do_submount(nd->path.mnt, nd->path.dentry, fh,
 				      fattr);
diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c
index 7a61fdb..92d5d63 100644
--- a/fs/nfs/nfs4namespace.c
+++ b/fs/nfs/nfs4namespace.c
@@ -14,7 +14,6 @@
 #include <linux/slab.h>
 #include <linux/string.h>
 #include <linux/sunrpc/clnt.h>
-#include <linux/sunrpc/rpc_pipe_fs.h>
 #include <linux/vfs.h>
 #include <linux/inet.h>
 #include "internal.h"
@@ -99,14 +98,13 @@ static int nfs4_validate_fspath(const struct vfsmount *mnt_parent,
 }
 
 static size_t nfs_parse_server_name(char *string, size_t len,
-		struct sockaddr *sa, size_t salen)
+		struct sockaddr *sa, size_t salen, struct vfsmount *rpcmount)
 {
 	ssize_t ret;
 
 	ret = rpc_pton(string, len, sa, salen);
 	if (ret == 0) {
-		ret = nfs_dns_resolve_name(string, len, sa, salen,
-				init_rpc_pipefs);
+		ret = nfs_dns_resolve_name(string, len, sa, salen, rpcmount);
 		if (ret < 0)
 			ret = 0;
 	}
@@ -115,7 +113,8 @@ static size_t nfs_parse_server_name(char *string, size_t len,
 
 static struct vfsmount *try_location(struct nfs_clone_mount *mountdata,
 				     char *page, char *page2,
-				     const struct nfs4_fs_location *location)
+				     const struct nfs4_fs_location *location,
+				     struct vfsmount *rpcmount)
 {
 	const size_t addr_bufsize = sizeof(struct sockaddr_storage);
 	struct vfsmount *mnt = ERR_PTR(-ENOENT);
@@ -143,7 +142,7 @@ static struct vfsmount *try_location(struct nfs_clone_mount *mountdata,
 			continue;
 
 		mountdata->addrlen = nfs_parse_server_name(buf->data, buf->len,
-				mountdata->addr, addr_bufsize);
+				mountdata->addr, addr_bufsize, rpcmount);
 		if (mountdata->addrlen == 0)
 			continue;
 
@@ -174,7 +173,8 @@ static struct vfsmount *try_location(struct nfs_clone_mount *mountdata,
  */
 static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
 					    const struct dentry *dentry,
-					    const struct nfs4_fs_locations *locations)
+					    const struct nfs4_fs_locations *locations,
+					    struct vfsmount *rpcmount)
 {
 	struct vfsmount *mnt = ERR_PTR(-ENOENT);
 	struct nfs_clone_mount mountdata = {
@@ -213,7 +213,7 @@ static struct vfsmount *nfs_follow_referral(const struct vfsmount *mnt_parent,
 		    location->rootpath.ncomponents == 0)
 			continue;
 
-		mnt = try_location(&mountdata, page, page2, location);
+		mnt = try_location(&mountdata, page, page2, location, rpcmount);
 		if (!IS_ERR(mnt))
 			break;
 	}
@@ -231,7 +231,9 @@ out:
  * @dentry - dentry of referral
  *
  */
-struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentry *dentry)
+struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent,
+		struct dentry *dentry,
+		struct vfsmount *rpcmount)
 {
 	struct vfsmount *mnt = ERR_PTR(-ENOMEM);
 	struct dentry *parent;
@@ -264,7 +266,7 @@ struct vfsmount *nfs_do_refmount(const struct vfsmount *mnt_parent, struct dentr
 	    fs_locations->fs_path.ncomponents <= 0)
 		goto out_free;
 
-	mnt = nfs_follow_referral(mnt_parent, dentry, fs_locations);
+	mnt = nfs_follow_referral(mnt_parent, dentry, fs_locations, rpcmount);
 out_free:
 	__free_page(page);
 	kfree(fs_locations);
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index b68c860..be4852b 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -35,6 +35,7 @@
 #include <linux/sunrpc/metrics.h>
 #include <linux/sunrpc/xprtsock.h>
 #include <linux/sunrpc/xprtrdma.h>
+#include <linux/sunrpc/rpc_pipe_fs.h>
 #include <linux/nfs_fs.h>
 #include <linux/nfs_mount.h>
 #include <linux/nfs4_mount.h>
@@ -106,6 +107,7 @@ enum {
 	Opt_lookupcache,
 	Opt_fscache_uniq,
 	Opt_local_lock,
+	Opt_rpcmount,
 
 	/* Special mount options */
 	Opt_userspace, Opt_deprecated, Opt_sloppy,
@@ -178,6 +180,7 @@ static const match_table_t nfs_mount_option_tokens = {
 	{ Opt_lookupcache, "lookupcache=%s" },
 	{ Opt_fscache_uniq, "fsc=%s" },
 	{ Opt_local_lock, "local_lock=%s" },
+	{ Opt_rpcmount, "rpcmount=%s" },
 
 	{ Opt_err, NULL }
 };
@@ -1486,6 +1489,13 @@ static int nfs_parse_mount_options(char *raw,
 				return 0;
 			};
 			break;
+		case Opt_rpcmount:
+			string = match_strdup(args);
+			if (string == NULL)
+				goto out_nomem;
+			kfree(mnt->rpcmount);
+			mnt->rpcmount = string;
+			break;
 
 		/*
 		 * Special options
@@ -1646,11 +1656,19 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args,
 	request.salen = args->mount_server.addrlen;
 	nfs_set_port(request.sap, &args->mount_server.port, 0);
 
+	request.rpcmount = get_rpc_pipefs(args->rpcmount);
+	if (IS_ERR(request.rpcmount)) {
+		dfprintk(MOUNT,	"NFS: unable get rpc_pipefs mount point, "
+				"error %ld\n", PTR_ERR(request.rpcmount));
+		return PTR_ERR(request.rpcmount);
+	}
+
 	/*
 	 * Now ask the mount server to map our export path
 	 * to a file handle.
 	 */
 	status = nfs_mount(&request);
+	mntput(request.rpcmount);
 	if (status != 0) {
 		dfprintk(MOUNT, "NFS: unable to mount server %s, error %d\n",
 				request.hostname, status);
@@ -2355,6 +2373,7 @@ out:
 	kfree(data->nfs_server.hostname);
 	kfree(data->mount_server.hostname);
 	kfree(data->fscache_uniq);
+	kfree(data->rpcmount);
 	security_free_mnt_opts(&data->lsm_opts);
 out_free_fh:
 	nfs_free_fhandle(mntfh);
@@ -2962,6 +2981,7 @@ out:
 	kfree(data->nfs_server.export_path);
 	kfree(data->nfs_server.hostname);
 	kfree(data->fscache_uniq);
+	kfree(data->rpcmount);
 out_free_data:
 	kfree(data);
 	dprintk("<-- nfs4_get_sb() = %d%s\n", error,
diff --git a/include/linux/nfs_fs_sb.h b/include/linux/nfs_fs_sb.h
index b197563..ad8d913 100644
--- a/include/linux/nfs_fs_sb.h
+++ b/include/linux/nfs_fs_sb.h
@@ -36,6 +36,7 @@ struct nfs_client {
 	struct list_head	cl_share_link;	/* link in global client list */
 	struct list_head	cl_superblocks;	/* List of nfs_server structs */
 
+	struct vfsmount		*cl_rpcmount;	/* rpc_pipefs mount point */
 	struct rpc_clnt *	cl_rpcclient;
 	const struct nfs_rpc_ops *rpc_ops;	/* NFS protocol vector */
 	int			cl_proto;	/* Network transport protocol */
-- 
1.7.3.4


^ permalink raw reply related

* [PATCH v3 05/16] sunrpc: get rpc_pipefs mount point for svc_serv from callers
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs, David S. Miller, Rob Landley, Al Viro,
	containers, netdev, linux-kernel, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas@openvz.org>

