* [PATCH net-next RFC 2/5] xen-netback: Change TX path from grant copy to mapping
From: Zoltan Kiss @ 2013-10-30 0:50 UTC (permalink / raw)
To: ian.campbell, wei.liu2, xen-devel, netdev, linux-kernel,
jonathan.davies
Cc: Zoltan Kiss
In-Reply-To: <1383094220-14775-1-git-send-email-zoltan.kiss@citrix.com>
This patch changes the grant copy on the TX patch to grant mapping
Signed-off-by: Zoltan Kiss <zoltan.kiss@citrix.com>
---
drivers/net/xen-netback/interface.c | 39 +++++-
drivers/net/xen-netback/netback.c | 241 +++++++++++++----------------------
2 files changed, 126 insertions(+), 154 deletions(-)
diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
index f5c3c57..fb16ede 100644
--- a/drivers/net/xen-netback/interface.c
+++ b/drivers/net/xen-netback/interface.c
@@ -336,8 +336,20 @@ struct xenvif *xenvif_alloc(struct device *parent, domid_t domid,
vif->pending_prod = MAX_PENDING_REQS;
for (i = 0; i < MAX_PENDING_REQS; i++)
vif->pending_ring[i] = i;
- for (i = 0; i < MAX_PENDING_REQS; i++)
- vif->mmap_pages[i] = NULL;
+ err = alloc_xenballooned_pages(MAX_PENDING_REQS,
+ vif->mmap_pages,
+ false);
+ if (err) {
+ netdev_err(dev, "Could not reserve mmap_pages\n");
+ return NULL;
+ }
+ for (i = 0; i < MAX_PENDING_REQS; i++) {
+ vif->pending_tx_info[i].callback_struct = (struct ubuf_info)
+ { .callback = xenvif_zerocopy_callback,
+ .ctx = NULL,
+ .desc = i };
+ vif->grant_tx_handle[i] = NETBACK_INVALID_HANDLE;
+ }
/*
* Initialise a dummy MAC address. We choose the numerically
@@ -481,11 +493,34 @@ void xenvif_disconnect(struct xenvif *vif)
void xenvif_free(struct xenvif *vif)
{
+ int i;
+
netif_napi_del(&vif->napi);
unregister_netdev(vif->dev);
free_netdev(vif->dev);
+ /* FIXME: This is a workaround for 2 problems:
+ * - the guest sent a packet just before teardown, and it is still not
+ * delivered
+ * - the stack forgot to notify us that we can give back a slot
+ * For the first problem we shouldn't do this, as the skb might still
+ * access that page. I will come up with a more robust solution later.
+ * The second is definitely a bug, it leaks a slot from the ring
+ * forever.
+ * Classic kernel patchset has delayed copy for that, we might want to
+ * reuse that?
+ */
+ for (i = 0; i < MAX_PENDING_REQS; ++i) {
+ if (vif->grant_tx_handle[i] != NETBACK_INVALID_HANDLE) {
+ netdev_err(vif->dev,
+ "Page still granted! Index: %x\n", i);
+ xenvif_idx_unmap(vif, i);
+ }
+ }
+
+ free_xenballooned_pages(MAX_PENDING_REQS, vif->mmap_pages);
+
module_put(THIS_MODULE);
}
diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c
index 10470dc..e544e9f 100644
--- a/drivers/net/xen-netback/netback.c
+++ b/drivers/net/xen-netback/netback.c
@@ -883,10 +883,10 @@ static inline void xenvif_tx_create_gop(struct xenvif *vif, u16 pending_idx,
}
-static struct gnttab_copy *xenvif_get_requests(struct xenvif *vif,
+static struct gnttab_map_grant_ref *xenvif_get_requests(struct xenvif *vif,
struct sk_buff *skb,
struct xen_netif_tx_request *txp,
- struct gnttab_copy *gop)
+ struct gnttab_map_grant_ref *gop)
{
struct skb_shared_info *shinfo = skb_shinfo(skb);
skb_frag_t *frags = shinfo->frags;
@@ -907,83 +907,12 @@ static struct gnttab_copy *xenvif_get_requests(struct xenvif *vif,
/* Skip first skb fragment if it is on same page as header fragment. */
start = (frag_get_pending_idx(&shinfo->frags[0]) == pending_idx);
- /* Coalesce tx requests, at this point the packet passed in
- * should be <= 64K. Any packets larger than 64K have been
- * handled in xenvif_count_requests().
- */
- for (shinfo->nr_frags = slot = start; slot < nr_slots;
- shinfo->nr_frags++) {
- struct pending_tx_info *pending_tx_info =
- vif->pending_tx_info;
-
- page = alloc_page(GFP_ATOMIC|__GFP_COLD);
- if (!page)
- goto err;
-
- dst_offset = 0;
- first = NULL;
- while (dst_offset < PAGE_SIZE && slot < nr_slots) {
- gop->flags = GNTCOPY_source_gref;
-
- gop->source.u.ref = txp->gref;
- gop->source.domid = vif->domid;
- gop->source.offset = txp->offset;
-
- gop->dest.domid = DOMID_SELF;
-
- gop->dest.offset = dst_offset;
- gop->dest.u.gmfn = virt_to_mfn(page_address(page));
-
- if (dst_offset + txp->size > PAGE_SIZE) {
- /* This page can only merge a portion
- * of tx request. Do not increment any
- * pointer / counter here. The txp
- * will be dealt with in future
- * rounds, eventually hitting the
- * `else` branch.
- */
- gop->len = PAGE_SIZE - dst_offset;
- txp->offset += gop->len;
- txp->size -= gop->len;
- dst_offset += gop->len; /* quit loop */
- } else {
- /* This tx request can be merged in the page */
- gop->len = txp->size;
- dst_offset += gop->len;
-
+ for (shinfo->nr_frags = start; shinfo->nr_frags < nr_slots;
+ shinfo->nr_frags++, txp++, gop++) {
index = pending_index(vif->pending_cons++);
-
pending_idx = vif->pending_ring[index];
-
- memcpy(&pending_tx_info[pending_idx].req, txp,
- sizeof(*txp));
-
- /* Poison these fields, corresponding
- * fields for head tx req will be set
- * to correct values after the loop.
- */
- vif->mmap_pages[pending_idx] = (void *)(~0UL);
- pending_tx_info[pending_idx].head =
- INVALID_PENDING_RING_IDX;
-
- if (!first) {
- first = &pending_tx_info[pending_idx];
- start_idx = index;
- head_idx = pending_idx;
- }
-
- txp++;
- slot++;
- }
-
- gop++;
- }
-
- first->req.offset = 0;
- first->req.size = dst_offset;
- first->head = start_idx;
- vif->mmap_pages[head_idx] = page;
- frag_set_pending_idx(&frags[shinfo->nr_frags], head_idx);
+ xenvif_tx_create_gop(vif, pending_idx, txp, gop);
+ frag_set_pending_idx(&frags[shinfo->nr_frags], pending_idx);
}
BUG_ON(shinfo->nr_frags > MAX_SKB_FRAGS);
@@ -1005,9 +934,9 @@ err:
static int xenvif_tx_check_gop(struct xenvif *vif,
struct sk_buff *skb,
- struct gnttab_copy **gopp)
+ struct gnttab_map_grant_ref **gopp)
{
- struct gnttab_copy *gop = *gopp;
+ struct gnttab_map_grant_ref *gop = *gopp;
u16 pending_idx = *((u16 *)skb->data);
struct skb_shared_info *shinfo = skb_shinfo(skb);
struct pending_tx_info *tx_info;
@@ -1019,6 +948,16 @@ static int xenvif_tx_check_gop(struct xenvif *vif,
err = gop->status;
if (unlikely(err))
xenvif_idx_release(vif, pending_idx, XEN_NETIF_RSP_ERROR);
+ else {
+ if (vif->grant_tx_handle[pending_idx] !=
+ NETBACK_INVALID_HANDLE) {
+ netdev_err(vif->dev,
+ "Stale mapped handle! pending_idx %x handle %x\n",
+ pending_idx, vif->grant_tx_handle[pending_idx]);
+ xenvif_fatal_tx_err(vif);
+ }
+ vif->grant_tx_handle[pending_idx] = gop->handle;
+ }
/* Skip first skb fragment if it is on same page as header fragment. */
start = (frag_get_pending_idx(&shinfo->frags[0]) == pending_idx);
@@ -1032,18 +971,24 @@ static int xenvif_tx_check_gop(struct xenvif *vif,
head = tx_info->head;
/* Check error status: if okay then remember grant handle. */
- do {
newerr = (++gop)->status;
- if (newerr)
- break;
- peek = vif->pending_ring[pending_index(++head)];
- } while (!pending_tx_is_head(vif, peek));
if (likely(!newerr)) {
+ if (vif->grant_tx_handle[pending_idx] !=
+ NETBACK_INVALID_HANDLE) {
+ netdev_err(vif->dev,
+ "Stale mapped handle! pending_idx %x handle %x\n",
+ pending_idx,
+ vif->grant_tx_handle[pending_idx]);
+ xenvif_fatal_tx_err(vif);
+ }
+ vif->grant_tx_handle[pending_idx] = gop->handle;
/* Had a previous error? Invalidate this fragment. */
- if (unlikely(err))
+ if (unlikely(err)) {
+ xenvif_idx_unmap(vif, pending_idx);
xenvif_idx_release(vif, pending_idx,
XEN_NETIF_RSP_OKAY);
+ }
continue;
}
@@ -1056,9 +1001,11 @@ static int xenvif_tx_check_gop(struct xenvif *vif,
/* First error: invalidate header and preceding fragments. */
pending_idx = *((u16 *)skb->data);
+ xenvif_idx_unmap(vif, pending_idx);
xenvif_idx_release(vif, pending_idx, XEN_NETIF_RSP_OKAY);
for (j = start; j < i; j++) {
pending_idx = frag_get_pending_idx(&shinfo->frags[j]);
+ xenvif_idx_unmap(vif, pending_idx);
xenvif_idx_release(vif, pending_idx,
XEN_NETIF_RSP_OKAY);
}
@@ -1071,7 +1018,8 @@ static int xenvif_tx_check_gop(struct xenvif *vif,
return err;
}
-static void xenvif_fill_frags(struct xenvif *vif, struct sk_buff *skb)
+static void xenvif_fill_frags(struct xenvif *vif, struct sk_buff *skb,
+ u16 prev_pending_idx)
{
struct skb_shared_info *shinfo = skb_shinfo(skb);
int nr_frags = shinfo->nr_frags;
@@ -1085,6 +1033,15 @@ static void xenvif_fill_frags(struct xenvif *vif, struct sk_buff *skb)
pending_idx = frag_get_pending_idx(frag);
+ /* If this is not the first frag, chain it to the previous*/
+ vif->pending_tx_info[pending_idx].callback_struct.ctx = NULL;
+ if (pending_idx != prev_pending_idx) {
+ vif->pending_tx_info[prev_pending_idx].callback_struct.ctx =
+ &(vif->pending_tx_info[pending_idx].callback_struct);
+ prev_pending_idx = pending_idx;
+ }
+
+
txp = &vif->pending_tx_info[pending_idx].req;
page = virt_to_page(idx_to_kaddr(vif, pending_idx));
__skb_fill_page_desc(skb, i, page, txp->offset, txp->size);
@@ -1092,10 +1049,15 @@ static void xenvif_fill_frags(struct xenvif *vif, struct sk_buff *skb)
skb->data_len += txp->size;
skb->truesize += txp->size;
- /* Take an extra reference to offset xenvif_idx_release */
+ /* Take an extra reference to offset network stack's put_page */
get_page(vif->mmap_pages[pending_idx]);
- xenvif_idx_release(vif, pending_idx, XEN_NETIF_RSP_OKAY);
}
+ /* FIXME: __skb_fill_page_desc set this to true because page->pfmemalloc
+ * overlaps with "index", and "mapping" is not set. I think mapping
+ * should be set. If delivered to local stack, it would drop this
+ * skb in sk_filter unless the socket has the right to use it.
+ */
+ skb->pfmemalloc = false;
}
static int xenvif_get_extras(struct xenvif *vif,
@@ -1426,7 +1388,7 @@ static bool tx_credit_exceeded(struct xenvif *vif, unsigned size)
static unsigned xenvif_tx_build_gops(struct xenvif *vif)
{
- struct gnttab_copy *gop = vif->tx_copy_ops, *request_gop;
+ struct gnttab_map_grant_ref *gop = vif->tx_map_ops, *request_gop;
struct sk_buff *skb;
int ret;
@@ -1533,30 +1495,10 @@ static unsigned xenvif_tx_build_gops(struct xenvif *vif)
}
}
- /* XXX could copy straight to head */
- page = xenvif_alloc_page(vif, pending_idx);
- if (!page) {
- kfree_skb(skb);
- xenvif_tx_err(vif, &txreq, idx);
- break;
- }
-
- 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;
+ xenvif_tx_create_gop(vif, pending_idx, &txreq, gop);
gop++;
- memcpy(&vif->pending_tx_info[pending_idx].req,
- &txreq, sizeof(txreq));
- vif->pending_tx_info[pending_idx].head = index;
*((u16 *)skb->data) = pending_idx;
__skb_put(skb, data_len);
@@ -1585,17 +1527,17 @@ static unsigned xenvif_tx_build_gops(struct xenvif *vif)
vif->tx.req_cons = idx;
- if ((gop-vif->tx_copy_ops) >= ARRAY_SIZE(vif->tx_copy_ops))
+ if ((gop-vif->tx_map_ops) >= ARRAY_SIZE(vif->tx_map_ops))
break;
}
- return gop - vif->tx_copy_ops;
+ return gop - vif->tx_map_ops;
}
static int xenvif_tx_submit(struct xenvif *vif, int budget)
{
- struct gnttab_copy *gop = vif->tx_copy_ops;
+ struct gnttab_map_grant_ref *gop = vif->tx_map_ops;
struct sk_buff *skb;
int work_done = 0;
@@ -1620,14 +1562,25 @@ static int xenvif_tx_submit(struct xenvif *vif, int budget)
memcpy(skb->data,
(void *)(idx_to_kaddr(vif, pending_idx)|txp->offset),
data_len);
+ vif->pending_tx_info[pending_idx].callback_struct.ctx = NULL;
if (data_len < txp->size) {
/* Append the packet payload as a fragment. */
txp->offset += data_len;
txp->size -= data_len;
- } else {
+ skb_shinfo(skb)->destructor_arg =
+ &vif->pending_tx_info[pending_idx].callback_struct;
+ } else if (!skb_shinfo(skb)->nr_frags) {
/* Schedule a response immediately. */
+ skb_shinfo(skb)->destructor_arg = NULL;
+ xenvif_idx_unmap(vif, pending_idx);
xenvif_idx_release(vif, pending_idx,
XEN_NETIF_RSP_OKAY);
+ } else {
+ /* FIXME: first request fits linear space, I don't know
+ * if any guest would do that, but I think it's possible
+ */
+ skb_shinfo(skb)->destructor_arg =
+ &vif->pending_tx_info[pending_idx].callback_struct;
}
if (txp->flags & XEN_NETTXF_csum_blank)
@@ -1635,13 +1588,19 @@ static int xenvif_tx_submit(struct xenvif *vif, int budget)
else if (txp->flags & XEN_NETTXF_data_validated)
skb->ip_summed = CHECKSUM_UNNECESSARY;
- xenvif_fill_frags(vif, skb);
+ xenvif_fill_frags(vif, skb, pending_idx);
if (skb_is_nonlinear(skb) && skb_headlen(skb) < PKT_PROT_LEN) {
int target = min_t(int, skb->len, PKT_PROT_LEN);
__pskb_pull_tail(skb, target - skb_headlen(skb));
}
+ /* Set this flag after __pskb_pull_tail, as it can trigger
+ * skb_copy_ubufs, while we are still in control of the skb
+ */
+ if (skb_shinfo(skb)->destructor_arg)
+ skb_shinfo(skb)->tx_flags |= SKBTX_DEV_ZEROCOPY;
+
skb->dev = vif->dev;
skb->protocol = eth_type_trans(skb, skb->dev);
skb_reset_network_header(skb);
@@ -1770,17 +1729,25 @@ static inline void xenvif_tx_action_dealloc(struct xenvif *vif)
int xenvif_tx_action(struct xenvif *vif, int budget)
{
unsigned nr_gops;
- int work_done;
+ int work_done, ret;
if (unlikely(!tx_work_todo(vif)))
return 0;
+ xenvif_tx_action_dealloc(vif);
+
nr_gops = xenvif_tx_build_gops(vif);
if (nr_gops == 0)
return 0;
- gnttab_batch_copy(vif->tx_copy_ops, nr_gops);
+ if (nr_gops) {
+ ret = gnttab_map_refs(vif->tx_map_ops,
+ NULL,
+ vif->pages_to_gnt,
+ nr_gops);
+ BUG_ON(ret);
+ }
work_done = xenvif_tx_submit(vif, nr_gops);
@@ -1791,45 +1758,13 @@ static void xenvif_idx_release(struct xenvif *vif, u16 pending_idx,
u8 status)
{
struct pending_tx_info *pending_tx_info;
- pending_ring_idx_t head;
+ pending_ring_idx_t index;
u16 peek; /* peek into next tx request */
- BUG_ON(vif->mmap_pages[pending_idx] == (void *)(~0UL));
-
- /* Already complete? */
- if (vif->mmap_pages[pending_idx] == NULL)
- return;
-
- pending_tx_info = &vif->pending_tx_info[pending_idx];
-
- head = pending_tx_info->head;
-
- BUG_ON(!pending_tx_is_head(vif, head));
- BUG_ON(vif->pending_ring[pending_index(head)] != pending_idx);
-
- do {
- pending_ring_idx_t index;
- pending_ring_idx_t idx = pending_index(head);
- u16 info_idx = vif->pending_ring[idx];
-
- pending_tx_info = &vif->pending_tx_info[info_idx];
+ pending_tx_info = &vif->pending_tx_info[pending_idx];
make_tx_response(vif, &pending_tx_info->req, status);
-
- /* Setting any number other than
- * INVALID_PENDING_RING_IDX indicates this slot is
- * starting a new packet / ending a previous packet.
- */
- pending_tx_info->head = 0;
-
index = pending_index(vif->pending_prod++);
- vif->pending_ring[index] = vif->pending_ring[info_idx];
-
- peek = vif->pending_ring[pending_index(++head)];
-
- } while (!pending_tx_is_head(vif, peek));
-
- put_page(vif->mmap_pages[pending_idx]);
- vif->mmap_pages[pending_idx] = NULL;
+ vif->pending_ring[index] = pending_idx;
}
void xenvif_idx_unmap(struct xenvif *vif, u16 pending_idx)
@@ -1904,6 +1839,8 @@ static inline int rx_work_todo(struct xenvif *vif)
static inline int tx_work_todo(struct xenvif *vif)
{
+ if (vif->dealloc_cons != vif->dealloc_prod)
+ return 1;
if (likely(RING_HAS_UNCONSUMED_REQUESTS(&vif->tx)) &&
(nr_pending_reqs(vif) + XEN_NETBK_LEGACY_SLOTS_MAX
^ permalink raw reply related
* [PATCH net-next RFC 1/5] xen-netback: Introduce TX grant map definitions
From: Zoltan Kiss @ 2013-10-30 0:50 UTC (permalink / raw)
To: ian.campbell, wei.liu2, xen-devel, netdev, linux-kernel,
jonathan.davies
Cc: Zoltan Kiss
In-Reply-To: <1383094220-14775-1-git-send-email-zoltan.kiss@citrix.com>
This patch contains the new definitions necessary for grant mapping.
Signed-off-by: Zoltan Kiss <zoltan.kiss@citrix.com>
---
drivers/net/xen-netback/common.h | 22 ++++++
drivers/net/xen-netback/interface.c | 1 +
drivers/net/xen-netback/netback.c | 134 +++++++++++++++++++++++++++++++++++
3 files changed, 157 insertions(+)
diff --git a/drivers/net/xen-netback/common.h b/drivers/net/xen-netback/common.h
index 55b8dec..36a3fda 100644
--- a/drivers/net/xen-netback/common.h
+++ b/drivers/net/xen-netback/common.h
@@ -79,6 +79,11 @@ struct pending_tx_info {
* if it is head of one or more tx
* reqs
*/
+ /* callback data for released SKBs. The callback is always
+ * xenvif_zerocopy_callback, ctx points to the next fragment, desc
+ * contains the pending_idx
+ */
+ struct ubuf_info callback_struct;
};
#define XEN_NETIF_TX_RING_SIZE __CONST_RING_SIZE(xen_netif_tx, PAGE_SIZE)
@@ -101,6 +106,8 @@ struct xenvif_rx_meta {
#define MAX_PENDING_REQS 256
+#define NETBACK_INVALID_HANDLE -1
+
struct xenvif {
/* Unique identifier for this interface. */
domid_t domid;
@@ -119,13 +126,22 @@ struct xenvif {
pending_ring_idx_t pending_cons;
u16 pending_ring[MAX_PENDING_REQS];
struct pending_tx_info pending_tx_info[MAX_PENDING_REQS];
+ grant_handle_t grant_tx_handle[MAX_PENDING_REQS];
/* Coalescing tx requests before copying makes number of grant
* copy ops greater or equal to number of slots required. In
* worst case a tx request consumes 2 gnttab_copy.
*/
struct gnttab_copy tx_copy_ops[2*MAX_PENDING_REQS];
+ struct gnttab_unmap_grant_ref tx_unmap_ops[MAX_PENDING_REQS];
+ struct gnttab_map_grant_ref tx_map_ops[MAX_PENDING_REQS];
+ /* passed to gnttab_[un]map_refs with pages under (un)mapping */
+ struct page *pages_to_gnt[MAX_PENDING_REQS];
+ spinlock_t dealloc_lock;
+ pending_ring_idx_t dealloc_prod;
+ pending_ring_idx_t dealloc_cons;
+ u16 dealloc_ring[MAX_PENDING_REQS];
/* Use kthread for guest RX */
struct task_struct *task;
@@ -226,6 +242,12 @@ void xenvif_rx_action(struct xenvif *vif);
int xenvif_kthread(void *data);
+/* Callback from stack when TX packet can be released */
+void xenvif_zerocopy_callback(struct ubuf_info *ubuf, bool zerocopy_success);
+
+/* Unmap a pending page, usually has to be called before xenvif_idx_release */
+void xenvif_idx_unmap(struct xenvif *vif, u16 pending_idx);
+
extern bool separate_tx_rx_irq;
#endif /* __XEN_NETBACK__COMMON_H__ */
diff --git a/drivers/net/xen-netback/interface.c b/drivers/net/xen-netback/interface.c
index e4aa267..f5c3c57 100644
--- a/drivers/net/xen-netback/interface.c
+++ b/drivers/net/xen-netback/interface.c
@@ -37,6 +37,7 @@
#include <xen/events.h>
#include <asm/xen/hypercall.h>
+#include <xen/balloon.h>
#define XENVIF_QUEUE_LENGTH 32
#define XENVIF_NAPI_WEIGHT 64
diff --git a/drivers/net/xen-netback/netback.c b/drivers/net/xen-netback/netback.c
index 828fdab..10470dc 100644
--- a/drivers/net/xen-netback/netback.c
+++ b/drivers/net/xen-netback/netback.c
@@ -869,6 +869,20 @@ static struct page *xenvif_alloc_page(struct xenvif *vif,
return page;
}
+static inline void xenvif_tx_create_gop(struct xenvif *vif, u16 pending_idx,
+ struct xen_netif_tx_request *txp,
+ struct gnttab_map_grant_ref *gop)
+{
+ vif->pages_to_gnt[gop-vif->tx_map_ops] = vif->mmap_pages[pending_idx];
+ gnttab_set_map_op(gop, idx_to_kaddr(vif, pending_idx),
+ GNTMAP_host_map | GNTMAP_readonly,
+ txp->gref, vif->domid);
+
+ memcpy(&vif->pending_tx_info[pending_idx].req, txp,
+ sizeof(*txp));
+
+}
+
static struct gnttab_copy *xenvif_get_requests(struct xenvif *vif,
struct sk_buff *skb,
struct xen_netif_tx_request *txp,
@@ -1652,6 +1666,106 @@ static int xenvif_tx_submit(struct xenvif *vif, int budget)
return work_done;
}
+void xenvif_zerocopy_callback(struct ubuf_info *ubuf, bool zerocopy_success)
+{
+ unsigned long flags;
+ pending_ring_idx_t index;
+ u16 pending_idx = ubuf->desc;
+ struct pending_tx_info *temp =
+ container_of(ubuf, struct pending_tx_info, callback_struct);
+ struct xenvif *vif =
+ container_of(temp - pending_idx, struct xenvif,
+ pending_tx_info[0]);
+
+ spin_lock_irqsave(&vif->dealloc_lock, flags);
+ do {
+ pending_idx = ubuf->desc;
+ ubuf = (struct ubuf_info *) ubuf->ctx;
+ index = pending_index(vif->dealloc_prod);
+ vif->dealloc_ring[index] = pending_idx;
+ /* Sync with xenvif_tx_action_dealloc:
+ * insert idx then incr producer.
+ */
+ smp_wmb();
+ vif->dealloc_prod++;
+ napi_schedule(&vif->napi);
+ } while (ubuf);
+ spin_unlock_irqrestore(&vif->dealloc_lock, flags);
+}
+
+static inline void xenvif_tx_action_dealloc(struct xenvif *vif)
+{
+ struct gnttab_unmap_grant_ref *gop;
+ pending_ring_idx_t dc, dp;
+ u16 pending_idx, pending_idx_release[MAX_PENDING_REQS];
+ unsigned int i = 0;
+
+ dc = vif->dealloc_cons;
+ gop = vif->tx_unmap_ops;
+
+ /* Free up any grants we have finished using */
+ do {
+ dp = vif->dealloc_prod;
+
+ /* Ensure we see all indices enqueued by netif_idx_release(). */
+ smp_rmb();
+
+ while (dc != dp) {
+
+ pending_idx_release[gop-vif->tx_unmap_ops] =
+ pending_idx =
+ vif->dealloc_ring[pending_index(dc++)];
+
+ /* Already unmapped? */
+ if (vif->grant_tx_handle[pending_idx] ==
+ NETBACK_INVALID_HANDLE) {
+ netdev_err(vif->dev,
+ "Trying to unmap invalid handle! "
+ "pending_idx: %x\n", pending_idx);
+ continue;
+ }
+
+ vif->pages_to_gnt[gop-vif->tx_unmap_ops] =
+ vif->mmap_pages[pending_idx];
+ gnttab_set_unmap_op(gop,
+ idx_to_kaddr(vif, pending_idx),
+ GNTMAP_host_map,
+ vif->grant_tx_handle[pending_idx]);
+ vif->grant_tx_handle[pending_idx] =
+ NETBACK_INVALID_HANDLE;
+ ++gop;
+ }
+
+ } while (dp != vif->dealloc_prod);
+
+ vif->dealloc_cons = dc;
+
+ if (gop - vif->tx_unmap_ops > 0) {
+ int ret;
+ ret = gnttab_unmap_refs(vif->tx_unmap_ops,
+ NULL,
+ vif->pages_to_gnt,
+ gop - vif->tx_unmap_ops);
+ if (ret) {
+ netdev_err(vif->dev, "Unmap fail: nr_ops %x ret %d\n",
+ gop - vif->tx_unmap_ops, ret);
+ for (i = 0; i < gop - vif->tx_unmap_ops; ++i) {
+ netdev_err(vif->dev,
+ " host_addr: %llx handle: %x status: %d\n",
+ gop[i].host_addr,
+ gop[i].handle,
+ gop[i].status);
+ }
+ BUG();
+ }
+ }
+
+ for (i = 0; i < gop - vif->tx_unmap_ops; ++i)
+ xenvif_idx_release(vif, pending_idx_release[i],
+ XEN_NETIF_RSP_OKAY);
+}
+
+
/* Called after netfront has transmitted */
int xenvif_tx_action(struct xenvif *vif, int budget)
{
@@ -1718,6 +1832,26 @@ static void xenvif_idx_release(struct xenvif *vif, u16 pending_idx,
vif->mmap_pages[pending_idx] = NULL;
}
+void xenvif_idx_unmap(struct xenvif *vif, u16 pending_idx)
+{
+ int ret;
+ if (vif->grant_tx_handle[pending_idx] == NETBACK_INVALID_HANDLE) {
+ netdev_err(vif->dev,
+ "Trying to unmap invalid handle! pending_idx: %x\n",
+ pending_idx);
+ return;
+ }
+ gnttab_set_unmap_op(&vif->tx_unmap_ops[0],
+ idx_to_kaddr(vif, pending_idx),
+ GNTMAP_host_map,
+ vif->grant_tx_handle[pending_idx]);
+ ret = gnttab_unmap_refs(vif->tx_unmap_ops,
+ NULL,
+ &vif->mmap_pages[pending_idx],
+ 1);
+ BUG_ON(ret);
+ vif->grant_tx_handle[pending_idx] = NETBACK_INVALID_HANDLE;
+}
static void make_tx_response(struct xenvif *vif,
struct xen_netif_tx_request *txp,
^ permalink raw reply related
* [PATCH net-next RFC 0/5] xen-netback: TX grant mapping instead of copy
From: Zoltan Kiss @ 2013-10-30 0:50 UTC (permalink / raw)
To: ian.campbell, wei.liu2, xen-devel, netdev, linux-kernel,
jonathan.davies
Cc: Zoltan Kiss
A long known problem of the upstream netback implementation that on the TX
path (from guest to Dom0) it copies the whole packet from guest memory into
Dom0. That simply became a bottleneck with 10Gb NICs, and generally it's a
huge perfomance penalty. The classic kernel version of netback used grant
mapping, and to get notified when the page can be unmapped, it used page
destructors. Unfortunately that destructor is not an upstreamable solution.
Ian Campbell's skb fragment destructor patch series
(http://lwn.net/Articles/491522/) tried to solve this problem, however it
seems to be very invasive on the network stack's code, and therefore haven't
progressed very well.
This patch series use SKBTX_DEV_ZEROCOPY flags to tell the stack it needs to
know when the skb is freed up. That is the way KVM solved the same problem,
and based on my initial tests it can do the same for us. Avoiding the extra
copy boosted up TX throughput from 6.8 Gbps to 7.9 (I used a slower
Interlagos box, both Dom0 and guest on upstream kernel, on the same NUMA node,
running iperf 2.0.5, and the remote end was a bare metal box on the same 10Gb
switch)
Based on my investigations the packet get only copied if it is delivered to
Dom0 stack, which is due to this patch:
https://lkml.org/lkml/2012/7/20/363
That's a bit unfortunate, but as far as I know for the huge majority this use
case is not too important. There are a couple of things which need more
polishing, see the FIXME comments. I will run some more extensive tests, but
in the meantime I would like to hear comments about what I've done so far.
I've tried to broke it down to smaller patches, with mixed results, so I
welcome suggestions on that part as well:
1/5: Introduce TX grant map definitions
2/5: Change TX path from grant copy to mapping
3/5: Remove old TX grant copy definitons
4/5: Fix indentations
5/5: Change RX path for mapped SKB fragments
Signed-off-by: Zoltan Kiss <zoltan.kiss@citrix.com>
^ permalink raw reply
* [PATCH net-next 11/11] openvswitch: TCP flags matching support.
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Jarno Rajahalme, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Jarno Rajahalme <jrajahalme@nicira.com>
tcp_flags=flags/mask
Bitwise match on TCP flags. The flags and mask are 16-bit num‐
bers written in decimal or in hexadecimal prefixed by 0x. Each
1-bit in mask requires that the corresponding bit in port must
match. Each 0-bit in mask causes the corresponding bit to be
ignored.
TCP protocol currently defines 9 flag bits, and additional 3
bits are reserved (must be transmitted as zero), see RFCs 793,
3168, and 3540. The flag bits are, numbering from the least
significant bit:
0: FIN No more data from sender.
1: SYN Synchronize sequence numbers.
2: RST Reset the connection.
3: PSH Push function.
4: ACK Acknowledgement field significant.
5: URG Urgent pointer field significant.
6: ECE ECN Echo.
7: CWR Congestion Windows Reduced.
8: NS Nonce Sum.
9-11: Reserved.
12-15: Not matchable, must be zero.
Signed-off-by: Jarno Rajahalme <jrajahalme@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
include/uapi/linux/openvswitch.h | 1 +
net/openvswitch/flow.c | 2 ++
net/openvswitch/flow.h | 2 ++
net/openvswitch/flow_netlink.c | 31 +++++++++++++++++++++++++++++--
4 files changed, 34 insertions(+), 2 deletions(-)
diff --git a/include/uapi/linux/openvswitch.h b/include/uapi/linux/openvswitch.h
index 2cc4644..d120f9f 100644
--- a/include/uapi/linux/openvswitch.h
+++ b/include/uapi/linux/openvswitch.h
@@ -271,6 +271,7 @@ enum ovs_key_attr {
OVS_KEY_ATTR_SKB_MARK, /* u32 skb mark */
OVS_KEY_ATTR_TUNNEL, /* Nested set of ovs_tunnel attributes */
OVS_KEY_ATTR_SCTP, /* struct ovs_key_sctp */
+ OVS_KEY_ATTR_TCP_FLAGS, /* be16 TCP flags. */
#ifdef __KERNEL__
OVS_KEY_ATTR_IPV4_TUNNEL, /* struct ovs_key_ipv4_tunnel */
diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
index 6ee7a41..863418f 100644
--- a/net/openvswitch/flow.c
+++ b/net/openvswitch/flow.c
@@ -482,6 +482,7 @@ int ovs_flow_extract(struct sk_buff *skb, u16 in_port, struct sw_flow_key *key)
struct tcphdr *tcp = tcp_hdr(skb);
key->ipv4.tp.src = tcp->source;
key->ipv4.tp.dst = tcp->dest;
+ key->ipv4.tp.flags = TCP_FLAGS_BE16(tcp);
}
} else if (key->ip.proto == IPPROTO_UDP) {
if (udphdr_ok(skb)) {
@@ -550,6 +551,7 @@ int ovs_flow_extract(struct sk_buff *skb, u16 in_port, struct sw_flow_key *key)
struct tcphdr *tcp = tcp_hdr(skb);
key->ipv6.tp.src = tcp->source;
key->ipv6.tp.dst = tcp->dest;
+ key->ipv6.tp.flags = TCP_FLAGS_BE16(tcp);
}
} else if (key->ip.proto == NEXTHDR_UDP) {
if (udphdr_ok(skb)) {
diff --git a/net/openvswitch/flow.h b/net/openvswitch/flow.h
index a5504de..d2d12ad 100644
--- a/net/openvswitch/flow.h
+++ b/net/openvswitch/flow.h
@@ -94,6 +94,7 @@ struct sw_flow_key {
struct {
__be16 src; /* TCP/UDP/SCTP source port. */
__be16 dst; /* TCP/UDP/SCTP destination port. */
+ __be16 flags; /* TCP flags. */
} tp;
struct {
u8 sha[ETH_ALEN]; /* ARP source hardware address. */
@@ -110,6 +111,7 @@ struct sw_flow_key {
struct {
__be16 src; /* TCP/UDP/SCTP source port. */
__be16 dst; /* TCP/UDP/SCTP destination port. */
+ __be16 flags; /* TCP flags. */
} tp;
struct {
struct in6_addr target; /* ND target address. */
diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c
index e04649c..2bc1bc1 100644
--- a/net/openvswitch/flow_netlink.c
+++ b/net/openvswitch/flow_netlink.c
@@ -114,6 +114,7 @@ static bool match_validate(const struct sw_flow_match *match,
mask_allowed &= ~((1 << OVS_KEY_ATTR_IPV4)
| (1 << OVS_KEY_ATTR_IPV6)
| (1 << OVS_KEY_ATTR_TCP)
+ | (1 << OVS_KEY_ATTR_TCP_FLAGS)
| (1 << OVS_KEY_ATTR_UDP)
| (1 << OVS_KEY_ATTR_SCTP)
| (1 << OVS_KEY_ATTR_ICMP)
@@ -154,8 +155,11 @@ static bool match_validate(const struct sw_flow_match *match,
if (match->key->ip.proto == IPPROTO_TCP) {
key_expected |= 1 << OVS_KEY_ATTR_TCP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
+ key_expected |= 1 << OVS_KEY_ATTR_TCP_FLAGS;
+ if (match->mask && (match->mask->key.ip.proto == 0xff)) {
mask_allowed |= 1 << OVS_KEY_ATTR_TCP;
+ mask_allowed |= 1 << OVS_KEY_ATTR_TCP_FLAGS;
+ }
}
if (match->key->ip.proto == IPPROTO_ICMP) {
@@ -186,8 +190,11 @@ static bool match_validate(const struct sw_flow_match *match,
if (match->key->ip.proto == IPPROTO_TCP) {
key_expected |= 1 << OVS_KEY_ATTR_TCP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
+ key_expected |= 1 << OVS_KEY_ATTR_TCP_FLAGS;
+ if (match->mask && (match->mask->key.ip.proto == 0xff)) {
mask_allowed |= 1 << OVS_KEY_ATTR_TCP;
+ mask_allowed |= 1 << OVS_KEY_ATTR_TCP_FLAGS;
+ }
}
if (match->key->ip.proto == IPPROTO_ICMPV6) {
@@ -235,6 +242,7 @@ static const int ovs_key_lens[OVS_KEY_ATTR_MAX + 1] = {
[OVS_KEY_ATTR_IPV4] = sizeof(struct ovs_key_ipv4),
[OVS_KEY_ATTR_IPV6] = sizeof(struct ovs_key_ipv6),
[OVS_KEY_ATTR_TCP] = sizeof(struct ovs_key_tcp),
+ [OVS_KEY_ATTR_TCP_FLAGS] = sizeof(__be16),
[OVS_KEY_ATTR_UDP] = sizeof(struct ovs_key_udp),
[OVS_KEY_ATTR_SCTP] = sizeof(struct ovs_key_sctp),
[OVS_KEY_ATTR_ICMP] = sizeof(struct ovs_key_icmp),
@@ -634,6 +642,19 @@ static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
attrs &= ~(1 << OVS_KEY_ATTR_TCP);
}
+ if (attrs & (1 << OVS_KEY_ATTR_TCP_FLAGS)) {
+ if (orig_attrs & (1 << OVS_KEY_ATTR_IPV4)) {
+ SW_FLOW_KEY_PUT(match, ipv4.tp.flags,
+ nla_get_be16(a[OVS_KEY_ATTR_TCP_FLAGS]),
+ is_mask);
+ } else {
+ SW_FLOW_KEY_PUT(match, ipv6.tp.flags,
+ nla_get_be16(a[OVS_KEY_ATTR_TCP_FLAGS]),
+ is_mask);
+ }
+ attrs &= ~(1 << OVS_KEY_ATTR_TCP_FLAGS);
+ }
+
if (attrs & (1 << OVS_KEY_ATTR_UDP)) {
const struct ovs_key_udp *udp_key;
@@ -1004,9 +1025,15 @@ int ovs_nla_put_flow(const struct sw_flow_key *swkey,
if (swkey->eth.type == htons(ETH_P_IP)) {
tcp_key->tcp_src = output->ipv4.tp.src;
tcp_key->tcp_dst = output->ipv4.tp.dst;
+ if (nla_put_be16(skb, OVS_KEY_ATTR_TCP_FLAGS,
+ output->ipv4.tp.flags))
+ goto nla_put_failure;
} else if (swkey->eth.type == htons(ETH_P_IPV6)) {
tcp_key->tcp_src = output->ipv6.tp.src;
tcp_key->tcp_dst = output->ipv6.tp.dst;
+ if (nla_put_be16(skb, OVS_KEY_ATTR_TCP_FLAGS,
+ output->ipv6.tp.flags))
+ goto nla_put_failure;
}
} else if (swkey->ip.proto == IPPROTO_UDP) {
struct ovs_key_udp *udp_key;
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 07/11] openvswitch: collect mega flow mask stats
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Andy Zhou, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Andy Zhou <azhou@nicira.com>
Collect mega flow mask stats. ovs-dpctl show command can be used to
display them for debugging and performance tuning.
