* Re: bonding: time limits too tight in bond_ab_arp_inspect
From: Chris Friesen @ 2012-08-22 18:58 UTC (permalink / raw)
To: Jay Vosburgh; +Cc: Jiri Bohac, Andy Gospodarek, netdev, Petr Tesarik
In-Reply-To: <24655.1345660922@death.nxdomain>
On 08/22/2012 12:42 PM, Jay Vosburgh wrote:
> Chris Friesen<chris.friesen@genband.com> wrote:
>
>> On 08/22/2012 11:45 AM, Jiri Bohac wrote:
>>
>>> This code is run from bond_activebackup_arp_mon() about
>>> delta_in_ticks jiffies after the previous ARP probe has been
>>> sent. If the delayed work gets executed exactly in delta_in_ticks
>>> jiffies, there is a chance the slave will be brought up. If the
>>> delayed work runs one jiffy later, the slave will stay down.
>
> Presumably the ARP reply is coming back in less than one jiffy,
> then, so the slave_last_rx() value is the same jiffy as when the
> _inspect was previously called?
>
>> <snip>
>>
>>> Should they perhaps all be increased by, say, delta_in_ticks/2, to make this
>>> less dependent on the current scheduling latencies?
>>
>> We have been using a patch that tracks the arpmon requested sleep time vs
>> the actual sleep time and adds any scheduling latency to the allowed
>> delta. That way if we sleep too long due to scheduling latency it doesn't
>> affect the calculation.
>
> How much scheduling latency do you see?
>
> Is that really better than just permitting a bit more slack in
> the timing window?
We hit enough latency that it triggered arpmon to falsely mark multiple
links as lost. This triggered our system maintenance code to go into a
"oh no we can't talk to the outside world" secenario, which does fairly
intrusive things to try and bring connectivity back up. Basically a bad
thing to happen just because of a random scheduler latency spike.
I should note that we added this some time back and are still running
older kernels so I have no idea what latency on modern kernels is like.
Chris
^ permalink raw reply
* Re: [PATCH 2/3] x86_64: Define 128-bit memory-mapped I/O operations
From: Ben Hutchings @ 2012-08-22 19:01 UTC (permalink / raw)
To: Linus Torvalds
Cc: H. Peter Anvin, David Laight, Benjamin LaHaise, David Miller,
tglx, mingo, netdev, linux-net-drivers, x86
In-Reply-To: <CA+55aFyBVRHk6ZcWxEPoUPRmFYyW6yihbi26vS-SAKey0nv9yA@mail.gmail.com>
On Wed, 2012-08-22 at 11:28 -0700, Linus Torvalds wrote:
> On Wed, Aug 22, 2012 at 11:11 AM, Ben Hutchings
> <bhutchings@solarflare.com> wrote:
> >
> > Right, I think it's been made pretty clear that it's going to be
> > dependent on more than just architecture.
>
> Well, it's entirely possible that the 128-bit case will work correctly
> on all x86-64 hardware out there on PCIe.
>
> I just can't guarantee it, because we certainly have had issues with
> hw doing odd things before. But maybe PCIe really is well-specified
> enough, and maybe nobody has done a odd PCIe bridges, and maybe every
> time some 128-bit access is split, the bus in question still always
> remembers the original 128-bit size in the transaction. It's not at
> all impossible. I just wouldn't *guarantee* it.
Even then, everything works out OK if the particular MMIO write I'm
concerned about *is* split - just as long as the resulting operations
are in ascending address order. This is not true on all systems if we
enable write-combining (without a fence in the middle, which defeats the
purpose), and I think hpa was saying that it may not be the case with
SSE writes either.
> And to some degree, for high-end server-only hardware in particular,
> it really *is* acceptable to say "If you have odd hardware, odd things
> will happen". So for this particular driver, maybe the right approach
> is simply to say "we require that your fabric works right". And see if
> anybody ever complains.
Maybe. At the moment reordering tends to cause the hardware to complain
that we sent an invalid sequence of DMA descriptors, but that's only
because we're not being as smart as we could about using TX push. I
don't want to run the risk of sending out corrupted packets (with
offloaded checksums, so they're not that obviously invalid) on the wire.
Ben.
> The 100ns may be worth those kinds of "you'd better not have old/crap
> hardware" decisions. It's not acceptable for some drivers (a driver
> for some consumer ATA chip might not want to make that kind of choice,
> and say "whatever, we'll be really conservative), but "Quod licet
> Jovi, non licet bovi".
>
> The fact that something might not be *guaranteed* to always work
> doesn't necessarily mean that it is always the wrong thing to do..
--
Ben Hutchings, Staff Engineer, Solarflare
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
* bonding: why zero out bond_dev->dev_addr when last slave removed?
From: Chris Friesen @ 2012-08-22 20:48 UTC (permalink / raw)
To: Jay Vosburgh, Andy Gospodarek, netdev
Hi all,
I've got a couple questions about the bonding driver.
I was wondering about the rationale for zeroing out bond_dev->dev_addr
when the last slave is removed from the bond.
Assuming bond0 is currently using the mac address of eth0 then doing
ifenslave -d bond0 eth0
ifenslave -d bond0 eth1
ifenslave bond0 eth1
ifenslave bond0 eth0
ends up changing the MAC address of the bond link. Given that the bond
itself stays up during this time, why don't we let the bond device keep
it's previous MAC address (at least if fail_over_mac is zero)?
Is this to account for the case where we move the bond to totally
different devices such that the old MAC no longer belongs to one of the
slaves but instead belongs to a NIC outside the bond?
Chris
--
Chris Friesen
Software Designer
3500 Carling Avenue
Ottawa, Ontario K2H 8E9
www.genband.com
^ permalink raw reply
* Re: [PATCH 2/3] x86_64: Define 128-bit memory-mapped I/O operations
From: David Miller @ 2012-08-22 21:14 UTC (permalink / raw)
To: hpa; +Cc: torvalds, bhutchings, tglx, mingo, netdev, linux-net-drivers, x86
In-Reply-To: <20120821.211427.1832042852041589162.davem@davemloft.net>
From: David Miller <davem@davemloft.net>
Date: Tue, 21 Aug 2012 21:14:27 -0700 (PDT)
> From: "H. Peter Anvin" <hpa@zytor.com>
> Date: Tue, 21 Aug 2012 20:59:26 -0700
>
>> kernel_fpu_end() would still have to re-enable preemption (and
>> preemption would have to check the work flag), but that should be cheap.
>>
>> We could allow the FPU in the kernel to have preemption, if we allocated
>> space for two xstates per thread instead of one. That is, however, a
>> fair hunk of memory.
>
> Once you have done the first FPU save for the sake of the kernel, you
> can minimize what you save for any deeper nesting because the kernel
> only cares about a very limited part of that FPU state not the whole
> 1K thing.
>
> Those bits you can save by hand with a bunch of explicit stores of the
> XMM registers, or something like that.
BTW, just to clarify, I'm not saying that we should save the FPU on
every trap where we find the FPU enabled or anything stupid like that.