Signed-off-by: Kirill A. Shutemov <kas@openvz.org>
---
 fs/lockd/svc.c             |    4 +++-
 fs/nfs/callback.c          |    4 +++-
 fs/nfsd/nfssvc.c           |    6 ++++--
 include/linux/sunrpc/svc.h |    8 ++++----
 net/sunrpc/svc.c           |   18 +++++++++---------
 5 files changed, 23 insertions(+), 17 deletions(-)

diff --git a/fs/lockd/svc.c b/fs/lockd/svc.c
index abfff9d..32310b1 100644
--- a/fs/lockd/svc.c
+++ b/fs/lockd/svc.c
@@ -31,6 +31,7 @@
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/svc.h>
 #include <linux/sunrpc/svcsock.h>
+#include <linux/sunrpc/rpc_pipe_fs.h>
 #include <net/ip.h>
 #include <linux/lockd/lockd.h>
 #include <linux/nfs.h>
@@ -269,7 +270,8 @@ int lockd_up(void)
 			"lockd_up: no pid, %d users??\n", nlmsvc_users);
 
 	error = -ENOMEM;
-	serv = svc_create(&nlmsvc_program, LOCKD_BUFSIZE, NULL);
+	serv = svc_create(&nlmsvc_program, init_rpc_pipefs, LOCKD_BUFSIZE,
+			NULL);
 	if (!serv) {
 		printk(KERN_WARNING "lockd_up: create service failed\n");
 		goto out;
diff --git a/fs/nfs/callback.c b/fs/nfs/callback.c
index 1990165..7a535c8 100644
--- a/fs/nfs/callback.c
+++ b/fs/nfs/callback.c
@@ -16,6 +16,7 @@
 #include <linux/freezer.h>
 #include <linux/kthread.h>
 #include <linux/sunrpc/svcauth_gss.h>
+#include <linux/sunrpc/rpc_pipe_fs.h>
 #include <linux/sunrpc/bc_xprt.h>
 
 #include <net/inet_sock.h>
@@ -290,7 +291,8 @@ int nfs_callback_up(u32 minorversion, struct rpc_xprt *xprt)
 		nfs_callback_bc_serv(minorversion, xprt, cb_info);
 		goto out;
 	}
-	serv = svc_create(&nfs4_callback_program, NFS4_CALLBACK_BUFSIZE, NULL);
+	serv = svc_create(&nfs4_callback_program, init_rpc_pipefs,
+			NFS4_CALLBACK_BUFSIZE, NULL);
 	if (!serv) {
 		ret = -ENOMEM;
 		goto out_err;
diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index 2bae1d8..d96c32b 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -13,6 +13,7 @@
 
 #include <linux/sunrpc/stats.h>
 #include <linux/sunrpc/svcsock.h>
+#include <linux/sunrpc/rpc_pipe_fs.h>
 #include <linux/lockd/bind.h>
 #include <linux/nfsacl.h>
 #include <linux/seq_file.h>
@@ -331,8 +332,9 @@ int nfsd_create_serv(void)
 	}
 	nfsd_reset_versions();
 
-	nfsd_serv = svc_create_pooled(&nfsd_program, nfsd_max_blksize,
-				      nfsd_last_thread, nfsd, THIS_MODULE);
+	nfsd_serv = svc_create_pooled(&nfsd_program, init_rpc_pipefs,
+				      nfsd_max_blksize, nfsd_last_thread, nfsd,
+				      THIS_MODULE);
 	if (nfsd_serv == NULL)
 		return -ENOMEM;
 
diff --git a/include/linux/sunrpc/svc.h b/include/linux/sunrpc/svc.h
index 534ea8e..ad30e5d 100644
--- a/include/linux/sunrpc/svc.h
+++ b/include/linux/sunrpc/svc.h
@@ -400,13 +400,13 @@ struct svc_procedure {
 /*
  * Function prototypes.
  */
-struct svc_serv *svc_create(struct svc_program *, unsigned int,
-			    void (*shutdown)(struct svc_serv *));
+struct svc_serv *svc_create(struct svc_program *, struct vfsmount *,
+			    unsigned int, void (*shutdown)(struct svc_serv *));
 struct svc_rqst *svc_prepare_thread(struct svc_serv *serv,
 					struct svc_pool *pool);
 void		   svc_exit_thread(struct svc_rqst *);
-struct svc_serv *  svc_create_pooled(struct svc_program *, unsigned int,
-			void (*shutdown)(struct svc_serv *),
+struct svc_serv *  svc_create_pooled(struct svc_program *, struct vfsmount *,
+			unsigned int, void (*shutdown)(struct svc_serv *),
 			svc_thread_fn, struct module *);
 int		   svc_set_num_threads(struct svc_serv *, struct svc_pool *, int);
 int		   svc_pool_stats_open(struct svc_serv *serv, struct file *file);
diff --git a/net/sunrpc/svc.c b/net/sunrpc/svc.c
index 8cc6e79..8472798 100644
--- a/net/sunrpc/svc.c
+++ b/net/sunrpc/svc.c
@@ -28,7 +28,6 @@
 #include <linux/sunrpc/svcsock.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/bc_xprt.h>
-#include <linux/sunrpc/rpc_pipe_fs.h>
 
 #define RPCDBG_FACILITY	RPCDBG_SVCDSP
 
@@ -361,7 +360,8 @@ svc_pool_for_cpu(struct svc_serv *serv, int cpu)
  * Create an RPC service
  */
 static struct svc_serv *
-__svc_create(struct svc_program *prog, unsigned int bufsize, int npools,
+__svc_create(struct svc_program *prog, struct vfsmount *rpcmount,
+	     unsigned int bufsize, int npools,
 	     void (*shutdown)(struct svc_serv *serv))
 {
 	struct svc_serv	*serv;
@@ -373,7 +373,7 @@ __svc_create(struct svc_program *prog, unsigned int bufsize, int npools,
 		return NULL;
 	serv->sv_name      = prog->pg_name;
 	serv->sv_program   = prog;
-	serv->sv_rpcmount  = mntget(init_rpc_pipefs);
+	serv->sv_rpcmount  = mntget(rpcmount);
 	serv->sv_nrthreads = 1;
 	serv->sv_stats     = prog->pg_stats;
 	if (bufsize > RPCSVC_MAXPAYLOAD)
@@ -429,22 +429,22 @@ __svc_create(struct svc_program *prog, unsigned int bufsize, int npools,
 }
 
 struct svc_serv *
-svc_create(struct svc_program *prog, unsigned int bufsize,
-	   void (*shutdown)(struct svc_serv *serv))
+svc_create(struct svc_program *prog, struct vfsmount *rpcmount,
+	   unsigned int bufsize, void (*shutdown)(struct svc_serv *serv))
 {
-	return __svc_create(prog, bufsize, /*npools*/1, shutdown);
+	return __svc_create(prog, rpcmount, bufsize, /*npools*/1, shutdown);
 }
 EXPORT_SYMBOL_GPL(svc_create);
 
 struct svc_serv *
-svc_create_pooled(struct svc_program *prog, unsigned int bufsize,
-		  void (*shutdown)(struct svc_serv *serv),
+svc_create_pooled(struct svc_program *prog, struct vfsmount *rpcmount,
+		  unsigned int bufsize, void (*shutdown)(struct svc_serv *serv),
 		  svc_thread_fn func, struct module *mod)
 {
 	struct svc_serv *serv;
 	unsigned int npools = svc_pool_map_get();
 
-	serv = __svc_create(prog, bufsize, npools, shutdown);
+	serv = __svc_create(prog, rpcmount, bufsize, npools, shutdown);
 
 	if (serv != NULL) {
 		serv->sv_function = func;
-- 
1.7.3.4


^ permalink raw reply related

* [PATCH v3 02/16] sunrpc: introduce init_rpc_pipefs
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs, David S. Miller, Rob Landley, Al Viro,
	containers, netdev, linux-kernel, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas@openvz.org>

Introduce global variable init_rpc_pipefs and use it instead of
rpc_get_mount()/rpc_put_mount().

Signed-off-by: Kirill A. Shutemov <kas@openvz.org>
---
 fs/nfs/cache_lib.c                 |    6 +++---
 include/linux/sunrpc/rpc_pipe_fs.h |    4 ++--
 net/sunrpc/clnt.c                  |   10 ++++------
 net/sunrpc/rpc_pipe.c              |   21 +++++----------------
 4 files changed, 14 insertions(+), 27 deletions(-)

diff --git a/fs/nfs/cache_lib.c b/fs/nfs/cache_lib.c
index 8469031..dd7ca5f 100644
--- a/fs/nfs/cache_lib.c
+++ b/fs/nfs/cache_lib.c
@@ -117,7 +117,7 @@ int nfs_cache_register(struct cache_detail *cd)
 	struct vfsmount *mnt;
 	int ret;
 
-	mnt = rpc_get_mount();
+	mnt = mntget(init_rpc_pipefs);
 	if (IS_ERR(mnt))
 		return PTR_ERR(mnt);
 	ret = vfs_path_lookup(mnt->mnt_root, mnt, "/cache", 0, &nd);
@@ -129,13 +129,13 @@ int nfs_cache_register(struct cache_detail *cd)
 	if (!ret)
 		return ret;
 err:
-	rpc_put_mount();
+	mntput(mnt);
 	return ret;
 }
 
 void nfs_cache_unregister(struct cache_detail *cd)
 {
 	sunrpc_cache_unregister_pipefs(cd);
-	rpc_put_mount();
+	mntput(init_rpc_pipefs);
 }
 
diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h
index cf14db9..b09bfa5 100644
--- a/include/linux/sunrpc/rpc_pipe_fs.h
+++ b/include/linux/sunrpc/rpc_pipe_fs.h
@@ -44,6 +44,8 @@ RPC_I(struct inode *inode)
 	return container_of(inode, struct rpc_inode, vfs_inode);
 }
 
+extern struct vfsmount *init_rpc_pipefs;
+
 extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *);
 
 struct rpc_clnt;
@@ -60,8 +62,6 @@ extern void rpc_remove_cache_dir(struct dentry *);
 extern struct dentry *rpc_mkpipe(struct dentry *, const char *, void *,
 				 const struct rpc_pipe_ops *, int flags);
 extern int rpc_unlink(struct dentry *);
-extern struct vfsmount *rpc_get_mount(void);
-extern void rpc_put_mount(void);
 extern int register_rpc_pipefs(void);
 extern void unregister_rpc_pipefs(void);
 
diff --git a/net/sunrpc/clnt.c b/net/sunrpc/clnt.c
index 57d344c..f3812d0 100644
--- a/net/sunrpc/clnt.c
+++ b/net/sunrpc/clnt.c
@@ -112,9 +112,7 @@ rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name)
 	if (dir_name == NULL)
 		return 0;
 
-	path.mnt = rpc_get_mount();
-	if (IS_ERR(path.mnt))
-		return PTR_ERR(path.mnt);
+	path.mnt = mntget(init_rpc_pipefs);
 	error = vfs_path_lookup(path.mnt->mnt_root, path.mnt, dir_name, 0, &nd);
 	if (error)
 		goto err;
@@ -140,7 +138,7 @@ rpc_setup_pipedir(struct rpc_clnt *clnt, char *dir_name)
 err_path_put:
 	path_put(&nd.path);
 err:
-	rpc_put_mount();
+	mntput(path.mnt);
 	return error;
 }
 
@@ -251,7 +249,7 @@ static struct rpc_clnt * rpc_new_client(const struct rpc_create_args *args, stru
 out_no_auth:
 	if (!IS_ERR(clnt->cl_path.dentry)) {
 		rpc_remove_client_dir(clnt->cl_path.dentry);
-		rpc_put_mount();
+		mntput(clnt->cl_path.mnt);
 	}
 out_no_path:
 	kfree(clnt->cl_principal);
@@ -472,7 +470,7 @@ rpc_free_client(struct rpc_clnt *clnt)
 			clnt->cl_protname, clnt->cl_server);
 	if (!IS_ERR(clnt->cl_path.dentry)) {
 		rpc_remove_client_dir(clnt->cl_path.dentry);
-		rpc_put_mount();
+		mntput(clnt->cl_path.mnt);
 	}
 	if (clnt->cl_parent != clnt) {
 		rpc_release_client(clnt->cl_parent);
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index 9ab9355..484c9a3 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -28,7 +28,8 @@
 #include <linux/sunrpc/rpc_pipe_fs.h>
 #include <linux/sunrpc/cache.h>
 
-static struct vfsmount *rpc_mnt __read_mostly;
+struct vfsmount *init_rpc_pipefs __read_mostly;
+EXPORT_SYMBOL_GPL(init_rpc_pipefs);
 
 static struct file_system_type rpc_pipe_fs_type;
 
@@ -420,18 +421,6 @@ struct rpc_filelist {
 	umode_t mode;
 };
 
-struct vfsmount *rpc_get_mount(void)
-{
-	return mntget(rpc_mnt);
-}
-EXPORT_SYMBOL_GPL(rpc_get_mount);
-
-void rpc_put_mount(void)
-{
-	mntput(rpc_mnt);
-}
-EXPORT_SYMBOL_GPL(rpc_put_mount);
-
 static int rpc_delete_dentry(const struct dentry *dentry)
 {
 	return 1;
@@ -1068,9 +1057,9 @@ int register_rpc_pipefs(void)
 	if (err)
 		goto destroy_cache;
 
-	rpc_mnt = kern_mount(&rpc_pipe_fs_type);
-	if (IS_ERR(rpc_mnt)) {
-		err = PTR_ERR(rpc_mnt);
+	init_rpc_pipefs = kern_mount(&rpc_pipe_fs_type);
+	if (IS_ERR(init_rpc_pipefs)) {
+		err = PTR_ERR(init_rpc_pipefs);
 		goto unregister_fs;
 	}
 
-- 
1.7.3.4


^ permalink raw reply related

* [PATCH v3 01/16] sunrpc: mount rpc_pipefs on initialization
From: Kirill A. Shutemov @ 2011-01-14 13:48 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs, David S. Miller, Rob Landley, Al Viro,
	containers, netdev, linux-kernel, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas@openvz.org>

Mount rpc_pipefs on register_rpc_pipefs() and replace
rpc_get_mount()/rpc_put_mount() implementation with mntget()/mntput().

This commit introduces temprorary regression: there is no way to
remove the module. It will be fixed at the end of the patchset.

Signed-off-by: Kirill A. Shutemov <kas@openvz.org>
---
 net/sunrpc/rpc_pipe.c |   26 +++++++++++++++-----------
 1 files changed, 15 insertions(+), 11 deletions(-)

diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index 72bc536..9ab9355 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -29,7 +29,6 @@
 #include <linux/sunrpc/cache.h>
 
 static struct vfsmount *rpc_mnt __read_mostly;
-static int rpc_mount_count;
 
 static struct file_system_type rpc_pipe_fs_type;
 
@@ -423,18 +422,13 @@ struct rpc_filelist {
 
 struct vfsmount *rpc_get_mount(void)
 {
-	int err;
-
-	err = simple_pin_fs(&rpc_pipe_fs_type, &rpc_mnt, &rpc_mount_count);
-	if (err != 0)
-		return ERR_PTR(err);
-	return rpc_mnt;
+	return mntget(rpc_mnt);
 }
 EXPORT_SYMBOL_GPL(rpc_get_mount);
 
 void rpc_put_mount(void)
 {
-	simple_release_fs(&rpc_mnt, &rpc_mount_count);
+	mntput(rpc_mnt);
 }
 EXPORT_SYMBOL_GPL(rpc_put_mount);
 
@@ -1071,12 +1065,22 @@ int register_rpc_pipefs(void)
 	if (!rpc_inode_cachep)
 		return -ENOMEM;
 	err = register_filesystem(&rpc_pipe_fs_type);
-	if (err) {
-		kmem_cache_destroy(rpc_inode_cachep);
-		return err;
+	if (err)
+		goto destroy_cache;
+
+	rpc_mnt = kern_mount(&rpc_pipe_fs_type);
+	if (IS_ERR(rpc_mnt)) {
+		err = PTR_ERR(rpc_mnt);
+		goto unregister_fs;
 	}
 
 	return 0;
+
+unregister_fs:
+	unregister_filesystem(&rpc_pipe_fs_type);
+destroy_cache:
+	kmem_cache_destroy(rpc_inode_cachep);
+	return err;
 }
 
 void unregister_rpc_pipefs(void)
-- 
1.7.3.4


^ permalink raw reply related

* [PATCH v3 16/16] Rework get_rpc_pipefs and introduce put_rpc_pipefs()
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs-u79uwXL29TY76Z2rM5mHXA,
	David S. Miller, Rob Landley, Al Viro,
	containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA,
	netdev-u79uwXL29TY76Z2rM5mHXA,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>

Now sunrpc module can be removed normally.