Signed-off-by: Andy Zhou <azhou@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
include/uapi/linux/openvswitch.h | 17 ++++++++++++++---
net/openvswitch/datapath.c | 38 +++++++++++++++++++++++++++++++-------
net/openvswitch/datapath.h | 4 ++++
net/openvswitch/flow_table.c | 16 +++++++++++++++-
net/openvswitch/flow_table.h | 4 +++-
5 files changed, 67 insertions(+), 12 deletions(-)
diff --git a/include/uapi/linux/openvswitch.h b/include/uapi/linux/openvswitch.h
index a74d375..2cc4644 100644
--- a/include/uapi/linux/openvswitch.h
+++ b/include/uapi/linux/openvswitch.h
@@ -63,15 +63,18 @@ enum ovs_datapath_cmd {
* not be sent.
* @OVS_DP_ATTR_STATS: Statistics about packets that have passed through the
* datapath. Always present in notifications.
+ * @OVS_DP_ATTR_MEGAFLOW_STATS: Statistics about mega flow masks usage for the
+ * datapath. Always present in notifications.
*
* These attributes follow the &struct ovs_header within the Generic Netlink
* payload for %OVS_DP_* commands.
*/
enum ovs_datapath_attr {
OVS_DP_ATTR_UNSPEC,
- OVS_DP_ATTR_NAME, /* name of dp_ifindex netdev */
- OVS_DP_ATTR_UPCALL_PID, /* Netlink PID to receive upcalls */
- OVS_DP_ATTR_STATS, /* struct ovs_dp_stats */
+ OVS_DP_ATTR_NAME, /* name of dp_ifindex netdev */
+ OVS_DP_ATTR_UPCALL_PID, /* Netlink PID to receive upcalls */
+ OVS_DP_ATTR_STATS, /* struct ovs_dp_stats */
+ OVS_DP_ATTR_MEGAFLOW_STATS, /* struct ovs_dp_megaflow_stats */
__OVS_DP_ATTR_MAX
};
@@ -84,6 +87,14 @@ struct ovs_dp_stats {
__u64 n_flows; /* Number of flows present */
};
+struct ovs_dp_megaflow_stats {
+ __u64 n_mask_hit; /* Number of masks used for flow lookups. */
+ __u32 n_masks; /* Number of masks for the datapath. */
+ __u32 pad0; /* Pad for future expension. */
+ __u64 pad1; /* Pad for future expension. */
+ __u64 pad2; /* Pad for future expension. */
+};
+
struct ovs_vport_stats {
__u64 rx_packets; /* total packets received */
__u64 tx_packets; /* total packets transmitted */
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index cf27097..5bc5a4e 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -221,6 +221,7 @@ void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
struct dp_stats_percpu *stats;
struct sw_flow_key key;
u64 *stats_counter;
+ u32 n_mask_hit;
int error;
stats = this_cpu_ptr(dp->stats_percpu);
@@ -233,7 +234,7 @@ void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
}
/* Look up flow. */
- flow = ovs_flow_tbl_lookup(&dp->table, &key);
+ flow = ovs_flow_tbl_lookup(&dp->table, &key, &n_mask_hit);
if (unlikely(!flow)) {
struct dp_upcall_info upcall;
@@ -258,6 +259,7 @@ out:
/* Update datapath statistics. */
u64_stats_update_begin(&stats->sync);
(*stats_counter)++;
+ stats->n_mask_hit += n_mask_hit;
u64_stats_update_end(&stats->sync);
}
@@ -563,13 +565,18 @@ static struct genl_ops dp_packet_genl_ops[] = {
}
};
-static void get_dp_stats(struct datapath *dp, struct ovs_dp_stats *stats)
+static void get_dp_stats(struct datapath *dp, struct ovs_dp_stats *stats,
+ struct ovs_dp_megaflow_stats *mega_stats)
{
int i;
+ memset(mega_stats, 0, sizeof(*mega_stats));
+
stats->n_flows = ovs_flow_tbl_count(&dp->table);
+ mega_stats->n_masks = ovs_flow_tbl_num_masks(&dp->table);
stats->n_hit = stats->n_missed = stats->n_lost = 0;
+
for_each_possible_cpu(i) {
const struct dp_stats_percpu *percpu_stats;
struct dp_stats_percpu local_stats;
@@ -585,6 +592,7 @@ static void get_dp_stats(struct datapath *dp, struct ovs_dp_stats *stats)
stats->n_hit += local_stats.n_hit;
stats->n_missed += local_stats.n_missed;
stats->n_lost += local_stats.n_lost;
+ mega_stats->n_mask_hit += local_stats.n_mask_hit;
}
}
@@ -743,6 +751,14 @@ static struct sk_buff *ovs_flow_cmd_build_info(struct sw_flow *flow,
return skb;
}
+static struct sw_flow *__ovs_flow_tbl_lookup(struct flow_table *tbl,
+ const struct sw_flow_key *key)
+{
+ u32 __always_unused n_mask_hit;
+
+ return ovs_flow_tbl_lookup(tbl, key, &n_mask_hit);
+}
+
static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
{
struct nlattr **a = info->attrs;
@@ -793,7 +809,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
goto err_unlock_ovs;
/* Check if this is a duplicate flow */
- flow = ovs_flow_tbl_lookup(&dp->table, &key);
+ flow = __ovs_flow_tbl_lookup(&dp->table, &key);
if (!flow) {
/* Bail out if we're not allowed to create a new flow. */
error = -ENOENT;
@@ -905,7 +921,7 @@ static int ovs_flow_cmd_get(struct sk_buff *skb, struct genl_info *info)
goto unlock;
}
- flow = ovs_flow_tbl_lookup(&dp->table, &key);
+ flow = __ovs_flow_tbl_lookup(&dp->table, &key);
if (!flow || !ovs_flow_cmp_unmasked_key(flow, &match)) {
err = -ENOENT;
goto unlock;
@@ -953,7 +969,7 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
if (err)
goto unlock;
- flow = ovs_flow_tbl_lookup(&dp->table, &key);
+ flow = __ovs_flow_tbl_lookup(&dp->table, &key);
if (!flow || !ovs_flow_cmp_unmasked_key(flow, &match)) {
err = -ENOENT;
goto unlock;
@@ -1067,6 +1083,7 @@ static size_t ovs_dp_cmd_msg_size(void)
msgsize += nla_total_size(IFNAMSIZ);
msgsize += nla_total_size(sizeof(struct ovs_dp_stats));
+ msgsize += nla_total_size(sizeof(struct ovs_dp_megaflow_stats));
return msgsize;
}
@@ -1076,6 +1093,7 @@ static int ovs_dp_cmd_fill_info(struct datapath *dp, struct sk_buff *skb,
{
struct ovs_header *ovs_header;
struct ovs_dp_stats dp_stats;
+ struct ovs_dp_megaflow_stats dp_megaflow_stats;
int err;
ovs_header = genlmsg_put(skb, portid, seq, &dp_datapath_genl_family,
@@ -1091,8 +1109,14 @@ static int ovs_dp_cmd_fill_info(struct datapath *dp, struct sk_buff *skb,
if (err)
goto nla_put_failure;
- get_dp_stats(dp, &dp_stats);
- if (nla_put(skb, OVS_DP_ATTR_STATS, sizeof(struct ovs_dp_stats), &dp_stats))
+ get_dp_stats(dp, &dp_stats, &dp_megaflow_stats);
+ if (nla_put(skb, OVS_DP_ATTR_STATS, sizeof(struct ovs_dp_stats),
+ &dp_stats))
+ goto nla_put_failure;
+
+ if (nla_put(skb, OVS_DP_ATTR_MEGAFLOW_STATS,
+ sizeof(struct ovs_dp_megaflow_stats),
+ &dp_megaflow_stats))
goto nla_put_failure;
return genlmsg_end(skb, ovs_header);
diff --git a/net/openvswitch/datapath.h b/net/openvswitch/datapath.h
index acfd4af..d3d14a58 100644
--- a/net/openvswitch/datapath.h
+++ b/net/openvswitch/datapath.h
@@ -46,11 +46,15 @@
* @n_lost: Number of received packets that had no matching flow in the flow
* table that could not be sent to userspace (normally due to an overflow in
* one of the datapath's queues).
+ * @n_mask_hit: Number of masks looked up for flow match.
+ * @n_mask_hit / (@n_hit + @n_missed) will be the average masks looked
+ * up per packet.
*/
struct dp_stats_percpu {
u64 n_hit;
u64 n_missed;
u64 n_lost;
+ u64 n_mask_hit;
struct u64_stats_sync sync;
};
diff --git a/net/openvswitch/flow_table.c b/net/openvswitch/flow_table.c
index 036e019..536b4d2 100644
--- a/net/openvswitch/flow_table.c
+++ b/net/openvswitch/flow_table.c
@@ -430,13 +430,16 @@ static struct sw_flow *masked_flow_lookup(struct table_instance *ti,
}
struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *tbl,
- const struct sw_flow_key *key)
+ const struct sw_flow_key *key,
+ u32 *n_mask_hit)
{
struct table_instance *ti = rcu_dereference(tbl->ti);
struct sw_flow_mask *mask;
struct sw_flow *flow;
+ *n_mask_hit = 0;
list_for_each_entry_rcu(mask, &tbl->mask_list, list) {
+ (*n_mask_hit)++;
flow = masked_flow_lookup(ti, key, mask);
if (flow) /* Found */
return flow;
@@ -444,6 +447,17 @@ struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *tbl,
return NULL;
}
+int ovs_flow_tbl_num_masks(const struct flow_table *table)
+{
+ struct sw_flow_mask *mask;
+ int num = 0;
+
+ list_for_each_entry(mask, &table->mask_list, list)
+ num++;
+
+ return num;
+}
+
static struct table_instance *table_instance_expand(struct table_instance *ti)
{
return table_instance_rehash(ti, ti->n_buckets * 2);
diff --git a/net/openvswitch/flow_table.h b/net/openvswitch/flow_table.h
index 4db5f78..fbe45d5 100644
--- a/net/openvswitch/flow_table.h
+++ b/net/openvswitch/flow_table.h
@@ -66,10 +66,12 @@ int ovs_flow_tbl_flush(struct flow_table *flow_table);
int ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow,
struct sw_flow_mask *mask);
void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow);
+int ovs_flow_tbl_num_masks(const struct flow_table *table);
struct sw_flow *ovs_flow_tbl_dump_next(struct table_instance *table,
u32 *bucket, u32 *idx);
struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *,
- const struct sw_flow_key *);
+ const struct sw_flow_key *,
+ u32 *n_mask_hit);
bool ovs_flow_cmp_unmasked_key(const struct sw_flow *flow,
struct sw_flow_match *match);
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 10/11] openvswitch: Widen TCP flags handling.
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Jarno Rajahalme, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Jarno Rajahalme <jrajahalme@nicira.com>
Widen TCP flags handling from 7 bits (uint8_t) to 12 bits (uint16_t).
The kernel interface remains at 8 bits, which makes no functional
difference now, as none of the higher bits is currently of interest
to the userspace.
Signed-off-by: Jarno Rajahalme <jrajahalme@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
net/openvswitch/datapath.c | 5 +++--
net/openvswitch/flow.c | 8 +++-----
net/openvswitch/flow.h | 2 +-
3 files changed, 7 insertions(+), 8 deletions(-)
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 8c8875b..376b2ad 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -673,8 +673,9 @@ static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
goto nla_put_failure;
}
- if (flow_stats.tcp_flags &&
- nla_put_u8(skb, OVS_FLOW_ATTR_TCP_FLAGS, flow_stats.tcp_flags))
+ if ((u8)ntohs(flow_stats.tcp_flags) &&
+ nla_put_u8(skb, OVS_FLOW_ATTR_TCP_FLAGS,
+ (u8)ntohs(flow_stats.tcp_flags)))
goto nla_put_failure;
/* If OVS_FLOW_ATTR_ACTIONS doesn't fit, skip dumping the actions if
diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
index 2d82995..6ee7a41 100644
--- a/net/openvswitch/flow.c
+++ b/net/openvswitch/flow.c
@@ -59,20 +59,18 @@ u64 ovs_flow_used_time(unsigned long flow_jiffies)
return cur_ms - idle_ms;
}
-#define TCP_FLAGS_OFFSET 13
-#define TCP_FLAG_MASK 0x3f
+#define TCP_FLAGS_BE16(tp) (*(__be16 *)&tcp_flag_word(tp) & htons(0x0FFF))
void ovs_flow_stats_update(struct sw_flow *flow, struct sk_buff *skb)
{
struct sw_flow_stats *stats = &flow->stats[smp_processor_id()];
- u8 tcp_flags = 0;
+ __be16 tcp_flags = 0;
if ((flow->key.eth.type == htons(ETH_P_IP) ||
flow->key.eth.type == htons(ETH_P_IPV6)) &&
flow->key.ip.proto == IPPROTO_TCP &&
likely(skb->len >= skb_transport_offset(skb) + sizeof(struct tcphdr))) {
- u8 *tcp = (u8 *)tcp_hdr(skb);
- tcp_flags = *(tcp + TCP_FLAGS_OFFSET) & TCP_FLAG_MASK;
+ tcp_flags = TCP_FLAGS_BE16(tcp_hdr(skb));
}
spin_lock(&stats->lock);
diff --git a/net/openvswitch/flow.h b/net/openvswitch/flow.h
index b844252..a5504de 100644
--- a/net/openvswitch/flow.h
+++ b/net/openvswitch/flow.h
@@ -150,7 +150,7 @@ struct sw_flow_stats {
u64 byte_count; /* Number of bytes matched. */
unsigned long used; /* Last used time (in jiffies). */
spinlock_t lock; /* Lock for atomic stats update. */
- u8 tcp_flags; /* Union of seen TCP flags. */
+ __be16 tcp_flags; /* Union of seen TCP flags. */
} ____cacheline_aligned_in_smp;
struct sw_flow {
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 09/11] openvswitch: Enable all GSO features on internal port.
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Pravin B Shelar, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Pravin B Shelar <pshelar@nicira.com>
OVS already can handle all types of segmentation offloads that
are supported by the kernel.
Following patch specifically enables UDP and IPV6 segmentation
offloads.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
net/openvswitch/vport-internal_dev.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/net/openvswitch/vport-internal_dev.c b/net/openvswitch/vport-internal_dev.c
index 98d3edb..729c687 100644
--- a/net/openvswitch/vport-internal_dev.c
+++ b/net/openvswitch/vport-internal_dev.c
@@ -134,7 +134,7 @@ static void do_setup(struct net_device *netdev)
netdev->tx_queue_len = 0;
netdev->features = NETIF_F_LLTX | NETIF_F_SG | NETIF_F_FRAGLIST |
- NETIF_F_HIGHDMA | NETIF_F_HW_CSUM | NETIF_F_TSO;
+ NETIF_F_HIGHDMA | NETIF_F_HW_CSUM | NETIF_F_GSO_SOFTWARE;
netdev->vlan_features = netdev->features;
netdev->features |= NETIF_F_HW_VLAN_CTAG_TX;
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 08/11] openvswitch: Per cpu flow stats.
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Pravin B Shelar, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Pravin B Shelar <pshelar@nicira.com>
With mega flow implementation ovs flow can be shared between
multiple CPUs which makes stats updates highly contended
operation. Following patch allocates separate stats for each
CPU to make stats update scalable.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
net/openvswitch/datapath.c | 54 +++++++++++++----------------------
net/openvswitch/flow.c | 68 +++++++++++++++++++++++++++++++++++++++-----
net/openvswitch/flow.h | 20 ++++++++-----
net/openvswitch/flow_table.c | 14 +++++++--
4 files changed, 105 insertions(+), 51 deletions(-)
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 5bc5a4e..8c8875b 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -251,9 +251,9 @@ void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
OVS_CB(skb)->flow = flow;
OVS_CB(skb)->pkt_key = &key;
- stats_counter = &stats->n_hit;
- ovs_flow_used(OVS_CB(skb)->flow, skb);
+ ovs_flow_stats_update(OVS_CB(skb)->flow, skb);
ovs_execute_actions(dp, skb);
+ stats_counter = &stats->n_hit;
out:
/* Update datapath statistics. */
@@ -453,14 +453,6 @@ out:
return err;
}
-static void clear_stats(struct sw_flow *flow)
-{
- flow->used = 0;
- flow->tcp_flags = 0;
- flow->packet_count = 0;
- flow->byte_count = 0;
-}
-
static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
{
struct ovs_header *ovs_header = info->userhdr;
@@ -634,11 +626,9 @@ static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
{
const int skb_orig_len = skb->len;
struct nlattr *start;
- struct ovs_flow_stats stats;
+ struct sw_flow_stats flow_stats;
struct ovs_header *ovs_header;
struct nlattr *nla;
- unsigned long used;
- u8 tcp_flags;
int err;
ovs_header = genlmsg_put(skb, portid, seq, &dp_flow_genl_family, flags, cmd);
@@ -667,24 +657,24 @@ static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
nla_nest_end(skb, nla);
- spin_lock_bh(&flow->lock);
- used = flow->used;
- stats.n_packets = flow->packet_count;
- stats.n_bytes = flow->byte_count;
- tcp_flags = flow->tcp_flags;
- spin_unlock_bh(&flow->lock);
-
- if (used &&
- nla_put_u64(skb, OVS_FLOW_ATTR_USED, ovs_flow_used_time(used)))
+ ovs_flow_stats_get(flow, &flow_stats);
+ if (flow_stats.used &&
+ nla_put_u64(skb, OVS_FLOW_ATTR_USED, ovs_flow_used_time(flow_stats.used)))
goto nla_put_failure;
- if (stats.n_packets &&
- nla_put(skb, OVS_FLOW_ATTR_STATS,
- sizeof(struct ovs_flow_stats), &stats))
- goto nla_put_failure;
+ if (flow_stats.packet_count) {
+ struct ovs_flow_stats stats = {
+ .n_packets = flow_stats.packet_count,
+ .n_bytes = flow_stats.byte_count,
+ };
- if (tcp_flags &&
- nla_put_u8(skb, OVS_FLOW_ATTR_TCP_FLAGS, tcp_flags))
+ if (nla_put(skb, OVS_FLOW_ATTR_STATS,
+ sizeof(struct ovs_flow_stats), &stats))
+ goto nla_put_failure;
+ }
+
+ if (flow_stats.tcp_flags &&
+ nla_put_u8(skb, OVS_FLOW_ATTR_TCP_FLAGS, flow_stats.tcp_flags))
goto nla_put_failure;
/* If OVS_FLOW_ATTR_ACTIONS doesn't fit, skip dumping the actions if
@@ -822,7 +812,6 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
error = PTR_ERR(flow);
goto err_unlock_ovs;
}
- clear_stats(flow);
flow->key = masked_key;
flow->unmasked_key = key;
@@ -868,11 +857,8 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
info->snd_seq, OVS_FLOW_CMD_NEW);
/* Clear stats. */
- if (a[OVS_FLOW_ATTR_CLEAR]) {
- spin_lock_bh(&flow->lock);
- clear_stats(flow);
- spin_unlock_bh(&flow->lock);
- }
+ if (a[OVS_FLOW_ATTR_CLEAR])
+ ovs_flow_stats_clear(flow);
}
ovs_unlock();
diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
index 617810f..2d82995 100644
--- a/net/openvswitch/flow.c
+++ b/net/openvswitch/flow.c
@@ -35,6 +35,7 @@
#include <linux/ip.h>
#include <linux/ipv6.h>
#include <linux/sctp.h>
+#include <linux/smp.h>
#include <linux/tcp.h>
#include <linux/udp.h>
#include <linux/icmp.h>
@@ -61,8 +62,9 @@ u64 ovs_flow_used_time(unsigned long flow_jiffies)
#define TCP_FLAGS_OFFSET 13
#define TCP_FLAG_MASK 0x3f
-void ovs_flow_used(struct sw_flow *flow, struct sk_buff *skb)
+void ovs_flow_stats_update(struct sw_flow *flow, struct sk_buff *skb)
{
+ struct sw_flow_stats *stats = &flow->stats[smp_processor_id()];
u8 tcp_flags = 0;
if ((flow->key.eth.type == htons(ETH_P_IP) ||
@@ -73,12 +75,64 @@ void ovs_flow_used(struct sw_flow *flow, struct sk_buff *skb)
tcp_flags = *(tcp + TCP_FLAGS_OFFSET) & TCP_FLAG_MASK;
}
- spin_lock(&flow->lock);
- flow->used = jiffies;
- flow->packet_count++;
- flow->byte_count += skb->len;
- flow->tcp_flags |= tcp_flags;
- spin_unlock(&flow->lock);
+ spin_lock(&stats->lock);
+ stats->used = jiffies;
+ stats->packet_count++;
+ stats->byte_count += skb->len;
+ stats->tcp_flags |= tcp_flags;
+ spin_unlock(&stats->lock);
+}
+
+void ovs_flow_stats_get(struct sw_flow *flow, struct sw_flow_stats *res)
+{
+ int cpu, cur_cpu;
+
+ memset(res, 0, sizeof(*res));
+
+ cur_cpu = get_cpu();
+ for_each_possible_cpu(cpu) {
+ struct sw_flow_stats *stats = &flow->stats[cpu];
+
+ if (cpu == cur_cpu)
+ local_bh_disable();
+
+ spin_lock(&stats->lock);
+ if (time_after(stats->used, res->used))
+ res->used = stats->used;
+ res->packet_count += stats->packet_count;
+ res->byte_count += stats->byte_count;
+ res->tcp_flags |= stats->tcp_flags;
+ spin_unlock(&stats->lock);
+
+ if (cpu == cur_cpu)
+ local_bh_enable();
+
+ }
+ put_cpu();
+}
+
+void ovs_flow_stats_clear(struct sw_flow *flow)
+{
+ int cpu, cur_cpu;
+
+ cur_cpu = get_cpu();
+ for_each_possible_cpu(cpu) {
+ struct sw_flow_stats *stats = &flow->stats[cpu];
+
+ if (cpu == cur_cpu)
+ local_bh_disable();
+
+ spin_lock(&stats->lock);
+ stats->used = 0;
+ stats->packet_count = 0;
+ stats->byte_count = 0;
+ stats->tcp_flags = 0;
+ spin_unlock(&stats->lock);
+
+ if (cpu == cur_cpu)
+ local_bh_enable();
+ }
+ put_cpu();
}
static int check_header(struct sk_buff *skb, int len)
diff --git a/net/openvswitch/flow.h b/net/openvswitch/flow.h
index 098fd1d..b844252 100644
--- a/net/openvswitch/flow.h
+++ b/net/openvswitch/flow.h
@@ -19,6 +19,7 @@
#ifndef FLOW_H
#define FLOW_H 1
+#include <linux/cache.h>
#include <linux/kernel.h>
#include <linux/netlink.h>
#include <linux/openvswitch.h>
@@ -144,6 +145,14 @@ struct sw_flow_actions {
struct nlattr actions[];
};
+struct sw_flow_stats {
+ u64 packet_count; /* Number of packets matched. */
+ u64 byte_count; /* Number of bytes matched. */
+ unsigned long used; /* Last used time (in jiffies). */
+ spinlock_t lock; /* Lock for atomic stats update. */
+ u8 tcp_flags; /* Union of seen TCP flags. */
+} ____cacheline_aligned_in_smp;
+
struct sw_flow {
struct rcu_head rcu;
struct hlist_node hash_node[2];
@@ -153,12 +162,7 @@ struct sw_flow {
struct sw_flow_key unmasked_key;
struct sw_flow_mask *mask;
struct sw_flow_actions __rcu *sf_acts;
-
- spinlock_t lock; /* Lock for values below. */
- unsigned long used; /* Last used time (in jiffies). */
- u64 packet_count; /* Number of packets matched. */
- u64 byte_count; /* Number of bytes matched. */
- u8 tcp_flags; /* Union of seen TCP flags. */
+ struct sw_flow_stats stats[];
};
struct arp_eth_header {
@@ -175,7 +179,9 @@ struct arp_eth_header {
unsigned char ar_tip[4]; /* target IP address */
} __packed;
-void ovs_flow_used(struct sw_flow *, struct sk_buff *);
+void ovs_flow_stats_update(struct sw_flow *flow, struct sk_buff *skb);
+void ovs_flow_stats_get(struct sw_flow *flow, struct sw_flow_stats *res);
+void ovs_flow_stats_clear(struct sw_flow *flow);
u64 ovs_flow_used_time(unsigned long flow_jiffies);
int ovs_flow_extract(struct sk_buff *, u16 in_port, struct sw_flow_key *);
diff --git a/net/openvswitch/flow_table.c b/net/openvswitch/flow_table.c
index 536b4d2..feb1b9b 100644
--- a/net/openvswitch/flow_table.c
+++ b/net/openvswitch/flow_table.c
@@ -75,15 +75,19 @@ void ovs_flow_mask_key(struct sw_flow_key *dst, const struct sw_flow_key *src,
struct sw_flow *ovs_flow_alloc(void)
{
struct sw_flow *flow;
+ int cpu;
flow = kmem_cache_alloc(flow_cache, GFP_KERNEL);
if (!flow)
return ERR_PTR(-ENOMEM);
- spin_lock_init(&flow->lock);
flow->sf_acts = NULL;
flow->mask = NULL;
+ memset(flow->stats, 0, num_possible_cpus() * sizeof(struct sw_flow_stats));
+ for_each_possible_cpu(cpu)
+ spin_lock_init(&flow->stats[cpu].lock);
+
return flow;
}
@@ -574,11 +578,15 @@ int ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow,
* Returns zero if successful or a negative error code. */
int ovs_flow_init(void)
{
+ int flow_size;
+
BUILD_BUG_ON(__alignof__(struct sw_flow_key) % __alignof__(long));
BUILD_BUG_ON(sizeof(struct sw_flow_key) % sizeof(long));
- flow_cache = kmem_cache_create("sw_flow", sizeof(struct sw_flow), 0,
- 0, NULL);
+ flow_size = sizeof(struct sw_flow) +
+ (num_possible_cpus() * sizeof(struct sw_flow_stats));
+
+ flow_cache = kmem_cache_create("sw_flow", flow_size, 0, 0, NULL);
if (flow_cache == NULL)
return -ENOMEM;
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 06/11] openvswitch: Simplify mega-flow APIs.
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Pravin B Shelar, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Pravin B Shelar <pshelar@nicira.com>
Hides mega-flow implementation in flow_table.c rather than
datapath.c.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
net/openvswitch/datapath.c | 27 +++------
net/openvswitch/flow_table.c | 138 +++++++++++++++++++++++++------------------
net/openvswitch/flow_table.h | 12 +---
3 files changed, 89 insertions(+), 88 deletions(-)
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 60b9be3..cf27097 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -161,7 +161,7 @@ static void destroy_dp_rcu(struct rcu_head *rcu)
{
struct datapath *dp = container_of(rcu, struct datapath, rcu);
- ovs_flow_tbl_destroy(&dp->table, false);
+ ovs_flow_tbl_destroy(&dp->table);
free_percpu(dp->stats_percpu);
release_net(ovs_dp_get_net(dp));
kfree(dp->ports);
@@ -795,8 +795,6 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
/* Check if this is a duplicate flow */
flow = ovs_flow_tbl_lookup(&dp->table, &key);
if (!flow) {
- struct sw_flow_mask *mask_p;
-
/* Bail out if we're not allowed to create a new flow. */
error = -ENOENT;
if (info->genlhdr->cmd == OVS_FLOW_CMD_SET)
@@ -812,25 +810,14 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
flow->key = masked_key;
flow->unmasked_key = key;
-
- /* Make sure mask is unique in the system */
- mask_p = ovs_sw_flow_mask_find(&dp->table, &mask);
- if (!mask_p) {
- /* Allocate a new mask if none exsits. */
- mask_p = ovs_sw_flow_mask_alloc();
- if (!mask_p)
- goto err_flow_free;
- mask_p->key = mask.key;
- mask_p->range = mask.range;
- ovs_sw_flow_mask_insert(&dp->table, mask_p);
- }
-
- ovs_sw_flow_mask_add_ref(mask_p);
- flow->mask = mask_p;
rcu_assign_pointer(flow->sf_acts, acts);
/* Put flow in bucket. */
- ovs_flow_tbl_insert(&dp->table, flow);
+ error = ovs_flow_tbl_insert(&dp->table, flow, &mask);
+ if (error) {
+ acts = NULL;
+ goto err_flow_free;
+ }
reply = ovs_flow_cmd_build_info(flow, dp, info->snd_portid,
info->snd_seq, OVS_FLOW_CMD_NEW);
@@ -1236,7 +1223,7 @@ err_destroy_ports_array:
err_destroy_percpu:
free_percpu(dp->stats_percpu);
err_destroy_table:
- ovs_flow_tbl_destroy(&dp->table, false);
+ ovs_flow_tbl_destroy(&dp->table);
err_free_dp:
release_net(ovs_dp_get_net(dp));
kfree(dp);
diff --git a/net/openvswitch/flow_table.c b/net/openvswitch/flow_table.c
index 1c7e773..036e019 100644
--- a/net/openvswitch/flow_table.c
+++ b/net/openvswitch/flow_table.c
@@ -128,12 +128,36 @@ static void rcu_free_flow_callback(struct rcu_head *rcu)
flow_free(flow);
}
+static void rcu_free_sw_flow_mask_cb(struct rcu_head *rcu)
+{
+ struct sw_flow_mask *mask = container_of(rcu, struct sw_flow_mask, rcu);
+
+ kfree(mask);
+}
+
+static void flow_mask_del_ref(struct sw_flow_mask *mask, bool deferred)
+{
+ if (!mask)
+ return;
+
+ BUG_ON(!mask->ref_count);
+ mask->ref_count--;
+
+ if (!mask->ref_count) {
+ list_del_rcu(&mask->list);
+ if (deferred)
+ call_rcu(&mask->rcu, rcu_free_sw_flow_mask_cb);
+ else
+ kfree(mask);
+ }
+}
+
void ovs_flow_free(struct sw_flow *flow, bool deferred)
{
if (!flow)
return;
- ovs_sw_flow_mask_del_ref(flow->mask, deferred);
+ flow_mask_del_ref(flow->mask, deferred);
if (deferred)
call_rcu(&flow->rcu, rcu_free_flow_callback);
@@ -225,11 +249,11 @@ static void table_instance_destroy(struct table_instance *ti, bool deferred)
__table_instance_destroy(ti);
}
-void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred)
+void ovs_flow_tbl_destroy(struct flow_table *table)
{
struct table_instance *ti = ovsl_dereference(table->ti);
- table_instance_destroy(ti, deferred);
+ table_instance_destroy(ti, false);
}
struct sw_flow *ovs_flow_tbl_dump_next(struct table_instance *ti,
@@ -304,7 +328,7 @@ static struct table_instance *table_instance_rehash(struct table_instance *ti,
new_ti = table_instance_alloc(n_buckets);
if (!new_ti)
- return ERR_PTR(-ENOMEM);
+ return NULL;
flow_table_copy_flows(ti, new_ti);
@@ -425,32 +449,6 @@ static struct table_instance *table_instance_expand(struct table_instance *ti)
return table_instance_rehash(ti, ti->n_buckets * 2);
}
-void ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow)
-{
- struct table_instance *ti = NULL;
- struct table_instance *new_ti = NULL;
-
- ti = ovsl_dereference(table->ti);
-
- /* Expand table, if necessary, to make room. */
- if (table->count > ti->n_buckets)
- new_ti = table_instance_expand(ti);
- else if (time_after(jiffies, table->last_rehash + REHASH_INTERVAL))
- new_ti = table_instance_rehash(ti, ti->n_buckets);
-
- if (new_ti && !IS_ERR(new_ti)) {
- rcu_assign_pointer(table->ti, new_ti);
- ovs_flow_tbl_destroy(table, true);
- ti = ovsl_dereference(table->ti);
- table->last_rehash = jiffies;
- }
-
- flow->hash = flow_hash(&flow->key, flow->mask->range.start,
- flow->mask->range.end);
- table_instance_insert(ti, flow);
- table->count++;
-}
-
void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow)
{
struct table_instance *ti = ovsl_dereference(table->ti);
@@ -460,7 +458,7 @@ void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow)
table->count--;
}
-struct sw_flow_mask *ovs_sw_flow_mask_alloc(void)
+static struct sw_flow_mask *mask_alloc(void)
{
struct sw_flow_mask *mask;
@@ -471,35 +469,11 @@ struct sw_flow_mask *ovs_sw_flow_mask_alloc(void)
return mask;
}
-void ovs_sw_flow_mask_add_ref(struct sw_flow_mask *mask)
+static void mask_add_ref(struct sw_flow_mask *mask)
{
mask->ref_count++;
}
-static void rcu_free_sw_flow_mask_cb(struct rcu_head *rcu)
-{
- struct sw_flow_mask *mask = container_of(rcu, struct sw_flow_mask, rcu);
-
- kfree(mask);
-}
-
-void ovs_sw_flow_mask_del_ref(struct sw_flow_mask *mask, bool deferred)
-{
- if (!mask)
- return;
-
- BUG_ON(!mask->ref_count);
- mask->ref_count--;
-
- if (!mask->ref_count) {
- list_del_rcu(&mask->list);
- if (deferred)
- call_rcu(&mask->rcu, rcu_free_sw_flow_mask_cb);
- else
- kfree(mask);
- }
-}
-
static bool mask_equal(const struct sw_flow_mask *a,
const struct sw_flow_mask *b)
{
@@ -511,7 +485,7 @@ static bool mask_equal(const struct sw_flow_mask *a,
&& (memcmp(a_, b_, range_n_bytes(&a->range)) == 0);
}
-struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *tbl,
+static struct sw_flow_mask *flow_mask_find(const struct flow_table *tbl,
const struct sw_flow_mask *mask)
{
struct list_head *ml;
@@ -531,9 +505,55 @@ struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *tbl,
* The caller needs to make sure that 'mask' is not the same
* as any masks that are already on the list.
*/
-void ovs_sw_flow_mask_insert(struct flow_table *tbl, struct sw_flow_mask *mask)
+static int flow_mask_insert(struct flow_table *tbl, struct sw_flow *flow,
+ struct sw_flow_mask *new)
+{
+ struct sw_flow_mask *mask;
+ mask = flow_mask_find(tbl, new);
+ if (!mask) {
+ /* Allocate a new mask if none exsits. */
+ mask = mask_alloc();
+ if (!mask)
+ return -ENOMEM;
+ mask->key = new->key;
+ mask->range = new->range;
+ list_add_rcu(&mask->list, &tbl->mask_list);
+ }
+
+ mask_add_ref(mask);
+ flow->mask = mask;
+ return 0;
+}
+
+int ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow,
+ struct sw_flow_mask *mask)
{
- list_add_rcu(&mask->list, &tbl->mask_list);
+ struct table_instance *new_ti = NULL;
+ struct table_instance *ti;
+ int err;
+
+ err = flow_mask_insert(table, flow, mask);
+ if (err)
+ return err;
+
+ flow->hash = flow_hash(&flow->key, flow->mask->range.start,
+ flow->mask->range.end);
+ ti = ovsl_dereference(table->ti);
+ table_instance_insert(ti, flow);
+ table->count++;
+
+ /* Expand table, if necessary, to make room. */
+ if (table->count > ti->n_buckets)
+ new_ti = table_instance_expand(ti);
+ else if (time_after(jiffies, table->last_rehash + REHASH_INTERVAL))
+ new_ti = table_instance_rehash(ti, ti->n_buckets);
+
+ if (new_ti) {
+ rcu_assign_pointer(table->ti, new_ti);
+ table_instance_destroy(ti, true);
+ table->last_rehash = jiffies;
+ }
+ return 0;
}
/* Initializes the flow module.
diff --git a/net/openvswitch/flow_table.h b/net/openvswitch/flow_table.h
index 5d1abe5..4db5f78 100644
--- a/net/openvswitch/flow_table.h
+++ b/net/openvswitch/flow_table.h
@@ -60,10 +60,11 @@ void ovs_flow_free(struct sw_flow *, bool deferred);
int ovs_flow_tbl_init(struct flow_table *);
int ovs_flow_tbl_count(struct flow_table *table);
-void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred);
+void ovs_flow_tbl_destroy(struct flow_table *table);
int ovs_flow_tbl_flush(struct flow_table *flow_table);
-void ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow);
+int ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow,
+ struct sw_flow_mask *mask);
void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow);
struct sw_flow *ovs_flow_tbl_dump_next(struct table_instance *table,
u32 *bucket, u32 *idx);
@@ -73,13 +74,6 @@ struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *,
bool ovs_flow_cmp_unmasked_key(const struct sw_flow *flow,
struct sw_flow_match *match);
-struct sw_flow_mask *ovs_sw_flow_mask_alloc(void);
-void ovs_sw_flow_mask_add_ref(struct sw_flow_mask *);
-void ovs_sw_flow_mask_del_ref(struct sw_flow_mask *, bool deferred);
-void ovs_sw_flow_mask_insert(struct flow_table *, struct sw_flow_mask *);
-struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *,
- const struct sw_flow_mask *);
void ovs_flow_mask_key(struct sw_flow_key *dst, const struct sw_flow_key *src,
const struct sw_flow_mask *mask);
-
#endif /* flow_table.h */
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 05/11] openvswitch: Move mega-flow list out of rehashing struct.
