* Re: [PATCH v6 2/9] ethtool: enable GSO and GRO by default
From: David Miller @ 2011-02-16 2:48 UTC (permalink / raw)
To: mirq-linux; +Cc: netdev, bhutchings
In-Reply-To: <d256f661690245d75bec50f5a6acafcefcae1d8a.1297823573.git.mirq-linux@rere.qmqm.pl>
You can't just update specific patches, you have to freshly resend
the entire series again if you want me to apply your work.
^ permalink raw reply
* [PATCH v6 5/9] net: Introduce new feature setting ops
From: Michał Mirosław @ 2011-02-16 2:43 UTC (permalink / raw)
To: netdev; +Cc: Ben Hutchings, David Miller
In-Reply-To: <33a801222e3b4254ff15de0fc5799de82f05aea8.1297594674.git.mirq-linux@rere.qmqm.pl>
This introduces a new framework to handle device features setting.
It consists of:
- new fields in struct net_device:
+ hw_features - features that hw/driver supports toggling
+ wanted_features - features that user wants enabled, when possible
- new netdev_ops:
+ feat = ndo_fix_features(dev, feat) - API checking constraints for
enabling features or their combinations
+ ndo_set_features(dev) - API updating hardware state to match
changed dev->features
- new ethtool commands:
+ ETHTOOL_GFEATURES/ETHTOOL_SFEATURES: get/set dev->wanted_features
and trigger device reconfiguration if resulting dev->features
changed
+ ETHTOOL_GSTRINGS(ETH_SS_FEATURES): get feature bits names (meaning)
Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
---
[updated to not generate a warning when registering non-SG capable device]
[rebased against current net-next]
include/linux/ethtool.h | 85 ++++++++++++++++++++++++++++++
include/linux/netdevice.h | 37 +++++++++++++-
net/core/dev.c | 46 ++++++++++++++--
net/core/ethtool.c | 125 ++++++++++++++++++++++++++++++++++++++++++++-
4 files changed, 283 insertions(+), 10 deletions(-)
diff --git a/include/linux/ethtool.h b/include/linux/ethtool.h
index 1908929..806e716 100644
--- a/include/linux/ethtool.h
+++ b/include/linux/ethtool.h
@@ -251,6 +251,7 @@ enum ethtool_stringset {
ETH_SS_STATS,
ETH_SS_PRIV_FLAGS,
ETH_SS_NTUPLE_FILTERS,
+ ETH_SS_FEATURES,
};
/* for passing string sets for data tagging */
@@ -523,6 +524,87 @@ struct ethtool_flash {
char data[ETHTOOL_FLASH_MAX_FILENAME];
};
+/* for returning and changing feature sets */
+
+/**
+ * struct ethtool_get_features_block - block with state of 32 features
+ * @available: mask of changeable features
+ * @requested: mask of features requested to be enabled if possible
+ * @active: mask of currently enabled features
+ * @never_changed: mask of features not changeable for any device
+ */
+struct ethtool_get_features_block {
+ __u32 available;
+ __u32 requested;
+ __u32 active;
+ __u32 never_changed;
+};
+
+/**
+ * struct ethtool_gfeatures - command to get state of device's features
+ * @cmd: command number = %ETHTOOL_GFEATURES
+ * @size: in: number of elements in the features[] array;
+ * out: number of elements in features[] needed to hold all features
+ * @features: state of features
+ */
+struct ethtool_gfeatures {
+ __u32 cmd;
+ __u32 size;
+ struct ethtool_get_features_block features[0];
+};
+
+/**
+ * struct ethtool_set_features_block - block with request for 32 features
+ * @valid: mask of features to be changed
+ * @requested: values of features to be changed
+ */
+struct ethtool_set_features_block {
+ __u32 valid;
+ __u32 requested;
+};
+
+/**
+ * struct ethtool_sfeatures - command to request change in device's features
+ * @cmd: command number = %ETHTOOL_SFEATURES
+ * @size: array size of the features[] array
+ * @features: feature change masks
+ */
+struct ethtool_sfeatures {
+ __u32 cmd;
+ __u32 size;
+ struct ethtool_set_features_block features[0];
+};
+
+/*
+ * %ETHTOOL_SFEATURES changes features present in features[].valid to the
+ * values of corresponding bits in features[].requested. Bits in .requested
+ * not set in .valid or not changeable are ignored.
+ *
+ * Returns %EINVAL when .valid contains undefined or never-changable bits
+ * or size is not equal to required number of features words (32-bit blocks).
+ * Returns >= 0 if request was completed; bits set in the value mean:
+ * %ETHTOOL_F_UNSUPPORTED - there were bits set in .valid that are not
+ * changeable (not present in %ETHTOOL_GFEATURES' features[].available)
+ * those bits were ignored.
+ * %ETHTOOL_F_WISH - some or all changes requested were recorded but the
+ * resulting state of bits masked by .valid is not equal to .requested.
+ * Probably there are other device-specific constraints on some features
+ * in the set. When %ETHTOOL_F_UNSUPPORTED is set, .valid is considered
+ * here as though ignored bits were cleared.
+ *
+ * Meaning of bits in the masks are obtained by %ETHTOOL_GSSET_INFO (number of
+ * bits in the arrays - always multiple of 32) and %ETHTOOL_GSTRINGS commands
+ * for ETH_SS_FEATURES string set. First entry in the table corresponds to least
+ * significant bit in features[0] fields. Empty strings mark undefined features.
+ */
+enum ethtool_sfeatures_retval_bits {
+ ETHTOOL_F_UNSUPPORTED__BIT,
+ ETHTOOL_F_WISH__BIT,
+};
+
+#define ETHTOOL_F_UNSUPPORTED (1 << ETHTOOL_F_UNSUPPORTED__BIT)
+#define ETHTOOL_F_WISH (1 << ETHTOOL_F_WISH__BIT)
+
#ifdef __KERNEL__
#include <linux/rculist.h>
@@ -744,6 +826,9 @@ struct ethtool_ops {
#define ETHTOOL_GRXFHINDIR 0x00000038 /* Get RX flow hash indir'n table */
#define ETHTOOL_SRXFHINDIR 0x00000039 /* Set RX flow hash indir'n table */
+#define ETHTOOL_GFEATURES 0x0000003a /* Get device offload settings */
+#define ETHTOOL_SFEATURES 0x0000003b /* Change device offload settings */
+
/* compatibility with older code */
#define SPARC_ETH_GSET ETHTOOL_GSET
#define SPARC_ETH_SSET ETHTOOL_SSET
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index dede3fd..85f67e2 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -791,6 +791,18 @@ struct netdev_tc_txq {
*
* int (*ndo_del_slave)(struct net_device *dev, struct net_device *slave_dev);
* Called to release previously enslaved netdev.
+ *
+ * Feature/offload setting functions.
+ * u32 (*ndo_fix_features)(struct net_device *dev, u32 features);
+ * Adjusts the requested feature flags according to device-specific
+ * constraints, and returns the resulting flags. Must not modify
+ * the device state.
+ *
+ * int (*ndo_set_features)(struct net_device *dev, u32 features);
+ * Called to update device configuration to new features. Passed
+ * feature set might be less than what was returned by ndo_fix_features()).
+ * Must return >0 or -errno if it changed dev->features itself.
+ *
*/
#define HAVE_NET_DEVICE_OPS
struct net_device_ops {
@@ -874,6 +886,10 @@ struct net_device_ops {
struct net_device *slave_dev);
int (*ndo_del_slave)(struct net_device *dev,
struct net_device *slave_dev);
+ u32 (*ndo_fix_features)(struct net_device *dev,
+ u32 features);
+ int (*ndo_set_features)(struct net_device *dev,
+ u32 features);
};
/*
@@ -925,12 +941,18 @@ struct net_device {
struct list_head napi_list;
struct list_head unreg_list;
- /* Net device features */
+ /* currently active device features */
u32 features;
-
+ /* user-changeable features */
+ u32 hw_features;
+ /* user-requested features */
+ u32 wanted_features;
/* VLAN feature mask */
u32 vlan_features;
+ /* Net device feature bits; if you change something,
+ * also update netdev_features_strings[] in ethtool.c */
+
#define NETIF_F_SG 1 /* Scatter/gather IO. */
#define NETIF_F_IP_CSUM 2 /* Can checksum TCP/UDP over IPv4. */
#define NETIF_F_NO_CSUM 4 /* Does not require checksum. F.e. loopack. */
@@ -966,6 +988,12 @@ struct net_device {
#define NETIF_F_TSO6 (SKB_GSO_TCPV6 << NETIF_F_GSO_SHIFT)
#define NETIF_F_FSO (SKB_GSO_FCOE << NETIF_F_GSO_SHIFT)
+ /* Features valid for ethtool to change */
+ /* = all defined minus driver/device-class-related */
+#define NETIF_F_NEVER_CHANGE (NETIF_F_HIGHDMA | NETIF_F_VLAN_CHALLENGED | \
+ NETIF_F_LLTX | NETIF_F_NETNS_LOCAL)
+#define NETIF_F_ETHTOOL_BITS (0x1f3fffff & ~NETIF_F_NEVER_CHANGE)
+
/* List of features with software fallbacks. */
#define NETIF_F_GSO_SOFTWARE (NETIF_F_TSO | NETIF_F_TSO_ECN | \
NETIF_F_TSO6 | NETIF_F_UFO)
@@ -2428,8 +2456,13 @@ extern char *netdev_drivername(const struct net_device *dev, char *buffer, int l
extern void linkwatch_run_queue(void);
+static inline u32 netdev_get_wanted_features(struct net_device *dev)
+{
+ return (dev->features & ~dev->hw_features) | dev->wanted_features;
+}
u32 netdev_increment_features(u32 all, u32 one, u32 mask);
u32 netdev_fix_features(struct net_device *dev, u32 features);
+void netdev_update_features(struct net_device *dev);
void netif_stacked_transfer_operstate(const struct net_device *rootdev,
struct net_device *dev);
diff --git a/net/core/dev.c b/net/core/dev.c
index 8686f6f..4f69439 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -5302,6 +5302,37 @@ u32 netdev_fix_features(struct net_device *dev, u32 features)
}
EXPORT_SYMBOL(netdev_fix_features);
+void netdev_update_features(struct net_device *dev)
+{
+ u32 features;
+ int err = 0;
+
+ features = netdev_get_wanted_features(dev);
+
+ if (dev->netdev_ops->ndo_fix_features)
+ features = dev->netdev_ops->ndo_fix_features(dev, features);
+
+ /* driver might be less strict about feature dependencies */
+ features = netdev_fix_features(dev, features);
+
+ if (dev->features == features)
+ return;
+
+ netdev_info(dev, "Features changed: 0x%08x -> 0x%08x\n",
+ dev->features, features);
+
+ if (dev->netdev_ops->ndo_set_features)
+ err = dev->netdev_ops->ndo_set_features(dev, features);
+
+ if (!err)
+ dev->features = features;
+ else if (err < 0)
+ netdev_err(dev,
+ "set_features() failed (%d); wanted 0x%08x, left 0x%08x\n",
+ err, features, dev->features);
+}
+EXPORT_SYMBOL(netdev_update_features);
+
/**
* netif_stacked_transfer_operstate - transfer operstate
* @rootdev: the root or lower level device to transfer state from
@@ -5436,15 +5467,18 @@ int register_netdevice(struct net_device *dev)
if (dev->iflink == -1)
dev->iflink = dev->ifindex;
- /* Enable software offloads by default - will be stripped in
- * netdev_fix_features() if not supported. */
- dev->features |= NETIF_F_SOFT_FEATURES;
+ /* Transfer changeable features to wanted_features and enable
+ * software offloads (GSO and GRO).
+ */
+ dev->hw_features |= NETIF_F_SOFT_FEATURES;
+ dev->wanted_features = (dev->features & dev->hw_features)
+ | NETIF_F_SOFT_FEATURES;
/* Avoid warning from netdev_fix_features() for GSO without SG */
- if (!(dev->features & NETIF_F_SG))
- dev->features &= ~NETIF_F_GSO;
+ if (!(dev->wanted_features & NETIF_F_SG))
+ dev->wanted_features &= ~NETIF_F_GSO;
- dev->features = netdev_fix_features(dev, dev->features);
+ netdev_update_features(dev);
/* Enable GRO and NETIF_F_HIGHDMA for vlans by default,
* vlan_dev_init() will do the dev->features check, so these features
diff --git a/net/core/ethtool.c b/net/core/ethtool.c
index c3fb8f9..9577396 100644
--- a/net/core/ethtool.c
+++ b/net/core/ethtool.c
@@ -172,10 +172,120 @@ EXPORT_SYMBOL(ethtool_ntuple_flush);
/* Handlers for each ethtool command */
+#define ETHTOOL_DEV_FEATURE_WORDS 1
+
+static int ethtool_get_features(struct net_device *dev, void __user *useraddr)
+{
+ struct ethtool_gfeatures cmd = {
+ .cmd = ETHTOOL_GFEATURES,
+ .size = ETHTOOL_DEV_FEATURE_WORDS,
+ };
+ struct ethtool_get_features_block features[ETHTOOL_DEV_FEATURE_WORDS] = {
+ {
+ .available = dev->hw_features,
+ .requested = dev->wanted_features,
+ .active = dev->features,
+ .never_changed = NETIF_F_NEVER_CHANGE,
+ },
+ };
+ u32 __user *sizeaddr;
+ u32 copy_size;
+
+ sizeaddr = useraddr + offsetof(struct ethtool_gfeatures, size);
+ if (get_user(copy_size, sizeaddr))
+ return -EFAULT;
+
+ if (copy_size > ETHTOOL_DEV_FEATURE_WORDS)
+ copy_size = ETHTOOL_DEV_FEATURE_WORDS;
+
+ if (copy_to_user(useraddr, &cmd, sizeof(cmd)))
+ return -EFAULT;
+ useraddr += sizeof(cmd);
+ if (copy_to_user(useraddr, features, copy_size * sizeof(*features)))
+ return -EFAULT;
+
+ return 0;
+}
+
+static int ethtool_set_features(struct net_device *dev, void __user *useraddr)
+{
+ struct ethtool_sfeatures cmd;
+ struct ethtool_set_features_block features[ETHTOOL_DEV_FEATURE_WORDS];
+ int ret = 0;
+
+ if (copy_from_user(&cmd, useraddr, sizeof(cmd)))
+ return -EFAULT;
+ useraddr += sizeof(cmd);
+
+ if (cmd.size != ETHTOOL_DEV_FEATURE_WORDS)
+ return -EINVAL;
+
+ if (copy_from_user(features, useraddr, sizeof(features)))
+ return -EFAULT;
+
+ if (features[0].valid & ~NETIF_F_ETHTOOL_BITS)
+ return -EINVAL;
+
+ if (features[0].valid & ~dev->hw_features) {
+ features[0].valid &= dev->hw_features;
+ ret |= ETHTOOL_F_UNSUPPORTED;
+ }
+
+ dev->wanted_features &= ~features[0].valid;
+ dev->wanted_features |= features[0].valid & features[0].requested;
+ netdev_update_features(dev);
+
+ if ((dev->wanted_features ^ dev->features) & features[0].valid)
+ ret |= ETHTOOL_F_WISH;
+
+ return ret;
+}
+
+static const char netdev_features_strings[ETHTOOL_DEV_FEATURE_WORDS * 32][ETH_GSTRING_LEN] = {
+ /* NETIF_F_SG */ "tx-scatter-gather",
+ /* NETIF_F_IP_CSUM */ "tx-checksum-ipv4",
+ /* NETIF_F_NO_CSUM */ "tx-checksum-unneeded",
+ /* NETIF_F_HW_CSUM */ "tx-checksum-ip-generic",
+ /* NETIF_F_IPV6_CSUM */ "tx_checksum-ipv6",
+ /* NETIF_F_HIGHDMA */ "highdma",
+ /* NETIF_F_FRAGLIST */ "tx-scatter-gather-fraglist",
+ /* NETIF_F_HW_VLAN_TX */ "tx-vlan-hw-insert",
+
+ /* NETIF_F_HW_VLAN_RX */ "rx-vlan-hw-parse",
+ /* NETIF_F_HW_VLAN_FILTER */ "rx-vlan-filter",
+ /* NETIF_F_VLAN_CHALLENGED */ "vlan-challenged",
+ /* NETIF_F_GSO */ "tx-generic-segmentation",
+ /* NETIF_F_LLTX */ "tx-lockless",
+ /* NETIF_F_NETNS_LOCAL */ "netns-local",
+ /* NETIF_F_GRO */ "rx-gro",
+ /* NETIF_F_LRO */ "rx-lro",
+
+ /* NETIF_F_TSO */ "tx-tcp-segmentation",
+ /* NETIF_F_UFO */ "tx-udp-fragmentation",
+ /* NETIF_F_GSO_ROBUST */ "tx-gso-robust",
+ /* NETIF_F_TSO_ECN */ "tx-tcp-ecn-segmentation",
+ /* NETIF_F_TSO6 */ "tx-tcp6-segmentation",
+ /* NETIF_F_FSO */ "tx-fcoe-segmentation",
+ "",
+ "",
+
+ /* NETIF_F_FCOE_CRC */ "tx-checksum-fcoe-crc",
+ /* NETIF_F_SCTP_CSUM */ "tx-checksum-sctp",
+ /* NETIF_F_FCOE_MTU */ "fcoe-mtu",
+ /* NETIF_F_NTUPLE */ "rx-ntuple-filter",
+ /* NETIF_F_RXHASH */ "rx-hashing",
+ "",
+ "",
+ "",
+};
+
static int __ethtool_get_sset_count(struct net_device *dev, int sset)
{
const struct ethtool_ops *ops = dev->ethtool_ops;
+ if (sset == ETH_SS_FEATURES)
+ return ARRAY_SIZE(netdev_features_strings);
+
if (ops && ops->get_sset_count && ops->get_strings)
return ops->get_sset_count(dev, sset);
else
@@ -187,8 +297,12 @@ static void __ethtool_get_strings(struct net_device *dev,
{
const struct ethtool_ops *ops = dev->ethtool_ops;
- /* ops->get_strings is valid because checked earlier */
- ops->get_strings(dev, stringset, data);
+ if (stringset == ETH_SS_FEATURES)
+ memcpy(data, netdev_features_strings,
+ sizeof(netdev_features_strings));
+ else
+ /* ops->get_strings is valid because checked earlier */
+ ops->get_strings(dev, stringset, data);
}
static u32 ethtool_get_feature_mask(u32 eth_cmd)
@@ -1533,6 +1647,7 @@ int dev_ethtool(struct net *net, struct ifreq *ifr)
case ETHTOOL_GRXCLSRLCNT:
case ETHTOOL_GRXCLSRULE:
case ETHTOOL_GRXCLSRLALL:
+ case ETHTOOL_GFEATURES:
break;
default:
if (!capable(CAP_NET_ADMIN))
@@ -1678,6 +1793,12 @@ int dev_ethtool(struct net *net, struct ifreq *ifr)
case ETHTOOL_SRXFHINDIR:
rc = ethtool_set_rxfh_indir(dev, useraddr);
break;
+ case ETHTOOL_GFEATURES:
+ rc = ethtool_get_features(dev, useraddr);
+ break;
+ case ETHTOOL_SFEATURES:
+ rc = ethtool_set_features(dev, useraddr);
+ break;
case ETHTOOL_GTXCSUM:
case ETHTOOL_GSG:
case ETHTOOL_GTSO:
--
1.7.2.3
^ permalink raw reply related
* [PATCH v6 2/9] ethtool: enable GSO and GRO by default
From: Michał Mirosław @ 2011-02-16 2:43 UTC (permalink / raw)
To: netdev; +Cc: Ben Hutchings, David Miller
In-Reply-To: <20110216024237.GA6434@rere.qmqm.pl>
Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
---
[updated to not generate a warning when registering non-SG capable device]
include/linux/netdevice.h | 3 +++
net/core/dev.c | 18 ++++++++++++++----
2 files changed, 17 insertions(+), 4 deletions(-)
diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
index d08ef65..168e3ad 100644
--- a/include/linux/netdevice.h
+++ b/include/linux/netdevice.h
@@ -984,6 +984,9 @@ struct net_device {
NETIF_F_SG | NETIF_F_HIGHDMA | \
NETIF_F_FRAGLIST)
+ /* changeable features with no special hardware requirements */
+#define NETIF_F_SOFT_FEATURES (NETIF_F_GSO | NETIF_F_GRO)
+
/* Interface index. Unique device identifier */
int ifindex;
int iflink;
diff --git a/net/core/dev.c b/net/core/dev.c
index 4580460..8686f6f 100644
--- a/net/core/dev.c
+++ b/net/core/dev.c
@@ -5274,6 +5274,12 @@ u32 netdev_fix_features(struct net_device *dev, u32 features)
features &= ~NETIF_F_TSO;
}
+ /* Software GSO depends on SG. */
+ if ((features & NETIF_F_GSO) && !(features & NETIF_F_SG)) {
+ netdev_info(dev, "Dropping NETIF_F_GSO since no SG feature.\n");
+ features &= ~NETIF_F_GSO;
+ }
+
/* UFO needs SG and checksumming */
if (features & NETIF_F_UFO) {
/* maybe split UFO into V4 and V6? */
@@ -5430,12 +5436,16 @@ int register_netdevice(struct net_device *dev)
if (dev->iflink == -1)
dev->iflink = dev->ifindex;
+ /* Enable software offloads by default - will be stripped in
+ * netdev_fix_features() if not supported. */
+ dev->features |= NETIF_F_SOFT_FEATURES;
+
+ /* Avoid warning from netdev_fix_features() for GSO without SG */
+ if (!(dev->features & NETIF_F_SG))
+ dev->features &= ~NETIF_F_GSO;
+
dev->features = netdev_fix_features(dev, dev->features);
- /* Enable software GSO if SG is supported. */
- if (dev->features & NETIF_F_SG)
- dev->features |= NETIF_F_GSO;
-
/* Enable GRO and NETIF_F_HIGHDMA for vlans by default,
* vlan_dev_init() will do the dev->features check, so these features
* are enabled only if supported by underlying device.
--
1.7.2.3
^ permalink raw reply related
* Re: [PATCH v5 RESEND 2/9] ethtool: enable GSO and GRO by default
From: Michał Mirosław @ 2011-02-16 2:42 UTC (permalink / raw)
To: David Miller; +Cc: netdev, bhutchings
In-Reply-To: <20110215.140043.59699086.davem@davemloft.net>
On Tue, Feb 15, 2011 at 02:00:43PM -0800, David Miller wrote:
> From: Michał Mirosław <mirq-linux@rere.qmqm.pl>
> Date: Tue, 15 Feb 2011 22:46:49 +0100
> > On Sun, Feb 13, 2011 at 10:50:23AM -0800, David Miller wrote:
> >> From: Michał Mirosław <mirq-linux@rere.qmqm.pl>
> >> Date: Sun, 13 Feb 2011 12:11:45 +0100 (CET)
> >> > Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
> >> This is not appropriate.
> >>
> >> Now, every driver that lacks SG support will spit out that warning
> >> message in netdev_fix_features().
> >>
> >> That's why the check is there conditionalizing NETIF_F_GSO on
> >> NETIF_F_SG in register_netdevice().
> > I think all those messages should be converted to DEBUG level. Those
> > conditions are constant and can be better described in
> > Documentation/networking/ or ethtool manpage. Preferably along the
> > device-specific conditions when implemented.
> >
> > Or I could just drop the message for the GSO case as it's something new here
> > anyway (I added it to make it consistent with handling of other features).
> The messages exist to let driver authors know they've constructed an
> illegal set of feature bits.
Only multiple checksum flags are actual driver bugs - those are handled
in register_netdevice(). netdev_fix_features() represents limitations of core
networking code . Hardware might as well handle TSO without SG or SG without
checksumming. What's the point in duplicating those restrictions in drivers?
> Since you're now adding the GSO bit yourself, you should perform
> due diligence and prevent the illegal combination yourself.
>
> This has no other impact on the other messages and cases, which
> definitely should stay intact.
>
> Your change is just wrong and knowingly introduces useless log
> messages, please just fix it up.
The messages from netdev_fix_features() will also be triggered by user
disabling required feature (ie. HW checksum for most of TX offloads)
for a device or its slave or sometimes on MTU changes (ie. for jme and
possibly others). Even after I fix triggering of "GSO needs SG"
message in register_netdevice(), this and other messages will show up
more often because of the way the new scheme works.
Best Regards,
Michał Mirosław
^ permalink raw reply
* Re: [kernel] net.ipv6.route.max_size
From: Mr. James W. Laferriere @ 2011-02-16 1:29 UTC (permalink / raw)
To: Lukasz Trabinski; +Cc: Linux Kernel Maillist, NetDev
In-Reply-To: <alpine.LFD.2.00.1102142024460.6561@oceanic.wsisiz.edu.pl>
Hello Lukasz ,
Since I have not seen anyone else suggest this or to do so I'll forward
this onto the NetDev <netdev@vger.kernel.org> list for their perusal as well .
Hth , JimL
On Mon, 14 Feb 2011, Lukasz Trabinski wrote:
> Hello
>
> IPv6 route table size is still growing. In stanard linux box configuration
> with IPv6 BGP session (with ipv6 world table) and BGP application like
> Quagga,
> we have problem like this:
>
> 2011/02/12 18:43:52 ZEBRA: netlink-cmd error: Cannot allocate memory,
> type=RTM_NEWROUTE(24), seq=11124600, pid=0
>
> This problem shows us that's all prefixes comes from BGP can't be install
> into FIB
>
> IN BGP application we can see, that's we recive 4484 prefixes from ipv6
> neighbor:
>
> cosmos> show bgp summary
> Neighbor V AS MsgRcvd MsgSent TblVer InQ OutQ Up/Down
> State/PfxRcd
> 2001:1a68:0:c::1
> 4 24724 431945 70172 0 0 0 2d01h18m 4484
>
>
> I suggest that net->ipv6.sysctl.ip6_rt_max_size should be increase.
> Below _simple_ patch for this problem. I don't understand why in ipv4/route
> we have different method to count ip4_rt_max_size? My value 65536 in near
> future will be also wrong. I know also, that we can change it by sysctl, but
> I thing that it's not good way to fix it.
>
>
> --- route.c~ 2011-02-14 20:26:12.000000000 +0100
> +++ route.c 2011-02-14 20:30:04.000000000 +0100
> @@ -2708,7 +2708,7 @@
> #endif
>
> net->ipv6.sysctl.flush_delay = 0;
> - net->ipv6.sysctl.ip6_rt_max_size = 4096;
> + net->ipv6.sysctl.ip6_rt_max_size = 65536;
> net->ipv6.sysctl.ip6_rt_gc_min_interval = HZ / 2;
> net->ipv6.sysctl.ip6_rt_gc_timeout = 60*HZ;
> net->ipv6.sysctl.ip6_rt_gc_interval = 30*HZ;
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
> the body of a message to majordomo@vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
> Please read the FAQ at http://www.tux.org/lkml/
>
--
+------------------------------------------------------------------+
| James W. Laferriere | System Techniques | Give me VMS |
| Network&System Engineer | 3237 Holden Road | Give me Linux |
| babydr@baby-dragons.com | Fairbanks, AK. 99709 | only on AXP |
+------------------------------------------------------------------+
^ permalink raw reply
* Re: Multicast snooping fixes and suggestions
From: Stephen Hemminger @ 2011-02-15 23:41 UTC (permalink / raw)
To: Linus Lüssing, Herbert Xu
Cc: David S. Miller, bridge, netdev, linux-kernel
In-Reply-To: <1297811961-19249-1-git-send-email-linus.luessing@web.de>
On Wed, 16 Feb 2011 00:19:16 +0100
Linus Lüssing <linus.luessing@web.de> wrote:
> Hello everyone,
>
> While testing the (very awesome!) bridge igmp/mld snooping support I came across
> two issues which are breaking IPv6 multicast snooping and IPv6
> non-link-local multicast on bridges with multicast snooping support enabled
> in general. The first two patches shall fix these issues.
>
> The third one addresses a potential bug on little endian machines which I noticed
> during this little code reviewing. This patch is untested though, feedback welcome.
>
> The fourth and fifth patch are a suggestion to also permit using the bridge multicast
> snooping feature for link local multimedia multicast traffic. Therefore
> using the transient multicast flag instead of the non-link-local scope criteria
> seems to be a suitable solution at least for IPv6, in my opinion. Let me know what
> you think about it.
>
> Thanks for reviewing these patches.
>
> Cheers,
> Linus
These look correct. Did you test them with real traffic?
^ permalink raw reply
* [PATCH 4/5] ipv6: Add IPv6 multicast address flag defines
From: Linus Lüssing @ 2011-02-15 23:19 UTC (permalink / raw)
To: Stephen Hemminger, David S. Miller, bridge
Cc: netdev, linux-kernel, Herbert Xu, Linus Lüssing
In-Reply-To: <1297811961-19249-1-git-send-email-linus.luessing@web.de>
This commit adds the missing IPv6 multicast address flag defines to
complement the already existing multicast address scope defines and to
be able to check these flags nicely in the future.
Signed-off-by: Linus Lüssing <linus.luessing@web.de>
---
include/net/ipv6.h | 12 ++++++++++++
1 files changed, 12 insertions(+), 0 deletions(-)
diff --git a/include/net/ipv6.h b/include/net/ipv6.h
index 4a3cd2c..96e50e0 100644
--- a/include/net/ipv6.h
+++ b/include/net/ipv6.h
@@ -89,6 +89,18 @@
#define IPV6_ADDR_SCOPE_GLOBAL 0x0e
/*
+ * Addr flags
+ */
+#ifdef __KERNEL__
+#define IPV6_ADDR_MC_FLAG_TRANSIENT(a) \
+ ((a)->s6_addr[1] & 0x10)
+#define IPV6_ADDR_MC_FLAG_PREFIX(a) \
+ ((a)->s6_addr[1] & 0x20)
+#define IPV6_ADDR_MC_FLAG_RENDEZVOUS(a) \
+ ((a)->s6_addr[1] & 0x40)
+#endif
+
+/*
* fragmentation header
*/
--
1.7.2.3
^ permalink raw reply related
* [PATCH 5/5] bridge: Allow mcast snooping for transient link local addresses too
From: Linus Lüssing @ 2011-02-15 23:19 UTC (permalink / raw)
To: Stephen Hemminger, David S. Miller, bridge
Cc: netdev, linux-kernel, Herbert Xu, Linus Lüssing
In-Reply-To: <1297811961-19249-1-git-send-email-linus.luessing@web.de>
Currently the multicast bridge snooping support is not active for
link local multicast. I assume this has been done to leave
important multicast data untouched, like IPv6 Neighborhood Discovery.
In larger, bridged, local networks it could however be desirable to
optimize for instance local multicast audio/video streaming too.
With the transient flag in IPv6 multicast addresses we have an easy
way to optimize such multimedia traffic without tempering with the
high priority multicast data from well-known addresses.
This patch alters the multicast bridge snooping for IPv6, to take
effect for transient multicast addresses instead of non-link-local
addresses.
Signed-off-by: Linus Lüssing <linus.luessing@web.de>
---
net/bridge/br_multicast.c | 9 ++++-----
1 files changed, 4 insertions(+), 5 deletions(-)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index e8fdaab..b5eb28a 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -37,10 +37,9 @@
rcu_dereference_protected(X, lockdep_is_held(&br->multicast_lock))
#if defined(CONFIG_IPV6) || defined(CONFIG_IPV6_MODULE)
-static inline int ipv6_is_local_multicast(const struct in6_addr *addr)
+static inline int ipv6_is_transient_multicast(const struct in6_addr *addr)
{
- if (ipv6_addr_is_multicast(addr) &&
- IPV6_ADDR_MC_SCOPE(addr) <= IPV6_ADDR_SCOPE_LINKLOCAL)
+ if (ipv6_addr_is_multicast(addr) && IPV6_ADDR_MC_FLAG_TRANSIENT(addr))
return 1;
return 0;
}
@@ -781,7 +780,7 @@ static int br_ip6_multicast_add_group(struct net_bridge *br,
{
struct br_ip br_group;
- if (ipv6_is_local_multicast(group))
+ if (!ipv6_is_transient_multicast(group))
return 0;
ipv6_addr_copy(&br_group.u.ip6, group);
@@ -1342,7 +1341,7 @@ static void br_ip6_multicast_leave_group(struct net_bridge *br,
{
struct br_ip br_group;
- if (ipv6_is_local_multicast(group))
+ if (!ipv6_is_transient_multicast(group))
return;
ipv6_addr_copy(&br_group.u.ip6, group);
--
1.7.2.3
^ permalink raw reply related
* [PATCH 3/5] bridge: Add missing ntohs()s for MLDv2 report parsing
From: Linus Lüssing @ 2011-02-15 23:19 UTC (permalink / raw)
To: Stephen Hemminger, David S. Miller, bridge
Cc: netdev, linux-kernel, Herbert Xu, Linus Lüssing
In-Reply-To: <1297811961-19249-1-git-send-email-linus.luessing@web.de>
The nsrcs number is 2 Byte wide, therefore we need to call ntohs()
before using it.
Signed-off-by: Linus Lüssing <linus.luessing@web.de>
---
net/bridge/br_multicast.c | 5 +++--
1 files changed, 3 insertions(+), 2 deletions(-)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index 45dcf10..e8fdaab 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -1021,11 +1021,12 @@ static int br_ip6_multicast_mld2_report(struct net_bridge *br,
if (!pskb_may_pull(skb,
len + sizeof(*grec) +
- sizeof(struct in6_addr) * (*nsrcs)))
+ sizeof(struct in6_addr) * ntohs(*nsrcs)))
return -EINVAL;
grec = (struct mld2_grec *)(skb->data + len);
- len += sizeof(*grec) + sizeof(struct in6_addr) * (*nsrcs);
+ len += sizeof(*grec) +
+ sizeof(struct in6_addr) * ntohs(*nsrcs);
/* We treat these as MLDv1 reports for now. */
switch (grec->grec_type) {
--
1.7.2.3
^ permalink raw reply related
* [PATCH 2/5] bridge: Fix IPv6 multicast snooping by correcting offset in MLDv2 report
From: Linus Lüssing @ 2011-02-15 23:19 UTC (permalink / raw)
To: Stephen Hemminger, David S. Miller, bridge
Cc: netdev, linux-kernel, Herbert Xu, Linus Lüssing
In-Reply-To: <1297811961-19249-1-git-send-email-linus.luessing@web.de>
We actually want a pointer to the grec_nsrcr and not the following
field. Otherwise we can get very high values for *nsrcs as the first two
bytes of the IPv6 multicast address are being used instead, leading to
a failing pskb_may_pull() which results in MLDv2 reports not being
parsed.
Signed-off-by: Linus Lüssing <linus.luessing@web.de>
---
net/bridge/br_multicast.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index 135d929..45dcf10 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -1014,7 +1014,7 @@ static int br_ip6_multicast_mld2_report(struct net_bridge *br,
nsrcs = skb_header_pointer(skb,
len + offsetof(struct mld2_grec,
- grec_mca),
+ grec_nsrcs),
sizeof(_nsrcs), &_nsrcs);
if (!nsrcs)
return -EINVAL;
--
1.7.2.3
^ permalink raw reply related
* [PATCH 1/5] bridge: Fix IPv6 multicast snooping by storing correct protocol type
From: Linus Lüssing @ 2011-02-15 23:19 UTC (permalink / raw)
To: Stephen Hemminger, David S. Miller, bridge
Cc: netdev, linux-kernel, Herbert Xu, Linus Lüssing
In-Reply-To: <1297811961-19249-1-git-send-email-linus.luessing@web.de>
The protocol type for IPv6 entries in the hash table for multicast
bridge snooping is falsely set to ETH_P_IP, marking it as an IPv4
address, instead of setting it to ETH_P_IPV6, which results in negative
look-ups in the hash table later.
Signed-off-by: Linus Lüssing <linus.luessing@web.de>
---
net/bridge/br_multicast.c | 2 +-
1 files changed, 1 insertions(+), 1 deletions(-)
diff --git a/net/bridge/br_multicast.c b/net/bridge/br_multicast.c
index f701a21..135d929 100644
--- a/net/bridge/br_multicast.c
+++ b/net/bridge/br_multicast.c
@@ -785,7 +785,7 @@ static int br_ip6_multicast_add_group(struct net_bridge *br,
return 0;
ipv6_addr_copy(&br_group.u.ip6, group);
- br_group.proto = htons(ETH_P_IP);
+ br_group.proto = htons(ETH_P_IPV6);
return br_multicast_add_group(br, port, &br_group);
}
--
1.7.2.3
^ permalink raw reply related
* Multicast snooping fixes and suggestions
From: Linus Lüssing @ 2011-02-15 23:19 UTC (permalink / raw)
To: Stephen Hemminger, David S. Miller, bridge
Cc: netdev, linux-kernel, Herbert Xu
Hello everyone,
While testing the (very awesome!) bridge igmp/mld snooping support I came across
two issues which are breaking IPv6 multicast snooping and IPv6
non-link-local multicast on bridges with multicast snooping support enabled
in general. The first two patches shall fix these issues.
The third one addresses a potential bug on little endian machines which I noticed
during this little code reviewing. This patch is untested though, feedback welcome.
The fourth and fifth patch are a suggestion to also permit using the bridge multicast
snooping feature for link local multimedia multicast traffic. Therefore
using the transient multicast flag instead of the non-link-local scope criteria
seems to be a suitable solution at least for IPv6, in my opinion. Let me know what
you think about it.
Thanks for reviewing these patches.
Cheers,
Linus
^ permalink raw reply
* Re: [PATCH] core: dev: don't call BUG() on bad input
From: Stephen Hemminger @ 2011-02-15 23:02 UTC (permalink / raw)
To: Vasiliy Kulikov
Cc: linux-kernel, David S. Miller, Eric Dumazet, Tom Herbert,
Changli Gao, Jesse Gross, netdev
In-Reply-To: <1297680967-11893-1-git-send-email-segoon@openwall.com>
On Mon, 14 Feb 2011 13:56:06 +0300
Vasiliy Kulikov <segoon@openwall.com> wrote:
> alloc_netdev() may be called with too long name (more that IFNAMSIZ bytes).
> Currently this leads to BUG(). Other insane inputs (bad txqs, rxqs) and
> even OOM don't lead to BUG(). Made alloc_netdev() return NULL, like on
> other errors.
The only way alloc_netdev could be called with a name too long
was if some driver was incorrectly written. It is not something
that can be exercised by user space.
Please leave the BUG() so the driver will show up in
kernel oops logs etc.
^ permalink raw reply
* [PATCH net-2.6/stable] tg3: Restrict phy ioctl access
From: Matt Carlson @ 2011-02-15 22:51 UTC (permalink / raw)
To: davem; +Cc: netdev, mcarlson, stable
If management firmware is present and the device is down, the firmware
will assume control of the phy. If a phy access were allowed from the
host, it will collide with firmware phy accesses, resulting in
unpredictable behavior. This patch fixes the problem by disallowing phy
accesses during the problematic condition.
Upstream commit ID f746a3136a61ae535c5d0b49a9418fa21edc61b5
Signed-off-by: Matt Carlson <mcarlson@broadcom.com>
Reviewed-by: Michael Chan <mchan@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
---
drivers/net/tg3.c | 8 ++++++--
1 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/drivers/net/tg3.c b/drivers/net/tg3.c
index 93b32d3..06c0e503 100644
--- a/drivers/net/tg3.c
+++ b/drivers/net/tg3.c
@@ -11158,7 +11158,9 @@ static int tg3_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
if (tp->phy_flags & TG3_PHYFLG_PHY_SERDES)
break; /* We have no PHY */
- if (tp->phy_flags & TG3_PHYFLG_IS_LOW_POWER)
+ if ((tp->phy_flags & TG3_PHYFLG_IS_LOW_POWER) ||
+ ((tp->tg3_flags & TG3_FLAG_ENABLE_ASF) &&
+ !netif_running(dev)))
return -EAGAIN;
spin_lock_bh(&tp->lock);
@@ -11174,7 +11176,9 @@ static int tg3_ioctl(struct net_device *dev, struct ifreq *ifr, int cmd)
if (tp->phy_flags & TG3_PHYFLG_PHY_SERDES)
break; /* We have no PHY */
- if (tp->phy_flags & TG3_PHYFLG_IS_LOW_POWER)
+ if ((tp->phy_flags & TG3_PHYFLG_IS_LOW_POWER) ||
+ ((tp->tg3_flags & TG3_FLAG_ENABLE_ASF) &&
+ !netif_running(dev)))
return -EAGAIN;
spin_lock_bh(&tp->lock);
--
1.7.3.4
^ permalink raw reply related
* Re: [PATCH] bnx2x: Support for managing RX indirection table
From: Tom Herbert @ 2011-02-15 22:50 UTC (permalink / raw)
To: Ben Hutchings; +Cc: davem, eilong, netdev
In-Reply-To: <1297788529.2584.12.camel@bwh-desktop>
>> + u32 rx_indir_table[128];
>
> Shouldn't the dimension be TSTORM_INDIRECTION_TABLE_SIZE?
>
It's not a defined constant, so the alternative would be to malloc it
which seems like overkill to me.
Broadcom guys: are there any adapters or configuration of bnx2x where
the indirection table would be greater than 128?
Tom
^ permalink raw reply
* Re: any way to reset all marked connections when using CONNMARK?
From: Jan Engelhardt @ 2011-02-15 22:42 UTC (permalink / raw)
To: Chris Friesen; +Cc: Andrew Beverley, netdev, netfilter-devel, netfilter
In-Reply-To: <4D5AF8CA.8080908@genband.com>
On Tuesday 2011-02-15 23:06, Chris Friesen wrote:
>On 02/12/2011 05:42 AM, Andrew Beverley wrote:
>> On Fri, 2011-02-11 at 15:24 -0600, Chris Friesen wrote:
>>> It has been proposed to add a custom patch to clear the mark for all
>>> marked connections--is there a better way of doing this?
>>>
>>
>> How about using the conntrack userspace command-line interface?
>>
>> conntrack -U -m 0
>>
>> looks like it should do what you want.
>
>Oooh, that looks perfect.
>
>For some reason the online man pages for conntrack don't seem to be up
>to date...they don't have the -U option. It's listed in the package
>itself though.
What online manpages are you referring to?
^ permalink raw reply
* Re: [PATCH] ipvs: fix timer in get_curr_sync_buff
From: Simon Horman @ 2011-02-15 22:33 UTC (permalink / raw)
To: Julian Anastasov; +Cc: lvs-devel, netdev, Tinggong Wang
In-Reply-To: <alpine.LFD.2.00.1102090214460.3494@ja.ssi.bg>
On Wed, Feb 09, 2011 at 02:21:59AM +0200, Julian Anastasov wrote:
>
> From: Tinggong Wang <wangtinggong@gmail.com>
>
> Fix get_curr_sync_buff to keep buffer for 2 seconds
> as intended, not just for the current jiffie. By this way
> we will sync more connection structures with single packet.
>
> Signed-off-by: Tinggong Wang <wangtinggong@gmail.com>
> Signed-off-by: Julian Anastasov <ja@ssi.bg>
> ---
>
> Patch is against lvs-test-2.6 from 09-FEB-2011
> but should apply cleanly to net-next.
Thanks, I have pushed this change into the master branch of lvs-test-2.6.
I am waiting for one or two more patches to settle down before
sending my next pull request to Patrick.
^ permalink raw reply
* Re: [PATCH] ipvs: remove extra lookups for ICMP packets
From: Simon Horman @ 2011-02-15 22:32 UTC (permalink / raw)
To: Julian Anastasov; +Cc: lvs-devel, netdev
In-Reply-To: <alpine.LFD.2.00.1102090222020.3494@ja.ssi.bg>
On Wed, Feb 09, 2011 at 02:26:38AM +0200, Julian Anastasov wrote:
>
> Remove code that should not be called anymore.
> Now when ip_vs_out handles replies for local clients at
> LOCAL_IN hook we do not need to call conn_out_get and
> handle_response_icmp from ip_vs_in_icmp* because such
> lookups were already performed for the ICMP packet and no
> connection was found.
>
> Signed-off-by: Julian Anastasov <ja@ssi.bg>
> ---
>
> Patch is against lvs-test-2.6 from 09-FEB-2011
> but should apply cleanly to net-next.
Thanks, I have pushed this change into the master branch of lvs-test-2.6.
I am waiting for one or two more patches to settle down before
sending my next pull request to Patrick.
^ permalink raw reply
* Re: [Bugme-new] [Bug 27212] New: Warning kmemcheck: Caught 64-bit read from uninitialized memory in netlink_broadcast_filtered
From: David Rientjes @ 2011-02-15 22:21 UTC (permalink / raw)
To: Pekka Enberg
Cc: Eric Dumazet, Andrew Morton, netdev, bugzilla-daemon,
bugme-daemon, casteyde.christian, Changli Gao, Vegard Nossum,
David Miller, linux-kernel, Christoph Lameter
In-Reply-To: <AANLkTinKTdeO5mnE38n==aZMz5P1XKKM+kykhoBu0=9k@mail.gmail.com>
On Tue, 15 Feb 2011, Pekka Enberg wrote:
> > [PATCH] slub: fix kmemcheck calls to match ksize() hints
> >
> > Recent use of ksize() in network stack (commit ca44ac38 : net: don't
> > reallocate skb->head unless the current one hasn't the needed extra size
> > or is shared) triggers kmemcheck warnings, because ksize() can return
> > more space than kmemcheck is aware of.
> >
> > Pekka Enberg noticed SLAB+kmemcheck is doing the right thing, while SLUB
> > +kmemcheck doesnt.
> >
> > Bugzilla reference #27212
> >
> > Reported-by: Christian Casteyde <casteyde.christian@free.fr>
> > Suggested-by: Pekka Enberg <penberg@kernel.org>
> > Signed-off-by: Eric Dumazet <eric.dumazet@gmail.com>
> > CC: David Miller <davem@davemloft.net>
> > CC: Changli Gao <xiaosuo@gmail.com>
> > CC: Andrew Morton <akpm@linux-foundation.org>
Acked-by: David Rientjes <rientjes@google.com>
^ permalink raw reply
* Re: any way to reset all marked connections when using CONNMARK?
From: Chris Friesen @ 2011-02-15 22:06 UTC (permalink / raw)
To: Andrew Beverley; +Cc: netdev, netfilter-devel, netfilter
In-Reply-To: <1297510935.1770.38.camel@andybev>
On 02/12/2011 05:42 AM, Andrew Beverley wrote:
> On Fri, 2011-02-11 at 15:24 -0600, Chris Friesen wrote:
>> It has been proposed to add a custom patch to clear the mark for all
>> marked connections--is there a better way of doing this?
>>
>
> How about using the conntrack userspace command-line interface?
>
> conntrack -U -m 0
>
> looks like it should do what you want.
Oooh, that looks perfect.
For some reason the online man pages for conntrack don't seem to be up
to date...they don't have the -U option. It's listed in the package
itself though.
Thanks,
Chris
--
Chris Friesen
Software Developer
GENBAND
chris.friesen@genband.com
www.genband.com
^ permalink raw reply
* Re: [PATCH v5 RESEND 2/9] ethtool: enable GSO and GRO by default
From: David Miller @ 2011-02-15 22:00 UTC (permalink / raw)
To: mirq-linux; +Cc: netdev, bhutchings
In-Reply-To: <20110215214649.GA23866@rere.qmqm.pl>
From: Michał Mirosław <mirq-linux@rere.qmqm.pl>
Date: Tue, 15 Feb 2011 22:46:49 +0100
> On Sun, Feb 13, 2011 at 10:50:23AM -0800, David Miller wrote:
>> From: Michał Mirosław <mirq-linux@rere.qmqm.pl>
>> Date: Sun, 13 Feb 2011 12:11:45 +0100 (CET)
>> > Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
>> This is not appropriate.
>>
>> Now, every driver that lacks SG support will spit out that warning
>> message in netdev_fix_features().
>>
>> That's why the check is there conditionalizing NETIF_F_GSO on
>> NETIF_F_SG in register_netdevice().
>
> I think all those messages should be converted to DEBUG level. Those
> conditions are constant and can be better described in
> Documentation/networking/ or ethtool manpage. Preferably along the
> device-specific conditions when implemented.
>
> Or I could just drop the message for the GSO case as it's something new here
> anyway (I added it to make it consistent with handling of other features).
The messages exist to let driver authors know they've constructed an
illegal set of feature bits.
Since you're now adding the GSO bit yourself, you should perform
due diligence and prevent the illegal combination yourself.
This has no other impact on the other messages and cases, which
definitely should stay intact.
Your change is just wrong and knowingly introduces useless log
messages, please just fix it up.
^ permalink raw reply
* Re: [PATCH v5 RESEND 2/9] ethtool: enable GSO and GRO by default
From: Michał Mirosław @ 2011-02-15 21:46 UTC (permalink / raw)
To: David Miller; +Cc: netdev, bhutchings
In-Reply-To: <20110213.105023.68138742.davem@davemloft.net>
On Sun, Feb 13, 2011 at 10:50:23AM -0800, David Miller wrote:
> From: Michał Mirosław <mirq-linux@rere.qmqm.pl>
> Date: Sun, 13 Feb 2011 12:11:45 +0100 (CET)
> > Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
> This is not appropriate.
>
> Now, every driver that lacks SG support will spit out that warning
> message in netdev_fix_features().
>
> That's why the check is there conditionalizing NETIF_F_GSO on
> NETIF_F_SG in register_netdevice().
I think all those messages should be converted to DEBUG level. Those
conditions are constant and can be better described in
Documentation/networking/ or ethtool manpage. Preferably along the
device-specific conditions when implemented.
Or I could just drop the message for the GSO case as it's something new here
anyway (I added it to make it consistent with handling of other features).
Best Regards,
Michał Mirosław
^ permalink raw reply
* Re: [PATCH v2] xen network backend driver
From: Konrad Rzeszutek Wilk @ 2011-02-15 21:35 UTC (permalink / raw)
To: Ian Campbell
Cc: netdev@vger.kernel.org, xen-devel, Jeremy Fitzhardinge,
Ben Hutchings, Herbert Xu
In-Reply-To: <1297160635.9149.21.camel@zakaz.uk.xensource.com>
> Signed-off-by: Ian Campbell <ian.campbell@citrix.com>
Hey Ian,
I took a look at and provided some input. I got lost with the
GSO, credit code, fragments, and the host of the other features
that can get negotiated. Will need to re-educate myself on the
networking code some more.
Sure changed a lot since 2.6.18..
Would it make sense to split the review in the netback and netfront
in two different patchsets (you might need to overlap the headers
that define the operations .. which is OK)?
>
> diff --git a/drivers/net/Kconfig b/drivers/net/Kconfig
> index cbf0635..1c77e18 100644
> --- a/drivers/net/Kconfig
> +++ b/drivers/net/Kconfig
> @@ -2963,12 +2963,38 @@ config XEN_NETDEV_FRONTEND
> select XEN_XENBUS_FRONTEND
> default y
> help
> - The network device frontend driver allows the kernel to
> - access network devices exported exported by a virtual
> - machine containing a physical network device driver. The
> - frontend driver is intended for unprivileged guest domains;
> - if you are compiling a kernel for a Xen guest, you almost
> - certainly want to enable this.
> + This driver provides support for Xen paravirtual network
> + devices exported by a Xen network driver domain (often
> + domain 0).
> +
> + The corresponding Linux backend driver is enabled by the
> + CONFIG_XEN_NETDEV_BACKEND option.
> +
> + If you are compiling a kernel for use as Xen guest, you
> + should say Y here. To compile this driver as a module, chose
> + M here: the module will be called xen-netfront.
> +
> +config XEN_NETDEV_BACKEND
> + tristate "Xen backend network device"
> + depends on XEN_BACKEND
> + help
> + This driver allows the kernel to act as a Xen network driver
> + domain which exports paravirtual network devices to other
> + Xen domains. These devices can be accessed by any operating
> + system that implements a compatible front end.
> +
> + The corresponding Linux frontend driver is enabled by the
> + CONFIG_XEN_NETDEV_FRONTEND configuration option.
> +
> + The backend driver presents a standard network device
> + endpoint for each paravirtual network device to the driver
> + domain network stack. These can then be bridged or routed
> + etc in order to provide full network connectivity.
> +
> + If you are compiling a kernel to run in a Xen network driver
> + domain (often this is domain 0) you should say Y here. To
> + compile this driver as a module, chose M here: the module
> + will be called xen-netback.
>
> config ISERIES_VETH
> tristate "iSeries Virtual Ethernet driver support"
> diff --git a/drivers/net/Makefile b/drivers/net/Makefile
> index b90738d..145dfd7 100644
> --- a/drivers/net/Makefile
> +++ b/drivers/net/Makefile
> @@ -171,6 +171,7 @@ obj-$(CONFIG_SLIP) += slip.o
> obj-$(CONFIG_SLHC) += slhc.o
>
> obj-$(CONFIG_XEN_NETDEV_FRONTEND) += xen-netfront.o
> +obj-$(CONFIG_XEN_NETDEV_BACKEND) += xen-netback/
>
> obj-$(CONFIG_DUMMY) += dummy.o
> obj-$(CONFIG_IFB) += ifb.o
> diff --git a/drivers/net/xen-netback/Makefile b/drivers/net/xen-netback/Makefile
> new file mode 100644
> index 0000000..e346e81
> --- /dev/null
> +++ b/drivers/net/xen-netback/Makefile
> @@ -0,0 +1,3 @@
> +obj-$(CONFIG_XEN_NETDEV_BACKEND) := xen-netback.o
> +
> +xen-netback-y := netback.o xenbus.o interface.o
> diff --git a/drivers/net/xen-netback/common.h b/drivers/net/xen-netback/common.h
> new file mode 100644
> index 0000000..03196ab
> --- /dev/null
> +++ b/drivers/net/xen-netback/common.h
> @@ -0,0 +1,147 @@
> +/*
> + * 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 Software 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.
> + */
> +
> +#ifndef __XEN_NETBACK__COMMON_H__
> +#define __XEN_NETBACK__COMMON_H__
> +
> +#define pr_fmt(fmt) KBUILD_MODNAME ":%s: " fmt, __func__
> +
> +#include <linux/module.h>
> +#include <linux/interrupt.h>
> +#include <linux/slab.h>
> +#include <linux/ip.h>
> +#include <linux/in.h>
> +#include <linux/io.h>
> +#include <linux/netdevice.h>
> +#include <linux/etherdevice.h>
> +#include <linux/wait.h>
> +#include <linux/sched.h>
> +
> +#include <xen/interface/io/netif.h>
> +#include <asm/pgalloc.h>
I don't think you need that file. Yeah, tested and it
compiles fine.
> +#include <xen/interface/grant_table.h>
> +#include <xen/grant_table.h>
> +#include <xen/xenbus.h>
> +
> +struct xen_netbk;
> +
> +struct xenvif {
> + /* Unique identifier for this interface. */
> + domid_t domid;
> + unsigned int handle;
> +
> + /* */
Looks like there was a comment there, but it went away?
> + struct xen_netbk *netbk;
> +
> + u8 fe_dev_addr[6];
> +
> + /* Physical parameters of the comms window. */
> + grant_handle_t tx_shmem_handle;
> + grant_ref_t tx_shmem_ref;
> + grant_handle_t rx_shmem_handle;
> + grant_ref_t rx_shmem_ref;
> + unsigned int irq;
> +
> + /* The shared rings and indexes. */
> + struct xen_netif_tx_back_ring tx;
> + struct xen_netif_rx_back_ring rx;
> + struct vm_struct *tx_comms_area;
> + struct vm_struct *rx_comms_area;
> +
> + /* Flags that must not be set in dev->features */
> + int features_disabled;
> +
> + /* Frontend feature information. */
> + u8 can_sg:1;
> + u8 gso:1;
> + u8 gso_prefix:1;
> + u8 csum:1;
> +
> + /* Internal feature information. */
> + u8 can_queue:1; /* can queue packets for receiver? */
> +
> + /* Allow xenvif_start_xmit() to peek ahead in the rx request
> + * ring. This is a prediction of what rx_req_cons will be once
> + * all queued skbs are put on the ring. */
> + RING_IDX rx_req_cons_peek;
> +
> + /* Transmit shaping: allow 'credit_bytes' every 'credit_usec'. */
> + unsigned long credit_bytes;
> + unsigned long credit_usec;
> + unsigned long remaining_credit;
> + struct timer_list credit_timeout;
> +
> + /* Statistics */
> + int rx_gso_checksum_fixup;
> +
> + /* Miscellaneous private stuff. */
> + struct list_head list; /* scheduling list */
> + atomic_t refcnt;
> + struct net_device *dev;
> + struct net_device_stats stats;
> +
> + unsigned int carrier;
> +
> + wait_queue_head_t waiting_to_free;
> +};
> +
> +#define XEN_NETIF_TX_RING_SIZE __RING_SIZE((struct xen_netif_tx_sring *)0, PAGE_SIZE)
> +#define XEN_NETIF_RX_RING_SIZE __RING_SIZE((struct xen_netif_rx_sring *)0, PAGE_SIZE)
> +
> +struct xenvif *xenvif_alloc(struct device *parent,
> + domid_t domid,
> + unsigned int handle);
> +
> +int xenvif_connect(struct xenvif *vif, unsigned long tx_ring_ref,
> + unsigned long rx_ring_ref, unsigned int evtchn);
> +void xenvif_disconnect(struct xenvif *vif);
> +
> +void xenvif_get(struct xenvif *vif);
> +void xenvif_put(struct xenvif *vif);
> +
> +int xenvif_xenbus_init(void);
> +
> +int xenvif_schedulable(struct xenvif *vif);
> +
> +void xenvif_schedule_work(struct xenvif *vif);
> +
> +int xenvif_queue_full(struct xenvif *vif);
> +
> +/* (De)Register a xenvif with the netback backend. */
> +void xen_netbk_add_xenvif(struct xenvif *vif);
> +void xen_netbk_remove_xenvif(struct xenvif *vif);
> +
> +/* */
> +void xen_netbk_schedule_xenvif(struct xenvif *vif);
> +void xen_netbk_deschedule_xenfif(struct xenvif *vif);
> +
> +/* */
> +unsigned int xen_netbk_count_skb_slots(struct xenvif *vif, struct sk_buff *skb);
> +
> +/* */
> +void xen_netbk_queue_tx_skb(struct xenvif *vif, struct sk_buff *skb);
> +
> +#endif /* __XEN_NETBACK__COMMON_H__ */
> diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
> new file mode 100644
> index 0000000..98a992d
> --- /dev/null
> +++ b/drivers/net/xen-netback/interface.c
> @@ -0,0 +1,550 @@
> +/*
> + * Network-device interface management.
> + *
> + * Copyright (c) 2004-2005, Keir Fraser
> + *
> + * 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 Software 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.
> + */
> +
> +#include "common.h"
> +
> +#include <linux/ethtool.h>
> +#include <linux/rtnetlink.h>
> +
> +#include <xen/events.h>
> +#include <asm/xen/hypercall.h>
> +
> +#define XENVIF_QUEUE_LENGTH 32
> +
> +void xenvif_get(struct xenvif *vif)
> +{
> + atomic_inc(&vif->refcnt);
> +}
> +
> +void xenvif_put(struct xenvif *vif)
> +{
> + if (atomic_dec_and_test(&vif->refcnt))
> + wake_up(&vif->waiting_to_free);
> +}
> +
> +static int xenvif_max_required_rx_slots(struct xenvif *vif)
> +{
> + int max = DIV_ROUND_UP(vif->dev->mtu, PAGE_SIZE);
> +
> + if (vif->can_sg || vif->gso || vif->gso_prefix)
> + max += MAX_SKB_FRAGS + 1; /* extra_info + frags */
> +
> + return max;
> +}
> +
> +int xenvif_queue_full(struct xenvif *vif)
> +{
> + RING_IDX peek = vif->rx_req_cons_peek;
> + RING_IDX needed = xenvif_max_required_rx_slots(vif);
> +
> + return ((vif->rx.sring->req_prod - peek) < needed) ||
> + ((vif->rx.rsp_prod_pvt + XEN_NETIF_RX_RING_SIZE - peek) < needed);
> +}
> +
> +/*
> + * Implement our own carrier flag: the network stack's version causes delays
> + * when the carrier is re-enabled (in particular, dev_activate() may not
> + * immediately be called, which can cause packet loss; also the etherbridge
> + * can be rather lazy in activating its port).
> + */
> +static void xenvif_carrier_on(struct xenvif *vif)
> +{
> + vif->carrier = 1;
> +}
> +static void xenvif_carrier_off(struct xenvif *vif)
> +{
> + vif->carrier = 0;
> +}
> +static int xenvif_carrier_ok(struct xenvif *vif)
> +{
> + return vif->carrier;
> +}
> +
> +int xenvif_schedulable(struct xenvif *vif)
> +{
> + return netif_running(vif->dev) && xenvif_carrier_ok(vif);
> +}
> +
> +static irqreturn_t xenvif_interrupt(int irq, void *dev_id)
> +{
> + struct xenvif *vif = dev_id;
> +
> + if (vif->netbk == NULL)
> + return IRQ_NONE;
> +
> + xen_netbk_schedule_xenvif(vif);
> +
> + if (xenvif_schedulable(vif) && !xenvif_queue_full(vif))
> + netif_wake_queue(vif->dev);
> +
> + return IRQ_HANDLED;
> +}
> +
> +static int xenvif_start_xmit(struct sk_buff *skb, struct net_device *dev)
> +{
> + struct xenvif *vif = netdev_priv(dev);
> +
> + BUG_ON(skb->dev != dev);
> +
> + if (vif->netbk == NULL)
> + goto drop;
> +
> + /* Drop the packet if the target domain has no receive buffers. */
> + if (unlikely(!xenvif_schedulable(vif) || xenvif_queue_full(vif)))
> + goto drop;
> +
> + /* Reserve ring slots for the worst-case number of fragments. */
> + vif->rx_req_cons_peek += xen_netbk_count_skb_slots(vif, skb);
> + xenvif_get(vif);
> +
> + if (vif->can_queue && xenvif_queue_full(vif)) {
> + vif->rx.sring->req_event = vif->rx_req_cons_peek +
> + xenvif_max_required_rx_slots(vif);
> + mb(); /* request notification /then/ check & stop the queue */
> + if (xenvif_queue_full(vif))
> + netif_stop_queue(dev);
> + }
> +
> + xen_netbk_queue_tx_skb(vif, skb);
> +
> + return 0;
> +
> + drop:
> + vif->stats.tx_dropped++;
> + dev_kfree_skb(skb);
> + return 0;
> +}
> +
> +static struct net_device_stats *xenvif_get_stats(struct net_device *dev)
> +{
> + struct xenvif *vif = netdev_priv(dev);
> + return &vif->stats;
> +}
> +
> +void xenvif_schedule_work(struct xenvif *vif)
> +{
> + int more_to_do;
> +
> + RING_FINAL_CHECK_FOR_REQUESTS(&vif->tx, more_to_do);
> +
> + if (more_to_do)
> + xen_netbk_schedule_xenvif(vif);
> +}
> +
> +
> +static void xenvif_up(struct xenvif *vif)
> +{
> + xen_netbk_add_xenvif(vif);
> + enable_irq(vif->irq);
> + xenvif_schedule_work(vif);
> +}
> +
> +static void xenvif_down(struct xenvif *vif)
> +{
> + disable_irq(vif->irq);
> + xen_netbk_deschedule_xenfif(vif);
> + xen_netbk_remove_xenvif(vif);
> +}
> +
> +static int xenvif_open(struct net_device *dev)
> +{
> + struct xenvif *vif = netdev_priv(dev);
> + if (xenvif_carrier_ok(vif)) {
> + xenvif_up(vif);
> + netif_start_queue(dev);
> + }
> + return 0;
> +}
> +
> +static int xenvif_close(struct net_device *dev)
> +{
> + struct xenvif *vif = netdev_priv(dev);
> + if (xenvif_carrier_ok(vif))
> + xenvif_down(vif);
> + netif_stop_queue(dev);
> + return 0;
> +}
> +
> +static int xenvif_change_mtu(struct net_device *dev, int mtu)
> +{
> + struct xenvif *vif = netdev_priv(dev);
> + int max = vif->can_sg ? 65535 - ETH_HLEN : ETH_DATA_LEN;
> +
> + if (mtu > max)
> + return -EINVAL;
> + dev->mtu = mtu;
> + return 0;
> +}
> +
> +static void xenvif_set_features(struct xenvif *vif)
> +{
> + struct net_device *dev = vif->dev;
> + int features = dev->features;
> +
> + if (vif->can_sg)
> + features |= NETIF_F_SG;
> + if (vif->gso || vif->gso_prefix)
> + features |= NETIF_F_TSO;
> + if (vif->csum)
> + features |= NETIF_F_IP_CSUM;
> +
> + features &= ~(vif->features_disabled);
> +
> + if (!(features & NETIF_F_SG) && dev->mtu > ETH_DATA_LEN)
> + dev->mtu = ETH_DATA_LEN;
> +
> + dev->features = features;
> +}
> +
> +static int xenvif_set_tx_csum(struct net_device *dev, u32 data)
> +{
> + struct xenvif *vif = netdev_priv(dev);
> + if (data) {
> + if (!vif->csum)
> + return -EOPNOTSUPP;
> + vif->features_disabled &= ~NETIF_F_IP_CSUM;
> + } else {
> + vif->features_disabled |= NETIF_F_IP_CSUM;
> + }
> +
> + xenvif_set_features(vif);
> + return 0;
> +}
> +
> +static int xenvif_set_sg(struct net_device *dev, u32 data)
> +{
> + struct xenvif *vif = netdev_priv(dev);
> + if (data) {
> + if (!vif->can_sg)
> + return -EOPNOTSUPP;
> + vif->features_disabled &= ~NETIF_F_SG;
> + } else {
> + vif->features_disabled |= NETIF_F_SG;
> + }
> +
> + xenvif_set_features(vif);
> + return 0;
> +}
> +
> +static int xenvif_set_tso(struct net_device *dev, u32 data)
> +{
> + struct xenvif *vif = netdev_priv(dev);
> + if (data) {
> + if (!vif->gso && !vif->gso_prefix)
> + return -EOPNOTSUPP;
> + vif->features_disabled &= ~NETIF_F_TSO;
> + } else {
> + vif->features_disabled |= NETIF_F_TSO;
> + }
> +
> + xenvif_set_features(vif);
> + return 0;
> +}
> +
> +static const struct xenvif_stat {
> + char name[ETH_GSTRING_LEN];
> + u16 offset;
> +} xenvif_stats[] = {
> + {
> + "rx_gso_checksum_fixup",
> + offsetof(struct xenvif, rx_gso_checksum_fixup)
> + },
> +};
> +
> +static int xenvif_get_sset_count(struct net_device *dev, int string_set)
> +{
> + switch (string_set) {
> + case ETH_SS_STATS:
> + return ARRAY_SIZE(xenvif_stats);
> + default:
> + return -EINVAL;
> + }
> +}
> +
> +static void xenvif_get_ethtool_stats(struct net_device *dev,
> + struct ethtool_stats *stats, u64 * data)
> +{
> + void *vif = netdev_priv(dev);
> + int i;
> +
> + for (i = 0; i < ARRAY_SIZE(xenvif_stats); i++)
> + data[i] = *(int *)(vif + xenvif_stats[i].offset);
> +}
> +
> +static void xenvif_get_strings(struct net_device *dev, u32 stringset, u8 * data)
> +{
> + int i;
> +
> + switch (stringset) {
> + case ETH_SS_STATS:
> + for (i = 0; i < ARRAY_SIZE(xenvif_stats); i++)
> + memcpy(data + i * ETH_GSTRING_LEN,
> + xenvif_stats[i].name, ETH_GSTRING_LEN);
> + break;
> + }
> +}
> +
> +static struct ethtool_ops xenvif_ethtool_ops = {
> + .get_tx_csum = ethtool_op_get_tx_csum,
> + .set_tx_csum = xenvif_set_tx_csum,
> + .get_sg = ethtool_op_get_sg,
> + .set_sg = xenvif_set_sg,
> + .get_tso = ethtool_op_get_tso,
> + .set_tso = xenvif_set_tso,
> + .get_link = ethtool_op_get_link,
> +
> + .get_sset_count = xenvif_get_sset_count,
> + .get_ethtool_stats = xenvif_get_ethtool_stats,
> + .get_strings = xenvif_get_strings,
> +};
> +
> +static struct net_device_ops xenvif_netdev_ops = {
> + .ndo_start_xmit = xenvif_start_xmit,
> + .ndo_get_stats = xenvif_get_stats,
> + .ndo_open = xenvif_open,
> + .ndo_stop = xenvif_close,
> + .ndo_change_mtu = xenvif_change_mtu,
> +};
> +
> +struct xenvif *xenvif_alloc(struct device *parent, domid_t domid,
> + unsigned int handle)
> +{
> + int err = 0;
> + struct net_device *dev;
> + struct xenvif *vif;
> + char name[IFNAMSIZ] = {};
> +
> + snprintf(name, IFNAMSIZ - 1, "vif%u.%u", domid, handle);
> + dev = alloc_netdev(sizeof(struct xenvif), name, ether_setup);
> + if (dev == NULL) {
> + pr_debug("Could not allocate netdev\n");
pr_warn?
> + return ERR_PTR(-ENOMEM);
> + }
> +
> + SET_NETDEV_DEV(dev, parent);
> +
> + vif = netdev_priv(dev);
> + memset(vif, 0, sizeof(*vif));
> + vif->domid = domid;
> + vif->handle = handle;
> + vif->netbk = NULL;
> + vif->can_sg = 1;
> + vif->csum = 1;
> + atomic_set(&vif->refcnt, 1);
> + init_waitqueue_head(&vif->waiting_to_free);
> + vif->dev = dev;
> + INIT_LIST_HEAD(&vif->list);
> +
> + xenvif_carrier_off(vif);
> +
> + vif->credit_bytes = vif->remaining_credit = ~0UL;
> + vif->credit_usec = 0UL;
> + init_timer(&vif->credit_timeout);
> + /* Initialize 'expires' now: it's used to track the credit window. */
> + vif->credit_timeout.expires = jiffies;
> +
> + dev->netdev_ops = &xenvif_netdev_ops;
> + xenvif_set_features(vif);
> + SET_ETHTOOL_OPS(dev, &xenvif_ethtool_ops);
> +
> + dev->tx_queue_len = XENVIF_QUEUE_LENGTH;
> +
> + /*
> + * Initialise a dummy MAC address. We choose the numerically
> + * largest non-broadcast address to prevent the address getting
> + * stolen by an Ethernet bridge for STP purposes.
> + * (FE:FF:FF:FF:FF:FF)
> + */
> + memset(dev->dev_addr, 0xFF, ETH_ALEN);
> + dev->dev_addr[0] &= ~0x01;
> +
> + rtnl_lock();
> + err = register_netdevice(dev);
> + rtnl_unlock();
> + if (err) {
> + pr_debug("Could not register new net device %s: err=%d\n",
> + dev->name, err);
pr_warn?
> + free_netdev(dev);
> + return ERR_PTR(err);
> + }
> +
> + pr_debug("Successfully created xenvif\n");
> + return vif;
> +}
> +
> +static int map_frontend_pages(struct xenvif *vif,
> + grant_ref_t tx_ring_ref,
> + grant_ref_t rx_ring_ref)
> +{
> + struct gnttab_map_grant_ref op;
> +
> + gnttab_set_map_op(&op, (unsigned long)vif->tx_comms_area->addr,
> + GNTMAP_host_map, tx_ring_ref, vif->domid);
> +
> + if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
> + BUG();
How about something less severe? Say return the error code?
> +
> + if (op.status) {
> + pr_debug("Gnttab failure mapping tx_ring_ref!\n");
pr_warn.
> + return op.status;
> + }
> +
> + vif->tx_shmem_ref = tx_ring_ref;
> + vif->tx_shmem_handle = op.handle;
> +
> + gnttab_set_map_op(&op, (unsigned long)vif->rx_comms_area->addr,
> + GNTMAP_host_map, rx_ring_ref, vif->domid);
> +
> + if (HYPERVISOR_grant_table_op(GNTTABOP_map_grant_ref, &op, 1))
> + BUG();
Ditto.. or perhaps tie it in with the check below.
> +
> + if (op.status) {
> + struct gnttab_unmap_grant_ref unop;
> +
> + gnttab_set_unmap_op(&unop,
> + (unsigned long)vif->tx_comms_area->addr,
> + GNTMAP_host_map, vif->tx_shmem_handle);
> + HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &unop, 1);
> + pr_debug("Gnttab failure mapping rx_ring_ref!\n");
pr_warn I think.
> + return op.status;
> + }
> +
> + vif->rx_shmem_ref = rx_ring_ref;
> + vif->rx_shmem_handle = op.handle;
> +
> + return 0;
> +}
> +
> +static void unmap_frontend_pages(struct xenvif *vif)
> +{
> + struct gnttab_unmap_grant_ref op;
> +
> + gnttab_set_unmap_op(&op, (unsigned long)vif->tx_comms_area->addr,
> + GNTMAP_host_map, vif->tx_shmem_handle);
> +
> + if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
> + BUG();
Well, we could ignore it and try
> +
> + gnttab_set_unmap_op(&op, (unsigned long)vif->rx_comms_area->addr,
> + GNTMAP_host_map, vif->rx_shmem_handle);
> +
> + if (HYPERVISOR_grant_table_op(GNTTABOP_unmap_grant_ref, &op, 1))
to do this and _then_ later report failure in doing it?
> + BUG();
> +}
> +
> +int xenvif_connect(struct xenvif *vif, unsigned long tx_ring_ref,
> + unsigned long rx_ring_ref, unsigned int evtchn)
> +{
> + int err = -ENOMEM;
> + struct xen_netif_tx_sring *txs;
> + struct xen_netif_rx_sring *rxs;
> +
> + /* Already connected through? */
> + if (vif->irq)
> + return 0;
> +
> + xenvif_set_features(vif);
> +
> + vif->tx_comms_area = alloc_vm_area(PAGE_SIZE);
> + if (vif->tx_comms_area == NULL)
> + return -ENOMEM;
> + vif->rx_comms_area = alloc_vm_area(PAGE_SIZE);
> + if (vif->rx_comms_area == NULL)
> + goto err_rx;
> +
> + err = map_frontend_pages(vif, tx_ring_ref, rx_ring_ref);
> + if (err)
> + goto err_map;
> +
> + err = bind_interdomain_evtchn_to_irqhandler(
> + vif->domid, evtchn, xenvif_interrupt, 0,
> + vif->dev->name, vif);
> + if (err < 0)
> + goto err_hypervisor;
> + vif->irq = err;
> + disable_irq(vif->irq);
> +
> + txs = (struct xen_netif_tx_sring *)vif->tx_comms_area->addr;
> + BACK_RING_INIT(&vif->tx, txs, PAGE_SIZE);
> +
> + rxs = (struct xen_netif_rx_sring *)
> + ((char *)vif->rx_comms_area->addr);
> + BACK_RING_INIT(&vif->rx, rxs, PAGE_SIZE);
> +
> + vif->rx_req_cons_peek = 0;
> +
> + xenvif_get(vif);
> +
> + rtnl_lock();
> + xenvif_carrier_on(vif);
> + if (netif_running(vif->dev))
> + xenvif_up(vif);
> + rtnl_unlock();
> +
> + return 0;
> +err_hypervisor:
> + unmap_frontend_pages(vif);
> +err_map:
> + free_vm_area(vif->rx_comms_area);
> +err_rx:
> + free_vm_area(vif->tx_comms_area);
> + return err;
> +}
> +
> +void xenvif_disconnect(struct xenvif *vif)
> +{
> + if (xenvif_carrier_ok(vif)) {
> + rtnl_lock();
> + xenvif_carrier_off(vif);
> + netif_carrier_off(vif->dev); /* discard queued packets */
> + if (netif_running(vif->dev))
> + xenvif_down(vif);
> + rtnl_unlock();
> + xenvif_put(vif);
> + }
> +
> + atomic_dec(&vif->refcnt);
> + wait_event(vif->waiting_to_free, atomic_read(&vif->refcnt) == 0);
> +
> + del_timer_sync(&vif->credit_timeout);
> +
> + if (vif->irq)
> + unbind_from_irqhandler(vif->irq, vif);
> +
> + unregister_netdev(vif->dev);
> +
> + if (vif->tx.sring) {
> + unmap_frontend_pages(vif);
> + free_vm_area(vif->tx_comms_area);
> + free_vm_area(vif->rx_comms_area);
> + }
> +
> + free_netdev(vif->dev);
> +}
> diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c
> new file mode 100644
> index 0000000..fbddf3d
> --- /dev/null
> +++ b/drivers/net/xen-netback/netback.c
> @@ -0,0 +1,1618 @@
> +/*
> + * Back-end of the driver for virtual network devices. This portion of the
> + * driver exports a 'unified' network-device interface that can be accessed
> + * by any operating system that implements a compatible front end. A
> + * reference front-end implementation can be found in:
> + * drivers/net/xen-netfront.c
> + *
> + * Copyright (c) 2002-2005, K A Fraser
> + *
> + * 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 Software 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.
> + */
> +
> +#include "common.h"
> +
> +#include <linux/kthread.h>
> +#include <linux/if_vlan.h>
> +#include <linux/udp.h>
> +
> +#include <net/tcp.h>
> +
> +#include <xen/events.h>
> +#include <xen/interface/memory.h>
> +
> +#include <asm/xen/hypercall.h>
> +#include <asm/xen/page.h>
> +
> +struct pending_tx_info {
> + struct xen_netif_tx_request req;
> + struct xenvif *vif;
> +};
> +typedef unsigned int pending_ring_idx_t;
> +
> +struct netbk_rx_meta {
> + int id;
> + int size;
> + int gso_size;
> +};
> +
> +#define MAX_PENDING_REQS 256
> +
> +#define MAX_BUFFER_OFFSET PAGE_SIZE
Why not use PAGE_SIZE instead of MAX_BUFFER_OFFSET?
> +
> +/* extra field used in struct page */
> +union page_ext {
> + struct {
> +#if BITS_PER_LONG < 64
> +#define IDX_WIDTH 8
> +#define GROUP_WIDTH (BITS_PER_LONG - IDX_WIDTH)
> + unsigned int group:GROUP_WIDTH;
> + unsigned int idx:IDX_WIDTH;
> +#else
> + unsigned int group, idx;
> +#endif
> + } e;
> + void *mapping;
> +};
> +
> +struct xen_netbk {
> + wait_queue_head_t wq;
> + struct task_struct *task;
> +
> + struct sk_buff_head rx_queue;
> + struct sk_buff_head tx_queue;
> +
> + struct timer_list net_timer;
> +
> + struct page *mmap_pages[MAX_PENDING_REQS];
> +
> + pending_ring_idx_t pending_prod;
> + pending_ring_idx_t pending_cons;
> + struct list_head net_schedule_list;
> +
> + /* Protect the net_schedule_list in netif. */
> + spinlock_t net_schedule_list_lock;
> +
> + atomic_t netfront_count;
> +
> + struct pending_tx_info pending_tx_info[MAX_PENDING_REQS];
> + struct gnttab_copy tx_copy_ops[MAX_PENDING_REQS];
> +
> + u16 pending_ring[MAX_PENDING_REQS];
> +
> + /*
> + * Each head or fragment can be up to 4096 bytes. Given
> + * MAX_BUFFER_OFFSET of 4096 the worst case is that each
> + * head/fragment uses 2 copy operation.
For an MTU of 9000 won't we have two fragments and one head?
> + */
> + struct gnttab_copy grant_copy_op[2*XEN_NETIF_RX_RING_SIZE];
> + unsigned char rx_notify[NR_IRQS];
So a 2KB array on which we poke a value most of the time (if not all)
past the nr_irq_gsi.. Is there a better way of doing this?
> + u16 notify_list[XEN_NETIF_RX_RING_SIZE];
> + struct netbk_rx_meta meta[2*XEN_NETIF_RX_RING_SIZE];
> +};
> +
> +static struct xen_netbk *xen_netbk;
> +static int xen_netbk_group_nr;
> +
> +void xen_netbk_add_xenvif(struct xenvif *vif)
> +{
> + int i;
> + int min_netfront_count;
> + int min_group = 0;
> + struct xen_netbk *netbk;
> +
> + min_netfront_count = atomic_read(&xen_netbk[0].netfront_count);
> + for (i = 0; i < xen_netbk_group_nr; i++) {
> + int netfront_count = atomic_read(&xen_netbk[i].netfront_count);
> + if (netfront_count < min_netfront_count) {
> + min_group = i;
> + min_netfront_count = netfront_count;
> + }
> + }
> +
> + netbk = &xen_netbk[min_group];
> +
> + vif->netbk = netbk;
> + atomic_inc(&netbk->netfront_count);
> +}
> +
> +void xen_netbk_remove_xenvif(struct xenvif *vif)
> +{
> + struct xen_netbk *netbk = vif->netbk;
> + vif->netbk = NULL;
> + atomic_dec(&netbk->netfront_count);
> +}
> +
> +static void xen_netbk_idx_release(struct xen_netbk *netbk, u16 pending_idx);
> +static void make_tx_response(struct xenvif *vif,
> + struct xen_netif_tx_request *txp,
> + s8 st);
> +static struct xen_netif_rx_response *make_rx_response(struct xenvif *vif,
> + u16 id,
> + s8 st,
> + u16 offset,
> + u16 size,
> + u16 flags);
> +
> +static inline unsigned long idx_to_pfn(struct xen_netbk *netbk,
> + unsigned int idx)
> +{
> + return page_to_pfn(netbk->mmap_pages[idx]);
> +}
> +
> +static inline unsigned long idx_to_kaddr(struct xen_netbk *netbk,
> + unsigned int idx)
> +{
> + return (unsigned long)pfn_to_kaddr(idx_to_pfn(netbk, idx));
> +}
> +
> +/* extra field used in struct page */
> +static inline void set_page_ext(struct page *pg, struct xen_netbk *netbk,
> + unsigned int idx)
> +{
> + unsigned int group = netbk - xen_netbk;
> + union page_ext ext = { .e = { .group = group + 1, .idx = idx } };
> +
> + BUILD_BUG_ON(sizeof(ext) > sizeof(ext.mapping));
> + pg->mapping = ext.mapping;
> +}
> +
> +static int get_page_ext(struct page *pg,
> + unsigned int *pgroup, unsigned int *pidx)
> +{
> + union page_ext ext = { .mapping = pg->mapping };
> + struct xen_netbk *netbk;
> + unsigned int group, idx;
> +
> + group = ext.e.group - 1;
> +
> + if (group < 0 || group >= xen_netbk_group_nr)
> + return 0;
> +
> + netbk = &xen_netbk[group];
> +
> + idx = ext.e.idx;
> +
> + if ((idx < 0) || (idx >= MAX_PENDING_REQS))
> + return 0;
> +
> + if (netbk->mmap_pages[idx] != pg)
> + return 0;
> +
> + *pgroup = group;
> + *pidx = idx;
> +
> + return 1;
> +}
> +
> +/*
> + * This is the amount of packet we copy rather than map, so that the
> + * guest can't fiddle with the contents of the headers while we do
> + * packet processing on them (netfilter, routing, etc).
> + */
> +#define PKT_PROT_LEN (ETH_HLEN + \
> + VLAN_HLEN + \
> + sizeof(struct iphdr) + MAX_IPOPTLEN + \
> + sizeof(struct tcphdr) + MAX_TCP_OPTION_SPACE)
> +
> +static inline pending_ring_idx_t pending_index(unsigned i)
> +{
> + return i & (MAX_PENDING_REQS-1);
> +}
> +
> +static inline pending_ring_idx_t nr_pending_reqs(struct xen_netbk *netbk)
> +{
> + return MAX_PENDING_REQS -
> + netbk->pending_prod + netbk->pending_cons;
> +}
> +
> +static void xen_netbk_kick_thread(struct xen_netbk *netbk)
> +{
> + wake_up(&netbk->wq);
> +}
> +
> +/*
> + * Returns true if we should start a new receive buffer instead of
> + * adding 'size' bytes to a buffer which currently contains 'offset'
> + * bytes.
> + */
> +static bool start_new_rx_buffer(int offset, unsigned long size, int head)
> +{
> + /* simple case: we have completely filled the current buffer. */
> + if (offset == MAX_BUFFER_OFFSET)
> + return true;
> +
> + /*
> + * complex case: start a fresh buffer if the current frag
> + * would overflow the current buffer but only if:
> + * (i) this frag would fit completely in the next buffer
> + * and (ii) there is already some data in the current buffer
> + * and (iii) this is not the head buffer.
> + *
> + * Where:
> + * - (i) stops us splitting a frag into two copies
> + * unless the frag is too large for a single buffer.
> + * - (ii) stops us from leaving a buffer pointlessly empty.
> + * - (iii) stops us leaving the first buffer
> + * empty. Strictly speaking this is already covered
> + * by (ii) but is explicitly checked because
> + * netfront relies on the first buffer being
> + * non-empty and can crash otherwise.
> + *
> + * This means we will effectively linearise small
> + * frags but do not needlessly split large buffers
> + * into multiple copies tend to give large frags their
> + * own buffers as before.
> + */
> + if ((offset + size > MAX_BUFFER_OFFSET) &&
> + (size <= MAX_BUFFER_OFFSET) && offset && !head)
> + return true;
> +
> + return false;
> +}
> +
> +/*
> + * Figure out how many ring slots we're going to need to send @skb to
> + * the guest. This function is essentially a dry run of
> + * netbk_gop_frag_copy.
> + */
> +unsigned int xen_netbk_count_skb_slots(struct xenvif *vif, struct sk_buff *skb)
> +{
> + unsigned int count;
> + int i, copy_off;
> +
> + count = DIV_ROUND_UP(
> + offset_in_page(skb->data)+skb_headlen(skb), PAGE_SIZE);
> +
> + copy_off = skb_headlen(skb) % PAGE_SIZE;
> +
> + if (skb_shinfo(skb)->gso_size)
> + count++;
> +
> + for (i = 0; i < skb_shinfo(skb)->nr_frags; i++) {
> + unsigned long size = skb_shinfo(skb)->frags[i].size;
> + unsigned long bytes;
> + while (size > 0) {
> + BUG_ON(copy_off > MAX_BUFFER_OFFSET);
> +
> + if (start_new_rx_buffer(copy_off, size, 0)) {
> + count++;
> + copy_off = 0;
> + }
> +
> + bytes = size;
> + if (copy_off + bytes > MAX_BUFFER_OFFSET)
> + bytes = MAX_BUFFER_OFFSET - copy_off;
> +
> + copy_off += bytes;
> + size -= bytes;
> + }
> + }
> + return count;
> +}
> +
> +struct netrx_pending_operations {
> + unsigned copy_prod, copy_cons;
> + unsigned meta_prod, meta_cons;
> + struct gnttab_copy *copy;
> + struct netbk_rx_meta *meta;
> + int copy_off;
> + grant_ref_t copy_gref;
> +};
> +
> +static struct netbk_rx_meta *get_next_rx_buffer(struct xenvif *vif,
> + struct netrx_pending_operations *npo)
> +{
> + struct netbk_rx_meta *meta;
> + struct xen_netif_rx_request *req;
> +
> + req = RING_GET_REQUEST(&vif->rx, vif->rx.req_cons++);
> +
> + meta = npo->meta + npo->meta_prod++;
> + meta->gso_size = 0;
> + meta->size = 0;
> + meta->id = req->id;
> +
> + npo->copy_off = 0;
> + npo->copy_gref = req->gref;
> +
> + return meta;
> +}
> +
> +/*
> + * Set up the grant operations for this fragment. If it's a flipping
> + * interface, we also set up the unmap request from here.
> + */
> +static void netbk_gop_frag_copy(struct xenvif *vif, struct sk_buff *skb,
> + struct netrx_pending_operations *npo,
> + struct page *page, unsigned long size,
> + unsigned long offset, int *head)
> +{
> + struct gnttab_copy *copy_gop;
> + struct netbk_rx_meta *meta;
> + /*
> + * These variables a used iff get_page_ext returns true,
> + * in which case they are guaranteed to be initialized.
> + */
> + unsigned int uninitialized_var(group), uninitialized_var(idx);
> + int foreign = get_page_ext(page, &group, &idx);
> + unsigned long bytes;
> +
> + /* Data must not cross a page boundary. */
> + BUG_ON(size + offset > PAGE_SIZE);
> +
> + meta = npo->meta + npo->meta_prod - 1;
> +
> + while (size > 0) {
> + BUG_ON(npo->copy_off > MAX_BUFFER_OFFSET);
> +
> + if (start_new_rx_buffer(npo->copy_off, size, *head)) {
> + /*
> + * Netfront requires there to be some data in the head
> + * buffer.
> + */
> + BUG_ON(*head);
What if we just WARN?
> +
> + meta = get_next_rx_buffer(vif, npo);
> + }
> +
> + bytes = size;
> + if (npo->copy_off + bytes > MAX_BUFFER_OFFSET)
> + bytes = MAX_BUFFER_OFFSET - npo->copy_off;
> +
> + copy_gop = npo->copy + npo->copy_prod++;
> + copy_gop->flags = GNTCOPY_dest_gref;
> + if (foreign) {
> + struct xen_netbk *netbk = &xen_netbk[group];
> + struct pending_tx_info *src_pend;
> +
> + src_pend = &netbk->pending_tx_info[idx];
> +
> + copy_gop->source.domid = src_pend->vif->domid;
> + copy_gop->source.u.ref = src_pend->req.gref;
> + copy_gop->flags |= GNTCOPY_source_gref;
> + } else {
> + void *vaddr = page_address(page);
> + copy_gop->source.domid = DOMID_SELF;
> + copy_gop->source.u.gmfn = virt_to_mfn(vaddr);
> + }
> + copy_gop->source.offset = offset;
> + copy_gop->dest.domid = vif->domid;
> +
> + copy_gop->dest.offset = npo->copy_off;
> + copy_gop->dest.u.ref = npo->copy_gref;
> + copy_gop->len = bytes;
> +
> + npo->copy_off += bytes;
> + meta->size += bytes;
> +
> + offset += bytes;
> + size -= bytes;
> +
> + /* Leave a gap for the GSO descriptor. */
> + if (*head && skb_shinfo(skb)->gso_size && !vif->gso_prefix)
> + vif->rx.req_cons++;
> +
> + *head = 0; /* There must be something in this buffer now. */
> +
> + }
> +}
> +
> +/*
> + * Prepare an SKB to be transmitted to the frontend.
> + *
> + * This function is responsible for allocating grant operations, meta
> + * structures, etc.
> + *
> + * It returns the number of meta structures consumed. The number of
> + * ring slots used is always equal to the number of meta slots used
> + * plus the number of GSO descriptors used. Currently, we use either
> + * zero GSO descriptors (for non-GSO packets) or one descriptor (for
> + * frontend-side LRO).
> + */
> +static int netbk_gop_skb(struct sk_buff *skb,
> + struct netrx_pending_operations *npo)
> +{
> + struct xenvif *vif = netdev_priv(skb->dev);
> + int nr_frags = skb_shinfo(skb)->nr_frags;
> + int i;
> + struct xen_netif_rx_request *req;
> + struct netbk_rx_meta *meta;
> + unsigned char *data;
> + int head = 1;
> + int old_meta_prod;
> +
> + old_meta_prod = npo->meta_prod;
> +
> + /* Set up a GSO prefix descriptor, if necessary */
> + if (skb_shinfo(skb)->gso_size && vif->gso_prefix) {
> + req = RING_GET_REQUEST(&vif->rx, vif->rx.req_cons++);
> + meta = npo->meta + npo->meta_prod++;
> + meta->gso_size = skb_shinfo(skb)->gso_size;
> + meta->size = 0;
> + meta->id = req->id;
> + }
> +
> + req = RING_GET_REQUEST(&vif->rx, vif->rx.req_cons++);
> + meta = npo->meta + npo->meta_prod++;
> +
> + if (!vif->gso_prefix)
> + meta->gso_size = skb_shinfo(skb)->gso_size;
> + else
> + meta->gso_size = 0;
> +
> + meta->size = 0;
> + meta->id = req->id;
> + npo->copy_off = 0;
> + npo->copy_gref = req->gref;
> +
> + data = skb->data;
> + while (data < skb_tail_pointer(skb)) {
> + unsigned int offset = offset_in_page(data);
> + unsigned int len = PAGE_SIZE - offset;
> +
> + if (data + len > skb_tail_pointer(skb))
> + len = skb_tail_pointer(skb) - data;
> +
> + netbk_gop_frag_copy(vif, skb, npo,
> + virt_to_page(data), len, offset, &head);
> + data += len;
> + }
> +
> + for (i = 0; i < nr_frags; i++) {
> + netbk_gop_frag_copy(vif, skb, npo,
> + skb_shinfo(skb)->frags[i].page,
> + skb_shinfo(skb)->frags[i].size,
> + skb_shinfo(skb)->frags[i].page_offset,
> + &head);
> + }
> +
> + return npo->meta_prod - old_meta_prod;
> +}
> +
> +/*
> + * This is a twin to netbk_gop_skb. Assume that netbk_gop_skb was
> + * used to set up the operations on the top of
> + * netrx_pending_operations, which have since been done. Check that
> + * they didn't give any errors and advance over them.
> + */
> +static int netbk_check_gop(int nr_meta_slots, domid_t domid,
> + struct netrx_pending_operations *npo)
> +{
> + struct gnttab_copy *copy_op;
> + int status = XEN_NETIF_RSP_OKAY;
> + int i;
> +
> + for (i = 0; i < nr_meta_slots; i++) {
> + copy_op = npo->copy + npo->copy_cons++;
> + if (copy_op->status != GNTST_okay) {
> + pr_debug("Bad status %d from copy to DOM%d.\n",
> + copy_op->status, domid);
pr_warn or pr_info?
> + status = XEN_NETIF_RSP_ERROR;
should we just break here?
> + }
> + }
> +
> + return status;
> +}
> +
> +static void netbk_add_frag_responses(struct xenvif *vif, int status,
> + struct netbk_rx_meta *meta,
> + int nr_meta_slots)
> +{
> + int i;
> + unsigned long offset;
> +
> + /* No fragments used */
> + if (nr_meta_slots <= 1)
> + return;
> +
> + nr_meta_slots--;
> +
> + for (i = 0; i < nr_meta_slots; i++) {
> + int flags;
> + if (i == nr_meta_slots - 1)
> + flags = 0;
> + else
> + flags = XEN_NETRXF_more_data;
> +
> + offset = 0;
> + make_rx_response(vif, meta[i].id, status, offset,
> + meta[i].size, flags);
> + }
> +}
> +
> +struct skb_cb_overlay {
> + int meta_slots_used;
> +};
> +
> +static void xen_netbk_rx_action(struct xen_netbk *netbk)
> +{
> + struct xenvif *vif = NULL;
> + s8 status;
> + u16 irq, flags;
> + struct xen_netif_rx_response *resp;
> + struct sk_buff_head rxq;
> + struct sk_buff *skb;
> + int notify_nr = 0;
> + int ret;
> + int nr_frags;
> + int count;
> + unsigned long offset;
> + struct skb_cb_overlay *sco;
> +
> + struct netrx_pending_operations npo = {
> + .copy = netbk->grant_copy_op,
> + .meta = netbk->meta,
> + };
> +
> + skb_queue_head_init(&rxq);
> +
> + count = 0;
> +
> + while ((skb = skb_dequeue(&netbk->rx_queue)) != NULL) {
> + vif = netdev_priv(skb->dev);
> + nr_frags = skb_shinfo(skb)->nr_frags;
> +
> + sco = (struct skb_cb_overlay *)skb->cb;
> + sco->meta_slots_used = netbk_gop_skb(skb, &npo);
> +
> + count += nr_frags + 1;
> +
> + __skb_queue_tail(&rxq, skb);
> +
> + /* Filled the batch queue? */
> + if (count + MAX_SKB_FRAGS >= XEN_NETIF_RX_RING_SIZE)
> + break;
> + }
> +
> + BUG_ON(npo.meta_prod > ARRAY_SIZE(netbk->meta));
> +
> + if (!npo.copy_prod)
> + return;
> +
> + BUG_ON(npo.copy_prod > ARRAY_SIZE(netbk->grant_copy_op));
> + ret = HYPERVISOR_grant_table_op(GNTTABOP_copy, &netbk->grant_copy_op,
> + npo.copy_prod);
> + BUG_ON(ret != 0);
> +
> + while ((skb = __skb_dequeue(&rxq)) != NULL) {
> + sco = (struct skb_cb_overlay *)skb->cb;
> +
> + vif = netdev_priv(skb->dev);
> +
> + if (netbk->meta[npo.meta_cons].gso_size && vif->gso_prefix) {
> + resp = RING_GET_RESPONSE(&vif->rx,
> + vif->rx.rsp_prod_pvt++);
> +
> + resp->flags = XEN_NETRXF_gso_prefix | XEN_NETRXF_more_data;
> +
> + resp->offset = netbk->meta[npo.meta_cons].gso_size;
> + resp->id = netbk->meta[npo.meta_cons].id;
> + resp->status = sco->meta_slots_used;
> +
> + npo.meta_cons++;
> + sco->meta_slots_used--;
> + }
> +
> +
> + vif->stats.tx_bytes += skb->len;
> + vif->stats.tx_packets++;
> +
> + status = netbk_check_gop(sco->meta_slots_used,
> + vif->domid, &npo);
> +
> + if (sco->meta_slots_used == 1)
> + flags = 0;
> + else
> + flags = XEN_NETRXF_more_data;
> +
> + if (skb->ip_summed == CHECKSUM_PARTIAL) /* local packet? */
> + flags |= XEN_NETRXF_csum_blank | XEN_NETRXF_data_validated;
> + else if (skb->ip_summed == CHECKSUM_UNNECESSARY)
> + /* remote but checksummed. */
> + flags |= XEN_NETRXF_data_validated;
> +
> + offset = 0;
> + resp = make_rx_response(vif, netbk->meta[npo.meta_cons].id,
> + status, offset,
> + netbk->meta[npo.meta_cons].size,
> + flags);
> +
> + if (netbk->meta[npo.meta_cons].gso_size && !vif->gso_prefix) {
> + struct xen_netif_extra_info *gso =
> + (struct xen_netif_extra_info *)
> + RING_GET_RESPONSE(&vif->rx,
> + vif->rx.rsp_prod_pvt++);
> +
> + resp->flags |= XEN_NETRXF_extra_info;
> +
> + gso->u.gso.size = netbk->meta[npo.meta_cons].gso_size;
> + gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4;
> + gso->u.gso.pad = 0;
> + gso->u.gso.features = 0;
> +
> + gso->type = XEN_NETIF_EXTRA_TYPE_GSO;
> + gso->flags = 0;
> + }
> +
> + netbk_add_frag_responses(vif, status,
> + netbk->meta + npo.meta_cons + 1,
> + sco->meta_slots_used);
> +
> + RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&vif->rx, ret);
> + irq = vif->irq;
> + if (ret && !netbk->rx_notify[irq]) {
> + netbk->rx_notify[irq] = 1;
> + netbk->notify_list[notify_nr++] = irq;
> + }
> +
> + if (netif_queue_stopped(vif->dev) &&
> + xenvif_schedulable(vif) &&
> + !xenvif_queue_full(vif))
> + netif_wake_queue(vif->dev);
> +
> + xenvif_put(vif);
> + npo.meta_cons += sco->meta_slots_used;
> + dev_kfree_skb(skb);
> + }
> +
> + while (notify_nr != 0) {
> + irq = netbk->notify_list[--notify_nr];
> + netbk->rx_notify[irq] = 0;
> + notify_remote_via_irq(irq);
> + }
> +
> + /* More work to do? */
> + if (!skb_queue_empty(&netbk->rx_queue) &&
> + !timer_pending(&netbk->net_timer))
> + xen_netbk_kick_thread(netbk);
> +}
> +
> +void xen_netbk_queue_tx_skb(struct xenvif *vif, struct sk_buff *skb)
> +{
> + struct xen_netbk *netbk = vif->netbk;
> +
> + skb_queue_tail(&netbk->rx_queue, skb);
> +
> + xen_netbk_kick_thread(netbk);
> +}
> +
> +static void xen_netbk_alarm(unsigned long data)
> +{
> + struct xen_netbk *netbk = (struct xen_netbk *)data;
> + xen_netbk_kick_thread(netbk);
> +}
> +
> +static int __on_net_schedule_list(struct xenvif *vif)
> +{
> + return !list_empty(&vif->list);
> +}
> +
> +/* Must be called with net_schedule_list_lock held */
> +static void remove_from_net_schedule_list(struct xenvif *vif)
> +{
> + if (likely(__on_net_schedule_list(vif))) {
> + list_del_init(&vif->list);
> + xenvif_put(vif);
> + }
> +}
> +
> +static struct xenvif *poll_net_schedule_list(struct xen_netbk *netbk)
> +{
> + struct xenvif *vif = NULL;
> +
> + spin_lock_irq(&netbk->net_schedule_list_lock);
> + if (list_empty(&netbk->net_schedule_list))
> + goto out;
> +
> + vif = list_first_entry(&netbk->net_schedule_list,
> + struct xenvif, list);
> + if (!vif)
> + goto out;
> +
> + xenvif_get(vif);
> +
> + remove_from_net_schedule_list(vif);
> +out:
> + spin_unlock_irq(&netbk->net_schedule_list_lock);
> + return vif;
> +}
> +
> +void xen_netbk_schedule_xenvif(struct xenvif *vif)
> +{
> + unsigned long flags;
> +
> + struct xen_netbk *netbk = vif->netbk;
> + if (__on_net_schedule_list(vif))
> + goto kick;
> +
> + spin_lock_irqsave(&netbk->net_schedule_list_lock, flags);
> + if (!__on_net_schedule_list(vif) &&
> + likely(xenvif_schedulable(vif))) {
> + list_add_tail(&vif->list, &netbk->net_schedule_list);
> + xenvif_get(vif);
> + }
> + spin_unlock_irqrestore(&netbk->net_schedule_list_lock, flags);
> +
> +kick:
> + smp_mb();
> + if ((nr_pending_reqs(netbk) < (MAX_PENDING_REQS/2)) &&
Would it make sense to make this a runtime knob to increase/decrease
the batching count?
> + !list_empty(&netbk->net_schedule_list))
> + xen_netbk_kick_thread(netbk);
> +}
> +
> +void xen_netbk_deschedule_xenfif(struct xenvif *vif)
> +{
> + struct xen_netbk *netbk = vif->netbk;
> + spin_lock_irq(&netbk->net_schedule_list_lock);
> + remove_from_net_schedule_list(vif);
> + spin_unlock_irq(&netbk->net_schedule_list_lock);
> +}
> +
> +static void tx_add_credit(struct xenvif *vif)
> +{
> + unsigned long max_burst, max_credit;
> +
> + /*
> + * Allow a burst big enough to transmit a jumbo packet of up to 128kB.
> + * Otherwise the interface can seize up due to insufficient credit.
> + */
> + max_burst = RING_GET_REQUEST(&vif->tx, vif->tx.req_cons)->size;
> + max_burst = min(max_burst, 131072UL);
> + max_burst = max(max_burst, vif->credit_bytes);
> +
> + /* Take care that adding a new chunk of credit doesn't wrap to zero. */
> + max_credit = vif->remaining_credit + vif->credit_bytes;
> + if (max_credit < vif->remaining_credit)
> + max_credit = ULONG_MAX; /* wrapped: clamp to ULONG_MAX */
> +
> + vif->remaining_credit = min(max_credit, max_burst);
> +}
> +
> +static void tx_credit_callback(unsigned long data)
> +{
> + struct xenvif *vif = (struct xenvif *)data;
> + tx_add_credit(vif);
> + xenvif_schedule_work(vif);
> +}
> +
> +static void netbk_tx_err(struct xenvif *vif,
> + struct xen_netif_tx_request *txp, RING_IDX end)
> +{
> + RING_IDX cons = vif->tx.req_cons;
> +
> + do {
> + make_tx_response(vif, txp, XEN_NETIF_RSP_ERROR);
> + if (cons >= end)
> + break;
> + txp = RING_GET_REQUEST(&vif->tx, cons++);
> + } while (1);
> + vif->tx.req_cons = cons;
> + xenvif_schedule_work(vif);
> + xenvif_put(vif);
> +}
> +
> +static int netbk_count_requests(struct xenvif *vif,
> + struct xen_netif_tx_request *first,
> + struct xen_netif_tx_request *txp,
> + int work_to_do)
> +{
> + RING_IDX cons = vif->tx.req_cons;
> + int frags = 0;
> +
> + if (!(first->flags & XEN_NETTXF_more_data))
> + return 0;
> +
> + do {
> + if (frags >= work_to_do) {
> + pr_debug("Need more frags\n");
> + return -frags;
> + }
> +
> + if (unlikely(frags >= MAX_SKB_FRAGS)) {
> + pr_debug("Too many frags\n");
> + return -frags;
> + }
> +
> + memcpy(txp, RING_GET_REQUEST(&vif->tx, cons + frags),
> + sizeof(*txp));
> + if (txp->size > first->size) {
> + pr_debug("Frags galore\n");
> + return -frags;
> + }
> +
> + first->size -= txp->size;
> + frags++;
> +
> + if (unlikely((txp->offset + txp->size) > PAGE_SIZE)) {
> + pr_debug("txp->offset: %x, size: %u\n",
> + txp->offset, txp->size);
> + return -frags;
> + }
> + } while ((txp++)->flags & XEN_NETTXF_more_data);
> + return frags;
> +}
> +
> +static struct page *xen_netbk_alloc_page(struct xen_netbk *netbk,
> + struct sk_buff *skb,
> + unsigned long pending_idx)
> +{
> + struct page *page;
> + page = alloc_page(GFP_KERNEL|__GFP_COLD);
> + if (!page)
> + return NULL;
> + set_page_ext(page, netbk, pending_idx);
> + netbk->mmap_pages[pending_idx] = page;
> + return page;
> +}
> +
> +static struct gnttab_copy *xen_netbk_get_requests(struct xen_netbk *netbk,
> + struct xenvif *vif,
> + struct sk_buff *skb,
> + struct xen_netif_tx_request *txp,
> + struct gnttab_copy *gop)
> +{
> + struct skb_shared_info *shinfo = skb_shinfo(skb);
> + skb_frag_t *frags = shinfo->frags;
> + unsigned long pending_idx = *((u16 *)skb->data);
> + int i, start;
> +
> + /* Skip first skb fragment if it is on same page as header fragment. */
> + start = ((unsigned long)shinfo->frags[0].page == pending_idx);
> +
> + for (i = start; i < shinfo->nr_frags; i++, txp++) {
> + struct page *page;
> + pending_ring_idx_t index;
> + struct pending_tx_info *pending_tx_info =
> + netbk->pending_tx_info;
> +
> + index = pending_index(netbk->pending_cons++);
> + pending_idx = netbk->pending_ring[index];
> + page = xen_netbk_alloc_page(netbk, skb, pending_idx);
> + if (!page)
> + return NULL;
> +
> + netbk->mmap_pages[pending_idx] = page;
> +
> + gop->source.u.ref = txp->gref;
> + gop->source.domid = vif->domid;
> + gop->source.offset = txp->offset;
> +
> + gop->dest.u.gmfn = virt_to_mfn(page_address(page));
> + gop->dest.domid = DOMID_SELF;
> + gop->dest.offset = txp->offset;
> +
> + gop->len = txp->size;
> + gop->flags = GNTCOPY_source_gref;
> +
> + gop++;
> +
> + memcpy(&pending_tx_info[pending_idx].req, txp, sizeof(*txp));
> + xenvif_get(vif);
> + pending_tx_info[pending_idx].vif = vif;
> + frags[i].page = (void *)pending_idx;
> + }
> +
> + return gop;
> +}
> +
> +static int xen_netbk_tx_check_gop(struct xen_netbk *netbk,
> + struct sk_buff *skb,
> + struct gnttab_copy **gopp)
> +{
> + struct gnttab_copy *gop = *gopp;
> + int pending_idx = *((u16 *)skb->data);
> + struct pending_tx_info *pending_tx_info = netbk->pending_tx_info;
> + struct xenvif *vif = pending_tx_info[pending_idx].vif;
> + struct xen_netif_tx_request *txp;
> + struct skb_shared_info *shinfo = skb_shinfo(skb);
> + int nr_frags = shinfo->nr_frags;
> + int i, err, start;
> +
> + /* Check status of header. */
> + err = gop->status;
> + if (unlikely(err)) {
> + pending_ring_idx_t index;
> + index = pending_index(netbk->pending_prod++);
> + txp = &pending_tx_info[pending_idx].req;
> + make_tx_response(vif, txp, XEN_NETIF_RSP_ERROR);
> + netbk->pending_ring[index] = pending_idx;
> + xenvif_put(vif);
> + }
> +
> + /* Skip first skb fragment if it is on same page as header fragment. */
> + start = ((unsigned long)shinfo->frags[0].page == pending_idx);
> +
> + for (i = start; i < nr_frags; i++) {
> + int j, newerr;
> + pending_ring_idx_t index;
> +
> + pending_idx = (unsigned long)shinfo->frags[i].page;
> +
> + /* Check error status: if okay then remember grant handle. */
> + newerr = (++gop)->status;
> + if (likely(!newerr)) {
> + /* Had a previous error? Invalidate this fragment. */
> + if (unlikely(err))
> + xen_netbk_idx_release(netbk, pending_idx);
> + continue;
> + }
> +
> + /* Error on this fragment: respond to client with an error. */
> + txp = &netbk->pending_tx_info[pending_idx].req;
> + make_tx_response(vif, txp, XEN_NETIF_RSP_ERROR);
> + index = pending_index(netbk->pending_prod++);
> + netbk->pending_ring[index] = pending_idx;
> + xenvif_put(vif);
> +
> + /* Not the first error? Preceding frags already invalidated. */
> + if (err)
> + continue;
> +
> + /* First error: invalidate header and preceding fragments. */
> + pending_idx = *((u16 *)skb->data);
> + xen_netbk_idx_release(netbk, pending_idx);
> + for (j = start; j < i; j++) {
> + pending_idx = (unsigned long)shinfo->frags[i].page;
> + xen_netbk_idx_release(netbk, pending_idx);
> + }
> +
> + /* Remember the error: invalidate all subsequent fragments. */
> + err = newerr;
> + }
> +
> + *gopp = gop + 1;
> + return err;
> +}
> +
> +static void xen_netbk_fill_frags(struct xen_netbk *netbk, struct sk_buff *skb)
> +{
> + struct skb_shared_info *shinfo = skb_shinfo(skb);
> + int nr_frags = shinfo->nr_frags;
> + int i;
> +
> + for (i = 0; i < nr_frags; i++) {
> + skb_frag_t *frag = shinfo->frags + i;
> + struct xen_netif_tx_request *txp;
> + unsigned long pending_idx;
> +
> + pending_idx = (unsigned long)frag->page;
> +
> + txp = &netbk->pending_tx_info[pending_idx].req;
> + frag->page = virt_to_page(idx_to_kaddr(netbk, pending_idx));
> + frag->size = txp->size;
> + frag->page_offset = txp->offset;
> +
> + skb->len += txp->size;
> + skb->data_len += txp->size;
> + skb->truesize += txp->size;
> +
> + /* Take an extra reference to offset xen_netbk_idx_release */
> + get_page(netbk->mmap_pages[pending_idx]);
> + xen_netbk_idx_release(netbk, pending_idx);
> + }
> +}
> +
> +static int xen_netbk_get_extras(struct xenvif *vif,
> + struct xen_netif_extra_info *extras,
> + int work_to_do)
> +{
> + struct xen_netif_extra_info extra;
> + RING_IDX cons = vif->tx.req_cons;
> +
> + do {
> + if (unlikely(work_to_do-- <= 0)) {
> + pr_debug("Missing extra info\n");
> + return -EBADR;
> + }
> +
> + memcpy(&extra, RING_GET_REQUEST(&vif->tx, cons),
> + sizeof(extra));
> + if (unlikely(!extra.type ||
> + extra.type >= XEN_NETIF_EXTRA_TYPE_MAX)) {
> + vif->tx.req_cons = ++cons;
> + pr_debug("Invalid extra type: %d\n", extra.type);
> + return -EINVAL;
> + }
> +
> + memcpy(&extras[extra.type - 1], &extra, sizeof(extra));
> + vif->tx.req_cons = ++cons;
> + } while (extra.flags & XEN_NETIF_EXTRA_FLAG_MORE);
> +
> + return work_to_do;
> +}
> +
> +static int netbk_set_skb_gso(struct sk_buff *skb,
> + struct xen_netif_extra_info *gso)
> +{
> + if (!gso->u.gso.size) {
> + pr_debug("GSO size must not be zero.\n");
> + return -EINVAL;
> + }
> +
> + /* Currently only TCPv4 S.O. is supported. */
> + if (gso->u.gso.type != XEN_NETIF_GSO_TYPE_TCPV4) {
> + pr_debug("Bad GSO type %d.\n", gso->u.gso.type);
> + return -EINVAL;
> + }
> +
> + skb_shinfo(skb)->gso_size = gso->u.gso.size;
> + skb_shinfo(skb)->gso_type = SKB_GSO_TCPV4;
> +
> + /* Header must be checked, and gso_segs computed. */
> + skb_shinfo(skb)->gso_type |= SKB_GSO_DODGY;
> + skb_shinfo(skb)->gso_segs = 0;
> +
> + return 0;
> +}
> +
> +static int checksum_setup(struct xenvif *vif, struct sk_buff *skb)
> +{
> + struct iphdr *iph;
> + unsigned char *th;
> + int err = -EPROTO;
> + int recalculate_partial_csum = 0;
> +
> + /*
> + * A GSO SKB must be CHECKSUM_PARTIAL. However some buggy
> + * peers can fail to set NETRXF_csum_blank when sending a GSO
> + * frame. In this case force the SKB to CHECKSUM_PARTIAL and
> + * recalculate the partial checksum.
> + */
> + if (skb->ip_summed != CHECKSUM_PARTIAL && skb_is_gso(skb)) {
> + vif->rx_gso_checksum_fixup++;
> + skb->ip_summed = CHECKSUM_PARTIAL;
> + recalculate_partial_csum = 1;
> + }
> +
> + /* A non-CHECKSUM_PARTIAL SKB does not require setup. */
> + if (skb->ip_summed != CHECKSUM_PARTIAL)
> + return 0;
> +
> + if (skb->protocol != htons(ETH_P_IP))
> + goto out;
> +
> + iph = (void *)skb->data;
> + th = skb->data + 4 * iph->ihl;
> + if (th >= skb_tail_pointer(skb))
> + goto out;
> +
> + skb->csum_start = th - skb->head;
> + switch (iph->protocol) {
> + case IPPROTO_TCP:
> + skb->csum_offset = offsetof(struct tcphdr, check);
> +
> + if (recalculate_partial_csum) {
> + struct tcphdr *tcph = (struct tcphdr *)th;
> + tcph->check = ~csum_tcpudp_magic(iph->saddr, iph->daddr,
> + skb->len - iph->ihl*4,
> + IPPROTO_TCP, 0);
> + }
> + break;
> + case IPPROTO_UDP:
> + skb->csum_offset = offsetof(struct udphdr, check);
> +
> + if (recalculate_partial_csum) {
> + struct udphdr *udph = (struct udphdr *)th;
> + udph->check = ~csum_tcpudp_magic(iph->saddr, iph->daddr,
> + skb->len - iph->ihl*4,
> + IPPROTO_UDP, 0);
> + }
> + break;
> + default:
> + if (net_ratelimit())
> + printk(KERN_ERR "Attempting to checksum a non-"
> + "TCP/UDP packet, dropping a protocol"
> + " %d packet", iph->protocol);
> + goto out;
> + }
> +
> + if ((th + skb->csum_offset + 2) > skb_tail_pointer(skb))
> + goto out;
> +
> + err = 0;
> +
> +out:
> + return err;
> +}
> +
> +static bool tx_credit_exceeded(struct xenvif *vif, unsigned size)
> +{
> + unsigned long now = jiffies;
> + unsigned long next_credit =
> + vif->credit_timeout.expires +
> + msecs_to_jiffies(vif->credit_usec / 1000);
> +
> + /* Timer could already be pending in rare cases. */
> + if (timer_pending(&vif->credit_timeout))
> + return true;
> +
> + /* Passed the point where we can replenish credit? */
> + if (time_after_eq(now, next_credit)) {
> + vif->credit_timeout.expires = now;
> + tx_add_credit(vif);
> + }
> +
> + /* Still too big to send right now? Set a callback. */
> + if (size > vif->remaining_credit) {
> + vif->credit_timeout.data =
> + (unsigned long)vif;
> + vif->credit_timeout.function =
> + tx_credit_callback;
> + mod_timer(&vif->credit_timeout,
> + next_credit);
> +
> + return true;
> + }
> +
> + return false;
> +}
> +
> +static unsigned xen_netbk_tx_build_gops(struct xen_netbk *netbk)
> +{
> + struct gnttab_copy *gop = netbk->tx_copy_ops, *request_gop;
> + struct sk_buff *skb;
> + int ret;
> +
> + while (((nr_pending_reqs(netbk) + MAX_SKB_FRAGS) < MAX_PENDING_REQS) &&
> + !list_empty(&netbk->net_schedule_list)) {
> + struct xenvif *vif;
> + struct xen_netif_tx_request txreq;
> + struct xen_netif_tx_request txfrags[MAX_SKB_FRAGS];
> + struct page *page;
> + struct xen_netif_extra_info extras[XEN_NETIF_EXTRA_TYPE_MAX-1];
> + u16 pending_idx;
> + RING_IDX idx;
> + int work_to_do;
> + unsigned int data_len;
> + pending_ring_idx_t index;
> +
> + /* Get a netif from the list with work to do. */
> + vif = poll_net_schedule_list(netbk);
> + if (!vif)
> + continue;
> +
> + RING_FINAL_CHECK_FOR_REQUESTS(&vif->tx, work_to_do);
> + if (!work_to_do) {
> + xenvif_put(vif);
> + continue;
> + }
> +
> + idx = vif->tx.req_cons;
> + rmb(); /* Ensure that we see the request before we copy it. */
> + memcpy(&txreq, RING_GET_REQUEST(&vif->tx, idx), sizeof(txreq));
> +
> + /* Credit-based scheduling. */
> + if (txreq.size > vif->remaining_credit &&
> + tx_credit_exceeded(vif, txreq.size)) {
> + xenvif_put(vif);
> + continue;
> + }
> +
> + vif->remaining_credit -= txreq.size;
> +
> + work_to_do--;
> + vif->tx.req_cons = ++idx;
> +
> + memset(extras, 0, sizeof(extras));
> + if (txreq.flags & XEN_NETTXF_extra_info) {
> + work_to_do = xen_netbk_get_extras(vif, extras,
> + work_to_do);
> + idx = vif->tx.req_cons;
> + if (unlikely(work_to_do < 0)) {
> + netbk_tx_err(vif, &txreq, idx);
> + continue;
> + }
> + }
> +
> + ret = netbk_count_requests(vif, &txreq, txfrags, work_to_do);
> + if (unlikely(ret < 0)) {
> + netbk_tx_err(vif, &txreq, idx - ret);
> + continue;
> + }
> + idx += ret;
> +
> + if (unlikely(txreq.size < ETH_HLEN)) {
> + pr_debug("Bad packet size: %d\n", txreq.size);
> + netbk_tx_err(vif, &txreq, idx);
> + continue;
> + }
> +
> + /* No crossing a page as the payload mustn't fragment. */
> + if (unlikely((txreq.offset + txreq.size) > PAGE_SIZE)) {
> + pr_debug("txreq.offset: %x, size: %u, end: %lu\n",
> + txreq.offset, txreq.size,
> + (txreq.offset&~PAGE_MASK) + txreq.size);
> + netbk_tx_err(vif, &txreq, idx);
> + continue;
> + }
> +
> + index = pending_index(netbk->pending_cons);
> + pending_idx = netbk->pending_ring[index];
> +
> + data_len = (txreq.size > PKT_PROT_LEN &&
> + ret < MAX_SKB_FRAGS) ?
> + PKT_PROT_LEN : txreq.size;
> +
> + skb = alloc_skb(data_len + NET_SKB_PAD + NET_IP_ALIGN,
> + GFP_ATOMIC | __GFP_NOWARN);
> + if (unlikely(skb == NULL)) {
> + pr_debug("Can't allocate a skb in start_xmit.\n");
> + netbk_tx_err(vif, &txreq, idx);
> + break;
> + }
> +
> + /* Packets passed to netif_rx() must have some headroom. */
> + skb_reserve(skb, NET_SKB_PAD + NET_IP_ALIGN);
> +
> + if (extras[XEN_NETIF_EXTRA_TYPE_GSO - 1].type) {
> + struct xen_netif_extra_info *gso;
> + gso = &extras[XEN_NETIF_EXTRA_TYPE_GSO - 1];
> +
> + if (netbk_set_skb_gso(skb, gso)) {
> + kfree_skb(skb);
> + netbk_tx_err(vif, &txreq, idx);
> + continue;
> + }
> + }
> +
> + /* XXX could copy straight to head */
> + page = xen_netbk_alloc_page(netbk, skb, pending_idx);
> + if (!page) {
> + kfree_skb(skb);
> + netbk_tx_err(vif, &txreq, idx);
> + continue;
> + }
> +
> + netbk->mmap_pages[pending_idx] = page;
> +
> + gop->source.u.ref = txreq.gref;
> + gop->source.domid = vif->domid;
> + gop->source.offset = txreq.offset;
> +
> + gop->dest.u.gmfn = virt_to_mfn(page_address(page));
> + gop->dest.domid = DOMID_SELF;
> + gop->dest.offset = txreq.offset;
> +
> + gop->len = txreq.size;
> + gop->flags = GNTCOPY_source_gref;
> +
> + gop++;
> +
> + memcpy(&netbk->pending_tx_info[pending_idx].req,
> + &txreq, sizeof(txreq));
> + netbk->pending_tx_info[pending_idx].vif = vif;
> + *((u16 *)skb->data) = pending_idx;
> +
> + __skb_put(skb, data_len);
> +
> + skb_shinfo(skb)->nr_frags = ret;
> + if (data_len < txreq.size) {
> + skb_shinfo(skb)->nr_frags++;
> + skb_shinfo(skb)->frags[0].page =
> + (void *)(unsigned long)pending_idx;
> + } else {
> + /* Discriminate from any valid pending_idx value. */
> + skb_shinfo(skb)->frags[0].page = (void *)~0UL;
> + }
> +
> + __skb_queue_tail(&netbk->tx_queue, skb);
> +
> + netbk->pending_cons++;
> +
> + request_gop = xen_netbk_get_requests(netbk, vif,
> + skb, txfrags, gop);
> + if (request_gop == NULL) {
> + kfree_skb(skb);
> + netbk_tx_err(vif, &txreq, idx);
> + continue;
> + }
> + gop = request_gop;
> +
> + vif->tx.req_cons = idx;
> + xenvif_schedule_work(vif);
> +
> + if ((gop-netbk->tx_copy_ops) >= ARRAY_SIZE(netbk->tx_copy_ops))
> + break;
> + }
> +
> + return gop - netbk->tx_copy_ops;
> +}
> +
> +static void xen_netbk_tx_submit(struct xen_netbk *netbk)
> +{
> + struct gnttab_copy *gop = netbk->tx_copy_ops;
> + struct sk_buff *skb;
> +
> + while ((skb = __skb_dequeue(&netbk->tx_queue)) != NULL) {
> + struct xen_netif_tx_request *txp;
> + struct xenvif *vif;
> + u16 pending_idx;
> + unsigned data_len;
> +
> + pending_idx = *((u16 *)skb->data);
> + vif = netbk->pending_tx_info[pending_idx].vif;
> + txp = &netbk->pending_tx_info[pending_idx].req;
> +
> + /* Check the remap error code. */
> + if (unlikely(xen_netbk_tx_check_gop(netbk, skb, &gop))) {
> + pr_debug("netback grant failed.\n");
> + skb_shinfo(skb)->nr_frags = 0;
> + kfree_skb(skb);
> + continue;
> + }
> +
> + data_len = skb->len;
> + memcpy(skb->data,
> + (void *)(idx_to_kaddr(netbk, pending_idx)|txp->offset),
> + data_len);
> + if (data_len < txp->size) {
> + /* Append the packet payload as a fragment. */
> + txp->offset += data_len;
> + txp->size -= data_len;
> + } else {
> + /* Schedule a response immediately. */
> + xen_netbk_idx_release(netbk, pending_idx);
> + }
> +
> + if (txp->flags & XEN_NETTXF_csum_blank)
> + skb->ip_summed = CHECKSUM_PARTIAL;
> + else if (txp->flags & XEN_NETTXF_data_validated)
> + skb->ip_summed = CHECKSUM_UNNECESSARY;
> +
> + xen_netbk_fill_frags(netbk, skb);
> +
> + /*
> + * If the initial fragment was < PKT_PROT_LEN then
> + * pull through some bytes from the other fragments to
> + * increase the linear region to PKT_PROT_LEN bytes.
> + */
> + if (skb_headlen(skb) < PKT_PROT_LEN && skb_is_nonlinear(skb)) {
> + int target = min_t(int, skb->len, PKT_PROT_LEN);
> + __pskb_pull_tail(skb, target - skb_headlen(skb));
> + }
> +
> + skb->dev = vif->dev;
> + skb->protocol = eth_type_trans(skb, skb->dev);
> +
> + if (checksum_setup(vif, skb)) {
> + pr_debug("Can't setup checksum in net_tx_action\n");
> + kfree_skb(skb);
> + continue;
> + }
> +
> + vif->stats.rx_bytes += skb->len;
> + vif->stats.rx_packets++;
> +
> + netif_rx_ni(skb);
> + vif->dev->last_rx = jiffies;
> + }
> +}
> +
> +/* Called after netfront has transmitted */
> +static void xen_netbk_tx_action(struct xen_netbk *netbk)
> +{
> + unsigned nr_gops;
> + int ret;
> +
> + nr_gops = xen_netbk_tx_build_gops(netbk);
> +
> + if (nr_gops == 0)
> + return;
> + ret = HYPERVISOR_grant_table_op(GNTTABOP_copy,
> + netbk->tx_copy_ops, nr_gops);
> + BUG_ON(ret);
> +
> + xen_netbk_tx_submit(netbk);
> +
> +}
> +
> +static void xen_netbk_idx_release(struct xen_netbk *netbk, u16 pending_idx)
> +{
> + struct xenvif *vif;
> + struct pending_tx_info *pending_tx_info;
> + pending_ring_idx_t index;
> +
> + /* Already complete? */
> + if (netbk->mmap_pages[pending_idx] == NULL)
> + return;
> +
> + pending_tx_info = &netbk->pending_tx_info[pending_idx];
> +
> + vif = pending_tx_info->vif;
> +
> + make_tx_response(vif, &pending_tx_info->req, XEN_NETIF_RSP_OKAY);
> +
> + index = pending_index(netbk->pending_prod++);
> + netbk->pending_ring[index] = pending_idx;
> +
> + xenvif_put(vif);
> +
> + netbk->mmap_pages[pending_idx]->mapping = 0;
> + put_page(netbk->mmap_pages[pending_idx]);
> + netbk->mmap_pages[pending_idx] = NULL;
> +}
> +
> +static void make_tx_response(struct xenvif *vif,
> + struct xen_netif_tx_request *txp,
> + s8 st)
> +{
> + RING_IDX i = vif->tx.rsp_prod_pvt;
> + struct xen_netif_tx_response *resp;
> + int notify;
> +
> + resp = RING_GET_RESPONSE(&vif->tx, i);
> + resp->id = txp->id;
> + resp->status = st;
> +
> + if (txp->flags & XEN_NETTXF_extra_info)
> + RING_GET_RESPONSE(&vif->tx, ++i)->status = XEN_NETIF_RSP_NULL;
> +
> + vif->tx.rsp_prod_pvt = ++i;
> + RING_PUSH_RESPONSES_AND_CHECK_NOTIFY(&vif->tx, notify);
> + if (notify)
> + notify_remote_via_irq(vif->irq);
> +}
> +
> +static struct xen_netif_rx_response *make_rx_response(struct xenvif *vif,
> + u16 id,
> + s8 st,
> + u16 offset,
> + u16 size,
> + u16 flags)
> +{
> + RING_IDX i = vif->rx.rsp_prod_pvt;
> + struct xen_netif_rx_response *resp;
> +
> + resp = RING_GET_RESPONSE(&vif->rx, i);
> + resp->offset = offset;
> + resp->flags = flags;
> + resp->id = id;
> + resp->status = (s16)size;
> + if (st < 0)
> + resp->status = (s16)st;
> +
> + vif->rx.rsp_prod_pvt = ++i;
> +
> + return resp;
> +}
> +
> +static inline int rx_work_todo(struct xen_netbk *netbk)
> +{
> + return !skb_queue_empty(&netbk->rx_queue);
> +}
> +
> +static inline int tx_work_todo(struct xen_netbk *netbk)
> +{
> +
> + if (((nr_pending_reqs(netbk) + MAX_SKB_FRAGS) < MAX_PENDING_REQS) &&
> + !list_empty(&netbk->net_schedule_list))
> + return 1;
> +
> + return 0;
> +}
> +
> +static int xen_netbk_kthread(void *data)
> +{
> + struct xen_netbk *netbk = (struct xen_netbk *)data;
> + while (!kthread_should_stop()) {
> + wait_event_interruptible(netbk->wq,
> + rx_work_todo(netbk)
> + || tx_work_todo(netbk)
> + || kthread_should_stop());
> + cond_resched();
> +
> + if (kthread_should_stop())
> + break;
> +
> + if (rx_work_todo(netbk))
> + xen_netbk_rx_action(netbk);
> +
> + if (tx_work_todo(netbk))
> + xen_netbk_tx_action(netbk);
> + }
> +
> + return 0;
> +}
> +
> +static int __init netback_init(void)
> +{
> + int i;
> + int rc = 0;
> + int group;
> +
> + if (!xen_pv_domain())
> + return -ENODEV;
> +
> + xen_netbk_group_nr = num_online_cpus();
> + xen_netbk = vmalloc(sizeof(struct xen_netbk) * xen_netbk_group_nr);
> + if (!xen_netbk) {
> + printk(KERN_ALERT "%s: out of memory\n", __func__);
> + return -ENOMEM;
> + }
> + memset(xen_netbk, 0, sizeof(struct xen_netbk) * xen_netbk_group_nr);
> +
> + for (group = 0; group < xen_netbk_group_nr; group++) {
> + struct xen_netbk *netbk = &xen_netbk[group];
> + skb_queue_head_init(&netbk->rx_queue);
> + skb_queue_head_init(&netbk->tx_queue);
> +
> + init_timer(&netbk->net_timer);
> + netbk->net_timer.data = (unsigned long)netbk;
> + netbk->net_timer.function = xen_netbk_alarm;
> +
> + netbk->pending_cons = 0;
> + netbk->pending_prod = MAX_PENDING_REQS;
> + for (i = 0; i < MAX_PENDING_REQS; i++)
> + netbk->pending_ring[i] = i;
> +
> + init_waitqueue_head(&netbk->wq);
> + netbk->task = kthread_create(xen_netbk_kthread,
> + (void *)netbk,
> + "netback/%u", group);
> +
> + if (IS_ERR(netbk->task)) {
> + printk(KERN_ALERT "kthread_run() fails at netback\n");
> + del_timer(&netbk->net_timer);
> + rc = PTR_ERR(netbk->task);
> + goto failed_init;
> + }
> +
> + kthread_bind(netbk->task, group);
> +
> + INIT_LIST_HEAD(&netbk->net_schedule_list);
> +
> + spin_lock_init(&netbk->net_schedule_list_lock);
> +
> + atomic_set(&netbk->netfront_count, 0);
> +
> + wake_up_process(netbk->task);
> + }
> +
> + rc = xenvif_xenbus_init();
> + if (rc)
> + goto failed_init;
> +
> + return 0;
> +
> +failed_init:
> + for (i = 0; i < group; i++) {
> + struct xen_netbk *netbk = &xen_netbk[i];
> + int j;
> + for (j = 0; j < MAX_PENDING_REQS; j++) {
> + if (netbk->mmap_pages[i])
> + __free_page(netbk->mmap_pages[i]);
> + }
> + del_timer(&netbk->net_timer);
> + kthread_stop(netbk->task);
> + }
> + vfree(xen_netbk);
> + return rc;
> +
> +}
> +
> +module_init(netback_init);
> +
> +MODULE_LICENSE("Dual BSD/GPL");
> diff --git a/drivers/net/xen-netback/xenbus.c b/drivers/net/xen-netback/xenbus.c
> new file mode 100644
> index 0000000..22b8c35
> --- /dev/null
> +++ b/drivers/net/xen-netback/xenbus.c
> @@ -0,0 +1,490 @@
> +/*
> + * Xenbus code for netif backend
> + *
> + * Copyright (C) 2005 Rusty Russell <rusty@rustcorp.com.au>
> + * Copyright (C) 2005 XenSource Ltd
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License as published by
> + * the Free Software Foundation; either version 2 of the License, or
> + * (at your option) any later version.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program; if not, write to the Free Software
> + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
> +*/
> +
> +#include "common.h"
> +
> +struct backend_info {
> + struct xenbus_device *dev;
> + struct xenvif *vif;
> + enum xenbus_state frontend_state;
> + struct xenbus_watch hotplug_status_watch;
> + int have_hotplug_status_watch:1;
> +};
> +
> +static int connect_rings(struct backend_info *);
> +static void connect(struct backend_info *);
> +static void backend_create_xenvif(struct backend_info *be);
> +static void unregister_hotplug_status_watch(struct backend_info *be);
> +
> +static int netback_remove(struct xenbus_device *dev)
> +{
> + struct backend_info *be = dev_get_drvdata(&dev->dev);
> +
> + unregister_hotplug_status_watch(be);
> + if (be->vif) {
> + kobject_uevent(&dev->dev.kobj, KOBJ_OFFLINE);
> + xenbus_rm(XBT_NIL, dev->nodename, "hotplug-status");
> + xenvif_disconnect(be->vif);
> + be->vif = NULL;
> + }
> + kfree(be);
> + dev_set_drvdata(&dev->dev, NULL);
> + return 0;
> +}
> +
> +
> +/**
> + * Entry point to this code when a new device is created. Allocate the basic
> + * structures and switch to InitWait.
> + */
> +static int netback_probe(struct xenbus_device *dev,
> + const struct xenbus_device_id *id)
> +{
> + const char *message;
> + struct xenbus_transaction xbt;
> + int err;
> + int sg;
> + struct backend_info *be = kzalloc(sizeof(struct backend_info),
> + GFP_KERNEL);
> + if (!be) {
> + xenbus_dev_fatal(dev, -ENOMEM,
> + "allocating backend structure");
> + return -ENOMEM;
> + }
> +
> + be->dev = dev;
> + dev_set_drvdata(&dev->dev, be);
> +
> + sg = 1;
> +
> + do {
> + err = xenbus_transaction_start(&xbt);
> + if (err) {
> + xenbus_dev_fatal(dev, err, "starting transaction");
> + goto fail;
> + }
> +
> + err = xenbus_printf(xbt, dev->nodename, "feature-sg", "%d", sg);
> + if (err) {
> + message = "writing feature-sg";
> + goto abort_transaction;
> + }
> +
> + err = xenbus_printf(xbt, dev->nodename, "feature-gso-tcpv4",
> + "%d", sg);
> + if (err) {
> + message = "writing feature-gso-tcpv4";
> + goto abort_transaction;
> + }
> +
> + /* We support rx-copy path. */
> + err = xenbus_printf(xbt, dev->nodename,
> + "feature-rx-copy", "%d", 1);
> + if (err) {
> + message = "writing feature-rx-copy";
> + goto abort_transaction;
> + }
> +
> + /*
> + * We don't support rx-flip path (except old guests who don't
> + * grok this feature flag).
> + */
> + err = xenbus_printf(xbt, dev->nodename,
> + "feature-rx-flip", "%d", 0);
> + if (err) {
> + message = "writing feature-rx-flip";
> + goto abort_transaction;
> + }
> +
> + err = xenbus_transaction_end(xbt, 0);
> + } while (err == -EAGAIN);
> +
> + if (err) {
> + xenbus_dev_fatal(dev, err, "completing transaction");
> + goto fail;
> + }
> +
> + err = xenbus_switch_state(dev, XenbusStateInitWait);
> + if (err)
> + goto fail;
> +
> + /* This kicks hotplug scripts, so do it immediately. */
> + backend_create_xenvif(be);
> +
> + return 0;
> +
> +abort_transaction:
> + xenbus_transaction_end(xbt, 1);
> + xenbus_dev_fatal(dev, err, "%s", message);
> +fail:
> + pr_debug("failed");
> + netback_remove(dev);
> + return err;
> +}
> +
> +
> +/*
> + * Handle the creation of the hotplug script environment. We add the script
> + * and vif variables to the environment, for the benefit of the vif-* hotplug
> + * scripts.
> + */
> +static int netback_uevent(struct xenbus_device *xdev,
> + struct kobj_uevent_env *env)
> +{
> + struct backend_info *be = dev_get_drvdata(&xdev->dev);
> + char *val;
> +
> + val = xenbus_read(XBT_NIL, xdev->nodename, "script", NULL);
> + if (IS_ERR(val)) {
> + int err = PTR_ERR(val);
> + xenbus_dev_fatal(xdev, err, "reading script");
> + return err;
> + } else {
> + if (add_uevent_var(env, "script=%s", val)) {
> + kfree(val);
> + return -ENOMEM;
> + }
> + kfree(val);
> + }
> +
> + if (!be || !be->vif)
> + return 0;
> +
> + return add_uevent_var(env, "vif=%s", be->vif->dev->name);
> +}
> +
> +
> +static void backend_create_xenvif(struct backend_info *be)
> +{
> + int err;
> + long handle;
> + struct xenbus_device *dev = be->dev;
> +
> + if (be->vif != NULL)
> + return;
> +
> + err = xenbus_scanf(XBT_NIL, dev->nodename, "handle", "%li", &handle);
> + if (err != 1) {
> + xenbus_dev_fatal(dev, err, "reading handle");
> + return;
> + }
> +
> + be->vif = xenvif_alloc(&dev->dev, dev->otherend_id, handle);
> + if (IS_ERR(be->vif)) {
> + err = PTR_ERR(be->vif);
> + be->vif = NULL;
> + xenbus_dev_fatal(dev, err, "creating interface");
> + return;
> + }
> +
> + kobject_uevent(&dev->dev.kobj, KOBJ_ONLINE);
> +}
> +
> +
> +static void disconnect_backend(struct xenbus_device *dev)
> +{
> + struct backend_info *be = dev_get_drvdata(&dev->dev);
> +
> + if (be->vif) {
> + xenbus_rm(XBT_NIL, dev->nodename, "hotplug-status");
> + xenvif_disconnect(be->vif);
> + be->vif = NULL;
> + }
> +}
> +
> +/**
> + * Callback received when the frontend's state changes.
> + */
> +static void frontend_changed(struct xenbus_device *dev,
> + enum xenbus_state frontend_state)
> +{
> + struct backend_info *be = dev_get_drvdata(&dev->dev);
> +
> + pr_debug("frontend state %s", xenbus_strstate(frontend_state));
> +
> + be->frontend_state = frontend_state;
> +
> + switch (frontend_state) {
> + case XenbusStateInitialising:
> + if (dev->state == XenbusStateClosed) {
> + printk(KERN_INFO "%s: %s: prepare for reconnect\n",
> + __func__, dev->nodename);
> + xenbus_switch_state(dev, XenbusStateInitWait);
> + }
> + break;
> +
> + case XenbusStateInitialised:
> + break;
> +
> + case XenbusStateConnected:
> + if (dev->state == XenbusStateConnected)
> + break;
> + backend_create_xenvif(be);
> + if (be->vif)
> + connect(be);
> + break;
> +
> + case XenbusStateClosing:
> + if (be->vif)
> + kobject_uevent(&dev->dev.kobj, KOBJ_OFFLINE);
> + disconnect_backend(dev);
> + xenbus_switch_state(dev, XenbusStateClosing);
> + break;
> +
> + case XenbusStateClosed:
> + xenbus_switch_state(dev, XenbusStateClosed);
> + if (xenbus_dev_is_online(dev))
> + break;
> + /* fall through if not online */
> + case XenbusStateUnknown:
> + device_unregister(&dev->dev);
> + break;
> +
> + default:
> + xenbus_dev_fatal(dev, -EINVAL, "saw state %d at frontend",
> + frontend_state);
> + break;
> + }
> +}
> +
> +
> +static void xen_net_read_rate(struct xenbus_device *dev,
> + unsigned long *bytes, unsigned long *usec)
> +{
> + char *s, *e;
> + unsigned long b, u;
> + char *ratestr;
> +
> + /* Default to unlimited bandwidth. */
> + *bytes = ~0UL;
> + *usec = 0;
> +
> + ratestr = xenbus_read(XBT_NIL, dev->nodename, "rate", NULL);
> + if (IS_ERR(ratestr))
> + return;
> +
> + s = ratestr;
> + b = simple_strtoul(s, &e, 10);
> + if ((s == e) || (*e != ','))
> + goto fail;
> +
> + s = e + 1;
> + u = simple_strtoul(s, &e, 10);
> + if ((s == e) || (*e != '\0'))
> + goto fail;
> +
> + *bytes = b;
> + *usec = u;
> +
> + kfree(ratestr);
> + return;
> +
> + fail:
> + pr_warn("Failed to parse network rate limit. Traffic unlimited.\n");
> + kfree(ratestr);
> +}
> +
> +static int xen_net_read_mac(struct xenbus_device *dev, u8 mac[])
> +{
> + char *s, *e, *macstr;
> + int i;
> +
> + macstr = s = xenbus_read(XBT_NIL, dev->nodename, "mac", NULL);
> + if (IS_ERR(macstr))
> + return PTR_ERR(macstr);
> +
> + for (i = 0; i < ETH_ALEN; i++) {
> + mac[i] = simple_strtoul(s, &e, 16);
> + if ((s == e) || (*e != ((i == ETH_ALEN-1) ? '\0' : ':'))) {
> + kfree(macstr);
> + return -ENOENT;
> + }
> + s = e+1;
> + }
> +
> + kfree(macstr);
> + return 0;
> +}
> +
> +static void unregister_hotplug_status_watch(struct backend_info *be)
> +{
> + if (be->have_hotplug_status_watch) {
> + unregister_xenbus_watch(&be->hotplug_status_watch);
> + kfree(be->hotplug_status_watch.node);
> + }
> + be->have_hotplug_status_watch = 0;
> +}
> +
> +static void hotplug_status_changed(struct xenbus_watch *watch,
> + const char **vec,
> + unsigned int vec_size)
> +{
> + struct backend_info *be = container_of(watch,
> + struct backend_info,
> + hotplug_status_watch);
> + char *str;
> + unsigned int len;
> +
> + str = xenbus_read(XBT_NIL, be->dev->nodename, "hotplug-status", &len);
> + if (IS_ERR(str))
> + return;
> + if (len == sizeof("connected")-1 && !memcmp(str, "connected", len)) {
> + xenbus_switch_state(be->dev, XenbusStateConnected);
> + /* Not interested in this watch anymore. */
> + unregister_hotplug_status_watch(be);
> + }
> + kfree(str);
> +}
> +
> +static void connect(struct backend_info *be)
> +{
> + int err;
> + struct xenbus_device *dev = be->dev;
> +
> + err = connect_rings(be);
> + if (err)
> + return;
> +
> + err = xen_net_read_mac(dev, be->vif->fe_dev_addr);
> + if (err) {
> + xenbus_dev_fatal(dev, err, "parsing %s/mac", dev->nodename);
> + return;
> + }
> +
> + xen_net_read_rate(dev, &be->vif->credit_bytes,
> + &be->vif->credit_usec);
> + be->vif->remaining_credit = be->vif->credit_bytes;
> +
> + unregister_hotplug_status_watch(be);
> + err = xenbus_watch_pathfmt(dev, &be->hotplug_status_watch,
> + hotplug_status_changed,
> + "%s/%s", dev->nodename, "hotplug-status");
> + if (err) {
> + /* Switch now, since we can't do a watch. */
> + xenbus_switch_state(dev, XenbusStateConnected);
> + } else {
> + be->have_hotplug_status_watch = 1;
> + }
> +
> + netif_wake_queue(be->vif->dev);
> +}
> +
> +
> +static int connect_rings(struct backend_info *be)
> +{
> + struct xenvif *vif = be->vif;
> + struct xenbus_device *dev = be->dev;
> + unsigned long tx_ring_ref, rx_ring_ref;
> + unsigned int evtchn, rx_copy;
> + int err;
> + int val;
> +
> + err = xenbus_gather(XBT_NIL, dev->otherend,
> + "tx-ring-ref", "%lu", &tx_ring_ref,
> + "rx-ring-ref", "%lu", &rx_ring_ref,
> + "event-channel", "%u", &evtchn, NULL);
> + if (err) {
> + xenbus_dev_fatal(dev, err,
> + "reading %s/ring-ref and event-channel",
> + dev->otherend);
> + return err;
> + }
> +
> + err = xenbus_scanf(XBT_NIL, dev->otherend, "request-rx-copy", "%u",
> + &rx_copy);
> + if (err == -ENOENT) {
> + err = 0;
> + rx_copy = 0;
> + }
> + if (err < 0) {
> + xenbus_dev_fatal(dev, err, "reading %s/request-rx-copy",
> + dev->otherend);
> + return err;
> + }
> + if (!rx_copy)
> + return -EOPNOTSUPP;
> +
> + if (vif->dev->tx_queue_len != 0) {
> + if (xenbus_scanf(XBT_NIL, dev->otherend,
> + "feature-rx-notify", "%d", &val) < 0)
> + val = 0;
> + if (val)
> + vif->can_queue = 1;
> + else
> + /* Must be non-zero for pfifo_fast to work. */
> + vif->dev->tx_queue_len = 1;
> + }
> +
> + if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-sg",
> + "%d", &val) < 0)
> + val = 0;
> + vif->can_sg = !!val;
> +
> + if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-gso-tcpv4",
> + "%d", &val) < 0)
> + val = 0;
> + vif->gso = !!val;
> +
> + if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-gso-tcpv4-prefix",
> + "%d", &val) < 0)
> + val = 0;
> + vif->gso_prefix = !!val;
> +
> + if (xenbus_scanf(XBT_NIL, dev->otherend, "feature-no-csum-offload",
> + "%d", &val) < 0)
> + val = 0;
> + vif->csum = !val;
Would it make sense to have a URL link or a short explanation of what each
feature provides?
> +
> + /* Map the shared frame, irq etc. */
> + err = xenvif_connect(vif, tx_ring_ref, rx_ring_ref, evtchn);
> + if (err) {
> + xenbus_dev_fatal(dev, err,
> + "mapping shared-frames %lu/%lu port %u",
> + tx_ring_ref, rx_ring_ref, evtchn);
> + return err;
> + }
> + return 0;
> +}
> +
> +
> +/* ** Driver Registration ** */
> +
> +
> +static const struct xenbus_device_id netback_ids[] = {
> + { "vif" },
> + { "" }
> +};
> +
> +
> +static struct xenbus_driver netback = {
> + .name = "vif",
> + .owner = THIS_MODULE,
> + .ids = netback_ids,
> + .probe = netback_probe,
> + .remove = netback_remove,
> + .uevent = netback_uevent,
> + .otherend_changed = frontend_changed,
> +};
> +
> +int xenvif_xenbus_init(void)
> +{
> + return xenbus_register_backend(&netback);
> +}
> diff --git a/drivers/net/xen-netfront.c b/drivers/net/xen-netfront.c
> index 458bb57..cc23d42 100644
> --- a/drivers/net/xen-netfront.c
> +++ b/drivers/net/xen-netfront.c
> @@ -356,7 +356,7 @@ static void xennet_tx_buf_gc(struct net_device *dev)
> struct xen_netif_tx_response *txrsp;
>
> txrsp = RING_GET_RESPONSE(&np->tx, cons);
> - if (txrsp->status == NETIF_RSP_NULL)
> + if (txrsp->status == XEN_NETIF_RSP_NULL)
> continue;
>
> id = txrsp->id;
> @@ -413,7 +413,7 @@ static void xennet_make_frags(struct sk_buff *skb, struct net_device *dev,
> larger than a page), split it it into page-sized chunks. */
> while (len > PAGE_SIZE - offset) {
> tx->size = PAGE_SIZE - offset;
> - tx->flags |= NETTXF_more_data;
> + tx->flags |= XEN_NETTXF_more_data;
> len -= tx->size;
> data += tx->size;
> offset = 0;
> @@ -439,7 +439,7 @@ static void xennet_make_frags(struct sk_buff *skb, struct net_device *dev,
> for (i = 0; i < frags; i++) {
> skb_frag_t *frag = skb_shinfo(skb)->frags + i;
>
> - tx->flags |= NETTXF_more_data;
> + tx->flags |= XEN_NETTXF_more_data;
>
> id = get_id_from_freelist(&np->tx_skb_freelist, np->tx_skbs);
> np->tx_skbs[id].skb = skb_get(skb);
> @@ -514,10 +514,10 @@ static int xennet_start_xmit(struct sk_buff *skb, struct net_device *dev)
> tx->flags = 0;
> if (skb->ip_summed == CHECKSUM_PARTIAL)
> /* local packet? */
> - tx->flags |= NETTXF_csum_blank | NETTXF_data_validated;
> + tx->flags |= XEN_NETTXF_csum_blank | XEN_NETTXF_data_validated;
> else if (skb->ip_summed == CHECKSUM_UNNECESSARY)
> /* remote but checksummed. */
> - tx->flags |= NETTXF_data_validated;
> + tx->flags |= XEN_NETTXF_data_validated;
>
> if (skb_shinfo(skb)->gso_size) {
> struct xen_netif_extra_info *gso;
> @@ -528,7 +528,7 @@ static int xennet_start_xmit(struct sk_buff *skb, struct net_device *dev)
> if (extra)
> extra->flags |= XEN_NETIF_EXTRA_FLAG_MORE;
> else
> - tx->flags |= NETTXF_extra_info;
> + tx->flags |= XEN_NETTXF_extra_info;
>
> gso->u.gso.size = skb_shinfo(skb)->gso_size;
> gso->u.gso.type = XEN_NETIF_GSO_TYPE_TCPV4;
> @@ -648,7 +648,7 @@ static int xennet_get_responses(struct netfront_info *np,
> int err = 0;
> unsigned long ret;
>
> - if (rx->flags & NETRXF_extra_info) {
> + if (rx->flags & XEN_NETRXF_extra_info) {
> err = xennet_get_extras(np, extras, rp);
> cons = np->rx.rsp_cons;
> }
> @@ -685,7 +685,7 @@ static int xennet_get_responses(struct netfront_info *np,
> __skb_queue_tail(list, skb);
>
> next:
> - if (!(rx->flags & NETRXF_more_data))
> + if (!(rx->flags & XEN_NETRXF_more_data))
> break;
>
> if (cons + frags == rp) {
> @@ -950,9 +950,9 @@ err:
> skb->truesize += skb->data_len - (RX_COPY_THRESHOLD - len);
> skb->len += skb->data_len;
>
> - if (rx->flags & NETRXF_csum_blank)
> + if (rx->flags & XEN_NETRXF_csum_blank)
> skb->ip_summed = CHECKSUM_PARTIAL;
> - else if (rx->flags & NETRXF_data_validated)
> + else if (rx->flags & XEN_NETRXF_data_validated)
> skb->ip_summed = CHECKSUM_UNNECESSARY;
>
> __skb_queue_tail(&rxq, skb);
> diff --git a/include/xen/interface/io/netif.h b/include/xen/interface/io/netif.h
> index 518481c..cb94668 100644
> --- a/include/xen/interface/io/netif.h
> +++ b/include/xen/interface/io/netif.h
> @@ -22,50 +22,50 @@
>
> /*
> * This is the 'wire' format for packets:
> - * Request 1: netif_tx_request -- NETTXF_* (any flags)
> - * [Request 2: netif_tx_extra] (only if request 1 has NETTXF_extra_info)
> - * [Request 3: netif_tx_extra] (only if request 2 has XEN_NETIF_EXTRA_MORE)
> - * Request 4: netif_tx_request -- NETTXF_more_data
> - * Request 5: netif_tx_request -- NETTXF_more_data
> + * Request 1: xen_netif_tx_request -- XEN_NETTXF_* (any flags)
> + * [Request 2: xen_netif_extra_info] (only if request 1 has XEN_NETTXF_extra_info)
> + * [Request 3: xen_netif_extra_info] (only if request 2 has XEN_NETIF_EXTRA_MORE)
> + * Request 4: xen_netif_tx_request -- XEN_NETTXF_more_data
> + * Request 5: xen_netif_tx_request -- XEN_NETTXF_more_data
> * ...
> - * Request N: netif_tx_request -- 0
> + * Request N: xen_netif_tx_request -- 0
> */
>
> /* Protocol checksum field is blank in the packet (hardware offload)? */
> -#define _NETTXF_csum_blank (0)
> -#define NETTXF_csum_blank (1U<<_NETTXF_csum_blank)
> +#define _XEN_NETTXF_csum_blank (0)
> +#define XEN_NETTXF_csum_blank (1U<<_XEN_NETTXF_csum_blank)
>
> /* Packet data has been validated against protocol checksum. */
> -#define _NETTXF_data_validated (1)
> -#define NETTXF_data_validated (1U<<_NETTXF_data_validated)
> +#define _XEN_NETTXF_data_validated (1)
> +#define XEN_NETTXF_data_validated (1U<<_XEN_NETTXF_data_validated)
>
> /* Packet continues in the next request descriptor. */
> -#define _NETTXF_more_data (2)
> -#define NETTXF_more_data (1U<<_NETTXF_more_data)
> +#define _XEN_NETTXF_more_data (2)
> +#define XEN_NETTXF_more_data (1U<<_XEN_NETTXF_more_data)
>
> /* Packet to be followed by extra descriptor(s). */
> -#define _NETTXF_extra_info (3)
> -#define NETTXF_extra_info (1U<<_NETTXF_extra_info)
> +#define _XEN_NETTXF_extra_info (3)
> +#define XEN_NETTXF_extra_info (1U<<_XEN_NETTXF_extra_info)
>
> struct xen_netif_tx_request {
> grant_ref_t gref; /* Reference to buffer page */
> uint16_t offset; /* Offset within buffer page */
> - uint16_t flags; /* NETTXF_* */
> + uint16_t flags; /* XEN_NETTXF_* */
> uint16_t id; /* Echoed in response message. */
> uint16_t size; /* Packet size in bytes. */
> };
>
> -/* Types of netif_extra_info descriptors. */
> -#define XEN_NETIF_EXTRA_TYPE_NONE (0) /* Never used - invalid */
> -#define XEN_NETIF_EXTRA_TYPE_GSO (1) /* u.gso */
> -#define XEN_NETIF_EXTRA_TYPE_MAX (2)
> +/* Types of xen_netif_extra_info descriptors. */
> +#define XEN_NETIF_EXTRA_TYPE_NONE (0) /* Never used - invalid */
> +#define XEN_NETIF_EXTRA_TYPE_GSO (1) /* u.gso */
> +#define XEN_NETIF_EXTRA_TYPE_MAX (2)
>
> -/* netif_extra_info flags. */
> -#define _XEN_NETIF_EXTRA_FLAG_MORE (0)
> -#define XEN_NETIF_EXTRA_FLAG_MORE (1U<<_XEN_NETIF_EXTRA_FLAG_MORE)
> +/* xen_netif_extra_info flags. */
> +#define _XEN_NETIF_EXTRA_FLAG_MORE (0)
> +#define XEN_NETIF_EXTRA_FLAG_MORE (1U<<_XEN_NETIF_EXTRA_FLAG_MORE)
>
> /* GSO types - only TCPv4 currently supported. */
> -#define XEN_NETIF_GSO_TYPE_TCPV4 (1)
> +#define XEN_NETIF_GSO_TYPE_TCPV4 (1)
>
> /*
> * This structure needs to fit within both netif_tx_request and
> @@ -107,7 +107,7 @@ struct xen_netif_extra_info {
>
> struct xen_netif_tx_response {
> uint16_t id;
> - int16_t status; /* NETIF_RSP_* */
> + int16_t status; /* XEN_NETIF_RSP_* */
> };
>
> struct xen_netif_rx_request {
> @@ -116,25 +116,29 @@ struct xen_netif_rx_request {
> };
>
> /* Packet data has been validated against protocol checksum. */
> -#define _NETRXF_data_validated (0)
> -#define NETRXF_data_validated (1U<<_NETRXF_data_validated)
> +#define _XEN_NETRXF_data_validated (0)
> +#define XEN_NETRXF_data_validated (1U<<_XEN_NETRXF_data_validated)
>
> /* Protocol checksum field is blank in the packet (hardware offload)? */
> -#define _NETRXF_csum_blank (1)
> -#define NETRXF_csum_blank (1U<<_NETRXF_csum_blank)
> +#define _XEN_NETRXF_csum_blank (1)
> +#define XEN_NETRXF_csum_blank (1U<<_XEN_NETRXF_csum_blank)
>
> /* Packet continues in the next request descriptor. */
> -#define _NETRXF_more_data (2)
> -#define NETRXF_more_data (1U<<_NETRXF_more_data)
> +#define _XEN_NETRXF_more_data (2)
> +#define XEN_NETRXF_more_data (1U<<_XEN_NETRXF_more_data)
>
> /* Packet to be followed by extra descriptor(s). */
> -#define _NETRXF_extra_info (3)
> -#define NETRXF_extra_info (1U<<_NETRXF_extra_info)
> +#define _XEN_NETRXF_extra_info (3)
> +#define XEN_NETRXF_extra_info (1U<<_XEN_NETRXF_extra_info)
> +
> +/* GSO Prefix descriptor. */
> +#define _XEN_NETRXF_gso_prefix (4)
> +#define XEN_NETRXF_gso_prefix (1U<<_XEN_NETRXF_gso_prefix)
>
> struct xen_netif_rx_response {
> uint16_t id;
> uint16_t offset; /* Offset in page of start of received packet */
> - uint16_t flags; /* NETRXF_* */
> + uint16_t flags; /* XEN_NETRXF_* */
> int16_t status; /* -ve: BLKIF_RSP_* ; +ve: Rx'ed pkt size. */
> };
>
> @@ -149,10 +153,10 @@ DEFINE_RING_TYPES(xen_netif_rx,
> struct xen_netif_rx_request,
> struct xen_netif_rx_response);
>
> -#define NETIF_RSP_DROPPED -2
> -#define NETIF_RSP_ERROR -1
> -#define NETIF_RSP_OKAY 0
> -/* No response: used for auxiliary requests (e.g., netif_tx_extra). */
> -#define NETIF_RSP_NULL 1
> +#define XEN_NETIF_RSP_DROPPED -2
> +#define XEN_NETIF_RSP_ERROR -1
> +#define XEN_NETIF_RSP_OKAY 0
> +/* No response: used for auxiliary requests (e.g., xen_netif_extra_info). */
> +#define XEN_NETIF_RSP_NULL 1
>
> #endif
>
>
^ permalink raw reply
* Re: [PATCH] mac80211: Add power to debugfs.
From: Ben Greear @ 2011-02-15 21:08 UTC (permalink / raw)
To: greearb; +Cc: netdev
In-Reply-To: <1297803874-9109-1-git-send-email-greearb@candelatech.com>
On 02/15/2011 01:04 PM, greearb@candelatech.com wrote:
> From: Ben Greear<greearb@candelatech.com>
Bleh..wrong mailing list..sorry.
Ben
>
> Signed-off-by: Ben Greear<greearb@candelatech.com>
> ---
> :100644 100644 1f02e59... 51f0d78... M net/mac80211/debugfs.c
> net/mac80211/debugfs.c | 6 ++++++
> 1 files changed, 6 insertions(+), 0 deletions(-)
>
> diff --git a/net/mac80211/debugfs.c b/net/mac80211/debugfs.c
> index 1f02e59..51f0d78 100644
> --- a/net/mac80211/debugfs.c
> +++ b/net/mac80211/debugfs.c
> @@ -60,6 +60,10 @@ static const struct file_operations name## _ops = { \
> debugfs_create_file(#name, mode, phyd, local,&name## _ops);
>
>
> +DEBUGFS_READONLY_FILE(user_power, "%d",
> + local->user_power_level);
> +DEBUGFS_READONLY_FILE(power, "%d",
> + local->hw.conf.power_level);
> DEBUGFS_READONLY_FILE(frequency, "%d",
> local->hw.conf.channel->center_freq);
> DEBUGFS_READONLY_FILE(total_ps_buffered, "%d",
> @@ -391,6 +395,8 @@ void debugfs_hw_add(struct ieee80211_local *local)
> DEBUGFS_ADD(uapsd_queues);
> DEBUGFS_ADD(uapsd_max_sp_len);
> DEBUGFS_ADD(channel_type);
> + DEBUGFS_ADD(user_power);
> + DEBUGFS_ADD(power);
>
> statsd = debugfs_create_dir("statistics", phyd);
>
--
Ben Greear <greearb@candelatech.com>
Candela Technologies Inc http://www.candelatech.com
^ permalink raw reply
* [PATCH] mac80211: Add power to debugfs.
From: greearb @ 2011-02-15 21:04 UTC (permalink / raw)
To: netdev; +Cc: Ben Greear
From: Ben Greear <greearb@candelatech.com>
Signed-off-by: Ben Greear <greearb@candelatech.com>
---
:100644 100644 1f02e59... 51f0d78... M net/mac80211/debugfs.c
net/mac80211/debugfs.c | 6 ++++++
1 files changed, 6 insertions(+), 0 deletions(-)
diff --git a/net/mac80211/debugfs.c b/net/mac80211/debugfs.c
index 1f02e59..51f0d78 100644
--- a/net/mac80211/debugfs.c
+++ b/net/mac80211/debugfs.c
@@ -60,6 +60,10 @@ static const struct file_operations name## _ops = { \
debugfs_create_file(#name, mode, phyd, local, &name## _ops);
+DEBUGFS_READONLY_FILE(user_power, "%d",
+ local->user_power_level);
+DEBUGFS_READONLY_FILE(power, "%d",
+ local->hw.conf.power_level);
DEBUGFS_READONLY_FILE(frequency, "%d",
local->hw.conf.channel->center_freq);
DEBUGFS_READONLY_FILE(total_ps_buffered, "%d",
@@ -391,6 +395,8 @@ void debugfs_hw_add(struct ieee80211_local *local)
DEBUGFS_ADD(uapsd_queues);
DEBUGFS_ADD(uapsd_max_sp_len);
DEBUGFS_ADD(channel_type);
+ DEBUGFS_ADD(user_power);
+ DEBUGFS_ADD(power);
statsd = debugfs_create_dir("statistics", phyd);
--
1.7.2.3
^ 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