Signed-off-by: Kirill A. Shutemov <kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>
---
 fs/nfs/client.c                    |    4 +-
 fs/nfs/super.c                     |    2 +-
 fs/nfsd/nfs4callback.c             |    2 +-
 fs/nfsd/nfssvc.c                   |    4 +-
 include/linux/sunrpc/rpc_pipe_fs.h |    1 +
 net/sunrpc/rpc_pipe.c              |   51 +++++++++++++++++++++++------------
 6 files changed, 40 insertions(+), 24 deletions(-)

diff --git a/fs/nfs/client.c b/fs/nfs/client.c
index 8dd0e99..0b8ac12 100644
--- a/fs/nfs/client.c
+++ b/fs/nfs/client.c
@@ -838,7 +838,7 @@ static int nfs_init_server(struct nfs_server *server,
 
 	/* Allocate or find a client reference we can use */
 	clp = nfs_get_client(&cl_init);
-	mntput(cl_init.rpcmount);
+	put_rpc_pipefs(cl_init.rpcmount);
 	if (IS_ERR(clp)) {
 		dprintk("<-- nfs_init_server() = error %ld\n", PTR_ERR(clp));
 		return PTR_ERR(clp);
@@ -1533,7 +1533,7 @@ static int nfs4_init_server(struct nfs_server *server,
 			&timeparms,
 			data->minorversion,
 			rpcmount);
-	mntput(rpcmount);
+	put_rpc_pipefs(rpcmount);
 	if (error < 0)
 		goto error;
 
diff --git a/fs/nfs/super.c b/fs/nfs/super.c
index be4852b..148843e 100644
--- a/fs/nfs/super.c
+++ b/fs/nfs/super.c
@@ -1668,7 +1668,7 @@ static int nfs_try_mount(struct nfs_parsed_mount_data *args,
 	 * to a file handle.
 	 */
 	status = nfs_mount(&request);
-	mntput(request.rpcmount);
+	put_rpc_pipefs(request.rpcmount);
 	if (status != 0) {
 		dfprintk(MOUNT, "NFS: unable to mount server %s, error %d\n",
 				request.hostname, status);
diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 61210b6..4d8383b 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -667,7 +667,7 @@ int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
 		return PTR_ERR(args.rpcmount);
 	/* Create RPC client */
 	client = rpc_create(&args);
-	mntput(args.rpcmount);
+	put_rpc_pipefs(args.rpcmount);
 	if (IS_ERR(client)) {
 		dprintk("NFSD: couldn't create callback client: %ld\n",
 			PTR_ERR(client));
diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index 7353420..cff49fc 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -227,7 +227,7 @@ static int nfsd_startup(unsigned short port, int nrservs)
 		goto out_racache;
 	}
 	ret = lockd_up(rpcmount);
-	mntput(rpcmount);
+	put_rpc_pipefs(rpcmount);
 	if (ret)
 		goto out_racache;
 	ret = nfs4_state_start();
@@ -346,7 +346,7 @@ int nfsd_create_serv(void)
 	nfsd_serv = svc_create_pooled(&nfsd_program, rpcmount,
 				      nfsd_max_blksize, nfsd_last_thread, nfsd,
 				      THIS_MODULE);
-	mntput(rpcmount);
+	put_rpc_pipefs(rpcmount);
 	if (nfsd_serv == NULL)
 		return -ENOMEM;
 
diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h
index a0b9c46..328b3da 100644
--- a/include/linux/sunrpc/rpc_pipe_fs.h
+++ b/include/linux/sunrpc/rpc_pipe_fs.h
@@ -45,6 +45,7 @@ RPC_I(struct inode *inode)
 }
 
 extern struct vfsmount *get_rpc_pipefs(const char *path);
+extern void put_rpc_pipefs(struct vfsmount *rpcmount);
 extern int rpc_pipefs_add_destroy_cb(struct super_block *sb,
 		void (*destroy_cb)(void *data), void *data);
 
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index 96973c0..82feacd 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -33,6 +33,7 @@
 #include <linux/sunrpc/cache.h>
 
 struct vfsmount *init_rpc_pipefs __read_mostly;
+static int init_rpc_pipefs_count;
 
 static struct file_system_type rpc_pipe_fs_type;
 
@@ -1005,22 +1006,35 @@ static int check_rpc_pipefs(struct vfsmount *mnt, void *arg)
 	return 1;
 }
 
-struct vfsmount *get_rpc_pipefs(const char *p)
+static struct vfsmount *find_rpc_pipefs(void)
 {
-	int error;
 	struct vfsmount *rpcmount = ERR_PTR(-EINVAL);
-	struct path path;
+	int err;
 
-	if (!p) {
-		iterate_mounts(check_rpc_pipefs, &rpcmount,
-				current->nsproxy->mnt_ns->root);
+	iterate_mounts(check_rpc_pipefs, &rpcmount,
+			current->nsproxy->mnt_ns->root);
 
-		if (IS_ERR(rpcmount) && (current->nsproxy->mnt_ns ==
-					init_task.nsproxy->mnt_ns))
-			return mntget(init_rpc_pipefs);
+	if (!IS_ERR(rpcmount))
+		return rpcmount;
 
+	if (current->nsproxy->mnt_ns != init_task.nsproxy->mnt_ns)
 		return rpcmount;
-	}
+
+	err = simple_pin_fs(&rpc_pipe_fs_type, &init_rpc_pipefs,
+			&init_rpc_pipefs_count);
+	if (err)
+		return ERR_PTR(err);
+	return init_rpc_pipefs;
+}
+
+struct vfsmount *get_rpc_pipefs(const char *p)
+{
+	int error;
+	struct vfsmount *rpcmount = ERR_PTR(-EINVAL);
+	struct path path;
+
+	if (!p)
+		return find_rpc_pipefs();
 
 	error = kern_path(p, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path);
 	if (error)
@@ -1033,6 +1047,15 @@ struct vfsmount *get_rpc_pipefs(const char *p)
 }
 EXPORT_SYMBOL_GPL(get_rpc_pipefs);
 
+void put_rpc_pipefs(struct vfsmount *rpcmount)
+{
+	if (rpcmount == init_rpc_pipefs)
+		simple_release_fs(&init_rpc_pipefs, &init_rpc_pipefs_count);
+	else
+		mntput(rpcmount);
+}
+EXPORT_SYMBOL_GPL(put_rpc_pipefs);
+
 struct destroy_cb {
 	struct list_head list;
 	void (*callback)(void *data);
@@ -1232,16 +1255,8 @@ int register_rpc_pipefs(void)
 	if (err)
 		goto destroy_cache;
 
-	init_rpc_pipefs = kern_mount(&rpc_pipe_fs_type);
-	if (IS_ERR(init_rpc_pipefs)) {
-		err = PTR_ERR(init_rpc_pipefs);
-		goto unregister_fs;
-	}
-
 	return 0;
 
-unregister_fs:
-	unregister_filesystem(&rpc_pipe_fs_type);
 destroy_cache:
 	kmem_cache_destroy(rpc_inode_cachep);
 	return err;
-- 
1.7.3.4

--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply related

* [PATCH v3 15/16] sunrpc: remove global init_rpc_pipefs
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs, David S. Miller, Rob Landley, Al Viro,
	containers, netdev, linux-kernel, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas@openvz.org>

Replace remaining init_rpc_pipefs references with get_rpc_pipefs(NULL)
and make init_rpc_pipefs static.

Signed-off-by: Kirill A. Shutemov <kas@openvz.org>
---
 fs/nfsd/nfs4callback.c             |    5 ++++-
 fs/nfsd/nfssvc.c                   |   16 ++++++++++++++--
 include/linux/sunrpc/rpc_pipe_fs.h |    2 --
 net/sunrpc/rpc_pipe.c              |    1 -
 4 files changed, 18 insertions(+), 6 deletions(-)

diff --git a/fs/nfsd/nfs4callback.c b/fs/nfsd/nfs4callback.c
index 9bec643..61210b6 100644
--- a/fs/nfsd/nfs4callback.c
+++ b/fs/nfsd/nfs4callback.c
@@ -647,7 +647,6 @@ int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
 		.version	= 0,
 		.authflavor	= clp->cl_flavor,
 		.flags		= (RPC_CLNT_CREATE_NOPING | RPC_CLNT_CREATE_QUIET),
-		.rpcmount	= init_rpc_pipefs,
 	};
 	struct rpc_clnt *client;
 
@@ -663,8 +662,12 @@ int setup_callback_client(struct nfs4_client *clp, struct nfs4_cb_conn *conn)
 		args.prognumber = clp->cl_cb_session->se_cb_prog;
 		args.protocol = XPRT_TRANSPORT_BC_TCP;
 	}
+	args.rpcmount = get_rpc_pipefs(NULL);
+	if (IS_ERR(args.rpcmount))
+		return PTR_ERR(args.rpcmount);
 	/* Create RPC client */
 	client = rpc_create(&args);
+	mntput(args.rpcmount);
 	if (IS_ERR(client)) {
 		dprintk("NFSD: couldn't create callback client: %ld\n",
 			PTR_ERR(client));
diff --git a/fs/nfsd/nfssvc.c b/fs/nfsd/nfssvc.c
index 17d78d3..7353420 100644
--- a/fs/nfsd/nfssvc.c
+++ b/fs/nfsd/nfssvc.c
@@ -206,6 +206,7 @@ static bool nfsd_up = false;
 static int nfsd_startup(unsigned short port, int nrservs)
 {
 	int ret;
+	struct vfsmount *rpcmount;
 
 	if (nfsd_up)
 		return 0;
@@ -220,7 +221,13 @@ static int nfsd_startup(unsigned short port, int nrservs)
 	ret = nfsd_init_socks(port);
 	if (ret)
 		goto out_racache;
-	ret = lockd_up(init_rpc_pipefs);
+	rpcmount = get_rpc_pipefs(NULL);
+	if (IS_ERR(rpcmount)) {
+		ret = PTR_ERR(rpcmount);
+		goto out_racache;
+	}
+	ret = lockd_up(rpcmount);
+	mntput(rpcmount);
 	if (ret)
 		goto out_racache;
 	ret = nfs4_state_start();
@@ -308,6 +315,7 @@ static void set_max_drc(void)
 int nfsd_create_serv(void)
 {
 	int err = 0;
+	struct vfsmount *rpcmount;
 
 	WARN_ON(!mutex_is_locked(&nfsd_mutex));
 	if (nfsd_serv) {
@@ -332,9 +340,13 @@ int nfsd_create_serv(void)
 	}
 	nfsd_reset_versions();
 
-	nfsd_serv = svc_create_pooled(&nfsd_program, init_rpc_pipefs,
+	rpcmount = get_rpc_pipefs(NULL);
+	if (IS_ERR(rpcmount))
+		return PTR_ERR(rpcmount);
+	nfsd_serv = svc_create_pooled(&nfsd_program, rpcmount,
 				      nfsd_max_blksize, nfsd_last_thread, nfsd,
 				      THIS_MODULE);
+	mntput(rpcmount);
 	if (nfsd_serv == NULL)
 		return -ENOMEM;
 
diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h
index 4a8830a..a0b9c46 100644
--- a/include/linux/sunrpc/rpc_pipe_fs.h
+++ b/include/linux/sunrpc/rpc_pipe_fs.h
@@ -44,8 +44,6 @@ RPC_I(struct inode *inode)
 	return container_of(inode, struct rpc_inode, vfs_inode);
 }
 
-extern struct vfsmount *init_rpc_pipefs;
-
 extern struct vfsmount *get_rpc_pipefs(const char *path);
 extern int rpc_pipefs_add_destroy_cb(struct super_block *sb,
 		void (*destroy_cb)(void *data), void *data);
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index 02416f1..96973c0 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -33,7 +33,6 @@
 #include <linux/sunrpc/cache.h>
 
 struct vfsmount *init_rpc_pipefs __read_mostly;
-EXPORT_SYMBOL_GPL(init_rpc_pipefs);
 
 static struct file_system_type rpc_pipe_fs_type;
 
-- 
1.7.3.4

^ permalink raw reply related

* [PATCH v3 14/16] sunrpc: make rpc_pipefs be mountable multiple times
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs-u79uwXL29TY76Z2rM5mHXA,
	David S. Miller, Rob Landley, Al Viro,
	containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA,
	netdev-u79uwXL29TY76Z2rM5mHXA,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>

To support containers, allow multiple independent instances of
rpc_pipefs. Use '-o newinstance' to create new of the filesystem.
The same semantics as with devpts.

Signed-off-by: Kirill A. Shutemov <kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>
---
 net/sunrpc/rpc_pipe.c |   80 ++++++++++++++++++++++++++++++++++++++++++++++++-
 1 files changed, 79 insertions(+), 1 deletions(-)

diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index b72cb01..02416f1 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -19,6 +19,7 @@
 #include <linux/nsproxy.h>
 #include <linux/mnt_namespace.h>
 #include <linux/fs_struct.h>
+#include <linux/parser.h>
 
 #include <asm/ioctls.h>
 #include <linux/fs.h>
@@ -41,6 +42,49 @@ static struct kmem_cache *rpc_inode_cachep __read_mostly;
 
 #define RPC_UPCALL_TIMEOUT (30*HZ)
 
+struct rpc_mount_opts {
+	int newinstance;
+};
+
+enum {
+	Opt_newinstance,
+
+	Opt_err
+};
+
+static const match_table_t tokens = {
+	{Opt_newinstance, "newinstance"},
+
+	{Opt_err, NULL}
+};
+
+static int
+parse_mount_options(char *data, struct rpc_mount_opts *opts)
+{
+	char *p;
+
+	opts->newinstance = 0;
+
+	while ((p = strsep(&data, ",")) != NULL) {
+		substring_t args[MAX_OPT_ARGS];
+		int token;
+
+		if (!*p)
+			continue;
+
+		token = match_token(p, tokens, args);
+		switch (token) {
+		case Opt_newinstance:
+			opts->newinstance = 1;
+			break;
+		default:
+			return -EINVAL;
+		}
+	}
+
+	return 0;
+}
+
 static void rpc_purge_list(struct rpc_inode *rpci, struct list_head *head,
 		void (*destroy_msg)(struct rpc_pipe_msg *), int err)
 {
@@ -1093,11 +1137,45 @@ rpc_fill_super(struct super_block *sb, void *data, int silent)
 	return 0;
 }
 
+static int
+compare_rpc_mnt_sb(struct super_block *s, void *p)
+{
+	if (init_rpc_pipefs)
+		return init_rpc_pipefs->mnt_sb == s;
+	return 0;
+}
+
 static struct dentry *
 rpc_mount(struct file_system_type *fs_type,
 		int flags, const char *dev_name, void *data)
 {
-	return mount_single(fs_type, flags, data, rpc_fill_super);
+	int error;
+	struct rpc_mount_opts opts;
+	struct super_block *s;
+
+	error = parse_mount_options(data, &opts);
+	if (error)
+		return ERR_PTR(error);
+
+	if (opts.newinstance)
+		s = sget(fs_type, NULL, set_anon_super, NULL);
+	else
+		s = sget(fs_type, compare_rpc_mnt_sb, set_anon_super, NULL);
+
+	if (IS_ERR(s))
+		return ERR_CAST(s);
+
+	if (!s->s_root) {
+		s->s_flags = flags;
+		error = rpc_fill_super(s, data, flags & MS_SILENT ? 1 : 0);
+		if (error) {
+			deactivate_locked_super(s);
+			return ERR_PTR(error);
+		}
+		s->s_flags |= MS_ACTIVE;
+	}
+
+	return dget(s->s_root);
 }
 
 static void rpc_kill_sb(struct super_block *sb)
-- 
1.7.3.4

--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply related

* [PATCH v3 12/16] sunrpc: introduce get_rpc_pipefs()
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs-u79uwXL29TY76Z2rM5mHXA,
	David S. Miller, Rob Landley, Al Viro,
	containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA,
	netdev-u79uwXL29TY76Z2rM5mHXA,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>

Get rpc_pipefs mount point by path. If path is NULL, we use first
rpc_pipefs from mount table visible from current root directory.

Signed-off-by: Kirill A. Shutemov <kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>
Acked-by: Rob Landley <rlandley-bzQdu9zFT3WakBO8gow8eQ@public.gmane.org>
---
 include/linux/sunrpc/rpc_pipe_fs.h |    1 +
 net/sunrpc/rpc_pipe.c              |   51 ++++++++++++++++++++++++++++++++++++
 2 files changed, 52 insertions(+), 0 deletions(-)

diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h
index f5216f1..4a8830a 100644
--- a/include/linux/sunrpc/rpc_pipe_fs.h
+++ b/include/linux/sunrpc/rpc_pipe_fs.h
@@ -46,6 +46,7 @@ RPC_I(struct inode *inode)
 
 extern struct vfsmount *init_rpc_pipefs;
 
+extern struct vfsmount *get_rpc_pipefs(const char *path);
 extern int rpc_pipefs_add_destroy_cb(struct super_block *sb,
 		void (*destroy_cb)(void *data), void *data);
 
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index 39511e3..b72cb01 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -16,6 +16,9 @@
 #include <linux/namei.h>
 #include <linux/fsnotify.h>
 #include <linux/kernel.h>
+#include <linux/nsproxy.h>
+#include <linux/mnt_namespace.h>
+#include <linux/fs_struct.h>
 
 #include <asm/ioctls.h>
 #include <linux/fs.h>
@@ -939,6 +942,54 @@ static const struct super_operations s_ops = {
 
 #define RPCAUTH_GSSMAGIC 0x67596969
 
+static int check_rpc_pipefs(struct vfsmount *mnt, void *arg)
+{
+	struct vfsmount **rpcmount = arg;
+	struct path path = {
+		.mnt = mnt,
+		.dentry = mnt->mnt_root,
+	};
+
+	if (!mnt->mnt_sb)
+		return 0;
+	if (mnt->mnt_sb->s_magic != RPCAUTH_GSSMAGIC)
+		return 0;
+
+	if (!path_is_under(&path, &current->fs->root))
+		return 0;
+
+	*rpcmount = mntget(mnt);
+	return 1;
+}
+
+struct vfsmount *get_rpc_pipefs(const char *p)
+{
+	int error;
+	struct vfsmount *rpcmount = ERR_PTR(-EINVAL);
+	struct path path;
+
+	if (!p) {
+		iterate_mounts(check_rpc_pipefs, &rpcmount,
+				current->nsproxy->mnt_ns->root);
+
+		if (IS_ERR(rpcmount) && (current->nsproxy->mnt_ns ==
+					init_task.nsproxy->mnt_ns))
+			return mntget(init_rpc_pipefs);
+
+		return rpcmount;
+	}
+
+	error = kern_path(p, LOOKUP_FOLLOW | LOOKUP_DIRECTORY, &path);
+	if (error)
+		return ERR_PTR(error);
+
+	check_rpc_pipefs(path.mnt, &rpcmount);
+	path_put(&path);
+
+	return rpcmount;
+}
+EXPORT_SYMBOL_GPL(get_rpc_pipefs);
+
 struct destroy_cb {
 	struct list_head list;
 	void (*callback)(void *data);
-- 
1.7.3.4

--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply related

* [PATCH v3 11/16] Export iterate_mounts symbol to be able to use from sunrpc module.
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs-u79uwXL29TY76Z2rM5mHXA,
	David S. Miller, Rob Landley, Al Viro,
	containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA,
	netdev-u79uwXL29TY76Z2rM5mHXA,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>

Signed-off-by: Kirill A. Shutemov <kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>
---
 fs/namespace.c |    1 +
 1 files changed, 1 insertions(+), 0 deletions(-)

diff --git a/fs/namespace.c b/fs/namespace.c
index 3ddfd90..f73054d 100644
--- a/fs/namespace.c
+++ b/fs/namespace.c
@@ -1498,6 +1498,7 @@ int iterate_mounts(int (*f)(struct vfsmount *, void *), void *arg,
 	}
 	return 0;
 }
+EXPORT_SYMBOL_GPL(iterate_mounts);
 
 static void cleanup_group_ids(struct vfsmount *mnt, struct vfsmount *end)
 {
-- 
1.7.3.4

--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply related

* [PATCH v3 10/16] nfs: per-rpc_pipefs dns cache
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs, David S. Miller, Rob Landley, Al Viro,
	containers, netdev, linux-kernel, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas@openvz.org>

Lazy initialization of dns cache: on first call nfs_dns_resolve_name().
Every rpc_pipefs has separate dns cache now.

Signed-off-by: Kirill A. Shutemov <kas@openvz.org>
---
 fs/nfs/cache_lib.c     |   17 ++-----
 fs/nfs/cache_lib.h     |    3 +-
 fs/nfs/dns_resolve.c   |  137 +++++++++++++++++++++++++++++++++++++-----------
 fs/nfs/dns_resolve.h   |   15 +-----
 fs/nfs/inode.c         |    9 +---
 fs/nfs/nfs4namespace.c |    4 +-
 6 files changed, 117 insertions(+), 68 deletions(-)

diff --git a/fs/nfs/cache_lib.c b/fs/nfs/cache_lib.c
index 0944d4e..9b99d9e 100644
--- a/fs/nfs/cache_lib.c
+++ b/fs/nfs/cache_lib.c
@@ -12,7 +12,6 @@
 #include <linux/namei.h>
 #include <linux/slab.h>
 #include <linux/sunrpc/cache.h>
-#include <linux/sunrpc/rpc_pipe_fs.h>
 
 #include "cache_lib.h"
 
@@ -111,25 +110,17 @@ int nfs_cache_wait_for_upcall(struct nfs_cache_defer_req *dreq)
 	return 0;
 }
 
-int nfs_cache_register(struct cache_detail *cd)
+int nfs_cache_register(struct cache_detail *cd, struct vfsmount *rpcmount)
 {
 	struct nameidata nd;
-	struct vfsmount *mnt;
 	int ret;
 
-	mnt = mntget(init_rpc_pipefs);
-	if (IS_ERR(mnt))
-		return PTR_ERR(mnt);
-	ret = vfs_path_lookup(mnt->mnt_root, mnt, "/cache", 0, &nd);
+	ret = vfs_path_lookup(rpcmount->mnt_root, rpcmount, "/cache", 0, &nd);
 	if (ret)
-		goto err;
-	ret = sunrpc_cache_register_pipefs(mnt, nd.path.dentry,
+		return ret;
+	ret = sunrpc_cache_register_pipefs(rpcmount, nd.path.dentry,
 			cd->name, 0600, cd);
 	path_put(&nd.path);
-	if (!ret)
-		return ret;
-err:
-	mntput(mnt);
 	return ret;
 }
 
diff --git a/fs/nfs/cache_lib.h b/fs/nfs/cache_lib.h
index 76f856e..1d4a0a5 100644
--- a/fs/nfs/cache_lib.h
+++ b/fs/nfs/cache_lib.h
@@ -23,5 +23,6 @@ extern struct nfs_cache_defer_req *nfs_cache_defer_req_alloc(void);
 extern void nfs_cache_defer_req_put(struct nfs_cache_defer_req *dreq);
 extern int nfs_cache_wait_for_upcall(struct nfs_cache_defer_req *dreq);
 
-extern int nfs_cache_register(struct cache_detail *cd);
+extern int nfs_cache_register(struct cache_detail *cd,
+		struct vfsmount *rpcmount);
 extern void nfs_cache_unregister(struct cache_detail *cd);
diff --git a/fs/nfs/dns_resolve.c b/fs/nfs/dns_resolve.c
index a6e711a..a832e64 100644
--- a/fs/nfs/dns_resolve.c
+++ b/fs/nfs/dns_resolve.c
@@ -12,7 +12,7 @@
 #include <linux/dns_resolver.h>
 
 ssize_t nfs_dns_resolve_name(char *name, size_t namelen,
-		struct sockaddr *sa, size_t salen)
+		struct sockaddr *sa, size_t salen, struct vfsmount *rpcmount)
 {
 	ssize_t ret;
 	char *ip_addr = NULL;
@@ -37,9 +37,11 @@ ssize_t nfs_dns_resolve_name(char *name, size_t namelen,
 #include <linux/socket.h>
 #include <linux/seq_file.h>
 #include <linux/inet.h>
+#include <linux/mount.h>
 #include <linux/sunrpc/clnt.h>
 #include <linux/sunrpc/cache.h>
 #include <linux/sunrpc/svcauth.h>
+#include <linux/sunrpc/rpc_pipe_fs.h>
 
 #include "dns_resolve.h"
 #include "cache_lib.h"
@@ -47,7 +49,13 @@ ssize_t nfs_dns_resolve_name(char *name, size_t namelen,
 #define NFS_DNS_HASHBITS 4
 #define NFS_DNS_HASHTBL_SIZE (1 << NFS_DNS_HASHBITS)
 
-static struct cache_head *nfs_dns_table[NFS_DNS_HASHTBL_SIZE];
+static DEFINE_SPINLOCK(nfs_dns_resolve_lock);
+static LIST_HEAD(nfs_dns_resolve_list);
+
+struct nfs_dns_resolve_list {
+	struct list_head list;
+	struct cache_detail *cd;
+};
 
 struct nfs_dns_ent {
 	struct cache_head h;
@@ -259,21 +267,6 @@ out:
 	return ret;
 }
 
-static struct cache_detail nfs_dns_resolve = {
-	.owner = THIS_MODULE,
-	.hash_size = NFS_DNS_HASHTBL_SIZE,
-	.hash_table = nfs_dns_table,
-	.name = "dns_resolve",
-	.cache_put = nfs_dns_ent_put,
-	.cache_upcall = nfs_dns_upcall,
-	.cache_parse = nfs_dns_parse,
-	.cache_show = nfs_dns_show,
-	.match = nfs_dns_match,
-	.init = nfs_dns_ent_init,
-	.update = nfs_dns_ent_update,
-	.alloc = nfs_dns_ent_alloc,
-};
-
 static int do_cache_lookup(struct cache_detail *cd,
 		struct nfs_dns_ent *key,
 		struct nfs_dns_ent **item,
@@ -336,37 +329,119 @@ out:
 	return ret;
 }
 
+static struct cache_detail *nfs_alloc_dns_resolve(void)
+{
+	struct cache_detail *dns_resolve;
+	struct cache_head **hash_table;
+
+	dns_resolve = kmalloc(sizeof(*dns_resolve), GFP_KERNEL);
+	if (!dns_resolve)
+		return NULL;
+
+	hash_table = kmalloc(sizeof(*hash_table) * NFS_DNS_HASHTBL_SIZE,
+			GFP_KERNEL);
+	if (!hash_table) {
+		kfree(dns_resolve);
+		return NULL;
+	}
+
+	dns_resolve->owner = THIS_MODULE;
+	dns_resolve->hash_size = NFS_DNS_HASHTBL_SIZE;
+	dns_resolve->hash_table = hash_table;
+	dns_resolve->name = "dns_resolve";
+	dns_resolve->cache_put = nfs_dns_ent_put;
+	dns_resolve->cache_upcall = nfs_dns_upcall;
+	dns_resolve->cache_parse = nfs_dns_parse;
+	dns_resolve->cache_show = nfs_dns_show;
+	dns_resolve->match = nfs_dns_match;
+	dns_resolve->init = nfs_dns_ent_init;
+	dns_resolve->update = nfs_dns_ent_update;
+	dns_resolve->alloc = nfs_dns_ent_alloc;
+
+	return dns_resolve;
+}
+
+static void nfs_free_dns_resolve(struct cache_detail *dns_resolve)
+{
+	kfree(dns_resolve->hash_table);
+	kfree(dns_resolve);
+}
+
+static struct cache_detail *nfs_get_dns_resolve(struct vfsmount *rpcmount)
+{
+	struct nfs_dns_resolve_list *dns_resolve;
+	int error = 0;
+
+	spin_lock(&nfs_dns_resolve_lock);
+	list_for_each_entry(dns_resolve, &nfs_dns_resolve_list, list) {
+		if (dns_resolve->cd->u.pipefs.mnt->mnt_sb != rpcmount->mnt_sb)
+			continue;
+
+		spin_unlock(&nfs_dns_resolve_lock);
+		return dns_resolve->cd;
+	}
+
+	dns_resolve = kmalloc(sizeof(*dns_resolve), GFP_KERNEL);
+	if (dns_resolve)
+		dns_resolve->cd = nfs_alloc_dns_resolve();
+	if (!dns_resolve || !dns_resolve->cd) {
+		error = -ENOMEM;
+		goto err;
+	}
+
+	error = nfs_cache_register(dns_resolve->cd, rpcmount);
+	if (error)
+		goto err;
+
+	INIT_LIST_HEAD(&dns_resolve->list);
+	list_add(&dns_resolve->list, &nfs_dns_resolve_list);
+	spin_unlock(&nfs_dns_resolve_lock);
+
+	return dns_resolve->cd;
+err:
+	spin_unlock(&nfs_dns_resolve_lock);
+	if (dns_resolve)
+		kfree(dns_resolve->cd);
+	kfree(dns_resolve);
+	return dns_resolve->cd;
+}
+
+static void nfs_dns_resolver_destroy(void *data)
+{
+	struct nfs_dns_resolve_list *dns_resolve = data;
+
+	spin_lock(&nfs_dns_resolve_lock);
+	nfs_cache_unregister(dns_resolve->cd);
+	nfs_free_dns_resolve(dns_resolve->cd);
+	list_del(&dns_resolve->list);
+	kfree(dns_resolve);
+	spin_unlock(&nfs_dns_resolve_lock);
+}
+
 ssize_t nfs_dns_resolve_name(char *name, size_t namelen,
-		struct sockaddr *sa, size_t salen)
+		struct sockaddr *sa, size_t salen, struct vfsmount *rpcmount)
 {
 	struct nfs_dns_ent key = {
 		.hostname = name,
 		.namelen = namelen,
 	};
+	struct cache_detail *dns_resolve;
 	struct nfs_dns_ent *item = NULL;
 	ssize_t ret;
 
-	ret = do_cache_lookup_wait(&nfs_dns_resolve, &key, &item);
+	dns_resolve = nfs_get_dns_resolve(rpcmount);
+	ret = do_cache_lookup_wait(dns_resolve, &key, &item);
 	if (ret == 0) {
 		if (salen >= item->addrlen) {
 			memcpy(sa, &item->addr, item->addrlen);
 			ret = item->addrlen;
 		} else
 			ret = -EOVERFLOW;
-		cache_put(&item->h, &nfs_dns_resolve);
+		cache_put(&item->h, dns_resolve);
+		rpc_pipefs_add_destroy_cb(rpcmount->mnt_sb,
+				nfs_dns_resolver_destroy, dns_resolve);
 	} else if (ret == -ENOENT)
 		ret = -ESRCH;
 	return ret;
 }
-
-int nfs_dns_resolver_init(void)
-{
-	return nfs_cache_register(&nfs_dns_resolve);
-}
-
-void nfs_dns_resolver_destroy(void)
-{
-	nfs_cache_unregister(&nfs_dns_resolve);
-}
-
 #endif
diff --git a/fs/nfs/dns_resolve.h b/fs/nfs/dns_resolve.h
index 199bb55..74ade60 100644
--- a/fs/nfs/dns_resolve.h
+++ b/fs/nfs/dns_resolve.h
@@ -7,20 +7,7 @@
 #define NFS_DNS_HOSTNAME_MAXLEN	(128)
 
 
-#ifdef CONFIG_NFS_USE_KERNEL_DNS
-static inline int nfs_dns_resolver_init(void)
-{
-	return 0;
-}
-
-static inline void nfs_dns_resolver_destroy(void)
-{}
-#else
-extern int nfs_dns_resolver_init(void);
-extern void nfs_dns_resolver_destroy(void);
-#endif
-
 extern ssize_t nfs_dns_resolve_name(char *name, size_t namelen,
-		struct sockaddr *sa, size_t salen);
+		struct sockaddr *sa, size_t salen, struct vfsmount *rpcmount);
 
 #endif
diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c
index ce00b70..414570b 100644
--- a/fs/nfs/inode.c
+++ b/fs/nfs/inode.c
@@ -1535,10 +1535,6 @@ static int __init init_nfs_fs(void)
 
 	err = nfs_idmap_init();
 	if (err < 0)
-		goto out9;
-
-	err = nfs_dns_resolver_init();
-	if (err < 0)
 		goto out8;
 
 	err = nfs_fscache_register();
@@ -1599,10 +1595,8 @@ out5:
 out6:
 	nfs_fscache_unregister();
 out7:
-	nfs_dns_resolver_destroy();
-out8:
 	nfs_idmap_quit();
-out9:
+out8:
 	return err;
 }
 
@@ -1614,7 +1608,6 @@ static void __exit exit_nfs_fs(void)
 	nfs_destroy_inodecache();
 	nfs_destroy_nfspagecache();
 	nfs_fscache_unregister();
-	nfs_dns_resolver_destroy();
 	nfs_idmap_quit();
 #ifdef CONFIG_PROC_FS
 	rpc_proc_unregister("nfs");
diff --git a/fs/nfs/nfs4namespace.c b/fs/nfs/nfs4namespace.c
index 3c2a172..7a61fdb 100644
--- a/fs/nfs/nfs4namespace.c
+++ b/fs/nfs/nfs4namespace.c
@@ -14,6 +14,7 @@
 #include <linux/slab.h>
 #include <linux/string.h>
 #include <linux/sunrpc/clnt.h>
+#include <linux/sunrpc/rpc_pipe_fs.h>
 #include <linux/vfs.h>
 #include <linux/inet.h>
 #include "internal.h"
@@ -104,7 +105,8 @@ static size_t nfs_parse_server_name(char *string, size_t len,
 
 	ret = rpc_pton(string, len, sa, salen);
 	if (ret == 0) {
-		ret = nfs_dns_resolve_name(string, len, sa, salen);
+		ret = nfs_dns_resolve_name(string, len, sa, salen,
+				init_rpc_pipefs);
 		if (ret < 0)
 			ret = 0;
 	}
-- 
1.7.3.4

^ permalink raw reply related

* [PATCH v3 09/16] sunrpc: introduce rpc_pipefs_add_destroy_cb()
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs, David S. Miller, Rob Landley, Al Viro,
	containers, netdev, linux-kernel, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas@openvz.org>

Add facility to do some action on destroying of rpc_pipefs superblock.

Signed-off-by: Kirill A. Shutemov <kas@openvz.org>
---
 include/linux/sunrpc/rpc_pipe_fs.h |    3 ++
 net/sunrpc/rpc_pipe.c              |   51 ++++++++++++++++++++++++++++++++++-
 2 files changed, 52 insertions(+), 2 deletions(-)

diff --git a/include/linux/sunrpc/rpc_pipe_fs.h b/include/linux/sunrpc/rpc_pipe_fs.h
index b09bfa5..f5216f1 100644
--- a/include/linux/sunrpc/rpc_pipe_fs.h
+++ b/include/linux/sunrpc/rpc_pipe_fs.h
@@ -46,6 +46,9 @@ RPC_I(struct inode *inode)
 
 extern struct vfsmount *init_rpc_pipefs;
 
+extern int rpc_pipefs_add_destroy_cb(struct super_block *sb,
+		void (*destroy_cb)(void *data), void *data);
+
 extern int rpc_queue_upcall(struct inode *, struct rpc_pipe_msg *);
 
 struct rpc_clnt;
diff --git a/net/sunrpc/rpc_pipe.c b/net/sunrpc/rpc_pipe.c
index 484c9a3..39511e3 100644
--- a/net/sunrpc/rpc_pipe.c
+++ b/net/sunrpc/rpc_pipe.c
@@ -939,6 +939,31 @@ static const struct super_operations s_ops = {
 
 #define RPCAUTH_GSSMAGIC 0x67596969
 
+struct destroy_cb {
+	struct list_head list;
+	void (*callback)(void *data);
+	void *data;
+};
+
+int rpc_pipefs_add_destroy_cb(struct super_block *sb,
+		void (*destroy_cb)(void *data), void *data)
+{
+	struct destroy_cb *dcb;
+	struct list_head *destroy_cb_list = sb->s_fs_info;
+
+	dcb = kmalloc(sizeof(*dcb), GFP_KERNEL);
+	if (!dcb)
+		return -ENOMEM;
+
+	dcb->callback = destroy_cb;
+	dcb->data = data;
+	INIT_LIST_HEAD(&dcb->list);
+	list_add(&dcb->list, destroy_cb_list);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(rpc_pipefs_add_destroy_cb);
+
 /*
  * We have a single directory with 1 node in it.
  */
@@ -1004,8 +1029,16 @@ rpc_fill_super(struct super_block *sb, void *data, int silent)
 		iput(inode);
 		return -ENOMEM;
 	}
-	if (rpc_populate(root, files, RPCAUTH_lockd, RPCAUTH_RootEOF, NULL))
+	/* List of destroy callbacks */
+	sb->s_fs_info = kmalloc(sizeof(struct list_head), GFP_KERNEL);
+	if (!sb->s_fs_info)
+		return -ENOMEM;
+	INIT_LIST_HEAD((struct list_head*) &sb->s_fs_info);
+	if (rpc_populate(root, files, RPCAUTH_lockd, RPCAUTH_RootEOF, NULL)) {
+		kfree(sb->s_fs_info);
 		return -ENOMEM;
+	}
+
 	return 0;
 }
 
@@ -1016,11 +1049,25 @@ rpc_mount(struct file_system_type *fs_type,
 	return mount_single(fs_type, flags, data, rpc_fill_super);
 }
 
+static void rpc_kill_sb(struct super_block *sb)
+{
+	struct list_head *destroy_cb_list = sb->s_fs_info;
+	struct destroy_cb *dcb, *tmp;
+
+	list_for_each_entry_safe(dcb, tmp, destroy_cb_list, list) {
+		dcb->callback(dcb->data);
+		list_del(&dcb->list);
+		kfree(dcb);
+	}
+	kfree(destroy_cb_list);
+	kill_litter_super(sb);
+}
+
 static struct file_system_type rpc_pipe_fs_type = {
 	.owner		= THIS_MODULE,
 	.name		= "rpc_pipefs",
 	.mount		= rpc_mount,
-	.kill_sb	= kill_litter_super,
+	.kill_sb	= rpc_kill_sb,
 };
 
 static void
-- 
1.7.3.4

^ permalink raw reply related

* [PATCH v3 08/16] sunrpc: tag pipefs field of cache_detail with rpc_pipefs mount point
From: Kirill A. Shutemov @ 2011-01-14 13:49 UTC (permalink / raw)
  To: Trond Myklebust, J. Bruce Fields, Neil Brown
  Cc: Pavel Emelyanov, linux-nfs-u79uwXL29TY76Z2rM5mHXA,
	David S. Miller, Rob Landley, Al Viro,
	containers-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA,
	netdev-u79uwXL29TY76Z2rM5mHXA,
	linux-kernel-u79uwXL29TY76Z2rM5mHXA, Kirill A. Shutemov
In-Reply-To: <1295012954-7769-1-git-send-email-kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>

Signed-off-by: Kirill A. Shutemov <kas-GEFAQzZX7r8dnm+yROfE0A@public.gmane.org>
---
 fs/nfs/cache_lib.c           |    3 +--
 include/linux/sunrpc/cache.h |    9 +++------
 net/sunrpc/cache.c           |   16 ++++++++++------
 3 files changed, 14 insertions(+), 14 deletions(-)

diff --git a/fs/nfs/cache_lib.c b/fs/nfs/cache_lib.c
index dd7ca5f..0944d4e 100644
--- a/fs/nfs/cache_lib.c
+++ b/fs/nfs/cache_lib.c
@@ -123,7 +123,7 @@ int nfs_cache_register(struct cache_detail *cd)
 	ret = vfs_path_lookup(mnt->mnt_root, mnt, "/cache", 0, &nd);
 	if (ret)
 		goto err;
-	ret = sunrpc_cache_register_pipefs(nd.path.dentry,
+	ret = sunrpc_cache_register_pipefs(mnt, nd.path.dentry,
 			cd->name, 0600, cd);
 	path_put(&nd.path);
 	if (!ret)
@@ -136,6 +136,5 @@ err:
 void nfs_cache_unregister(struct cache_detail *cd)
 {
 	sunrpc_cache_unregister_pipefs(cd);
-	mntput(init_rpc_pipefs);
 }
 
diff --git a/include/linux/sunrpc/cache.h b/include/linux/sunrpc/cache.h
index 78aa104..339bca3 100644
--- a/include/linux/sunrpc/cache.h
+++ b/include/linux/sunrpc/cache.h
@@ -65,10 +65,6 @@ struct cache_detail_procfs {
 	struct proc_dir_entry   *flush_ent, *channel_ent, *content_ent;
 };
 
-struct cache_detail_pipefs {
-	struct dentry *dir;
-};
-
 struct cache_detail {
 	struct module *		owner;
 	int			hash_size;
@@ -115,7 +111,7 @@ struct cache_detail {
 
 	union {
 		struct cache_detail_procfs procfs;
-		struct cache_detail_pipefs pipefs;
+		struct path pipefs;
 	} u;
 };
 
@@ -202,7 +198,8 @@ extern int cache_register_net(struct cache_detail *cd, struct net *net);
 extern void cache_unregister(struct cache_detail *cd);
 extern void cache_unregister_net(struct cache_detail *cd, struct net *net);
 
-extern int sunrpc_cache_register_pipefs(struct dentry *parent, const char *,
+extern int sunrpc_cache_register_pipefs(struct vfsmount *rpcmount,
+					struct dentry *parent, const char *,
 					mode_t, struct cache_detail *);
 extern void sunrpc_cache_unregister_pipefs(struct cache_detail *);
 
diff --git a/net/sunrpc/cache.c b/net/sunrpc/cache.c
index e433e75..ed50d49 100644
--- a/net/sunrpc/cache.c
+++ b/net/sunrpc/cache.c
@@ -28,6 +28,7 @@
 #include <linux/workqueue.h>
 #include <linux/mutex.h>
 #include <linux/pagemap.h>
+#include <linux/mount.h>
 #include <asm/ioctls.h>
 #include <linux/sunrpc/types.h>
 #include <linux/sunrpc/cache.h>
@@ -1753,7 +1754,8 @@ const struct file_operations cache_flush_operations_pipefs = {
 	.llseek		= no_llseek,
 };
 
-int sunrpc_cache_register_pipefs(struct dentry *parent,
+int sunrpc_cache_register_pipefs(struct vfsmount *rpcmount,
+				 struct dentry *parent,
 				 const char *name, mode_t umode,
 				 struct cache_detail *cd)
 {
@@ -1766,9 +1768,10 @@ int sunrpc_cache_register_pipefs(struct dentry *parent,
 	q.len = strlen(name);
 	q.hash = full_name_hash(q.name, q.len);
 	dir = rpc_create_cache_dir(parent, &q, umode, cd);
-	if (!IS_ERR(dir))
-		cd->u.pipefs.dir = dir;
-	else {
+	if (!IS_ERR(dir)) {
+		cd->u.pipefs.mnt = mntget(rpcmount);
+		cd->u.pipefs.dentry = dir;
+	} else {
 		sunrpc_destroy_cache_detail(cd);
 		ret = PTR_ERR(dir);
 	}
@@ -1778,8 +1781,9 @@ EXPORT_SYMBOL_GPL(sunrpc_cache_register_pipefs);
 
 void sunrpc_cache_unregister_pipefs(struct cache_detail *cd)
 {
-	rpc_remove_cache_dir(cd->u.pipefs.dir);
-	cd->u.pipefs.dir = NULL;
+	rpc_remove_cache_dir(cd->u.pipefs.dentry);
+	cd->u.pipefs.dentry = NULL;
+	mntput(cd->u.pipefs.mnt);
 	sunrpc_destroy_cache_detail(cd);
 }
 EXPORT_SYMBOL_GPL(sunrpc_cache_unregister_pipefs);
-- 
1.7.3.4

--
To unsubscribe from this list: send the line "unsubscribe linux-nfs" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

^ permalink raw reply related


This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox