* [PATCH 2/3] net-next: mediatek: change my email address
From: John Crispin @ 2016-05-05 9:28 UTC (permalink / raw)
To: David S. Miller
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, John Crispin,
linux-mediatek-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-kernel-u79uwXL29TY76Z2rM5mHXA, nbd-Vt+b4OUoWG0
In-Reply-To: <1462440526-52203-1-git-send-email-john-Pj+rj9U5foFAfugRpC6u6w@public.gmane.org>
The old address is no longer valid. Use the new one instead.
Signed-off-by: John Crispin <john-Pj+rj9U5foFAfugRpC6u6w@public.gmane.org>
---
drivers/net/ethernet/mediatek/mtk_eth_soc.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
index b709212..3da91df 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -7,7 +7,7 @@
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
- * Copyright (C) 2009-2016 John Crispin <blogic-p3rKhJxN3npAfugRpC6u6w@public.gmane.org>
+ * Copyright (C) 2009-2016 John Crispin <john-Pj+rj9U5foFAfugRpC6u6w@public.gmane.org>
* Copyright (C) 2009-2016 Felix Fietkau <nbd-p3rKhJxN3npAfugRpC6u6w@public.gmane.org>
* Copyright (C) 2013-2016 Michael Lee <igvtee-Re5JQEeQqe8AvxtiuMwx3w@public.gmane.org>
*/
@@ -1913,5 +1913,5 @@ static struct platform_driver mtk_driver = {
module_platform_driver(mtk_driver);
MODULE_LICENSE("GPL");
-MODULE_AUTHOR("John Crispin <blogic-p3rKhJxN3npAfugRpC6u6w@public.gmane.org>");
+MODULE_AUTHOR("John Crispin <john-Pj+rj9U5foFAfugRpC6u6w@public.gmane.org>");
MODULE_DESCRIPTION("Ethernet driver for MediaTek SoC");
--
1.7.10.4
^ permalink raw reply related
* [PATCH 3/3] net-next: lantiq: change my email address
From: John Crispin @ 2016-05-05 9:28 UTC (permalink / raw)
To: David S. Miller
Cc: netdev-u79uwXL29TY76Z2rM5mHXA, John Crispin,
linux-mediatek-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r,
linux-kernel-u79uwXL29TY76Z2rM5mHXA, nbd-Vt+b4OUoWG0
In-Reply-To: <1462440526-52203-1-git-send-email-john-Pj+rj9U5foFAfugRpC6u6w@public.gmane.org>
The old address is no longer valid. Use the new one instead.
Signed-off-by: John Crispin <john-Pj+rj9U5foFAfugRpC6u6w@public.gmane.org>
---
drivers/net/ethernet/lantiq_etop.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/lantiq_etop.c b/drivers/net/ethernet/lantiq_etop.c
index dc82b1b..fdd76b2 100644
--- a/drivers/net/ethernet/lantiq_etop.c
+++ b/drivers/net/ethernet/lantiq_etop.c
@@ -11,7 +11,7 @@
* You should have received a copy of the GNU General Public License
* along with this program; if not, see <http://www.gnu.org/licenses/>.
*
- * Copyright (C) 2011 John Crispin <blogic-p3rKhJxN3npAfugRpC6u6w@public.gmane.org>
+ * Copyright (C) 2011 John Crispin <john-Pj+rj9U5foFAfugRpC6u6w@public.gmane.org>
*/
#include <linux/kernel.h>
@@ -791,6 +791,6 @@ exit_ltq_etop(void)
module_init(init_ltq_etop);
module_exit(exit_ltq_etop);
-MODULE_AUTHOR("John Crispin <blogic-p3rKhJxN3npAfugRpC6u6w@public.gmane.org>");
+MODULE_AUTHOR("John Crispin <john-Pj+rj9U5foFAfugRpC6u6w@public.gmane.org>");
MODULE_DESCRIPTION("Lantiq SoC ETOP");
MODULE_LICENSE("GPL");
--
1.7.10.4
^ permalink raw reply related
* RE: [v10, 7/7] mmc: sdhci-of-esdhc: fix host version for T4240-R1.0-R2.0
From: Yangbo Lu @ 2016-05-05 9:41 UTC (permalink / raw)
To: Arnd Bergmann, Scott Wood,
linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org
Cc: Mark Rutland, devicetree-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
ulf.hansson-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org, Russell King,
Bhupesh Sharma, netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
Kumar Gala, linux-mmc-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Yang-Leo Li,
iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org,
Rob Herring, linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
Claudiu Manoil, Santosh Shilimkar, Xiaobo Xie,
linux-clk-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org
In-Reply-To: <4448030.Zs8fVnr3nm@wuerfel>
Hi Arnd,
> -----Original Message-----
> From: Arnd Bergmann [mailto:arnd-r2nGTMty4D4@public.gmane.org]
> Sent: Thursday, May 05, 2016 4:32 PM
> To: linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org
> Cc: Yangbo Lu; linux-mmc-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; devicetree-u79uwXL29TY76Z2rM5mHXA@public.gmane.org;
> linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org; linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org;
> linux-clk-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; iommu-cunTk1MwBs/ROKNJybVBZg@public.gmane.org
> foundation.org; netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; Mark Rutland;
> ulf.hansson-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org; Russell King; Bhupesh Sharma; Joerg Roedel;
> Santosh Shilimkar; Yang-Leo Li; Scott Wood; Rob Herring; Claudiu Manoil;
> Kumar Gala; Xiaobo Xie; Qiang Zhao
> Subject: Re: [v10, 7/7] mmc: sdhci-of-esdhc: fix host version for T4240-
> R1.0-R2.0
>
> On Thursday 05 May 2016 11:12:30 Yangbo Lu wrote:
> >
> > + fsl_guts_init();
> > + svr = fsl_guts_get_svr();
> > + if (svr) {
> > + esdhc->soc_ver = SVR_SOC_VER(svr);
> > + esdhc->soc_rev = SVR_REV(svr);
> > + } else {
> > + dev_err(&pdev->dev, "Failed to get SVR value!\n");
> > + }
> > +
> >
>
>
> Sorry for jumping in again after not participating in the discussion for
> the past few versions.
>
> What happened to my suggestion of making this a platform-independent
> interface to avoid the link time dependency?
>
> Specifically, why not add an exported function to drivers/base/soc.c that
> uses glob_match() for comparing a string in the device driver to the ID
> of the SoC that is set by whatever SoC identifying driver the platform
> has?
>
> Arnd
[Lu Yangbo-B47093] I think this has been discussed in v6.
You can find Scott's comments about this in below link.
https://patchwork.kernel.org/patch/8544501/
Thanks.
^ permalink raw reply
* [PATCH 10/12] net-next: mediatek: fix off by one in the TX ring allocation
From: John Crispin @ 2016-05-05 9:26 UTC (permalink / raw)
To: David S. Miller; +Cc: nbd, netdev, linux-mediatek, linux-kernel, John Crispin
In-Reply-To: <1462440385-51939-1-git-send-email-john@phrozen.org>
The TX ring setup has an off by one error causing it to not utilise all
descriptors. This has the side effect that we need to reset the next
pointer at runtime to make it work. Fix the off by one and remove the
code fixing the ring at runtime.
Signed-off-by: John Crispin <john@phrozen.org>
---
drivers/net/ethernet/mediatek/mtk_eth_soc.c | 3 +--
1 file changed, 1 insertion(+), 2 deletions(-)
diff --git a/drivers/net/ethernet/mediatek/mtk_eth_soc.c b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
index 12e9f1a..81155a1 100644
--- a/drivers/net/ethernet/mediatek/mtk_eth_soc.c
+++ b/drivers/net/ethernet/mediatek/mtk_eth_soc.c
@@ -919,7 +919,6 @@ static int mtk_poll_tx(struct mtk_eth *eth, int budget, bool *tx_again)
}
mtk_tx_unmap(eth->dev, tx_buf);
- ring->last_free->txd2 = next_cpu;
ring->last_free = desc;
atomic_inc(&ring->free_count);
@@ -1027,7 +1026,7 @@ static int mtk_tx_alloc(struct mtk_eth *eth)
atomic_set(&ring->free_count, MTK_DMA_SIZE - 2);
ring->next_free = &ring->dma[0];
- ring->last_free = &ring->dma[MTK_DMA_SIZE - 2];
+ ring->last_free = &ring->dma[MTK_DMA_SIZE - 1];
ring->thresh = MAX_SKB_FRAGS;
/* make sure that all changes to the dma ring are flushed before we
--
1.7.10.4
^ permalink raw reply related
* Re: [PATCH net-next v2] macvtap: add namespace support to the sysfs device class
From: Marc Angel @ 2016-05-05 10:12 UTC (permalink / raw)
To: David Miller; +Cc: netdev, Eric W. Biederman
In-Reply-To: <20160504.235308.1088954708869507214.davem@davemloft.net>
On Thu, May 5, 2016 at 4:53 AM, David Miller <davem@davemloft.net> wrote:
> From: Marc Angel <marc@arista.com>
> Date: Thu, 5 May 2016 01:33:09 +0100
>
>> On Wed, May 4, 2016 at 9:04 PM, David Miller <davem@davemloft.net> wrote:
>>> From: Marc Angel <marc@arista.com>
>>> Date: Tue, 3 May 2016 20:30:54 +0200
>>>
>>>> @@ -1274,6 +1285,7 @@ static int macvtap_device_event(struct notifier_block *unused,
>>>> unsigned long event, void *ptr)
>>>> {
>>>> struct net_device *dev = netdev_notifier_info_to_dev(ptr);
>>>> + const char *tap_name = kasprintf(GFP_KERNEL, "tap%d", dev->ifindex);
>>>> struct macvlan_dev *vlan;
>>>> struct device *classdev;
>>>> dev_t devt;
>>>
>>> This 'tap_name' buffer seems to be leaked in several code paths.
>>
>> Sorry about that... V3 should be better.
>
> V3 doesn't apply cleanly to net-next, please respin your patch against
> the current tree.
>
> Thanks.
I wasn't up-to-date, sorry again about that...
Thanks for being so patient.
^ permalink raw reply
* [PATCH net-next v4] macvtap: add namespace support to the sysfs device class
From: Marc Angel @ 2016-05-05 10:14 UTC (permalink / raw)
To: netdev; +Cc: ebiederm
In-Reply-To: <20160504.235308.1088954708869507214.davem@davemloft.net>
When creating macvtaps that are expected to have the same ifindex
in different network namespaces, only the first one will succeed.
The others will fail with a sysfs_warn_dup warning due to them trying
to create the following sysfs link (with 'NN' the ifindex of macvtapX):
/sys/class/macvtap/tapNN -> /sys/devices/virtual/net/macvtapX/tapNN
This is reproducible by running the following commands:
ip netns add ns1
ip netns add ns2
ip link add veth0 type veth peer name veth1
ip link set veth0 netns ns1
ip link set veth1 netns ns2
ip netns exec ns1 ip l add link veth0 macvtap0 type macvtap
ip netns exec ns2 ip l add link veth1 macvtap1 type macvtap
The last command will fail with "RTNETLINK answers: File exists" (along
with the kernel warning) but retrying it will work because the ifindex
was incremented.
The 'net' device class is isolated between network namespaces so each
one has its own hierarchy of net devices.
This isn't the case for the 'macvtap' device class.
The problem occurs half-way through the netdev registration, when
`macvtap_device_event` is called-back to create the 'tapNN' macvtap
class device under the 'macvtapX' net class device.
This patch adds namespace support to the 'macvtap' device class so
that /sys/class/macvtap is no longer shared between net namespaces.
However, making the macvtap sysfs class namespace-aware has the side
effect of changing /sys/devices/virtual/net/macvtapX/tapNN into
/sys/devices/virtual/net/macvtapX/macvtap/tapNN.
This is due to Commit 24b1442 ("Driver-core: Always create class
directories for classses that support namespaces") and the fact that
class devices supporting namespaces are really not supposed to be placed
directly under other class devices.
To avoid breaking userland, a tapNN symlink pointing to macvtap/tapNN is
created inside the macvtapX directory.
Signed-off-by: Marc Angel <marc@arista.com>
---
drivers/net/macvtap.c | 36 ++++++++++++++++++++++++++----------
1 file changed, 26 insertions(+), 10 deletions(-)
diff --git a/drivers/net/macvtap.c b/drivers/net/macvtap.c
index 74cb15a..22b85b0 100644
--- a/drivers/net/macvtap.c
+++ b/drivers/net/macvtap.c
@@ -129,7 +129,18 @@ static DEFINE_MUTEX(minor_lock);
static DEFINE_IDR(minor_idr);
#define GOODCOPY_LEN 128
-static struct class *macvtap_class;
+static const void *macvtap_net_namespace(struct device *d)
+{
+ struct net_device *dev = to_net_dev(d->parent);
+ return dev_net(dev);
+}
+
+static struct class macvtap_class = {
+ .name = "macvtap",
+ .owner = THIS_MODULE,
+ .ns_type = &net_ns_type_operations,
+ .namespace = macvtap_net_namespace,
+};
static struct cdev macvtap_cdev;
static const struct proto_ops macvtap_socket_ops;
@@ -1278,10 +1289,12 @@ static int macvtap_device_event(struct notifier_block *unused,
struct device *classdev;
dev_t devt;
int err;
+ char tap_name[IFNAMSIZ];
if (dev->rtnl_link_ops != &macvtap_link_ops)
return NOTIFY_DONE;
+ snprintf(tap_name, IFNAMSIZ, "tap%d", dev->ifindex);
vlan = netdev_priv(dev);
switch (event) {
@@ -1295,19 +1308,24 @@ static int macvtap_device_event(struct notifier_block *unused,
return notifier_from_errno(err);
devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
- classdev = device_create(macvtap_class, &dev->dev, devt,
- dev, "tap%d", dev->ifindex);
+ classdev = device_create(&macvtap_class, &dev->dev, devt,
+ dev, tap_name);
if (IS_ERR(classdev)) {
macvtap_free_minor(vlan);
return notifier_from_errno(PTR_ERR(classdev));
}
+ err = sysfs_create_link(&dev->dev.kobj, &classdev->kobj,
+ tap_name);
+ if (err)
+ return notifier_from_errno(err);
break;
case NETDEV_UNREGISTER:
/* vlan->minor == 0 if NETDEV_REGISTER above failed */
if (vlan->minor == 0)
break;
+ sysfs_remove_link(&dev->dev.kobj, tap_name);
devt = MKDEV(MAJOR(macvtap_major), vlan->minor);
- device_destroy(macvtap_class, devt);
+ device_destroy(&macvtap_class, devt);
macvtap_free_minor(vlan);
break;
}
@@ -1333,11 +1351,9 @@ static int macvtap_init(void)
if (err)
goto out2;
- macvtap_class = class_create(THIS_MODULE, "macvtap");
- if (IS_ERR(macvtap_class)) {
- err = PTR_ERR(macvtap_class);
+ err = class_register(&macvtap_class);
+ if (err)
goto out3;
- }
err = register_netdevice_notifier(&macvtap_notifier_block);
if (err)
@@ -1352,7 +1368,7 @@ static int macvtap_init(void)
out5:
unregister_netdevice_notifier(&macvtap_notifier_block);
out4:
- class_unregister(macvtap_class);
+ class_unregister(&macvtap_class);
out3:
cdev_del(&macvtap_cdev);
out2:
@@ -1366,7 +1382,7 @@ static void macvtap_exit(void)
{
rtnl_link_unregister(&macvtap_link_ops);
unregister_netdevice_notifier(&macvtap_notifier_block);
- class_unregister(macvtap_class);
+ class_unregister(&macvtap_class);
cdev_del(&macvtap_cdev);
unregister_chrdev_region(macvtap_major, MACVTAP_NUM_DEVS);
idr_destroy(&minor_idr);
--
2.8.0
^ permalink raw reply related
* RE: [PATCH 2/2] mac80211_hwsim: Allow managing radios from non-initial namespaces
From: David Laight @ 2016-05-05 10:16 UTC (permalink / raw)
To: 'Martin Willi', Johannes Berg
Cc: linux-wireless@vger.kernel.org, netdev@vger.kernel.org
In-Reply-To: <1462258398-6749-3-git-send-email-martin@strongswan.org>
From: Martin Willi
> Sent: 03 May 2016 07:53
> While wiphys can be moved into network namespaces over nl80211, the
> creation and removal of hwsim radios is currently limited to the initial
> namespace. This patch allows management of namespaced radios from the
> owning namespace by setting genetlink netnsok.
>
> To prevent two arbitrary namespaces from communicating over the simulated
> shared medium, radios are separated by netgroups. Each radio created in
> the same namespace lives in the same netgroup and hence can communicate
> with other radios in that group. When moving radios to other namespaces,
> the netgroup is preserved, so two radios having the same netgroup can
> communicate even if not in the same namespace; This allows a controlling
> namespace to create radios and move them to other namespaces for
> communication.
>
...
> + data->netgroup = *(int *)net_generic(net, hwsim_net_id);
Anything doing *(integer_type *) rings alarm bells.
I suspect you should be defining a structure that currently contains
one integer member.
Something (maybe a compile time assert) needs to check that buffer
space you are accessing (where ever it is) is large enough.
David
^ permalink raw reply
* Re: [v10, 7/7] mmc: sdhci-of-esdhc: fix host version for T4240-R1.0-R2.0
From: Arnd Bergmann @ 2016-05-05 11:10 UTC (permalink / raw)
To: linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r
Cc: Mark Rutland, ulf.hansson-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org,
Xiaobo Xie, linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-clk-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Qiang Zhao,
Russell King, Bhupesh Sharma, Claudiu Manoil,
devicetree-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Yangbo Lu,
Scott Wood, Rob Herring, Santosh Shilimkar,
netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-mmc-u79uwXL29TY76Z2rM5mHXA@public.gmane.org,
linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org, Yang-Leo Li,
iommu-cunTk1MwBs9QetFLy7KEm3xJsTq8ys+cHZ5vskTnxNA@public.gmane.org,
Kumar
In-Reply-To: <HE1PR04MB0889B52134A43C0E7A452B7BF87C0-6LN7OEpIatX1kPMWxTxe+c9NdZoXdze2vxpqHgZTriW3zl9H0oFU5g@public.gmane.org>
On Thursday 05 May 2016 09:41:32 Yangbo Lu wrote:
> > -----Original Message-----
> > From: Arnd Bergmann [mailto:arnd-r2nGTMty4D4@public.gmane.org]
> > Sent: Thursday, May 05, 2016 4:32 PM
> > To: linuxppc-dev-uLR06cmDAlY/bJ5BZ2RsiQ@public.gmane.org
> > Cc: Yangbo Lu; linux-mmc-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; devicetree-u79uwXL29TY76Z2rM5mHXA@public.gmane.org;
> > linux-arm-kernel-IAPFreCvJWM7uuMidbF8XUB+6BGkLq7r@public.gmane.org; linux-kernel-u79uwXL29TY76Z2rM5mHXA@public.gmane.org;
> > linux-clk-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; linux-i2c-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; iommu-cunTk1MwBs/ROKNJybVBZg@public.gmane.org
> > foundation.org; netdev-u79uwXL29TY76Z2rM5mHXA@public.gmane.org; Mark Rutland;
> > ulf.hansson-QSEj5FYQhm4dnm+yROfE0A@public.gmane.org; Russell King; Bhupesh Sharma; Joerg Roedel;
> > Santosh Shilimkar; Yang-Leo Li; Scott Wood; Rob Herring; Claudiu Manoil;
> > Kumar Gala; Xiaobo Xie; Qiang Zhao
> > Subject: Re: [v10, 7/7] mmc: sdhci-of-esdhc: fix host version for T4240-
> > R1.0-R2.0
> >
> > On Thursday 05 May 2016 11:12:30 Yangbo Lu wrote:
> > >
> > > + fsl_guts_init();
> > > + svr = fsl_guts_get_svr();
> > > + if (svr) {
> > > + esdhc->soc_ver = SVR_SOC_VER(svr);
> > > + esdhc->soc_rev = SVR_REV(svr);
> > > + } else {
> > > + dev_err(&pdev->dev, "Failed to get SVR value!\n");
> > > + }
> > > +
> > >
> >
> >
> > Sorry for jumping in again after not participating in the discussion for
> > the past few versions.
> >
> > What happened to my suggestion of making this a platform-independent
> > interface to avoid the link time dependency?
> >
> > Specifically, why not add an exported function to drivers/base/soc.c that
> > uses glob_match() for comparing a string in the device driver to the ID
> > of the SoC that is set by whatever SoC identifying driver the platform
> > has?
>
> [Lu Yangbo-B47093] I think this has been discussed in v6.
> You can find Scott's comments about this in below link.
> https://patchwork.kernel.org/patch/8544501/
Ah, thanks for bearing with me and digging this out again. Let me follow
up on Scott's older replies here then:
> >> IIRC, it is the same IP block as i.MX and Arnd's point is this won't
> >> even compile on !PPC. It is things like this that prevent sharing the
> >> driver.
>
> The whole point of using the MMIO SVR instead of the PPC SPR is so that
> it will work on ARM... The guts driver should build on any platform as
> long as OF is enabled, and if it doesn't find a node to bind to it will
> return 0 for SVR, and the eSDHC driver will continue (after printing an
> error that should be removed) without the ability to test for errata
> based on SVR.
It feels like a bad design to have to come up with a different
method for each SoC type here when they all do the same thing
and want to identify some variant of the chip to do device
specific quirks.
As far as I'm concerned, every driver in drivers/soc that needs to
export a symbol to be used by a device driver is an indication that
we don't have the right set of abstractions yet. There are cases
that are not worth abstracting because the functionality is rather
obscure and only a couple of drivers for one particular chip
ever need it.
Finding out the version of the SoC does not look like this case.
> > I think the first four patches take care of building for ARM,
> > but the problem remains if you want to enable COMPILE_TEST as
> > we need for certain automated checking.
>
> What specific problem is there with COMPILE_TEST?
COMPILE_TEST is solvable here and the way it is implemented in this
case (selecting FSL_GUTS from the driver) indeed looks like it works
correctly, but it's still awkward that this means building the
SoC specific ID stuff into the vmlinux binary for any driver that
uses something like that for a particular SoC.
> >> Dealing with Si revs is a common problem. We should have a
> >> common solution. There is soc_device for this purpose.
> >
> > Exactly. The last time this came up, I think we agreed to implement a
> > helper using glob_match() on the soc_device strings. Unfortunately
> > this hasn't happened then, but I'd still prefer that over yet another
> > vendor-specific way of dealing with the generic issue.
>
> soc_device would require encoding the SVR as a string and then decoding
> the string, which is more complicated and error prone than having
> platform-specific code test a platform-specific number.
You already need to encode it as a string to register the soc_device,
and the driver just needs to pass a glob string, so the only part that
is missing is the generic function that takes the string from the
driver and passes that to glob_match for the soc_device.
> And when would it get registered on arm64, which doesn't have
> platform code?
Whenever the soc driver is loaded, as is the case now. The match
function can return -EPROBE_DEFER if no SoC device is registered
yet.
Arnd
^ permalink raw reply
* Re: A couple of questions about the SKB fragments
From: Edward Cree @ 2016-05-05 11:13 UTC (permalink / raw)
To: matvejchikov, netdev; +Cc: Aleksey.Baulin
In-Reply-To: <CAKh5naYrZtf5y_Qw93G+GbxRVucsk4nbuiaguHcghtKEL=yQBw@mail.gmail.com>
On 05/05/16 08:40, Ilya Matveychikov wrote:
> Hello folks,
>
> While working with fragmented SKBs we've got stuck with the following:
> - is it possible for an SKB fragment in skb_shinfo(skb)->frag_list to
> be fragmented too (i.e. to have SKBs in frag_list)?
> - do skb->len and skb->data_len contain the whole SKB length,
> including the length of all fragments (not only the paged parts)?
>
> Is there any docs except the kernel sources itself to refer to?
davem has some docs up at http://vger.kernel.org/~davem/skb.html and
http://vger.kernel.org/~davem/skb_data.html
In particular note the following:
"The frag_list is used to maintain a chain of SKBs organized for
fragmentation purposes, it is _not_ used for maintaining paged data."
So my reading would suggest there is no way to multiple-layer-fragment
an SKB; the frags are page pointers and offsets, not entire sk_buff
structs in their own right.
-Ed
^ permalink raw reply
* [PATCH net-next 0/4] xen-netback: support for control ring
From: Paul Durrant @ 2016-05-05 11:19 UTC (permalink / raw)
To: xen-devel, netdev; +Cc: Paul Durrant
My recent patch to import an up-to-date include/xen/interface/io/netif.h
from the Xen Project brought in the necessary definitions to support the
new control shared ring and protocol. This patch series updates xen-netback
to support the new ring.
Patch #1 adds the necessary boilerplate to map the control ring and handle
messages. No implementation of the new protocol is included in this patch
so that it can be kept to a reasonable size.
Patch #2 adds the protocol implementation.
Patch #3 adds support for passing has values calculated by xen-netback to
capable frontends.
Patch #4 adds support for accepting hash values calculated by capable
frontends and using them the set the socket buffer hash.
^ permalink raw reply
* [PATCH net-next 4/4] xen-netback: use hash value from the frontend
From: Paul Durrant @ 2016-05-05 11:19 UTC (permalink / raw)
To: xen-devel, netdev; +Cc: Paul Durrant, Wei Liu
In-Reply-To: <1462447170-1815-1-git-send-email-paul.durrant@citrix.com>
My recent patch to include/xen/interface/io/netif.h defines a new extra
info type that can be used to pass hash values between backend and guest
frontend.
This patch adds code to xen-netback to use the value in a hash extra
info fragment passed from the guest frontend in a transmit-side
(i.e. netback receive side) packet to set the skb hash accordingly.
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
---
drivers/net/xen-netback/netback.c | 27 +++++++++++++++++++++++++++
1 file changed, 27 insertions(+)
diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c
index 7c72510..a5b5aad 100644
--- a/drivers/net/xen-netback/netback.c
+++ b/drivers/net/xen-netback/netback.c
@@ -1509,6 +1509,33 @@ static void xenvif_tx_build_gops(struct xenvif_queue *queue,
}
}
+ if (extras[XEN_NETIF_EXTRA_TYPE_HASH - 1].type) {
+ struct xen_netif_extra_info *extra;
+ enum pkt_hash_types type = PKT_HASH_TYPE_NONE;
+
+ extra = &extras[XEN_NETIF_EXTRA_TYPE_HASH - 1];
+
+ switch (extra->u.hash.type) {
+ case _XEN_NETIF_CTRL_HASH_TYPE_IPV4:
+ case _XEN_NETIF_CTRL_HASH_TYPE_IPV6:
+ type = PKT_HASH_TYPE_L3;
+ break;
+
+ case _XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP:
+ case _XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP:
+ type = PKT_HASH_TYPE_L4;
+ break;
+
+ default:
+ break;
+ }
+
+ if (type != PKT_HASH_TYPE_NONE)
+ skb_set_hash(skb,
+ *(u32 *)extra->u.hash.value,
+ type);
+ }
+
XENVIF_TX_CB(skb)->pending_idx = pending_idx;
__skb_put(skb, data_len);
--
2.1.4
^ permalink raw reply related
* [PATCH net-next 3/4] xen-netback: pass hash value to the frontend
From: Paul Durrant @ 2016-05-05 11:19 UTC (permalink / raw)
To: xen-devel, netdev; +Cc: Paul Durrant, Wei Liu
In-Reply-To: <1462447170-1815-1-git-send-email-paul.durrant@citrix.com>
My recent patch to include/xen/interface/io/netif.h defines a new extra
info type that can be used to pass hash values between backend and guest
frontend.
This patch adds code to xen-netback to pass hash values calculated for
guest receive-side packets (i.e. netback transmit side) to the frontend.
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
---
drivers/net/xen-netback/interface.c | 13 ++++++-
drivers/net/xen-netback/netback.c | 78 +++++++++++++++++++++++++++++++------
2 files changed, 77 insertions(+), 14 deletions(-)
diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
index e54b475..b2d945f 100644
--- a/drivers/net/xen-netback/interface.c
+++ b/drivers/net/xen-netback/interface.c
@@ -158,8 +158,17 @@ static u16 xenvif_select_queue(struct net_device *dev, struct sk_buff *skb,
struct xenvif *vif = netdev_priv(dev);
unsigned int size = vif->hash.size;
- if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE)
- return fallback(dev, skb) % dev->real_num_tx_queues;
+ if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE) {
+ u16 index = fallback(dev, skb) % dev->real_num_tx_queues;
+
+ /* Make sure there is no hash information in the socket
+ * buffer otherwise it would be incorrectly forwarded
+ * to the frontend.
+ */
+ skb_clear_hash(skb);
+
+ return index;
+ }
xenvif_set_skb_hash(vif, skb);
diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c
index 6509d11..7c72510 100644
--- a/drivers/net/xen-netback/netback.c
+++ b/drivers/net/xen-netback/netback.c
@@ -168,6 +168,8 @@ static bool xenvif_rx_ring_slots_available(struct xenvif_queue *queue)
needed = DIV_ROUND_UP(skb->len, XEN_PAGE_SIZE);
if (skb_is_gso(skb))
needed++;
+ if (skb->sw_hash)
+ needed++;
do {
prod = queue->rx.sring->req_prod;
@@ -285,6 +287,8 @@ struct gop_frag_copy {
struct xenvif_rx_meta *meta;
int head;
int gso_type;
+ int protocol;
+ int hash_present;
struct page *page;
};
@@ -331,8 +335,15 @@ static void xenvif_setup_copy_gop(unsigned long gfn,
npo->copy_off += *len;
info->meta->size += *len;
+ if (!info->head)
+ return;
+
/* Leave a gap for the GSO descriptor. */
- if (info->head && ((1 << info->gso_type) & queue->vif->gso_mask))
+ if ((1 << info->gso_type) & queue->vif->gso_mask)
+ queue->rx.req_cons++;
+
+ /* Leave a gap for the hash extra segment. */
+ if (info->hash_present)
queue->rx.req_cons++;
info->head = 0; /* There must be something in this buffer now */
@@ -367,6 +378,11 @@ static void xenvif_gop_frag_copy(struct xenvif_queue *queue, struct sk_buff *skb
.npo = npo,
.head = *head,
.gso_type = XEN_NETIF_GSO_TYPE_NONE,
+ /* xenvif_set_skb_hash() will have either set a s/w
+ * hash or cleared the hash depending on
+ * whether the the frontend wants a hash for this skb.
+ */
+ .hash_present = skb->sw_hash,
};
unsigned long bytes;
@@ -555,6 +571,7 @@ void xenvif_kick_thread(struct xenvif_queue *queue)
static void xenvif_rx_action(struct xenvif_queue *queue)
{
+ struct xenvif *vif = queue->vif;
s8 status;
u16 flags;
struct xen_netif_rx_response *resp;
@@ -590,9 +607,10 @@ static void xenvif_rx_action(struct xenvif_queue *queue)
gnttab_batch_copy(queue->grant_copy_op, npo.copy_prod);
while ((skb = __skb_dequeue(&rxq)) != NULL) {
+ struct xen_netif_extra_info *extra = NULL;
if ((1 << queue->meta[npo.meta_cons].gso_type) &
- queue->vif->gso_prefix_mask) {
+ vif->gso_prefix_mask) {
resp = RING_GET_RESPONSE(&queue->rx,
queue->rx.rsp_prod_pvt++);
@@ -610,7 +628,7 @@ static void xenvif_rx_action(struct xenvif_queue *queue)
queue->stats.tx_bytes += skb->len;
queue->stats.tx_packets++;
- status = xenvif_check_gop(queue->vif,
+ status = xenvif_check_gop(vif,
XENVIF_RX_CB(skb)->meta_slots_used,
&npo);
@@ -632,21 +650,57 @@ static void xenvif_rx_action(struct xenvif_queue *queue)
flags);
if ((1 << queue->meta[npo.meta_cons].gso_type) &
- queue->vif->gso_mask) {
- struct xen_netif_extra_info *gso =
- (struct xen_netif_extra_info *)
+ vif->gso_mask) {
+ extra = (struct xen_netif_extra_info *)
RING_GET_RESPONSE(&queue->rx,
queue->rx.rsp_prod_pvt++);
resp->flags |= XEN_NETRXF_extra_info;
- gso->u.gso.type = queue->meta[npo.meta_cons].gso_type;
- gso->u.gso.size = queue->meta[npo.meta_cons].gso_size;
- gso->u.gso.pad = 0;
- gso->u.gso.features = 0;
+ extra->u.gso.type = queue->meta[npo.meta_cons].gso_type;
+ extra->u.gso.size = queue->meta[npo.meta_cons].gso_size;
+ extra->u.gso.pad = 0;
+ extra->u.gso.features = 0;
+
+ extra->type = XEN_NETIF_EXTRA_TYPE_GSO;
+ extra->flags = 0;
+ }
+
+ if (skb->sw_hash) {
+ /* Since the skb got here via xenvif_select_queue()
+ * we know that the hash has been re-calculated
+ * according to a configuration set by the frontend
+ * and therefore we know that it is legitimate to
+ * pass it to the frontend.
+ */
+ if (resp->flags & XEN_NETRXF_extra_info)
+ extra->flags |= XEN_NETIF_EXTRA_FLAG_MORE;
+ else
+ resp->flags |= XEN_NETRXF_extra_info;
+
+ extra = (struct xen_netif_extra_info *)
+ RING_GET_RESPONSE(&queue->rx,
+ queue->rx.rsp_prod_pvt++);
- gso->type = XEN_NETIF_EXTRA_TYPE_GSO;
- gso->flags = 0;
+ extra->u.hash.algorithm =
+ XEN_NETIF_CTRL_HASH_ALGORITHM_TOEPLITZ;
+
+ if (skb->l4_hash)
+ extra->u.hash.type =
+ skb->protocol == htons(ETH_P_IP) ?
+ _XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP :
+ _XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP;
+ else
+ extra->u.hash.type =
+ skb->protocol == htons(ETH_P_IP) ?
+ _XEN_NETIF_CTRL_HASH_TYPE_IPV4 :
+ _XEN_NETIF_CTRL_HASH_TYPE_IPV6;
+
+ *(uint32_t *)extra->u.hash.value =
+ skb_get_hash_raw(skb);
+
+ extra->type = XEN_NETIF_EXTRA_TYPE_HASH;
+ extra->flags = 0;
}
xenvif_add_frag_responses(queue, status,
--
2.1.4
^ permalink raw reply related
* [PATCH net-next 2/4] xen-netback: add control protocol implementation
From: Paul Durrant @ 2016-05-05 11:19 UTC (permalink / raw)
To: xen-devel, netdev; +Cc: Paul Durrant, Wei Liu
In-Reply-To: <1462447170-1815-1-git-send-email-paul.durrant@citrix.com>
My recent patch to include/xen/interface/io/netif.h defines a new shared
ring (in addition to the rx and tx rings) for passing control messages
from a VM frontend driver to a backend driver.
A previous patch added the necessary boilerplate for mapping the control
ring from the frontend, should it be created. This patch adds
implementations for each of the defined protocol messages.
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
---
drivers/net/xen-netback/Makefile | 2 +-
drivers/net/xen-netback/common.h | 43 +++++
drivers/net/xen-netback/hash.c | 361 ++++++++++++++++++++++++++++++++++++
drivers/net/xen-netback/interface.c | 28 +++
drivers/net/xen-netback/netback.c | 49 ++++-
5 files changed, 480 insertions(+), 3 deletions(-)
create mode 100644 drivers/net/xen-netback/hash.c
diff --git a/drivers/net/xen-netback/Makefile b/drivers/net/xen-netback/Makefile
index e346e81..11e02be 100644
--- a/drivers/net/xen-netback/Makefile
+++ b/drivers/net/xen-netback/Makefile
@@ -1,3 +1,3 @@
obj-$(CONFIG_XEN_NETDEV_BACKEND) := xen-netback.o
-xen-netback-y := netback.o xenbus.o interface.o
+xen-netback-y := netback.o xenbus.o interface.o hash.o
diff --git a/drivers/net/xen-netback/common.h b/drivers/net/xen-netback/common.h
index 093a12a..4959716 100644
--- a/drivers/net/xen-netback/common.h
+++ b/drivers/net/xen-netback/common.h
@@ -220,6 +220,32 @@ struct xenvif_mcast_addr {
#define XEN_NETBK_MCAST_MAX 64
+#define XEN_NETBK_MAX_HASH_KEY_SIZE 40
+#define XEN_NETBK_MAX_HASH_MAPPING_SIZE 128
+#define XEN_NETBK_HASH_TAG_SIZE 40
+
+struct xenvif_hash_cache_entry {
+ u8 tag[XEN_NETBK_HASH_TAG_SIZE];
+ unsigned int len;
+ u32 val;
+ int seq;
+};
+
+struct xenvif_hash_cache {
+ rwlock_t lock;
+ struct xenvif_hash_cache_entry *entry;
+ atomic_t seq;
+};
+
+struct xenvif_hash {
+ unsigned int alg;
+ u32 flags;
+ u8 key[XEN_NETBK_MAX_HASH_KEY_SIZE];
+ u32 mapping[XEN_NETBK_MAX_HASH_MAPPING_SIZE];
+ unsigned int size;
+ struct xenvif_hash_cache cache;
+};
+
struct xenvif {
/* Unique identifier for this interface. */
domid_t domid;
@@ -251,6 +277,8 @@ struct xenvif {
unsigned int num_queues; /* active queues, resource allocated */
unsigned int stalled_queues;
+ struct xenvif_hash hash;
+
struct xenbus_watch credit_watch;
struct xenbus_watch mcast_ctrl_watch;
@@ -353,6 +381,7 @@ extern bool separate_tx_rx_irq;
extern unsigned int rx_drain_timeout_msecs;
extern unsigned int rx_stall_timeout_msecs;
extern unsigned int xenvif_max_queues;
+extern unsigned int xenvif_hash_cache_size;
#ifdef CONFIG_DEBUG_FS
extern struct dentry *xen_netback_dbg_root;
@@ -366,4 +395,18 @@ void xenvif_skb_zerocopy_complete(struct xenvif_queue *queue);
bool xenvif_mcast_match(struct xenvif *vif, const u8 *addr);
void xenvif_mcast_addr_list_free(struct xenvif *vif);
+/* Hash */
+int xenvif_init_hash(struct xenvif *vif);
+void xenvif_deinit_hash(struct xenvif *vif);
+
+u32 xenvif_set_hash_alg(struct xenvif *vif, u32 alg);
+u32 xenvif_get_hash_flags(struct xenvif *vif, u32 *flags);
+u32 xenvif_set_hash_flags(struct xenvif *vif, u32 flags);
+u32 xenvif_set_hash_key(struct xenvif *vif, u32 gref, u32 len);
+u32 xenvif_set_hash_mapping_size(struct xenvif *vif, u32 size);
+u32 xenvif_set_hash_mapping(struct xenvif *vif, u32 gref, u32 len,
+ u32 off);
+
+void xenvif_set_skb_hash(struct xenvif *vif, struct sk_buff *skb);
+
#endif /* __XEN_NETBACK__COMMON_H__ */
diff --git a/drivers/net/xen-netback/hash.c b/drivers/net/xen-netback/hash.c
new file mode 100644
index 0000000..054bfd9
--- /dev/null
+++ b/drivers/net/xen-netback/hash.c
@@ -0,0 +1,361 @@
+/*
+ * Copyright (c) 2016 Citrix Systems Inc.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of the GNU General Public License version 2
+ * as published by the Free Softare Foundation; or, when distributed
+ * separately from the Linux kernel or incorporated into other
+ * software packages, subject to the following license:
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a copy
+ * of this source file (the "Software"), to deal in the Software without
+ * restriction, including without limitation the rights to use, copy, modify,
+ * merge, publish, distribute, sublicense, and/or sell copies of the Software,
+ * and to permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * The above copyright notice and this permission notice shall be included in
+ * all copies or substantial portions of the Software.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+ * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+ * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
+ * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
+ * IN THE SOFTWARE.
+ */
+
+#define XEN_NETIF_DEFINE_TOEPLITZ
+
+#include "common.h"
+#include <linux/vmalloc.h>
+
+static u32 xenvif_new_hash(struct xenvif *vif, const u8 *data,
+ unsigned int len)
+{
+ struct xenvif_hash_cache_entry new, *entry, *oldest = NULL;
+ unsigned long flags;
+ u32 val;
+ unsigned int i;
+
+ val = xen_netif_toeplitz_hash(vif->hash.key,
+ sizeof(vif->hash.key),
+ data, len);
+
+ if (xenvif_hash_cache_size == 0)
+ goto out;
+
+ write_lock_irqsave(&vif->hash.cache.lock, flags);
+
+ /* Create a new cache entry */
+ memcpy(new.tag, data, len);
+ new.len = len;
+ new.val = val;
+ new.seq = atomic_inc_return(&vif->hash.cache.seq);
+
+ /* Find the oldest entry in the cache. */
+ for (i = 0; i < xenvif_hash_cache_size; i++) {
+ entry = &vif->hash.cache.entry[i];
+
+ if (!oldest ||
+ (new.seq - entry->seq > new.seq - oldest->seq))
+ oldest = entry;
+ }
+
+ /* Replace that entry with the new one. */
+ *oldest = new;
+
+ write_unlock_irqrestore(&vif->hash.cache.lock, flags);
+
+out:
+ return val;
+}
+
+static void xenvif_flush_hash(struct xenvif *vif)
+{
+ unsigned long flags;
+
+ if (xenvif_hash_cache_size == 0)
+ return;
+
+ write_lock_irqsave(&vif->hash.cache.lock, flags);
+
+ memset(vif->hash.cache.entry, 0, xenvif_hash_cache_size *
+ sizeof(struct xenvif_hash_cache_entry));
+
+ write_unlock_irqrestore(&vif->hash.cache.lock, flags);
+}
+
+static u32 xenvif_find_hash(struct xenvif *vif, const u8 *data,
+ unsigned int len)
+{
+ struct xenvif_hash_cache_entry *entry;
+ unsigned long flags;
+ unsigned int i;
+ bool found = false;
+
+ if (len >= XEN_NETBK_HASH_TAG_SIZE)
+ return 0;
+
+ if (xenvif_hash_cache_size == 0)
+ return xenvif_new_hash(vif, data, len);
+
+ read_lock_irqsave(&vif->hash.cache.lock, flags);
+
+ for (i = 0; i < xenvif_hash_cache_size; i++) {
+ entry = &vif->hash.cache.entry[i];
+
+ if (entry->len == len &&
+ memcmp(entry->tag, data, len) == 0) {
+ found = true;
+ break;
+ }
+ }
+
+ read_unlock_irqrestore(&vif->hash.cache.lock, flags);
+
+ if (!found)
+ return xenvif_new_hash(vif, data, len);
+
+ entry->seq = atomic_inc_return(&vif->hash.cache.seq);
+ return entry->val;
+}
+
+void xenvif_set_skb_hash(struct xenvif *vif, struct sk_buff *skb)
+{
+ struct flow_keys flow;
+ u32 hash = 0;
+ enum pkt_hash_types type = PKT_HASH_TYPE_NONE;
+ u32 flags = vif->hash.flags;
+ bool has_tcp_hdr;
+
+ /* Quick rejection test: If the network protocol doesn't
+ * correspond to any enabled hash type then there's no point
+ * in parsing the packet header.
+ */
+ switch (skb->protocol) {
+ case htons(ETH_P_IP):
+ if (flags & (XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP |
+ XEN_NETIF_CTRL_HASH_TYPE_IPV4))
+ break;
+
+ goto done;
+
+ case htons(ETH_P_IPV6):
+ if (flags & (XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP |
+ XEN_NETIF_CTRL_HASH_TYPE_IPV6))
+ break;
+
+ goto done;
+
+ default:
+ goto done;
+ }
+
+ memset(&flow, 0, sizeof(flow));
+ if (!skb_flow_dissect_flow_keys(skb, &flow, 0))
+ goto done;
+
+ has_tcp_hdr = (flow.basic.ip_proto == IPPROTO_TCP) &&
+ !(flow.control.flags & FLOW_DIS_IS_FRAGMENT);
+
+ switch (skb->protocol) {
+ case htons(ETH_P_IP):
+ if (has_tcp_hdr &&
+ (flags & XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP)) {
+ u8 data[12];
+
+ memcpy(&data[0], &flow.addrs.v4addrs.src, 4);
+ memcpy(&data[4], &flow.addrs.v4addrs.dst, 4);
+ memcpy(&data[8], &flow.ports.src, 2);
+ memcpy(&data[10], &flow.ports.dst, 2);
+
+ hash = xenvif_find_hash(vif, data, sizeof(data));
+ type = PKT_HASH_TYPE_L4;
+ } else if (flags & XEN_NETIF_CTRL_HASH_TYPE_IPV4) {
+ u8 data[8];
+
+ memcpy(&data[0], &flow.addrs.v4addrs.src, 4);
+ memcpy(&data[4], &flow.addrs.v4addrs.dst, 4);
+
+ hash = xenvif_find_hash(vif, data, sizeof(data));
+ type = PKT_HASH_TYPE_L3;
+ }
+
+ break;
+
+ case htons(ETH_P_IPV6):
+ if (has_tcp_hdr &&
+ (flags & XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP)) {
+ u8 data[36];
+
+ memcpy(&data[0], &flow.addrs.v6addrs.src, 16);
+ memcpy(&data[16], &flow.addrs.v6addrs.dst, 16);
+ memcpy(&data[32], &flow.ports.src, 2);
+ memcpy(&data[34], &flow.ports.dst, 2);
+
+ hash = xenvif_find_hash(vif, data, sizeof(data));
+ type = PKT_HASH_TYPE_L4;
+ } else if (flags & XEN_NETIF_CTRL_HASH_TYPE_IPV6) {
+ u8 data[32];
+
+ memcpy(&data[0], &flow.addrs.v6addrs.src, 16);
+ memcpy(&data[16], &flow.addrs.v6addrs.dst, 16);
+
+ hash = xenvif_find_hash(vif, data, sizeof(data));
+ type = PKT_HASH_TYPE_L3;
+ }
+
+ break;
+ }
+
+done:
+ if (type == PKT_HASH_TYPE_NONE)
+ skb_clear_hash(skb);
+ else
+ __skb_set_sw_hash(skb, hash, type == PKT_HASH_TYPE_L4);
+}
+
+u32 xenvif_set_hash_alg(struct xenvif *vif, u32 alg)
+{
+ switch (alg) {
+ case XEN_NETIF_CTRL_HASH_ALGORITHM_NONE:
+ case XEN_NETIF_CTRL_HASH_ALGORITHM_TOEPLITZ:
+ break;
+
+ default:
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+ }
+
+ vif->hash.alg = alg;
+
+ return XEN_NETIF_CTRL_STATUS_SUCCESS;
+}
+
+u32 xenvif_get_hash_flags(struct xenvif *vif, u32 *flags)
+{
+ if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE)
+ return XEN_NETIF_CTRL_STATUS_NOT_SUPPORTED;
+
+ *flags = XEN_NETIF_CTRL_HASH_TYPE_IPV4 |
+ XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP |
+ XEN_NETIF_CTRL_HASH_TYPE_IPV6 |
+ XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP;
+
+ return XEN_NETIF_CTRL_STATUS_SUCCESS;
+}
+
+u32 xenvif_set_hash_flags(struct xenvif *vif, u32 flags)
+{
+ if (flags & ~(XEN_NETIF_CTRL_HASH_TYPE_IPV4 |
+ XEN_NETIF_CTRL_HASH_TYPE_IPV4_TCP |
+ XEN_NETIF_CTRL_HASH_TYPE_IPV6 |
+ XEN_NETIF_CTRL_HASH_TYPE_IPV6_TCP))
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+
+ if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE)
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+
+ vif->hash.flags = flags;
+
+ return XEN_NETIF_CTRL_STATUS_SUCCESS;
+}
+
+u32 xenvif_set_hash_key(struct xenvif *vif, u32 gref, u32 len)
+{
+ u8 *key = vif->hash.key;
+ struct gnttab_copy copy_op = {
+ .source.u.ref = gref,
+ .source.domid = vif->domid,
+ .dest.u.gmfn = virt_to_gfn(key),
+ .dest.domid = DOMID_SELF,
+ .dest.offset = xen_offset_in_page(key),
+ .len = len,
+ .flags = GNTCOPY_source_gref
+ };
+
+ if (len > XEN_NETBK_MAX_HASH_KEY_SIZE)
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+
+ if (len != 0) {
+ gnttab_batch_copy(©_op, 1);
+
+ if (copy_op.status != GNTST_okay)
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+ }
+
+ /* Clear any remaining key octets */
+ if (len < XEN_NETBK_MAX_HASH_KEY_SIZE)
+ memset(key + len, 0, XEN_NETBK_MAX_HASH_KEY_SIZE - len);
+
+ xenvif_flush_hash(vif);
+
+ return XEN_NETIF_CTRL_STATUS_SUCCESS;
+}
+
+u32 xenvif_set_hash_mapping_size(struct xenvif *vif, u32 size)
+{
+ if (size > XEN_NETBK_MAX_HASH_MAPPING_SIZE)
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+
+ vif->hash.size = size;
+ memset(vif->hash.mapping, 0, sizeof(u32) * size);
+
+ return XEN_NETIF_CTRL_STATUS_SUCCESS;
+}
+
+u32 xenvif_set_hash_mapping(struct xenvif *vif, u32 gref, u32 len,
+ u32 off)
+{
+ u32 *mapping = &vif->hash.mapping[off];
+ struct gnttab_copy copy_op = {
+ .source.u.ref = gref,
+ .source.domid = vif->domid,
+ .dest.u.gmfn = virt_to_gfn(mapping),
+ .dest.domid = DOMID_SELF,
+ .dest.offset = xen_offset_in_page(mapping),
+ .len = len * sizeof(u32),
+ .flags = GNTCOPY_source_gref
+ };
+
+ if ((off + len > vif->hash.size) || copy_op.len > XEN_PAGE_SIZE)
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+
+ while (len-- != 0)
+ if (mapping[off++] >= vif->num_queues)
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+
+ if (len != 0) {
+ gnttab_batch_copy(©_op, 1);
+
+ if (copy_op.status != GNTST_okay)
+ return XEN_NETIF_CTRL_STATUS_INVALID_PARAMETER;
+ }
+
+ return XEN_NETIF_CTRL_STATUS_SUCCESS;
+}
+
+int xenvif_init_hash(struct xenvif *vif)
+{
+ if (xenvif_hash_cache_size == 0)
+ return 0;
+
+ vif->hash.cache.entry =
+ vzalloc(xenvif_hash_cache_size *
+ sizeof(struct xenvif_hash_cache_entry));
+
+ if (!vif->hash.cache.entry)
+ return -ENOMEM;
+
+ rwlock_init(&vif->hash.cache.lock);
+ return 0;
+}
+
+void xenvif_deinit_hash(struct xenvif *vif)
+{
+ if (vif->hash.cache.entry)
+ vfree(vif->hash.cache.entry);
+
+ memset(&vif->hash, 0, sizeof(vif->hash));
+}
diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
index 78a10d2..e54b475 100644
--- a/drivers/net/xen-netback/interface.c
+++ b/drivers/net/xen-netback/interface.c
@@ -151,6 +151,24 @@ void xenvif_wake_queue(struct xenvif_queue *queue)
netif_tx_wake_queue(netdev_get_tx_queue(dev, id));
}
+static u16 xenvif_select_queue(struct net_device *dev, struct sk_buff *skb,
+ void *accel_priv,
+ select_queue_fallback_t fallback)
+{
+ struct xenvif *vif = netdev_priv(dev);
+ unsigned int size = vif->hash.size;
+
+ if (vif->hash.alg == XEN_NETIF_CTRL_HASH_ALGORITHM_NONE)
+ return fallback(dev, skb) % dev->real_num_tx_queues;
+
+ xenvif_set_skb_hash(vif, skb);
+
+ if (size == 0)
+ return skb_get_hash_raw(skb) % dev->real_num_tx_queues;
+
+ return vif->hash.mapping[skb_get_hash_raw(skb) % size];
+}
+
static int xenvif_start_xmit(struct sk_buff *skb, struct net_device *dev)
{
struct xenvif *vif = netdev_priv(dev);
@@ -395,6 +413,7 @@ static const struct ethtool_ops xenvif_ethtool_ops = {
};
static const struct net_device_ops xenvif_netdev_ops = {
+ .ndo_select_queue = xenvif_select_queue,
.ndo_start_xmit = xenvif_start_xmit,
.ndo_get_stats = xenvif_get_stats,
.ndo_open = xenvif_open,
@@ -563,6 +582,10 @@ int xenvif_connect_ctrl(struct xenvif *vif, grant_ref_t ring_ref,
vif->ctrl_irq = err;
+ err = xenvif_init_hash(vif);
+ if (err < 0)
+ goto err_unbind;
+
task = kthread_create(xenvif_ctrl_kthread, (void *)vif,
"%s-control", dev->name);
if (IS_ERR(task)) {
@@ -579,6 +602,9 @@ int xenvif_connect_ctrl(struct xenvif *vif, grant_ref_t ring_ref,
return 0;
err_deinit:
+ xenvif_deinit_hash(vif);
+
+err_unbind:
unbind_from_irqhandler(vif->ctrl_irq, vif);
vif->ctrl_irq = 0;
@@ -749,6 +775,8 @@ void xenvif_disconnect_ctrl(struct xenvif *vif)
vif->ctrl_task = NULL;
}
+ xenvif_deinit_hash(vif);
+
if (vif->ctrl_irq) {
unbind_from_irqhandler(vif->ctrl_irq, vif);
vif->ctrl_irq = 0;
diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c
index 45aa40b..6509d11 100644
--- a/drivers/net/xen-netback/netback.c
+++ b/drivers/net/xen-netback/netback.c
@@ -89,6 +89,11 @@ module_param(fatal_skb_slots, uint, 0444);
*/
#define XEN_NETBACK_TX_COPY_LEN 128
+/* This is the maximum number of flows in the hash cache. */
+#define XENVIF_HASH_CACHE_SIZE_DEFAULT 64
+unsigned int xenvif_hash_cache_size = XENVIF_HASH_CACHE_SIZE_DEFAULT;
+module_param_named(hash_cache_size, xenvif_hash_cache_size, uint, 0644);
+MODULE_PARM_DESC(hash_cache_size, "Number of flows in the hash cache");
static void xenvif_idx_release(struct xenvif_queue *queue, u16 pending_idx,
u8 status);
@@ -2191,8 +2196,48 @@ static void push_ctrl_response(struct xenvif *vif)
static void process_ctrl_request(struct xenvif *vif,
const struct xen_netif_ctrl_request *req)
{
- make_ctrl_response(vif, req, XEN_NETIF_CTRL_STATUS_NOT_SUPPORTED,
- 0);
+ u32 status = XEN_NETIF_CTRL_STATUS_NOT_SUPPORTED;
+ u32 data = 0;
+
+ switch (req->type) {
+ case XEN_NETIF_CTRL_TYPE_SET_HASH_ALGORITHM:
+ status = xenvif_set_hash_alg(vif, req->data[0]);
+ break;
+
+ case XEN_NETIF_CTRL_TYPE_GET_HASH_FLAGS:
+ status = xenvif_get_hash_flags(vif, &data);
+ break;
+
+ case XEN_NETIF_CTRL_TYPE_SET_HASH_FLAGS:
+ status = xenvif_set_hash_flags(vif, req->data[0]);
+ break;
+
+ case XEN_NETIF_CTRL_TYPE_SET_HASH_KEY:
+ status = xenvif_set_hash_key(vif, req->data[0],
+ req->data[1]);
+ break;
+
+ case XEN_NETIF_CTRL_TYPE_GET_HASH_MAPPING_SIZE:
+ status = XEN_NETIF_CTRL_STATUS_SUCCESS;
+ data = XEN_NETBK_MAX_HASH_MAPPING_SIZE;
+ break;
+
+ case XEN_NETIF_CTRL_TYPE_SET_HASH_MAPPING_SIZE:
+ status = xenvif_set_hash_mapping_size(vif,
+ req->data[0]);
+ break;
+
+ case XEN_NETIF_CTRL_TYPE_SET_HASH_MAPPING:
+ status = xenvif_set_hash_mapping(vif, req->data[0],
+ req->data[1],
+ req->data[2]);
+ break;
+
+ default:
+ break;
+ }
+
+ make_ctrl_response(vif, req, status, data);
push_ctrl_response(vif);
}
--
2.1.4
^ permalink raw reply related
* [PATCH net-next 1/4] xen-netback: add control ring boilerplate
From: Paul Durrant @ 2016-05-05 11:19 UTC (permalink / raw)
To: xen-devel, netdev; +Cc: Paul Durrant, Wei Liu
In-Reply-To: <1462447170-1815-1-git-send-email-paul.durrant@citrix.com>
My recent patch to include/xen/interface/io/netif.h defines a new shared
ring (in addition to the rx and tx rings) for passing control messages
from a VM frontend driver to a backend driver.
This patch adds the necessary code to xen-netback to map this new shared
ring, should it be created by a frontend, but does not add implementations
for any of the defined protocol messages. These are added in a subsequent
patch for clarity.
Signed-off-by: Paul Durrant <paul.durrant@citrix.com>
Cc: Wei Liu <wei.liu2@citrix.com>
---
drivers/net/xen-netback/common.h | 28 +++++++---
drivers/net/xen-netback/interface.c | 101 +++++++++++++++++++++++++++++++++---
drivers/net/xen-netback/netback.c | 99 +++++++++++++++++++++++++++++++++--
drivers/net/xen-netback/xenbus.c | 75 ++++++++++++++++++++++----
4 files changed, 273 insertions(+), 30 deletions(-)
diff --git a/drivers/net/xen-netback/common.h b/drivers/net/xen-netback/common.h
index f44b388..093a12a 100644
--- a/drivers/net/xen-netback/common.h
+++ b/drivers/net/xen-netback/common.h
@@ -260,6 +260,11 @@ struct xenvif {
struct dentry *xenvif_dbg_root;
#endif
+ struct xen_netif_ctrl_back_ring ctrl;
+ struct task_struct *ctrl_task;
+ wait_queue_head_t ctrl_wq;
+ unsigned int ctrl_irq;
+
/* Miscellaneous private stuff. */
struct net_device *dev;
};
@@ -285,10 +290,15 @@ struct xenvif *xenvif_alloc(struct device *parent,
int xenvif_init_queue(struct xenvif_queue *queue);
void xenvif_deinit_queue(struct xenvif_queue *queue);
-int xenvif_connect(struct xenvif_queue *queue, unsigned long tx_ring_ref,
- unsigned long rx_ring_ref, unsigned int tx_evtchn,
- unsigned int rx_evtchn);
-void xenvif_disconnect(struct xenvif *vif);
+int xenvif_connect_data(struct xenvif_queue *queue,
+ unsigned long tx_ring_ref,
+ unsigned long rx_ring_ref,
+ unsigned int tx_evtchn,
+ unsigned int rx_evtchn);
+void xenvif_disconnect_data(struct xenvif *vif);
+int xenvif_connect_ctrl(struct xenvif *vif, grant_ref_t ring_ref,
+ unsigned int evtchn);
+void xenvif_disconnect_ctrl(struct xenvif *vif);
void xenvif_free(struct xenvif *vif);
int xenvif_xenbus_init(void);
@@ -300,10 +310,10 @@ int xenvif_queue_stopped(struct xenvif_queue *queue);
void xenvif_wake_queue(struct xenvif_queue *queue);
/* (Un)Map communication rings. */
-void xenvif_unmap_frontend_rings(struct xenvif_queue *queue);
-int xenvif_map_frontend_rings(struct xenvif_queue *queue,
- grant_ref_t tx_ring_ref,
- grant_ref_t rx_ring_ref);
+void xenvif_unmap_frontend_data_rings(struct xenvif_queue *queue);
+int xenvif_map_frontend_data_rings(struct xenvif_queue *queue,
+ grant_ref_t tx_ring_ref,
+ grant_ref_t rx_ring_ref);
/* Check for SKBs from frontend and schedule backend processing */
void xenvif_napi_schedule_or_enable_events(struct xenvif_queue *queue);
@@ -318,6 +328,8 @@ void xenvif_kick_thread(struct xenvif_queue *queue);
int xenvif_dealloc_kthread(void *data);
+int xenvif_ctrl_kthread(void *data);
+
void xenvif_rx_queue_tail(struct xenvif_queue *queue, struct sk_buff *skb);
void xenvif_carrier_on(struct xenvif *vif);
diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
index f5231a2..78a10d2 100644
--- a/drivers/net/xen-netback/interface.c
+++ b/drivers/net/xen-netback/interface.c
@@ -128,6 +128,15 @@ irqreturn_t xenvif_interrupt(int irq, void *dev_id)
return IRQ_HANDLED;
}
+irqreturn_t xenvif_ctrl_interrupt(int irq, void *dev_id)
+{
+ struct xenvif *vif = dev_id;
+
+ wake_up(&vif->ctrl_wq);
+
+ return IRQ_HANDLED;
+}
+
int xenvif_queue_stopped(struct xenvif_queue *queue)
{
struct net_device *dev = queue->vif->dev;
@@ -527,9 +536,66 @@ void xenvif_carrier_on(struct xenvif *vif)
rtnl_unlock();
}
-int xenvif_connect(struct xenvif_queue *queue, unsigned long tx_ring_ref,
- unsigned long rx_ring_ref, unsigned int tx_evtchn,
- unsigned int rx_evtchn)
+int xenvif_connect_ctrl(struct xenvif *vif, grant_ref_t ring_ref,
+ unsigned int evtchn)
+{
+ struct net_device *dev = vif->dev;
+ void *addr;
+ struct xen_netif_ctrl_sring *shared;
+ struct task_struct *task;
+ int err = -ENOMEM;
+
+ err = xenbus_map_ring_valloc(xenvif_to_xenbus_device(vif),
+ &ring_ref, 1, &addr);
+ if (err)
+ goto err;
+
+ shared = (struct xen_netif_ctrl_sring *)addr;
+ BACK_RING_INIT(&vif->ctrl, shared, XEN_PAGE_SIZE);
+
+ init_waitqueue_head(&vif->ctrl_wq);
+
+ err = bind_interdomain_evtchn_to_irqhandler(vif->domid, evtchn,
+ xenvif_ctrl_interrupt,
+ 0, dev->name, vif);
+ if (err < 0)
+ goto err_unmap;
+
+ vif->ctrl_irq = err;
+
+ task = kthread_create(xenvif_ctrl_kthread, (void *)vif,
+ "%s-control", dev->name);
+ if (IS_ERR(task)) {
+ pr_warn("Could not allocate kthread for %s\n", dev->name);
+ err = PTR_ERR(task);
+ goto err_deinit;
+ }
+
+ get_task_struct(task);
+ vif->ctrl_task = task;
+
+ wake_up_process(vif->ctrl_task);
+
+ return 0;
+
+err_deinit:
+ unbind_from_irqhandler(vif->ctrl_irq, vif);
+ vif->ctrl_irq = 0;
+
+err_unmap:
+ xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(vif),
+ vif->ctrl.sring);
+ vif->ctrl.sring = NULL;
+
+err:
+ return err;
+}
+
+int xenvif_connect_data(struct xenvif_queue *queue,
+ unsigned long tx_ring_ref,
+ unsigned long rx_ring_ref,
+ unsigned int tx_evtchn,
+ unsigned int rx_evtchn)
{
struct task_struct *task;
int err = -ENOMEM;
@@ -538,7 +604,8 @@ int xenvif_connect(struct xenvif_queue *queue, unsigned long tx_ring_ref,
BUG_ON(queue->task);
BUG_ON(queue->dealloc_task);
- err = xenvif_map_frontend_rings(queue, tx_ring_ref, rx_ring_ref);
+ err = xenvif_map_frontend_data_rings(queue, tx_ring_ref,
+ rx_ring_ref);
if (err < 0)
goto err;
@@ -614,7 +681,7 @@ err_tx_unbind:
unbind_from_irqhandler(queue->tx_irq, queue);
queue->tx_irq = 0;
err_unmap:
- xenvif_unmap_frontend_rings(queue);
+ xenvif_unmap_frontend_data_rings(queue);
netif_napi_del(&queue->napi);
err:
module_put(THIS_MODULE);
@@ -634,7 +701,7 @@ void xenvif_carrier_off(struct xenvif *vif)
rtnl_unlock();
}
-void xenvif_disconnect(struct xenvif *vif)
+void xenvif_disconnect_data(struct xenvif *vif)
{
struct xenvif_queue *queue = NULL;
unsigned int num_queues = vif->num_queues;
@@ -668,12 +735,32 @@ void xenvif_disconnect(struct xenvif *vif)
queue->tx_irq = 0;
}
- xenvif_unmap_frontend_rings(queue);
+ xenvif_unmap_frontend_data_rings(queue);
}
xenvif_mcast_addr_list_free(vif);
}
+void xenvif_disconnect_ctrl(struct xenvif *vif)
+{
+ if (vif->ctrl_task) {
+ kthread_stop(vif->ctrl_task);
+ put_task_struct(vif->ctrl_task);
+ vif->ctrl_task = NULL;
+ }
+
+ if (vif->ctrl_irq) {
+ unbind_from_irqhandler(vif->ctrl_irq, vif);
+ vif->ctrl_irq = 0;
+ }
+
+ if (vif->ctrl.sring) {
+ xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(vif),
+ vif->ctrl.sring);
+ vif->ctrl.sring = NULL;
+ }
+}
+
/* Reverse the relevant parts of xenvif_init_queue().
* Used for queue teardown from xenvif_free(), and on the
* error handling paths in xenbus.c:connect().
diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c
index b42f260..45aa40b 100644
--- a/drivers/net/xen-netback/netback.c
+++ b/drivers/net/xen-netback/netback.c
@@ -1925,7 +1925,7 @@ static inline bool tx_dealloc_work_todo(struct xenvif_queue *queue)
return queue->dealloc_cons != queue->dealloc_prod;
}
-void xenvif_unmap_frontend_rings(struct xenvif_queue *queue)
+void xenvif_unmap_frontend_data_rings(struct xenvif_queue *queue)
{
if (queue->tx.sring)
xenbus_unmap_ring_vfree(xenvif_to_xenbus_device(queue->vif),
@@ -1935,9 +1935,9 @@ void xenvif_unmap_frontend_rings(struct xenvif_queue *queue)
queue->rx.sring);
}
-int xenvif_map_frontend_rings(struct xenvif_queue *queue,
- grant_ref_t tx_ring_ref,
- grant_ref_t rx_ring_ref)
+int xenvif_map_frontend_data_rings(struct xenvif_queue *queue,
+ grant_ref_t tx_ring_ref,
+ grant_ref_t rx_ring_ref)
{
void *addr;
struct xen_netif_tx_sring *txs;
@@ -1964,7 +1964,7 @@ int xenvif_map_frontend_rings(struct xenvif_queue *queue,
return 0;
err:
- xenvif_unmap_frontend_rings(queue);
+ xenvif_unmap_frontend_data_rings(queue);
return err;
}
@@ -2163,6 +2163,95 @@ int xenvif_dealloc_kthread(void *data)
return 0;
}
+static void make_ctrl_response(struct xenvif *vif,
+ const struct xen_netif_ctrl_request *req,
+ u32 status, u32 data)
+{
+ RING_IDX idx = vif->ctrl.rsp_prod_pvt;
+ struct xen_netif_ctrl_response rsp = {
+ .id = req->id,
+ .type = req->type,
+ .status = status,
+ .data = data,
+ };
+
+ *RING_GET_RESPONSE(&vif->ctrl, idx) = rsp;
+ vif->ctrl.rsp_prod_pvt = ++idx;
+}
+
+static void push_ctrl_response(struct xenvif *vif)
+{
+ int notify;
+
+ RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&vif->ctrl, notify);
+ if (notify)
+ notify_remote_via_irq(vif->ctrl_irq);
+}
+
+static void process_ctrl_request(struct xenvif *vif,
+ const struct xen_netif_ctrl_request *req)
+{
+ make_ctrl_response(vif, req, XEN_NETIF_CTRL_STATUS_NOT_SUPPORTED,
+ 0);
+ push_ctrl_response(vif);
+}
+
+static void xenvif_ctrl_action(struct xenvif *vif)
+{
+ for (;;) {
+ RING_IDX req_prod, req_cons;
+
+ req_prod = vif->ctrl.sring->req_prod;
+ req_cons = vif->ctrl.req_cons;
+
+ /* Make sure we can see requests before we process them. */
+ rmb();
+
+ if (req_cons == req_prod)
+ break;
+
+ while (req_cons != req_prod) {
+ struct xen_netif_ctrl_request req;
+
+ RING_COPY_REQUEST(&vif->ctrl, req_cons, &req);
+ req_cons++;
+
+ process_ctrl_request(vif, &req);
+ }
+
+ vif->ctrl.req_cons = req_cons;
+ vif->ctrl.sring->req_event = req_cons + 1;
+ }
+}
+
+static bool xenvif_ctrl_work_todo(struct xenvif *vif)
+{
+ if (likely(RING_HAS_UNCONSUMED_REQUESTS(&vif->ctrl)))
+ return 1;
+
+ return 0;
+}
+
+int xenvif_ctrl_kthread(void *data)
+{
+ struct xenvif *vif = data;
+
+ for (;;) {
+ wait_event_interruptible(vif->ctrl_wq,
+ xenvif_ctrl_work_todo(vif) ||
+ kthread_should_stop());
+ if (kthread_should_stop())
+ break;
+
+ while (xenvif_ctrl_work_todo(vif))
+ xenvif_ctrl_action(vif);
+
+ cond_resched();
+ }
+
+ return 0;
+}
+
static int __init netback_init(void)
{
int rc = 0;
diff --git a/drivers/net/xen-netback/xenbus.c b/drivers/net/xen-netback/xenbus.c
index bd182cd..278f67b 100644
--- a/drivers/net/xen-netback/xenbus.c
+++ b/drivers/net/xen-netback/xenbus.c
@@ -38,7 +38,8 @@ struct backend_info {
const char *hotplug_script;
};
-static int connect_rings(struct backend_info *be, struct xenvif_queue *queue);
+static int connect_data_rings(struct backend_info *be,
+ struct xenvif_queue *queue);
static void connect(struct backend_info *be);
static int read_xenbus_vif_flags(struct backend_info *be);
static int backend_create_xenvif(struct backend_info *be);
@@ -367,6 +368,12 @@ static int netback_probe(struct xenbus_device *dev,
if (err)
pr_debug("Error writing multi-queue-max-queues\n");
+ err = xenbus_printf(XBT_NIL, dev->nodename,
+ "feature-ctrl-ring",
+ "%u", true);
+ if (err)
+ pr_debug("Error writing feature-ctrl-ring\n");
+
script = xenbus_read(XBT_NIL, dev->nodename, "script", NULL);
if (IS_ERR(script)) {
err = PTR_ERR(script);
@@ -457,7 +464,8 @@ static void backend_disconnect(struct backend_info *be)
#ifdef CONFIG_DEBUG_FS
xenvif_debugfs_delif(be->vif);
#endif /* CONFIG_DEBUG_FS */
- xenvif_disconnect(be->vif);
+ xenvif_disconnect_data(be->vif);
+ xenvif_disconnect_ctrl(be->vif);
}
}
@@ -825,6 +833,44 @@ static void hotplug_status_changed(struct xenbus_watch *watch,
kfree(str);
}
+static int connect_ctrl_ring(struct backend_info *be)
+{
+ struct xenbus_device *dev = be->dev;
+ struct xenvif *vif = be->vif;
+ unsigned int val;
+ grant_ref_t ring_ref;
+ unsigned int evtchn;
+ int err;
+
+ err = xenbus_gather(XBT_NIL, dev->otherend,
+ "ctrl-ring-ref", "%u", &val, NULL);
+ if (err)
+ return 0; /* The frontend does not have a control ring */
+
+ ring_ref = val;
+
+ err = xenbus_gather(XBT_NIL, dev->otherend,
+ "event-channel-ctrl", "%u", &val, NULL);
+ if (err) {
+ xenbus_dev_fatal(dev, err,
+ "reading %s/event-channel-ctrl",
+ dev->otherend);
+ return err;
+ }
+
+ evtchn = val;
+
+ err = xenvif_connect_ctrl(vif, ring_ref, evtchn);
+ if (err) {
+ xenbus_dev_fatal(dev, err,
+ "mapping shared-frame %u port %u",
+ ring_ref, evtchn);
+ return err;
+ }
+
+ return 0;
+}
+
static void connect(struct backend_info *be)
{
int err;
@@ -861,6 +907,12 @@ static void connect(struct backend_info *be)
xen_register_watchers(dev, be->vif);
read_xenbus_vif_flags(be);
+ err = connect_ctrl_ring(be);
+ if (err) {
+ xenbus_dev_fatal(dev, err, "connecting control ring");
+ return;
+ }
+
/* Use the number of queues requested by the frontend */
be->vif->queues = vzalloc(requested_num_queues *
sizeof(struct xenvif_queue));
@@ -896,11 +948,12 @@ static void connect(struct backend_info *be)
queue->remaining_credit = credit_bytes;
queue->credit_usec = credit_usec;
- err = connect_rings(be, queue);
+ err = connect_data_rings(be, queue);
if (err) {
- /* connect_rings() cleans up after itself on failure,
- * but we need to clean up after xenvif_init_queue() here,
- * and also clean up any previously initialised queues.
+ /* connect_data_rings() cleans up after itself on
+ * failure, but we need to clean up after
+ * xenvif_init_queue() here, and also clean up any
+ * previously initialised queues.
*/
xenvif_deinit_queue(queue);
be->vif->num_queues = queue_index;
@@ -935,15 +988,17 @@ static void connect(struct backend_info *be)
err:
if (be->vif->num_queues > 0)
- xenvif_disconnect(be->vif); /* Clean up existing queues */
+ xenvif_disconnect_data(be->vif); /* Clean up existing queues */
vfree(be->vif->queues);
be->vif->queues = NULL;
be->vif->num_queues = 0;
+ xenvif_disconnect_ctrl(be->vif);
return;
}
-static int connect_rings(struct backend_info *be, struct xenvif_queue *queue)
+static int connect_data_rings(struct backend_info *be,
+ struct xenvif_queue *queue)
{
struct xenbus_device *dev = be->dev;
unsigned int num_queues = queue->vif->num_queues;
@@ -1007,8 +1062,8 @@ static int connect_rings(struct backend_info *be, struct xenvif_queue *queue)
}
/* Map the shared frame, irq etc. */
- err = xenvif_connect(queue, tx_ring_ref, rx_ring_ref,
- tx_evtchn, rx_evtchn);
+ err = xenvif_connect_data(queue, tx_ring_ref, rx_ring_ref,
+ tx_evtchn, rx_evtchn);
if (err) {
xenbus_dev_fatal(dev, err,
"mapping shared-frames %lu/%lu port tx %u rx %u",
--
2.1.4
^ permalink raw reply related
* Re: [PATCH nf-next 0/9] netfilter: remove per-netns conntrack tables, part 1
From: Pablo Neira Ayuso @ 2016-05-05 11:54 UTC (permalink / raw)
To: Florian Westphal; +Cc: netfilter-devel, netdev
In-Reply-To: <1461863628-23350-1-git-send-email-fw@strlen.de>
On Thu, Apr 28, 2016 at 07:13:39PM +0200, Florian Westphal wrote:
> [ CCing netdev so netns folks can have a look too ]
>
> This patch series removes the per-netns connection tracking tables.
> All conntrack objects are then stored in one global global table.
>
> This avoids the infamous 'vmalloc' when lots of namespaces are used:
> We no longer allocate a new conntrack table for each namespace (with 64k
> size this saves 512kb of memory per netns).
>
> - net namespace address is made part of conntrack hash, to spread
> conntracks over entire table even if netns has overlapping ip addresses.
> - lookup and iterators net_eq() to skip conntracks living in a different
> namespace.
>
> Only the main conntrack table is converted here:
> NAT bysrc and expectation hashes are still per namespace (will be unified
> in a followup series). Also, this retains the per-namespace kmem cache
> for the conntrack objects. This will also be resolved in a followup series.
Series applied, thanks Florian.
^ permalink raw reply
* Re: [PATCH 2/3] net-next: mediatek: add fixed-phy support
From: Andrew Lunn @ 2016-05-05 12:02 UTC (permalink / raw)
To: John Crispin; +Cc: David S. Miller, nbd, netdev, linux-mediatek, linux-kernel
In-Reply-To: <1462439856-51788-3-git-send-email-john@phrozen.org>
On Thu, May 05, 2016 at 11:17:35AM +0200, John Crispin wrote:
> The MT7623 SoC has a builtin gigabit switch. If we want to use it, GMAC1
> needs to be configured using a fixed link speed and flow control settings.
> The easiest way to do this is to used the fixed-phy driver, allowing us to
> reuse the existing mdio polling code to setup the MAC.
>
> Signed-off-by: John Crispin <john@phrozen.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Looks good.
Andrew
^ permalink raw reply
* Re: [PATCH 3/3] net-next: mediatek: add RX delay support
From: Andrew Lunn @ 2016-05-05 12:13 UTC (permalink / raw)
To: John Crispin; +Cc: David S. Miller, nbd, netdev, linux-mediatek, linux-kernel
In-Reply-To: <1462439856-51788-4-git-send-email-john@phrozen.org>
On Thu, May 05, 2016 at 11:17:36AM +0200, John Crispin wrote:
> If an external Gigabit PHY is connected to either of the MACs we need to
> tell the to use a RX delay. Not doing so will result in heavy packet loss
> and/or data corruption of RX traffic.
Hi John
Is this comment correct? Reading the code, all this switch statement
does is select between RGMII, MII and RMII. It has nothing to do with
delay. I suspect the PHY is doing the delay, not the MAC, since you
pass the phy mode to of_phy_connect().
If my interpretation of the code is correct, you might also want to
handle PHY_INTERFACE_MODE_RGMII_TXID and PHY_INTERFACE_MODE_RGMII_ID
which are also RGMII modes.
Andrew
^ permalink raw reply
* Re: [REGRESSION] asix: Lots of asix_rx_fixup() errors and slow transmissions
From: Guodong Xu @ 2016-05-05 12:19 UTC (permalink / raw)
To: Dean Jenkins
Cc: John Stultz, lkml, Mark Craske, David S. Miller, YongQin Liu,
linux-usb, netdev, Ivan Vecera, David B. Robins
In-Reply-To: <572B0023.9060505@mentor.com>
On 5 May 2016 at 16:11, Dean Jenkins <Dean_Jenkins@mentor.com> wrote:
> On 05/05/16 00:45, John Stultz wrote:
>>
>> On Tue, May 3, 2016 at 3:54 AM, Dean Jenkins <Dean_Jenkins@mentor.com>
>> wrote:
>>>
>>> On 03/05/16 11:04, Guodong Xu wrote:
>>>>
>>>> did you test on ARM 64-bit system or ARM 32-bit? I ask because HiKey
>>>> is an ARM 64-bit system. I suggest we should be careful on that. I saw
>>>> similar issues when transferring to a 64-bit system in other net
>>>> drivers.
>>>
>>> We used 32-bit ARM and never tested on 64-bit ARM so I suggest that the
>>> commits need to be reviewed with 64-bit OS in mind.
>>>>
>>>>
>>>> Do you have any suggestion on this regard?
>>>
>>> Try testing on a Linux PC x86 32-bit OS which has has a kernel containing
>>> my
>>> ASIX commits. This will help to confirm whether the failure is related to
>>> 32-bit or 64-bit OS. Then try with Linux PC x86 64-bit OS, this should
>>> fail
>>> otherwise it points to something specific in your ARM 64-bit platform.
>>
>> Just as a sample point, I have managed to reproduce exactly this issue
>> on an x86_64 system by simply scp'ing a large file.
>
> Please tell us the x86_64 kernel version number that you used and which
> Linux Distribution it was ? This allows other people a chance to reproduce
> your observations.
>
>>
>> [ 417.819276] asix 1-5:1.0 eth1: asix_rx_fixup() Data Header
>> synchronisation was lost, remaining 988
>
> It is interesting that the reported "remaining" value is 988. Is 988 always
> shown ? I mean that do you see any other "remaining" values for the "Data
> Header synchronisation was lost" error message ?
>
>> [ 417.823415] asix 1-5:1.0 eth1: asix_rx_fixup() Bad Header Length
>> 0xef830347, offset 4
>
> The gap in the timestamps shows 417.823415 - 417.819276 = 0.004139 = 4ms
> which is a large gap in terms of USB 2.0 high speed communications. This gap
> is expected to be in the 100us range for consecutive URBs. So 4ms is
> strange.
>
> The expectation is that the "Data Header synchronisation was lost" error
> message resets the 32-bit header word synchronisation to the start of the
> next URB buffer. The "Bad Header Length, offset 4" is the expected outcome
> for the next URB because it is unlikely the 32-bit header word is at the
> start of URB buffer due to Ethernet frames spanning across URBs.
>>
>> [ 417.827502] asix 1-5:1.0 eth1: asix_rx_fixup() Bad Header Length
>> 0x31e2b348, offset 4
>
> Timestamps show the gap to be 4ms which is strange for USB 2.0 high speed,
> are you sure high speed mode is being used ?
>>
>> [ 417.843779] asix 1-5:1.0 eth1: asix_rx_fixup() Data Header
>> synchronisation was lost, remaining 988
>> [ 417.847921] asix 1-5:1.0 eth1: asix_rx_fixup() Bad Header Length
>> 0x8af91035, offset 4
>> [ 417.852004] asix 1-5:1.0 eth1: asix_rx_fixup() Bad Header Length
>> 0x8521fa03, offset 4
>> [ 418.273394] asix 1-5:1.0 eth1: asix_rx_fixup() Data Header
>> synchronisation was lost, remaining 988
>> [ 418.277532] asix 1-5:1.0 eth1: asix_rx_fixup() Bad Header Length
>> 0x33cd9c7c, offset 4
>> [ 418.281683] asix 1-5:1.0 eth1: asix_rx_fixup() Bad Header Length
>> 0x3d850896, offset 4
>> [ 418.286227] asix 1-5:1.0 eth1: asix_rx_fixup() Bad Header Length
>> 0x86443357, offset 4
>> [ 418.290319] asix 1-5:1.0 eth1: asix_rx_fixup() Bad Header Length
>> 0xee6c81d1, offset 4
>>
>> I don't have any 32bit x86 installs around so I'm not sure I can easly
>> test there, but its clear its not arm64 specific.
>
> I agree the issue is not specific to your ARM 64 bit platform.
>
> Please can you supply the output of ifconfig for the USB to Ethernet
> adaptor, your example above shows eth1 as the device.
>
> Please show the output of ifconfig eth1 before and after the issue is seen.
> This will show us whether the kernel logs any network errors and how many
> bytes have been transferred.
>
> After the issue is seen, please can you show us the output of "dmesg | grep
> asix" so that we can see status messages from the ASIX driver that the USB
> to Ethernet adaptor is using. In particular we need to check that USB high
> speed operation (480Mbps) is being used and not full speed operation
> (12Mbps).
Hi, Dean
I am not sure why do you insist 'not full speed'. Actually, the tests
I run on ARM-64bit is at USB full speed mode. I pasted my log here:
http://paste.ubuntu.com/16236442/
, which includes the information you requested above, ifconfig, dmesg.
The interval between two consecutive errors varies from 10 to 40ms.
> It is interesting that the reported "remaining" value is 988. Is 988 always
> shown ? I mean that do you see any other "remaining" values for the "Data
> Header synchronisation was lost" error message ?
Yes and No. When doing iperf test in TCP mode, always 988. I have
never seen other "remaining" value.
But,
1. I tried "ping -f -s 1400 [my.arm.64bit.board.ip]", but this cannot
trigger the error.
2. Tried iperf in UDP mode, I saw "Data Header synchronisation was
lost" remaining value is 984 (again, seemingly always in several
tries). Log is pasted here. http://paste.ubuntu.com/16236764/
-Guodong Xu
>
> Thanks,
>
> Regards,
> Dean
>
>>
>> thanks
>> -john
>
>
>
> --
> Dean Jenkins
> Embedded Software Engineer
> Linux Transportation Solutions
> Mentor Embedded Software Division
> Mentor Graphics (UK) Ltd.
>
^ permalink raw reply
* Re: [PATCH 00/12] net-next: mediatek: bug fixes and tx stall fixes
From: Andrew Lunn @ 2016-05-05 12:25 UTC (permalink / raw)
To: John Crispin; +Cc: David S. Miller, nbd, netdev, linux-mediatek, linux-kernel
In-Reply-To: <1462440385-51939-1-git-send-email-john@phrozen.org>
On Thu, May 05, 2016 at 11:26:13AM +0200, John Crispin wrote:
> This series is a collection of fixes for bugs that we stumble across while
> doing more performance testing and code level review.
Hi John
I just noticed the code uses kfree(mii_bus), rather than
mdiobus_free(mii_bus).
Andrew
^ permalink raw reply
* Re: A couple of questions about the SKB fragments
From: Sowmini Varadhan @ 2016-05-05 12:58 UTC (permalink / raw)
To: Edward Cree; +Cc: matvejchikov, netdev, Aleksey.Baulin
In-Reply-To: <572B2AEE.1020802@solarflare.com>
On (05/05/16 12:13), Edward Cree wrote:
> On 05/05/16 08:40, Ilya Matveychikov wrote:
> >
> > While working with fragmented SKBs we've got stuck with the following:
> > - is it possible for an SKB fragment in skb_shinfo(skb)->frag_list to
> > be fragmented too (i.e. to have SKBs in frag_list)?
> > - do skb->len and skb->data_len contain the whole SKB length,
> > including the length of all fragments (not only the paged parts)?
> >
> > Is there any docs except the kernel sources itself to refer to?
> davem has some docs up at http://vger.kernel.org/~davem/skb.html and
> http://vger.kernel.org/~davem/skb_data.html
> In particular note the following:
> "The frag_list is used to maintain a chain of SKBs organized for
> fragmentation purposes, it is _not_ used for maintaining paged data."
> So my reading would suggest there is no way to multiple-layer-fragment
> an SKB; the frags are page pointers and offsets, not entire sk_buff
> structs in their own right.
Actually (and perhaps I am mixing up skb_frag_t with the sk_buff in
skb_shinfo(skb)->frag_list) I think it is entirely possible in theory
for the frag_list to itself both skb_frag_t's as well as more sk_buffs
in the ->frag_list- at least that's why I understand fromcode
that recurses using skb_walk_frags(), such as tcp_md5_hash_skb_data().
AIUI, the statement about page pointers and offsets applies to the
skb_frag_t's in the skb_shared_info. The len and data_len should be
the sum-total for the whole skb, including skb_frag_t's and ->frag_list.
--Sowmini
^ permalink raw reply
* [PULL] virtio/qemu: fixes for 4.6
From: Michael S. Tsirkin @ 2016-05-05 13:04 UTC (permalink / raw)
To: Linus Torvalds
Cc: kvm, mst, netdev, linux-kernel, virtualization, dan.carpenter
The following changes since commit c3b46c73264b03000d1e18b22f5caf63332547c9:
Linux 4.6-rc4 (2016-04-17 19:13:32 -0700)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/mst/vhost.git tags/for_linus
for you to fetch changes up to e00f7bd221292b318d4d09c3f0c2c8af9b1e5edf:
virtio: Silence uninitialized variable warning (2016-05-01 15:50:08 +0300)
----------------------------------------------------------------
virtio/qemu: fixes for 4.6
A couple of fixes for virtio and for the new QEMU fw cfg driver.
Signed-off-by: Michael S. Tsirkin <mst@redhat.com>
----------------------------------------------------------------
Dan Carpenter (2):
firmware: qemu_fw_cfg.c: potential unintialized variable
virtio: Silence uninitialized variable warning
drivers/firmware/qemu_fw_cfg.c | 2 +-
drivers/virtio/virtio_ring.c | 2 +-
2 files changed, 2 insertions(+), 2 deletions(-)
^ permalink raw reply
* [patch] i40e: fix an uninitialized variable bug
From: Dan Carpenter @ 2016-05-05 13:18 UTC (permalink / raw)
To: Jeff Kirsher, Jesse Brandeburg
Cc: Shannon Nelson, Carolyn Wyborny, Don Skidmore, Bruce Allan,
John Ronciak, Mitch Williams, intel-wired-lan, netdev,
linux-kernel, kernel-janitors
We removed this initialization but it is required. Let's put it back.
Fixes: 895106a577c4 ('i40e: trivial fixes')
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
diff --git a/drivers/net/ethernet/intel/i40e/i40e_hmc.c b/drivers/net/ethernet/intel/i40e/i40e_hmc.c
index 5ebe12d..a7c7b1d 100644
--- a/drivers/net/ethernet/intel/i40e/i40e_hmc.c
+++ b/drivers/net/ethernet/intel/i40e/i40e_hmc.c
@@ -49,7 +49,7 @@ i40e_status i40e_add_sd_table_entry(struct i40e_hw *hw,
struct i40e_hmc_sd_entry *sd_entry;
bool dma_mem_alloc_done = false;
struct i40e_dma_mem mem;
- i40e_status ret_code;
+ i40e_status ret_code = I40E_SUCCESS;
u64 alloc_len;
if (NULL == hmc_info->sd_table.sd_entry) {
^ permalink raw reply related
* [patch 1/2] netxen: fix error handling in netxen_get_flash_block()
From: Dan Carpenter @ 2016-05-05 13:18 UTC (permalink / raw)
To: Manish Chopra
Cc: Sony Chacko, Rajesh Borundia, netdev, linux-kernel,
kernel-janitors
My static checker complained that "v" can be used unintialized if
netxen_rom_fast_read() returns -EIO. That function never actually
returns -1.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
diff --git a/drivers/net/ethernet/qlogic/netxen/netxen_nic_hw.c b/drivers/net/ethernet/qlogic/netxen/netxen_nic_hw.c
index db80eb1..a320541 100644
--- a/drivers/net/ethernet/qlogic/netxen/netxen_nic_hw.c
+++ b/drivers/net/ethernet/qlogic/netxen/netxen_nic_hw.c
@@ -1015,20 +1015,24 @@ static int netxen_get_flash_block(struct netxen_adapter *adapter, int base,
{
int i, v, addr;
__le32 *ptr32;
+ int ret;
addr = base;
ptr32 = buf;
for (i = 0; i < size / sizeof(u32); i++) {
- if (netxen_rom_fast_read(adapter, addr, &v) == -1)
- return -1;
+ ret = netxen_rom_fast_read(adapter, addr, &v);
+ if (ret)
+ return ret;
+
*ptr32 = cpu_to_le32(v);
ptr32++;
addr += sizeof(u32);
}
if ((char *)buf + size > (char *)ptr32) {
__le32 local;
- if (netxen_rom_fast_read(adapter, addr, &v) == -1)
- return -1;
+ ret = netxen_rom_fast_read(adapter, addr, &v);
+ if (ret)
+ return ret;
local = cpu_to_le32(v);
memcpy(ptr32, &local, (char *)buf + size - (char *)ptr32);
}
^ permalink raw reply related
* [patch 2/2] netxen: reversed condition in netxen_nic_set_link_parameters()
From: Dan Carpenter @ 2016-05-05 13:19 UTC (permalink / raw)
To: Manish Chopra
Cc: Sony Chacko, Rajesh Borundia, netdev, linux-kernel,
kernel-janitors
My static checker complains that we are using "autoneg" without
initializing it. The problem is the ->phy_read() condition is reversed
so we only set this on error instead of success.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
diff --git a/drivers/net/ethernet/qlogic/netxen/netxen_nic_hw.c b/drivers/net/ethernet/qlogic/netxen/netxen_nic_hw.c
index a320541..2b10f1b 100644
--- a/drivers/net/ethernet/qlogic/netxen/netxen_nic_hw.c
+++ b/drivers/net/ethernet/qlogic/netxen/netxen_nic_hw.c
@@ -1944,7 +1944,7 @@ void netxen_nic_set_link_parameters(struct netxen_adapter *adapter)
if (adapter->phy_read &&
adapter->phy_read(adapter,
NETXEN_NIU_GB_MII_MGMT_ADDR_AUTONEG,
- &autoneg) != 0)
+ &autoneg) == 0)
adapter->link_autoneg = autoneg;
} else
goto link_down;
^ permalink raw reply related
* [patch] netxen: netxen_rom_fast_read() doesn't return -1
From: Dan Carpenter @ 2016-05-05 13:20 UTC (permalink / raw)
To: Manish Chopra
Cc: Sony Chacko, Rajesh Borundia, netdev, linux-kernel,
kernel-janitors
The error handling is broken here. netxen_rom_fast_read() returns zero
on success and -EIO on error. It never returns -1.
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
diff --git a/drivers/net/ethernet/qlogic/netxen/netxen_nic_main.c b/drivers/net/ethernet/qlogic/netxen/netxen_nic_main.c
index fd362b6..9c6eed9 100644
--- a/drivers/net/ethernet/qlogic/netxen/netxen_nic_main.c
+++ b/drivers/net/ethernet/qlogic/netxen/netxen_nic_main.c
@@ -852,7 +852,8 @@ netxen_check_options(struct netxen_adapter *adapter)
ptr32 = (__le32 *)&serial_num;
offset = NX_FW_SERIAL_NUM_OFFSET;
for (i = 0; i < 8; i++) {
- if (netxen_rom_fast_read(adapter, offset, &val) == -1) {
+ err = netxen_rom_fast_read(adapter, offset, &val);
+ if (err) {
dev_err(&pdev->dev, "error reading board info\n");
adapter->driver_mismatch = 1;
return;
^ permalink raw reply related
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