xpc.h 40.6 KB
Newer Older
1 2 3 4 5
/*
 * This file is subject to the terms and conditions of the GNU General Public
 * License.  See the file "COPYING" in the main directory of this archive
 * for more details.
 *
6
 * Copyright (c) 2004-2008 Silicon Graphics, Inc.  All Rights Reserved.
7 8 9 10 11 12
 */

/*
 * Cross Partition Communication (XPC) structures and macros.
 */

13 14
#ifndef _DRIVERS_MISC_SGIXP_XPC_H
#define _DRIVERS_MISC_SGIXP_XPC_H
15 16 17 18

#include <linux/interrupt.h>
#include <linux/sysctl.h>
#include <linux/device.h>
J
Jes Sorensen 已提交
19 20
#include <linux/mutex.h>
#include <linux/completion.h>
21 22 23 24 25 26 27
#include <asm/pgtable.h>
#include <asm/processor.h>
#include <asm/sn/bte.h>
#include <asm/sn/clksupport.h>
#include <asm/sn/addrs.h>
#include <asm/sn/mspec.h>
#include <asm/sn/shub_mmr.h>
28
#include "xp.h"
29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56

/*
 * XPC Version numbers consist of a major and minor number. XPC can always
 * talk to versions with same major #, and never talk to versions with a
 * different major #.
 */
#define _XPC_VERSION(_maj, _min)	(((_maj) << 4) | ((_min) & 0xf))
#define XPC_VERSION_MAJOR(_v)		((_v) >> 4)
#define XPC_VERSION_MINOR(_v)		((_v) & 0xf)

/*
 * The next macros define word or bit representations for given
 * C-brick nasid in either the SAL provided bit array representing
 * nasids in the partition/machine or the AMO_t array used for
 * inter-partition initiation communications.
 *
 * For SN2 machines, C-Bricks are alway even numbered NASIDs.  As
 * such, some space will be saved by insisting that nasid information
 * passed from SAL always be packed for C-Bricks and the
 * cross-partition interrupts use the same packing scheme.
 */
#define XPC_NASID_W_INDEX(_n)	(((_n) / 64) / 2)
#define XPC_NASID_B_INDEX(_n)	(((_n) / 2) & (64 - 1))
#define XPC_NASID_IN_ARRAY(_n, _p) ((_p)[XPC_NASID_W_INDEX(_n)] & \
				    (1UL << XPC_NASID_B_INDEX(_n)))
#define XPC_NASID_FROM_W_B(_w, _b) (((_w) * 64 + (_b)) * 2)

#define XPC_HB_DEFAULT_INTERVAL		5	/* incr HB every x secs */
57
#define XPC_HB_CHECK_DEFAULT_INTERVAL	20	/* check HB every x secs */
58 59 60 61 62 63 64 65 66

/* define the process name of HB checker and the CPU it is pinned to */
#define XPC_HB_CHECK_THREAD_NAME	"xpc_hb"
#define XPC_HB_CHECK_CPU		0

/* define the process name of the discovery thread */
#define XPC_DISCOVERY_THREAD_NAME	"xpc_discovery"

/*
67
 * the reserved page
68
 *
69 70 71 72 73 74 75 76 77 78 79 80
 *   SAL reserves one page of memory per partition for XPC. Though a full page
 *   in length (16384 bytes), its starting address is not page aligned, but it
 *   is cacheline aligned. The reserved page consists of the following:
 *
 *   reserved page header
 *
 *     The first cacheline of the reserved page contains the header
 *     (struct xpc_rsvd_page). Before SAL initialization has completed,
 *     SAL has set up the following fields of the reserved page header:
 *     SAL_signature, SAL_version, partid, and nasids_size. The other
 *     fields are set up by XPC. (xpc_rsvd_page points to the local
 *     partition's reserved page.)
81
 *
82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103
 *   part_nasids mask
 *   mach_nasids mask
 *
 *     SAL also sets up two bitmaps (or masks), one that reflects the actual
 *     nasids in this partition (part_nasids), and the other that reflects
 *     the actual nasids in the entire machine (mach_nasids). We're only
 *     interested in the even numbered nasids (which contain the processors
 *     and/or memory), so we only need half as many bits to represent the
 *     nasids. The part_nasids mask is located starting at the first cacheline
 *     following the reserved page header. The mach_nasids mask follows right
 *     after the part_nasids mask. The size in bytes of each mask is reflected
 *     by the reserved page header field 'nasids_size'. (Local partition's
 *     mask pointers are xpc_part_nasids and xpc_mach_nasids.)
 *
 *   vars
 *   vars part
 *
 *     Immediately following the mach_nasids mask are the XPC variables
 *     required by other partitions. First are those that are generic to all
 *     partitions (vars), followed on the next available cacheline by those
 *     which are partition specific (vars part). These are setup by XPC.
 *     (Local partition's vars pointers are xpc_vars and xpc_vars_part.)
104 105 106 107
 *
 * Note: Until vars_pa is set, the partition XPC code has not been initialized.
 */
struct xpc_rsvd_page {
108 109 110
	u64 SAL_signature;	/* SAL: unique signature */
	u64 SAL_version;	/* SAL: version */
	u8 partid;		/* SAL: partition ID */
111
	u8 version;
112
	u8 pad1[6];		/* align to next u64 in cacheline */
113
	u64 vars_pa;		/* physical address of struct xpc_vars */
114 115 116
	struct timespec stamp;	/* time when reserved page was setup by XPC */
	u64 pad2[9];		/* align to last u64 in cacheline */
	u64 nasids_size;	/* SAL: size of each nasid mask in bytes */
117 118
};

119
#define XPC_RP_VERSION _XPC_VERSION(1, 1) /* version 1.1 of the reserved page */
120 121

#define XPC_SUPPORTS_RP_STAMP(_version) \
122
			(_version >= _XPC_VERSION(1, 1))
123 124 125 126 127 128 129 130 131 132 133 134 135

/*
 * compare stamps - the return value is:
 *
 *	< 0,	if stamp1 < stamp2
 *	= 0,	if stamp1 == stamp2
 *	> 0,	if stamp1 > stamp2
 */
static inline int
xpc_compare_stamps(struct timespec *stamp1, struct timespec *stamp2)
{
	int ret;

136 137
	ret = stamp1->tv_sec - stamp2->tv_sec;
	if (ret == 0)
138
		ret = stamp1->tv_nsec - stamp2->tv_nsec;
139

140 141
	return ret;
}
142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160

/*
 * Define the structures by which XPC variables can be exported to other
 * partitions. (There are two: struct xpc_vars and struct xpc_vars_part)
 */

/*
 * The following structure describes the partition generic variables
 * needed by other partitions in order to properly initialize.
 *
 * struct xpc_vars version number also applies to struct xpc_vars_part.
 * Changes to either structure and/or related functionality should be
 * reflected by incrementing either the major or minor version numbers
 * of struct xpc_vars.
 */
struct xpc_vars {
	u8 version;
	u64 heartbeat;
	u64 heartbeating_to_mask;
161
	u64 heartbeat_offline;	/* if 0, heartbeat should be changing */
162 163 164 165 166 167 168
	int act_nasid;
	int act_phys_cpuid;
	u64 vars_part_pa;
	u64 amos_page_pa;	/* paddr of page of AMOs from MSPEC driver */
	AMO_t *amos_page;	/* vaddr of page of AMOs from MSPEC driver */
};

169
#define XPC_V_VERSION _XPC_VERSION(3, 1)    /* version 3.1 of the cross vars */
170 171

#define XPC_SUPPORTS_DISENGAGE_REQUEST(_version) \
172
			(_version >= _XPC_VERSION(3, 1))
173 174

static inline int
175
xpc_hb_allowed(short partid, struct xpc_vars *vars)
176 177 178 179 180
{
	return ((vars->heartbeating_to_mask & (1UL << partid)) != 0);
}

static inline void
181
xpc_allow_hb(short partid, struct xpc_vars *vars)
182 183 184 185 186 187 188
{
	u64 old_mask, new_mask;

	do {
		old_mask = vars->heartbeating_to_mask;
		new_mask = (old_mask | (1UL << partid));
	} while (cmpxchg(&vars->heartbeating_to_mask, old_mask, new_mask) !=
189
		 old_mask);
190 191 192
}

static inline void
193
xpc_disallow_hb(short partid, struct xpc_vars *vars)
194 195 196 197 198 199 200
{
	u64 old_mask, new_mask;

	do {
		old_mask = vars->heartbeating_to_mask;
		new_mask = (old_mask & ~(1UL << partid));
	} while (cmpxchg(&vars->heartbeating_to_mask, old_mask, new_mask) !=
201
		 old_mask);
202 203 204 205 206
}

/*
 * The AMOs page consists of a number of AMO variables which are divided into
 * four groups, The first two groups are used to identify an IRQ's sender.
207
 * These two groups consist of 64 and 128 AMO variables respectively. The last
208 209 210 211 212
 * two groups, consisting of just one AMO variable each, are used to identify
 * the remote partitions that are currently engaged (from the viewpoint of
 * the XPC running on the remote partition).
 */
#define XPC_NOTIFY_IRQ_AMOS	   0
213
#define XPC_ACTIVATE_IRQ_AMOS	   (XPC_NOTIFY_IRQ_AMOS + XP_MAX_NPARTITIONS_SN2)
214 215 216
#define XPC_ENGAGED_PARTITIONS_AMO (XPC_ACTIVATE_IRQ_AMOS + XP_NASID_MASK_WORDS)
#define XPC_DISENGAGE_REQUEST_AMO  (XPC_ENGAGED_PARTITIONS_AMO + 1)

217 218 219 220 221 222 223 224 225 226 227
/*
 * The following structure describes the per partition specific variables.
 *
 * An array of these structures, one per partition, will be defined. As a
 * partition becomes active XPC will copy the array entry corresponding to
 * itself from that partition. It is desirable that the size of this
 * structure evenly divide into a cacheline, such that none of the entries
 * in this array crosses a cacheline boundary. As it is now, each entry
 * occupies half a cacheline.
 */
struct xpc_vars_part {
228
	u64 magic;
229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250

	u64 openclose_args_pa;	/* physical address of open and close args */
	u64 GPs_pa;		/* physical address of Get/Put values */

	u64 IPI_amo_pa;		/* physical address of IPI AMO_t structure */
	int IPI_nasid;		/* nasid of where to send IPIs */
	int IPI_phys_cpuid;	/* physical CPU ID of where to send IPIs */

	u8 nchannels;		/* #of defined channels supported */

	u8 reserved[23];	/* pad to a full 64 bytes */
};

/*
 * The vars_part MAGIC numbers play a part in the first contact protocol.
 *
 * MAGIC1 indicates that the per partition specific variables for a remote
 * partition have been initialized by this partition.
 *
 * MAGIC2 indicates that this partition has pulled the remote partititions
 * per partition variables that pertain to this partition.
 */
251 252
#define XPC_VP_MAGIC1	0x0053524156435058L   /* 'XPCVARS\0'L (little endian) */
#define XPC_VP_MAGIC2	0x0073726176435058L   /* 'XPCvars\0'L (little endian) */
253

254 255 256
/* the reserved page sizes and offsets */

#define XPC_RP_HEADER_SIZE	L1_CACHE_ALIGN(sizeof(struct xpc_rsvd_page))
257
#define XPC_RP_VARS_SIZE	L1_CACHE_ALIGN(sizeof(struct xpc_vars))
258

259
#define XPC_RP_PART_NASIDS(_rp) ((u64 *)((u8 *)(_rp) + XPC_RP_HEADER_SIZE))
260
#define XPC_RP_MACH_NASIDS(_rp) (XPC_RP_PART_NASIDS(_rp) + xp_nasid_mask_words)
261 262 263 264
#define XPC_RP_VARS(_rp)	((struct xpc_vars *)(XPC_RP_MACH_NASIDS(_rp) + \
				    xp_nasid_mask_words))
#define XPC_RP_VARS_PART(_rp)	((struct xpc_vars_part *) \
				    ((u8 *)XPC_RP_VARS(_rp) + XPC_RP_VARS_SIZE))
265

266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282
/*
 * Functions registered by add_timer() or called by kernel_thread() only
 * allow for a single 64-bit argument. The following macros can be used to
 * pack and unpack two (32-bit, 16-bit or 8-bit) arguments into or out from
 * the passed argument.
 */
#define XPC_PACK_ARGS(_arg1, _arg2) \
			((((u64) _arg1) & 0xffffffff) | \
			((((u64) _arg2) & 0xffffffff) << 32))

#define XPC_UNPACK_ARG1(_args)	(((u64) _args) & 0xffffffff)
#define XPC_UNPACK_ARG2(_args)	((((u64) _args) >> 32) & 0xffffffff)

/*
 * Define a Get/Put value pair (pointers) used with a message queue.
 */
struct xpc_gp {
283 284
	s64 get;		/* Get value */
	s64 put;		/* Put value */
285 286 287
};

#define XPC_GP_SIZE \
288
		L1_CACHE_ALIGN(sizeof(struct xpc_gp) * XPC_MAX_NCHANNELS)
289 290 291 292 293 294 295 296 297 298 299 300 301 302

/*
 * Define a structure that contains arguments associated with opening and
 * closing a channel.
 */
struct xpc_openclose_args {
	u16 reason;		/* reason why channel is closing */
	u16 msg_size;		/* sizeof each message entry */
	u16 remote_nentries;	/* #of message entries in remote msg queue */
	u16 local_nentries;	/* #of message entries in local msg queue */
	u64 local_msgqueue_pa;	/* physical address of local message queue */
};

#define XPC_OPENCLOSE_ARGS_SIZE \
303 304
	      L1_CACHE_ALIGN(sizeof(struct xpc_openclose_args) * \
	      XPC_MAX_NCHANNELS)
305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321

/* struct xpc_msg flags */

#define	XPC_M_DONE		0x01	/* msg has been received/consumed */
#define	XPC_M_READY		0x02	/* msg is ready to be sent */
#define	XPC_M_INTERRUPT		0x04	/* send interrupt when msg consumed */

#define XPC_MSG_ADDRESS(_payload) \
		((struct xpc_msg *)((u8 *)(_payload) - XPC_MSG_PAYLOAD_OFFSET))

/*
 * Defines notify entry.
 *
 * This is used to notify a message's sender that their message was received
 * and consumed by the intended recipient.
 */
struct xpc_notify {
322
	u8 type;		/* type of notification */
323 324

	/* the following two fields are only used if type == XPC_N_CALL */
325 326
	xpc_notify_func func;	/* user's notify function */
	void *key;		/* pointer to user's key */
327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411
};

/* struct xpc_notify type of notification */

#define	XPC_N_CALL		0x01	/* notify function provided by user */

/*
 * Define the structure that manages all the stuff required by a channel. In
 * particular, they are used to manage the messages sent across the channel.
 *
 * This structure is private to a partition, and is NOT shared across the
 * partition boundary.
 *
 * There is an array of these structures for each remote partition. It is
 * allocated at the time a partition becomes active. The array contains one
 * of these structures for each potential channel connection to that partition.
 *
 * Each of these structures manages two message queues (circular buffers).
 * They are allocated at the time a channel connection is made. One of
 * these message queues (local_msgqueue) holds the locally created messages
 * that are destined for the remote partition. The other of these message
 * queues (remote_msgqueue) is a locally cached copy of the remote partition's
 * own local_msgqueue.
 *
 * The following is a description of the Get/Put pointers used to manage these
 * two message queues. Consider the local_msgqueue to be on one partition
 * and the remote_msgqueue to be its cached copy on another partition. A
 * description of what each of the lettered areas contains is included.
 *
 *
 *                     local_msgqueue      remote_msgqueue
 *
 *                        |/////////|      |/////////|
 *    w_remote_GP.get --> +---------+      |/////////|
 *                        |    F    |      |/////////|
 *     remote_GP.get  --> +---------+      +---------+ <-- local_GP->get
 *                        |         |      |         |
 *                        |         |      |    E    |
 *                        |         |      |         |
 *                        |         |      +---------+ <-- w_local_GP.get
 *                        |    B    |      |/////////|
 *                        |         |      |////D////|
 *                        |         |      |/////////|
 *                        |         |      +---------+ <-- w_remote_GP.put
 *                        |         |      |////C////|
 *      local_GP->put --> +---------+      +---------+ <-- remote_GP.put
 *                        |         |      |/////////|
 *                        |    A    |      |/////////|
 *                        |         |      |/////////|
 *     w_local_GP.put --> +---------+      |/////////|
 *                        |/////////|      |/////////|
 *
 *
 *	    ( remote_GP.[get|put] are cached copies of the remote
 *	      partition's local_GP->[get|put], and thus their values can
 *	      lag behind their counterparts on the remote partition. )
 *
 *
 *  A - Messages that have been allocated, but have not yet been sent to the
 *	remote partition.
 *
 *  B - Messages that have been sent, but have not yet been acknowledged by the
 *      remote partition as having been received.
 *
 *  C - Area that needs to be prepared for the copying of sent messages, by
 *	the clearing of the message flags of any previously received messages.
 *
 *  D - Area into which sent messages are to be copied from the remote
 *	partition's local_msgqueue and then delivered to their intended
 *	recipients. [ To allow for a multi-message copy, another pointer
 *	(next_msg_to_pull) has been added to keep track of the next message
 *	number needing to be copied (pulled). It chases after w_remote_GP.put.
 *	Any messages lying between w_local_GP.get and next_msg_to_pull have
 *	been copied and are ready to be delivered. ]
 *
 *  E - Messages that have been copied and delivered, but have not yet been
 *	acknowledged by the recipient as having been received.
 *
 *  F - Messages that have been acknowledged, but XPC has not yet notified the
 *	sender that the message was received by its intended recipient.
 *	This is also an area that needs to be prepared for the allocating of
 *	new messages, by the clearing of the message flags of the acknowledged
 *	messages.
 */
struct xpc_channel {
412
	short partid;		/* ID of remote partition connected */
413 414
	spinlock_t lock;	/* lock for updating this structure */
	u32 flags;		/* general flags */
415

416
	enum xp_retval reason;	/* reason why channel is disconnect'g */
417
	int reason_line;	/* line# disconnect initiated from */
418

419
	u16 number;		/* channel # */
420

421 422 423
	u16 msg_size;		/* sizeof each msg entry */
	u16 local_nentries;	/* #of msg entries in local msg queue */
	u16 remote_nentries;	/* #of msg entries in remote msg queue */
424 425 426 427

	void *local_msgqueue_base;	/* base address of kmalloc'd space */
	struct xpc_msg *local_msgqueue;	/* local message queue */
	void *remote_msgqueue_base;	/* base address of kmalloc'd space */
428 429
	struct xpc_msg *remote_msgqueue; /* cached copy of remote partition's */
					 /* local message queue */
430 431
	u64 remote_msgqueue_pa;	/* phys addr of remote partition's */
				/* local message queue */
432

433
	atomic_t references;	/* #of external references to queues */
434

435 436
	atomic_t n_on_msg_allocate_wq;	/* #on msg allocation wait queue */
	wait_queue_head_t msg_allocate_wq;	/* msg allocation wait queue */
437

438 439
	u8 delayed_IPI_flags;	/* IPI flags received, but delayed */
				/* action until channel disconnected */
440

441 442
	/* queue of msg senders who want to be notified when msg received */

443
	atomic_t n_to_notify;	/* #of msg senders to notify */
444
	struct xpc_notify *notify_queue;    /* notify queue for messages sent */
445

446 447
	xpc_channel_func func;	/* user's channel function */
	void *key;		/* pointer to user's key */
448

J
Jes Sorensen 已提交
449
	struct mutex msg_to_pull_mutex;	/* next msg to pull serialization */
450
	struct completion wdisconnect_wait;    /* wait for channel disconnect */
451

452 453
	struct xpc_openclose_args *local_openclose_args; /* args passed on */
					     /* opening or closing of channel */
454 455 456 457 458 459 460

	/* various flavors of local and remote Get/Put values */

	struct xpc_gp *local_GP;	/* local Get/Put values */
	struct xpc_gp remote_GP;	/* remote Get/Put values */
	struct xpc_gp w_local_GP;	/* working local Get/Put values */
	struct xpc_gp w_remote_GP;	/* working remote Get/Put values */
461
	s64 next_msg_to_pull;	/* Put value of next msg to pull */
462 463 464 465

	/* kthread management related fields */

	atomic_t kthreads_assigned;	/* #of kthreads assigned to channel */
466 467
	u32 kthreads_assigned_limit;	/* limit on #of kthreads assigned */
	atomic_t kthreads_idle;	/* #of kthreads idle waiting for work */
468 469 470 471 472 473 474 475 476
	u32 kthreads_idle_limit;	/* limit on #of kthreads idle */
	atomic_t kthreads_active;	/* #of kthreads actively working */

	wait_queue_head_t idle_wq;	/* idle kthread wait queue */

} ____cacheline_aligned;

/* struct xpc_channel flags */

477
#define	XPC_C_WASCONNECTED	0x00000001	/* channel was connected */
478

479 480
#define	XPC_C_ROPENREPLY	0x00000002	/* remote open channel reply */
#define	XPC_C_OPENREPLY		0x00000004	/* local open channel reply */
481
#define	XPC_C_ROPENREQUEST	0x00000008     /* remote open channel request */
482
#define	XPC_C_OPENREQUEST	0x00000010	/* local open channel request */
483

484 485
#define	XPC_C_SETUP		0x00000020 /* channel's msgqueues are alloc'd */
#define	XPC_C_CONNECTEDCALLOUT	0x00000040     /* connected callout initiated */
486
#define	XPC_C_CONNECTEDCALLOUT_MADE \
487
				0x00000080     /* connected callout completed */
488 489
#define	XPC_C_CONNECTED		0x00000100	/* local channel is connected */
#define	XPC_C_CONNECTING	0x00000200	/* channel is being connected */
490

491 492
#define	XPC_C_RCLOSEREPLY	0x00000400	/* remote close channel reply */
#define	XPC_C_CLOSEREPLY	0x00000800	/* local close channel reply */
493 494
#define	XPC_C_RCLOSEREQUEST	0x00001000    /* remote close channel request */
#define	XPC_C_CLOSEREQUEST	0x00002000     /* local close channel request */
495

496
#define	XPC_C_DISCONNECTED	0x00004000	/* channel is disconnected */
497
#define	XPC_C_DISCONNECTING	0x00008000   /* channel is being disconnected */
498
#define	XPC_C_DISCONNECTINGCALLOUT \
499
				0x00010000 /* disconnecting callout initiated */
500
#define	XPC_C_DISCONNECTINGCALLOUT_MADE \
501 502
				0x00020000 /* disconnecting callout completed */
#define	XPC_C_WDISCONNECT	0x00040000  /* waiting for channel disconnect */
503 504 505 506 507 508 509 510 511 512

/*
 * Manages channels on a partition basis. There is one of these structures
 * for each partition (a partition will never utilize the structure that
 * represents itself).
 */
struct xpc_partition {

	/* XPC HB infrastructure */

513
	u8 remote_rp_version;	/* version# of partition's rsvd pg */
514
	struct timespec remote_rp_stamp; /* time when rsvd pg was initialized */
515 516
	u64 remote_rp_pa;	/* phys addr of partition's rsvd pg */
	u64 remote_vars_pa;	/* phys addr of partition's vars */
517
	u64 remote_vars_part_pa;	/* phys addr of partition's vars part */
518
	u64 last_heartbeat;	/* HB at last read */
519
	u64 remote_amos_page_pa;	/* phys addr of partition's amos page */
520
	int remote_act_nasid;	/* active part's act/deact nasid */
521
	int remote_act_phys_cpuid;	/* active part's act/deact phys cpuid */
522 523 524 525
	u32 act_IRQ_rcvd;	/* IRQs since activation */
	spinlock_t act_lock;	/* protect updating of act_state */
	u8 act_state;		/* from XPC HB viewpoint */
	u8 remote_vars_version;	/* version# of partition's vars */
526
	enum xp_retval reason;	/* reason partition is deactivating */
527 528 529 530
	int reason_line;	/* line# deactivation initiated from */
	int reactivate_nasid;	/* nasid in partition to reactivate */

	unsigned long disengage_request_timeout;	/* timeout in jiffies */
531 532
	struct timer_list disengage_request_timer;

533 534
	/* XPC infrastructure referencing and teardown control */

535
	u8 setup_state;		/* infrastructure setup state */
536
	wait_queue_head_t teardown_wq;	/* kthread waiting to teardown infra */
537
	atomic_t references;	/* #of references to infrastructure */
538 539 540 541 542 543 544 545

	/*
	 * NONE OF THE PRECEDING FIELDS OF THIS STRUCTURE WILL BE CLEARED WHEN
	 * XPC SETS UP THE NECESSARY INFRASTRUCTURE TO SUPPORT CROSS PARTITION
	 * COMMUNICATION. ALL OF THE FOLLOWING FIELDS WILL BE CLEARED. (THE
	 * 'nchannels' FIELD MUST BE THE FIRST OF THE FIELDS TO BE CLEARED.)
	 */

546
	u8 nchannels;		/* #of defined channels supported */
547 548
	atomic_t nchannels_active;  /* #of channels that are not DISCONNECTED */
	atomic_t nchannels_engaged;  /* #of channels engaged with remote part */
549 550 551 552 553
	struct xpc_channel *channels;	/* array of channel structures */

	void *local_GPs_base;	/* base address of kmalloc'd space */
	struct xpc_gp *local_GPs;	/* local Get/Put values */
	void *remote_GPs_base;	/* base address of kmalloc'd space */
554 555
	struct xpc_gp *remote_GPs;	/* copy of remote partition's local */
					/* Get/Put values */
556 557
	u64 remote_GPs_pa;	/* phys address of remote partition's local */
				/* Get/Put values */
558 559 560

	/* fields used to pass args when opening or closing a channel */

561 562 563 564 565
	void *local_openclose_args_base;   /* base address of kmalloc'd space */
	struct xpc_openclose_args *local_openclose_args;      /* local's args */
	void *remote_openclose_args_base;  /* base address of kmalloc'd space */
	struct xpc_openclose_args *remote_openclose_args; /* copy of remote's */
							  /* args */
566
	u64 remote_openclose_args_pa;	/* phys addr of remote's args */
567 568 569

	/* IPI sending, receiving and handling related fields */

570 571
	int remote_IPI_nasid;	/* nasid of where to send IPIs */
	int remote_IPI_phys_cpuid;	/* phys CPU ID of where to send IPIs */
572
	AMO_t *remote_IPI_amo_va;    /* address of remote IPI AMO_t structure */
573

574 575 576 577
	AMO_t *local_IPI_amo_va;	/* address of IPI AMO_t structure */
	u64 local_IPI_amo;	/* IPI amo flags yet to be handled */
	char IPI_owner[8];	/* IPI owner's name */
	struct timer_list dropped_IPI_timer;	/* dropped IPI timer */
578

579
	spinlock_t IPI_lock;	/* IPI handler lock */
580 581 582 583

	/* channel manager related fields */

	atomic_t channel_mgr_requests;	/* #of requests to activate chan mgr */
584
	wait_queue_head_t channel_mgr_wq;	/* channel mgr's wait queue */
585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612

} ____cacheline_aligned;

/* struct xpc_partition act_state values (for XPC HB) */

#define	XPC_P_INACTIVE		0x00	/* partition is not active */
#define XPC_P_ACTIVATION_REQ	0x01	/* created thread to activate */
#define XPC_P_ACTIVATING	0x02	/* activation thread started */
#define XPC_P_ACTIVE		0x03	/* xpc_partition_up() was called */
#define XPC_P_DEACTIVATING	0x04	/* partition deactivation initiated */

#define XPC_DEACTIVATE_PARTITION(_p, _reason) \
			xpc_deactivate_partition(__LINE__, (_p), (_reason))

/* struct xpc_partition setup_state values */

#define XPC_P_UNSET		0x00	/* infrastructure was never setup */
#define XPC_P_SETUP		0x01	/* infrastructure is setup */
#define XPC_P_WTEARDOWN		0x02	/* waiting to teardown infrastructure */
#define XPC_P_TORNDOWN		0x03	/* infrastructure is torndown */

/*
 * struct xpc_partition IPI_timer #of seconds to wait before checking for
 * dropped IPIs. These occur whenever an IPI amo write doesn't complete until
 * after the IPI was received.
 */
#define XPC_P_DROPPED_IPI_WAIT	(0.25 * HZ)

613
/* number of seconds to wait for other partitions to disengage */
614
#define XPC_DISENGAGE_REQUEST_DEFAULT_TIMELIMIT	90
615 616 617 618

/* interval in seconds to print 'waiting disengagement' messages */
#define XPC_DISENGAGE_PRINTMSG_INTERVAL		10

619
#define XPC_PARTID(_p)	((short)((_p) - &xpc_partitions[0]))
620 621 622 623

/* found in xp_main.c */
extern struct xpc_registration xpc_registrations[];

624
/* found in xpc_main.c */
625 626
extern struct device *xpc_part;
extern struct device *xpc_chan;
627
extern int xpc_disengage_request_timelimit;
628
extern int xpc_disengage_request_timedout;
A
Al Viro 已提交
629
extern irqreturn_t xpc_notify_IRQ_handler(int, void *);
630
extern void xpc_dropped_IPI_check(struct xpc_partition *);
631
extern void xpc_activate_partition(struct xpc_partition *);
632
extern void xpc_activate_kthreads(struct xpc_channel *, int);
633
extern void xpc_create_kthreads(struct xpc_channel *, int, int);
634 635 636 637 638 639 640
extern void xpc_disconnect_wait(int);

/* found in xpc_partition.c */
extern int xpc_exiting;
extern struct xpc_vars *xpc_vars;
extern struct xpc_rsvd_page *xpc_rsvd_page;
extern struct xpc_vars_part *xpc_vars_part;
641
extern struct xpc_partition *xpc_partitions;
642 643 644
extern char *xpc_remote_copy_buffer;
extern void *xpc_remote_copy_buffer_base;
extern void *xpc_kmalloc_cacheline_aligned(size_t, gfp_t, void **);
645 646 647 648
extern struct xpc_rsvd_page *xpc_rsvd_page_init(void);
extern void xpc_allow_IPI_ops(void);
extern void xpc_restrict_IPI_ops(void);
extern int xpc_identify_act_IRQ_sender(void);
649
extern int xpc_partition_disengaged(struct xpc_partition *);
650
extern enum xp_retval xpc_mark_partition_active(struct xpc_partition *);
651 652 653 654
extern void xpc_mark_partition_inactive(struct xpc_partition *);
extern void xpc_discovery(void);
extern void xpc_check_remote_hb(void);
extern void xpc_deactivate_partition(const int, struct xpc_partition *,
655
				     enum xp_retval);
656
extern enum xp_retval xpc_initiate_partid_to_nasids(short, void *);
657 658 659 660

/* found in xpc_channel.c */
extern void xpc_initiate_connect(int);
extern void xpc_initiate_disconnect(int);
661 662 663
extern enum xp_retval xpc_initiate_allocate(short, int, u32, void **);
extern enum xp_retval xpc_initiate_send(short, int, void *);
extern enum xp_retval xpc_initiate_send_notify(short, int, void *,
664
					       xpc_notify_func, void *);
665
extern void xpc_initiate_received(short, int, void *);
666 667
extern enum xp_retval xpc_setup_infrastructure(struct xpc_partition *);
extern enum xp_retval xpc_pull_remote_vars_part(struct xpc_partition *);
668 669 670 671
extern void xpc_process_channel_activity(struct xpc_partition *);
extern void xpc_connected_callout(struct xpc_channel *);
extern void xpc_deliver_msg(struct xpc_channel *);
extern void xpc_disconnect_channel(const int, struct xpc_channel *,
672 673 674
				   enum xp_retval, unsigned long *);
extern void xpc_disconnect_callout(struct xpc_channel *, enum xp_retval);
extern void xpc_partition_going_down(struct xpc_partition *, enum xp_retval);
675 676 677 678 679
extern void xpc_teardown_infrastructure(struct xpc_partition *);

static inline void
xpc_wakeup_channel_mgr(struct xpc_partition *part)
{
680
	if (atomic_inc_return(&part->channel_mgr_requests) == 1)
681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699
		wake_up(&part->channel_mgr_wq);
}

/*
 * These next two inlines are used to keep us from tearing down a channel's
 * msg queues while a thread may be referencing them.
 */
static inline void
xpc_msgqueue_ref(struct xpc_channel *ch)
{
	atomic_inc(&ch->references);
}

static inline void
xpc_msgqueue_deref(struct xpc_channel *ch)
{
	s32 refs = atomic_dec_return(&ch->references);

	DBUG_ON(refs < 0);
700
	if (refs == 0)
701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716
		xpc_wakeup_channel_mgr(&xpc_partitions[ch->partid]);
}

#define XPC_DISCONNECT_CHANNEL(_ch, _reason, _irqflgs) \
		xpc_disconnect_channel(__LINE__, _ch, _reason, _irqflgs)

/*
 * These two inlines are used to keep us from tearing down a partition's
 * setup infrastructure while a thread may be referencing it.
 */
static inline void
xpc_part_deref(struct xpc_partition *part)
{
	s32 refs = atomic_dec_return(&part->references);

	DBUG_ON(refs < 0);
717
	if (refs == 0 && part->setup_state == XPC_P_WTEARDOWN)
718 719 720 721 722 723 724 725 726 727
		wake_up(&part->teardown_wq);
}

static inline int
xpc_part_ref(struct xpc_partition *part)
{
	int setup;

	atomic_inc(&part->references);
	setup = (part->setup_state == XPC_P_SETUP);
728
	if (!setup)
729
		xpc_part_deref(part);
730

731 732 733 734 735 736 737 738 739 740 741 742 743 744
	return setup;
}

/*
 * The following macro is to be used for the setting of the reason and
 * reason_line fields in both the struct xpc_channel and struct xpc_partition
 * structures.
 */
#define XPC_SET_REASON(_p, _reason, _line) \
	{ \
		(_p)->reason = _reason; \
		(_p)->reason_line = _line; \
	}

745 746 747 748 749 750 751 752 753
/*
 * This next set of inlines are used to keep track of when a partition is
 * potentially engaged in accessing memory belonging to another partition.
 */

static inline void
xpc_mark_partition_engaged(struct xpc_partition *part)
{
	unsigned long irq_flags;
754 755 756
	AMO_t *amo = (AMO_t *)__va(part->remote_amos_page_pa +
				   (XPC_ENGAGED_PARTITIONS_AMO *
				    sizeof(AMO_t)));
757 758 759 760

	local_irq_save(irq_flags);

	/* set bit corresponding to our partid in remote partition's AMO */
761 762
	FETCHOP_STORE_OP(TO_AMO((u64)&amo->variable), FETCHOP_OR,
			 (1UL << sn_partition_id));
763 764 765 766 767 768
	/*
	 * We must always use the nofault function regardless of whether we
	 * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we
	 * didn't, we'd never know that the other partition is down and would
	 * keep sending IPIs and AMOs to it until the heartbeat times out.
	 */
769 770 771
	(void)xp_nofault_PIOR((u64 *)GLOBAL_MMR_ADDR(NASID_GET(&amo->
							       variable),
						     xp_nofault_PIOR_target));
772 773 774 775 776 777 778 779

	local_irq_restore(irq_flags);
}

static inline void
xpc_mark_partition_disengaged(struct xpc_partition *part)
{
	unsigned long irq_flags;
780 781 782
	AMO_t *amo = (AMO_t *)__va(part->remote_amos_page_pa +
				   (XPC_ENGAGED_PARTITIONS_AMO *
				    sizeof(AMO_t)));
783 784 785 786

	local_irq_save(irq_flags);

	/* clear bit corresponding to our partid in remote partition's AMO */
787 788
	FETCHOP_STORE_OP(TO_AMO((u64)&amo->variable), FETCHOP_AND,
			 ~(1UL << sn_partition_id));
789 790 791 792 793 794
	/*
	 * We must always use the nofault function regardless of whether we
	 * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we
	 * didn't, we'd never know that the other partition is down and would
	 * keep sending IPIs and AMOs to it until the heartbeat times out.
	 */
795 796 797
	(void)xp_nofault_PIOR((u64 *)GLOBAL_MMR_ADDR(NASID_GET(&amo->
							       variable),
						     xp_nofault_PIOR_target));
798 799 800 801 802 803 804 805

	local_irq_restore(irq_flags);
}

static inline void
xpc_request_partition_disengage(struct xpc_partition *part)
{
	unsigned long irq_flags;
806 807
	AMO_t *amo = (AMO_t *)__va(part->remote_amos_page_pa +
				   (XPC_DISENGAGE_REQUEST_AMO * sizeof(AMO_t)));
808 809 810 811

	local_irq_save(irq_flags);

	/* set bit corresponding to our partid in remote partition's AMO */
812 813
	FETCHOP_STORE_OP(TO_AMO((u64)&amo->variable), FETCHOP_OR,
			 (1UL << sn_partition_id));
814 815 816 817 818 819
	/*
	 * We must always use the nofault function regardless of whether we
	 * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we
	 * didn't, we'd never know that the other partition is down and would
	 * keep sending IPIs and AMOs to it until the heartbeat times out.
	 */
820 821 822
	(void)xp_nofault_PIOR((u64 *)GLOBAL_MMR_ADDR(NASID_GET(&amo->
							       variable),
						     xp_nofault_PIOR_target));
823 824 825 826 827 828 829 830

	local_irq_restore(irq_flags);
}

static inline void
xpc_cancel_partition_disengage_request(struct xpc_partition *part)
{
	unsigned long irq_flags;
831 832
	AMO_t *amo = (AMO_t *)__va(part->remote_amos_page_pa +
				   (XPC_DISENGAGE_REQUEST_AMO * sizeof(AMO_t)));
833 834 835 836

	local_irq_save(irq_flags);

	/* clear bit corresponding to our partid in remote partition's AMO */
837 838
	FETCHOP_STORE_OP(TO_AMO((u64)&amo->variable), FETCHOP_AND,
			 ~(1UL << sn_partition_id));
839 840 841 842 843 844
	/*
	 * We must always use the nofault function regardless of whether we
	 * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we
	 * didn't, we'd never know that the other partition is down and would
	 * keep sending IPIs and AMOs to it until the heartbeat times out.
	 */
845 846 847
	(void)xp_nofault_PIOR((u64 *)GLOBAL_MMR_ADDR(NASID_GET(&amo->
							       variable),
						     xp_nofault_PIOR_target));
848 849 850 851 852 853 854 855 856 857

	local_irq_restore(irq_flags);
}

static inline u64
xpc_partition_engaged(u64 partid_mask)
{
	AMO_t *amo = xpc_vars->amos_page + XPC_ENGAGED_PARTITIONS_AMO;

	/* return our partition's AMO variable ANDed with partid_mask */
858 859
	return (FETCHOP_LOAD_OP(TO_AMO((u64)&amo->variable), FETCHOP_LOAD) &
		partid_mask);
860 861 862 863 864 865 866 867
}

static inline u64
xpc_partition_disengage_requested(u64 partid_mask)
{
	AMO_t *amo = xpc_vars->amos_page + XPC_DISENGAGE_REQUEST_AMO;

	/* return our partition's AMO variable ANDed with partid_mask */
868 869
	return (FETCHOP_LOAD_OP(TO_AMO((u64)&amo->variable), FETCHOP_LOAD) &
		partid_mask);
870 871 872 873 874 875 876 877
}

static inline void
xpc_clear_partition_engaged(u64 partid_mask)
{
	AMO_t *amo = xpc_vars->amos_page + XPC_ENGAGED_PARTITIONS_AMO;

	/* clear bit(s) based on partid_mask in our partition's AMO */
878 879
	FETCHOP_STORE_OP(TO_AMO((u64)&amo->variable), FETCHOP_AND,
			 ~partid_mask);
880 881 882 883 884 885 886 887
}

static inline void
xpc_clear_partition_disengage_request(u64 partid_mask)
{
	AMO_t *amo = xpc_vars->amos_page + XPC_DISENGAGE_REQUEST_AMO;

	/* clear bit(s) based on partid_mask in our partition's AMO */
888 889
	FETCHOP_STORE_OP(TO_AMO((u64)&amo->variable), FETCHOP_AND,
			 ~partid_mask);
890 891
}

892 893 894 895 896 897 898 899 900 901
/*
 * The following set of macros and inlines are used for the sending and
 * receiving of IPIs (also known as IRQs). There are two flavors of IPIs,
 * one that is associated with partition activity (SGI_XPC_ACTIVATE) and
 * the other that is associated with channel activity (SGI_XPC_NOTIFY).
 */

static inline u64
xpc_IPI_receive(AMO_t *amo)
{
902
	return FETCHOP_LOAD_OP(TO_AMO((u64)&amo->variable), FETCHOP_CLEAR);
903 904
}

905
static inline enum xp_retval
906 907 908 909 910 911 912
xpc_IPI_send(AMO_t *amo, u64 flag, int nasid, int phys_cpuid, int vector)
{
	int ret = 0;
	unsigned long irq_flags;

	local_irq_save(irq_flags);

913
	FETCHOP_STORE_OP(TO_AMO((u64)&amo->variable), FETCHOP_OR, flag);
914 915 916 917 918 919 920 921
	sn_send_IPI_phys(nasid, phys_cpuid, vector, 0);

	/*
	 * We must always use the nofault function regardless of whether we
	 * are on a Shub 1.1 system or a Shub 1.2 slice 0xc processor. If we
	 * didn't, we'd never know that the other partition is down and would
	 * keep sending IPIs and AMOs to it until the heartbeat times out.
	 */
922 923
	ret = xp_nofault_PIOR((u64 *)GLOBAL_MMR_ADDR(NASID_GET(&amo->variable),
						     xp_nofault_PIOR_target));
924 925 926

	local_irq_restore(irq_flags);

927
	return ((ret == 0) ? xpSuccess : xpPioReadError);
928 929 930 931 932 933 934 935 936 937
}

/*
 * IPIs associated with SGI_XPC_ACTIVATE IRQ.
 */

/*
 * Flag the appropriate AMO variable and send an IPI to the specified node.
 */
static inline void
938
xpc_activate_IRQ_send(u64 amos_page_pa, int from_nasid, int to_nasid,
939
		      int to_phys_cpuid)
940 941 942
{
	int w_index = XPC_NASID_W_INDEX(from_nasid);
	int b_index = XPC_NASID_B_INDEX(from_nasid);
943 944
	AMO_t *amos = (AMO_t *)__va(amos_page_pa +
				    (XPC_ACTIVATE_IRQ_AMOS * sizeof(AMO_t)));
945

946 947
	(void)xpc_IPI_send(&amos[w_index], (1UL << b_index), to_nasid,
			   to_phys_cpuid, SGI_XPC_ACTIVATE);
948 949 950 951 952 953
}

static inline void
xpc_IPI_send_activate(struct xpc_vars *vars)
{
	xpc_activate_IRQ_send(vars->amos_page_pa, cnodeid_to_nasid(0),
954
			      vars->act_nasid, vars->act_phys_cpuid);
955 956 957 958 959 960
}

static inline void
xpc_IPI_send_activated(struct xpc_partition *part)
{
	xpc_activate_IRQ_send(part->remote_amos_page_pa, cnodeid_to_nasid(0),
961 962
			      part->remote_act_nasid,
			      part->remote_act_phys_cpuid);
963 964 965 966 967 968
}

static inline void
xpc_IPI_send_reactivate(struct xpc_partition *part)
{
	xpc_activate_IRQ_send(xpc_vars->amos_page_pa, part->reactivate_nasid,
969
			      xpc_vars->act_nasid, xpc_vars->act_phys_cpuid);
970 971
}

972 973 974 975
static inline void
xpc_IPI_send_disengage(struct xpc_partition *part)
{
	xpc_activate_IRQ_send(part->remote_amos_page_pa, cnodeid_to_nasid(0),
976 977
			      part->remote_act_nasid,
			      part->remote_act_phys_cpuid);
978 979
}

980 981 982 983 984 985 986 987 988 989 990 991 992
/*
 * IPIs associated with SGI_XPC_NOTIFY IRQ.
 */

/*
 * Send an IPI to the remote partition that is associated with the
 * specified channel.
 */
#define XPC_NOTIFY_IRQ_SEND(_ch, _ipi_f, _irq_f) \
		xpc_notify_IRQ_send(_ch, _ipi_f, #_ipi_f, _irq_f)

static inline void
xpc_notify_IRQ_send(struct xpc_channel *ch, u8 ipi_flag, char *ipi_flag_string,
993
		    unsigned long *irq_flags)
994 995
{
	struct xpc_partition *part = &xpc_partitions[ch->partid];
996
	enum xp_retval ret;
997 998 999

	if (likely(part->act_state != XPC_P_DEACTIVATING)) {
		ret = xpc_IPI_send(part->remote_IPI_amo_va,
1000 1001 1002
				   (u64)ipi_flag << (ch->number * 8),
				   part->remote_IPI_nasid,
				   part->remote_IPI_phys_cpuid, SGI_XPC_NOTIFY);
1003 1004
		dev_dbg(xpc_chan, "%s sent to partid=%d, channel=%d, ret=%d\n",
			ipi_flag_string, ch->partid, ch->number, ret);
1005
		if (unlikely(ret != xpSuccess)) {
1006
			if (irq_flags != NULL)
1007 1008
				spin_unlock_irqrestore(&ch->lock, *irq_flags);
			XPC_DEACTIVATE_PARTITION(part, ret);
1009
			if (irq_flags != NULL)
1010 1011 1012 1013 1014 1015 1016 1017 1018 1019 1020 1021 1022 1023 1024
				spin_lock_irqsave(&ch->lock, *irq_flags);
		}
	}
}

/*
 * Make it look like the remote partition, which is associated with the
 * specified channel, sent us an IPI. This faked IPI will be handled
 * by xpc_dropped_IPI_check().
 */
#define XPC_NOTIFY_IRQ_SEND_LOCAL(_ch, _ipi_f) \
		xpc_notify_IRQ_send_local(_ch, _ipi_f, #_ipi_f)

static inline void
xpc_notify_IRQ_send_local(struct xpc_channel *ch, u8 ipi_flag,
1025
			  char *ipi_flag_string)
1026 1027 1028
{
	struct xpc_partition *part = &xpc_partitions[ch->partid];

1029 1030
	FETCHOP_STORE_OP(TO_AMO((u64)&part->local_IPI_amo_va->variable),
			 FETCHOP_OR, ((u64)ipi_flag << (ch->number * 8)));
1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050
	dev_dbg(xpc_chan, "%s sent local from partid=%d, channel=%d\n",
		ipi_flag_string, ch->partid, ch->number);
}

/*
 * The sending and receiving of IPIs includes the setting of an AMO variable
 * to indicate the reason the IPI was sent. The 64-bit variable is divided
 * up into eight bytes, ordered from right to left. Byte zero pertains to
 * channel 0, byte one to channel 1, and so on. Each byte is described by
 * the following IPI flags.
 */

#define	XPC_IPI_CLOSEREQUEST	0x01
#define	XPC_IPI_CLOSEREPLY	0x02
#define	XPC_IPI_OPENREQUEST	0x04
#define	XPC_IPI_OPENREPLY	0x08
#define	XPC_IPI_MSGREQUEST	0x10

/* given an AMO variable and a channel#, get its associated IPI flags */
#define XPC_GET_IPI_FLAGS(_amo, _c)	((u8) (((_amo) >> ((_c) * 8)) & 0xff))
1051
#define XPC_SET_IPI_FLAGS(_amo, _c, _f)	(_amo) |= ((u64) (_f) << ((_c) * 8))
1052

1053 1054
#define	XPC_ANY_OPENCLOSE_IPI_FLAGS_SET(_amo) ((_amo) & 0x0f0f0f0f0f0f0f0fUL)
#define XPC_ANY_MSG_IPI_FLAGS_SET(_amo)       ((_amo) & 0x1010101010101010UL)
1055 1056 1057 1058 1059 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107

static inline void
xpc_IPI_send_closerequest(struct xpc_channel *ch, unsigned long *irq_flags)
{
	struct xpc_openclose_args *args = ch->local_openclose_args;

	args->reason = ch->reason;

	XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_CLOSEREQUEST, irq_flags);
}

static inline void
xpc_IPI_send_closereply(struct xpc_channel *ch, unsigned long *irq_flags)
{
	XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_CLOSEREPLY, irq_flags);
}

static inline void
xpc_IPI_send_openrequest(struct xpc_channel *ch, unsigned long *irq_flags)
{
	struct xpc_openclose_args *args = ch->local_openclose_args;

	args->msg_size = ch->msg_size;
	args->local_nentries = ch->local_nentries;

	XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_OPENREQUEST, irq_flags);
}

static inline void
xpc_IPI_send_openreply(struct xpc_channel *ch, unsigned long *irq_flags)
{
	struct xpc_openclose_args *args = ch->local_openclose_args;

	args->remote_nentries = ch->remote_nentries;
	args->local_nentries = ch->local_nentries;
	args->local_msgqueue_pa = __pa(ch->local_msgqueue);

	XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_OPENREPLY, irq_flags);
}

static inline void
xpc_IPI_send_msgrequest(struct xpc_channel *ch)
{
	XPC_NOTIFY_IRQ_SEND(ch, XPC_IPI_MSGREQUEST, NULL);
}

static inline void
xpc_IPI_send_local_msgrequest(struct xpc_channel *ch)
{
	XPC_NOTIFY_IRQ_SEND_LOCAL(ch, XPC_IPI_MSGREQUEST);
}

/*
1108
>>> this block comment needs to be moved and re-written.
1109 1110 1111 1112 1113 1114
 * Memory for XPC's AMO variables is allocated by the MSPEC driver. These
 * pages are located in the lowest granule. The lowest granule uses 4k pages
 * for cached references and an alternate TLB handler to never provide a
 * cacheable mapping for the entire region. This will prevent speculative
 * reading of cached copies of our lines from being issued which will cause
 * a PI FSB Protocol error to be generated by the SHUB. For XPC, we need 64
1115
 * AMO variables (based on xp_max_npartitions) for message notification and an
1116 1117
 * additional 128 AMO variables (based on XP_NASID_MASK_WORDS) for partition
 * activation and 2 AMO variables for partition deactivation.
1118 1119
 */
static inline AMO_t *
1120
xpc_IPI_init(int index)
1121
{
1122
	AMO_t *amo = xpc_vars->amos_page + index;
1123

1124
	(void)xpc_IPI_receive(amo);	/* clear AMO variable */
1125
	return amo;
1126 1127
}

1128
static inline enum xp_retval
1129 1130
xpc_map_bte_errors(bte_result_t error)
{
1131
	return ((error == BTE_SUCCESS) ? xpSuccess : xpBteCopyError);
1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144
}

/*
 * Check to see if there is any channel activity to/from the specified
 * partition.
 */
static inline void
xpc_check_for_channel_activity(struct xpc_partition *part)
{
	u64 IPI_amo;
	unsigned long irq_flags;

	IPI_amo = xpc_IPI_receive(part->local_IPI_amo_va);
1145
	if (IPI_amo == 0)
1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157
		return;

	spin_lock_irqsave(&part->IPI_lock, irq_flags);
	part->local_IPI_amo |= IPI_amo;
	spin_unlock_irqrestore(&part->IPI_lock, irq_flags);

	dev_dbg(xpc_chan, "received IPI from partid=%d, IPI_amo=0x%lx\n",
		XPC_PARTID(part), IPI_amo);

	xpc_wakeup_channel_mgr(part);
}

1158
#endif /* _DRIVERS_MISC_SGIXP_XPC_H */