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Pravin B Shelar, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Pravin B Shelar <pshelar@nicira.com>
ovs-flow rehash does not touch mega flow list. Following patch
moves it dp struct datapath. Avoid one extra indirection for
accessing mega-flow list head on every packet receive.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
net/openvswitch/datapath.c | 77 ++++------------
net/openvswitch/datapath.h | 6 +-
net/openvswitch/flow_table.c | 205 ++++++++++++++++++++++++++-----------------
net/openvswitch/flow_table.h | 32 +++----
4 files changed, 155 insertions(+), 165 deletions(-)
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 72e6874..60b9be3 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -59,8 +59,6 @@
#include "vport-internal_dev.h"
#include "vport-netdev.h"
-#define REHASH_FLOW_INTERVAL (10 * 60 * HZ)
-
int ovs_net_id __read_mostly;
static void ovs_notify(struct sk_buff *skb, struct genl_info *info,
@@ -163,7 +161,7 @@ static void destroy_dp_rcu(struct rcu_head *rcu)
{
struct datapath *dp = container_of(rcu, struct datapath, rcu);
- ovs_flow_tbl_destroy((__force struct flow_table *)dp->table, false);
+ ovs_flow_tbl_destroy(&dp->table, false);
free_percpu(dp->stats_percpu);
release_net(ovs_dp_get_net(dp));
kfree(dp->ports);
@@ -235,7 +233,7 @@ void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
}
/* Look up flow. */
- flow = ovs_flow_tbl_lookup(rcu_dereference(dp->table), &key);
+ flow = ovs_flow_tbl_lookup(&dp->table, &key);
if (unlikely(!flow)) {
struct dp_upcall_info upcall;
@@ -453,23 +451,6 @@ out:
return err;
}
-/* Called with ovs_mutex. */
-static int flush_flows(struct datapath *dp)
-{
- struct flow_table *old_table;
- struct flow_table *new_table;
-
- old_table = ovsl_dereference(dp->table);
- new_table = ovs_flow_tbl_alloc(TBL_MIN_BUCKETS);
- if (!new_table)
- return -ENOMEM;
-
- rcu_assign_pointer(dp->table, new_table);
-
- ovs_flow_tbl_destroy(old_table, true);
- return 0;
-}
-
static void clear_stats(struct sw_flow *flow)
{
flow->used = 0;
@@ -584,11 +565,9 @@ static struct genl_ops dp_packet_genl_ops[] = {
static void get_dp_stats(struct datapath *dp, struct ovs_dp_stats *stats)
{
- struct flow_table *table;
int i;
- table = rcu_dereference_check(dp->table, lockdep_ovsl_is_held());
- stats->n_flows = ovs_flow_tbl_count(table);
+ stats->n_flows = ovs_flow_tbl_count(&dp->table);
stats->n_hit = stats->n_missed = stats->n_lost = 0;
for_each_possible_cpu(i) {
@@ -773,7 +752,6 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
struct sw_flow_mask mask;
struct sk_buff *reply;
struct datapath *dp;
- struct flow_table *table;
struct sw_flow_actions *acts = NULL;
struct sw_flow_match match;
int error;
@@ -814,12 +792,9 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
if (!dp)
goto err_unlock_ovs;
- table = ovsl_dereference(dp->table);
-
/* Check if this is a duplicate flow */
- flow = ovs_flow_tbl_lookup(table, &key);
+ flow = ovs_flow_tbl_lookup(&dp->table, &key);
if (!flow) {
- struct flow_table *new_table = NULL;
struct sw_flow_mask *mask_p;
/* Bail out if we're not allowed to create a new flow. */
@@ -827,19 +802,6 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
if (info->genlhdr->cmd == OVS_FLOW_CMD_SET)
goto err_unlock_ovs;
- /* Expand table, if necessary, to make room. */
- if (ovs_flow_tbl_need_to_expand(table))
- new_table = ovs_flow_tbl_expand(table);
- else if (time_after(jiffies, dp->last_rehash + REHASH_FLOW_INTERVAL))
- new_table = ovs_flow_tbl_rehash(table);
-
- if (new_table && !IS_ERR(new_table)) {
- rcu_assign_pointer(dp->table, new_table);
- ovs_flow_tbl_destroy(table, true);
- table = ovsl_dereference(dp->table);
- dp->last_rehash = jiffies;
- }
-
/* Allocate flow. */
flow = ovs_flow_alloc();
if (IS_ERR(flow)) {
@@ -852,7 +814,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
flow->unmasked_key = key;
/* Make sure mask is unique in the system */
- mask_p = ovs_sw_flow_mask_find(table, &mask);
+ mask_p = ovs_sw_flow_mask_find(&dp->table, &mask);
if (!mask_p) {
/* Allocate a new mask if none exsits. */
mask_p = ovs_sw_flow_mask_alloc();
@@ -860,7 +822,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
goto err_flow_free;
mask_p->key = mask.key;
mask_p->range = mask.range;
- ovs_sw_flow_mask_insert(table, mask_p);
+ ovs_sw_flow_mask_insert(&dp->table, mask_p);
}
ovs_sw_flow_mask_add_ref(mask_p);
@@ -868,7 +830,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
rcu_assign_pointer(flow->sf_acts, acts);
/* Put flow in bucket. */
- ovs_flow_tbl_insert(table, flow);
+ ovs_flow_tbl_insert(&dp->table, flow);
reply = ovs_flow_cmd_build_info(flow, dp, info->snd_portid,
info->snd_seq, OVS_FLOW_CMD_NEW);
@@ -936,7 +898,6 @@ static int ovs_flow_cmd_get(struct sk_buff *skb, struct genl_info *info)
struct sk_buff *reply;
struct sw_flow *flow;
struct datapath *dp;
- struct flow_table *table;
struct sw_flow_match match;
int err;
@@ -957,8 +918,7 @@ static int ovs_flow_cmd_get(struct sk_buff *skb, struct genl_info *info)
goto unlock;
}
- table = ovsl_dereference(dp->table);
- flow = ovs_flow_tbl_lookup(table, &key);
+ flow = ovs_flow_tbl_lookup(&dp->table, &key);
if (!flow || !ovs_flow_cmp_unmasked_key(flow, &match)) {
err = -ENOENT;
goto unlock;
@@ -986,7 +946,6 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
struct sk_buff *reply;
struct sw_flow *flow;
struct datapath *dp;
- struct flow_table *table;
struct sw_flow_match match;
int err;
@@ -998,7 +957,7 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
}
if (!a[OVS_FLOW_ATTR_KEY]) {
- err = flush_flows(dp);
+ err = ovs_flow_tbl_flush(&dp->table);
goto unlock;
}
@@ -1007,8 +966,7 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
if (err)
goto unlock;
- table = ovsl_dereference(dp->table);
- flow = ovs_flow_tbl_lookup(table, &key);
+ flow = ovs_flow_tbl_lookup(&dp->table, &key);
if (!flow || !ovs_flow_cmp_unmasked_key(flow, &match)) {
err = -ENOENT;
goto unlock;
@@ -1020,7 +978,7 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
goto unlock;
}
- ovs_flow_tbl_remove(table, flow);
+ ovs_flow_tbl_remove(&dp->table, flow);
err = ovs_flow_cmd_fill_info(flow, dp, reply, info->snd_portid,
info->snd_seq, 0, OVS_FLOW_CMD_DEL);
@@ -1039,8 +997,8 @@ unlock:
static int ovs_flow_cmd_dump(struct sk_buff *skb, struct netlink_callback *cb)
{
struct ovs_header *ovs_header = genlmsg_data(nlmsg_data(cb->nlh));
+ struct table_instance *ti;
struct datapath *dp;
- struct flow_table *table;
rcu_read_lock();
dp = get_dp(sock_net(skb->sk), ovs_header->dp_ifindex);
@@ -1049,14 +1007,14 @@ static int ovs_flow_cmd_dump(struct sk_buff *skb, struct netlink_callback *cb)
return -ENODEV;
}
- table = rcu_dereference(dp->table);
+ ti = rcu_dereference(dp->table.ti);
for (;;) {
struct sw_flow *flow;
u32 bucket, obj;
bucket = cb->args[0];
obj = cb->args[1];
- flow = ovs_flow_tbl_dump_next(table, &bucket, &obj);
+ flow = ovs_flow_tbl_dump_next(ti, &bucket, &obj);
if (!flow)
break;
@@ -1220,9 +1178,8 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
ovs_dp_set_net(dp, hold_net(sock_net(skb->sk)));
/* Allocate table. */
- err = -ENOMEM;
- rcu_assign_pointer(dp->table, ovs_flow_tbl_alloc(TBL_MIN_BUCKETS));
- if (!dp->table)
+ err = ovs_flow_tbl_init(&dp->table);
+ if (err)
goto err_free_dp;
dp->stats_percpu = alloc_percpu(struct dp_stats_percpu);
@@ -1279,7 +1236,7 @@ err_destroy_ports_array:
err_destroy_percpu:
free_percpu(dp->stats_percpu);
err_destroy_table:
- ovs_flow_tbl_destroy(ovsl_dereference(dp->table), false);
+ ovs_flow_tbl_destroy(&dp->table, false);
err_free_dp:
release_net(ovs_dp_get_net(dp));
kfree(dp);
diff --git a/net/openvswitch/datapath.h b/net/openvswitch/datapath.h
index a6982ef..acfd4af 100644
--- a/net/openvswitch/datapath.h
+++ b/net/openvswitch/datapath.h
@@ -58,12 +58,11 @@ struct dp_stats_percpu {
* struct datapath - datapath for flow-based packet switching
* @rcu: RCU callback head for deferred destruction.
* @list_node: Element in global 'dps' list.
- * @table: Current flow table. Protected by ovs_mutex and RCU.
+ * @table: flow table.
* @ports: Hash table for ports. %OVSP_LOCAL port always exists. Protected by
* ovs_mutex and RCU.
* @stats_percpu: Per-CPU datapath statistics.
* @net: Reference to net namespace.
- * @last_rehash: Timestamp of last rehash.
*
* Context: See the comment on locking at the top of datapath.c for additional
* locking information.
@@ -73,7 +72,7 @@ struct datapath {
struct list_head list_node;
/* Flow table. */
- struct flow_table __rcu *table;
+ struct flow_table table;
/* Switch ports. */
struct hlist_head *ports;
@@ -85,7 +84,6 @@ struct datapath {
/* Network namespace ref. */
struct net *net;
#endif
- unsigned long last_rehash;
};
/**
diff --git a/net/openvswitch/flow_table.c b/net/openvswitch/flow_table.c
index dcadb75..1c7e773 100644
--- a/net/openvswitch/flow_table.c
+++ b/net/openvswitch/flow_table.c
@@ -44,6 +44,11 @@
#include <net/ipv6.h>
#include <net/ndisc.h>
+#include "datapath.h"
+
+#define TBL_MIN_BUCKETS 1024
+#define REHASH_INTERVAL (10 * 60 * HZ)
+
static struct kmem_cache *flow_cache;
static u16 range_n_bytes(const struct sw_flow_key_range *range)
@@ -82,6 +87,11 @@ struct sw_flow *ovs_flow_alloc(void)
return flow;
}
+int ovs_flow_tbl_count(struct flow_table *table)
+{
+ return table->count;
+}
+
static struct flex_array *alloc_buckets(unsigned int n_buckets)
{
struct flex_array *buckets;
@@ -136,18 +146,18 @@ static void free_buckets(struct flex_array *buckets)
flex_array_free(buckets);
}
-static void __flow_tbl_destroy(struct flow_table *table)
+static void __table_instance_destroy(struct table_instance *ti)
{
int i;
- if (table->keep_flows)
+ if (ti->keep_flows)
goto skip_flows;
- for (i = 0; i < table->n_buckets; i++) {
+ for (i = 0; i < ti->n_buckets; i++) {
struct sw_flow *flow;
- struct hlist_head *head = flex_array_get(table->buckets, i);
+ struct hlist_head *head = flex_array_get(ti->buckets, i);
struct hlist_node *n;
- int ver = table->node_ver;
+ int ver = ti->node_ver;
hlist_for_each_entry_safe(flow, n, head, hash_node[ver]) {
hlist_del(&flow->hash_node[ver]);
@@ -155,74 +165,74 @@ static void __flow_tbl_destroy(struct flow_table *table)
}
}
- BUG_ON(!list_empty(table->mask_list));
- kfree(table->mask_list);
-
skip_flows:
- free_buckets(table->buckets);
- kfree(table);
+ free_buckets(ti->buckets);
+ kfree(ti);
}
-static struct flow_table *__flow_tbl_alloc(int new_size)
+static struct table_instance *table_instance_alloc(int new_size)
{
- struct flow_table *table = kmalloc(sizeof(*table), GFP_KERNEL);
+ struct table_instance *ti = kmalloc(sizeof(*ti), GFP_KERNEL);
- if (!table)
+ if (!ti)
return NULL;
- table->buckets = alloc_buckets(new_size);
+ ti->buckets = alloc_buckets(new_size);
- if (!table->buckets) {
- kfree(table);
+ if (!ti->buckets) {
+ kfree(ti);
return NULL;
}
- table->n_buckets = new_size;
- table->count = 0;
- table->node_ver = 0;
- table->keep_flows = false;
- get_random_bytes(&table->hash_seed, sizeof(u32));
- table->mask_list = NULL;
+ ti->n_buckets = new_size;
+ ti->node_ver = 0;
+ ti->keep_flows = false;
+ get_random_bytes(&ti->hash_seed, sizeof(u32));
- return table;
+ return ti;
}
-struct flow_table *ovs_flow_tbl_alloc(int new_size)
+int ovs_flow_tbl_init(struct flow_table *table)
{
- struct flow_table *table = __flow_tbl_alloc(new_size);
+ struct table_instance *ti;
- if (!table)
- return NULL;
+ ti = table_instance_alloc(TBL_MIN_BUCKETS);
- table->mask_list = kmalloc(sizeof(struct list_head), GFP_KERNEL);
- if (!table->mask_list) {
- table->keep_flows = true;
- __flow_tbl_destroy(table);
- return NULL;
- }
- INIT_LIST_HEAD(table->mask_list);
+ if (!ti)
+ return -ENOMEM;
- return table;
+ rcu_assign_pointer(table->ti, ti);
+ INIT_LIST_HEAD(&table->mask_list);
+ table->last_rehash = jiffies;
+ table->count = 0;
+ return 0;
}
static void flow_tbl_destroy_rcu_cb(struct rcu_head *rcu)
{
- struct flow_table *table = container_of(rcu, struct flow_table, rcu);
+ struct table_instance *ti = container_of(rcu, struct table_instance, rcu);
- __flow_tbl_destroy(table);
+ __table_instance_destroy(ti);
}
-void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred)
+static void table_instance_destroy(struct table_instance *ti, bool deferred)
{
- if (!table)
+ if (!ti)
return;
if (deferred)
- call_rcu(&table->rcu, flow_tbl_destroy_rcu_cb);
+ call_rcu(&ti->rcu, flow_tbl_destroy_rcu_cb);
else
- __flow_tbl_destroy(table);
+ __table_instance_destroy(ti);
+}
+
+void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred)
+{
+ struct table_instance *ti = ovsl_dereference(table->ti);
+
+ table_instance_destroy(ti, deferred);
}
-struct sw_flow *ovs_flow_tbl_dump_next(struct flow_table *table,
+struct sw_flow *ovs_flow_tbl_dump_next(struct table_instance *ti,
u32 *bucket, u32 *last)
{
struct sw_flow *flow;
@@ -230,10 +240,10 @@ struct sw_flow *ovs_flow_tbl_dump_next(struct flow_table *table,
int ver;
int i;
- ver = table->node_ver;
- while (*bucket < table->n_buckets) {
+ ver = ti->node_ver;
+ while (*bucket < ti->n_buckets) {
i = 0;
- head = flex_array_get(table->buckets, *bucket);
+ head = flex_array_get(ti->buckets, *bucket);
hlist_for_each_entry_rcu(flow, head, hash_node[ver]) {
if (i < *last) {
i++;
@@ -249,25 +259,23 @@ struct sw_flow *ovs_flow_tbl_dump_next(struct flow_table *table,
return NULL;
}
-static struct hlist_head *find_bucket(struct flow_table *table, u32 hash)
+static struct hlist_head *find_bucket(struct table_instance *ti, u32 hash)
{
- hash = jhash_1word(hash, table->hash_seed);
- return flex_array_get(table->buckets,
- (hash & (table->n_buckets - 1)));
+ hash = jhash_1word(hash, ti->hash_seed);
+ return flex_array_get(ti->buckets,
+ (hash & (ti->n_buckets - 1)));
}
-static void __tbl_insert(struct flow_table *table, struct sw_flow *flow)
+static void table_instance_insert(struct table_instance *ti, struct sw_flow *flow)
{
struct hlist_head *head;
- head = find_bucket(table, flow->hash);
- hlist_add_head_rcu(&flow->hash_node[table->node_ver], head);
-
- table->count++;
+ head = find_bucket(ti, flow->hash);
+ hlist_add_head_rcu(&flow->hash_node[ti->node_ver], head);
}
-static void flow_table_copy_flows(struct flow_table *old,
- struct flow_table *new)
+static void flow_table_copy_flows(struct table_instance *old,
+ struct table_instance *new)
{
int old_ver;
int i;
@@ -283,35 +291,42 @@ static void flow_table_copy_flows(struct flow_table *old,
head = flex_array_get(old->buckets, i);
hlist_for_each_entry(flow, head, hash_node[old_ver])
- __tbl_insert(new, flow);
+ table_instance_insert(new, flow);
}
- new->mask_list = old->mask_list;
old->keep_flows = true;
}
-static struct flow_table *__flow_tbl_rehash(struct flow_table *table,
+static struct table_instance *table_instance_rehash(struct table_instance *ti,
int n_buckets)
{
- struct flow_table *new_table;
+ struct table_instance *new_ti;
- new_table = __flow_tbl_alloc(n_buckets);
- if (!new_table)
+ new_ti = table_instance_alloc(n_buckets);
+ if (!new_ti)
return ERR_PTR(-ENOMEM);
- flow_table_copy_flows(table, new_table);
+ flow_table_copy_flows(ti, new_ti);
- return new_table;
+ return new_ti;
}
-struct flow_table *ovs_flow_tbl_rehash(struct flow_table *table)
+int ovs_flow_tbl_flush(struct flow_table *flow_table)
{
- return __flow_tbl_rehash(table, table->n_buckets);
-}
+ struct table_instance *old_ti;
+ struct table_instance *new_ti;
-struct flow_table *ovs_flow_tbl_expand(struct flow_table *table)
-{
- return __flow_tbl_rehash(table, table->n_buckets * 2);
+ old_ti = ovsl_dereference(flow_table->ti);
+ new_ti = table_instance_alloc(TBL_MIN_BUCKETS);
+ if (!new_ti)
+ return -ENOMEM;
+
+ rcu_assign_pointer(flow_table->ti, new_ti);
+ flow_table->last_rehash = jiffies;
+ flow_table->count = 0;
+
+ table_instance_destroy(old_ti, true);
+ return 0;
}
static u32 flow_hash(const struct sw_flow_key *key, int key_start,
@@ -367,7 +382,7 @@ bool ovs_flow_cmp_unmasked_key(const struct sw_flow *flow,
return cmp_key(&flow->unmasked_key, key, key_start, key_end);
}
-static struct sw_flow *masked_flow_lookup(struct flow_table *table,
+static struct sw_flow *masked_flow_lookup(struct table_instance *ti,
const struct sw_flow_key *unmasked,
struct sw_flow_mask *mask)
{
@@ -380,8 +395,8 @@ static struct sw_flow *masked_flow_lookup(struct flow_table *table,
ovs_flow_mask_key(&masked_key, unmasked, mask);
hash = flow_hash(&masked_key, key_start, key_end);
- head = find_bucket(table, hash);
- hlist_for_each_entry_rcu(flow, head, hash_node[table->node_ver]) {
+ head = find_bucket(ti, hash);
+ hlist_for_each_entry_rcu(flow, head, hash_node[ti->node_ver]) {
if (flow->mask == mask &&
flow_cmp_masked_key(flow, &masked_key,
key_start, key_end))
@@ -393,29 +408,55 @@ static struct sw_flow *masked_flow_lookup(struct flow_table *table,
struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *tbl,
const struct sw_flow_key *key)
{
- struct sw_flow *flow = NULL;
+ struct table_instance *ti = rcu_dereference(tbl->ti);
struct sw_flow_mask *mask;
+ struct sw_flow *flow;
- list_for_each_entry_rcu(mask, tbl->mask_list, list) {
- flow = masked_flow_lookup(tbl, key, mask);
+ list_for_each_entry_rcu(mask, &tbl->mask_list, list) {
+ flow = masked_flow_lookup(ti, key, mask);
if (flow) /* Found */
- break;
+ return flow;
}
+ return NULL;
+}
- return flow;
+static struct table_instance *table_instance_expand(struct table_instance *ti)
+{
+ return table_instance_rehash(ti, ti->n_buckets * 2);
}
void ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow)
{
+ struct table_instance *ti = NULL;
+ struct table_instance *new_ti = NULL;
+
+ ti = ovsl_dereference(table->ti);
+
+ /* Expand table, if necessary, to make room. */
+ if (table->count > ti->n_buckets)
+ new_ti = table_instance_expand(ti);
+ else if (time_after(jiffies, table->last_rehash + REHASH_INTERVAL))
+ new_ti = table_instance_rehash(ti, ti->n_buckets);
+
+ if (new_ti && !IS_ERR(new_ti)) {
+ rcu_assign_pointer(table->ti, new_ti);
+ ovs_flow_tbl_destroy(table, true);
+ ti = ovsl_dereference(table->ti);
+ table->last_rehash = jiffies;
+ }
+
flow->hash = flow_hash(&flow->key, flow->mask->range.start,
flow->mask->range.end);
- __tbl_insert(table, flow);
+ table_instance_insert(ti, flow);
+ table->count++;
}
void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow)
{
+ struct table_instance *ti = ovsl_dereference(table->ti);
+
BUG_ON(table->count == 0);
- hlist_del_rcu(&flow->hash_node[table->node_ver]);
+ hlist_del_rcu(&flow->hash_node[ti->node_ver]);
table->count--;
}
@@ -475,7 +516,7 @@ struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *tbl,
{
struct list_head *ml;
- list_for_each(ml, tbl->mask_list) {
+ list_for_each(ml, &tbl->mask_list) {
struct sw_flow_mask *m;
m = container_of(ml, struct sw_flow_mask, list);
if (mask_equal(mask, m))
@@ -492,7 +533,7 @@ struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *tbl,
*/
void ovs_sw_flow_mask_insert(struct flow_table *tbl, struct sw_flow_mask *mask)
{
- list_add_rcu(&mask->list, tbl->mask_list);
+ list_add_rcu(&mask->list, &tbl->mask_list);
}
/* Initializes the flow module.
diff --git a/net/openvswitch/flow_table.h b/net/openvswitch/flow_table.h
index d7a1144..5d1abe5 100644
--- a/net/openvswitch/flow_table.h
+++ b/net/openvswitch/flow_table.h
@@ -36,42 +36,36 @@
#include "flow.h"
-#define TBL_MIN_BUCKETS 1024
-
-struct flow_table {
+struct table_instance {
struct flex_array *buckets;
- unsigned int count, n_buckets;
+ unsigned int n_buckets;
struct rcu_head rcu;
- struct list_head *mask_list;
int node_ver;
u32 hash_seed;
bool keep_flows;
};
+struct flow_table {
+ struct table_instance __rcu *ti;
+ struct list_head mask_list;
+ unsigned long last_rehash;
+ unsigned int count;
+};
+
int ovs_flow_init(void);
void ovs_flow_exit(void);
struct sw_flow *ovs_flow_alloc(void);
void ovs_flow_free(struct sw_flow *, bool deferred);
-static inline int ovs_flow_tbl_count(struct flow_table *table)
-{
- return table->count;
-}
-
-static inline int ovs_flow_tbl_need_to_expand(struct flow_table *table)
-{
- return (table->count > table->n_buckets);
-}
-
-struct flow_table *ovs_flow_tbl_alloc(int new_size);
-struct flow_table *ovs_flow_tbl_expand(struct flow_table *table);
-struct flow_table *ovs_flow_tbl_rehash(struct flow_table *table);
+int ovs_flow_tbl_init(struct flow_table *);
+int ovs_flow_tbl_count(struct flow_table *table);
void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred);
+int ovs_flow_tbl_flush(struct flow_table *flow_table);
void ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow);
void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow);
-struct sw_flow *ovs_flow_tbl_dump_next(struct flow_table *table,
+struct sw_flow *ovs_flow_tbl_dump_next(struct table_instance *table,
u32 *bucket, u32 *idx);
struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *,
const struct sw_flow_key *);
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 04/11] openvswitch: Restructure datapath.c and flow.c
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Pravin B Shelar, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Pravin B Shelar <pshelar@nicira.com>
Over the time datapath.c and flow.c has became pretty large files.
Following patch restructures functionality of component into three
different components:
flow.c: contains flow extract.
flow_netlink.c: netlink flow api.
flow_table.c: flow table api.
This patch restructures code without changing logic.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
net/openvswitch/Makefile | 2 +
net/openvswitch/datapath.c | 528 +------------
net/openvswitch/datapath.h | 1 +
net/openvswitch/flow.c | 1605 +---------------------------------------
net/openvswitch/flow.h | 128 +---
net/openvswitch/flow_netlink.c | 1603 +++++++++++++++++++++++++++++++++++++++
net/openvswitch/flow_netlink.h | 60 ++
net/openvswitch/flow_table.c | 517 +++++++++++++
net/openvswitch/flow_table.h | 91 +++
9 files changed, 2354 insertions(+), 2181 deletions(-)
create mode 100644 net/openvswitch/flow_netlink.c
create mode 100644 net/openvswitch/flow_netlink.h
create mode 100644 net/openvswitch/flow_table.c
create mode 100644 net/openvswitch/flow_table.h
diff --git a/net/openvswitch/Makefile b/net/openvswitch/Makefile
index ea36e99..3591cb5 100644
--- a/net/openvswitch/Makefile
+++ b/net/openvswitch/Makefile
@@ -9,6 +9,8 @@ openvswitch-y := \
datapath.o \
dp_notify.o \
flow.o \
+ flow_netlink.o \
+ flow_table.o \
vport.o \
vport-internal_dev.o \
vport-netdev.o
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 2e1a9c2..72e6874 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -55,10 +55,10 @@
#include "datapath.h"
#include "flow.h"
+#include "flow_netlink.h"
#include "vport-internal_dev.h"
#include "vport-netdev.h"
-
#define REHASH_FLOW_INTERVAL (10 * 60 * HZ)
int ovs_net_id __read_mostly;
@@ -235,7 +235,7 @@ void ovs_dp_process_received_packet(struct vport *p, struct sk_buff *skb)
}
/* Look up flow. */
- flow = ovs_flow_lookup(rcu_dereference(dp->table), &key);
+ flow = ovs_flow_tbl_lookup(rcu_dereference(dp->table), &key);
if (unlikely(!flow)) {
struct dp_upcall_info upcall;
@@ -433,7 +433,7 @@ static int queue_userspace_packet(struct net *net, int dp_ifindex,
upcall->dp_ifindex = dp_ifindex;
nla = nla_nest_start(user_skb, OVS_PACKET_ATTR_KEY);
- ovs_flow_to_nlattrs(upcall_info->key, upcall_info->key, user_skb);
+ ovs_nla_put_flow(upcall_info->key, upcall_info->key, user_skb);
nla_nest_end(user_skb, nla);
if (upcall_info->userdata)
@@ -470,381 +470,6 @@ static int flush_flows(struct datapath *dp)
return 0;
}
-static struct nlattr *reserve_sfa_size(struct sw_flow_actions **sfa, int attr_len)
-{
-
- struct sw_flow_actions *acts;
- int new_acts_size;
- int req_size = NLA_ALIGN(attr_len);
- int next_offset = offsetof(struct sw_flow_actions, actions) +
- (*sfa)->actions_len;
-
- if (req_size <= (ksize(*sfa) - next_offset))
- goto out;
-
- new_acts_size = ksize(*sfa) * 2;
-
- if (new_acts_size > MAX_ACTIONS_BUFSIZE) {
- if ((MAX_ACTIONS_BUFSIZE - next_offset) < req_size)
- return ERR_PTR(-EMSGSIZE);
- new_acts_size = MAX_ACTIONS_BUFSIZE;
- }
-
- acts = ovs_flow_actions_alloc(new_acts_size);
- if (IS_ERR(acts))
- return (void *)acts;
-
- memcpy(acts->actions, (*sfa)->actions, (*sfa)->actions_len);
- acts->actions_len = (*sfa)->actions_len;
- kfree(*sfa);
- *sfa = acts;
-
-out:
- (*sfa)->actions_len += req_size;
- return (struct nlattr *) ((unsigned char *)(*sfa) + next_offset);
-}
-
-static int add_action(struct sw_flow_actions **sfa, int attrtype, void *data, int len)
-{
- struct nlattr *a;
-
- a = reserve_sfa_size(sfa, nla_attr_size(len));
- if (IS_ERR(a))
- return PTR_ERR(a);
-
- a->nla_type = attrtype;
- a->nla_len = nla_attr_size(len);
-
- if (data)
- memcpy(nla_data(a), data, len);
- memset((unsigned char *) a + a->nla_len, 0, nla_padlen(len));
-
- return 0;
-}
-
-static inline int add_nested_action_start(struct sw_flow_actions **sfa, int attrtype)
-{
- int used = (*sfa)->actions_len;
- int err;
-
- err = add_action(sfa, attrtype, NULL, 0);
- if (err)
- return err;
-
- return used;
-}
-
-static inline void add_nested_action_end(struct sw_flow_actions *sfa, int st_offset)
-{
- struct nlattr *a = (struct nlattr *) ((unsigned char *)sfa->actions + st_offset);
-
- a->nla_len = sfa->actions_len - st_offset;
-}
-
-static int validate_and_copy_actions(const struct nlattr *attr,
- const struct sw_flow_key *key, int depth,
- struct sw_flow_actions **sfa);
-
-static int validate_and_copy_sample(const struct nlattr *attr,
- const struct sw_flow_key *key, int depth,
- struct sw_flow_actions **sfa)
-{
- const struct nlattr *attrs[OVS_SAMPLE_ATTR_MAX + 1];
- const struct nlattr *probability, *actions;
- const struct nlattr *a;
- int rem, start, err, st_acts;
-
- memset(attrs, 0, sizeof(attrs));
- nla_for_each_nested(a, attr, rem) {
- int type = nla_type(a);
- if (!type || type > OVS_SAMPLE_ATTR_MAX || attrs[type])
- return -EINVAL;
- attrs[type] = a;
- }
- if (rem)
- return -EINVAL;
-
- probability = attrs[OVS_SAMPLE_ATTR_PROBABILITY];
- if (!probability || nla_len(probability) != sizeof(u32))
- return -EINVAL;
-
- actions = attrs[OVS_SAMPLE_ATTR_ACTIONS];
- if (!actions || (nla_len(actions) && nla_len(actions) < NLA_HDRLEN))
- return -EINVAL;
-
- /* validation done, copy sample action. */
- start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SAMPLE);
- if (start < 0)
- return start;
- err = add_action(sfa, OVS_SAMPLE_ATTR_PROBABILITY, nla_data(probability), sizeof(u32));
- if (err)
- return err;
- st_acts = add_nested_action_start(sfa, OVS_SAMPLE_ATTR_ACTIONS);
- if (st_acts < 0)
- return st_acts;
-
- err = validate_and_copy_actions(actions, key, depth + 1, sfa);
- if (err)
- return err;
-
- add_nested_action_end(*sfa, st_acts);
- add_nested_action_end(*sfa, start);
-
- return 0;
-}
-
-static int validate_tp_port(const struct sw_flow_key *flow_key)
-{
- if (flow_key->eth.type == htons(ETH_P_IP)) {
- if (flow_key->ipv4.tp.src || flow_key->ipv4.tp.dst)
- return 0;
- } else if (flow_key->eth.type == htons(ETH_P_IPV6)) {
- if (flow_key->ipv6.tp.src || flow_key->ipv6.tp.dst)
- return 0;
- }
-
- return -EINVAL;
-}
-
-static int validate_and_copy_set_tun(const struct nlattr *attr,
- struct sw_flow_actions **sfa)
-{
- struct sw_flow_match match;
- struct sw_flow_key key;
- int err, start;
-
- ovs_match_init(&match, &key, NULL);
- err = ovs_ipv4_tun_from_nlattr(nla_data(attr), &match, false);
- if (err)
- return err;
-
- start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SET);
- if (start < 0)
- return start;
-
- err = add_action(sfa, OVS_KEY_ATTR_IPV4_TUNNEL, &match.key->tun_key,
- sizeof(match.key->tun_key));
- add_nested_action_end(*sfa, start);
-
- return err;
-}
-
-static int validate_set(const struct nlattr *a,
- const struct sw_flow_key *flow_key,
- struct sw_flow_actions **sfa,
- bool *set_tun)
-{
- const struct nlattr *ovs_key = nla_data(a);
- int key_type = nla_type(ovs_key);
-
- /* There can be only one key in a action */
- if (nla_total_size(nla_len(ovs_key)) != nla_len(a))
- return -EINVAL;
-
- if (key_type > OVS_KEY_ATTR_MAX ||
- (ovs_key_lens[key_type] != nla_len(ovs_key) &&
- ovs_key_lens[key_type] != -1))
- return -EINVAL;
-
- switch (key_type) {
- const struct ovs_key_ipv4 *ipv4_key;
- const struct ovs_key_ipv6 *ipv6_key;
- int err;
-
- case OVS_KEY_ATTR_PRIORITY:
- case OVS_KEY_ATTR_SKB_MARK:
- case OVS_KEY_ATTR_ETHERNET:
- break;
-
- case OVS_KEY_ATTR_TUNNEL:
- *set_tun = true;
- err = validate_and_copy_set_tun(a, sfa);
- if (err)
- return err;
- break;
-
- case OVS_KEY_ATTR_IPV4:
- if (flow_key->eth.type != htons(ETH_P_IP))
- return -EINVAL;
-
- if (!flow_key->ip.proto)
- return -EINVAL;
-
- ipv4_key = nla_data(ovs_key);
- if (ipv4_key->ipv4_proto != flow_key->ip.proto)
- return -EINVAL;
-
- if (ipv4_key->ipv4_frag != flow_key->ip.frag)
- return -EINVAL;
-
- break;
-
- case OVS_KEY_ATTR_IPV6:
- if (flow_key->eth.type != htons(ETH_P_IPV6))
- return -EINVAL;
-
- if (!flow_key->ip.proto)
- return -EINVAL;
-
- ipv6_key = nla_data(ovs_key);
- if (ipv6_key->ipv6_proto != flow_key->ip.proto)
- return -EINVAL;
-
- if (ipv6_key->ipv6_frag != flow_key->ip.frag)
- return -EINVAL;
-
- if (ntohl(ipv6_key->ipv6_label) & 0xFFF00000)
- return -EINVAL;
-
- break;
-
- case OVS_KEY_ATTR_TCP:
- if (flow_key->ip.proto != IPPROTO_TCP)
- return -EINVAL;
-
- return validate_tp_port(flow_key);
-
- case OVS_KEY_ATTR_UDP:
- if (flow_key->ip.proto != IPPROTO_UDP)
- return -EINVAL;
-
- return validate_tp_port(flow_key);
-
- case OVS_KEY_ATTR_SCTP:
- if (flow_key->ip.proto != IPPROTO_SCTP)
- return -EINVAL;
-
- return validate_tp_port(flow_key);
-
- default:
- return -EINVAL;
- }
-
- return 0;
-}
-
-static int validate_userspace(const struct nlattr *attr)
-{
- static const struct nla_policy userspace_policy[OVS_USERSPACE_ATTR_MAX + 1] = {
- [OVS_USERSPACE_ATTR_PID] = {.type = NLA_U32 },
- [OVS_USERSPACE_ATTR_USERDATA] = {.type = NLA_UNSPEC },
- };
- struct nlattr *a[OVS_USERSPACE_ATTR_MAX + 1];
- int error;
-
- error = nla_parse_nested(a, OVS_USERSPACE_ATTR_MAX,
- attr, userspace_policy);
- if (error)
- return error;
-
- if (!a[OVS_USERSPACE_ATTR_PID] ||
- !nla_get_u32(a[OVS_USERSPACE_ATTR_PID]))
- return -EINVAL;
-
- return 0;
-}
-
-static int copy_action(const struct nlattr *from,
- struct sw_flow_actions **sfa)
-{
- int totlen = NLA_ALIGN(from->nla_len);
- struct nlattr *to;
-
- to = reserve_sfa_size(sfa, from->nla_len);
- if (IS_ERR(to))
- return PTR_ERR(to);
-
- memcpy(to, from, totlen);
- return 0;
-}
-
-static int validate_and_copy_actions(const struct nlattr *attr,
- const struct sw_flow_key *key,
- int depth,
- struct sw_flow_actions **sfa)
-{
- const struct nlattr *a;
- int rem, err;
-
- if (depth >= SAMPLE_ACTION_DEPTH)
- return -EOVERFLOW;
-
- nla_for_each_nested(a, attr, rem) {
- /* Expected argument lengths, (u32)-1 for variable length. */
- static const u32 action_lens[OVS_ACTION_ATTR_MAX + 1] = {
- [OVS_ACTION_ATTR_OUTPUT] = sizeof(u32),
- [OVS_ACTION_ATTR_USERSPACE] = (u32)-1,
- [OVS_ACTION_ATTR_PUSH_VLAN] = sizeof(struct ovs_action_push_vlan),
- [OVS_ACTION_ATTR_POP_VLAN] = 0,
- [OVS_ACTION_ATTR_SET] = (u32)-1,
- [OVS_ACTION_ATTR_SAMPLE] = (u32)-1
- };
- const struct ovs_action_push_vlan *vlan;
- int type = nla_type(a);
- bool skip_copy;
-
- if (type > OVS_ACTION_ATTR_MAX ||
- (action_lens[type] != nla_len(a) &&
- action_lens[type] != (u32)-1))
- return -EINVAL;
-
- skip_copy = false;
- switch (type) {
- case OVS_ACTION_ATTR_UNSPEC:
- return -EINVAL;
-
- case OVS_ACTION_ATTR_USERSPACE:
- err = validate_userspace(a);
- if (err)
- return err;
- break;
-
- case OVS_ACTION_ATTR_OUTPUT:
- if (nla_get_u32(a) >= DP_MAX_PORTS)
- return -EINVAL;
- break;
-
-
- case OVS_ACTION_ATTR_POP_VLAN:
- break;
-
- case OVS_ACTION_ATTR_PUSH_VLAN:
- vlan = nla_data(a);
- if (vlan->vlan_tpid != htons(ETH_P_8021Q))
- return -EINVAL;
- if (!(vlan->vlan_tci & htons(VLAN_TAG_PRESENT)))
- return -EINVAL;
- break;
-
- case OVS_ACTION_ATTR_SET:
- err = validate_set(a, key, sfa, &skip_copy);
- if (err)
- return err;
- break;
-
- case OVS_ACTION_ATTR_SAMPLE:
- err = validate_and_copy_sample(a, key, depth, sfa);
- if (err)
- return err;
- skip_copy = true;
- break;
-
- default:
- return -EINVAL;
- }
- if (!skip_copy) {
- err = copy_action(a, sfa);
- if (err)
- return err;
- }
- }
-
- if (rem > 0)
- return -EINVAL;
-
- return 0;
-}
-
static void clear_stats(struct sw_flow *flow)
{
flow->used = 0;
@@ -900,15 +525,16 @@ static int ovs_packet_cmd_execute(struct sk_buff *skb, struct genl_info *info)
if (err)
goto err_flow_free;
- err = ovs_flow_metadata_from_nlattrs(flow, a[OVS_PACKET_ATTR_KEY]);
+ err = ovs_nla_get_flow_metadata(flow, a[OVS_PACKET_ATTR_KEY]);
if (err)
goto err_flow_free;
- acts = ovs_flow_actions_alloc(nla_len(a[OVS_PACKET_ATTR_ACTIONS]));
+ acts = ovs_nla_alloc_flow_actions(nla_len(a[OVS_PACKET_ATTR_ACTIONS]));
err = PTR_ERR(acts);
if (IS_ERR(acts))
goto err_flow_free;
- err = validate_and_copy_actions(a[OVS_PACKET_ATTR_ACTIONS], &flow->key, 0, &acts);
+ err = ovs_nla_copy_actions(a[OVS_PACKET_ATTR_ACTIONS],
+ &flow->key, 0, &acts);
rcu_assign_pointer(flow->sf_acts, acts);
if (err)
goto err_flow_free;
@@ -1003,100 +629,6 @@ static struct genl_multicast_group ovs_dp_flow_multicast_group = {
.name = OVS_FLOW_MCGROUP
};
-static int actions_to_attr(const struct nlattr *attr, int len, struct sk_buff *skb);
-static int sample_action_to_attr(const struct nlattr *attr, struct sk_buff *skb)
-{
- const struct nlattr *a;
- struct nlattr *start;
- int err = 0, rem;
-
- start = nla_nest_start(skb, OVS_ACTION_ATTR_SAMPLE);
- if (!start)
- return -EMSGSIZE;
-
- nla_for_each_nested(a, attr, rem) {
- int type = nla_type(a);
- struct nlattr *st_sample;
-
- switch (type) {
- case OVS_SAMPLE_ATTR_PROBABILITY:
- if (nla_put(skb, OVS_SAMPLE_ATTR_PROBABILITY, sizeof(u32), nla_data(a)))
- return -EMSGSIZE;
- break;
- case OVS_SAMPLE_ATTR_ACTIONS:
- st_sample = nla_nest_start(skb, OVS_SAMPLE_ATTR_ACTIONS);
- if (!st_sample)
- return -EMSGSIZE;
- err = actions_to_attr(nla_data(a), nla_len(a), skb);
- if (err)
- return err;
- nla_nest_end(skb, st_sample);
- break;
- }
- }
-
- nla_nest_end(skb, start);
- return err;
-}
-
-static int set_action_to_attr(const struct nlattr *a, struct sk_buff *skb)
-{
- const struct nlattr *ovs_key = nla_data(a);
- int key_type = nla_type(ovs_key);
- struct nlattr *start;
- int err;
-
- switch (key_type) {
- case OVS_KEY_ATTR_IPV4_TUNNEL:
- start = nla_nest_start(skb, OVS_ACTION_ATTR_SET);
- if (!start)
- return -EMSGSIZE;
-
- err = ovs_ipv4_tun_to_nlattr(skb, nla_data(ovs_key),
- nla_data(ovs_key));
- if (err)
- return err;
- nla_nest_end(skb, start);
- break;
- default:
- if (nla_put(skb, OVS_ACTION_ATTR_SET, nla_len(a), ovs_key))
- return -EMSGSIZE;
- break;
- }
-
- return 0;
-}
-
-static int actions_to_attr(const struct nlattr *attr, int len, struct sk_buff *skb)
-{
- const struct nlattr *a;
- int rem, err;
-
- nla_for_each_attr(a, attr, len, rem) {
- int type = nla_type(a);
-
- switch (type) {
- case OVS_ACTION_ATTR_SET:
- err = set_action_to_attr(a, skb);
- if (err)
- return err;
- break;
-
- case OVS_ACTION_ATTR_SAMPLE:
- err = sample_action_to_attr(a, skb);
- if (err)
- return err;
- break;
- default:
- if (nla_put(skb, type, nla_len(a), nla_data(a)))
- return -EMSGSIZE;
- break;
- }
- }
-
- return 0;
-}
-
static size_t ovs_flow_cmd_msg_size(const struct sw_flow_actions *acts)
{
return NLMSG_ALIGN(sizeof(struct ovs_header))
@@ -1133,8 +665,7 @@ static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
if (!nla)
goto nla_put_failure;
- err = ovs_flow_to_nlattrs(&flow->unmasked_key,
- &flow->unmasked_key, skb);
+ err = ovs_nla_put_flow(&flow->unmasked_key, &flow->unmasked_key, skb);
if (err)
goto error;
nla_nest_end(skb, nla);
@@ -1143,7 +674,7 @@ static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
if (!nla)
goto nla_put_failure;
- err = ovs_flow_to_nlattrs(&flow->key, &flow->mask->key, skb);
+ err = ovs_nla_put_flow(&flow->key, &flow->mask->key, skb);
if (err)
goto error;
@@ -1186,7 +717,8 @@ static int ovs_flow_cmd_fill_info(struct sw_flow *flow, struct datapath *dp,
sf_acts = rcu_dereference_check(flow->sf_acts,
lockdep_ovsl_is_held());
- err = actions_to_attr(sf_acts->actions, sf_acts->actions_len, skb);
+ err = ovs_nla_put_actions(sf_acts->actions,
+ sf_acts->actions_len, skb);
if (!err)
nla_nest_end(skb, start);
else {
@@ -1252,21 +784,21 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
goto error;
ovs_match_init(&match, &key, &mask);
- error = ovs_match_from_nlattrs(&match,
- a[OVS_FLOW_ATTR_KEY], a[OVS_FLOW_ATTR_MASK]);
+ error = ovs_nla_get_match(&match,
+ a[OVS_FLOW_ATTR_KEY], a[OVS_FLOW_ATTR_MASK]);
if (error)
goto error;
/* Validate actions. */
if (a[OVS_FLOW_ATTR_ACTIONS]) {
- acts = ovs_flow_actions_alloc(nla_len(a[OVS_FLOW_ATTR_ACTIONS]));
+ acts = ovs_nla_alloc_flow_actions(nla_len(a[OVS_FLOW_ATTR_ACTIONS]));
error = PTR_ERR(acts);
if (IS_ERR(acts))
goto error;
- ovs_flow_key_mask(&masked_key, &key, &mask);
- error = validate_and_copy_actions(a[OVS_FLOW_ATTR_ACTIONS],
- &masked_key, 0, &acts);
+ ovs_flow_mask_key(&masked_key, &key, &mask);
+ error = ovs_nla_copy_actions(a[OVS_FLOW_ATTR_ACTIONS],
+ &masked_key, 0, &acts);
if (error) {
OVS_NLERR("Flow actions may not be safe on all matching packets.\n");
goto err_kfree;
@@ -1285,7 +817,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
table = ovsl_dereference(dp->table);
/* Check if this is a duplicate flow */
- flow = ovs_flow_lookup(table, &key);
+ flow = ovs_flow_tbl_lookup(table, &key);
if (!flow) {
struct flow_table *new_table = NULL;
struct sw_flow_mask *mask_p;
@@ -1336,7 +868,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
rcu_assign_pointer(flow->sf_acts, acts);
/* Put flow in bucket. */
- ovs_flow_insert(table, flow);
+ ovs_flow_tbl_insert(table, flow);
reply = ovs_flow_cmd_build_info(flow, dp, info->snd_portid,
info->snd_seq, OVS_FLOW_CMD_NEW);
@@ -1357,7 +889,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
/* The unmasked key has to be the same for flow updates. */
error = -EINVAL;
- if (!ovs_flow_cmp_unmasked_key(flow, &key, match.range.end)) {
+ if (!ovs_flow_cmp_unmasked_key(flow, &match)) {
OVS_NLERR("Flow modification message rejected, unmasked key does not match.\n");
goto err_unlock_ovs;
}
@@ -1365,7 +897,7 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
/* Update actions. */
old_acts = ovsl_dereference(flow->sf_acts);
rcu_assign_pointer(flow->sf_acts, acts);
- ovs_flow_deferred_free_acts(old_acts);
+ ovs_nla_free_flow_actions(old_acts);
reply = ovs_flow_cmd_build_info(flow, dp, info->snd_portid,
info->snd_seq, OVS_FLOW_CMD_NEW);
@@ -1414,7 +946,7 @@ static int ovs_flow_cmd_get(struct sk_buff *skb, struct genl_info *info)
}
ovs_match_init(&match, &key, NULL);
- err = ovs_match_from_nlattrs(&match, a[OVS_FLOW_ATTR_KEY], NULL);
+ err = ovs_nla_get_match(&match, a[OVS_FLOW_ATTR_KEY], NULL);
if (err)
return err;
@@ -1426,8 +958,8 @@ static int ovs_flow_cmd_get(struct sk_buff *skb, struct genl_info *info)
}
table = ovsl_dereference(dp->table);
- flow = ovs_flow_lookup_unmasked_key(table, &match);
- if (!flow) {
+ flow = ovs_flow_tbl_lookup(table, &key);
+ if (!flow || !ovs_flow_cmp_unmasked_key(flow, &match)) {
err = -ENOENT;
goto unlock;
}
@@ -1471,13 +1003,13 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
}
ovs_match_init(&match, &key, NULL);
- err = ovs_match_from_nlattrs(&match, a[OVS_FLOW_ATTR_KEY], NULL);
+ err = ovs_nla_get_match(&match, a[OVS_FLOW_ATTR_KEY], NULL);
if (err)
goto unlock;
table = ovsl_dereference(dp->table);
- flow = ovs_flow_lookup_unmasked_key(table, &match);
- if (!flow) {
+ flow = ovs_flow_tbl_lookup(table, &key);
+ if (!flow || !ovs_flow_cmp_unmasked_key(flow, &match)) {
err = -ENOENT;
goto unlock;
}
@@ -1488,7 +1020,7 @@ static int ovs_flow_cmd_del(struct sk_buff *skb, struct genl_info *info)
goto unlock;
}
- ovs_flow_remove(table, flow);
+ ovs_flow_tbl_remove(table, flow);
err = ovs_flow_cmd_fill_info(flow, dp, reply, info->snd_portid,
info->snd_seq, 0, OVS_FLOW_CMD_DEL);
@@ -1524,7 +1056,7 @@ static int ovs_flow_cmd_dump(struct sk_buff *skb, struct netlink_callback *cb)
bucket = cb->args[0];
obj = cb->args[1];
- flow = ovs_flow_dump_next(table, &bucket, &obj);
+ flow = ovs_flow_tbl_dump_next(table, &bucket, &obj);
if (!flow)
break;
@@ -1700,7 +1232,7 @@ static int ovs_dp_cmd_new(struct sk_buff *skb, struct genl_info *info)
}
dp->ports = kmalloc(DP_VPORT_HASH_BUCKETS * sizeof(struct hlist_head),
- GFP_KERNEL);
+ GFP_KERNEL);
if (!dp->ports) {
err = -ENOMEM;
goto err_destroy_percpu;
diff --git a/net/openvswitch/datapath.h b/net/openvswitch/datapath.h
index 2c15541..a6982ef 100644
--- a/net/openvswitch/datapath.h
+++ b/net/openvswitch/datapath.h
@@ -27,6 +27,7 @@
#include <linux/u64_stats_sync.h>
#include "flow.h"
+#include "flow_table.h"
#include "vport.h"
#define DP_MAX_PORTS USHRT_MAX
diff --git a/net/openvswitch/flow.c b/net/openvswitch/flow.c
index 410db90..617810f 100644
--- a/net/openvswitch/flow.c
+++ b/net/openvswitch/flow.c
@@ -45,202 +45,40 @@
#include <net/ipv6.h>
#include <net/ndisc.h>
-static struct kmem_cache *flow_cache;
-
-static void ovs_sw_flow_mask_set(struct sw_flow_mask *mask,
- struct sw_flow_key_range *range, u8 val);
-
-static void update_range__(struct sw_flow_match *match,
- size_t offset, size_t size, bool is_mask)
+u64 ovs_flow_used_time(unsigned long flow_jiffies)
{
- struct sw_flow_key_range *range = NULL;
- size_t start = rounddown(offset, sizeof(long));
- size_t end = roundup(offset + size, sizeof(long));
-
- if (!is_mask)
- range = &match->range;
- else if (match->mask)
- range = &match->mask->range;
-
- if (!range)
- return;
-
- if (range->start == range->end) {
- range->start = start;
- range->end = end;
- return;
- }
-
- if (range->start > start)
- range->start = start;
+ struct timespec cur_ts;
+ u64 cur_ms, idle_ms;
- if (range->end < end)
- range->end = end;
-}
+ ktime_get_ts(&cur_ts);
+ idle_ms = jiffies_to_msecs(jiffies - flow_jiffies);
+ cur_ms = (u64)cur_ts.tv_sec * MSEC_PER_SEC +
+ cur_ts.tv_nsec / NSEC_PER_MSEC;
-#define SW_FLOW_KEY_PUT(match, field, value, is_mask) \
- do { \
- update_range__(match, offsetof(struct sw_flow_key, field), \
- sizeof((match)->key->field), is_mask); \
- if (is_mask) { \
- if ((match)->mask) \
- (match)->mask->key.field = value; \
- } else { \
- (match)->key->field = value; \
- } \
- } while (0)
-
-#define SW_FLOW_KEY_MEMCPY(match, field, value_p, len, is_mask) \
- do { \
- update_range__(match, offsetof(struct sw_flow_key, field), \
- len, is_mask); \
- if (is_mask) { \
- if ((match)->mask) \
- memcpy(&(match)->mask->key.field, value_p, len);\
- } else { \
- memcpy(&(match)->key->field, value_p, len); \
- } \
- } while (0)
-
-static u16 range_n_bytes(const struct sw_flow_key_range *range)
-{
- return range->end - range->start;
+ return cur_ms - idle_ms;
}
-void ovs_match_init(struct sw_flow_match *match,
- struct sw_flow_key *key,
- struct sw_flow_mask *mask)
-{
- memset(match, 0, sizeof(*match));
- match->key = key;
- match->mask = mask;
-
- memset(key, 0, sizeof(*key));
-
- if (mask) {
- memset(&mask->key, 0, sizeof(mask->key));
- mask->range.start = mask->range.end = 0;
- }
-}
+#define TCP_FLAGS_OFFSET 13
+#define TCP_FLAG_MASK 0x3f
-static bool ovs_match_validate(const struct sw_flow_match *match,
- u64 key_attrs, u64 mask_attrs)
+void ovs_flow_used(struct sw_flow *flow, struct sk_buff *skb)
{
- u64 key_expected = 1 << OVS_KEY_ATTR_ETHERNET;
- u64 mask_allowed = key_attrs; /* At most allow all key attributes */
-
- /* The following mask attributes allowed only if they
- * pass the validation tests. */
- mask_allowed &= ~((1 << OVS_KEY_ATTR_IPV4)
- | (1 << OVS_KEY_ATTR_IPV6)
- | (1 << OVS_KEY_ATTR_TCP)
- | (1 << OVS_KEY_ATTR_UDP)
- | (1 << OVS_KEY_ATTR_SCTP)
- | (1 << OVS_KEY_ATTR_ICMP)
- | (1 << OVS_KEY_ATTR_ICMPV6)
- | (1 << OVS_KEY_ATTR_ARP)
- | (1 << OVS_KEY_ATTR_ND));
-
- /* Always allowed mask fields. */
- mask_allowed |= ((1 << OVS_KEY_ATTR_TUNNEL)
- | (1 << OVS_KEY_ATTR_IN_PORT)
- | (1 << OVS_KEY_ATTR_ETHERTYPE));
-
- /* Check key attributes. */
- if (match->key->eth.type == htons(ETH_P_ARP)
- || match->key->eth.type == htons(ETH_P_RARP)) {
- key_expected |= 1 << OVS_KEY_ATTR_ARP;
- if (match->mask && (match->mask->key.eth.type == htons(0xffff)))
- mask_allowed |= 1 << OVS_KEY_ATTR_ARP;
- }
-
- if (match->key->eth.type == htons(ETH_P_IP)) {
- key_expected |= 1 << OVS_KEY_ATTR_IPV4;
- if (match->mask && (match->mask->key.eth.type == htons(0xffff)))
- mask_allowed |= 1 << OVS_KEY_ATTR_IPV4;
-
- if (match->key->ip.frag != OVS_FRAG_TYPE_LATER) {
- if (match->key->ip.proto == IPPROTO_UDP) {
- key_expected |= 1 << OVS_KEY_ATTR_UDP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
- mask_allowed |= 1 << OVS_KEY_ATTR_UDP;
- }
-
- if (match->key->ip.proto == IPPROTO_SCTP) {
- key_expected |= 1 << OVS_KEY_ATTR_SCTP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
- mask_allowed |= 1 << OVS_KEY_ATTR_SCTP;
- }
-
- if (match->key->ip.proto == IPPROTO_TCP) {
- key_expected |= 1 << OVS_KEY_ATTR_TCP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
- mask_allowed |= 1 << OVS_KEY_ATTR_TCP;
- }
-
- if (match->key->ip.proto == IPPROTO_ICMP) {
- key_expected |= 1 << OVS_KEY_ATTR_ICMP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
- mask_allowed |= 1 << OVS_KEY_ATTR_ICMP;
- }
- }
- }
-
- if (match->key->eth.type == htons(ETH_P_IPV6)) {
- key_expected |= 1 << OVS_KEY_ATTR_IPV6;
- if (match->mask && (match->mask->key.eth.type == htons(0xffff)))
- mask_allowed |= 1 << OVS_KEY_ATTR_IPV6;
-
- if (match->key->ip.frag != OVS_FRAG_TYPE_LATER) {
- if (match->key->ip.proto == IPPROTO_UDP) {
- key_expected |= 1 << OVS_KEY_ATTR_UDP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
- mask_allowed |= 1 << OVS_KEY_ATTR_UDP;
- }
-
- if (match->key->ip.proto == IPPROTO_SCTP) {
- key_expected |= 1 << OVS_KEY_ATTR_SCTP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
- mask_allowed |= 1 << OVS_KEY_ATTR_SCTP;
- }
-
- if (match->key->ip.proto == IPPROTO_TCP) {
- key_expected |= 1 << OVS_KEY_ATTR_TCP;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
- mask_allowed |= 1 << OVS_KEY_ATTR_TCP;
- }
-
- if (match->key->ip.proto == IPPROTO_ICMPV6) {
- key_expected |= 1 << OVS_KEY_ATTR_ICMPV6;
- if (match->mask && (match->mask->key.ip.proto == 0xff))
- mask_allowed |= 1 << OVS_KEY_ATTR_ICMPV6;
-
- if (match->key->ipv6.tp.src ==
- htons(NDISC_NEIGHBOUR_SOLICITATION) ||
- match->key->ipv6.tp.src == htons(NDISC_NEIGHBOUR_ADVERTISEMENT)) {
- key_expected |= 1 << OVS_KEY_ATTR_ND;
- if (match->mask && (match->mask->key.ipv6.tp.src == htons(0xffff)))
- mask_allowed |= 1 << OVS_KEY_ATTR_ND;
- }
- }
- }
- }
-
- if ((key_attrs & key_expected) != key_expected) {
- /* Key attributes check failed. */
- OVS_NLERR("Missing expected key attributes (key_attrs=%llx, expected=%llx).\n",
- key_attrs, key_expected);
- return false;
- }
+ u8 tcp_flags = 0;
- if ((mask_attrs & mask_allowed) != mask_attrs) {
- /* Mask attributes check failed. */
- OVS_NLERR("Contain more than allowed mask fields (mask_attrs=%llx, mask_allowed=%llx).\n",
- mask_attrs, mask_allowed);
- return false;
+ if ((flow->key.eth.type == htons(ETH_P_IP) ||
+ flow->key.eth.type == htons(ETH_P_IPV6)) &&
+ flow->key.ip.proto == IPPROTO_TCP &&
+ likely(skb->len >= skb_transport_offset(skb) + sizeof(struct tcphdr))) {
+ u8 *tcp = (u8 *)tcp_hdr(skb);
+ tcp_flags = *(tcp + TCP_FLAGS_OFFSET) & TCP_FLAG_MASK;
}
- return true;
+ spin_lock(&flow->lock);
+ flow->used = jiffies;
+ flow->packet_count++;
+ flow->byte_count += skb->len;
+ flow->tcp_flags |= tcp_flags;
+ spin_unlock(&flow->lock);
}
static int check_header(struct sk_buff *skb, int len)
@@ -311,19 +149,6 @@ static bool icmphdr_ok(struct sk_buff *skb)
sizeof(struct icmphdr));
}
-u64 ovs_flow_used_time(unsigned long flow_jiffies)
-{
- struct timespec cur_ts;
- u64 cur_ms, idle_ms;
-
- ktime_get_ts(&cur_ts);
- idle_ms = jiffies_to_msecs(jiffies - flow_jiffies);
- cur_ms = (u64)cur_ts.tv_sec * MSEC_PER_SEC +
- cur_ts.tv_nsec / NSEC_PER_MSEC;
-
- return cur_ms - idle_ms;
-}
-
static int parse_ipv6hdr(struct sk_buff *skb, struct sw_flow_key *key)
{
unsigned int nh_ofs = skb_network_offset(skb);
@@ -372,311 +197,6 @@ static bool icmp6hdr_ok(struct sk_buff *skb)
sizeof(struct icmp6hdr));
}
-void ovs_flow_key_mask(struct sw_flow_key *dst, const struct sw_flow_key *src,
- const struct sw_flow_mask *mask)
-{
- const long *m = (long *)((u8 *)&mask->key + mask->range.start);
- const long *s = (long *)((u8 *)src + mask->range.start);
- long *d = (long *)((u8 *)dst + mask->range.start);
- int i;
-
- /* The memory outside of the 'mask->range' are not set since
- * further operations on 'dst' only uses contents within
- * 'mask->range'.
- */
- for (i = 0; i < range_n_bytes(&mask->range); i += sizeof(long))
- *d++ = *s++ & *m++;
-}
-
-#define TCP_FLAGS_OFFSET 13
-#define TCP_FLAG_MASK 0x3f
-
-void ovs_flow_used(struct sw_flow *flow, struct sk_buff *skb)
-{
- u8 tcp_flags = 0;
-
- if ((flow->key.eth.type == htons(ETH_P_IP) ||
- flow->key.eth.type == htons(ETH_P_IPV6)) &&
- flow->key.ip.proto == IPPROTO_TCP &&
- likely(skb->len >= skb_transport_offset(skb) + sizeof(struct tcphdr))) {
- u8 *tcp = (u8 *)tcp_hdr(skb);
- tcp_flags = *(tcp + TCP_FLAGS_OFFSET) & TCP_FLAG_MASK;
- }
-
- spin_lock(&flow->lock);
- flow->used = jiffies;
- flow->packet_count++;
- flow->byte_count += skb->len;
- flow->tcp_flags |= tcp_flags;
- spin_unlock(&flow->lock);
-}
-
-struct sw_flow_actions *ovs_flow_actions_alloc(int size)
-{
- struct sw_flow_actions *sfa;
-
- if (size > MAX_ACTIONS_BUFSIZE)
- return ERR_PTR(-EINVAL);
-
- sfa = kmalloc(sizeof(*sfa) + size, GFP_KERNEL);
- if (!sfa)
- return ERR_PTR(-ENOMEM);
-
- sfa->actions_len = 0;
- return sfa;
-}
-
-struct sw_flow *ovs_flow_alloc(void)
-{
- struct sw_flow *flow;
-
- flow = kmem_cache_alloc(flow_cache, GFP_KERNEL);
- if (!flow)
- return ERR_PTR(-ENOMEM);
-
- spin_lock_init(&flow->lock);
- flow->sf_acts = NULL;
- flow->mask = NULL;
-
- return flow;
-}
-
-static struct hlist_head *find_bucket(struct flow_table *table, u32 hash)
-{
- hash = jhash_1word(hash, table->hash_seed);
- return flex_array_get(table->buckets,
- (hash & (table->n_buckets - 1)));
-}
-
-static struct flex_array *alloc_buckets(unsigned int n_buckets)
-{
- struct flex_array *buckets;
- int i, err;
-
- buckets = flex_array_alloc(sizeof(struct hlist_head),
- n_buckets, GFP_KERNEL);
- if (!buckets)
- return NULL;
-
- err = flex_array_prealloc(buckets, 0, n_buckets, GFP_KERNEL);
- if (err) {
- flex_array_free(buckets);
- return NULL;
- }
-
- for (i = 0; i < n_buckets; i++)
- INIT_HLIST_HEAD((struct hlist_head *)
- flex_array_get(buckets, i));
-
- return buckets;
-}
-
-static void free_buckets(struct flex_array *buckets)
-{
- flex_array_free(buckets);
-}
-
-static struct flow_table *__flow_tbl_alloc(int new_size)
-{
- struct flow_table *table = kmalloc(sizeof(*table), GFP_KERNEL);
-
- if (!table)
- return NULL;
-
- table->buckets = alloc_buckets(new_size);
-
- if (!table->buckets) {
- kfree(table);
- return NULL;
- }
- table->n_buckets = new_size;
- table->count = 0;
- table->node_ver = 0;
- table->keep_flows = false;
- get_random_bytes(&table->hash_seed, sizeof(u32));
- table->mask_list = NULL;
-
- return table;
-}
-
-static void __flow_tbl_destroy(struct flow_table *table)
-{
- int i;
-
- if (table->keep_flows)
- goto skip_flows;
-
- for (i = 0; i < table->n_buckets; i++) {
- struct sw_flow *flow;
- struct hlist_head *head = flex_array_get(table->buckets, i);
- struct hlist_node *n;
- int ver = table->node_ver;
-
- hlist_for_each_entry_safe(flow, n, head, hash_node[ver]) {
- hlist_del(&flow->hash_node[ver]);
- ovs_flow_free(flow, false);
- }
- }
-
- BUG_ON(!list_empty(table->mask_list));
- kfree(table->mask_list);
-
-skip_flows:
- free_buckets(table->buckets);
- kfree(table);
-}
-
-struct flow_table *ovs_flow_tbl_alloc(int new_size)
-{
- struct flow_table *table = __flow_tbl_alloc(new_size);
-
- if (!table)
- return NULL;
-
- table->mask_list = kmalloc(sizeof(struct list_head), GFP_KERNEL);
- if (!table->mask_list) {
- table->keep_flows = true;
- __flow_tbl_destroy(table);
- return NULL;
- }
- INIT_LIST_HEAD(table->mask_list);
-
- return table;
-}
-
-static void flow_tbl_destroy_rcu_cb(struct rcu_head *rcu)
-{
- struct flow_table *table = container_of(rcu, struct flow_table, rcu);
-
- __flow_tbl_destroy(table);
-}
-
-void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred)
-{
- if (!table)
- return;
-
- if (deferred)
- call_rcu(&table->rcu, flow_tbl_destroy_rcu_cb);
- else
- __flow_tbl_destroy(table);
-}
-
-struct sw_flow *ovs_flow_dump_next(struct flow_table *table, u32 *bucket, u32 *last)
-{
- struct sw_flow *flow;
- struct hlist_head *head;
- int ver;
- int i;
-
- ver = table->node_ver;
- while (*bucket < table->n_buckets) {
- i = 0;
- head = flex_array_get(table->buckets, *bucket);
- hlist_for_each_entry_rcu(flow, head, hash_node[ver]) {
- if (i < *last) {
- i++;
- continue;
- }
- *last = i + 1;
- return flow;
- }
- (*bucket)++;
- *last = 0;
- }
-
- return NULL;
-}
-
-static void __tbl_insert(struct flow_table *table, struct sw_flow *flow)
-{
- struct hlist_head *head;
-
- head = find_bucket(table, flow->hash);
- hlist_add_head_rcu(&flow->hash_node[table->node_ver], head);
-
- table->count++;
-}
-
-static void flow_table_copy_flows(struct flow_table *old, struct flow_table *new)
-{
- int old_ver;
- int i;
-
- old_ver = old->node_ver;
- new->node_ver = !old_ver;
-
- /* Insert in new table. */
- for (i = 0; i < old->n_buckets; i++) {
- struct sw_flow *flow;
- struct hlist_head *head;
-
- head = flex_array_get(old->buckets, i);
-
- hlist_for_each_entry(flow, head, hash_node[old_ver])
- __tbl_insert(new, flow);
- }
-
- new->mask_list = old->mask_list;
- old->keep_flows = true;
-}
-
-static struct flow_table *__flow_tbl_rehash(struct flow_table *table, int n_buckets)
-{
- struct flow_table *new_table;
-
- new_table = __flow_tbl_alloc(n_buckets);
- if (!new_table)
- return ERR_PTR(-ENOMEM);
-
- flow_table_copy_flows(table, new_table);
-
- return new_table;
-}
-
-struct flow_table *ovs_flow_tbl_rehash(struct flow_table *table)
-{
- return __flow_tbl_rehash(table, table->n_buckets);
-}
-
-struct flow_table *ovs_flow_tbl_expand(struct flow_table *table)
-{
- return __flow_tbl_rehash(table, table->n_buckets * 2);
-}
-
-static void __flow_free(struct sw_flow *flow)
-{
- kfree((struct sf_flow_acts __force *)flow->sf_acts);
- kmem_cache_free(flow_cache, flow);
-}
-
-static void rcu_free_flow_callback(struct rcu_head *rcu)
-{
- struct sw_flow *flow = container_of(rcu, struct sw_flow, rcu);
-
- __flow_free(flow);
-}
-
-void ovs_flow_free(struct sw_flow *flow, bool deferred)
-{
- if (!flow)
- return;
-
- ovs_sw_flow_mask_del_ref(flow->mask, deferred);
-
- if (deferred)
- call_rcu(&flow->rcu, rcu_free_flow_callback);
- else
- __flow_free(flow);
-}
-
-/* Schedules 'sf_acts' to be freed after the next RCU grace period.
- * The caller must hold rcu_read_lock for this to be sensible. */
-void ovs_flow_deferred_free_acts(struct sw_flow_actions *sf_acts)
-{
- kfree_rcu(sf_acts, rcu);
-}
-
static int parse_vlan(struct sk_buff *skb, struct sw_flow_key *key)
{
struct qtag_prefix {
@@ -1002,1080 +522,3 @@ int ovs_flow_extract(struct sk_buff *skb, u16 in_port, struct sw_flow_key *key)
return 0;
}
-
-static u32 ovs_flow_hash(const struct sw_flow_key *key, int key_start,
- int key_end)
-{
- u32 *hash_key = (u32 *)((u8 *)key + key_start);
- int hash_u32s = (key_end - key_start) >> 2;
-
- /* Make sure number of hash bytes are multiple of u32. */
- BUILD_BUG_ON(sizeof(long) % sizeof(u32));
-
- return jhash2(hash_key, hash_u32s, 0);
-}
-
-static int flow_key_start(const struct sw_flow_key *key)
-{
- if (key->tun_key.ipv4_dst)
- return 0;
- else
- return rounddown(offsetof(struct sw_flow_key, phy),
- sizeof(long));
-}
-
-static bool __cmp_key(const struct sw_flow_key *key1,
- const struct sw_flow_key *key2, int key_start, int key_end)
-{
- const long *cp1 = (long *)((u8 *)key1 + key_start);
- const long *cp2 = (long *)((u8 *)key2 + key_start);
- long diffs = 0;
- int i;
-
- for (i = key_start; i < key_end; i += sizeof(long))
- diffs |= *cp1++ ^ *cp2++;
-
- return diffs == 0;
-}
-
-static bool __flow_cmp_masked_key(const struct sw_flow *flow,
- const struct sw_flow_key *key, int key_start, int key_end)
-{
- return __cmp_key(&flow->key, key, key_start, key_end);
-}
-
-static bool __flow_cmp_unmasked_key(const struct sw_flow *flow,
- const struct sw_flow_key *key, int key_start, int key_end)
-{
- return __cmp_key(&flow->unmasked_key, key, key_start, key_end);
-}
-
-bool ovs_flow_cmp_unmasked_key(const struct sw_flow *flow,
- const struct sw_flow_key *key, int key_end)
-{
- int key_start;
- key_start = flow_key_start(key);
-
- return __flow_cmp_unmasked_key(flow, key, key_start, key_end);
-
-}
-
-struct sw_flow *ovs_flow_lookup_unmasked_key(struct flow_table *table,
- struct sw_flow_match *match)
-{
- struct sw_flow_key *unmasked = match->key;
- int key_end = match->range.end;
- struct sw_flow *flow;
-
- flow = ovs_flow_lookup(table, unmasked);
- if (flow && (!ovs_flow_cmp_unmasked_key(flow, unmasked, key_end)))
- flow = NULL;
-
- return flow;
-}
-
-static struct sw_flow *ovs_masked_flow_lookup(struct flow_table *table,
- const struct sw_flow_key *unmasked,
- struct sw_flow_mask *mask)
-{
- struct sw_flow *flow;
- struct hlist_head *head;
- int key_start = mask->range.start;
- int key_end = mask->range.end;
- u32 hash;
- struct sw_flow_key masked_key;
-
- ovs_flow_key_mask(&masked_key, unmasked, mask);
- hash = ovs_flow_hash(&masked_key, key_start, key_end);
- head = find_bucket(table, hash);
- hlist_for_each_entry_rcu(flow, head, hash_node[table->node_ver]) {
- if (flow->mask == mask &&
- __flow_cmp_masked_key(flow, &masked_key,
- key_start, key_end))
- return flow;
- }
- return NULL;
-}
-
-struct sw_flow *ovs_flow_lookup(struct flow_table *tbl,
- const struct sw_flow_key *key)
-{
- struct sw_flow *flow = NULL;
- struct sw_flow_mask *mask;
-
- list_for_each_entry_rcu(mask, tbl->mask_list, list) {
- flow = ovs_masked_flow_lookup(tbl, key, mask);
- if (flow) /* Found */
- break;
- }
-
- return flow;
-}
-
-
-void ovs_flow_insert(struct flow_table *table, struct sw_flow *flow)
-{
- flow->hash = ovs_flow_hash(&flow->key, flow->mask->range.start,
- flow->mask->range.end);
- __tbl_insert(table, flow);
-}
-
-void ovs_flow_remove(struct flow_table *table, struct sw_flow *flow)
-{
- BUG_ON(table->count == 0);
- hlist_del_rcu(&flow->hash_node[table->node_ver]);
- table->count--;
-}
-
-/* The size of the argument for each %OVS_KEY_ATTR_* Netlink attribute. */
-const int ovs_key_lens[OVS_KEY_ATTR_MAX + 1] = {
- [OVS_KEY_ATTR_ENCAP] = -1,
- [OVS_KEY_ATTR_PRIORITY] = sizeof(u32),
- [OVS_KEY_ATTR_IN_PORT] = sizeof(u32),
- [OVS_KEY_ATTR_SKB_MARK] = sizeof(u32),
- [OVS_KEY_ATTR_ETHERNET] = sizeof(struct ovs_key_ethernet),
- [OVS_KEY_ATTR_VLAN] = sizeof(__be16),
- [OVS_KEY_ATTR_ETHERTYPE] = sizeof(__be16),
- [OVS_KEY_ATTR_IPV4] = sizeof(struct ovs_key_ipv4),
- [OVS_KEY_ATTR_IPV6] = sizeof(struct ovs_key_ipv6),
- [OVS_KEY_ATTR_TCP] = sizeof(struct ovs_key_tcp),
- [OVS_KEY_ATTR_UDP] = sizeof(struct ovs_key_udp),
- [OVS_KEY_ATTR_SCTP] = sizeof(struct ovs_key_sctp),
- [OVS_KEY_ATTR_ICMP] = sizeof(struct ovs_key_icmp),
- [OVS_KEY_ATTR_ICMPV6] = sizeof(struct ovs_key_icmpv6),
- [OVS_KEY_ATTR_ARP] = sizeof(struct ovs_key_arp),
- [OVS_KEY_ATTR_ND] = sizeof(struct ovs_key_nd),
- [OVS_KEY_ATTR_TUNNEL] = -1,
-};
-
-static bool is_all_zero(const u8 *fp, size_t size)
-{
- int i;
-
- if (!fp)
- return false;
-
- for (i = 0; i < size; i++)
- if (fp[i])
- return false;
-
- return true;
-}
-
-static int __parse_flow_nlattrs(const struct nlattr *attr,
- const struct nlattr *a[],
- u64 *attrsp, bool nz)
-{
- const struct nlattr *nla;
- u32 attrs;
- int rem;
-
- attrs = *attrsp;
- nla_for_each_nested(nla, attr, rem) {
- u16 type = nla_type(nla);
- int expected_len;
-
- if (type > OVS_KEY_ATTR_MAX) {
- OVS_NLERR("Unknown key attribute (type=%d, max=%d).\n",
- type, OVS_KEY_ATTR_MAX);
- return -EINVAL;
- }
-
- if (attrs & (1 << type)) {
- OVS_NLERR("Duplicate key attribute (type %d).\n", type);
- return -EINVAL;
- }
-
- expected_len = ovs_key_lens[type];
- if (nla_len(nla) != expected_len && expected_len != -1) {
- OVS_NLERR("Key attribute has unexpected length (type=%d"
- ", length=%d, expected=%d).\n", type,
- nla_len(nla), expected_len);
- return -EINVAL;
- }
-
- if (!nz || !is_all_zero(nla_data(nla), expected_len)) {
- attrs |= 1 << type;
- a[type] = nla;
- }
- }
- if (rem) {
- OVS_NLERR("Message has %d unknown bytes.\n", rem);
- return -EINVAL;
- }
-
- *attrsp = attrs;
- return 0;
-}
-
-static int parse_flow_mask_nlattrs(const struct nlattr *attr,
- const struct nlattr *a[], u64 *attrsp)
-{
- return __parse_flow_nlattrs(attr, a, attrsp, true);
-}
-
-static int parse_flow_nlattrs(const struct nlattr *attr,
- const struct nlattr *a[], u64 *attrsp)
-{
- return __parse_flow_nlattrs(attr, a, attrsp, false);
-}
-
-int ovs_ipv4_tun_from_nlattr(const struct nlattr *attr,
- struct sw_flow_match *match, bool is_mask)
-{
- struct nlattr *a;
- int rem;
- bool ttl = false;
- __be16 tun_flags = 0;
-
- nla_for_each_nested(a, attr, rem) {
- int type = nla_type(a);
- static const u32 ovs_tunnel_key_lens[OVS_TUNNEL_KEY_ATTR_MAX + 1] = {
- [OVS_TUNNEL_KEY_ATTR_ID] = sizeof(u64),
- [OVS_TUNNEL_KEY_ATTR_IPV4_SRC] = sizeof(u32),
- [OVS_TUNNEL_KEY_ATTR_IPV4_DST] = sizeof(u32),
- [OVS_TUNNEL_KEY_ATTR_TOS] = 1,
- [OVS_TUNNEL_KEY_ATTR_TTL] = 1,
- [OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT] = 0,
- [OVS_TUNNEL_KEY_ATTR_CSUM] = 0,
- };
-
- if (type > OVS_TUNNEL_KEY_ATTR_MAX) {
- OVS_NLERR("Unknown IPv4 tunnel attribute (type=%d, max=%d).\n",
- type, OVS_TUNNEL_KEY_ATTR_MAX);
- return -EINVAL;
- }
-
- if (ovs_tunnel_key_lens[type] != nla_len(a)) {
- OVS_NLERR("IPv4 tunnel attribute type has unexpected "
- " length (type=%d, length=%d, expected=%d).\n",
- type, nla_len(a), ovs_tunnel_key_lens[type]);
- return -EINVAL;
- }
-
- switch (type) {
- case OVS_TUNNEL_KEY_ATTR_ID:
- SW_FLOW_KEY_PUT(match, tun_key.tun_id,
- nla_get_be64(a), is_mask);
- tun_flags |= TUNNEL_KEY;
- break;
- case OVS_TUNNEL_KEY_ATTR_IPV4_SRC:
- SW_FLOW_KEY_PUT(match, tun_key.ipv4_src,
- nla_get_be32(a), is_mask);
- break;
- case OVS_TUNNEL_KEY_ATTR_IPV4_DST:
- SW_FLOW_KEY_PUT(match, tun_key.ipv4_dst,
- nla_get_be32(a), is_mask);
- break;
- case OVS_TUNNEL_KEY_ATTR_TOS:
- SW_FLOW_KEY_PUT(match, tun_key.ipv4_tos,
- nla_get_u8(a), is_mask);
- break;
- case OVS_TUNNEL_KEY_ATTR_TTL:
- SW_FLOW_KEY_PUT(match, tun_key.ipv4_ttl,
- nla_get_u8(a), is_mask);
- ttl = true;
- break;
- case OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT:
- tun_flags |= TUNNEL_DONT_FRAGMENT;
- break;
- case OVS_TUNNEL_KEY_ATTR_CSUM:
- tun_flags |= TUNNEL_CSUM;
- break;
- default:
- return -EINVAL;
- }
- }
-
- SW_FLOW_KEY_PUT(match, tun_key.tun_flags, tun_flags, is_mask);
-
- if (rem > 0) {
- OVS_NLERR("IPv4 tunnel attribute has %d unknown bytes.\n", rem);
- return -EINVAL;
- }
-
- if (!is_mask) {
- if (!match->key->tun_key.ipv4_dst) {
- OVS_NLERR("IPv4 tunnel destination address is zero.\n");
- return -EINVAL;
- }
-
- if (!ttl) {
- OVS_NLERR("IPv4 tunnel TTL not specified.\n");
- return -EINVAL;
- }
- }
-
- return 0;
-}
-
-int ovs_ipv4_tun_to_nlattr(struct sk_buff *skb,
- const struct ovs_key_ipv4_tunnel *tun_key,
- const struct ovs_key_ipv4_tunnel *output)
-{
- struct nlattr *nla;
-
- nla = nla_nest_start(skb, OVS_KEY_ATTR_TUNNEL);
- if (!nla)
- return -EMSGSIZE;
-
- if (output->tun_flags & TUNNEL_KEY &&
- nla_put_be64(skb, OVS_TUNNEL_KEY_ATTR_ID, output->tun_id))
- return -EMSGSIZE;
- if (output->ipv4_src &&
- nla_put_be32(skb, OVS_TUNNEL_KEY_ATTR_IPV4_SRC, output->ipv4_src))
- return -EMSGSIZE;
- if (output->ipv4_dst &&
- nla_put_be32(skb, OVS_TUNNEL_KEY_ATTR_IPV4_DST, output->ipv4_dst))
- return -EMSGSIZE;
- if (output->ipv4_tos &&
- nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TOS, output->ipv4_tos))
- return -EMSGSIZE;
- if (nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TTL, output->ipv4_ttl))
- return -EMSGSIZE;
- if ((output->tun_flags & TUNNEL_DONT_FRAGMENT) &&
- nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT))
- return -EMSGSIZE;
- if ((output->tun_flags & TUNNEL_CSUM) &&
- nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_CSUM))
- return -EMSGSIZE;
-
- nla_nest_end(skb, nla);
- return 0;
-}
-
-static int metadata_from_nlattrs(struct sw_flow_match *match, u64 *attrs,
- const struct nlattr **a, bool is_mask)
-{
- if (*attrs & (1 << OVS_KEY_ATTR_PRIORITY)) {
- SW_FLOW_KEY_PUT(match, phy.priority,
- nla_get_u32(a[OVS_KEY_ATTR_PRIORITY]), is_mask);
- *attrs &= ~(1 << OVS_KEY_ATTR_PRIORITY);
- }
-
- if (*attrs & (1 << OVS_KEY_ATTR_IN_PORT)) {
- u32 in_port = nla_get_u32(a[OVS_KEY_ATTR_IN_PORT]);
-
- if (is_mask)
- in_port = 0xffffffff; /* Always exact match in_port. */
- else if (in_port >= DP_MAX_PORTS)
- return -EINVAL;
-
- SW_FLOW_KEY_PUT(match, phy.in_port, in_port, is_mask);
- *attrs &= ~(1 << OVS_KEY_ATTR_IN_PORT);
- } else if (!is_mask) {
- SW_FLOW_KEY_PUT(match, phy.in_port, DP_MAX_PORTS, is_mask);
- }
-
- if (*attrs & (1 << OVS_KEY_ATTR_SKB_MARK)) {
- uint32_t mark = nla_get_u32(a[OVS_KEY_ATTR_SKB_MARK]);
-
- SW_FLOW_KEY_PUT(match, phy.skb_mark, mark, is_mask);
- *attrs &= ~(1 << OVS_KEY_ATTR_SKB_MARK);
- }
- if (*attrs & (1 << OVS_KEY_ATTR_TUNNEL)) {
- if (ovs_ipv4_tun_from_nlattr(a[OVS_KEY_ATTR_TUNNEL], match,
- is_mask))
- return -EINVAL;
- *attrs &= ~(1 << OVS_KEY_ATTR_TUNNEL);
- }
- return 0;
-}
-
-static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
- const struct nlattr **a, bool is_mask)
-{
- int err;
- u64 orig_attrs = attrs;
-
- err = metadata_from_nlattrs(match, &attrs, a, is_mask);
- if (err)
- return err;
-
- if (attrs & (1 << OVS_KEY_ATTR_ETHERNET)) {
- const struct ovs_key_ethernet *eth_key;
-
- eth_key = nla_data(a[OVS_KEY_ATTR_ETHERNET]);
- SW_FLOW_KEY_MEMCPY(match, eth.src,
- eth_key->eth_src, ETH_ALEN, is_mask);
- SW_FLOW_KEY_MEMCPY(match, eth.dst,
- eth_key->eth_dst, ETH_ALEN, is_mask);
- attrs &= ~(1 << OVS_KEY_ATTR_ETHERNET);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_VLAN)) {
- __be16 tci;
-
- tci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
- if (!(tci & htons(VLAN_TAG_PRESENT))) {
- if (is_mask)
- OVS_NLERR("VLAN TCI mask does not have exact match for VLAN_TAG_PRESENT bit.\n");
- else
- OVS_NLERR("VLAN TCI does not have VLAN_TAG_PRESENT bit set.\n");
-
- return -EINVAL;
- }
-
- SW_FLOW_KEY_PUT(match, eth.tci, tci, is_mask);
- attrs &= ~(1 << OVS_KEY_ATTR_VLAN);
- } else if (!is_mask)
- SW_FLOW_KEY_PUT(match, eth.tci, htons(0xffff), true);
-
- if (attrs & (1 << OVS_KEY_ATTR_ETHERTYPE)) {
- __be16 eth_type;
-
- eth_type = nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]);
- if (is_mask) {
- /* Always exact match EtherType. */
- eth_type = htons(0xffff);
- } else if (ntohs(eth_type) < ETH_P_802_3_MIN) {
- OVS_NLERR("EtherType is less than minimum (type=%x, min=%x).\n",
- ntohs(eth_type), ETH_P_802_3_MIN);
- return -EINVAL;
- }
-
- SW_FLOW_KEY_PUT(match, eth.type, eth_type, is_mask);
- attrs &= ~(1 << OVS_KEY_ATTR_ETHERTYPE);
- } else if (!is_mask) {
- SW_FLOW_KEY_PUT(match, eth.type, htons(ETH_P_802_2), is_mask);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_IPV4)) {
- const struct ovs_key_ipv4 *ipv4_key;
-
- ipv4_key = nla_data(a[OVS_KEY_ATTR_IPV4]);
- if (!is_mask && ipv4_key->ipv4_frag > OVS_FRAG_TYPE_MAX) {
- OVS_NLERR("Unknown IPv4 fragment type (value=%d, max=%d).\n",
- ipv4_key->ipv4_frag, OVS_FRAG_TYPE_MAX);
- return -EINVAL;
- }
- SW_FLOW_KEY_PUT(match, ip.proto,
- ipv4_key->ipv4_proto, is_mask);
- SW_FLOW_KEY_PUT(match, ip.tos,
- ipv4_key->ipv4_tos, is_mask);
- SW_FLOW_KEY_PUT(match, ip.ttl,
- ipv4_key->ipv4_ttl, is_mask);
- SW_FLOW_KEY_PUT(match, ip.frag,
- ipv4_key->ipv4_frag, is_mask);
- SW_FLOW_KEY_PUT(match, ipv4.addr.src,
- ipv4_key->ipv4_src, is_mask);
- SW_FLOW_KEY_PUT(match, ipv4.addr.dst,
- ipv4_key->ipv4_dst, is_mask);
- attrs &= ~(1 << OVS_KEY_ATTR_IPV4);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_IPV6)) {
- const struct ovs_key_ipv6 *ipv6_key;
-
- ipv6_key = nla_data(a[OVS_KEY_ATTR_IPV6]);
- if (!is_mask && ipv6_key->ipv6_frag > OVS_FRAG_TYPE_MAX) {
- OVS_NLERR("Unknown IPv6 fragment type (value=%d, max=%d).\n",
- ipv6_key->ipv6_frag, OVS_FRAG_TYPE_MAX);
- return -EINVAL;
- }
- SW_FLOW_KEY_PUT(match, ipv6.label,
- ipv6_key->ipv6_label, is_mask);
- SW_FLOW_KEY_PUT(match, ip.proto,
- ipv6_key->ipv6_proto, is_mask);
- SW_FLOW_KEY_PUT(match, ip.tos,
- ipv6_key->ipv6_tclass, is_mask);
- SW_FLOW_KEY_PUT(match, ip.ttl,
- ipv6_key->ipv6_hlimit, is_mask);
- SW_FLOW_KEY_PUT(match, ip.frag,
- ipv6_key->ipv6_frag, is_mask);
- SW_FLOW_KEY_MEMCPY(match, ipv6.addr.src,
- ipv6_key->ipv6_src,
- sizeof(match->key->ipv6.addr.src),
- is_mask);
- SW_FLOW_KEY_MEMCPY(match, ipv6.addr.dst,
- ipv6_key->ipv6_dst,
- sizeof(match->key->ipv6.addr.dst),
- is_mask);
-
- attrs &= ~(1 << OVS_KEY_ATTR_IPV6);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_ARP)) {
- const struct ovs_key_arp *arp_key;
-
- arp_key = nla_data(a[OVS_KEY_ATTR_ARP]);
- if (!is_mask && (arp_key->arp_op & htons(0xff00))) {
- OVS_NLERR("Unknown ARP opcode (opcode=%d).\n",
- arp_key->arp_op);
- return -EINVAL;
- }
-
- SW_FLOW_KEY_PUT(match, ipv4.addr.src,
- arp_key->arp_sip, is_mask);
- SW_FLOW_KEY_PUT(match, ipv4.addr.dst,
- arp_key->arp_tip, is_mask);
- SW_FLOW_KEY_PUT(match, ip.proto,
- ntohs(arp_key->arp_op), is_mask);
- SW_FLOW_KEY_MEMCPY(match, ipv4.arp.sha,
- arp_key->arp_sha, ETH_ALEN, is_mask);
- SW_FLOW_KEY_MEMCPY(match, ipv4.arp.tha,
- arp_key->arp_tha, ETH_ALEN, is_mask);
-
- attrs &= ~(1 << OVS_KEY_ATTR_ARP);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_TCP)) {
- const struct ovs_key_tcp *tcp_key;
-
- tcp_key = nla_data(a[OVS_KEY_ATTR_TCP]);
- if (orig_attrs & (1 << OVS_KEY_ATTR_IPV4)) {
- SW_FLOW_KEY_PUT(match, ipv4.tp.src,
- tcp_key->tcp_src, is_mask);
- SW_FLOW_KEY_PUT(match, ipv4.tp.dst,
- tcp_key->tcp_dst, is_mask);
- } else {
- SW_FLOW_KEY_PUT(match, ipv6.tp.src,
- tcp_key->tcp_src, is_mask);
- SW_FLOW_KEY_PUT(match, ipv6.tp.dst,
- tcp_key->tcp_dst, is_mask);
- }
- attrs &= ~(1 << OVS_KEY_ATTR_TCP);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_UDP)) {
- const struct ovs_key_udp *udp_key;
-
- udp_key = nla_data(a[OVS_KEY_ATTR_UDP]);
- if (orig_attrs & (1 << OVS_KEY_ATTR_IPV4)) {
- SW_FLOW_KEY_PUT(match, ipv4.tp.src,
- udp_key->udp_src, is_mask);
- SW_FLOW_KEY_PUT(match, ipv4.tp.dst,
- udp_key->udp_dst, is_mask);
- } else {
- SW_FLOW_KEY_PUT(match, ipv6.tp.src,
- udp_key->udp_src, is_mask);
- SW_FLOW_KEY_PUT(match, ipv6.tp.dst,
- udp_key->udp_dst, is_mask);
- }
- attrs &= ~(1 << OVS_KEY_ATTR_UDP);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_SCTP)) {
- const struct ovs_key_sctp *sctp_key;
-
- sctp_key = nla_data(a[OVS_KEY_ATTR_SCTP]);
- if (orig_attrs & (1 << OVS_KEY_ATTR_IPV4)) {
- SW_FLOW_KEY_PUT(match, ipv4.tp.src,
- sctp_key->sctp_src, is_mask);
- SW_FLOW_KEY_PUT(match, ipv4.tp.dst,
- sctp_key->sctp_dst, is_mask);
- } else {
- SW_FLOW_KEY_PUT(match, ipv6.tp.src,
- sctp_key->sctp_src, is_mask);
- SW_FLOW_KEY_PUT(match, ipv6.tp.dst,
- sctp_key->sctp_dst, is_mask);
- }
- attrs &= ~(1 << OVS_KEY_ATTR_SCTP);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_ICMP)) {
- const struct ovs_key_icmp *icmp_key;
-
- icmp_key = nla_data(a[OVS_KEY_ATTR_ICMP]);
- SW_FLOW_KEY_PUT(match, ipv4.tp.src,
- htons(icmp_key->icmp_type), is_mask);
- SW_FLOW_KEY_PUT(match, ipv4.tp.dst,
- htons(icmp_key->icmp_code), is_mask);
- attrs &= ~(1 << OVS_KEY_ATTR_ICMP);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_ICMPV6)) {
- const struct ovs_key_icmpv6 *icmpv6_key;
-
- icmpv6_key = nla_data(a[OVS_KEY_ATTR_ICMPV6]);
- SW_FLOW_KEY_PUT(match, ipv6.tp.src,
- htons(icmpv6_key->icmpv6_type), is_mask);
- SW_FLOW_KEY_PUT(match, ipv6.tp.dst,
- htons(icmpv6_key->icmpv6_code), is_mask);
- attrs &= ~(1 << OVS_KEY_ATTR_ICMPV6);
- }
-
- if (attrs & (1 << OVS_KEY_ATTR_ND)) {
- const struct ovs_key_nd *nd_key;
-
- nd_key = nla_data(a[OVS_KEY_ATTR_ND]);
- SW_FLOW_KEY_MEMCPY(match, ipv6.nd.target,
- nd_key->nd_target,
- sizeof(match->key->ipv6.nd.target),
- is_mask);
- SW_FLOW_KEY_MEMCPY(match, ipv6.nd.sll,
- nd_key->nd_sll, ETH_ALEN, is_mask);
- SW_FLOW_KEY_MEMCPY(match, ipv6.nd.tll,
- nd_key->nd_tll, ETH_ALEN, is_mask);
- attrs &= ~(1 << OVS_KEY_ATTR_ND);
- }
-
- if (attrs != 0)
- return -EINVAL;
-
- return 0;
-}
-
-/**
- * ovs_match_from_nlattrs - parses Netlink attributes into a flow key and
- * mask. In case the 'mask' is NULL, the flow is treated as exact match
- * flow. Otherwise, it is treated as a wildcarded flow, except the mask
- * does not include any don't care bit.
- * @match: receives the extracted flow match information.
- * @key: Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink attribute
- * sequence. The fields should of the packet that triggered the creation
- * of this flow.
- * @mask: Optional. Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink
- * attribute specifies the mask field of the wildcarded flow.
- */
-int ovs_match_from_nlattrs(struct sw_flow_match *match,
- const struct nlattr *key,
- const struct nlattr *mask)
-{
- const struct nlattr *a[OVS_KEY_ATTR_MAX + 1];
- const struct nlattr *encap;
- u64 key_attrs = 0;
- u64 mask_attrs = 0;
- bool encap_valid = false;
- int err;
-
- err = parse_flow_nlattrs(key, a, &key_attrs);
- if (err)
- return err;
-
- if ((key_attrs & (1 << OVS_KEY_ATTR_ETHERNET)) &&
- (key_attrs & (1 << OVS_KEY_ATTR_ETHERTYPE)) &&
- (nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]) == htons(ETH_P_8021Q))) {
- __be16 tci;
-
- if (!((key_attrs & (1 << OVS_KEY_ATTR_VLAN)) &&
- (key_attrs & (1 << OVS_KEY_ATTR_ENCAP)))) {
- OVS_NLERR("Invalid Vlan frame.\n");
- return -EINVAL;
- }
-
- key_attrs &= ~(1 << OVS_KEY_ATTR_ETHERTYPE);
- tci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
- encap = a[OVS_KEY_ATTR_ENCAP];
- key_attrs &= ~(1 << OVS_KEY_ATTR_ENCAP);
- encap_valid = true;
-
- if (tci & htons(VLAN_TAG_PRESENT)) {
- err = parse_flow_nlattrs(encap, a, &key_attrs);
- if (err)
- return err;
- } else if (!tci) {
- /* Corner case for truncated 802.1Q header. */
- if (nla_len(encap)) {
- OVS_NLERR("Truncated 802.1Q header has non-zero encap attribute.\n");
- return -EINVAL;
- }
- } else {
- OVS_NLERR("Encap attribute is set for a non-VLAN frame.\n");
- return -EINVAL;
- }
- }
-
- err = ovs_key_from_nlattrs(match, key_attrs, a, false);
- if (err)
- return err;
-
- if (mask) {
- err = parse_flow_mask_nlattrs(mask, a, &mask_attrs);
- if (err)
- return err;
-
- if (mask_attrs & 1ULL << OVS_KEY_ATTR_ENCAP) {
- __be16 eth_type = 0;
- __be16 tci = 0;
-
- if (!encap_valid) {
- OVS_NLERR("Encap mask attribute is set for non-VLAN frame.\n");
- return -EINVAL;
- }
-
- mask_attrs &= ~(1 << OVS_KEY_ATTR_ENCAP);
- if (a[OVS_KEY_ATTR_ETHERTYPE])
- eth_type = nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]);
-
- if (eth_type == htons(0xffff)) {
- mask_attrs &= ~(1 << OVS_KEY_ATTR_ETHERTYPE);
- encap = a[OVS_KEY_ATTR_ENCAP];
- err = parse_flow_mask_nlattrs(encap, a, &mask_attrs);
- } else {
- OVS_NLERR("VLAN frames must have an exact match on the TPID (mask=%x).\n",
- ntohs(eth_type));
- return -EINVAL;
- }
-
- if (a[OVS_KEY_ATTR_VLAN])
- tci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
-
- if (!(tci & htons(VLAN_TAG_PRESENT))) {
- OVS_NLERR("VLAN tag present bit must have an exact match (tci_mask=%x).\n", ntohs(tci));
- return -EINVAL;
- }
- }
-
- err = ovs_key_from_nlattrs(match, mask_attrs, a, true);
- if (err)
- return err;
- } else {
- /* Populate exact match flow's key mask. */
- if (match->mask)
- ovs_sw_flow_mask_set(match->mask, &match->range, 0xff);
- }
-
- if (!ovs_match_validate(match, key_attrs, mask_attrs))
- return -EINVAL;
-
- return 0;
-}
-
-/**
- * ovs_flow_metadata_from_nlattrs - parses Netlink attributes into a flow key.
- * @flow: Receives extracted in_port, priority, tun_key and skb_mark.
- * @attr: Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink attribute
- * sequence.
- *
- * This parses a series of Netlink attributes that form a flow key, which must
- * take the same form accepted by flow_from_nlattrs(), but only enough of it to
- * get the metadata, that is, the parts of the flow key that cannot be
- * extracted from the packet itself.
- */
-
-int ovs_flow_metadata_from_nlattrs(struct sw_flow *flow,
- const struct nlattr *attr)
-{
- struct ovs_key_ipv4_tunnel *tun_key = &flow->key.tun_key;
- const struct nlattr *a[OVS_KEY_ATTR_MAX + 1];
- u64 attrs = 0;
- int err;
- struct sw_flow_match match;
-
- flow->key.phy.in_port = DP_MAX_PORTS;
- flow->key.phy.priority = 0;
- flow->key.phy.skb_mark = 0;
- memset(tun_key, 0, sizeof(flow->key.tun_key));
-
- err = parse_flow_nlattrs(attr, a, &attrs);
- if (err)
- return -EINVAL;
-
- memset(&match, 0, sizeof(match));
- match.key = &flow->key;
-
- err = metadata_from_nlattrs(&match, &attrs, a, false);
- if (err)
- return err;
-
- return 0;
-}
-
-int ovs_flow_to_nlattrs(const struct sw_flow_key *swkey,
- const struct sw_flow_key *output, struct sk_buff *skb)
-{
- struct ovs_key_ethernet *eth_key;
- struct nlattr *nla, *encap;
- bool is_mask = (swkey != output);
-
- if (nla_put_u32(skb, OVS_KEY_ATTR_PRIORITY, output->phy.priority))
- goto nla_put_failure;
-
- if ((swkey->tun_key.ipv4_dst || is_mask) &&
- ovs_ipv4_tun_to_nlattr(skb, &swkey->tun_key, &output->tun_key))
- goto nla_put_failure;
-
- if (swkey->phy.in_port == DP_MAX_PORTS) {
- if (is_mask && (output->phy.in_port == 0xffff))
- if (nla_put_u32(skb, OVS_KEY_ATTR_IN_PORT, 0xffffffff))
- goto nla_put_failure;
- } else {
- u16 upper_u16;
- upper_u16 = !is_mask ? 0 : 0xffff;
-
- if (nla_put_u32(skb, OVS_KEY_ATTR_IN_PORT,
- (upper_u16 << 16) | output->phy.in_port))
- goto nla_put_failure;
- }
-
- if (nla_put_u32(skb, OVS_KEY_ATTR_SKB_MARK, output->phy.skb_mark))
- goto nla_put_failure;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_ETHERNET, sizeof(*eth_key));
- if (!nla)
- goto nla_put_failure;
-
- eth_key = nla_data(nla);
- memcpy(eth_key->eth_src, output->eth.src, ETH_ALEN);
- memcpy(eth_key->eth_dst, output->eth.dst, ETH_ALEN);
-
- if (swkey->eth.tci || swkey->eth.type == htons(ETH_P_8021Q)) {
- __be16 eth_type;
- eth_type = !is_mask ? htons(ETH_P_8021Q) : htons(0xffff);
- if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE, eth_type) ||
- nla_put_be16(skb, OVS_KEY_ATTR_VLAN, output->eth.tci))
- goto nla_put_failure;
- encap = nla_nest_start(skb, OVS_KEY_ATTR_ENCAP);
- if (!swkey->eth.tci)
- goto unencap;
- } else
- encap = NULL;
-
- if (swkey->eth.type == htons(ETH_P_802_2)) {
- /*
- * Ethertype 802.2 is represented in the netlink with omitted
- * OVS_KEY_ATTR_ETHERTYPE in the flow key attribute, and
- * 0xffff in the mask attribute. Ethertype can also
- * be wildcarded.
- */
- if (is_mask && output->eth.type)
- if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE,
- output->eth.type))
- goto nla_put_failure;
- goto unencap;
- }
-
- if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE, output->eth.type))
- goto nla_put_failure;
-
- if (swkey->eth.type == htons(ETH_P_IP)) {
- struct ovs_key_ipv4 *ipv4_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_IPV4, sizeof(*ipv4_key));
- if (!nla)
- goto nla_put_failure;
- ipv4_key = nla_data(nla);
- ipv4_key->ipv4_src = output->ipv4.addr.src;
- ipv4_key->ipv4_dst = output->ipv4.addr.dst;
- ipv4_key->ipv4_proto = output->ip.proto;
- ipv4_key->ipv4_tos = output->ip.tos;
- ipv4_key->ipv4_ttl = output->ip.ttl;
- ipv4_key->ipv4_frag = output->ip.frag;
- } else if (swkey->eth.type == htons(ETH_P_IPV6)) {
- struct ovs_key_ipv6 *ipv6_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_IPV6, sizeof(*ipv6_key));
- if (!nla)
- goto nla_put_failure;
- ipv6_key = nla_data(nla);
- memcpy(ipv6_key->ipv6_src, &output->ipv6.addr.src,
- sizeof(ipv6_key->ipv6_src));
- memcpy(ipv6_key->ipv6_dst, &output->ipv6.addr.dst,
- sizeof(ipv6_key->ipv6_dst));
- ipv6_key->ipv6_label = output->ipv6.label;
- ipv6_key->ipv6_proto = output->ip.proto;
- ipv6_key->ipv6_tclass = output->ip.tos;
- ipv6_key->ipv6_hlimit = output->ip.ttl;
- ipv6_key->ipv6_frag = output->ip.frag;
- } else if (swkey->eth.type == htons(ETH_P_ARP) ||
- swkey->eth.type == htons(ETH_P_RARP)) {
- struct ovs_key_arp *arp_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_ARP, sizeof(*arp_key));
- if (!nla)
- goto nla_put_failure;
- arp_key = nla_data(nla);
- memset(arp_key, 0, sizeof(struct ovs_key_arp));
- arp_key->arp_sip = output->ipv4.addr.src;
- arp_key->arp_tip = output->ipv4.addr.dst;
- arp_key->arp_op = htons(output->ip.proto);
- memcpy(arp_key->arp_sha, output->ipv4.arp.sha, ETH_ALEN);
- memcpy(arp_key->arp_tha, output->ipv4.arp.tha, ETH_ALEN);
- }
-
- if ((swkey->eth.type == htons(ETH_P_IP) ||
- swkey->eth.type == htons(ETH_P_IPV6)) &&
- swkey->ip.frag != OVS_FRAG_TYPE_LATER) {
-
- if (swkey->ip.proto == IPPROTO_TCP) {
- struct ovs_key_tcp *tcp_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_TCP, sizeof(*tcp_key));
- if (!nla)
- goto nla_put_failure;
- tcp_key = nla_data(nla);
- if (swkey->eth.type == htons(ETH_P_IP)) {
- tcp_key->tcp_src = output->ipv4.tp.src;
- tcp_key->tcp_dst = output->ipv4.tp.dst;
- } else if (swkey->eth.type == htons(ETH_P_IPV6)) {
- tcp_key->tcp_src = output->ipv6.tp.src;
- tcp_key->tcp_dst = output->ipv6.tp.dst;
- }
- } else if (swkey->ip.proto == IPPROTO_UDP) {
- struct ovs_key_udp *udp_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_UDP, sizeof(*udp_key));
- if (!nla)
- goto nla_put_failure;
- udp_key = nla_data(nla);
- if (swkey->eth.type == htons(ETH_P_IP)) {
- udp_key->udp_src = output->ipv4.tp.src;
- udp_key->udp_dst = output->ipv4.tp.dst;
- } else if (swkey->eth.type == htons(ETH_P_IPV6)) {
- udp_key->udp_src = output->ipv6.tp.src;
- udp_key->udp_dst = output->ipv6.tp.dst;
- }
- } else if (swkey->ip.proto == IPPROTO_SCTP) {
- struct ovs_key_sctp *sctp_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_SCTP, sizeof(*sctp_key));
- if (!nla)
- goto nla_put_failure;
- sctp_key = nla_data(nla);
- if (swkey->eth.type == htons(ETH_P_IP)) {
- sctp_key->sctp_src = swkey->ipv4.tp.src;
- sctp_key->sctp_dst = swkey->ipv4.tp.dst;
- } else if (swkey->eth.type == htons(ETH_P_IPV6)) {
- sctp_key->sctp_src = swkey->ipv6.tp.src;
- sctp_key->sctp_dst = swkey->ipv6.tp.dst;
- }
- } else if (swkey->eth.type == htons(ETH_P_IP) &&
- swkey->ip.proto == IPPROTO_ICMP) {
- struct ovs_key_icmp *icmp_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_ICMP, sizeof(*icmp_key));
- if (!nla)
- goto nla_put_failure;
- icmp_key = nla_data(nla);
- icmp_key->icmp_type = ntohs(output->ipv4.tp.src);
- icmp_key->icmp_code = ntohs(output->ipv4.tp.dst);
- } else if (swkey->eth.type == htons(ETH_P_IPV6) &&
- swkey->ip.proto == IPPROTO_ICMPV6) {
- struct ovs_key_icmpv6 *icmpv6_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_ICMPV6,
- sizeof(*icmpv6_key));
- if (!nla)
- goto nla_put_failure;
- icmpv6_key = nla_data(nla);
- icmpv6_key->icmpv6_type = ntohs(output->ipv6.tp.src);
- icmpv6_key->icmpv6_code = ntohs(output->ipv6.tp.dst);
-
- if (icmpv6_key->icmpv6_type == NDISC_NEIGHBOUR_SOLICITATION ||
- icmpv6_key->icmpv6_type == NDISC_NEIGHBOUR_ADVERTISEMENT) {
- struct ovs_key_nd *nd_key;
-
- nla = nla_reserve(skb, OVS_KEY_ATTR_ND, sizeof(*nd_key));
- if (!nla)
- goto nla_put_failure;
- nd_key = nla_data(nla);
- memcpy(nd_key->nd_target, &output->ipv6.nd.target,
- sizeof(nd_key->nd_target));
- memcpy(nd_key->nd_sll, output->ipv6.nd.sll, ETH_ALEN);
- memcpy(nd_key->nd_tll, output->ipv6.nd.tll, ETH_ALEN);
- }
- }
- }
-
-unencap:
- if (encap)
- nla_nest_end(skb, encap);
-
- return 0;
-
-nla_put_failure:
- return -EMSGSIZE;
-}
-
-/* Initializes the flow module.
- * Returns zero if successful or a negative error code. */
-int ovs_flow_init(void)
-{
- BUILD_BUG_ON(__alignof__(struct sw_flow_key) % __alignof__(long));
- BUILD_BUG_ON(sizeof(struct sw_flow_key) % sizeof(long));
-
- flow_cache = kmem_cache_create("sw_flow", sizeof(struct sw_flow), 0,
- 0, NULL);
- if (flow_cache == NULL)
- return -ENOMEM;
-
- return 0;
-}
-
-/* Uninitializes the flow module. */
-void ovs_flow_exit(void)
-{
- kmem_cache_destroy(flow_cache);
-}
-
-struct sw_flow_mask *ovs_sw_flow_mask_alloc(void)
-{
- struct sw_flow_mask *mask;
-
- mask = kmalloc(sizeof(*mask), GFP_KERNEL);
- if (mask)
- mask->ref_count = 0;
-
- return mask;
-}
-
-void ovs_sw_flow_mask_add_ref(struct sw_flow_mask *mask)
-{
- mask->ref_count++;
-}
-
-void ovs_sw_flow_mask_del_ref(struct sw_flow_mask *mask, bool deferred)
-{
- if (!mask)
- return;
-
- BUG_ON(!mask->ref_count);
- mask->ref_count--;
-
- if (!mask->ref_count) {
- list_del_rcu(&mask->list);
- if (deferred)
- kfree_rcu(mask, rcu);
- else
- kfree(mask);
- }
-}
-
-static bool ovs_sw_flow_mask_equal(const struct sw_flow_mask *a,
- const struct sw_flow_mask *b)
-{
- u8 *a_ = (u8 *)&a->key + a->range.start;
- u8 *b_ = (u8 *)&b->key + b->range.start;
-
- return (a->range.end == b->range.end)
- && (a->range.start == b->range.start)
- && (memcmp(a_, b_, range_n_bytes(&a->range)) == 0);
-}
-
-struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *tbl,
- const struct sw_flow_mask *mask)
-{
- struct list_head *ml;
-
- list_for_each(ml, tbl->mask_list) {
- struct sw_flow_mask *m;
- m = container_of(ml, struct sw_flow_mask, list);
- if (ovs_sw_flow_mask_equal(mask, m))
- return m;
- }
-
- return NULL;
-}
-
-/**
- * add a new mask into the mask list.
- * The caller needs to make sure that 'mask' is not the same
- * as any masks that are already on the list.
- */
-void ovs_sw_flow_mask_insert(struct flow_table *tbl, struct sw_flow_mask *mask)
-{
- list_add_rcu(&mask->list, tbl->mask_list);
-}
-
-/**
- * Set 'range' fields in the mask to the value of 'val'.
- */
-static void ovs_sw_flow_mask_set(struct sw_flow_mask *mask,
- struct sw_flow_key_range *range, u8 val)
-{
- u8 *m = (u8 *)&mask->key + range->start;
-
- mask->range = *range;
- memset(m, val, range_n_bytes(range));
-}
diff --git a/net/openvswitch/flow.h b/net/openvswitch/flow.h
index 212fbf7..098fd1d 100644
--- a/net/openvswitch/flow.h
+++ b/net/openvswitch/flow.h
@@ -33,14 +33,6 @@
#include <net/inet_ecn.h>
struct sk_buff;
-struct sw_flow_mask;
-struct flow_table;
-
-struct sw_flow_actions {
- struct rcu_head rcu;
- u32 actions_len;
- struct nlattr actions[];
-};
/* Used to memset ovs_key_ipv4_tunnel padding. */
#define OVS_TUNNEL_KEY_SIZE \
@@ -127,6 +119,31 @@ struct sw_flow_key {
};
} __aligned(BITS_PER_LONG/8); /* Ensure that we can do comparisons as longs. */
+struct sw_flow_key_range {
+ size_t start;
+ size_t end;
+};
+
+struct sw_flow_mask {
+ int ref_count;
+ struct rcu_head rcu;
+ struct list_head list;
+ struct sw_flow_key_range range;
+ struct sw_flow_key key;
+};
+
+struct sw_flow_match {
+ struct sw_flow_key *key;
+ struct sw_flow_key_range range;
+ struct sw_flow_mask *mask;
+};
+
+struct sw_flow_actions {
+ struct rcu_head rcu;
+ u32 actions_len;
+ struct nlattr actions[];
+};
+
struct sw_flow {
struct rcu_head rcu;
struct hlist_node hash_node[2];
@@ -144,20 +161,6 @@ struct sw_flow {
u8 tcp_flags; /* Union of seen TCP flags. */
};
-struct sw_flow_key_range {
- size_t start;
- size_t end;
-};
-
-struct sw_flow_match {
- struct sw_flow_key *key;
- struct sw_flow_key_range range;
- struct sw_flow_mask *mask;
-};
-
-void ovs_match_init(struct sw_flow_match *match,
- struct sw_flow_key *key, struct sw_flow_mask *mask);
-
struct arp_eth_header {
__be16 ar_hrd; /* format of hardware address */
__be16 ar_pro; /* format of protocol address */
@@ -172,88 +175,9 @@ struct arp_eth_header {
unsigned char ar_tip[4]; /* target IP address */
} __packed;
-int ovs_flow_init(void);
-void ovs_flow_exit(void);
-
-struct sw_flow *ovs_flow_alloc(void);
-void ovs_flow_deferred_free(struct sw_flow *);
-void ovs_flow_free(struct sw_flow *, bool deferred);
-
-struct sw_flow_actions *ovs_flow_actions_alloc(int actions_len);
-void ovs_flow_deferred_free_acts(struct sw_flow_actions *);
-
-int ovs_flow_extract(struct sk_buff *, u16 in_port, struct sw_flow_key *);
void ovs_flow_used(struct sw_flow *, struct sk_buff *);
u64 ovs_flow_used_time(unsigned long flow_jiffies);
-int ovs_flow_to_nlattrs(const struct sw_flow_key *,
- const struct sw_flow_key *, struct sk_buff *);
-int ovs_match_from_nlattrs(struct sw_flow_match *match,
- const struct nlattr *,
- const struct nlattr *);
-int ovs_flow_metadata_from_nlattrs(struct sw_flow *flow,
- const struct nlattr *attr);
-#define MAX_ACTIONS_BUFSIZE (32 * 1024)
-#define TBL_MIN_BUCKETS 1024
-
-struct flow_table {
- struct flex_array *buckets;
- unsigned int count, n_buckets;
- struct rcu_head rcu;
- struct list_head *mask_list;
- int node_ver;
- u32 hash_seed;
- bool keep_flows;
-};
-
-static inline int ovs_flow_tbl_count(struct flow_table *table)
-{
- return table->count;
-}
-
-static inline int ovs_flow_tbl_need_to_expand(struct flow_table *table)
-{
- return (table->count > table->n_buckets);
-}
-
-struct sw_flow *ovs_flow_lookup(struct flow_table *,
- const struct sw_flow_key *);
-struct sw_flow *ovs_flow_lookup_unmasked_key(struct flow_table *table,
- struct sw_flow_match *match);
-
-void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred);
-struct flow_table *ovs_flow_tbl_alloc(int new_size);
-struct flow_table *ovs_flow_tbl_expand(struct flow_table *table);
-struct flow_table *ovs_flow_tbl_rehash(struct flow_table *table);
-
-void ovs_flow_insert(struct flow_table *table, struct sw_flow *flow);
-void ovs_flow_remove(struct flow_table *table, struct sw_flow *flow);
-
-struct sw_flow *ovs_flow_dump_next(struct flow_table *table, u32 *bucket, u32 *idx);
-extern const int ovs_key_lens[OVS_KEY_ATTR_MAX + 1];
-int ovs_ipv4_tun_from_nlattr(const struct nlattr *attr,
- struct sw_flow_match *match, bool is_mask);
-int ovs_ipv4_tun_to_nlattr(struct sk_buff *skb,
- const struct ovs_key_ipv4_tunnel *tun_key,
- const struct ovs_key_ipv4_tunnel *output);
-
-bool ovs_flow_cmp_unmasked_key(const struct sw_flow *flow,
- const struct sw_flow_key *key, int key_end);
-
-struct sw_flow_mask {
- int ref_count;
- struct rcu_head rcu;
- struct list_head list;
- struct sw_flow_key_range range;
- struct sw_flow_key key;
-};
+int ovs_flow_extract(struct sk_buff *, u16 in_port, struct sw_flow_key *);
-struct sw_flow_mask *ovs_sw_flow_mask_alloc(void);
-void ovs_sw_flow_mask_add_ref(struct sw_flow_mask *);
-void ovs_sw_flow_mask_del_ref(struct sw_flow_mask *, bool deferred);
-void ovs_sw_flow_mask_insert(struct flow_table *, struct sw_flow_mask *);
-struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *,
- const struct sw_flow_mask *);
-void ovs_flow_key_mask(struct sw_flow_key *dst, const struct sw_flow_key *src,
- const struct sw_flow_mask *mask);
#endif /* flow.h */
diff --git a/net/openvswitch/flow_netlink.c b/net/openvswitch/flow_netlink.c
new file mode 100644
index 0000000..e04649c
--- /dev/null
+++ b/net/openvswitch/flow_netlink.c
@@ -0,0 +1,1603 @@
+/*
+ * Copyright (c) 2007-2013 Nicira, Inc.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * 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., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
+ */
+
+#include "flow.h"
+#include "datapath.h"
+#include <linux/uaccess.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/if_ether.h>
+#include <linux/if_vlan.h>
+#include <net/llc_pdu.h>
+#include <linux/kernel.h>
+#include <linux/jhash.h>
+#include <linux/jiffies.h>
+#include <linux/llc.h>
+#include <linux/module.h>
+#include <linux/in.h>
+#include <linux/rcupdate.h>
+#include <linux/if_arp.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include <linux/sctp.h>
+#include <linux/tcp.h>
+#include <linux/udp.h>
+#include <linux/icmp.h>
+#include <linux/icmpv6.h>
+#include <linux/rculist.h>
+#include <net/ip.h>
+#include <net/ipv6.h>
+#include <net/ndisc.h>
+
+#include "flow_netlink.h"
+
+static void update_range__(struct sw_flow_match *match,
+ size_t offset, size_t size, bool is_mask)
+{
+ struct sw_flow_key_range *range = NULL;
+ size_t start = rounddown(offset, sizeof(long));
+ size_t end = roundup(offset + size, sizeof(long));
+
+ if (!is_mask)
+ range = &match->range;
+ else if (match->mask)
+ range = &match->mask->range;
+
+ if (!range)
+ return;
+
+ if (range->start == range->end) {
+ range->start = start;
+ range->end = end;
+ return;
+ }
+
+ if (range->start > start)
+ range->start = start;
+
+ if (range->end < end)
+ range->end = end;
+}
+
+#define SW_FLOW_KEY_PUT(match, field, value, is_mask) \
+ do { \
+ update_range__(match, offsetof(struct sw_flow_key, field), \
+ sizeof((match)->key->field), is_mask); \
+ if (is_mask) { \
+ if ((match)->mask) \
+ (match)->mask->key.field = value; \
+ } else { \
+ (match)->key->field = value; \
+ } \
+ } while (0)
+
+#define SW_FLOW_KEY_MEMCPY(match, field, value_p, len, is_mask) \
+ do { \
+ update_range__(match, offsetof(struct sw_flow_key, field), \
+ len, is_mask); \
+ if (is_mask) { \
+ if ((match)->mask) \
+ memcpy(&(match)->mask->key.field, value_p, len);\
+ } else { \
+ memcpy(&(match)->key->field, value_p, len); \
+ } \
+ } while (0)
+
+static u16 range_n_bytes(const struct sw_flow_key_range *range)
+{
+ return range->end - range->start;
+}
+
+static bool match_validate(const struct sw_flow_match *match,
+ u64 key_attrs, u64 mask_attrs)
+{
+ u64 key_expected = 1 << OVS_KEY_ATTR_ETHERNET;
+ u64 mask_allowed = key_attrs; /* At most allow all key attributes */
+
+ /* The following mask attributes allowed only if they
+ * pass the validation tests. */
+ mask_allowed &= ~((1 << OVS_KEY_ATTR_IPV4)
+ | (1 << OVS_KEY_ATTR_IPV6)
+ | (1 << OVS_KEY_ATTR_TCP)
+ | (1 << OVS_KEY_ATTR_UDP)
+ | (1 << OVS_KEY_ATTR_SCTP)
+ | (1 << OVS_KEY_ATTR_ICMP)
+ | (1 << OVS_KEY_ATTR_ICMPV6)
+ | (1 << OVS_KEY_ATTR_ARP)
+ | (1 << OVS_KEY_ATTR_ND));
+
+ /* Always allowed mask fields. */
+ mask_allowed |= ((1 << OVS_KEY_ATTR_TUNNEL)
+ | (1 << OVS_KEY_ATTR_IN_PORT)
+ | (1 << OVS_KEY_ATTR_ETHERTYPE));
+
+ /* Check key attributes. */
+ if (match->key->eth.type == htons(ETH_P_ARP)
+ || match->key->eth.type == htons(ETH_P_RARP)) {
+ key_expected |= 1 << OVS_KEY_ATTR_ARP;
+ if (match->mask && (match->mask->key.eth.type == htons(0xffff)))
+ mask_allowed |= 1 << OVS_KEY_ATTR_ARP;
+ }
+
+ if (match->key->eth.type == htons(ETH_P_IP)) {
+ key_expected |= 1 << OVS_KEY_ATTR_IPV4;
+ if (match->mask && (match->mask->key.eth.type == htons(0xffff)))
+ mask_allowed |= 1 << OVS_KEY_ATTR_IPV4;
+
+ if (match->key->ip.frag != OVS_FRAG_TYPE_LATER) {
+ if (match->key->ip.proto == IPPROTO_UDP) {
+ key_expected |= 1 << OVS_KEY_ATTR_UDP;
+ if (match->mask && (match->mask->key.ip.proto == 0xff))
+ mask_allowed |= 1 << OVS_KEY_ATTR_UDP;
+ }
+
+ if (match->key->ip.proto == IPPROTO_SCTP) {
+ key_expected |= 1 << OVS_KEY_ATTR_SCTP;
+ if (match->mask && (match->mask->key.ip.proto == 0xff))
+ mask_allowed |= 1 << OVS_KEY_ATTR_SCTP;
+ }
+
+ if (match->key->ip.proto == IPPROTO_TCP) {
+ key_expected |= 1 << OVS_KEY_ATTR_TCP;
+ if (match->mask && (match->mask->key.ip.proto == 0xff))
+ mask_allowed |= 1 << OVS_KEY_ATTR_TCP;
+ }
+
+ if (match->key->ip.proto == IPPROTO_ICMP) {
+ key_expected |= 1 << OVS_KEY_ATTR_ICMP;
+ if (match->mask && (match->mask->key.ip.proto == 0xff))
+ mask_allowed |= 1 << OVS_KEY_ATTR_ICMP;
+ }
+ }
+ }
+
+ if (match->key->eth.type == htons(ETH_P_IPV6)) {
+ key_expected |= 1 << OVS_KEY_ATTR_IPV6;
+ if (match->mask && (match->mask->key.eth.type == htons(0xffff)))
+ mask_allowed |= 1 << OVS_KEY_ATTR_IPV6;
+
+ if (match->key->ip.frag != OVS_FRAG_TYPE_LATER) {
+ if (match->key->ip.proto == IPPROTO_UDP) {
+ key_expected |= 1 << OVS_KEY_ATTR_UDP;
+ if (match->mask && (match->mask->key.ip.proto == 0xff))
+ mask_allowed |= 1 << OVS_KEY_ATTR_UDP;
+ }
+
+ if (match->key->ip.proto == IPPROTO_SCTP) {
+ key_expected |= 1 << OVS_KEY_ATTR_SCTP;
+ if (match->mask && (match->mask->key.ip.proto == 0xff))
+ mask_allowed |= 1 << OVS_KEY_ATTR_SCTP;
+ }
+
+ if (match->key->ip.proto == IPPROTO_TCP) {
+ key_expected |= 1 << OVS_KEY_ATTR_TCP;
+ if (match->mask && (match->mask->key.ip.proto == 0xff))
+ mask_allowed |= 1 << OVS_KEY_ATTR_TCP;
+ }
+
+ if (match->key->ip.proto == IPPROTO_ICMPV6) {
+ key_expected |= 1 << OVS_KEY_ATTR_ICMPV6;
+ if (match->mask && (match->mask->key.ip.proto == 0xff))
+ mask_allowed |= 1 << OVS_KEY_ATTR_ICMPV6;
+
+ if (match->key->ipv6.tp.src ==
+ htons(NDISC_NEIGHBOUR_SOLICITATION) ||
+ match->key->ipv6.tp.src == htons(NDISC_NEIGHBOUR_ADVERTISEMENT)) {
+ key_expected |= 1 << OVS_KEY_ATTR_ND;
+ if (match->mask && (match->mask->key.ipv6.tp.src == htons(0xffff)))
+ mask_allowed |= 1 << OVS_KEY_ATTR_ND;
+ }
+ }
+ }
+ }
+
+ if ((key_attrs & key_expected) != key_expected) {
+ /* Key attributes check failed. */
+ OVS_NLERR("Missing expected key attributes (key_attrs=%llx, expected=%llx).\n",
+ key_attrs, key_expected);
+ return false;
+ }
+
+ if ((mask_attrs & mask_allowed) != mask_attrs) {
+ /* Mask attributes check failed. */
+ OVS_NLERR("Contain more than allowed mask fields (mask_attrs=%llx, mask_allowed=%llx).\n",
+ mask_attrs, mask_allowed);
+ return false;
+ }
+
+ return true;
+}
+
+/* The size of the argument for each %OVS_KEY_ATTR_* Netlink attribute. */
+static const int ovs_key_lens[OVS_KEY_ATTR_MAX + 1] = {
+ [OVS_KEY_ATTR_ENCAP] = -1,
+ [OVS_KEY_ATTR_PRIORITY] = sizeof(u32),
+ [OVS_KEY_ATTR_IN_PORT] = sizeof(u32),
+ [OVS_KEY_ATTR_SKB_MARK] = sizeof(u32),
+ [OVS_KEY_ATTR_ETHERNET] = sizeof(struct ovs_key_ethernet),
+ [OVS_KEY_ATTR_VLAN] = sizeof(__be16),
+ [OVS_KEY_ATTR_ETHERTYPE] = sizeof(__be16),
+ [OVS_KEY_ATTR_IPV4] = sizeof(struct ovs_key_ipv4),
+ [OVS_KEY_ATTR_IPV6] = sizeof(struct ovs_key_ipv6),
+ [OVS_KEY_ATTR_TCP] = sizeof(struct ovs_key_tcp),
+ [OVS_KEY_ATTR_UDP] = sizeof(struct ovs_key_udp),
+ [OVS_KEY_ATTR_SCTP] = sizeof(struct ovs_key_sctp),
+ [OVS_KEY_ATTR_ICMP] = sizeof(struct ovs_key_icmp),
+ [OVS_KEY_ATTR_ICMPV6] = sizeof(struct ovs_key_icmpv6),
+ [OVS_KEY_ATTR_ARP] = sizeof(struct ovs_key_arp),
+ [OVS_KEY_ATTR_ND] = sizeof(struct ovs_key_nd),
+ [OVS_KEY_ATTR_TUNNEL] = -1,
+};
+
+static bool is_all_zero(const u8 *fp, size_t size)
+{
+ int i;
+
+ if (!fp)
+ return false;
+
+ for (i = 0; i < size; i++)
+ if (fp[i])
+ return false;
+
+ return true;
+}
+
+static int __parse_flow_nlattrs(const struct nlattr *attr,
+ const struct nlattr *a[],
+ u64 *attrsp, bool nz)
+{
+ const struct nlattr *nla;
+ u64 attrs;
+ int rem;
+
+ attrs = *attrsp;
+ nla_for_each_nested(nla, attr, rem) {
+ u16 type = nla_type(nla);
+ int expected_len;
+
+ if (type > OVS_KEY_ATTR_MAX) {
+ OVS_NLERR("Unknown key attribute (type=%d, max=%d).\n",
+ type, OVS_KEY_ATTR_MAX);
+ return -EINVAL;
+ }
+
+ if (attrs & (1 << type)) {
+ OVS_NLERR("Duplicate key attribute (type %d).\n", type);
+ return -EINVAL;
+ }
+
+ expected_len = ovs_key_lens[type];
+ if (nla_len(nla) != expected_len && expected_len != -1) {
+ OVS_NLERR("Key attribute has unexpected length (type=%d"
+ ", length=%d, expected=%d).\n", type,
+ nla_len(nla), expected_len);
+ return -EINVAL;
+ }
+
+ if (!nz || !is_all_zero(nla_data(nla), expected_len)) {
+ attrs |= 1 << type;
+ a[type] = nla;
+ }
+ }
+ if (rem) {
+ OVS_NLERR("Message has %d unknown bytes.\n", rem);
+ return -EINVAL;
+ }
+
+ *attrsp = attrs;
+ return 0;
+}
+
+static int parse_flow_mask_nlattrs(const struct nlattr *attr,
+ const struct nlattr *a[], u64 *attrsp)
+{
+ return __parse_flow_nlattrs(attr, a, attrsp, true);
+}
+
+static int parse_flow_nlattrs(const struct nlattr *attr,
+ const struct nlattr *a[], u64 *attrsp)
+{
+ return __parse_flow_nlattrs(attr, a, attrsp, false);
+}
+
+static int ipv4_tun_from_nlattr(const struct nlattr *attr,
+ struct sw_flow_match *match, bool is_mask)
+{
+ struct nlattr *a;
+ int rem;
+ bool ttl = false;
+ __be16 tun_flags = 0;
+
+ nla_for_each_nested(a, attr, rem) {
+ int type = nla_type(a);
+ static const u32 ovs_tunnel_key_lens[OVS_TUNNEL_KEY_ATTR_MAX + 1] = {
+ [OVS_TUNNEL_KEY_ATTR_ID] = sizeof(u64),
+ [OVS_TUNNEL_KEY_ATTR_IPV4_SRC] = sizeof(u32),
+ [OVS_TUNNEL_KEY_ATTR_IPV4_DST] = sizeof(u32),
+ [OVS_TUNNEL_KEY_ATTR_TOS] = 1,
+ [OVS_TUNNEL_KEY_ATTR_TTL] = 1,
+ [OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT] = 0,
+ [OVS_TUNNEL_KEY_ATTR_CSUM] = 0,
+ };
+
+ if (type > OVS_TUNNEL_KEY_ATTR_MAX) {
+ OVS_NLERR("Unknown IPv4 tunnel attribute (type=%d, max=%d).\n",
+ type, OVS_TUNNEL_KEY_ATTR_MAX);
+ return -EINVAL;
+ }
+
+ if (ovs_tunnel_key_lens[type] != nla_len(a)) {
+ OVS_NLERR("IPv4 tunnel attribute type has unexpected "
+ " length (type=%d, length=%d, expected=%d).\n",
+ type, nla_len(a), ovs_tunnel_key_lens[type]);
+ return -EINVAL;
+ }
+
+ switch (type) {
+ case OVS_TUNNEL_KEY_ATTR_ID:
+ SW_FLOW_KEY_PUT(match, tun_key.tun_id,
+ nla_get_be64(a), is_mask);
+ tun_flags |= TUNNEL_KEY;
+ break;
+ case OVS_TUNNEL_KEY_ATTR_IPV4_SRC:
+ SW_FLOW_KEY_PUT(match, tun_key.ipv4_src,
+ nla_get_be32(a), is_mask);
+ break;
+ case OVS_TUNNEL_KEY_ATTR_IPV4_DST:
+ SW_FLOW_KEY_PUT(match, tun_key.ipv4_dst,
+ nla_get_be32(a), is_mask);
+ break;
+ case OVS_TUNNEL_KEY_ATTR_TOS:
+ SW_FLOW_KEY_PUT(match, tun_key.ipv4_tos,
+ nla_get_u8(a), is_mask);
+ break;
+ case OVS_TUNNEL_KEY_ATTR_TTL:
+ SW_FLOW_KEY_PUT(match, tun_key.ipv4_ttl,
+ nla_get_u8(a), is_mask);
+ ttl = true;
+ break;
+ case OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT:
+ tun_flags |= TUNNEL_DONT_FRAGMENT;
+ break;
+ case OVS_TUNNEL_KEY_ATTR_CSUM:
+ tun_flags |= TUNNEL_CSUM;
+ break;
+ default:
+ return -EINVAL;
+ }
+ }
+
+ SW_FLOW_KEY_PUT(match, tun_key.tun_flags, tun_flags, is_mask);
+
+ if (rem > 0) {
+ OVS_NLERR("IPv4 tunnel attribute has %d unknown bytes.\n", rem);
+ return -EINVAL;
+ }
+
+ if (!is_mask) {
+ if (!match->key->tun_key.ipv4_dst) {
+ OVS_NLERR("IPv4 tunnel destination address is zero.\n");
+ return -EINVAL;
+ }
+
+ if (!ttl) {
+ OVS_NLERR("IPv4 tunnel TTL not specified.\n");
+ return -EINVAL;
+ }
+ }
+
+ return 0;
+}
+
+static int ipv4_tun_to_nlattr(struct sk_buff *skb,
+ const struct ovs_key_ipv4_tunnel *tun_key,
+ const struct ovs_key_ipv4_tunnel *output)
+{
+ struct nlattr *nla;
+
+ nla = nla_nest_start(skb, OVS_KEY_ATTR_TUNNEL);
+ if (!nla)
+ return -EMSGSIZE;
+
+ if (output->tun_flags & TUNNEL_KEY &&
+ nla_put_be64(skb, OVS_TUNNEL_KEY_ATTR_ID, output->tun_id))
+ return -EMSGSIZE;
+ if (output->ipv4_src &&
+ nla_put_be32(skb, OVS_TUNNEL_KEY_ATTR_IPV4_SRC, output->ipv4_src))
+ return -EMSGSIZE;
+ if (output->ipv4_dst &&
+ nla_put_be32(skb, OVS_TUNNEL_KEY_ATTR_IPV4_DST, output->ipv4_dst))
+ return -EMSGSIZE;
+ if (output->ipv4_tos &&
+ nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TOS, output->ipv4_tos))
+ return -EMSGSIZE;
+ if (nla_put_u8(skb, OVS_TUNNEL_KEY_ATTR_TTL, output->ipv4_ttl))
+ return -EMSGSIZE;
+ if ((output->tun_flags & TUNNEL_DONT_FRAGMENT) &&
+ nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_DONT_FRAGMENT))
+ return -EMSGSIZE;
+ if ((output->tun_flags & TUNNEL_CSUM) &&
+ nla_put_flag(skb, OVS_TUNNEL_KEY_ATTR_CSUM))
+ return -EMSGSIZE;
+
+ nla_nest_end(skb, nla);
+ return 0;
+}
+
+
+static int metadata_from_nlattrs(struct sw_flow_match *match, u64 *attrs,
+ const struct nlattr **a, bool is_mask)
+{
+ if (*attrs & (1 << OVS_KEY_ATTR_PRIORITY)) {
+ SW_FLOW_KEY_PUT(match, phy.priority,
+ nla_get_u32(a[OVS_KEY_ATTR_PRIORITY]), is_mask);
+ *attrs &= ~(1 << OVS_KEY_ATTR_PRIORITY);
+ }
+
+ if (*attrs & (1 << OVS_KEY_ATTR_IN_PORT)) {
+ u32 in_port = nla_get_u32(a[OVS_KEY_ATTR_IN_PORT]);
+
+ if (is_mask)
+ in_port = 0xffffffff; /* Always exact match in_port. */
+ else if (in_port >= DP_MAX_PORTS)
+ return -EINVAL;
+
+ SW_FLOW_KEY_PUT(match, phy.in_port, in_port, is_mask);
+ *attrs &= ~(1 << OVS_KEY_ATTR_IN_PORT);
+ } else if (!is_mask) {
+ SW_FLOW_KEY_PUT(match, phy.in_port, DP_MAX_PORTS, is_mask);
+ }
+
+ if (*attrs & (1 << OVS_KEY_ATTR_SKB_MARK)) {
+ uint32_t mark = nla_get_u32(a[OVS_KEY_ATTR_SKB_MARK]);
+
+ SW_FLOW_KEY_PUT(match, phy.skb_mark, mark, is_mask);
+ *attrs &= ~(1 << OVS_KEY_ATTR_SKB_MARK);
+ }
+ if (*attrs & (1 << OVS_KEY_ATTR_TUNNEL)) {
+ if (ipv4_tun_from_nlattr(a[OVS_KEY_ATTR_TUNNEL], match,
+ is_mask))
+ return -EINVAL;
+ *attrs &= ~(1 << OVS_KEY_ATTR_TUNNEL);
+ }
+ return 0;
+}
+
+static int ovs_key_from_nlattrs(struct sw_flow_match *match, u64 attrs,
+ const struct nlattr **a, bool is_mask)
+{
+ int err;
+ u64 orig_attrs = attrs;
+
+ err = metadata_from_nlattrs(match, &attrs, a, is_mask);
+ if (err)
+ return err;
+
+ if (attrs & (1 << OVS_KEY_ATTR_ETHERNET)) {
+ const struct ovs_key_ethernet *eth_key;
+
+ eth_key = nla_data(a[OVS_KEY_ATTR_ETHERNET]);
+ SW_FLOW_KEY_MEMCPY(match, eth.src,
+ eth_key->eth_src, ETH_ALEN, is_mask);
+ SW_FLOW_KEY_MEMCPY(match, eth.dst,
+ eth_key->eth_dst, ETH_ALEN, is_mask);
+ attrs &= ~(1 << OVS_KEY_ATTR_ETHERNET);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_VLAN)) {
+ __be16 tci;
+
+ tci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
+ if (!(tci & htons(VLAN_TAG_PRESENT))) {
+ if (is_mask)
+ OVS_NLERR("VLAN TCI mask does not have exact match for VLAN_TAG_PRESENT bit.\n");
+ else
+ OVS_NLERR("VLAN TCI does not have VLAN_TAG_PRESENT bit set.\n");
+
+ return -EINVAL;
+ }
+
+ SW_FLOW_KEY_PUT(match, eth.tci, tci, is_mask);
+ attrs &= ~(1 << OVS_KEY_ATTR_VLAN);
+ } else if (!is_mask)
+ SW_FLOW_KEY_PUT(match, eth.tci, htons(0xffff), true);
+
+ if (attrs & (1 << OVS_KEY_ATTR_ETHERTYPE)) {
+ __be16 eth_type;
+
+ eth_type = nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]);
+ if (is_mask) {
+ /* Always exact match EtherType. */
+ eth_type = htons(0xffff);
+ } else if (ntohs(eth_type) < ETH_P_802_3_MIN) {
+ OVS_NLERR("EtherType is less than minimum (type=%x, min=%x).\n",
+ ntohs(eth_type), ETH_P_802_3_MIN);
+ return -EINVAL;
+ }
+
+ SW_FLOW_KEY_PUT(match, eth.type, eth_type, is_mask);
+ attrs &= ~(1 << OVS_KEY_ATTR_ETHERTYPE);
+ } else if (!is_mask) {
+ SW_FLOW_KEY_PUT(match, eth.type, htons(ETH_P_802_2), is_mask);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_IPV4)) {
+ const struct ovs_key_ipv4 *ipv4_key;
+
+ ipv4_key = nla_data(a[OVS_KEY_ATTR_IPV4]);
+ if (!is_mask && ipv4_key->ipv4_frag > OVS_FRAG_TYPE_MAX) {
+ OVS_NLERR("Unknown IPv4 fragment type (value=%d, max=%d).\n",
+ ipv4_key->ipv4_frag, OVS_FRAG_TYPE_MAX);
+ return -EINVAL;
+ }
+ SW_FLOW_KEY_PUT(match, ip.proto,
+ ipv4_key->ipv4_proto, is_mask);
+ SW_FLOW_KEY_PUT(match, ip.tos,
+ ipv4_key->ipv4_tos, is_mask);
+ SW_FLOW_KEY_PUT(match, ip.ttl,
+ ipv4_key->ipv4_ttl, is_mask);
+ SW_FLOW_KEY_PUT(match, ip.frag,
+ ipv4_key->ipv4_frag, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv4.addr.src,
+ ipv4_key->ipv4_src, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv4.addr.dst,
+ ipv4_key->ipv4_dst, is_mask);
+ attrs &= ~(1 << OVS_KEY_ATTR_IPV4);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_IPV6)) {
+ const struct ovs_key_ipv6 *ipv6_key;
+
+ ipv6_key = nla_data(a[OVS_KEY_ATTR_IPV6]);
+ if (!is_mask && ipv6_key->ipv6_frag > OVS_FRAG_TYPE_MAX) {
+ OVS_NLERR("Unknown IPv6 fragment type (value=%d, max=%d).\n",
+ ipv6_key->ipv6_frag, OVS_FRAG_TYPE_MAX);
+ return -EINVAL;
+ }
+ SW_FLOW_KEY_PUT(match, ipv6.label,
+ ipv6_key->ipv6_label, is_mask);
+ SW_FLOW_KEY_PUT(match, ip.proto,
+ ipv6_key->ipv6_proto, is_mask);
+ SW_FLOW_KEY_PUT(match, ip.tos,
+ ipv6_key->ipv6_tclass, is_mask);
+ SW_FLOW_KEY_PUT(match, ip.ttl,
+ ipv6_key->ipv6_hlimit, is_mask);
+ SW_FLOW_KEY_PUT(match, ip.frag,
+ ipv6_key->ipv6_frag, is_mask);
+ SW_FLOW_KEY_MEMCPY(match, ipv6.addr.src,
+ ipv6_key->ipv6_src,
+ sizeof(match->key->ipv6.addr.src),
+ is_mask);
+ SW_FLOW_KEY_MEMCPY(match, ipv6.addr.dst,
+ ipv6_key->ipv6_dst,
+ sizeof(match->key->ipv6.addr.dst),
+ is_mask);
+
+ attrs &= ~(1 << OVS_KEY_ATTR_IPV6);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_ARP)) {
+ const struct ovs_key_arp *arp_key;
+
+ arp_key = nla_data(a[OVS_KEY_ATTR_ARP]);
+ if (!is_mask && (arp_key->arp_op & htons(0xff00))) {
+ OVS_NLERR("Unknown ARP opcode (opcode=%d).\n",
+ arp_key->arp_op);
+ return -EINVAL;
+ }
+
+ SW_FLOW_KEY_PUT(match, ipv4.addr.src,
+ arp_key->arp_sip, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv4.addr.dst,
+ arp_key->arp_tip, is_mask);
+ SW_FLOW_KEY_PUT(match, ip.proto,
+ ntohs(arp_key->arp_op), is_mask);
+ SW_FLOW_KEY_MEMCPY(match, ipv4.arp.sha,
+ arp_key->arp_sha, ETH_ALEN, is_mask);
+ SW_FLOW_KEY_MEMCPY(match, ipv4.arp.tha,
+ arp_key->arp_tha, ETH_ALEN, is_mask);
+
+ attrs &= ~(1 << OVS_KEY_ATTR_ARP);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_TCP)) {
+ const struct ovs_key_tcp *tcp_key;
+
+ tcp_key = nla_data(a[OVS_KEY_ATTR_TCP]);
+ if (orig_attrs & (1 << OVS_KEY_ATTR_IPV4)) {
+ SW_FLOW_KEY_PUT(match, ipv4.tp.src,
+ tcp_key->tcp_src, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv4.tp.dst,
+ tcp_key->tcp_dst, is_mask);
+ } else {
+ SW_FLOW_KEY_PUT(match, ipv6.tp.src,
+ tcp_key->tcp_src, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv6.tp.dst,
+ tcp_key->tcp_dst, is_mask);
+ }
+ attrs &= ~(1 << OVS_KEY_ATTR_TCP);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_UDP)) {
+ const struct ovs_key_udp *udp_key;
+
+ udp_key = nla_data(a[OVS_KEY_ATTR_UDP]);
+ if (orig_attrs & (1 << OVS_KEY_ATTR_IPV4)) {
+ SW_FLOW_KEY_PUT(match, ipv4.tp.src,
+ udp_key->udp_src, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv4.tp.dst,
+ udp_key->udp_dst, is_mask);
+ } else {
+ SW_FLOW_KEY_PUT(match, ipv6.tp.src,
+ udp_key->udp_src, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv6.tp.dst,
+ udp_key->udp_dst, is_mask);
+ }
+ attrs &= ~(1 << OVS_KEY_ATTR_UDP);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_SCTP)) {
+ const struct ovs_key_sctp *sctp_key;
+
+ sctp_key = nla_data(a[OVS_KEY_ATTR_SCTP]);
+ if (orig_attrs & (1 << OVS_KEY_ATTR_IPV4)) {
+ SW_FLOW_KEY_PUT(match, ipv4.tp.src,
+ sctp_key->sctp_src, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv4.tp.dst,
+ sctp_key->sctp_dst, is_mask);
+ } else {
+ SW_FLOW_KEY_PUT(match, ipv6.tp.src,
+ sctp_key->sctp_src, is_mask);
+ SW_FLOW_KEY_PUT(match, ipv6.tp.dst,
+ sctp_key->sctp_dst, is_mask);
+ }
+ attrs &= ~(1 << OVS_KEY_ATTR_SCTP);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_ICMP)) {
+ const struct ovs_key_icmp *icmp_key;
+
+ icmp_key = nla_data(a[OVS_KEY_ATTR_ICMP]);
+ SW_FLOW_KEY_PUT(match, ipv4.tp.src,
+ htons(icmp_key->icmp_type), is_mask);
+ SW_FLOW_KEY_PUT(match, ipv4.tp.dst,
+ htons(icmp_key->icmp_code), is_mask);
+ attrs &= ~(1 << OVS_KEY_ATTR_ICMP);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_ICMPV6)) {
+ const struct ovs_key_icmpv6 *icmpv6_key;
+
+ icmpv6_key = nla_data(a[OVS_KEY_ATTR_ICMPV6]);
+ SW_FLOW_KEY_PUT(match, ipv6.tp.src,
+ htons(icmpv6_key->icmpv6_type), is_mask);
+ SW_FLOW_KEY_PUT(match, ipv6.tp.dst,
+ htons(icmpv6_key->icmpv6_code), is_mask);
+ attrs &= ~(1 << OVS_KEY_ATTR_ICMPV6);
+ }
+
+ if (attrs & (1 << OVS_KEY_ATTR_ND)) {
+ const struct ovs_key_nd *nd_key;
+
+ nd_key = nla_data(a[OVS_KEY_ATTR_ND]);
+ SW_FLOW_KEY_MEMCPY(match, ipv6.nd.target,
+ nd_key->nd_target,
+ sizeof(match->key->ipv6.nd.target),
+ is_mask);
+ SW_FLOW_KEY_MEMCPY(match, ipv6.nd.sll,
+ nd_key->nd_sll, ETH_ALEN, is_mask);
+ SW_FLOW_KEY_MEMCPY(match, ipv6.nd.tll,
+ nd_key->nd_tll, ETH_ALEN, is_mask);
+ attrs &= ~(1 << OVS_KEY_ATTR_ND);
+ }
+
+ if (attrs != 0)
+ return -EINVAL;
+
+ return 0;
+}
+
+static void sw_flow_mask_set(struct sw_flow_mask *mask,
+ struct sw_flow_key_range *range, u8 val)
+{
+ u8 *m = (u8 *)&mask->key + range->start;
+
+ mask->range = *range;
+ memset(m, val, range_n_bytes(range));
+}
+
+/**
+ * ovs_nla_get_match - parses Netlink attributes into a flow key and
+ * mask. In case the 'mask' is NULL, the flow is treated as exact match
+ * flow. Otherwise, it is treated as a wildcarded flow, except the mask
+ * does not include any don't care bit.
+ * @match: receives the extracted flow match information.
+ * @key: Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink attribute
+ * sequence. The fields should of the packet that triggered the creation
+ * of this flow.
+ * @mask: Optional. Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink
+ * attribute specifies the mask field of the wildcarded flow.
+ */
+int ovs_nla_get_match(struct sw_flow_match *match,
+ const struct nlattr *key,
+ const struct nlattr *mask)
+{
+ const struct nlattr *a[OVS_KEY_ATTR_MAX + 1];
+ const struct nlattr *encap;
+ u64 key_attrs = 0;
+ u64 mask_attrs = 0;
+ bool encap_valid = false;
+ int err;
+
+ err = parse_flow_nlattrs(key, a, &key_attrs);
+ if (err)
+ return err;
+
+ if ((key_attrs & (1 << OVS_KEY_ATTR_ETHERNET)) &&
+ (key_attrs & (1 << OVS_KEY_ATTR_ETHERTYPE)) &&
+ (nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]) == htons(ETH_P_8021Q))) {
+ __be16 tci;
+
+ if (!((key_attrs & (1 << OVS_KEY_ATTR_VLAN)) &&
+ (key_attrs & (1 << OVS_KEY_ATTR_ENCAP)))) {
+ OVS_NLERR("Invalid Vlan frame.\n");
+ return -EINVAL;
+ }
+
+ key_attrs &= ~(1 << OVS_KEY_ATTR_ETHERTYPE);
+ tci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
+ encap = a[OVS_KEY_ATTR_ENCAP];
+ key_attrs &= ~(1 << OVS_KEY_ATTR_ENCAP);
+ encap_valid = true;
+
+ if (tci & htons(VLAN_TAG_PRESENT)) {
+ err = parse_flow_nlattrs(encap, a, &key_attrs);
+ if (err)
+ return err;
+ } else if (!tci) {
+ /* Corner case for truncated 802.1Q header. */
+ if (nla_len(encap)) {
+ OVS_NLERR("Truncated 802.1Q header has non-zero encap attribute.\n");
+ return -EINVAL;
+ }
+ } else {
+ OVS_NLERR("Encap attribute is set for a non-VLAN frame.\n");
+ return -EINVAL;
+ }
+ }
+
+ err = ovs_key_from_nlattrs(match, key_attrs, a, false);
+ if (err)
+ return err;
+
+ if (mask) {
+ err = parse_flow_mask_nlattrs(mask, a, &mask_attrs);
+ if (err)
+ return err;
+
+ if (mask_attrs & 1 << OVS_KEY_ATTR_ENCAP) {
+ __be16 eth_type = 0;
+ __be16 tci = 0;
+
+ if (!encap_valid) {
+ OVS_NLERR("Encap mask attribute is set for non-VLAN frame.\n");
+ return -EINVAL;
+ }
+
+ mask_attrs &= ~(1 << OVS_KEY_ATTR_ENCAP);
+ if (a[OVS_KEY_ATTR_ETHERTYPE])
+ eth_type = nla_get_be16(a[OVS_KEY_ATTR_ETHERTYPE]);
+
+ if (eth_type == htons(0xffff)) {
+ mask_attrs &= ~(1 << OVS_KEY_ATTR_ETHERTYPE);
+ encap = a[OVS_KEY_ATTR_ENCAP];
+ err = parse_flow_mask_nlattrs(encap, a, &mask_attrs);
+ } else {
+ OVS_NLERR("VLAN frames must have an exact match on the TPID (mask=%x).\n",
+ ntohs(eth_type));
+ return -EINVAL;
+ }
+
+ if (a[OVS_KEY_ATTR_VLAN])
+ tci = nla_get_be16(a[OVS_KEY_ATTR_VLAN]);
+
+ if (!(tci & htons(VLAN_TAG_PRESENT))) {
+ OVS_NLERR("VLAN tag present bit must have an exact match (tci_mask=%x).\n", ntohs(tci));
+ return -EINVAL;
+ }
+ }
+
+ err = ovs_key_from_nlattrs(match, mask_attrs, a, true);
+ if (err)
+ return err;
+ } else {
+ /* Populate exact match flow's key mask. */
+ if (match->mask)
+ sw_flow_mask_set(match->mask, &match->range, 0xff);
+ }
+
+ if (!match_validate(match, key_attrs, mask_attrs))
+ return -EINVAL;
+
+ return 0;
+}
+
+/**
+ * ovs_nla_get_flow_metadata - parses Netlink attributes into a flow key.
+ * @flow: Receives extracted in_port, priority, tun_key and skb_mark.
+ * @attr: Netlink attribute holding nested %OVS_KEY_ATTR_* Netlink attribute
+ * sequence.
+ *
+ * This parses a series of Netlink attributes that form a flow key, which must
+ * take the same form accepted by flow_from_nlattrs(), but only enough of it to
+ * get the metadata, that is, the parts of the flow key that cannot be
+ * extracted from the packet itself.
+ */
+
+int ovs_nla_get_flow_metadata(struct sw_flow *flow,
+ const struct nlattr *attr)
+{
+ struct ovs_key_ipv4_tunnel *tun_key = &flow->key.tun_key;
+ const struct nlattr *a[OVS_KEY_ATTR_MAX + 1];
+ u64 attrs = 0;
+ int err;
+ struct sw_flow_match match;
+
+ flow->key.phy.in_port = DP_MAX_PORTS;
+ flow->key.phy.priority = 0;
+ flow->key.phy.skb_mark = 0;
+ memset(tun_key, 0, sizeof(flow->key.tun_key));
+
+ err = parse_flow_nlattrs(attr, a, &attrs);
+ if (err)
+ return -EINVAL;
+
+ memset(&match, 0, sizeof(match));
+ match.key = &flow->key;
+
+ err = metadata_from_nlattrs(&match, &attrs, a, false);
+ if (err)
+ return err;
+
+ return 0;
+}
+
+int ovs_nla_put_flow(const struct sw_flow_key *swkey,
+ const struct sw_flow_key *output, struct sk_buff *skb)
+{
+ struct ovs_key_ethernet *eth_key;
+ struct nlattr *nla, *encap;
+ bool is_mask = (swkey != output);
+
+ if (nla_put_u32(skb, OVS_KEY_ATTR_PRIORITY, output->phy.priority))
+ goto nla_put_failure;
+
+ if ((swkey->tun_key.ipv4_dst || is_mask) &&
+ ipv4_tun_to_nlattr(skb, &swkey->tun_key, &output->tun_key))
+ goto nla_put_failure;
+
+ if (swkey->phy.in_port == DP_MAX_PORTS) {
+ if (is_mask && (output->phy.in_port == 0xffff))
+ if (nla_put_u32(skb, OVS_KEY_ATTR_IN_PORT, 0xffffffff))
+ goto nla_put_failure;
+ } else {
+ u16 upper_u16;
+ upper_u16 = !is_mask ? 0 : 0xffff;
+
+ if (nla_put_u32(skb, OVS_KEY_ATTR_IN_PORT,
+ (upper_u16 << 16) | output->phy.in_port))
+ goto nla_put_failure;
+ }
+
+ if (nla_put_u32(skb, OVS_KEY_ATTR_SKB_MARK, output->phy.skb_mark))
+ goto nla_put_failure;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_ETHERNET, sizeof(*eth_key));
+ if (!nla)
+ goto nla_put_failure;
+
+ eth_key = nla_data(nla);
+ memcpy(eth_key->eth_src, output->eth.src, ETH_ALEN);
+ memcpy(eth_key->eth_dst, output->eth.dst, ETH_ALEN);
+
+ if (swkey->eth.tci || swkey->eth.type == htons(ETH_P_8021Q)) {
+ __be16 eth_type;
+ eth_type = !is_mask ? htons(ETH_P_8021Q) : htons(0xffff);
+ if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE, eth_type) ||
+ nla_put_be16(skb, OVS_KEY_ATTR_VLAN, output->eth.tci))
+ goto nla_put_failure;
+ encap = nla_nest_start(skb, OVS_KEY_ATTR_ENCAP);
+ if (!swkey->eth.tci)
+ goto unencap;
+ } else
+ encap = NULL;
+
+ if (swkey->eth.type == htons(ETH_P_802_2)) {
+ /*
+ * Ethertype 802.2 is represented in the netlink with omitted
+ * OVS_KEY_ATTR_ETHERTYPE in the flow key attribute, and
+ * 0xffff in the mask attribute. Ethertype can also
+ * be wildcarded.
+ */
+ if (is_mask && output->eth.type)
+ if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE,
+ output->eth.type))
+ goto nla_put_failure;
+ goto unencap;
+ }
+
+ if (nla_put_be16(skb, OVS_KEY_ATTR_ETHERTYPE, output->eth.type))
+ goto nla_put_failure;
+
+ if (swkey->eth.type == htons(ETH_P_IP)) {
+ struct ovs_key_ipv4 *ipv4_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_IPV4, sizeof(*ipv4_key));
+ if (!nla)
+ goto nla_put_failure;
+ ipv4_key = nla_data(nla);
+ ipv4_key->ipv4_src = output->ipv4.addr.src;
+ ipv4_key->ipv4_dst = output->ipv4.addr.dst;
+ ipv4_key->ipv4_proto = output->ip.proto;
+ ipv4_key->ipv4_tos = output->ip.tos;
+ ipv4_key->ipv4_ttl = output->ip.ttl;
+ ipv4_key->ipv4_frag = output->ip.frag;
+ } else if (swkey->eth.type == htons(ETH_P_IPV6)) {
+ struct ovs_key_ipv6 *ipv6_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_IPV6, sizeof(*ipv6_key));
+ if (!nla)
+ goto nla_put_failure;
+ ipv6_key = nla_data(nla);
+ memcpy(ipv6_key->ipv6_src, &output->ipv6.addr.src,
+ sizeof(ipv6_key->ipv6_src));
+ memcpy(ipv6_key->ipv6_dst, &output->ipv6.addr.dst,
+ sizeof(ipv6_key->ipv6_dst));
+ ipv6_key->ipv6_label = output->ipv6.label;
+ ipv6_key->ipv6_proto = output->ip.proto;
+ ipv6_key->ipv6_tclass = output->ip.tos;
+ ipv6_key->ipv6_hlimit = output->ip.ttl;
+ ipv6_key->ipv6_frag = output->ip.frag;
+ } else if (swkey->eth.type == htons(ETH_P_ARP) ||
+ swkey->eth.type == htons(ETH_P_RARP)) {
+ struct ovs_key_arp *arp_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_ARP, sizeof(*arp_key));
+ if (!nla)
+ goto nla_put_failure;
+ arp_key = nla_data(nla);
+ memset(arp_key, 0, sizeof(struct ovs_key_arp));
+ arp_key->arp_sip = output->ipv4.addr.src;
+ arp_key->arp_tip = output->ipv4.addr.dst;
+ arp_key->arp_op = htons(output->ip.proto);
+ memcpy(arp_key->arp_sha, output->ipv4.arp.sha, ETH_ALEN);
+ memcpy(arp_key->arp_tha, output->ipv4.arp.tha, ETH_ALEN);
+ }
+
+ if ((swkey->eth.type == htons(ETH_P_IP) ||
+ swkey->eth.type == htons(ETH_P_IPV6)) &&
+ swkey->ip.frag != OVS_FRAG_TYPE_LATER) {
+
+ if (swkey->ip.proto == IPPROTO_TCP) {
+ struct ovs_key_tcp *tcp_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_TCP, sizeof(*tcp_key));
+ if (!nla)
+ goto nla_put_failure;
+ tcp_key = nla_data(nla);
+ if (swkey->eth.type == htons(ETH_P_IP)) {
+ tcp_key->tcp_src = output->ipv4.tp.src;
+ tcp_key->tcp_dst = output->ipv4.tp.dst;
+ } else if (swkey->eth.type == htons(ETH_P_IPV6)) {
+ tcp_key->tcp_src = output->ipv6.tp.src;
+ tcp_key->tcp_dst = output->ipv6.tp.dst;
+ }
+ } else if (swkey->ip.proto == IPPROTO_UDP) {
+ struct ovs_key_udp *udp_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_UDP, sizeof(*udp_key));
+ if (!nla)
+ goto nla_put_failure;
+ udp_key = nla_data(nla);
+ if (swkey->eth.type == htons(ETH_P_IP)) {
+ udp_key->udp_src = output->ipv4.tp.src;
+ udp_key->udp_dst = output->ipv4.tp.dst;
+ } else if (swkey->eth.type == htons(ETH_P_IPV6)) {
+ udp_key->udp_src = output->ipv6.tp.src;
+ udp_key->udp_dst = output->ipv6.tp.dst;
+ }
+ } else if (swkey->ip.proto == IPPROTO_SCTP) {
+ struct ovs_key_sctp *sctp_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_SCTP, sizeof(*sctp_key));
+ if (!nla)
+ goto nla_put_failure;
+ sctp_key = nla_data(nla);
+ if (swkey->eth.type == htons(ETH_P_IP)) {
+ sctp_key->sctp_src = swkey->ipv4.tp.src;
+ sctp_key->sctp_dst = swkey->ipv4.tp.dst;
+ } else if (swkey->eth.type == htons(ETH_P_IPV6)) {
+ sctp_key->sctp_src = swkey->ipv6.tp.src;
+ sctp_key->sctp_dst = swkey->ipv6.tp.dst;
+ }
+ } else if (swkey->eth.type == htons(ETH_P_IP) &&
+ swkey->ip.proto == IPPROTO_ICMP) {
+ struct ovs_key_icmp *icmp_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_ICMP, sizeof(*icmp_key));
+ if (!nla)
+ goto nla_put_failure;
+ icmp_key = nla_data(nla);
+ icmp_key->icmp_type = ntohs(output->ipv4.tp.src);
+ icmp_key->icmp_code = ntohs(output->ipv4.tp.dst);
+ } else if (swkey->eth.type == htons(ETH_P_IPV6) &&
+ swkey->ip.proto == IPPROTO_ICMPV6) {
+ struct ovs_key_icmpv6 *icmpv6_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_ICMPV6,
+ sizeof(*icmpv6_key));
+ if (!nla)
+ goto nla_put_failure;
+ icmpv6_key = nla_data(nla);
+ icmpv6_key->icmpv6_type = ntohs(output->ipv6.tp.src);
+ icmpv6_key->icmpv6_code = ntohs(output->ipv6.tp.dst);
+
+ if (icmpv6_key->icmpv6_type == NDISC_NEIGHBOUR_SOLICITATION ||
+ icmpv6_key->icmpv6_type == NDISC_NEIGHBOUR_ADVERTISEMENT) {
+ struct ovs_key_nd *nd_key;
+
+ nla = nla_reserve(skb, OVS_KEY_ATTR_ND, sizeof(*nd_key));
+ if (!nla)
+ goto nla_put_failure;
+ nd_key = nla_data(nla);
+ memcpy(nd_key->nd_target, &output->ipv6.nd.target,
+ sizeof(nd_key->nd_target));
+ memcpy(nd_key->nd_sll, output->ipv6.nd.sll, ETH_ALEN);
+ memcpy(nd_key->nd_tll, output->ipv6.nd.tll, ETH_ALEN);
+ }
+ }
+ }
+
+unencap:
+ if (encap)
+ nla_nest_end(skb, encap);
+
+ return 0;
+
+nla_put_failure:
+ return -EMSGSIZE;
+}
+
+#define MAX_ACTIONS_BUFSIZE (32 * 1024)
+
+struct sw_flow_actions *ovs_nla_alloc_flow_actions(int size)
+{
+ struct sw_flow_actions *sfa;
+
+ if (size > MAX_ACTIONS_BUFSIZE)
+ return ERR_PTR(-EINVAL);
+
+ sfa = kmalloc(sizeof(*sfa) + size, GFP_KERNEL);
+ if (!sfa)
+ return ERR_PTR(-ENOMEM);
+
+ sfa->actions_len = 0;
+ return sfa;
+}
+
+/* RCU callback used by ovs_nla_free_flow_actions. */
+static void rcu_free_acts_callback(struct rcu_head *rcu)
+{
+ struct sw_flow_actions *sf_acts = container_of(rcu,
+ struct sw_flow_actions, rcu);
+ kfree(sf_acts);
+}
+
+/* Schedules 'sf_acts' to be freed after the next RCU grace period.
+ * The caller must hold rcu_read_lock for this to be sensible. */
+void ovs_nla_free_flow_actions(struct sw_flow_actions *sf_acts)
+{
+ call_rcu(&sf_acts->rcu, rcu_free_acts_callback);
+}
+
+static struct nlattr *reserve_sfa_size(struct sw_flow_actions **sfa,
+ int attr_len)
+{
+
+ struct sw_flow_actions *acts;
+ int new_acts_size;
+ int req_size = NLA_ALIGN(attr_len);
+ int next_offset = offsetof(struct sw_flow_actions, actions) +
+ (*sfa)->actions_len;
+
+ if (req_size <= (ksize(*sfa) - next_offset))
+ goto out;
+
+ new_acts_size = ksize(*sfa) * 2;
+
+ if (new_acts_size > MAX_ACTIONS_BUFSIZE) {
+ if ((MAX_ACTIONS_BUFSIZE - next_offset) < req_size)
+ return ERR_PTR(-EMSGSIZE);
+ new_acts_size = MAX_ACTIONS_BUFSIZE;
+ }
+
+ acts = ovs_nla_alloc_flow_actions(new_acts_size);
+ if (IS_ERR(acts))
+ return (void *)acts;
+
+ memcpy(acts->actions, (*sfa)->actions, (*sfa)->actions_len);
+ acts->actions_len = (*sfa)->actions_len;
+ kfree(*sfa);
+ *sfa = acts;
+
+out:
+ (*sfa)->actions_len += req_size;
+ return (struct nlattr *) ((unsigned char *)(*sfa) + next_offset);
+}
+
+static int add_action(struct sw_flow_actions **sfa, int attrtype, void *data, int len)
+{
+ struct nlattr *a;
+
+ a = reserve_sfa_size(sfa, nla_attr_size(len));
+ if (IS_ERR(a))
+ return PTR_ERR(a);
+
+ a->nla_type = attrtype;
+ a->nla_len = nla_attr_size(len);
+
+ if (data)
+ memcpy(nla_data(a), data, len);
+ memset((unsigned char *) a + a->nla_len, 0, nla_padlen(len));
+
+ return 0;
+}
+
+static inline int add_nested_action_start(struct sw_flow_actions **sfa,
+ int attrtype)
+{
+ int used = (*sfa)->actions_len;
+ int err;
+
+ err = add_action(sfa, attrtype, NULL, 0);
+ if (err)
+ return err;
+
+ return used;
+}
+
+static inline void add_nested_action_end(struct sw_flow_actions *sfa,
+ int st_offset)
+{
+ struct nlattr *a = (struct nlattr *) ((unsigned char *)sfa->actions +
+ st_offset);
+
+ a->nla_len = sfa->actions_len - st_offset;
+}
+
+static int validate_and_copy_sample(const struct nlattr *attr,
+ const struct sw_flow_key *key, int depth,
+ struct sw_flow_actions **sfa)
+{
+ const struct nlattr *attrs[OVS_SAMPLE_ATTR_MAX + 1];
+ const struct nlattr *probability, *actions;
+ const struct nlattr *a;
+ int rem, start, err, st_acts;
+
+ memset(attrs, 0, sizeof(attrs));
+ nla_for_each_nested(a, attr, rem) {
+ int type = nla_type(a);
+ if (!type || type > OVS_SAMPLE_ATTR_MAX || attrs[type])
+ return -EINVAL;
+ attrs[type] = a;
+ }
+ if (rem)
+ return -EINVAL;
+
+ probability = attrs[OVS_SAMPLE_ATTR_PROBABILITY];
+ if (!probability || nla_len(probability) != sizeof(u32))
+ return -EINVAL;
+
+ actions = attrs[OVS_SAMPLE_ATTR_ACTIONS];
+ if (!actions || (nla_len(actions) && nla_len(actions) < NLA_HDRLEN))
+ return -EINVAL;
+
+ /* validation done, copy sample action. */
+ start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SAMPLE);
+ if (start < 0)
+ return start;
+ err = add_action(sfa, OVS_SAMPLE_ATTR_PROBABILITY,
+ nla_data(probability), sizeof(u32));
+ if (err)
+ return err;
+ st_acts = add_nested_action_start(sfa, OVS_SAMPLE_ATTR_ACTIONS);
+ if (st_acts < 0)
+ return st_acts;
+
+ err = ovs_nla_copy_actions(actions, key, depth + 1, sfa);
+ if (err)
+ return err;
+
+ add_nested_action_end(*sfa, st_acts);
+ add_nested_action_end(*sfa, start);
+
+ return 0;
+}
+
+static int validate_tp_port(const struct sw_flow_key *flow_key)
+{
+ if (flow_key->eth.type == htons(ETH_P_IP)) {
+ if (flow_key->ipv4.tp.src || flow_key->ipv4.tp.dst)
+ return 0;
+ } else if (flow_key->eth.type == htons(ETH_P_IPV6)) {
+ if (flow_key->ipv6.tp.src || flow_key->ipv6.tp.dst)
+ return 0;
+ }
+
+ return -EINVAL;
+}
+
+void ovs_match_init(struct sw_flow_match *match,
+ struct sw_flow_key *key,
+ struct sw_flow_mask *mask)
+{
+ memset(match, 0, sizeof(*match));
+ match->key = key;
+ match->mask = mask;
+
+ memset(key, 0, sizeof(*key));
+
+ if (mask) {
+ memset(&mask->key, 0, sizeof(mask->key));
+ mask->range.start = mask->range.end = 0;
+ }
+}
+
+static int validate_and_copy_set_tun(const struct nlattr *attr,
+ struct sw_flow_actions **sfa)
+{
+ struct sw_flow_match match;
+ struct sw_flow_key key;
+ int err, start;
+
+ ovs_match_init(&match, &key, NULL);
+ err = ipv4_tun_from_nlattr(nla_data(attr), &match, false);
+ if (err)
+ return err;
+
+ start = add_nested_action_start(sfa, OVS_ACTION_ATTR_SET);
+ if (start < 0)
+ return start;
+
+ err = add_action(sfa, OVS_KEY_ATTR_IPV4_TUNNEL, &match.key->tun_key,
+ sizeof(match.key->tun_key));
+ add_nested_action_end(*sfa, start);
+
+ return err;
+}
+
+static int validate_set(const struct nlattr *a,
+ const struct sw_flow_key *flow_key,
+ struct sw_flow_actions **sfa,
+ bool *set_tun)
+{
+ const struct nlattr *ovs_key = nla_data(a);
+ int key_type = nla_type(ovs_key);
+
+ /* There can be only one key in a action */
+ if (nla_total_size(nla_len(ovs_key)) != nla_len(a))
+ return -EINVAL;
+
+ if (key_type > OVS_KEY_ATTR_MAX ||
+ (ovs_key_lens[key_type] != nla_len(ovs_key) &&
+ ovs_key_lens[key_type] != -1))
+ return -EINVAL;
+
+ switch (key_type) {
+ const struct ovs_key_ipv4 *ipv4_key;
+ const struct ovs_key_ipv6 *ipv6_key;
+ int err;
+
+ case OVS_KEY_ATTR_PRIORITY:
+ case OVS_KEY_ATTR_SKB_MARK:
+ case OVS_KEY_ATTR_ETHERNET:
+ break;
+
+ case OVS_KEY_ATTR_TUNNEL:
+ *set_tun = true;
+ err = validate_and_copy_set_tun(a, sfa);
+ if (err)
+ return err;
+ break;
+
+ case OVS_KEY_ATTR_IPV4:
+ if (flow_key->eth.type != htons(ETH_P_IP))
+ return -EINVAL;
+
+ if (!flow_key->ip.proto)
+ return -EINVAL;
+
+ ipv4_key = nla_data(ovs_key);
+ if (ipv4_key->ipv4_proto != flow_key->ip.proto)
+ return -EINVAL;
+
+ if (ipv4_key->ipv4_frag != flow_key->ip.frag)
+ return -EINVAL;
+
+ break;
+
+ case OVS_KEY_ATTR_IPV6:
+ if (flow_key->eth.type != htons(ETH_P_IPV6))
+ return -EINVAL;
+
+ if (!flow_key->ip.proto)
+ return -EINVAL;
+
+ ipv6_key = nla_data(ovs_key);
+ if (ipv6_key->ipv6_proto != flow_key->ip.proto)
+ return -EINVAL;
+
+ if (ipv6_key->ipv6_frag != flow_key->ip.frag)
+ return -EINVAL;
+
+ if (ntohl(ipv6_key->ipv6_label) & 0xFFF00000)
+ return -EINVAL;
+
+ break;
+
+ case OVS_KEY_ATTR_TCP:
+ if (flow_key->ip.proto != IPPROTO_TCP)
+ return -EINVAL;
+
+ return validate_tp_port(flow_key);
+
+ case OVS_KEY_ATTR_UDP:
+ if (flow_key->ip.proto != IPPROTO_UDP)
+ return -EINVAL;
+
+ return validate_tp_port(flow_key);
+
+ case OVS_KEY_ATTR_SCTP:
+ if (flow_key->ip.proto != IPPROTO_SCTP)
+ return -EINVAL;
+
+ return validate_tp_port(flow_key);
+
+ default:
+ return -EINVAL;
+ }
+
+ return 0;
+}
+
+static int validate_userspace(const struct nlattr *attr)
+{
+ static const struct nla_policy userspace_policy[OVS_USERSPACE_ATTR_MAX + 1] = {
+ [OVS_USERSPACE_ATTR_PID] = {.type = NLA_U32 },
+ [OVS_USERSPACE_ATTR_USERDATA] = {.type = NLA_UNSPEC },
+ };
+ struct nlattr *a[OVS_USERSPACE_ATTR_MAX + 1];
+ int error;
+
+ error = nla_parse_nested(a, OVS_USERSPACE_ATTR_MAX,
+ attr, userspace_policy);
+ if (error)
+ return error;
+
+ if (!a[OVS_USERSPACE_ATTR_PID] ||
+ !nla_get_u32(a[OVS_USERSPACE_ATTR_PID]))
+ return -EINVAL;
+
+ return 0;
+}
+
+static int copy_action(const struct nlattr *from,
+ struct sw_flow_actions **sfa)
+{
+ int totlen = NLA_ALIGN(from->nla_len);
+ struct nlattr *to;
+
+ to = reserve_sfa_size(sfa, from->nla_len);
+ if (IS_ERR(to))
+ return PTR_ERR(to);
+
+ memcpy(to, from, totlen);
+ return 0;
+}
+
+int ovs_nla_copy_actions(const struct nlattr *attr,
+ const struct sw_flow_key *key,
+ int depth,
+ struct sw_flow_actions **sfa)
+{
+ const struct nlattr *a;
+ int rem, err;
+
+ if (depth >= SAMPLE_ACTION_DEPTH)
+ return -EOVERFLOW;
+
+ nla_for_each_nested(a, attr, rem) {
+ /* Expected argument lengths, (u32)-1 for variable length. */
+ static const u32 action_lens[OVS_ACTION_ATTR_MAX + 1] = {
+ [OVS_ACTION_ATTR_OUTPUT] = sizeof(u32),
+ [OVS_ACTION_ATTR_USERSPACE] = (u32)-1,
+ [OVS_ACTION_ATTR_PUSH_VLAN] = sizeof(struct ovs_action_push_vlan),
+ [OVS_ACTION_ATTR_POP_VLAN] = 0,
+ [OVS_ACTION_ATTR_SET] = (u32)-1,
+ [OVS_ACTION_ATTR_SAMPLE] = (u32)-1
+ };
+ const struct ovs_action_push_vlan *vlan;
+ int type = nla_type(a);
+ bool skip_copy;
+
+ if (type > OVS_ACTION_ATTR_MAX ||
+ (action_lens[type] != nla_len(a) &&
+ action_lens[type] != (u32)-1))
+ return -EINVAL;
+
+ skip_copy = false;
+ switch (type) {
+ case OVS_ACTION_ATTR_UNSPEC:
+ return -EINVAL;
+
+ case OVS_ACTION_ATTR_USERSPACE:
+ err = validate_userspace(a);
+ if (err)
+ return err;
+ break;
+
+ case OVS_ACTION_ATTR_OUTPUT:
+ if (nla_get_u32(a) >= DP_MAX_PORTS)
+ return -EINVAL;
+ break;
+
+
+ case OVS_ACTION_ATTR_POP_VLAN:
+ break;
+
+ case OVS_ACTION_ATTR_PUSH_VLAN:
+ vlan = nla_data(a);
+ if (vlan->vlan_tpid != htons(ETH_P_8021Q))
+ return -EINVAL;
+ if (!(vlan->vlan_tci & htons(VLAN_TAG_PRESENT)))
+ return -EINVAL;
+ break;
+
+ case OVS_ACTION_ATTR_SET:
+ err = validate_set(a, key, sfa, &skip_copy);
+ if (err)
+ return err;
+ break;
+
+ case OVS_ACTION_ATTR_SAMPLE:
+ err = validate_and_copy_sample(a, key, depth, sfa);
+ if (err)
+ return err;
+ skip_copy = true;
+ break;
+
+ default:
+ return -EINVAL;
+ }
+ if (!skip_copy) {
+ err = copy_action(a, sfa);
+ if (err)
+ return err;
+ }
+ }
+
+ if (rem > 0)
+ return -EINVAL;
+
+ return 0;
+}
+
+static int sample_action_to_attr(const struct nlattr *attr, struct sk_buff *skb)
+{
+ const struct nlattr *a;
+ struct nlattr *start;
+ int err = 0, rem;
+
+ start = nla_nest_start(skb, OVS_ACTION_ATTR_SAMPLE);
+ if (!start)
+ return -EMSGSIZE;
+
+ nla_for_each_nested(a, attr, rem) {
+ int type = nla_type(a);
+ struct nlattr *st_sample;
+
+ switch (type) {
+ case OVS_SAMPLE_ATTR_PROBABILITY:
+ if (nla_put(skb, OVS_SAMPLE_ATTR_PROBABILITY,
+ sizeof(u32), nla_data(a)))
+ return -EMSGSIZE;
+ break;
+ case OVS_SAMPLE_ATTR_ACTIONS:
+ st_sample = nla_nest_start(skb, OVS_SAMPLE_ATTR_ACTIONS);
+ if (!st_sample)
+ return -EMSGSIZE;
+ err = ovs_nla_put_actions(nla_data(a), nla_len(a), skb);
+ if (err)
+ return err;
+ nla_nest_end(skb, st_sample);
+ break;
+ }
+ }
+
+ nla_nest_end(skb, start);
+ return err;
+}
+
+static int set_action_to_attr(const struct nlattr *a, struct sk_buff *skb)
+{
+ const struct nlattr *ovs_key = nla_data(a);
+ int key_type = nla_type(ovs_key);
+ struct nlattr *start;
+ int err;
+
+ switch (key_type) {
+ case OVS_KEY_ATTR_IPV4_TUNNEL:
+ start = nla_nest_start(skb, OVS_ACTION_ATTR_SET);
+ if (!start)
+ return -EMSGSIZE;
+
+ err = ipv4_tun_to_nlattr(skb, nla_data(ovs_key),
+ nla_data(ovs_key));
+ if (err)
+ return err;
+ nla_nest_end(skb, start);
+ break;
+ default:
+ if (nla_put(skb, OVS_ACTION_ATTR_SET, nla_len(a), ovs_key))
+ return -EMSGSIZE;
+ break;
+ }
+
+ return 0;
+}
+
+int ovs_nla_put_actions(const struct nlattr *attr, int len, struct sk_buff *skb)
+{
+ const struct nlattr *a;
+ int rem, err;
+
+ nla_for_each_attr(a, attr, len, rem) {
+ int type = nla_type(a);
+
+ switch (type) {
+ case OVS_ACTION_ATTR_SET:
+ err = set_action_to_attr(a, skb);
+ if (err)
+ return err;
+ break;
+
+ case OVS_ACTION_ATTR_SAMPLE:
+ err = sample_action_to_attr(a, skb);
+ if (err)
+ return err;
+ break;
+ default:
+ if (nla_put(skb, type, nla_len(a), nla_data(a)))
+ return -EMSGSIZE;
+ break;
+ }
+ }
+
+ return 0;
+}
diff --git a/net/openvswitch/flow_netlink.h b/net/openvswitch/flow_netlink.h
new file mode 100644
index 0000000..4401510
--- /dev/null
+++ b/net/openvswitch/flow_netlink.h
@@ -0,0 +1,60 @@
+/*
+ * Copyright (c) 2007-2013 Nicira, Inc.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * 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., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
+ */
+
+
+#ifndef FLOW_NETLINK_H
+#define FLOW_NETLINK_H 1
+
+#include <linux/kernel.h>
+#include <linux/netlink.h>
+#include <linux/openvswitch.h>
+#include <linux/spinlock.h>
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/if_ether.h>
+#include <linux/in6.h>
+#include <linux/jiffies.h>
+#include <linux/time.h>
+#include <linux/flex_array.h>
+
+#include <net/inet_ecn.h>
+#include <net/ip_tunnels.h>
+
+#include "flow.h"
+
+void ovs_match_init(struct sw_flow_match *match,
+ struct sw_flow_key *key, struct sw_flow_mask *mask);
+
+int ovs_nla_put_flow(const struct sw_flow_key *,
+ const struct sw_flow_key *, struct sk_buff *);
+int ovs_nla_get_flow_metadata(struct sw_flow *flow,
+ const struct nlattr *attr);
+int ovs_nla_get_match(struct sw_flow_match *match,
+ const struct nlattr *,
+ const struct nlattr *);
+
+int ovs_nla_copy_actions(const struct nlattr *attr,
+ const struct sw_flow_key *key, int depth,
+ struct sw_flow_actions **sfa);
+int ovs_nla_put_actions(const struct nlattr *attr,
+ int len, struct sk_buff *skb);
+
+struct sw_flow_actions *ovs_nla_alloc_flow_actions(int actions_len);
+void ovs_nla_free_flow_actions(struct sw_flow_actions *);
+
+#endif /* flow_netlink.h */
diff --git a/net/openvswitch/flow_table.c b/net/openvswitch/flow_table.c
new file mode 100644
index 0000000..dcadb75
--- /dev/null
+++ b/net/openvswitch/flow_table.c
@@ -0,0 +1,517 @@
+/*
+ * Copyright (c) 2007-2013 Nicira, Inc.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * 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., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
+ */
+
+#include "flow.h"
+#include "datapath.h"
+#include <linux/uaccess.h>
+#include <linux/netdevice.h>
+#include <linux/etherdevice.h>
+#include <linux/if_ether.h>
+#include <linux/if_vlan.h>
+#include <net/llc_pdu.h>
+#include <linux/kernel.h>
+#include <linux/jhash.h>
+#include <linux/jiffies.h>
+#include <linux/llc.h>
+#include <linux/module.h>
+#include <linux/in.h>
+#include <linux/rcupdate.h>
+#include <linux/if_arp.h>
+#include <linux/ip.h>
+#include <linux/ipv6.h>
+#include <linux/sctp.h>
+#include <linux/tcp.h>
+#include <linux/udp.h>
+#include <linux/icmp.h>
+#include <linux/icmpv6.h>
+#include <linux/rculist.h>
+#include <net/ip.h>
+#include <net/ipv6.h>
+#include <net/ndisc.h>
+
+static struct kmem_cache *flow_cache;
+
+static u16 range_n_bytes(const struct sw_flow_key_range *range)
+{
+ return range->end - range->start;
+}
+
+void ovs_flow_mask_key(struct sw_flow_key *dst, const struct sw_flow_key *src,
+ const struct sw_flow_mask *mask)
+{
+ const long *m = (long *)((u8 *)&mask->key + mask->range.start);
+ const long *s = (long *)((u8 *)src + mask->range.start);
+ long *d = (long *)((u8 *)dst + mask->range.start);
+ int i;
+
+ /* The memory outside of the 'mask->range' are not set since
+ * further operations on 'dst' only uses contents within
+ * 'mask->range'.
+ */
+ for (i = 0; i < range_n_bytes(&mask->range); i += sizeof(long))
+ *d++ = *s++ & *m++;
+}
+
+struct sw_flow *ovs_flow_alloc(void)
+{
+ struct sw_flow *flow;
+
+ flow = kmem_cache_alloc(flow_cache, GFP_KERNEL);
+ if (!flow)
+ return ERR_PTR(-ENOMEM);
+
+ spin_lock_init(&flow->lock);
+ flow->sf_acts = NULL;
+ flow->mask = NULL;
+
+ return flow;
+}
+
+static struct flex_array *alloc_buckets(unsigned int n_buckets)
+{
+ struct flex_array *buckets;
+ int i, err;
+
+ buckets = flex_array_alloc(sizeof(struct hlist_head),
+ n_buckets, GFP_KERNEL);
+ if (!buckets)
+ return NULL;
+
+ err = flex_array_prealloc(buckets, 0, n_buckets, GFP_KERNEL);
+ if (err) {
+ flex_array_free(buckets);
+ return NULL;
+ }
+
+ for (i = 0; i < n_buckets; i++)
+ INIT_HLIST_HEAD((struct hlist_head *)
+ flex_array_get(buckets, i));
+
+ return buckets;
+}
+
+static void flow_free(struct sw_flow *flow)
+{
+ kfree((struct sf_flow_acts __force *)flow->sf_acts);
+ kmem_cache_free(flow_cache, flow);
+}
+
+static void rcu_free_flow_callback(struct rcu_head *rcu)
+{
+ struct sw_flow *flow = container_of(rcu, struct sw_flow, rcu);
+
+ flow_free(flow);
+}
+
+void ovs_flow_free(struct sw_flow *flow, bool deferred)
+{
+ if (!flow)
+ return;
+
+ ovs_sw_flow_mask_del_ref(flow->mask, deferred);
+
+ if (deferred)
+ call_rcu(&flow->rcu, rcu_free_flow_callback);
+ else
+ flow_free(flow);
+}
+
+static void free_buckets(struct flex_array *buckets)
+{
+ flex_array_free(buckets);
+}
+
+static void __flow_tbl_destroy(struct flow_table *table)
+{
+ int i;
+
+ if (table->keep_flows)
+ goto skip_flows;
+
+ for (i = 0; i < table->n_buckets; i++) {
+ struct sw_flow *flow;
+ struct hlist_head *head = flex_array_get(table->buckets, i);
+ struct hlist_node *n;
+ int ver = table->node_ver;
+
+ hlist_for_each_entry_safe(flow, n, head, hash_node[ver]) {
+ hlist_del(&flow->hash_node[ver]);
+ ovs_flow_free(flow, false);
+ }
+ }
+
+ BUG_ON(!list_empty(table->mask_list));
+ kfree(table->mask_list);
+
+skip_flows:
+ free_buckets(table->buckets);
+ kfree(table);
+}
+
+static struct flow_table *__flow_tbl_alloc(int new_size)
+{
+ struct flow_table *table = kmalloc(sizeof(*table), GFP_KERNEL);
+
+ if (!table)
+ return NULL;
+
+ table->buckets = alloc_buckets(new_size);
+
+ if (!table->buckets) {
+ kfree(table);
+ return NULL;
+ }
+ table->n_buckets = new_size;
+ table->count = 0;
+ table->node_ver = 0;
+ table->keep_flows = false;
+ get_random_bytes(&table->hash_seed, sizeof(u32));
+ table->mask_list = NULL;
+
+ return table;
+}
+
+struct flow_table *ovs_flow_tbl_alloc(int new_size)
+{
+ struct flow_table *table = __flow_tbl_alloc(new_size);
+
+ if (!table)
+ return NULL;
+
+ table->mask_list = kmalloc(sizeof(struct list_head), GFP_KERNEL);
+ if (!table->mask_list) {
+ table->keep_flows = true;
+ __flow_tbl_destroy(table);
+ return NULL;
+ }
+ INIT_LIST_HEAD(table->mask_list);
+
+ return table;
+}
+
+static void flow_tbl_destroy_rcu_cb(struct rcu_head *rcu)
+{
+ struct flow_table *table = container_of(rcu, struct flow_table, rcu);
+
+ __flow_tbl_destroy(table);
+}
+
+void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred)
+{
+ if (!table)
+ return;
+
+ if (deferred)
+ call_rcu(&table->rcu, flow_tbl_destroy_rcu_cb);
+ else
+ __flow_tbl_destroy(table);
+}
+
+struct sw_flow *ovs_flow_tbl_dump_next(struct flow_table *table,
+ u32 *bucket, u32 *last)
+{
+ struct sw_flow *flow;
+ struct hlist_head *head;
+ int ver;
+ int i;
+
+ ver = table->node_ver;
+ while (*bucket < table->n_buckets) {
+ i = 0;
+ head = flex_array_get(table->buckets, *bucket);
+ hlist_for_each_entry_rcu(flow, head, hash_node[ver]) {
+ if (i < *last) {
+ i++;
+ continue;
+ }
+ *last = i + 1;
+ return flow;
+ }
+ (*bucket)++;
+ *last = 0;
+ }
+
+ return NULL;
+}
+
+static struct hlist_head *find_bucket(struct flow_table *table, u32 hash)
+{
+ hash = jhash_1word(hash, table->hash_seed);
+ return flex_array_get(table->buckets,
+ (hash & (table->n_buckets - 1)));
+}
+
+static void __tbl_insert(struct flow_table *table, struct sw_flow *flow)
+{
+ struct hlist_head *head;
+
+ head = find_bucket(table, flow->hash);
+ hlist_add_head_rcu(&flow->hash_node[table->node_ver], head);
+
+ table->count++;
+}
+
+static void flow_table_copy_flows(struct flow_table *old,
+ struct flow_table *new)
+{
+ int old_ver;
+ int i;
+
+ old_ver = old->node_ver;
+ new->node_ver = !old_ver;
+
+ /* Insert in new table. */
+ for (i = 0; i < old->n_buckets; i++) {
+ struct sw_flow *flow;
+ struct hlist_head *head;
+
+ head = flex_array_get(old->buckets, i);
+
+ hlist_for_each_entry(flow, head, hash_node[old_ver])
+ __tbl_insert(new, flow);
+ }
+
+ new->mask_list = old->mask_list;
+ old->keep_flows = true;
+}
+
+static struct flow_table *__flow_tbl_rehash(struct flow_table *table,
+ int n_buckets)
+{
+ struct flow_table *new_table;
+
+ new_table = __flow_tbl_alloc(n_buckets);
+ if (!new_table)
+ return ERR_PTR(-ENOMEM);
+
+ flow_table_copy_flows(table, new_table);
+
+ return new_table;
+}
+
+struct flow_table *ovs_flow_tbl_rehash(struct flow_table *table)
+{
+ return __flow_tbl_rehash(table, table->n_buckets);
+}
+
+struct flow_table *ovs_flow_tbl_expand(struct flow_table *table)
+{
+ return __flow_tbl_rehash(table, table->n_buckets * 2);
+}
+
+static u32 flow_hash(const struct sw_flow_key *key, int key_start,
+ int key_end)
+{
+ u32 *hash_key = (u32 *)((u8 *)key + key_start);
+ int hash_u32s = (key_end - key_start) >> 2;
+
+ /* Make sure number of hash bytes are multiple of u32. */
+ BUILD_BUG_ON(sizeof(long) % sizeof(u32));
+
+ return jhash2(hash_key, hash_u32s, 0);
+}
+
+static int flow_key_start(const struct sw_flow_key *key)
+{
+ if (key->tun_key.ipv4_dst)
+ return 0;
+ else
+ return rounddown(offsetof(struct sw_flow_key, phy),
+ sizeof(long));
+}
+
+static bool cmp_key(const struct sw_flow_key *key1,
+ const struct sw_flow_key *key2,
+ int key_start, int key_end)
+{
+ const long *cp1 = (long *)((u8 *)key1 + key_start);
+ const long *cp2 = (long *)((u8 *)key2 + key_start);
+ long diffs = 0;
+ int i;
+
+ for (i = key_start; i < key_end; i += sizeof(long))
+ diffs |= *cp1++ ^ *cp2++;
+
+ return diffs == 0;
+}
+
+static bool flow_cmp_masked_key(const struct sw_flow *flow,
+ const struct sw_flow_key *key,
+ int key_start, int key_end)
+{
+ return cmp_key(&flow->key, key, key_start, key_end);
+}
+
+bool ovs_flow_cmp_unmasked_key(const struct sw_flow *flow,
+ struct sw_flow_match *match)
+{
+ struct sw_flow_key *key = match->key;
+ int key_start = flow_key_start(key);
+ int key_end = match->range.end;
+
+ return cmp_key(&flow->unmasked_key, key, key_start, key_end);
+}
+
+static struct sw_flow *masked_flow_lookup(struct flow_table *table,
+ const struct sw_flow_key *unmasked,
+ struct sw_flow_mask *mask)
+{
+ struct sw_flow *flow;
+ struct hlist_head *head;
+ int key_start = mask->range.start;
+ int key_end = mask->range.end;
+ u32 hash;
+ struct sw_flow_key masked_key;
+
+ ovs_flow_mask_key(&masked_key, unmasked, mask);
+ hash = flow_hash(&masked_key, key_start, key_end);
+ head = find_bucket(table, hash);
+ hlist_for_each_entry_rcu(flow, head, hash_node[table->node_ver]) {
+ if (flow->mask == mask &&
+ flow_cmp_masked_key(flow, &masked_key,
+ key_start, key_end))
+ return flow;
+ }
+ return NULL;
+}
+
+struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *tbl,
+ const struct sw_flow_key *key)
+{
+ struct sw_flow *flow = NULL;
+ struct sw_flow_mask *mask;
+
+ list_for_each_entry_rcu(mask, tbl->mask_list, list) {
+ flow = masked_flow_lookup(tbl, key, mask);
+ if (flow) /* Found */
+ break;
+ }
+
+ return flow;
+}
+
+void ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow)
+{
+ flow->hash = flow_hash(&flow->key, flow->mask->range.start,
+ flow->mask->range.end);
+ __tbl_insert(table, flow);
+}
+
+void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow)
+{
+ BUG_ON(table->count == 0);
+ hlist_del_rcu(&flow->hash_node[table->node_ver]);
+ table->count--;
+}
+
+struct sw_flow_mask *ovs_sw_flow_mask_alloc(void)
+{
+ struct sw_flow_mask *mask;
+
+ mask = kmalloc(sizeof(*mask), GFP_KERNEL);
+ if (mask)
+ mask->ref_count = 0;
+
+ return mask;
+}
+
+void ovs_sw_flow_mask_add_ref(struct sw_flow_mask *mask)
+{
+ mask->ref_count++;
+}
+
+static void rcu_free_sw_flow_mask_cb(struct rcu_head *rcu)
+{
+ struct sw_flow_mask *mask = container_of(rcu, struct sw_flow_mask, rcu);
+
+ kfree(mask);
+}
+
+void ovs_sw_flow_mask_del_ref(struct sw_flow_mask *mask, bool deferred)
+{
+ if (!mask)
+ return;
+
+ BUG_ON(!mask->ref_count);
+ mask->ref_count--;
+
+ if (!mask->ref_count) {
+ list_del_rcu(&mask->list);
+ if (deferred)
+ call_rcu(&mask->rcu, rcu_free_sw_flow_mask_cb);
+ else
+ kfree(mask);
+ }
+}
+
+static bool mask_equal(const struct sw_flow_mask *a,
+ const struct sw_flow_mask *b)
+{
+ u8 *a_ = (u8 *)&a->key + a->range.start;
+ u8 *b_ = (u8 *)&b->key + b->range.start;
+
+ return (a->range.end == b->range.end)
+ && (a->range.start == b->range.start)
+ && (memcmp(a_, b_, range_n_bytes(&a->range)) == 0);
+}
+
+struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *tbl,
+ const struct sw_flow_mask *mask)
+{
+ struct list_head *ml;
+
+ list_for_each(ml, tbl->mask_list) {
+ struct sw_flow_mask *m;
+ m = container_of(ml, struct sw_flow_mask, list);
+ if (mask_equal(mask, m))
+ return m;
+ }
+
+ return NULL;
+}
+
+/**
+ * add a new mask into the mask list.
+ * The caller needs to make sure that 'mask' is not the same
+ * as any masks that are already on the list.
+ */
+void ovs_sw_flow_mask_insert(struct flow_table *tbl, struct sw_flow_mask *mask)
+{
+ list_add_rcu(&mask->list, tbl->mask_list);
+}
+
+/* Initializes the flow module.
+ * Returns zero if successful or a negative error code. */
+int ovs_flow_init(void)
+{
+ BUILD_BUG_ON(__alignof__(struct sw_flow_key) % __alignof__(long));
+ BUILD_BUG_ON(sizeof(struct sw_flow_key) % sizeof(long));
+
+ flow_cache = kmem_cache_create("sw_flow", sizeof(struct sw_flow), 0,
+ 0, NULL);
+ if (flow_cache == NULL)
+ return -ENOMEM;
+
+ return 0;
+}
+
+/* Uninitializes the flow module. */
+void ovs_flow_exit(void)
+{
+ kmem_cache_destroy(flow_cache);
+}
diff --git a/net/openvswitch/flow_table.h b/net/openvswitch/flow_table.h
new file mode 100644
index 0000000..d7a1144
--- /dev/null
+++ b/net/openvswitch/flow_table.h
@@ -0,0 +1,91 @@
+/*
+ * Copyright (c) 2007-2013 Nicira, Inc.
+ *
+ * This program is free software; you can redistribute it and/or
+ * modify it under the terms of version 2 of the GNU General Public
+ * License as published by the Free Software Foundation.
+ *
+ * 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., 51 Franklin Street, Fifth Floor, Boston, MA
+ * 02110-1301, USA
+ */
+
+#ifndef FLOW_TABLE_H
+#define FLOW_TABLE_H 1
+
+#include <linux/kernel.h>
+#include <linux/netlink.h>
+#include <linux/openvswitch.h>
+#include <linux/spinlock.h>
+#include <linux/types.h>
+#include <linux/rcupdate.h>
+#include <linux/if_ether.h>
+#include <linux/in6.h>
+#include <linux/jiffies.h>
+#include <linux/time.h>
+#include <linux/flex_array.h>
+
+#include <net/inet_ecn.h>
+#include <net/ip_tunnels.h>
+
+#include "flow.h"
+
+#define TBL_MIN_BUCKETS 1024
+
+struct flow_table {
+ struct flex_array *buckets;
+ unsigned int count, n_buckets;
+ struct rcu_head rcu;
+ struct list_head *mask_list;
+ int node_ver;
+ u32 hash_seed;
+ bool keep_flows;
+};
+
+int ovs_flow_init(void);
+void ovs_flow_exit(void);
+
+struct sw_flow *ovs_flow_alloc(void);
+void ovs_flow_free(struct sw_flow *, bool deferred);
+
+static inline int ovs_flow_tbl_count(struct flow_table *table)
+{
+ return table->count;
+}
+
+static inline int ovs_flow_tbl_need_to_expand(struct flow_table *table)
+{
+ return (table->count > table->n_buckets);
+}
+
+struct flow_table *ovs_flow_tbl_alloc(int new_size);
+struct flow_table *ovs_flow_tbl_expand(struct flow_table *table);
+struct flow_table *ovs_flow_tbl_rehash(struct flow_table *table);
+void ovs_flow_tbl_destroy(struct flow_table *table, bool deferred);
+
+void ovs_flow_tbl_insert(struct flow_table *table, struct sw_flow *flow);
+void ovs_flow_tbl_remove(struct flow_table *table, struct sw_flow *flow);
+struct sw_flow *ovs_flow_tbl_dump_next(struct flow_table *table,
+ u32 *bucket, u32 *idx);
+struct sw_flow *ovs_flow_tbl_lookup(struct flow_table *,
+ const struct sw_flow_key *);
+
+bool ovs_flow_cmp_unmasked_key(const struct sw_flow *flow,
+ struct sw_flow_match *match);
+
+struct sw_flow_mask *ovs_sw_flow_mask_alloc(void);
+void ovs_sw_flow_mask_add_ref(struct sw_flow_mask *);
+void ovs_sw_flow_mask_del_ref(struct sw_flow_mask *, bool deferred);
+void ovs_sw_flow_mask_insert(struct flow_table *, struct sw_flow_mask *);
+struct sw_flow_mask *ovs_sw_flow_mask_find(const struct flow_table *,
+ const struct sw_flow_mask *);
+void ovs_flow_mask_key(struct sw_flow_key *dst, const struct sw_flow_key *src,
+ const struct sw_flow_mask *mask);
+
+#endif /* flow_table.h */
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 03/11] openvswitch: remove duplicated include from vport-gre.c
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Wei Yongjun, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Remove duplicated include.
Signed-off-by: Wei Yongjun <yongjun_wei@trendmicro.com.cn>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
net/openvswitch/vport-gre.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/net/openvswitch/vport-gre.c b/net/openvswitch/vport-gre.c
index c99dea5..a3d6951 100644
--- a/net/openvswitch/vport-gre.c
+++ b/net/openvswitch/vport-gre.c
@@ -24,8 +24,6 @@
#include <linux/if_tunnel.h>
#include <linux/if_vlan.h>
#include <linux/in.h>
-#include <linux/if_vlan.h>
-#include <linux/in.h>
#include <linux/in_route.h>
#include <linux/inetdevice.h>
#include <linux/jhash.h>
--
1.8.3.2
^ permalink raw reply related
* [PATCH net-next 01/11] openvswitch: Move flow table rehashing to flow install.
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev, Pravin B Shelar, Jesse Gross
In-Reply-To: <1383092544-50599-1-git-send-email-jesse@nicira.com>
From: Pravin B Shelar <pshelar@nicira.com>
Rehashing in ovs-workqueue can cause ovs-mutex lock contentions
in case of heavy flow setups where both needs ovs-mutex. So by
moving rehashing to flow-setup we can eliminate contention.
This also simplify ovs locking and reduces dependence on
workqueue.
Signed-off-by: Pravin B Shelar <pshelar@nicira.com>
Signed-off-by: Jesse Gross <jesse@nicira.com>
---
net/openvswitch/datapath.c | 50 ++++++++++------------------------------------
net/openvswitch/datapath.h | 2 ++
2 files changed, 13 insertions(+), 39 deletions(-)
diff --git a/net/openvswitch/datapath.c b/net/openvswitch/datapath.c
index 2aa13bd..2e1a9c2 100644
--- a/net/openvswitch/datapath.c
+++ b/net/openvswitch/datapath.c
@@ -60,8 +60,6 @@
#define REHASH_FLOW_INTERVAL (10 * 60 * HZ)
-static void rehash_flow_table(struct work_struct *work);
-static DECLARE_DELAYED_WORK(rehash_flow_wq, rehash_flow_table);
int ovs_net_id __read_mostly;
@@ -1289,22 +1287,25 @@ static int ovs_flow_cmd_new_or_set(struct sk_buff *skb, struct genl_info *info)
/* Check if this is a duplicate flow */
flow = ovs_flow_lookup(table, &key);
if (!flow) {
+ struct flow_table *new_table = NULL;
struct sw_flow_mask *mask_p;
+
/* Bail out if we're not allowed to create a new flow. */
error = -ENOENT;
if (info->genlhdr->cmd == OVS_FLOW_CMD_SET)
goto err_unlock_ovs;
/* Expand table, if necessary, to make room. */
- if (ovs_flow_tbl_need_to_expand(table)) {
- struct flow_table *new_table;
-
+ if (ovs_flow_tbl_need_to_expand(table))
new_table = ovs_flow_tbl_expand(table);
- if (!IS_ERR(new_table)) {
- rcu_assign_pointer(dp->table, new_table);
- ovs_flow_tbl_destroy(table, true);
- table = ovsl_dereference(dp->table);
- }
+ else if (time_after(jiffies, dp->last_rehash + REHASH_FLOW_INTERVAL))
+ new_table = ovs_flow_tbl_rehash(table);
+
+ if (new_table && !IS_ERR(new_table)) {
+ rcu_assign_pointer(dp->table, new_table);
+ ovs_flow_tbl_destroy(table, true);
+ table = ovsl_dereference(dp->table);
+ dp->last_rehash = jiffies;
}
/* Allocate flow. */
@@ -2336,32 +2337,6 @@ error:
return err;
}
-static void rehash_flow_table(struct work_struct *work)
-{
- struct datapath *dp;
- struct net *net;
-
- ovs_lock();
- rtnl_lock();
- for_each_net(net) {
- struct ovs_net *ovs_net = net_generic(net, ovs_net_id);
-
- list_for_each_entry(dp, &ovs_net->dps, list_node) {
- struct flow_table *old_table = ovsl_dereference(dp->table);
- struct flow_table *new_table;
-
- new_table = ovs_flow_tbl_rehash(old_table);
- if (!IS_ERR(new_table)) {
- rcu_assign_pointer(dp->table, new_table);
- ovs_flow_tbl_destroy(old_table, true);
- }
- }
- }
- rtnl_unlock();
- ovs_unlock();
- schedule_delayed_work(&rehash_flow_wq, REHASH_FLOW_INTERVAL);
-}
-
static int __net_init ovs_init_net(struct net *net)
{
struct ovs_net *ovs_net = net_generic(net, ovs_net_id);
@@ -2419,8 +2394,6 @@ static int __init dp_init(void)
if (err < 0)
goto error_unreg_notifier;
- schedule_delayed_work(&rehash_flow_wq, REHASH_FLOW_INTERVAL);
-
return 0;
error_unreg_notifier:
@@ -2437,7 +2410,6 @@ error:
static void dp_cleanup(void)
{
- cancel_delayed_work_sync(&rehash_flow_wq);
dp_unregister_genl(ARRAY_SIZE(dp_genl_families));
unregister_netdevice_notifier(&ovs_dp_device_notifier);
unregister_pernet_device(&ovs_net_ops);
diff --git a/net/openvswitch/datapath.h b/net/openvswitch/datapath.h
index 4d109c1..2c15541 100644
--- a/net/openvswitch/datapath.h
+++ b/net/openvswitch/datapath.h
@@ -62,6 +62,7 @@ struct dp_stats_percpu {
* ovs_mutex and RCU.
* @stats_percpu: Per-CPU datapath statistics.
* @net: Reference to net namespace.
+ * @last_rehash: Timestamp of last rehash.
*
* Context: See the comment on locking at the top of datapath.c for additional
* locking information.
@@ -83,6 +84,7 @@ struct datapath {
/* Network namespace ref. */
struct net *net;
#endif
+ unsigned long last_rehash;
};
/**
--
1.8.3.2
^ permalink raw reply related
* [GIT net-next] Open vSwitch
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller; +Cc: netdev, dev
A set of updates for net-next/3.13. Major changes are:
* Restructure flow handling code to be more logically organized and
easier to read.
* Previously flow state was effectively per-CPU but this is no longer
true with the addition of wildcarded flows (megaflows). While good
for flow setup rates, it is bad for stats updates. Stats are now
per-CPU again to get the best of both worlds.
* Rehashing of the flow table is moved from a workqueue to flow
installation time. Before, heavy load could block the workqueue for
excessive periods of time.
* Additional debugging information is provided to help diagnose megaflows.
* It's now possible to match on TCP flags.
The following changes since commit 272b98c6455f00884f0350f775c5342358ebb73f:
Linux 3.12-rc1 (2013-09-16 16:17:51 -0400)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/jesse/openvswitch.git master
for you to fetch changes up to c7e8b9659587449b21ee68d7aee0cacadf650cce:
openvswitch: TCP flags matching support. (2013-10-23 02:19:50 -0700)
----------------------------------------------------------------
Andy Zhou (1):
openvswitch: collect mega flow mask stats
Jarno Rajahalme (2):
openvswitch: Widen TCP flags handling.
openvswitch: TCP flags matching support.
Pravin B Shelar (6):
openvswitch: Move flow table rehashing to flow install.
openvswitch: Restructure datapath.c and flow.c
openvswitch: Move mega-flow list out of rehashing struct.
openvswitch: Simplify mega-flow APIs.
openvswitch: Per cpu flow stats.
openvswitch: Enable all GSO features on internal port.
Wei Yongjun (2):
openvswitch: remove duplicated include from vport-vxlan.c
openvswitch: remove duplicated include from vport-gre.c
include/uapi/linux/openvswitch.h | 18 +-
net/openvswitch/Makefile | 2 +
net/openvswitch/datapath.c | 721 ++-------------
net/openvswitch/datapath.h | 9 +-
net/openvswitch/flow.c | 1631 ++--------------------------------
net/openvswitch/flow.h | 148 +--
net/openvswitch/flow_netlink.c | 1630 +++++++++++++++++++++++++++++++++
net/openvswitch/flow_netlink.h | 60 ++
net/openvswitch/flow_table.c | 600 +++++++++++++
net/openvswitch/flow_table.h | 81 ++
net/openvswitch/vport-gre.c | 2 -
net/openvswitch/vport-internal_dev.c | 2 +-
net/openvswitch/vport-vxlan.c | 1 -
13 files changed, 2590 insertions(+), 2315 deletions(-)
create mode 100644 net/openvswitch/flow_netlink.c
create mode 100644 net/openvswitch/flow_netlink.h
create mode 100644 net/openvswitch/flow_table.c
create mode 100644 net/openvswitch/flow_table.h
^ permalink raw reply
* [PATCH net-next 02/11] openvswitch: remove duplicated include from vport-vxlan.c
From: Jesse Gross @ 2013-10-30 0:22 UTC (permalink / raw)
To: David Miller
Cc: dev-yBygre7rU0TnMu66kgdUjQ, netdev-u79uwXL29TY76Z2rM5mHXA,
Wei Yongjun
In-Reply-To: <1383092544-50599-1-git-send-email-jesse-l0M0P4e3n4LQT0dZR+AlfA@public.gmane.org>
From: Wei Yongjun <yongjun_wei-zrsr2BFq86L20UzCJQGyNP8+0UxHXcjY@public.gmane.org>
Remove duplicated include.
Signed-off-by: Wei Yongjun <yongjun_wei-zrsr2BFq86L20UzCJQGyNP8+0UxHXcjY@public.gmane.org>
Signed-off-by: Jesse Gross <jesse-l0M0P4e3n4LQT0dZR+AlfA@public.gmane.org>
---
net/openvswitch/vport-vxlan.c | 1 -
1 file changed, 1 deletion(-)
diff --git a/net/openvswitch/vport-vxlan.c b/net/openvswitch/vport-vxlan.c
index a481c03..b0da394 100644
--- a/net/openvswitch/vport-vxlan.c
+++ b/net/openvswitch/vport-vxlan.c
@@ -29,7 +29,6 @@
#include <net/ip.h>
#include <net/udp.h>
#include <net/ip_tunnels.h>
-#include <net/udp.h>
#include <net/rtnetlink.h>
#include <net/route.h>
#include <net/dsfield.h>
--
1.8.3.2
^ permalink raw reply related
* named network namespace -- setns() with Invalid argument (errno 22)
From: Dilip Daya @ 2013-10-30 0:16 UTC (permalink / raw)
To: netdev; +Cc: Eric W. Biederman
Hi All,
Is the following intended behavior for adding "nested" named network namespaces ?
Steps to reproduce:
# uname -r
3.10.1
# /sbin/ip -V
ip utility, iproute2-ss130903
Existing network namespaces:
# ip netns list
NETNS0
NETNS1
List of named network namespace objects with inode/permissions:
# ls -li /var/run/netns/
total 0
4026532310 -r--r--r-- 1 root root 0 Oct 29 09:11 NETNS0
4026532366 -r--r--r-- 1 root root 0 Oct 29 09:13 NETNS1
Enter existing named network namespace:
# ip netns exec NETNS0 bash
List network devices for named netns:
# ls -l /sys/class/net/
total 0
lrwxrwxrwx 1 root root 0 Oct 29 12:25 lo -> ../../devices/virtual/net/lo/
List of named network namespace objects with inode/permissions:
# ls -li /var/run/netns/
total 0
4026532310 -r--r--r-- 1 root root 0 Oct 29 09:11 NETNS0
4026532366 -r--r--r-- 1 root root 0 Oct 29 09:13 NETNS1
# ip netns add NETNS0a <<< adding NETNS0a from within NETNS0
# ls -li /var/run/netns/
total 0
4026532310 -r--r--r-- 1 root root 0 Oct 29 09:11 NETNS0
4026532366 -r--r--r-- 1 root root 0 Oct 29 09:13 NETNS1
4026532423 -r--r--r-- 1 root root 0 Oct 29 12:28 NETNS0a
^^^^^^^^^^ ^^^^^^^^^^
inode permissions
# ip netns exec NETNS0a ls -l /sys/class/net/
total 0
lrwxrwxrwx 1 root root 0 Oct 29 12:28 lo -> ../../devices/virtual/net/lo
# exit <<< exiting from NETNS0
Listing from host/default namespace:
# ls -li /var/run/netns/
total 0
4026532310 -r--r--r-- 1 root root 0 Oct 29 09:11 NETNS0
4026532366 -r--r--r-- 1 root root 0 Oct 29 09:13 NETNS1
964863 ---------- 1 root root 0 Oct 29 12:28 NETNS0a <<< NULL permissions
^^^^^^ ^^^^^^^^^^
Re-enter NETNS0:
# ip netns exec NETNS0 bash
# ls -li /var/run/netns/
total 0
4026532310 -r--r--r-- 1 root root 0 Oct 29 09:11 NETNS0
4026532366 -r--r--r-- 1 root root 0 Oct 29 09:13 NETNS1
964863 ---------- 1 root root 0 Oct 29 12:28 NETNS0a <<< NULL permissions
^^^^^^^^^^
# ip netns exec NETNS0a ls -l /sys/class/net/
seting the network namespace "NETNS0a" failed: Invalid argument
=> It seems the bash shell that created the nested named netns is the only
one that can view/enter the nested named netns. All other attempts from
either another bash shell or host/default namespace will get a different
inode with NULL permissions. Once the initial bash shell that created the
nested named netns exists the nested netns is rendered unusable due to
NULL permissions on its inode. setns() Invalid argument (errno 22) seems
to be due to NULL permissions on /var/run/netns/<netnsName> object.
Thanks.
^ permalink raw reply
* From Mr. Charles Kabore.
From: Mr. Charles Kabore @ 2013-10-30 0:08 UTC (permalink / raw)
>From Mr. Charles Kabore.
Ouagadougou,
Burkina Faso.
Dear Friend,
Good day to you. I must say sorry for interfering into your privacy without seeking for permission. It's because of urgency of this matter. Firstly, I am Mr. Charles Kabore a lawyer and personal confidant to Abdullah Senussi who was the intelligence chief of Colonel Muammar Gaddafi. I need your urgent assistance in transferring the sum of ($39.5) million to your account within 14 banking days from a bank in Burkina Faso. This money belongs to my master Abdullah Senussi and was deposited in the bank on the name of his son. The urgent need for the transfer of this fund is to avoid confiscation by the Libyan government as they quest the seizure of every related assets belonging to Late Colonel Muammar Gaddafi and his aides. I am contacting you in a good faith so that the bank will release the
money to you for safe keeping/investments till the release of my master who is now in custody.
So if you are capable of receiving this huge amount of money,let me have a positive response from you via return mail for more personal discussions on how we are going to go about it. but if you are not convinced to handle this deal with me then do me a favor by deleting this message from your mail box and help me seal the confidentiality of this matter. thank you for your understanding and hope to hear from you soon.
Best.Regards,
Mr. Charles Kabore.
^ permalink raw reply
* Re: ipv6 fragmentation-related panic in netfilter
From: Patrick McHardy @ 2013-10-30 0:07 UTC (permalink / raw)
To: Tomas Hlavacek; +Cc: netdev, netfilter-devel
In-Reply-To: <2060a7d2-c307-4e30-b1d4-0bd26c904d6f@gmail.com>
On Tue, Oct 29, 2013 at 10:07:59PM +0100, Tomas Hlavacek wrote:
> Hi!
>
> I have encountered following condition on 3 distinct hosts in last
> few days. Hosts are failing several times a day (4 to 7 times) and
> it usually happens roughly at the same time. Affected hosts has
> almost exactly the same HW, but different kernel versions from
> Debian (Wheezy) default 3.2 up to 3.11.6.
>
>
> KERNEL: /usr/src/vmlinux DUMPFILE:
> dump.201310291545 [PARTIAL DUMP]
> CPUS: 16
> DATE: Tue Oct 29 15:45:11 2013
> UPTIME: 06:04:17
> LOAD AVERAGE: 0.04, 0.25, 0.32
> TASKS: 211
> NODENAME: fw03a
> RELEASE: 3.11.6
> VERSION: #2 SMP Mon Oct 28 20:29:03 CET 2013
> MACHINE: x86_64 (2393 Mhz)
> MEMORY: 12 GB
> PANIC: PID: 0
> COMMAND: "swapper/1"
> TASK: ffff8801b90ac7b0 (1 of 16) [THREAD_INFO: ffff8801b90b4000]
> CPU: 1
> STATE: TASK_RUNNING (PANIC)
>
> crash> bt
> PID: 0 TASK: ffff8801b90ac7b0 CPU: 1 COMMAND: "swapper/1"
> #0 [ffff8801bfc235d0] machine_kexec at ffffffff81032f68
> #1 [ffff8801bfc23610] crash_kexec at ffffffff8109e055
> #2 [ffff8801bfc236e0] oops_end at ffffffff81005e90
> #3 [ffff8801bfc23700] do_invalid_op at ffffffff81003004
> #4 [ffff8801bfc237a0] invalid_op at ffffffff8142b368
> [exception RIP: pskb_expand_head+596]
> RIP: ffffffff81333c74 RSP: ffff8801bfc23850 RFLAGS: 00010202
> RAX: 0000000000000003 RBX: ffff8801b6d99080 RCX: 0000000000000020
> RDX: 00000000000005f4 RSI: 0000000000000000 RDI: ffff8801b6d99080
> RBP: 0000000040115833 R8: 00000000000002c0 R9: ffff8801b8cf2c00
> R10: 000000000000ffff R11: 00000000197033fe R12: 0000000000000000
> R13: ffff880337b59a00 R14: ffffffffa03fb160 R15: ffff880337b59a00
> ORIG_RAX: ffffffffffffffff CS: 0010 SS: 0018
> #5 [ffff8801bfc23858] __nf_conntrack_confirm at ffffffffa03ace16
> [nf_conntrack]
> #6 [ffff8801bfc238c8] vlan_netlink_fini at ffffffffa03fb160 [8021q]
> #7 [ffff8801bfc23928] dev_queue_xmit at ffffffff81342d79
> #8 [ffff8801bfc23978] ip6_finish_output2 at ffffffff813d26ee
> #9 [ffff8801bfc239c8] ip6_forward at ffffffff813d44be
> #10 [ffff8801bfc23a48] __ipv6_conntrack_in at ffffffffa034f7b6
> [nf_conntrack_ipv6]
> #11 [ffff8801bfc23a98] nf_iterate at ffffffff8136ba0d
> #12 [ffff8801bfc23af8] nf_hook_slow at ffffffff8136baae
> #13 [ffff8801bfc23b68] nf_ct_frag6_output at ffffffffa039decf
> [nf_defrag_ipv6]
> #14 [ffff8801bfc23bd8] ipv6_defrag at ffffffffa039d0c1 [nf_defrag_ipv6]
> #15 [ffff8801bfc23c18] nf_iterate at ffffffff8136ba0d
> #16 [ffff8801bfc23c78] nf_hook_slow at ffffffff8136baae
> #17 [ffff8801bfc23ce8] ipv6_rcv at ffffffff813d59f5
> #18 [ffff8801bfc23d38] __netif_receive_skb_core at ffffffff813410db
> #19 [ffff8801bfc23db8] napi_gro_receive at ffffffff81341d88
> #20 [ffff8801bfc23dd8] igb_poll at ffffffffa0035867 [igb]
> #21 [ffff8801bfc23e88] net_rx_action at ffffffff81341ac9
> #22 [ffff8801bfc23ed8] __do_softirq at ffffffff81049fb6
> #23 [ffff8801bfc23f38] call_softirq at ffffffff8142b4fc
> #24 [ffff8801bfc23f50] do_softirq at ffffffff8100481d
> #25 [ffff8801bfc23f80] do_IRQ at ffffffff810043bb
> --- <IRQ stack> ---
> #26 [ffff8801b90b5db8] ret_from_intr at ffffffff81429baa
> [exception RIP: cpuidle_enter_state+86]
> RIP: ffffffff813107a6 RSP: ffff8801b90b5e68 RFLAGS: 00000216
> RAX: 000000000007ff2b RBX: 0000000140523c4c RCX: 0000000000000018
> RDX: 0000000225c17d03 RSI: 0000000000000000 RDI: ffffffff81812600
> RBP: 0000000000000004 R8: 0000000000000018 R9: 00000000000006cf
> R10: 0000000000000001 R11: 0000000000000006 R12: 0000000100523c4e
> R13: 0000000000000000 R14: ffffffff81066415 R15: 0000000000000086
> ORIG_RAX: ffffffffffffff94 CS: 0010 SS: 0018
> #27 [ffff8801b90b5eb0] cpuidle_idle_call at ffffffff813108ce
> #28 [ffff8801b90b5ee0] arch_cpu_idle at ffffffff8100b769
> #29 [ffff8801b90b5ef0] cpu_startup_entry at ffffffff81086b1d
> #30 [ffff8801b90b5f30] start_secondary at ffffffff8102af40
>
> I am investigating at the moment. All suggestions/help would be
> appreciated.
The problem is that the reassembled packet is referenced by the individual
fragments, so we trigger the BUG_ON in pskb_expand_head(). In this
particular case the case we BUG() on is actually OK, but I'm looking at
a way we can fix this without special casing. Hope to have a patch for
testing in the next hours.
^ permalink raw reply
* Re: [PATCH v2 net-next] net: introduce gro_frag_list_enable sysctl
From: Ben Hutchings @ 2013-10-30 0:06 UTC (permalink / raw)
To: David Miller
Cc: eric.dumazet, christoph.paasch, herbert, netdev, hkchu, mwdalton
In-Reply-To: <20131029.194446.2215574000648693370.davem@davemloft.net>
On Tue, 2013-10-29 at 19:44 -0400, David Miller wrote:
> From: Eric Dumazet <eric.dumazet@gmail.com>
> Date: Tue, 29 Oct 2013 08:12:35 -0700
>
> > From: Eric Dumazet <edumazet@google.com>
> >
> > Christoph Paasch and Jerry Chu reported crashes in skb_segment() caused
> > by commit 8a29111c7ca6 ("net: gro: allow to build full sized skb")
> >
> > (Jerry is working on adding native GRO support for tunnels)
> >
> > skb_segment() only deals with a frag_list chain containing MSS sized
> > fragments.
> >
> > This patch adds support any kind of frag, and adds a new sysctl,
> > as clearly the GRO layer should avoid building frag_list skbs
> > on a router, as the segmentation is adding cpu overhead.
> >
> > Note that we could try to reuse page fragments instead of doing
> > copy to linear skbs, but this requires a fair amount of work,
> > and possible truesize nightmares, as we do not track individual
> > (per page fragment) truesizes.
> >
> > /proc/sys/net/core/gro_frag_list_enable possible values are :
> >
> > 0 : GRO layer is not allowed to use frag_list to extend skb capacity
> > 1 : GRO layer is allowed to use frag_list, but skb_segment()
> > automatically sets the sysctl to 0.
> > 2 : GRO is allowed to use frag_list, and skb_segment() wont
> > clear the sysctl.
> >
> > Default value is 1 : automatic discovery
> >
> > Reported-by: Christoph Paasch <christoph.paasch@uclouvain.be>
> > Reported-by: Jerry Chu <hkchu@google.com>
> > Cc: Michael Dalton <mwdalton@google.com>
> > Signed-off-by: Eric Dumazet <edumazet@google.com>
> > ---
> > v2: added missing sysctl definition in skbuff.c
>
> I do not like the idea of packet actions indirectly changing sysctl
> values, even if you document it sufficiently as you have here.
>
> Plus this puts the sysctl change logic in a fast path.
>
> I would suggest instead making it change in response to changes to
> ip_forward, as we do with per-device LRO settings. This means that,
> like ip_forward, you should also make this sysctl a global + devinet
> per-device sysctl.
>
> You might even emit a pr_info() when this logic triggers, and if you
> are ambitious enough keep track of the previous GRO sysctl state so
> you can restore it if ip_forward is set back to zero.
Speaking of which: insteading of disabling LRO once, we really ought to
keep count of the forwarders (IPv4 routing, IPv6 routing, bridge) and
use that to mask out LRO in netdev_fix_features().
I think that the forwarder count is also needed for this.
Ben.
--
Ben Hutchings, Staff Engineer, Solarflare
Not speaking for my employer; that's the marketing department's job.
They asked us to note that Solarflare product names are trademarked.
^ permalink raw reply
* Re: [PATCH 00/19] Enable various Renesas drivers on all ARM platforms
From: Simon Horman @ 2013-10-30 0:05 UTC (permalink / raw)
To: Mark Brown
Cc: linux-fbdev, Wolfram Sang, Linus Walleij, Guennadi Liakhovetski,
Thierry Reding, linux-mtd, Laurent Pinchart, Laurent Pinchart,
Vinod Koul, Joerg Roedel, linux-sh, Magnus Damm, Eduardo Valentin,
Tomi Valkeinen, linux-serial, linux-input, Zhang Rui, Chris Ball,
Jean-Christophe Plagniol-Villard, linux-media, linux-pwm,
Samuel Ortiz, linux-pm, Ian Molton
In-Reply-To: <20131029175834.GD20251@sirena.org.uk>
On Tue, Oct 29, 2013 at 10:58:34AM -0700, Mark Brown wrote:
> On Tue, Oct 29, 2013 at 06:29:59PM +0100, Laurent Pinchart wrote:
> > On Tuesday 29 October 2013 10:23:31 Mark Brown wrote:
> > > On Tue, Oct 29, 2013 at 06:05:53PM +0100, Laurent Pinchart wrote:
> > > > The first one is that I can't compile-test all those drivers on all
> > > > architectures. The spi-sh-msiof driver, for instance, uses
> > > > io(read|write)(16|
>
> > > Which architectures are these and is there not a symbol we can depend on
> > > for them?
>
> > arch/cris for instance. We can use readl/writel instead (maybe it would be
> > time to rationalize and document the I/O accessors across all architectures,
> > but that's another topic).
>
> It'd certainly be sensible, or adding a config option to depend on if
> you rely on these functions.
>
> > My point is that there might be other issues that I won't be able to easily
> > catch. This would break compilation for everybody for no reason, as the
> > drivers are useless on non-SuperH, non-ARM platforms. That's why I believe
> > COMPILE_TEST would be a better option as a first step.
>
> Yes, it would - please do that. Note that it won't stop anyone running
> into build issues on other architectures though, it's just about
> stopping Kconfig noise.
FWIW, I am happy with using COMPILE_TEST for this series.
^ permalink raw reply
* Re: [PATCH v2 net-next] net: introduce gro_frag_list_enable sysctl
From: Jerry Chu @ 2013-10-30 0:03 UTC (permalink / raw)
To: Eric Dumazet
Cc: Christoph Paasch, David Miller, Herbert Xu, netdev,
Michael Dalton
In-Reply-To: <1383059555.5464.33.camel@edumazet-glaptop.roam.corp.google.com>
On Tue, Oct 29, 2013 at 8:12 AM, Eric Dumazet <eric.dumazet@gmail.com> wrote:
> From: Eric Dumazet <edumazet@google.com>
>
> Christoph Paasch and Jerry Chu reported crashes in skb_segment() caused
> by commit 8a29111c7ca6 ("net: gro: allow to build full sized skb")
>
> (Jerry is working on adding native GRO support for tunnels)
>
> skb_segment() only deals with a frag_list chain containing MSS sized
> fragments.
>
> This patch adds support any kind of frag, and adds a new sysctl,
> as clearly the GRO layer should avoid building frag_list skbs
> on a router, as the segmentation is adding cpu overhead.
>
> Note that we could try to reuse page fragments instead of doing
> copy to linear skbs, but this requires a fair amount of work,
> and possible truesize nightmares, as we do not track individual
> (per page fragment) truesizes.
>
> /proc/sys/net/core/gro_frag_list_enable possible values are :
>
> 0 : GRO layer is not allowed to use frag_list to extend skb capacity
> 1 : GRO layer is allowed to use frag_list, but skb_segment()
> automatically sets the sysctl to 0.
> 2 : GRO is allowed to use frag_list, and skb_segment() wont
> clear the sysctl.
>
> Default value is 1 : automatic discovery
>
> Reported-by: Christoph Paasch <christoph.paasch@uclouvain.be>
> Reported-by: Jerry Chu <hkchu@google.com>
> Cc: Michael Dalton <mwdalton@google.com>
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> ---
> v2: added missing sysctl definition in skbuff.c
>
> Documentation/sysctl/net.txt | 19 +++++++++++++++++++
> include/linux/netdevice.h | 1 +
> net/core/skbuff.c | 31 ++++++++++++++++++++++---------
> net/core/sysctl_net_core.c | 10 ++++++++++
> 4 files changed, 52 insertions(+), 9 deletions(-)
>
> diff --git a/Documentation/sysctl/net.txt b/Documentation/sysctl/net.txt
> index 9a0319a82470..8778568ae64e 100644
> --- a/Documentation/sysctl/net.txt
> +++ b/Documentation/sysctl/net.txt
> @@ -87,6 +87,25 @@ sysctl.net.busy_read globally.
> Will increase power usage.
> Default: 0 (off)
>
> +gro_frag_list_enable
> +--------------------
> +
> +GRO layer can build full size GRO packets (~64K of payload) if it is allowed
> +to extend skb using the frag_list pointer. However, this strategy is a win
> +on hosts, where TCP flows are terminated. For a router, using frag_list
> +skbs is not a win because we have to segment skbs before transmit,
> +as most NIC drivers do not support frag_list.
> +As soon as one frag_list skb has to be segmented, this sysctl is automatically
> +changed from 1 to 0.
> +If the value is set to 2, kernel wont change it.
> +
> +Choices : 0 (off),
> + 1 (on, with automatic change to 0)
> + 2 (on, permanent)
> +
> +Default: 1 (on, with automatic downgrade on a router)
> +
> +
> rmem_default
> ------------
>
> diff --git a/include/linux/netdevice.h b/include/linux/netdevice.h
> index 27f62f746621..b82ff52f301e 100644
> --- a/include/linux/netdevice.h
> +++ b/include/linux/netdevice.h
> @@ -2807,6 +2807,7 @@ extern int netdev_max_backlog;
> extern int netdev_tstamp_prequeue;
> extern int weight_p;
> extern int bpf_jit_enable;
> +extern int sysctl_gro_frag_list_enable;
>
> bool netdev_has_upper_dev(struct net_device *dev, struct net_device *upper_dev);
> bool netdev_has_any_upper_dev(struct net_device *dev);
> diff --git a/net/core/skbuff.c b/net/core/skbuff.c
> index 0ab32faa520f..e089cd2782e5 100644
> --- a/net/core/skbuff.c
> +++ b/net/core/skbuff.c
> @@ -74,6 +74,8 @@
> struct kmem_cache *skbuff_head_cache __read_mostly;
> static struct kmem_cache *skbuff_fclone_cache __read_mostly;
>
> +int sysctl_gro_frag_list_enable __read_mostly = 1;
> +
> static void sock_pipe_buf_release(struct pipe_inode_info *pipe,
> struct pipe_buffer *buf)
> {
> @@ -2761,7 +2763,7 @@ struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features)
> unsigned int len;
> __be16 proto;
> bool csum;
> - int sg = !!(features & NETIF_F_SG);
> + bool sg = !!(features & NETIF_F_SG);
> int nfrags = skb_shinfo(skb)->nr_frags;
> int err = -ENOMEM;
> int i = 0;
> @@ -2793,7 +2795,13 @@ struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features)
> hsize = len;
>
> if (!hsize && i >= nfrags) {
> - BUG_ON(fskb->len != len);
> + if (fskb->len != len) {
> + if (sysctl_gro_frag_list_enable == 1)
> + sysctl_gro_frag_list_enable = 0;
> + hsize = len;
> + sg = false;
> + goto do_linear;
> + }
>
> pos += len;
> nskb = skb_clone(fskb, GFP_ATOMIC);
> @@ -2812,6 +2820,7 @@ struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features)
> skb_release_head_state(nskb);
> __skb_push(nskb, doffset);
> } else {
> +do_linear:
> nskb = __alloc_skb(hsize + doffset + headroom,
> GFP_ATOMIC, skb_alloc_rx_flag(skb),
> NUMA_NO_NODE);
> @@ -2838,9 +2847,6 @@ struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features)
> nskb->data - tnl_hlen,
> doffset + tnl_hlen);
>
> - if (fskb != skb_shinfo(skb)->frag_list)
> - goto perform_csum_check;
> -
> if (!sg) {
> nskb->ip_summed = CHECKSUM_NONE;
> nskb->csum = skb_copy_and_csum_bits(skb, offset,
> @@ -2849,6 +2855,9 @@ struct sk_buff *skb_segment(struct sk_buff *skb, netdev_features_t features)
> continue;
> }
>
> + if (fskb != skb_shinfo(skb)->frag_list)
> + goto perform_csum_check;
> +
> frag = skb_shinfo(nskb)->frags;
>
> skb_copy_from_linear_data_offset(skb, offset,
> @@ -2944,9 +2953,11 @@ int skb_gro_receive(struct sk_buff **head, struct sk_buff *skb)
> int i = skbinfo->nr_frags;
> int nr_frags = pinfo->nr_frags + i;
>
> - if (nr_frags > MAX_SKB_FRAGS)
> + if (unlikely(nr_frags > MAX_SKB_FRAGS)) {
> + if (!sysctl_gro_frag_list_enable)
> + return -E2BIG;
> goto merge;
> -
> + }
> offset -= headlen;
> pinfo->nr_frags = nr_frags;
> skbinfo->nr_frags = 0;
> @@ -2977,9 +2988,11 @@ int skb_gro_receive(struct sk_buff **head, struct sk_buff *skb)
> unsigned int first_size = headlen - offset;
> unsigned int first_offset;
>
> - if (nr_frags + 1 + skbinfo->nr_frags > MAX_SKB_FRAGS)
> + if (unlikely(nr_frags + 1 + skbinfo->nr_frags > MAX_SKB_FRAGS)) {
> + if (!sysctl_gro_frag_list_enable)
> + return -E2BIG;
> goto merge;
> -
> + }
> first_offset = skb->data -
> (unsigned char *)page_address(page) +
> offset;
> diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c
> index cca444190907..2d6aaf6d5838 100644
> --- a/net/core/sysctl_net_core.c
> +++ b/net/core/sysctl_net_core.c
> @@ -24,6 +24,7 @@
>
> static int zero = 0;
> static int one = 1;
> +static int two = 2;
> static int ushort_max = USHRT_MAX;
>
> #ifdef CONFIG_RPS
> @@ -360,6 +361,15 @@ static struct ctl_table net_core_table[] = {
> .mode = 0644,
> .proc_handler = proc_dointvec
> },
> + {
> + .procname = "gro_frag_list_enable",
> + .data = &sysctl_gro_frag_list_enable,
> + .maxlen = sizeof(int),
> + .mode = 0644,
> + .proc_handler = proc_dointvec_minmax,
> + .extra1 = &zero,
> + .extra2 = &two,
> + },
> { }
> };
>
>
>
I've tested all three values of sysctl_gro_frag_list_enable with my
yet-to-submitted GRE-GRO patch running on a router. The BUG panic is
gone and the patch seems to work as documented. Also setting the value
to 0 or 1 does seem to save a bit of CPU time on the forwarding path
and also help throughput a little (a combined of 5-8%).
Tested by: Jerry Chu <hkchu@google.com>
^ permalink raw reply
* [PATCH 2/2] ibm emac: Fix locking for enable/disable eob irq
From: Alistair Popple @ 2013-10-29 23:50 UTC (permalink / raw)
To: netdev, davem; +Cc: benh, Alistair Popple
In-Reply-To: <1383090638-21526-1-git-send-email-alistair@popple.id.au>
Calls to mal_enable_eob_irq perform a read-write-modify of a dcr to
enable device irqs which is protected by a spin lock. However calls to
mal_disable_eob_irq do not take the corresponding lock.
This patch resolves the problem by ensuring that calls to
mal_disable_eob_irq also take the lock.
Signed-off-by: Alistair Popple <alistair@popple.id.au>
---
drivers/net/ethernet/ibm/emac/mal.c | 4 ++++
1 file changed, 4 insertions(+)
diff --git a/drivers/net/ethernet/ibm/emac/mal.c b/drivers/net/ethernet/ibm/emac/mal.c
index dcb623a..2d2e1dd 100644
--- a/drivers/net/ethernet/ibm/emac/mal.c
+++ b/drivers/net/ethernet/ibm/emac/mal.c
@@ -263,7 +263,9 @@ static inline void mal_schedule_poll(struct mal_instance *mal)
{
if (likely(napi_schedule_prep(&mal->napi))) {
MAL_DBG2(mal, "schedule_poll" NL);
+ spin_lock(&mal->lock);
mal_disable_eob_irq(mal);
+ spin_unlock(&mal->lock);
__napi_schedule(&mal->napi);
} else
MAL_DBG2(mal, "already in poll" NL);
@@ -445,7 +447,9 @@ static int mal_poll(struct napi_struct *napi, int budget)
if (!napi_reschedule(napi))
goto more_work;
+ spin_lock_irqsave(&mal->lock, flags);
mal_disable_eob_irq(mal);
+ spin_unlock_irqrestore(&mal->lock, flags);
goto again;
}
mc->ops->poll_tx(mc->dev);
--
1.7.10.4
^ permalink raw reply related
* [PATCH 1/2] ibm emac: Don't call napi_complete if napi_reschedule failed
From: Alistair Popple @ 2013-10-29 23:50 UTC (permalink / raw)
To: netdev, davem; +Cc: benh, Alistair Popple
This patch fixes a bug which would trigger the BUG_ON() at
net/core/dev.c:4156. It was found that this was due to continuing
processing in the current poll call even when the call to
napi_reschedule failed, indicating the device was already on the
polling list. This resulted in an extra call to napi_complete which
triggered the BUG_ON().
This patch ensures that we only contine processing rotting packets in
the current mal_poll call if we are not already on the polling list.
Signed-off-by: Alistair Popple <alistair@popple.id.au>
---
drivers/net/ethernet/ibm/emac/mal.c | 12 ++++--------
1 file changed, 4 insertions(+), 8 deletions(-)
diff --git a/drivers/net/ethernet/ibm/emac/mal.c b/drivers/net/ethernet/ibm/emac/mal.c
index 856ea66..dcb623a 100644
--- a/drivers/net/ethernet/ibm/emac/mal.c
+++ b/drivers/net/ethernet/ibm/emac/mal.c
@@ -442,15 +442,11 @@ static int mal_poll(struct napi_struct *napi, int budget)
if (unlikely(mc->ops->peek_rx(mc->dev) ||
test_bit(MAL_COMMAC_RX_STOPPED, &mc->flags))) {
MAL_DBG2(mal, "rotting packet" NL);
- if (napi_reschedule(napi))
- mal_disable_eob_irq(mal);
- else
- MAL_DBG2(mal, "already in poll list" NL);
-
- if (budget > 0)
- goto again;
- else
+ if (!napi_reschedule(napi))
goto more_work;
+
+ mal_disable_eob_irq(mal);
+ goto again;
}
mc->ops->poll_tx(mc->dev);
}
--
1.7.10.4
^ permalink raw reply related
* Re: [PATCH] ibm emac: Don't call napi_complete if napi_reschedule failed
From: Alistair Popple @ 2013-10-29 23:53 UTC (permalink / raw)
To: David Miller; +Cc: netdev, benh
In-Reply-To: <20131025.191301.1496864269923138314.davem@davemloft.net>
On Fri, 25 Oct 2013 19:13:01 David Miller wrote:
> The mistaken napi_complete() call and the locking change are two different
> bugs, fix them in two different patches.
OK - I've resubmitted as two separate patches.
^ permalink raw reply
* Re: [PATCH v2 net-next] net: introduce gro_frag_list_enable sysctl
From: David Miller @ 2013-10-29 23:44 UTC (permalink / raw)
To: eric.dumazet; +Cc: christoph.paasch, herbert, netdev, hkchu, mwdalton
In-Reply-To: <1383059555.5464.33.camel@edumazet-glaptop.roam.corp.google.com>
From: Eric Dumazet <eric.dumazet@gmail.com>
Date: Tue, 29 Oct 2013 08:12:35 -0700
> From: Eric Dumazet <edumazet@google.com>
>
> Christoph Paasch and Jerry Chu reported crashes in skb_segment() caused
> by commit 8a29111c7ca6 ("net: gro: allow to build full sized skb")
>
> (Jerry is working on adding native GRO support for tunnels)
>
> skb_segment() only deals with a frag_list chain containing MSS sized
> fragments.
>
> This patch adds support any kind of frag, and adds a new sysctl,
> as clearly the GRO layer should avoid building frag_list skbs
> on a router, as the segmentation is adding cpu overhead.
>
> Note that we could try to reuse page fragments instead of doing
> copy to linear skbs, but this requires a fair amount of work,
> and possible truesize nightmares, as we do not track individual
> (per page fragment) truesizes.
>
> /proc/sys/net/core/gro_frag_list_enable possible values are :
>
> 0 : GRO layer is not allowed to use frag_list to extend skb capacity
> 1 : GRO layer is allowed to use frag_list, but skb_segment()
> automatically sets the sysctl to 0.
> 2 : GRO is allowed to use frag_list, and skb_segment() wont
> clear the sysctl.
>
> Default value is 1 : automatic discovery
>
> Reported-by: Christoph Paasch <christoph.paasch@uclouvain.be>
> Reported-by: Jerry Chu <hkchu@google.com>
> Cc: Michael Dalton <mwdalton@google.com>
> Signed-off-by: Eric Dumazet <edumazet@google.com>
> ---
> v2: added missing sysctl definition in skbuff.c
I do not like the idea of packet actions indirectly changing sysctl
values, even if you document it sufficiently as you have here.
Plus this puts the sysctl change logic in a fast path.
I would suggest instead making it change in response to changes to
ip_forward, as we do with per-device LRO settings. This means that,
like ip_forward, you should also make this sysctl a global + devinet
per-device sysctl.
You might even emit a pr_info() when this logic triggers, and if you
are ambitious enough keep track of the previous GRO sysctl state so
you can restore it if ip_forward is set back to zero.
Thanks.
^ permalink raw reply
page: next (older) | prev (newer) | latest
- recent:[subjects (threaded)|topics (new)|topics (active)]
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox