[deliverable/linux.git] / drivers / net / xen-netback / common.h

/*
 * This program is free software; you can redistribute it and/or
 * modify it under the terms of the GNU General Public License version 2
 * as published by the Free Software Foundation; or, when distributed
 * separately from the Linux kernel or incorporated into other
 * software packages, subject to the following license:
 *
 * Permission is hereby granted, free of charge, to any person obtaining a copy
 * of this source file (the "Software"), to deal in the Software without
 * restriction, including without limitation the rights to use, copy, modify,
 * merge, publish, distribute, sublicense, and/or sell copies of the Software,
 * and to permit persons to whom the Software is furnished to do so, subject to
 * the following conditions:
 *
 * The above copyright notice and this permission notice shall be included in
 * all copies or substantial portions of the Software.
 *
 * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
 * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
 * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
 * AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
 * LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
 * FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
 * IN THE SOFTWARE.
 */

#ifndef __XEN_NETBACK__COMMON_H__
#define __XEN_NETBACK__COMMON_H__

#define pr_fmt(fmt) KBUILD_MODNAME ":%s: " fmt, __func__

#include <linux/module.h>
#include <linux/interrupt.h>
#include <linux/slab.h>
#include <linux/ip.h>
#include <linux/in.h>
#include <linux/io.h>
#include <linux/netdevice.h>
#include <linux/etherdevice.h>
#include <linux/wait.h>
#include <linux/sched.h>

#include <xen/interface/io/netif.h>
#include <xen/interface/grant_table.h>
#include <xen/grant_table.h>
#include <xen/xenbus.h>

typedef unsigned int pending_ring_idx_t;
#define INVALID_PENDING_RING_IDX (~0U)

struct pending_tx_info {
	struct xen_netif_tx_request req; /* tx request */
	/* Callback data for released SKBs. The callback is always
	 * xenvif_zerocopy_callback, desc contains the pending_idx, which is
	 * also an index in pending_tx_info array. It is initialized in
	 * xenvif_alloc and it never changes.
	 * skb_shinfo(skb)->destructor_arg points to the first mapped slot's
	 * callback_struct in this array of struct pending_tx_info's, then ctx
	 * to the next, or NULL if there is no more slot for this skb.
	 * ubuf_to_vif is a helper which finds the struct xenvif from a pointer
	 * to this field.
	 */
	struct ubuf_info callback_struct;
};

#define XEN_NETIF_TX_RING_SIZE __CONST_RING_SIZE(xen_netif_tx, PAGE_SIZE)
#define XEN_NETIF_RX_RING_SIZE __CONST_RING_SIZE(xen_netif_rx, PAGE_SIZE)

struct xenvif_rx_meta {
	int id;
	int size;
	int gso_type;
	int gso_size;
};

#define GSO_BIT(type) \
	(1 << XEN_NETIF_GSO_TYPE_ ## type)

/* Discriminate from any valid pending_idx value. */
#define INVALID_PENDING_IDX 0xFFFF

#define MAX_BUFFER_OFFSET PAGE_SIZE

#define MAX_PENDING_REQS XEN_NETIF_TX_RING_SIZE

/* It's possible for an skb to have a maximal number of frags
 * but still be less than MAX_BUFFER_OFFSET in size. Thus the
 * worst-case number of copy operations is MAX_SKB_FRAGS per
 * ring slot.
 */
#define MAX_GRANT_COPY_OPS (MAX_SKB_FRAGS * XEN_NETIF_RX_RING_SIZE)

#define NETBACK_INVALID_HANDLE -1

/* To avoid confusion, we define XEN_NETBK_LEGACY_SLOTS_MAX indicating
 * the maximum slots a valid packet can use. Now this value is defined
 * to be XEN_NETIF_NR_SLOTS_MIN, which is supposed to be supported by
 * all backend.
 */
#define XEN_NETBK_LEGACY_SLOTS_MAX XEN_NETIF_NR_SLOTS_MIN

struct xenvif {
	/* Unique identifier for this interface. */
	domid_t          domid;
	unsigned int     handle;

	/* Is this interface disabled? True when backend discovers
	 * frontend is rogue.
	 */
	bool disabled;

	/* Use NAPI for guest TX */
	struct napi_struct napi;
	/* When feature-split-event-channels = 0, tx_irq = rx_irq. */
	unsigned int tx_irq;
	/* Only used when feature-split-event-channels = 1 */
	char tx_irq_name[IFNAMSIZ+4]; /* DEVNAME-tx */
	struct xen_netif_tx_back_ring tx;
	struct sk_buff_head tx_queue;
	struct page *mmap_pages[MAX_PENDING_REQS];
	pending_ring_idx_t pending_prod;
	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];

	struct gnttab_copy tx_copy_ops[MAX_PENDING_REQS];
	struct gnttab_map_grant_ref tx_map_ops[MAX_PENDING_REQS];
	struct gnttab_unmap_grant_ref tx_unmap_ops[MAX_PENDING_REQS];
	/* passed to gnttab_[un]map_refs with pages under (un)mapping */
	struct page *pages_to_map[MAX_PENDING_REQS];
	struct page *pages_to_unmap[MAX_PENDING_REQS];

	/* This prevents zerocopy callbacks  to race over dealloc_ring */
	spinlock_t callback_lock;
	/* This prevents dealloc thread and NAPI instance to race over response
	 * creation and pending_ring in xenvif_idx_release. In xenvif_tx_err
	 * it only protect response creation
	 */
	spinlock_t response_lock;
	pending_ring_idx_t dealloc_prod;
	pending_ring_idx_t dealloc_cons;
	u16 dealloc_ring[MAX_PENDING_REQS];
	struct task_struct *dealloc_task;
	wait_queue_head_t dealloc_wq;

	/* Use kthread for guest RX */
	struct task_struct *task;
	wait_queue_head_t wq;
	/* When feature-split-event-channels = 0, tx_irq = rx_irq. */
	unsigned int rx_irq;
	/* Only used when feature-split-event-channels = 1 */
	char rx_irq_name[IFNAMSIZ+4]; /* DEVNAME-rx */
	struct xen_netif_rx_back_ring rx;
	struct sk_buff_head rx_queue;
	RING_IDX rx_last_skb_slots;
	bool rx_queue_purge;

	struct timer_list wake_queue;

	/* This array is allocated seperately as it is large */
	struct gnttab_copy *grant_copy_op;

	/* We create one meta structure per ring request we consume, so
	 * the maximum number is the same as the ring size.
	 */
	struct xenvif_rx_meta meta[XEN_NETIF_RX_RING_SIZE];

	u8               fe_dev_addr[6];

	/* Frontend feature information. */
	int gso_mask;
	int gso_prefix_mask;

	u8 can_sg:1;
	u8 ip_csum:1;
	u8 ipv6_csum:1;

	/* Internal feature information. */
	u8 can_queue:1;	    /* can queue packets for receiver? */

	/* Transmit shaping: allow 'credit_bytes' every 'credit_usec'. */
	unsigned long   credit_bytes;
	unsigned long   credit_usec;
	unsigned long   remaining_credit;
	struct timer_list credit_timeout;
	u64 credit_window_start;

	/* Statistics */
	unsigned long rx_gso_checksum_fixup;
	unsigned long tx_zerocopy_sent;
	unsigned long tx_zerocopy_success;
	unsigned long tx_zerocopy_fail;
	unsigned long tx_frag_overflow;

	/* Miscellaneous private stuff. */
	struct net_device *dev;
};

static inline struct xenbus_device *xenvif_to_xenbus_device(struct xenvif *vif)
{
	return to_xenbus_device(vif->dev->dev.parent);
}

struct xenvif *xenvif_alloc(struct device *parent,
			    domid_t domid,
			    unsigned int handle);

int xenvif_connect(struct xenvif *vif, unsigned long tx_ring_ref,
		   unsigned long rx_ring_ref, unsigned int tx_evtchn,
		   unsigned int rx_evtchn);
void xenvif_disconnect(struct xenvif *vif);
void xenvif_free(struct xenvif *vif);

int xenvif_xenbus_init(void);
void xenvif_xenbus_fini(void);

int xenvif_schedulable(struct xenvif *vif);

int xenvif_must_stop_queue(struct xenvif *vif);

/* (Un)Map communication rings. */
void xenvif_unmap_frontend_rings(struct xenvif *vif);
int xenvif_map_frontend_rings(struct xenvif *vif,
			      grant_ref_t tx_ring_ref,
			      grant_ref_t rx_ring_ref);

/* Check for SKBs from frontend and schedule backend processing */
void xenvif_napi_schedule_or_enable_events(struct xenvif *vif);

/* Prevent the device from generating any further traffic. */
void xenvif_carrier_off(struct xenvif *vif);

int xenvif_tx_action(struct xenvif *vif, int budget);

int xenvif_kthread_guest_rx(void *data);
void xenvif_kick_thread(struct xenvif *vif);

int xenvif_dealloc_kthread(void *data);

/* Determine whether the needed number of slots (req) are available,
 * and set req_event if not.
 */
bool xenvif_rx_ring_slots_available(struct xenvif *vif, int needed);

void xenvif_stop_queue(struct xenvif *vif);

/* Callback from stack when TX packet can be released */
void xenvif_zerocopy_callback(struct ubuf_info *ubuf, bool zerocopy_success);

/* Unmap a pending page and release it back to the guest */
void xenvif_idx_unmap(struct xenvif *vif, u16 pending_idx);

static inline pending_ring_idx_t nr_pending_reqs(struct xenvif *vif)
{
	return MAX_PENDING_REQS -
		vif->pending_prod + vif->pending_cons;
}

/* Callback from stack when TX packet can be released */
void xenvif_zerocopy_callback(struct ubuf_info *ubuf, bool zerocopy_success);

extern bool separate_tx_rx_irq;

extern unsigned int rx_drain_timeout_msecs;
extern unsigned int rx_drain_timeout_jiffies;

#endif /* __XEN_NETBACK__COMMON_H__ */
Commit	Line	Data
f942dc25 IC	1	/*
	2	* This program is free software; you can redistribute it and/or
	3	* modify it under the terms of the GNU General Public License version 2
	4	* as published by the Free Software Foundation; or, when distributed
	5	* separately from the Linux kernel or incorporated into other
	6	* software packages, subject to the following license:
	7	*
	8	* Permission is hereby granted, free of charge, to any person obtaining a copy
	9	* of this source file (the "Software"), to deal in the Software without
	10	* restriction, including without limitation the rights to use, copy, modify,
	11	* merge, publish, distribute, sublicense, and/or sell copies of the Software,
	12	* and to permit persons to whom the Software is furnished to do so, subject to
	13	* the following conditions:
	14	*
	15	* The above copyright notice and this permission notice shall be included in
	16	* all copies or substantial portions of the Software.
	17	*
	18	* THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
	19	* IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
	20	* FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
	21	* AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
	22	* LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING
	23	* FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
	24	* IN THE SOFTWARE.
	25	*/
	26
	27	#ifndef __XEN_NETBACK__COMMON_H__
	28	#define __XEN_NETBACK__COMMON_H__
	29
	30	#define pr_fmt(fmt) KBUILD_MODNAME ":%s: " fmt, __func__
	31
	32	#include <linux/module.h>
	33	#include <linux/interrupt.h>
	34	#include <linux/slab.h>
	35	#include <linux/ip.h>
	36	#include <linux/in.h>
	37	#include <linux/io.h>
	38	#include <linux/netdevice.h>
	39	#include <linux/etherdevice.h>
	40	#include <linux/wait.h>
	41	#include <linux/sched.h>
	42
	43	#include <xen/interface/io/netif.h>
	44	#include <xen/interface/grant_table.h>
	45	#include <xen/grant_table.h>
	46	#include <xen/xenbus.h>
	47
b3f980bd WL	48	typedef unsigned int pending_ring_idx_t;
	49	#define INVALID_PENDING_RING_IDX (~0U)
	50
b3f980bd	51	struct pending_tx_info {
62bad319	52	struct xen_netif_tx_request req; /* tx request */
f53c3fe8 ZK	53	/* Callback data for released SKBs. The callback is always
	54	* xenvif_zerocopy_callback, desc contains the pending_idx, which is
	55	* also an index in pending_tx_info array. It is initialized in
	56	* xenvif_alloc and it never changes.
	57	* skb_shinfo(skb)->destructor_arg points to the first mapped slot's
	58	* callback_struct in this array of struct pending_tx_info's, then ctx
	59	* to the next, or NULL if there is no more slot for this skb.
	60	* ubuf_to_vif is a helper which finds the struct xenvif from a pointer
	61	* to this field.
	62	*/
	63	struct ubuf_info callback_struct;
b3f980bd WL	64	};
	65
	66	#define XEN_NETIF_TX_RING_SIZE __CONST_RING_SIZE(xen_netif_tx, PAGE_SIZE)
	67	#define XEN_NETIF_RX_RING_SIZE __CONST_RING_SIZE(xen_netif_rx, PAGE_SIZE)
	68
	69	struct xenvif_rx_meta {
	70	int id;
	71	int size;
82cada22	72	int gso_type;
b3f980bd WL	73	int gso_size;
	74	};
	75
82cada22 PD	76	#define GSO_BIT(type) \
	77	(1 << XEN_NETIF_GSO_TYPE_ ## type)
	78
b3f980bd WL	79	/* Discriminate from any valid pending_idx value. */
	80	#define INVALID_PENDING_IDX 0xFFFF
	81
	82	#define MAX_BUFFER_OFFSET PAGE_SIZE
	83
869b9b19	84	#define MAX_PENDING_REQS XEN_NETIF_TX_RING_SIZE
f942dc25	85
ac3d5ac2 PD	86	/* It's possible for an skb to have a maximal number of frags
	87	* but still be less than MAX_BUFFER_OFFSET in size. Thus the
	88	* worst-case number of copy operations is MAX_SKB_FRAGS per
	89	* ring slot.
	90	*/
	91	#define MAX_GRANT_COPY_OPS (MAX_SKB_FRAGS * XEN_NETIF_RX_RING_SIZE)
	92
121fa4b7 ZK	93	#define NETBACK_INVALID_HANDLE -1
	94
	95	/* To avoid confusion, we define XEN_NETBK_LEGACY_SLOTS_MAX indicating
	96	* the maximum slots a valid packet can use. Now this value is defined
	97	* to be XEN_NETIF_NR_SLOTS_MIN, which is supposed to be supported by
	98	* all backend.
	99	*/
	100	#define XEN_NETBK_LEGACY_SLOTS_MAX XEN_NETIF_NR_SLOTS_MIN
	101
f942dc25 IC	102	struct xenvif {
	103	/* Unique identifier for this interface. */
	104	domid_t domid;
	105	unsigned int handle;
	106
e9d8b2c2 WL	107	/* Is this interface disabled? True when backend discovers
	108	* frontend is rogue.
	109	*/
	110	bool disabled;
	111
b3f980bd WL	112	/* Use NAPI for guest TX */
	113	struct napi_struct napi;
	114	/* When feature-split-event-channels = 0, tx_irq = rx_irq. */
	115	unsigned int tx_irq;
	116	/* Only used when feature-split-event-channels = 1 */
	117	char tx_irq_name[IFNAMSIZ+4]; /* DEVNAME-tx */
	118	struct xen_netif_tx_back_ring tx;
	119	struct sk_buff_head tx_queue;
	120	struct page *mmap_pages[MAX_PENDING_REQS];
	121	pending_ring_idx_t pending_prod;
	122	pending_ring_idx_t pending_cons;
	123	u16 pending_ring[MAX_PENDING_REQS];
	124	struct pending_tx_info pending_tx_info[MAX_PENDING_REQS];
f53c3fe8	125	grant_handle_t grant_tx_handle[MAX_PENDING_REQS];
b3f980bd	126
bdab8275	127	struct gnttab_copy tx_copy_ops[MAX_PENDING_REQS];
f53c3fe8 ZK	128	struct gnttab_map_grant_ref tx_map_ops[MAX_PENDING_REQS];
	129	struct gnttab_unmap_grant_ref tx_unmap_ops[MAX_PENDING_REQS];
	130	/* passed to gnttab_[un]map_refs with pages under (un)mapping */
	131	struct page *pages_to_map[MAX_PENDING_REQS];
	132	struct page *pages_to_unmap[MAX_PENDING_REQS];
	133
	134	/* This prevents zerocopy callbacks to race over dealloc_ring */
	135	spinlock_t callback_lock;
	136	/* This prevents dealloc thread and NAPI instance to race over response
	137	* creation and pending_ring in xenvif_idx_release. In xenvif_tx_err
	138	* it only protect response creation
	139	*/
	140	spinlock_t response_lock;
	141	pending_ring_idx_t dealloc_prod;
	142	pending_ring_idx_t dealloc_cons;
	143	u16 dealloc_ring[MAX_PENDING_REQS];
	144	struct task_struct *dealloc_task;
	145	wait_queue_head_t dealloc_wq;
f942dc25	146
b3f980bd WL	147	/* Use kthread for guest RX */
	148	struct task_struct *task;
	149	wait_queue_head_t wq;
e1f00a69	150	/* When feature-split-event-channels = 0, tx_irq = rx_irq. */
e1f00a69 WL	151	unsigned int rx_irq;
e1f00a69 WL	152	/* Only used when feature-split-event-channels = 1 */
e1f00a69	153	char rx_irq_name[IFNAMSIZ+4]; /* DEVNAME-rx */
b3f980bd WL	154	struct xen_netif_rx_back_ring rx;
b3f980bd WL	155	struct sk_buff_head rx_queue;
9ab9831b	156	RING_IDX rx_last_skb_slots;
09350788 ZK	157	bool rx_queue_purge;
	158
	159	struct timer_list wake_queue;
b3f980bd	160
ac3d5ac2 PD	161	/* This array is allocated seperately as it is large */
ac3d5ac2 PD	162	struct gnttab_copy *grant_copy_op;
f942dc25	163
ac3d5ac2 PD	164	/* We create one meta structure per ring request we consume, so
	165	* the maximum number is the same as the ring size.
	166	*/
	167	struct xenvif_rx_meta meta[XEN_NETIF_RX_RING_SIZE];
b3f980bd WL	168
b3f980bd WL	169	u8 fe_dev_addr[6];
f942dc25	170
f942dc25	171	/* Frontend feature information. */
82cada22 PD	172	int gso_mask;
	173	int gso_prefix_mask;
	174
f942dc25	175	u8 can_sg:1;
146c8a77 PD	176	u8 ip_csum:1;
146c8a77 PD	177	u8 ipv6_csum:1;
f942dc25 IC	178
	179	/* Internal feature information. */
	180	u8 can_queue:1; /* can queue packets for receiver? */
	181
f942dc25 IC	182	/* Transmit shaping: allow 'credit_bytes' every 'credit_usec'. */
	183	unsigned long credit_bytes;
	184	unsigned long credit_usec;
	185	unsigned long remaining_credit;
	186	struct timer_list credit_timeout;
059dfa6a	187	u64 credit_window_start;
f942dc25 IC	188
	189	/* Statistics */
	190	unsigned long rx_gso_checksum_fixup;
1bb332af ZK	191	unsigned long tx_zerocopy_sent;
	192	unsigned long tx_zerocopy_success;
	193	unsigned long tx_zerocopy_fail;
e3377f36	194	unsigned long tx_frag_overflow;
f942dc25 IC	195
f942dc25 IC	196	/* Miscellaneous private stuff. */
f942dc25	197	struct net_device *dev;
f942dc25 IC	198	};
f942dc25 IC	199
c9d63699 DV	200	static inline struct xenbus_device xenvif_to_xenbus_device(struct xenvif vif)
	201	{
	202	return to_xenbus_device(vif->dev->dev.parent);
	203	}
	204
f942dc25 IC	205	struct xenvif xenvif_alloc(struct device parent,
	206	domid_t domid,
	207	unsigned int handle);
	208
	209	int xenvif_connect(struct xenvif *vif, unsigned long tx_ring_ref,
e1f00a69 WL	210	unsigned long rx_ring_ref, unsigned int tx_evtchn,
e1f00a69 WL	211	unsigned int rx_evtchn);
f942dc25	212	void xenvif_disconnect(struct xenvif *vif);
279f438e	213	void xenvif_free(struct xenvif *vif);
f942dc25	214
f942dc25	215	int xenvif_xenbus_init(void);
b103f358	216	void xenvif_xenbus_fini(void);
f942dc25 IC	217
	218	int xenvif_schedulable(struct xenvif *vif);
	219
7376419a	220	int xenvif_must_stop_queue(struct xenvif *vif);
f942dc25 IC	221
f942dc25 IC	222	/* (Un)Map communication rings. */
7376419a WL	223	void xenvif_unmap_frontend_rings(struct xenvif *vif);
	224	int xenvif_map_frontend_rings(struct xenvif *vif,
	225	grant_ref_t tx_ring_ref,
	226	grant_ref_t rx_ring_ref);
f942dc25	227
f942dc25	228	/* Check for SKBs from frontend and schedule backend processing */
0d08fceb	229	void xenvif_napi_schedule_or_enable_events(struct xenvif *vif);
f942dc25	230
48856286 IC	231	/* Prevent the device from generating any further traffic. */
	232	void xenvif_carrier_off(struct xenvif *vif);
	233
7376419a	234	int xenvif_tx_action(struct xenvif *vif, int budget);
b3f980bd	235
121fa4b7	236	int xenvif_kthread_guest_rx(void *data);
ca2f09f2 PD	237	void xenvif_kick_thread(struct xenvif *vif);
ca2f09f2 PD	238
f53c3fe8 ZK	239	int xenvif_dealloc_kthread(void *data);
f53c3fe8 ZK	240
ca2f09f2 PD	241	/* Determine whether the needed number of slots (req) are available,
	242	* and set req_event if not.
	243	*/
	244	bool xenvif_rx_ring_slots_available(struct xenvif *vif, int needed);
	245
	246	void xenvif_stop_queue(struct xenvif *vif);
b3f980bd	247
f53c3fe8 ZK	248	/* Callback from stack when TX packet can be released */
	249	void xenvif_zerocopy_callback(struct ubuf_info *ubuf, bool zerocopy_success);
	250
	251	/* Unmap a pending page and release it back to the guest */
	252	void xenvif_idx_unmap(struct xenvif *vif, u16 pending_idx);
	253
121fa4b7 ZK	254	static inline pending_ring_idx_t nr_pending_reqs(struct xenvif *vif)
	255	{
	256	return MAX_PENDING_REQS -
	257	vif->pending_prod + vif->pending_cons;
	258	}
	259
3e2234b3 ZK	260	/* Callback from stack when TX packet can be released */
	261	void xenvif_zerocopy_callback(struct ubuf_info *ubuf, bool zerocopy_success);
	262
e1f00a69 WL	263	extern bool separate_tx_rx_irq;
e1f00a69 WL	264
09350788 ZK	265	extern unsigned int rx_drain_timeout_msecs;
	266	extern unsigned int rx_drain_timeout_jiffies;
	267
f942dc25	268	#endif /* __XEN_NETBACK__COMMON_H__ */