Definitely keep the kern_fpu_begin()/kern_fpu_end() type markers
around FPU usage, but allow some kind of nesting facility.
Here's one idea. Anyone using the existing kern_fpu_*() markers get
the existing behavior. Only one level of kernel FPU usage is allowed.
But a new interface allows specification of a state-save mask. And it
is only users of this interface for which we allow nesting past the
first FPU user.
If this is the first kernel FPU user, we always do the full fxsave or
whatever to push out the full state. For any level of kernel FPU
nesting we save only what is in the save-mask, by hand.
^ permalink raw reply
* Re: bonding: why zero out bond_dev->dev_addr when last slave removed?
From: Jay Vosburgh @ 2012-08-22 21:20 UTC (permalink / raw)
To: Chris Friesen; +Cc: Andy Gospodarek, netdev
In-Reply-To: <5035458C.9000804@genband.com>
Chris Friesen <chris.friesen@genband.com> wrote:
[...]
>I was wondering about the rationale for zeroing out bond_dev->dev_addr
>when the last slave is removed from the bond.
>
>Assuming bond0 is currently using the mac address of eth0 then doing
>
>ifenslave -d bond0 eth0
>ifenslave -d bond0 eth1
>ifenslave bond0 eth1
>ifenslave bond0 eth0
>
>ends up changing the MAC address of the bond link. Given that the bond
>itself stays up during this time, why don't we let the bond device keep
>it's previous MAC address (at least if fail_over_mac is zero)?
>
>Is this to account for the case where we move the bond to totally
>different devices such that the old MAC no longer belongs to one of the
>slaves but instead belongs to a NIC outside the bond?
More or less. When the last slave is removed from the bond,
several properties of the bond are reset to an initial state, one of
which is that the bond's MAC is reset to all zeroes.
If fail_over_mac is set to "active," then the bond must use the
MAC of one of its slaves, as the implication is that the slaves cannot
or should not change their MAC addresses. In this case, I believe the
bond's MAC will be set to the first active slave's MAC, so holding over
the bond's MAC makes no practical difference.
For fail_over_mac=follow, if the bond kept its MAC, then it
would be possible for both slaves to end up with the same MAC (because
the follow logic will set the first slave to the bond's MAC when it is
made active, and the second slave will keep its MAC after enslavement,
and those could be the same MAC if the second slave is the one the bond
originally got its MAC from).
-J
---
-Jay Vosburgh, IBM Linux Technology Center, fubar@us.ibm.com
^ permalink raw reply
* Re: [PATCH 00/18] netfilter: IPv6 NAT
From: Patrick McHardy @ 2012-08-22 21:23 UTC (permalink / raw)
To: David Miller; +Cc: netfilter-devel, netdev
In-Reply-To: <20120822.022801.740821969974372752.davem@davemloft.net>
On Wed, 22 Aug 2012, David Miller wrote:
> From: Patrick McHardy <kaber@trash.net>
> Date: Mon, 20 Aug 2012 05:39:48 +0200
>
>> Following is the latest IPv6 NAT patchset, based on -rc2.
>
> Feel free to push patch #1 via the nf tree when give the
> final version of this to Pablo, and you can add my ACK to
> it as well if you like:
>
> Acked-by: David S. Miller <davem@davemloft.net>
Thanks, will do. The NAT patches depend on a couple of fixes in Pablo's
latest submission, if you could merge net.git into net-next.git, I can
push them to Pablo without creating conflicts. Thanks!
^ permalink raw reply
* Re: [net-next 0/6][pull request] Intel Wired LAN Driver Updates
From: David Miller @ 2012-08-22 21:24 UTC (permalink / raw)
To: jeffrey.t.kirsher; +Cc: netdev, gospo, sassmann
In-Reply-To: <1345538275-1690-1-git-send-email-jeffrey.t.kirsher@intel.com>
From: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Date: Tue, 21 Aug 2012 01:37:49 -0700
> This series contains updates to ethtool.h, e1000, e1000e, and igb to
> implement MDI/MDIx control.
>
> The following are changes since commit 1d76efe1577b4323609b1bcbfafa8b731eda071a:
> team: add support for non-ethernet devices
> and are available in the git repository at:
> git://git.kernel.org/pub/scm/linux/kernel/git/jkirsher/net-next master
Pulled, thanks Jeff.
^ permalink raw reply
* Re: [PATCH 2/3] x86_64: Define 128-bit memory-mapped I/O operations
From: Linus Torvalds @ 2012-08-22 21:28 UTC (permalink / raw)
To: David Miller; +Cc: hpa, bhutchings, tglx, mingo, netdev, linux-net-drivers, x86
In-Reply-To: <20120822.141433.730254311852927123.davem@davemloft.net>
On Wed, Aug 22, 2012 at 2:14 PM, David Miller <davem@davemloft.net> wrote:
>
> BTW, just to clarify, I'm not saying that we should save the FPU on
> every trap where we find the FPU enabled or anything stupid like that.
>
> Definitely keep the kern_fpu_begin()/kern_fpu_end() type markers
> around FPU usage, but allow some kind of nesting facility.
So nesting shouldn't be horrible, but the thing that really screws
with people like the crypto use is not nesting, but the fact that
sometimes you can't save at all, and the whole "kernel_fpu_possible()"
or whatever we call the checking function.
IOW, in [soft]irq context, to avoid races with the irq happening as
the process is going to do something with the FPU state, we don't
allow saving and changing state, because that would mean that the
normal FP state paths would have to be irq-safe, and they aren't.
And once you have to have that fpu possible check, if it happens to
also disallow nested use, I doubt that's going to really affect
anybody. The code has to take the case of "I'm not allowed to change
FPU state" case into account regardless.
Linus
^ permalink raw reply
* Re: [PATCH 2/3] x86_64: Define 128-bit memory-mapped I/O operations
From: David Miller @ 2012-08-22 21:38 UTC (permalink / raw)
To: torvalds; +Cc: hpa, bhutchings, tglx, mingo, netdev, linux-net-drivers, x86
In-Reply-To: <CA+55aFxxSFMiohu80BZCObA0APPe08h1-7eAZ_BLoAZDWqqv0Q@mail.gmail.com>
From: Linus Torvalds <torvalds@linux-foundation.org>
Date: Wed, 22 Aug 2012 14:28:50 -0700
> On Wed, Aug 22, 2012 at 2:14 PM, David Miller <davem@davemloft.net> wrote:
>>
>> BTW, just to clarify, I'm not saying that we should save the FPU on
>> every trap where we find the FPU enabled or anything stupid like that.
>>
>> Definitely keep the kern_fpu_begin()/kern_fpu_end() type markers
>> around FPU usage, but allow some kind of nesting facility.
>
> So nesting shouldn't be horrible, but the thing that really screws
> with people like the crypto use is not nesting, but the fact that
> sometimes you can't save at all, and the whole "kernel_fpu_possible()"
> or whatever we call the checking function.
>
> IOW, in [soft]irq context, to avoid races with the irq happening as
> the process is going to do something with the FPU state, we don't
> allow saving and changing state, because that would mean that the
> normal FP state paths would have to be irq-safe, and they aren't.
>
> And once you have to have that fpu possible check, if it happens to
> also disallow nested use, I doubt that's going to really affect
> anybody. The code has to take the case of "I'm not allowed to change
> FPU state" case into account regardless.
I don't think you really have to do anything special to handle
interrupts properly.
Let's assume that we use some variable length save area at the end of
thread_info to do this nested saving.
When you are asked for FPU usage, you first figure out how much you're
going to save.
Then you advance the allocation pointer in the thread_info, and save
into the space you allocated.
If an interrupt wants to use the FPU, that should be fine as well.
Whether the interrupt FPU save does it's save after you did, or
before, it should work out fine.
I suppose you might have some issues in determining whether we need to
do the full fxsave stuff or not. There could be a state bit for that,
or similar.
Another idea, instead of doing this in thread_info, is to do it on the
local stack. That way if we're in an interrupt, we'll use that
interrupt type's kernel stack.
You might be able to get away with always doing the full FPU
save/restore in that situation.
^ permalink raw reply
* Re: [PATCH 00/18] netfilter: IPv6 NAT
From: David Miller @ 2012-08-22 21:42 UTC (permalink / raw)
To: kaber; +Cc: netfilter-devel, netdev
In-Reply-To: <Pine.GSO.4.63.1208222322060.25423@stinky-local.trash.net>
From: Patrick McHardy <kaber@trash.net>
Date: Wed, 22 Aug 2012 23:23:53 +0200 (MEST)
> The NAT patches depend on a couple of fixes in Pablo's latest
> submission, if you could merge net.git into net-next.git, I can push
> them to Pablo without creating conflicts. Thanks!
I just did that right now, should show up on kernel.org shortly.
^ permalink raw reply
* [PATCH] netdev/phy: add MDIO bus multiplexer driven by a memory-mapped FPGA
From: Timur Tabi @ 2012-08-22 21:45 UTC (permalink / raw)
To: Andy Fleming, David Miller, netdev, david.daney
An FPGA controls which sub-bus is connected to the master MDIO bus. The
FPGA must be memory-mapped and contain only 8-bit registers (which keeps
things simple).
Tested on a Freescale P5020DS board which uses the "PIXIS" FPGA attached
to the localbus.
Signed-off-by: Timur Tabi <timur@freescale.com>
---
.../devicetree/bindings/net/mdio-mux-fpga.txt | 74 ++++++++
drivers/net/phy/Kconfig | 13 ++
drivers/net/phy/Makefile | 1 +
drivers/net/phy/mdio-mux-fpga.c | 186 ++++++++++++++++++++
4 files changed, 274 insertions(+), 0 deletions(-)
create mode 100644 Documentation/devicetree/bindings/net/mdio-mux-fpga.txt
create mode 100644 drivers/net/phy/mdio-mux-fpga.c
diff --git a/Documentation/devicetree/bindings/net/mdio-mux-fpga.txt b/Documentation/devicetree/bindings/net/mdio-mux-fpga.txt
new file mode 100644
index 0000000..ef567c6
--- /dev/null
+++ b/Documentation/devicetree/bindings/net/mdio-mux-fpga.txt
@@ -0,0 +1,74 @@
+Properties for an MDIO bus multiplexer/switch controlled by an FPGA register.
+
+This is a special case of a MDIO bus multiplexer. An FPGA register is used
+to control which child bus is connected.
+
+Required properties in addition to the generic multiplexer properties:
+
+- compatible : string, must contain "mdio-mux-fpga"
+
+- mdio-mux-device : phandle, points to the FPGA (or similar) node. This
+ must be a memory-mapped device with 8-bit registers.
+
+- mdio-mux-register : integer, contains the offset of the register that
+ controls the bus multiplexer.
+
+- mdio-mux-mask : integer, contains an 8-bit mask that specifies which
+ bits in the register control the actual bus multiplexer. The
+ 'reg' property of each child mdio-mux node must be constrained by
+ this mask.
+
+Example:
+
+The FPGA node defines a memory-mapped FPGA with a register space of 0x30 bytes.
+For the "EMI2" MDIO bus, register 9 (BRDCFG1) controls the mux on that bus.
+A bitmask of 0x6 means that bits 1 and 2 (bit 0 is lsb) are the bits on
+BRDCFG1 that control the actual mux.
+
+ /* The FPGA node */
+ fpga: board-control@3,0 {
+ compatible = "fsl,p5020ds-fpga", "fsl,fpga-ngpixis";
+ reg = <3 0 0x30>;
+ };
+
+ /* The parent MDIO bus. */
+ xmdio0: mdio@f1000 {
+ #address-cells = <1>;
+ #size-cells = <0>;
+ compatible = "fsl,fman-xmdio";
+ reg = <0xf1000 0x1000>;
+ interrupts = <100 1 0 0>;
+ };
+
+ mdio-mux-emi2 {
+ compatible = "mdio-mux-fpga", "mdio-mux";
+ mdio-parent-bus = <&xmdio0>;
+ #address-cells = <1>;
+ #size-cells = <0>;
+ mdio-mux-device = <&fpga>;
+ mdio-mux-register = <9>; // BRDCFG1
+ mdio-mux-mask = <0x6>; // EMI2
+
+ emi2_slot1: mdio@0 { // Slot 1 XAUI (FM2)
+ reg = <0>;
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ phy_xgmii_slot1: ethernet-phy@0 {
+ compatible = "ethernet-phy-ieee802.3-c45";
+ reg = <4>;
+ };
+ };
+
+ emi2_slot2: mdio@2 { // Slot 2 XAUI (FM1)
+ reg = <2>;
+ #address-cells = <1>;
+ #size-cells = <0>;
+
+ phy_xgmii_slot2: ethernet-phy@4 {
+ compatible = "ethernet-phy-ieee802.3-c45";
+ reg = <0>;
+ };
+ };
+ };
+
diff --git a/drivers/net/phy/Kconfig b/drivers/net/phy/Kconfig
index 3090dc6..c3fc957 100644
--- a/drivers/net/phy/Kconfig
+++ b/drivers/net/phy/Kconfig
@@ -159,6 +159,19 @@ config MDIO_BUS_MUX_GPIO
several child MDIO busses to a parent bus. Child bus
selection is under the control of GPIO lines.
+config MDIO_BUS_MUX_FPGA
+ tristate "Support for FPGA-controlled MDIO bus multiplexers"
+ depends on OF_MDIO
+ select MDIO_BUS_MUX
+ help
+ This module provides a driver for MDIO bus multiplexers that
+ are controlled via a simple memory-mapped FPGA device. The
+ multiplexer connects one of several child MDIO busses to a parent
+ bus. Child bus selection is under the control of one of the
+ FPGA's registers.
+
+ Currently, only 8-bit registers are supported.
+
endif # PHYLIB
config MICREL_KS8995MA
diff --git a/drivers/net/phy/Makefile b/drivers/net/phy/Makefile
index 6d2dc6c..3bf4d7a 100644
--- a/drivers/net/phy/Makefile
+++ b/drivers/net/phy/Makefile
@@ -28,3 +28,4 @@ obj-$(CONFIG_MICREL_KS8995MA) += spi_ks8995.o
obj-$(CONFIG_AMD_PHY) += amd.o
obj-$(CONFIG_MDIO_BUS_MUX) += mdio-mux.o
obj-$(CONFIG_MDIO_BUS_MUX_GPIO) += mdio-mux-gpio.o
+obj-$(CONFIG_MDIO_BUS_MUX_FPGA) += mdio-mux-fpga.o
diff --git a/drivers/net/phy/mdio-mux-fpga.c b/drivers/net/phy/mdio-mux-fpga.c
new file mode 100644
index 0000000..7b4e69c
--- /dev/null
+++ b/drivers/net/phy/mdio-mux-fpga.c
@@ -0,0 +1,186 @@
+/*
+ * FPGA MDIO MUX driver
+ *
+ * This driver supports
+ * Author: Timur Tabi <timur@freescale.com>
+ *
+ * Copyright 2012 Freescale Semiconductor, Inc.
+ *
+ * This file is licensed under the terms of the GNU General Public License
+ * version 2. This program is licensed "as is" without any warranty of any
+ * kind, whether express or implied.
+ */
+
+#include <linux/platform_device.h>
+#include <linux/device.h>
+#include <linux/of_mdio.h>
+#include <linux/module.h>
+#include <linux/init.h>
+#include <linux/phy.h>
+#include <linux/mdio-mux.h>
+
+struct mdio_mux_fpga_state {
+ void *mux_handle;
+ phys_addr_t phys;
+ unsigned int offset;
+ uint8_t mask;
+};
+
+/*
+ * MDIO multiplexing switch function
+ *
+ * This function is called by the mdio-mux layer when it thinks the mdio bus
+ * multiplexer needs to switch.
+ *
+ * 'current_child' is the current value of the mux register (masked via
+ * s->mask).
+ *
+ * 'desired_child' is the value of the 'reg' property of the target child MDIO
+ * node.
+ *
+ * The first time this function is called, current_child == -1.
+ *
+ * If current_child == desired_child, then the mux is already set to the
+ * correct bus.
+ */
+static int mdio_mux_fpga_switch_fn(int current_child, int desired_child,
+ void *data)
+{
+ struct mdio_mux_fpga_state *s = data;
+
+ if (current_child ^ desired_child) {
+ void *p = ioremap(s->phys + s->offset, 1);
+ uint8_t x;
+
+ if (!p)
+ return -ENOMEM;
+
+ x = ioread8(p);
+ iowrite8((x & ~s->mask) | desired_child, p);
+
+ iounmap(p);
+ }
+
+ return 0;
+}
+
+static int __devinit mdio_mux_fpga_probe(struct platform_device *pdev)
+{
+ struct device_node *np2, *np = pdev->dev.of_node;
+ struct mdio_mux_fpga_state *s;
+ struct resource res;
+ const __be32 *iprop;
+ int len, ret;
+
+ dev_dbg(&pdev->dev, "probing node %s\n", np->full_name);
+
+ s = devm_kzalloc(&pdev->dev, sizeof(*s), GFP_KERNEL);
+ if (!s)
+ return -ENOMEM;
+
+ iprop = of_get_property(np, "mdio-mux-device", &len);
+ if (!iprop || len != sizeof(phandle)) {
+ dev_err(&pdev->dev, "missing mdio-mux-device property\n");
+ return -ENODEV;
+ }
+ np2 = of_find_node_by_phandle(be32_to_cpup(iprop));
+ if (!np2) {
+ dev_err(&pdev->dev, "mdio-mux-device points to invalid node\n");
+ return -ENODEV;
+ }
+
+ ret = of_address_to_resource(np2, 0, &res);
+ if (ret) {
+ dev_err(&pdev->dev, "cannot obtain memory map for node %s\n",
+ np2->full_name);
+ return ret;
+ }
+ s->phys = res.start;
+
+ iprop = of_get_property(np, "mdio-mux-register", &len);
+ if (!iprop || len != sizeof(uint32_t)) {
+ dev_err(&pdev->dev, "missing mdio-mux-register property\n");
+ return -EINVAL;
+ }
+ s->offset = be32_to_cpup(iprop);
+ if (s->offset >= resource_size(&res)) {
+ dev_err(&pdev->dev, "mdio-mux-register value %u is too large\n",
+ s->offset);
+ return -EINVAL;
+ }
+
+ iprop = of_get_property(np, "mdio-mux-mask", &len);
+ if (!iprop || len != sizeof(uint32_t)) {
+ dev_err(&pdev->dev, "missing mdio-mux-mask property\n");
+ return -ENODEV;
+ }
+ if (be32_to_cpup(iprop) > 255) {
+ dev_err(&pdev->dev, "only 8-bit registers are supported\n");
+ return -EINVAL;
+ }
+ s->mask = be32_to_cpup(iprop);
+
+ /*
+ * Verify that the 'reg' property of each child MDIO bus does not
+ * set any bits outside of the 'mask'.
+ */
+ for_each_available_child_of_node(np, np2) {
+ iprop = of_get_property(np2, "reg", &len);
+ if (!iprop || len != sizeof(uint32_t)) {
+ dev_err(&pdev->dev, "mdio-mux child node %s is "
+ "missing a 'reg' property\n", np2->full_name);
+ return -ENODEV;
+ }
+ if (be32_to_cpup(iprop) & ~s->mask) {
+ dev_err(&pdev->dev, "mdio-mux child node %s has "
+ "a 'reg' value with unmasked bits\n",
+ np2->full_name);
+ return -ENODEV;
+ }
+ }
+
+ ret = mdio_mux_init(&pdev->dev, mdio_mux_fpga_switch_fn,
+ &s->mux_handle, s);
+ if (ret) {
+ dev_err(&pdev->dev, "failed to register mdio-mux bus %s\n",
+ np->full_name);
+ return ret;
+ }
+
+ pdev->dev.platform_data = s;
+
+ return 0;
+}
+
+static int __devexit mdio_mux_fpga_remove(struct platform_device *pdev)
+{
+ struct mdio_mux_fpga_state *s = dev_get_platdata(&pdev->dev);
+
+ mdio_mux_uninit(s->mux_handle);
+
+ return 0;
+}
+
+static struct of_device_id mdio_mux_fpga_match[] = {
+ {
+ .compatible = "mdio-mux-fpga",
+ },
+ {},
+};
+MODULE_DEVICE_TABLE(of, mdio_mux_fpga_match);
+
+static struct platform_driver mdio_mux_fpga_driver = {
+ .driver = {
+ .name = "mdio-mux-fpga",
+ .owner = THIS_MODULE,
+ .of_match_table = mdio_mux_fpga_match,
+ },
+ .probe = mdio_mux_fpga_probe,
+ .remove = __devexit_p(mdio_mux_fpga_remove),
+};
+
+module_platform_driver(mdio_mux_fpga_driver);
+
+MODULE_AUTHOR("Timur Tabi <timur@freescale.com>");
+MODULE_DESCRIPTION("FPGA MDIO MUX driver");
+MODULE_LICENSE("GPL v2");
--
1.7.3.4
^ permalink raw reply related
* Re: [PATCH 06/18] netfilter: add protocol independant NAT core
From: Patrick McHardy @ 2012-08-22 22:13 UTC (permalink / raw)
To: Jan Engelhardt; +Cc: netfilter-devel, netdev
In-Reply-To: <alpine.LNX.2.01.1208201011480.6101@frira.zrqbmnf.qr>
On Mon, 20 Aug 2012, Jan Engelhardt wrote:
> On Monday 2012-08-20 05:39, Patrick McHardy wrote:
>>
>> enum ctattr_nat {
>> CTA_NAT_UNSPEC,
>> - CTA_NAT_MINIP,
>> - CTA_NAT_MAXIP,
>> + CTA_NAT_V4_MINIP,
>> +#define CTA_NAT_MINIP CTA_NAT_V4_MINIP
>> + CTA_NAT_V4_MAXIP,
>> +#define CTA_NAT_MAXIP CTA_NAT_V4_MAXIP
>> CTA_NAT_PROTO,
>> __CTA_NAT_MAX
>> };
>
> One could also
>
> enum ctattr_nat {
> ...
> __CTA_NAT_MAX,
>
> CTA_NAT_MINIP = CTA_NAT_V4_MINIP,
> CTA_NAT_MAXIP = CTA_NAT_V4_MAXIP,
> };
>
> to provide the old names.
Sure. Doesn't really matter since defines are not used consistently
(in which case you could use them for #ifdefs).
>> diff --git a/net/ipv4/netfilter/Kconfig b/net/ipv4/netfilter/Kconfig
>> index fcc543c..33372a1 100644
>> --- a/net/ipv4/netfilter/Kconfig
>> +++ b/net/ipv4/netfilter/Kconfig
> [...]
>>
>> -config NF_NAT_NEEDED
>> - bool
>> - depends on NF_NAT
>> - default y
>> -
>
> Could add "if NF_NAT_IPV4".."endif" block as appropriate around here,
> to save on all the extra "depend on NF_NAT_IPV4" clauses.
Agreed.
>> +static int nf_nat_ipv4_in_range(const struct nf_conntrack_tuple *t,
>> + const struct nf_nat_range *range)
>> +{
>> + return ntohl(t->src.u3.ip) >= ntohl(range->min_addr.ip) &&
>> + ntohl(t->src.u3.ip) <= ntohl(range->max_addr.ip);
>> +}
>
> static bool ..
Changed.
>> +static bool nf_nat_ipv4_manip_pkt(struct sk_buff *skb,
>> + unsigned int iphdroff,
>> + const struct nf_nat_l4proto *l4proto,
>> + const struct nf_conntrack_tuple *target,
>> + enum nf_nat_manip_type maniptype)
>> +{
>> + struct iphdr *iph;
>> + unsigned int hdroff;
>> +
>> + if (!skb_make_writable(skb, iphdroff + sizeof(*iph)))
>> + return false;
>> +
>> + iph = (void *)skb->data + iphdroff;
>
> Is iph = ip_hdr(skb), hdroff = iphdroff+skb_iphdrlen(iph) not usable here?
No, we're also translating the inner packets in ICMP error messages.
>> + hdroff = iphdroff + iph->ihl * 4;
>> +
>> + if (!l4proto->manip_pkt(skb, &nf_nat_l3proto_ipv4, iphdroff, hdroff,
>> + target, maniptype))
>> + return false;
>> + iph = (void *)skb->data + iphdroff;
>
> Is trying to avoid some GNU extensions a worthwhile goal? If so,
> iph = (struct iphdr *)(skb->data + iphdroff) should be used, like in:
I don't get your point.
>> +static void nf_nat_ipv4_csum_update(struct sk_buff *skb,
>> + unsigned int iphdroff, __sum16 *check,
>> + const struct nf_conntrack_tuple *t,
>> + enum nf_nat_manip_type maniptype)
>> +{
>> + struct iphdr *iph = (struct iphdr *)(skb->data + iphdroff);
>> [...]
>> +}
>
>
>
>> +static void nf_nat_ipv4_csum_recalc(struct sk_buff *skb,
>> + u8 proto, void *data, __sum16 *check,
>> + int datalen, int oldlen)
>> +{
>> + const struct iphdr *iph = ip_hdr(skb);
>> + struct rtable *rt = skb_rtable(skb);
>> +
>> + if (skb->ip_summed != CHECKSUM_PARTIAL) {
>> + if (!(rt->rt_flags & RTCF_LOCAL) &&
>> + (!skb->dev || skb->dev->features & NETIF_F_V4_CSUM)) {
>> + skb->ip_summed = CHECKSUM_PARTIAL;
>> + skb->csum_start = skb_headroom(skb) +
>> + skb_network_offset(skb) +
>> + ip_hdrlen(skb);
>> + skb->csum_offset = (void *)check - data;
>> + *check = ~csum_tcpudp_magic(iph->saddr, iph->daddr,
>> + datalen, proto, 0);
>> + } else {
>> + *check = 0;
>> + *check = csum_tcpudp_magic(iph->saddr, iph->daddr,
>> + datalen, proto,
>> + csum_partial(data, datalen,
>> + 0));
>> + if (proto == IPPROTO_UDP && !*check)
>> + *check = CSUM_MANGLED_0;
>> + }
>> + } else
>> + inet_proto_csum_replace2(check, skb,
>> + htons(oldlen), htons(datalen), 1);
>> +}
>
> Here is a style factory trick: invert the condition such that the
> simple case is first, and the big one becomes an else if
> with a reduced indent:
This is existing code, I don't want to bloat the diff by unnecessarily
rearranging it.
>> +static void __exit nf_nat_l3proto_ipv4_exit(void)
>> +{
>> + nf_nat_l3proto_unregister(&nf_nat_l3proto_ipv4);
>> + nf_nat_l4proto_unregister(NFPROTO_IPV4, &nf_nat_l4proto_icmp);
>> +}
>> +
>> +MODULE_LICENSE("GPL");
>> +MODULE_ALIAS("nf-nat-" __stringify(AF_INET));
>
> Technically, this would have to be NFPROTO_IPV4, though GNU C has yet
> to gain an extension to stringify enum constants..
I'm aware of that.
>> + /* 1) If this srcip/proto/src-proto-part is currently mapped,
>> + * and that same mapping gives a unique tuple within the given
>> + * range, use that.
>> + *
>> + * This is only required for source (ie. NAT/masq) mappings.
>> + * So far, we don't do local source mappings, so multiple
>> + * manips not an issue.
> manips are not an issue.
>
>
>> - /* nf_conntrack_alter_reply might re-allocate extension area */
>> + /* nf_conntrack_alter_reply might re-allocate exntension aera */
>
> extension was correct :)
Fixed, thanks.
>> + .target = xt_snat_target_v1,
>> + .targetsize = sizeof(struct nf_nat_range),
>> + .table = "nat",
>> + .hooks = (1 << NF_INET_POST_ROUTING) |
>> + (1 << NF_INET_LOCAL_OUT),
>> + .me = THIS_MODULE,
>
> .family = NFPROTO_UNSPEC,
>
> Just for completeness.
This is obvious.
^ permalink raw reply
* Re: [PATCH 09/18] netfilter: ipv6: add IPv6 NAT support
From: Patrick McHardy @ 2012-08-22 22:15 UTC (permalink / raw)
To: Jan Engelhardt; +Cc: netfilter-devel, netdev
In-Reply-To: <alpine.LNX.2.01.1208201100340.6101@frira.zrqbmnf.qr>
On Mon, 20 Aug 2012, Jan Engelhardt wrote:
> On Monday 2012-08-20 05:39, Patrick McHardy wrote:
>> +static struct nf_hook_ops nf_nat_ipv6_ops[] __read_mostly = {
>> + /* Before packet filtering, change destination */
>> + {
>> + .hook = nf_nat_ipv6_in,
>> + .owner = THIS_MODULE,
>> + .pf = NFPROTO_IPV6,
>> + .hooknum = NF_INET_PRE_ROUTING,
>> + .priority = NF_IP_PRI_NAT_DST,
>
> NF_IP6_PRI_NAT_DST
>
>> + .hook = nf_nat_ipv6_out,
>> + .owner = THIS_MODULE,
>> + .pf = NFPROTO_IPV6,
>> + .hooknum = NF_INET_POST_ROUTING,
>> + .priority = NF_IP_PRI_NAT_SRC,
>
> IP6 too... (2 more occurrences)
Fixed, thanks.
>> +static void nf_nat_ipv6_csum_recalc(struct sk_buff *skb,
>> + u8 proto, void *data, __sum16 *check,
>> + int datalen, int oldlen)
>> +{
>> + const struct ipv6hdr *ipv6h = ipv6_hdr(skb);
>> + struct rt6_info *rt = (struct rt6_info *)skb_dst(skb);
>> +
>> + if (skb->ip_summed != CHECKSUM_PARTIAL) {
>
> Maybe invert to == CHECKSUM_PARTIAL like in p06/18.
I didn't change the other patch.
^ permalink raw reply
* Re: [PATCH 11/11] netlink: add documentation for memory mapped I/O
From: Patrick McHardy @ 2012-08-22 22:19 UTC (permalink / raw)
To: Jan Engelhardt; +Cc: Florian.Westphal, netdev, netfilter-devel
In-Reply-To: <alpine.LNX.2.01.1208201224430.8114@frira.zrqbmnf.qr>
On Mon, 20 Aug 2012, Jan Engelhardt wrote:
> On Monday 2012-08-20 08:18, Patrick McHardy wrote:
>> +
>> +RX and TX rings
>> +----------------
>> +
>> +Each ring contains a number of continous memory blocks, containing frames of
>> +fixed size dependant on the parameters used for ring setup.
>
> dependent
> [...]
All fixed, thanks.
^ permalink raw reply
* Re: [PATCH 05/19] netfilter: nf_conntrack_ipv6: improve fragmentation handling
From: Patrick McHardy @ 2012-08-22 22:21 UTC (permalink / raw)
To: Jesper Dangaard Brouer; +Cc: netfilter-devel, netdev
In-Reply-To: <1345468385.3069.251.camel@localhost>
On Mon, 20 Aug 2012, Jesper Dangaard Brouer wrote:
> On Sun, 2012-08-19 at 21:44 +0200, Patrick McHardy wrote:
>
>> Could you send me your patch so I get a better picture of what you're
>> doing exactly?
>
> Okay, just posted the patchset.
>
> Specifically look at patch:
> [PATCH 3/3] ipvs: Complete IPv6 fragment handling for IPVS
>
> Where I use the hook to copy the fw mark from the reasm SKB packet to
> the SKB fragments. (Perhaps, this could be done else were in the
> netfilter framework).
Thanks, I'll have a look at this tommorrow.
^ permalink raw reply
* Re: [PATCH 00/18] netfilter: IPv6 NAT
From: Patrick McHardy @ 2012-08-22 22:21 UTC (permalink / raw)
To: David Miller; +Cc: netfilter-devel, netdev
In-Reply-To: <20120822.144208.2245712403707914143.davem@davemloft.net>
On Wed, 22 Aug 2012, David Miller wrote:
> From: Patrick McHardy <kaber@trash.net>
> Date: Wed, 22 Aug 2012 23:23:53 +0200 (MEST)
>
>> The NAT patches depend on a couple of fixes in Pablo's latest
>> submission, if you could merge net.git into net-next.git, I can push
>> them to Pablo without creating conflicts. Thanks!
>
> I just did that right now, should show up on kernel.org shortly.
Thanks Dave.
^ permalink raw reply
* Re: [PATCH 00/18] netfilter: IPv6 NAT
From: Pablo Neira Ayuso @ 2012-08-22 22:22 UTC (permalink / raw)
To: David Miller; +Cc: kaber, netfilter-devel, netdev
In-Reply-To: <20120822.144208.2245712403707914143.davem@davemloft.net>
On Wed, Aug 22, 2012 at 02:42:08PM -0700, David Miller wrote:
> From: Patrick McHardy <kaber@trash.net>
> Date: Wed, 22 Aug 2012 23:23:53 +0200 (MEST)
>
> > The NAT patches depend on a couple of fixes in Pablo's latest
> > submission, if you could merge net.git into net-next.git, I can push
> > them to Pablo without creating conflicts. Thanks!
>
> I just did that right now, should show up on kernel.org shortly.
First, I'll pass you a Netfilter batch with 10 patches asap.
Then, I'll refresh my tree and apply Patrick's NAT patches.
^ permalink raw reply
* Re: [PATCH] netdev/phy: add MDIO bus multiplexer driven by a memory-mapped FPGA
From: David Daney @ 2012-08-22 22:24 UTC (permalink / raw)
To: Timur Tabi, devicetree-discuss@lists.ozlabs.org
Cc: Andy Fleming, David Miller, netdev, david.daney
In-Reply-To: <1345671954-6398-1-git-send-email-timur@freescale.com>
On 08/22/2012 02:45 PM, Timur Tabi wrote:
> An FPGA controls which sub-bus is connected to the master MDIO bus. The
> FPGA must be memory-mapped and contain only 8-bit registers (which keeps
> things simple).
>
> Tested on a Freescale P5020DS board which uses the "PIXIS" FPGA attached
> to the localbus.
>
> Signed-off-by: Timur Tabi <timur@freescale.com>
> ---
> .../devicetree/bindings/net/mdio-mux-fpga.txt | 74 ++++++++
> drivers/net/phy/Kconfig | 13 ++
> drivers/net/phy/Makefile | 1 +
> drivers/net/phy/mdio-mux-fpga.c | 186 ++++++++++++++++++++
I am fine with the general concept of the patch, so I am going to start
a Bike Shedding session with it over the names of some of the things here.
I wonder if *fpga is really a good name for this. It is a general
purpose multiplexer with a memory mapped control register. I would call
it something like mdio-mux-mmioreg.
> 4 files changed, 274 insertions(+), 0 deletions(-)
> create mode 100644 Documentation/devicetree/bindings/net/mdio-mux-fpga.txt
> create mode 100644 drivers/net/phy/mdio-mux-fpga.c
>
> diff --git a/Documentation/devicetree/bindings/net/mdio-mux-fpga.txt b/Documentation/devicetree/bindings/net/mdio-mux-fpga.txt
> new file mode 100644
> index 0000000..ef567c6
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/net/mdio-mux-fpga.txt
> @@ -0,0 +1,74 @@
> +Properties for an MDIO bus multiplexer/switch controlled by an FPGA register.
> +
> +This is a special case of a MDIO bus multiplexer. An FPGA register is used
> +to control which child bus is connected.
> +
> +Required properties in addition to the generic multiplexer properties:
> +
> +- compatible : string, must contain "mdio-mux-fpga"
> +
> +- mdio-mux-device : phandle, points to the FPGA (or similar) node. This
> + must be a memory-mapped device with 8-bit registers.
You shouldn't need this. Just make the multiplexer a child of FPGA node
to indicate where it lives.
> +
> +- mdio-mux-register : integer, contains the offset of the register that
> + controls the bus multiplexer.
This should just be the normal "reg" properly
> +
> +- mdio-mux-mask : integer, contains an 8-bit mask that specifies which
> + bits in the register control the actual bus multiplexer. The
> + 'reg' property of each child mdio-mux node must be constrained by
> + this mask.
> +
"reg-mask" ??
Do you need a shift too?
David Daney
^ permalink raw reply
* Re: [PATCH 3/3] netfilter: replace list_for_each_continue_rcu with new interface
From: Pablo Neira Ayuso @ 2012-08-22 22:32 UTC (permalink / raw)
To: Michael Wang
Cc: LKML, netdev@vger.kernel.org, netfilter, coreteam,
netfilter-devel, David Miller, kaber, paulmck@linux.vnet.ibm.com
In-Reply-To: <502DC9A3.2070703@linux.vnet.ibm.com>
On Fri, Aug 17, 2012 at 12:33:39PM +0800, Michael Wang wrote:
> From: Michael Wang <wangyun@linux.vnet.ibm.com>
>
> This patch replaces list_for_each_continue_rcu() with
> list_for_each_entry_continue_rcu() to allow removing
> list_for_each_continue_rcu().
Applied, thanks.
^ permalink raw reply
* Re: [PATCH net-next] ipv4/netfilter: remove unnecessary goto statement for error recovery
From: Pablo Neira Ayuso @ 2012-08-22 22:33 UTC (permalink / raw)
To: Jean Sacren; +Cc: netfilter-devel, netdev
In-Reply-To: <1345425092-11905-1-git-send-email-sakiwit@gmail.com>
On Sun, Aug 19, 2012 at 07:11:32PM -0600, Jean Sacren wrote:
> Usually it's a good practice to use goto statement for error recovery
> when initializing the module. This approach could be an overkill if:
>
> 1) there is only one fail case;
> 2) success and failure use the same return statement.
>
> For a cleaner approach, remove the unnecessary goto statement and
> directly implement error recovery.
Applied, thanks Jean.
^ permalink raw reply
* Re: [PATCH] netdev/phy: add MDIO bus multiplexer driven by a memory-mapped FPGA
From: Timur Tabi @ 2012-08-22 22:38 UTC (permalink / raw)
To: David Daney
Cc: devicetree-discuss@lists.ozlabs.org, Andy Fleming, David Miller,
netdev, david.daney
In-Reply-To: <50355C05.6000407@gmail.com>
David Daney wrote:
> I wonder if *fpga is really a good name for this. It is a general
> purpose multiplexer with a memory mapped control register. I would call
> it something like mdio-mux-mmioreg.
At one point, I thought of using mdio-mux-bitbang, but -mmioreg is better.
Thanks.
>> +- mdio-mux-device : phandle, points to the FPGA (or similar) node. This
>> + must be a memory-mapped device with 8-bit registers.
>
> You shouldn't need this. Just make the multiplexer a child of FPGA node
> to indicate where it lives.
The problem is that we don't normally consider the FPGA node to be a bus,
so its child nodes won't get probed. That's why I have this:
compatible = "mdio-mux-fpga", "mdio-mux";
^^^^^^^^
This allows me to have multiple mdio-mux parent nodes (which I do, since I
have multiple mdio bus muxes), and they all get registered and probed
properly because I also do this:
static const struct of_device_id of_device_ids[] __devinitconst = {
{
.compatible = "simple-bus"
},
{
.compatible = "fsl,srio",
},
...
{
.compatible = "mdio-mux",
},
{}
};
The .compatible = "mdio-mux" is what causes all of the mdio-mux nodes to
be registered. Therefore, it's simpler if all the mdio-mux nodes are root
nodes.
>> +
>> +- mdio-mux-register : integer, contains the offset of the register that
>> + controls the bus multiplexer.
>
> This should just be the normal "reg" properly
Ok.
>> +- mdio-mux-mask : integer, contains an 8-bit mask that specifies which
>> + bits in the register control the actual bus multiplexer. The
>> + 'reg' property of each child mdio-mux node must be constrained by
>> + this mask.
>> +
>
> "reg-mask" ??
Ok.
>
> Do you need a shift too?
The 'reg' property of the mdio bus child nodes should take the shift into
account. That's why, in the example, I have mask=0x6 and reg=0 or reg=2.
There's even code in the driver to make sure that the 'reg' values are
constrained to the mask.
--
Timur Tabi
Linux kernel developer at Freescale
^ permalink raw reply
* Re: [PATCH 06/18] netfilter: add protocol independant NAT core
From: Jan Engelhardt @ 2012-08-22 22:48 UTC (permalink / raw)
To: Patrick McHardy; +Cc: netfilter-devel, netdev
In-Reply-To: <Pine.GSO.4.63.1208230008220.5113@stinky-local.trash.net>
On Thursday 2012-08-23 00:13, Patrick McHardy wrote:
>>> + iph = (void *)skb->data + iphdroff;
>>
>> Is trying to avoid some GNU extensions a worthwhile goal? If so,
>> iph = (struct iphdr *)(skb->data + iphdroff) should be used, like in:
>
> I don't get your point.
You are doing arithmetic with a void* pointer, which is a GNU extension.
Should we try to limit "unnecessary excess usage" of GNU features?
You could do arithmetic with the char* that skb->data is:
iph = (void *)(skb->data + iphdroff);
or, if more clarity is desired, the more verbose form
iph = (struct iphdr *)(skb->data + iphdroff);
Does this seem like something worthwhile to passively pursue?
^ permalink raw reply
* Re: [PATCH] netdev/phy: add MDIO bus multiplexer driven by a memory-mapped FPGA
From: David Daney @ 2012-08-22 22:52 UTC (permalink / raw)
To: Timur Tabi
Cc: David Daney, devicetree-discuss@lists.ozlabs.org, Andy Fleming,
David Miller, netdev, david.daney
In-Reply-To: <50355F6D.4070002@freescale.com>
On 08/22/2012 03:38 PM, Timur Tabi wrote:
> David Daney wrote:
>
>> I wonder if *fpga is really a good name for this. It is a general
>> purpose multiplexer with a memory mapped control register. I would call
>> it something like mdio-mux-mmioreg.
>
> At one point, I thought of using mdio-mux-bitbang, but -mmioreg is better.
> Thanks.
>
>>> +- mdio-mux-device : phandle, points to the FPGA (or similar) node. This
>>> + must be a memory-mapped device with 8-bit registers.
>>
>> You shouldn't need this. Just make the multiplexer a child of FPGA node
>> to indicate where it lives.
>
> The problem is that we don't normally consider the FPGA node to be a bus,
> so its child nodes won't get probed. That's why I have this:
>
That would seem to be a mistake/error.
You should be able to arrive at any directly addressable register by
walking down the tree to the children and applying any "ranges"
properties at each node. The OF infrastructure will take care of
resolving all the addresses and you get rid of much of the code you
added to duplicate its function.
^ permalink raw reply
* Re: [PATCH 1/1] ipv4: ipmr_expire_timer causes crash when removing net namespace
From: Francesco Ruggeri @ 2012-08-22 22:56 UTC (permalink / raw)
To: Eric Dumazet; +Cc: netdev, Eric W. Biederman
In-Reply-To: <1345526991.5158.373.camel@edumazet-glaptop>
On Mon, Aug 20, 2012 at 10:29 PM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> On Mon, 2012-08-20 at 17:15 -0700, Francesco Ruggeri wrote:
>> I
>> +static void ipmr_free_table(struct mr_table *mrt)
>> +{
>> + del_timer(&mrt->ipmr_expire_timer);
>> + mroute_clean_tables(mrt);
>> + kfree(mrt);
>> +}
>
> Seems racy to me.
>
> del_timer() doesnt make sure timer is completely disabled.
>
> Probably need spin_lock_bh(&mfc_unres_lock) /
> spin_unlock_bh(&mfc_unres_lock), and maybe del_timer_sync()
>
I see your point, del_timer_sync should be used to take care of any
pending timers.
I am not sure about the need for further locking though. This function
simply replaces a direct call to kfree(mrt), so I assume by this point
today's code already makes sure no one is going to access this
structure (including to start the timer), or we have a bigger problem
than just the timer.
Maybe someone familiar with ipmr can comment.
Francesco
>
>
^ permalink raw reply
* [PATCH 00/10] netfilter updates for net-next (batch 1)
From: pablo @ 2012-08-22 23:38 UTC (permalink / raw)
To: netfilter-devel; +Cc: davem, netdev
From: Pablo Neira Ayuso <pablo@netfilter.org>
Hi David,
This is the first batch of Netfilter and IPVS updates for your
net-next tree. Mostly cleanups for the Netfilter side. They are:
* Remove unnecessary RTNL locking now that we have support
for namespace in nf_conntrack, from Patrick McHardy.
* Cleanup to eliminate unnecessary goto in the initialization
path of several Netfilter tables, from Jean Sacren.
* Another cleanup from Wu Fengguang, this time to PTR_RET instead
of if IS_ERR then return PTR_ERR.
* Use list_for_each_entry_continue_rcu in nf_iterate, from
Michael Wang.
* Add pmtu_disc sysctl option to disable PMTU in their tunneling
transmitter, from Julian Anastasov.
* Generalize application protocol registration in IPVS and modify
IPVS FTP helper to use it, from Julian Anastasov.
* update Kconfig. The IPVS FTP helper depends on the Netfilter FTP
helper for NAT support, from Julian Anastasov.
* Add logic to update PMTU for IPIP packets in IPVS, again
from Julian Anastasov.
* A couple of sparse warning fixes for IPVS and Netfilter from
Claudiu Ghioc and Patrick McHardy respectively.
Patrick's IPv6 NAT changes will follow after this batch, I need
to flush this batch first before refreshing my tree.
You can pull these changes from:
git://1984.lsi.us.es/nf-next
Thanks!
Claudiu Ghioc (1):
ipvs: fixed sparse warning
Jean Sacren (1):
netfilter: remove unnecessary goto statement for error recovery
Julian Anastasov (4):
ipvs: ip_vs_ftp depends on nf_conntrack_ftp helper
ipvs: generalize app registration in netns
ipvs: implement passive PMTUD for IPIP packets
ipvs: add pmtu_disc option to disable IP DF for TUN packets
Michael Wang (1):
netfilter: replace list_for_each_continue_rcu with new interface
Patrick McHardy (2):
netfilter: sparse endian fixes
netfilter: nf_conntrack: remove unnecessary RTNL locking
Wu Fengguang (1):
netfilter: PTR_RET can be used
include/net/ip_vs.h | 16 ++++--
net/bridge/netfilter/ebtable_filter.c | 4 +-
net/bridge/netfilter/ebtable_nat.c | 4 +-
net/ipv4/netfilter/iptable_filter.c | 10 +---
net/ipv4/netfilter/iptable_mangle.c | 10 +---
net/ipv4/netfilter/iptable_raw.c | 10 +---
net/ipv4/netfilter/iptable_security.c | 5 +-
net/ipv6/netfilter/ip6table_filter.c | 4 +-
net/ipv6/netfilter/ip6table_mangle.c | 4 +-
net/ipv6/netfilter/ip6table_raw.c | 4 +-
net/ipv6/netfilter/ip6table_security.c | 5 +-
net/netfilter/core.c | 10 ++--
net/netfilter/ipvs/Kconfig | 3 +-
net/netfilter/ipvs/ip_vs_app.c | 58 ++++++++++++++++------
net/netfilter/ipvs/ip_vs_core.c | 76 +++++++++++++++++++++++++++--
net/netfilter/ipvs/ip_vs_ctl.c | 16 ++++--
net/netfilter/ipvs/ip_vs_ftp.c | 21 ++------
net/netfilter/ipvs/ip_vs_xmit.c | 83 ++++++++++++++++++++++----------
net/netfilter/nf_conntrack_proto.c | 5 --
net/netfilter/nfnetlink_acct.c | 4 +-
net/netfilter/nfnetlink_cthelper.c | 2 +-
net/netfilter/xt_NFQUEUE.c | 8 +--
net/netfilter/xt_osf.c | 2 +-
23 files changed, 232 insertions(+), 132 deletions(-)
--
1.7.10.4
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox