1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
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
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
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
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
|
/*
* Linux MegaRAID driver for SAS based RAID controllers
*
* Copyright (c) 2009-2011 LSI Corporation.
*
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
*
* This program is distributed in the hope that it will be useful,
* but WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
* GNU General Public License for more details.
*
* You should have received a copy of the GNU General Public License
* along with this program; if not, write to the Free Software
* Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*
* FILE: megaraid_sas_fp.c
*
* Authors: LSI Corporation
* Sumant Patro
* Varad Talamacki
* Manoj Jose
*
* Send feedback to: <megaraidlinux@lsi.com>
*
* Mail to: LSI Corporation, 1621 Barber Lane, Milpitas, CA 95035
* ATTN: Linuxraid
*/
#include <linux/kernel.h>
#include <linux/types.h>
#include <linux/pci.h>
#include <linux/list.h>
#include <linux/moduleparam.h>
#include <linux/module.h>
#include <linux/spinlock.h>
#include <linux/interrupt.h>
#include <linux/delay.h>
#include <linux/uio.h>
#include <linux/uaccess.h>
#include <linux/fs.h>
#include <linux/compat.h>
#include <linux/blkdev.h>
#include <linux/poll.h>
#include <scsi/scsi.h>
#include <scsi/scsi_cmnd.h>
#include <scsi/scsi_device.h>
#include <scsi/scsi_host.h>
#include "megaraid_sas_fusion.h"
#include "megaraid_sas.h"
#include <asm/div64.h>
#define ABS_DIFF(a, b) (((a) > (b)) ? ((a) - (b)) : ((b) - (a)))
#define MR_LD_STATE_OPTIMAL 3
#define FALSE 0
#define TRUE 1
/* Prototypes */
void
mr_update_load_balance_params(struct MR_FW_RAID_MAP_ALL *map,
struct LD_LOAD_BALANCE_INFO *lbInfo);
u32 mega_mod64(u64 dividend, u32 divisor)
{
u64 d;
u32 remainder;
if (!divisor)
printk(KERN_ERR "megasas : DIVISOR is zero, in div fn\n");
d = dividend;
remainder = do_div(d, divisor);
return remainder;
}
/**
* @param dividend : Dividend
* @param divisor : Divisor
*
* @return quotient
**/
u64 mega_div64_32(uint64_t dividend, uint32_t divisor)
{
u32 remainder;
u64 d;
if (!divisor)
printk(KERN_ERR "megasas : DIVISOR is zero in mod fn\n");
d = dividend;
remainder = do_div(d, divisor);
return d;
}
struct MR_LD_RAID *MR_LdRaidGet(u32 ld, struct MR_FW_RAID_MAP_ALL *map)
{
return &map->raidMap.ldSpanMap[ld].ldRaid;
}
static struct MR_SPAN_BLOCK_INFO *MR_LdSpanInfoGet(u32 ld,
struct MR_FW_RAID_MAP_ALL
*map)
{
return &map->raidMap.ldSpanMap[ld].spanBlock[0];
}
static u8 MR_LdDataArmGet(u32 ld, u32 armIdx, struct MR_FW_RAID_MAP_ALL *map)
{
return map->raidMap.ldSpanMap[ld].dataArmMap[armIdx];
}
static u16 MR_ArPdGet(u32 ar, u32 arm, struct MR_FW_RAID_MAP_ALL *map)
{
return map->raidMap.arMapInfo[ar].pd[arm];
}
static u16 MR_LdSpanArrayGet(u32 ld, u32 span, struct MR_FW_RAID_MAP_ALL *map)
{
return map->raidMap.ldSpanMap[ld].spanBlock[span].span.arrayRef;
}
static u16 MR_PdDevHandleGet(u32 pd, struct MR_FW_RAID_MAP_ALL *map)
{
return map->raidMap.devHndlInfo[pd].curDevHdl;
}
u16 MR_GetLDTgtId(u32 ld, struct MR_FW_RAID_MAP_ALL *map)
{
return map->raidMap.ldSpanMap[ld].ldRaid.targetId;
}
u16 MR_TargetIdToLdGet(u32 ldTgtId, struct MR_FW_RAID_MAP_ALL *map)
{
return map->raidMap.ldTgtIdToLd[ldTgtId];
}
static struct MR_LD_SPAN *MR_LdSpanPtrGet(u32 ld, u32 span,
struct MR_FW_RAID_MAP_ALL *map)
{
return &map->raidMap.ldSpanMap[ld].spanBlock[span].span;
}
/*
* This function will validate Map info data provided by FW
*/
u8 MR_ValidateMapInfo(struct MR_FW_RAID_MAP_ALL *map,
struct LD_LOAD_BALANCE_INFO *lbInfo)
{
struct MR_FW_RAID_MAP *pFwRaidMap = &map->raidMap;
if (pFwRaidMap->totalSize !=
(sizeof(struct MR_FW_RAID_MAP) -sizeof(struct MR_LD_SPAN_MAP) +
(sizeof(struct MR_LD_SPAN_MAP) *pFwRaidMap->ldCount))) {
printk(KERN_ERR "megasas: map info structure size 0x%x is not matching with ld count\n",
(unsigned int)((sizeof(struct MR_FW_RAID_MAP) -
sizeof(struct MR_LD_SPAN_MAP)) +
(sizeof(struct MR_LD_SPAN_MAP) *
pFwRaidMap->ldCount)));
printk(KERN_ERR "megasas: span map %x, pFwRaidMap->totalSize "
": %x\n", (unsigned int)sizeof(struct MR_LD_SPAN_MAP),
pFwRaidMap->totalSize);
return 0;
}
mr_update_load_balance_params(map, lbInfo);
return 1;
}
u32 MR_GetSpanBlock(u32 ld, u64 row, u64 *span_blk,
struct MR_FW_RAID_MAP_ALL *map, int *div_error)
{
struct MR_SPAN_BLOCK_INFO *pSpanBlock = MR_LdSpanInfoGet(ld, map);
struct MR_QUAD_ELEMENT *quad;
struct MR_LD_RAID *raid = MR_LdRaidGet(ld, map);
u32 span, j;
for (span = 0; span < raid->spanDepth; span++, pSpanBlock++) {
for (j = 0; j < pSpanBlock->block_span_info.noElements; j++) {
quad = &pSpanBlock->block_span_info.quad[j];
if (quad->diff == 0) {
*div_error = 1;
return span;
}
if (quad->logStart <= row && row <= quad->logEnd &&
(mega_mod64(row-quad->logStart, quad->diff)) == 0) {
if (span_blk != NULL) {
u64 blk, debugBlk;
blk =
mega_div64_32(
(row-quad->logStart),
quad->diff);
debugBlk = blk;
blk = (blk + quad->offsetInSpan) <<
raid->stripeShift;
*span_blk = blk;
}
return span;
}
}
}
return span;
}
/*
******************************************************************************
*
* This routine calculates the arm, span and block for the specified stripe and
* reference in stripe.
*
* Inputs :
*
* ld - Logical drive number
* stripRow - Stripe number
* stripRef - Reference in stripe
*
* Outputs :
*
* span - Span number
* block - Absolute Block number in the physical disk
*/
u8 MR_GetPhyParams(struct megasas_instance *instance, u32 ld, u64 stripRow,
u16 stripRef, u64 *pdBlock, u16 *pDevHandle,
struct RAID_CONTEXT *pRAID_Context,
struct MR_FW_RAID_MAP_ALL *map)
{
struct MR_LD_RAID *raid = MR_LdRaidGet(ld, map);
u32 pd, arRef;
u8 physArm, span;
u64 row;
u8 retval = TRUE;
int error_code = 0;
row = mega_div64_32(stripRow, raid->rowDataSize);
if (raid->level == 6) {
/* logical arm within row */
u32 logArm = mega_mod64(stripRow, raid->rowDataSize);
u32 rowMod, armQ, arm;
if (raid->rowSize == 0)
return FALSE;
/* get logical row mod */
rowMod = mega_mod64(row, raid->rowSize);
armQ = raid->rowSize-1-rowMod; /* index of Q drive */
arm = armQ+1+logArm; /* data always logically follows Q */
if (arm >= raid->rowSize) /* handle wrap condition */
arm -= raid->rowSize;
physArm = (u8)arm;
} else {
if (raid->modFactor == 0)
return FALSE;
physArm = MR_LdDataArmGet(ld, mega_mod64(stripRow,
raid->modFactor),
map);
}
if (raid->spanDepth == 1) {
span = 0;
*pdBlock = row << raid->stripeShift;
} else {
span = (u8)MR_GetSpanBlock(ld, row, pdBlock, map, &error_code);
if (error_code == 1)
return FALSE;
}
/* Get the array on which this span is present */
arRef = MR_LdSpanArrayGet(ld, span, map);
pd = MR_ArPdGet(arRef, physArm, map); /* Get the pd */
if (pd != MR_PD_INVALID)
/* Get dev handle from Pd. */
*pDevHandle = MR_PdDevHandleGet(pd, map);
else {
*pDevHandle = MR_PD_INVALID; /* set dev handle as invalid. */
if ((raid->level >= 5) &&
((instance->pdev->device != PCI_DEVICE_ID_LSI_INVADER) ||
(instance->pdev->device == PCI_DEVICE_ID_LSI_INVADER &&
raid->regTypeReqOnRead != REGION_TYPE_UNUSED)))
pRAID_Context->regLockFlags = REGION_TYPE_EXCLUSIVE;
else if (raid->level == 1) {
/* Get alternate Pd. */
pd = MR_ArPdGet(arRef, physArm + 1, map);
if (pd != MR_PD_INVALID)
/* Get dev handle from Pd */
*pDevHandle = MR_PdDevHandleGet(pd, map);
}
}
*pdBlock += stripRef + MR_LdSpanPtrGet(ld, span, map)->startBlk;
pRAID_Context->spanArm = (span << RAID_CTX_SPANARM_SPAN_SHIFT) |
physArm;
return retval;
}
/*
******************************************************************************
*
* MR_BuildRaidContext function
*
* This function will initiate command processing. The start/end row and strip
* information is calculated then the lock is acquired.
* This function will return 0 if region lock was acquired OR return num strips
*/
u8
MR_BuildRaidContext(struct megasas_instance *instance,
struct IO_REQUEST_INFO *io_info,
struct RAID_CONTEXT *pRAID_Context,
struct MR_FW_RAID_MAP_ALL *map)
{
struct MR_LD_RAID *raid;
u32 ld, stripSize, stripe_mask;
u64 endLba, endStrip, endRow, start_row, start_strip;
u64 regStart;
u32 regSize;
u8 num_strips, numRows;
u16 ref_in_start_stripe, ref_in_end_stripe;
u64 ldStartBlock;
u32 numBlocks, ldTgtId;
u8 isRead;
u8 retval = 0;
ldStartBlock = io_info->ldStartBlock;
numBlocks = io_info->numBlocks;
ldTgtId = io_info->ldTgtId;
isRead = io_info->isRead;
ld = MR_TargetIdToLdGet(ldTgtId, map);
raid = MR_LdRaidGet(ld, map);
stripSize = 1 << raid->stripeShift;
stripe_mask = stripSize-1;
/*
* calculate starting row and stripe, and number of strips and rows
*/
start_strip = ldStartBlock >> raid->stripeShift;
ref_in_start_stripe = (u16)(ldStartBlock & stripe_mask);
endLba = ldStartBlock + numBlocks - 1;
ref_in_end_stripe = (u16)(endLba & stripe_mask);
endStrip = endLba >> raid->stripeShift;
num_strips = (u8)(endStrip - start_strip + 1); /* End strip */
if (raid->rowDataSize == 0)
return FALSE;
start_row = mega_div64_32(start_strip, raid->rowDataSize);
endRow = mega_div64_32(endStrip, raid->rowDataSize);
numRows = (u8)(endRow - start_row + 1);
/*
* calculate region info.
*/
/* assume region is at the start of the first row */
regStart = start_row << raid->stripeShift;
/* assume this IO needs the full row - we'll adjust if not true */
regSize = stripSize;
/* Check if we can send this I/O via FastPath */
if (raid->capability.fpCapable) {
if (isRead)
io_info->fpOkForIo = (raid->capability.fpReadCapable &&
((num_strips == 1) ||
raid->capability.
fpReadAcrossStripe));
else
io_info->fpOkForIo = (raid->capability.fpWriteCapable &&
((num_strips == 1) ||
raid->capability.
fpWriteAcrossStripe));
} else
io_info->fpOkForIo = FALSE;
if (numRows == 1) {
/* single-strip IOs can always lock only the data needed */
if (num_strips == 1) {
regStart += ref_in_start_stripe;
regSize = numBlocks;
}
/* multi-strip IOs always need to full stripe locked */
} else {
if (start_strip == (start_row + 1) * raid->rowDataSize - 1) {
/* If the start strip is the last in the start row */
regStart += ref_in_start_stripe;
regSize = stripSize - ref_in_start_stripe;
/* initialize count to sectors from startref to end
of strip */
}
if (numRows > 2)
/* Add complete rows in the middle of the transfer */
regSize += (numRows-2) << raid->stripeShift;
/* if IO ends within first strip of last row */
if (endStrip == endRow*raid->rowDataSize)
regSize += ref_in_end_stripe+1;
else
regSize += stripSize;
}
pRAID_Context->timeoutValue = map->raidMap.fpPdIoTimeoutSec;
if (instance->pdev->device == PCI_DEVICE_ID_LSI_INVADER)
pRAID_Context->regLockFlags = (isRead) ?
raid->regTypeReqOnRead : raid->regTypeReqOnWrite;
else
pRAID_Context->regLockFlags = (isRead) ?
REGION_TYPE_SHARED_READ : raid->regTypeReqOnWrite;
pRAID_Context->VirtualDiskTgtId = raid->targetId;
pRAID_Context->regLockRowLBA = regStart;
pRAID_Context->regLockLength = regSize;
pRAID_Context->configSeqNum = raid->seqNum;
/*Get Phy Params only if FP capable, or else leave it to MR firmware
to do the calculation.*/
if (io_info->fpOkForIo) {
retval = MR_GetPhyParams(instance, ld, start_strip,
ref_in_start_stripe,
&io_info->pdBlock,
&io_info->devHandle, pRAID_Context,
map);
/* If IO on an invalid Pd, then FP i snot possible */
if (io_info->devHandle == MR_PD_INVALID)
io_info->fpOkForIo = FALSE;
return retval;
} else if (isRead) {
uint stripIdx;
for (stripIdx = 0; stripIdx < num_strips; stripIdx++) {
if (!MR_GetPhyParams(instance, ld,
start_strip + stripIdx,
ref_in_start_stripe,
&io_info->pdBlock,
&io_info->devHandle,
pRAID_Context, map))
return TRUE;
}
}
return TRUE;
}
void
mr_update_load_balance_params(struct MR_FW_RAID_MAP_ALL *map,
struct LD_LOAD_BALANCE_INFO *lbInfo)
{
int ldCount;
u16 ld;
struct MR_LD_RAID *raid;
for (ldCount = 0; ldCount < MAX_LOGICAL_DRIVES; ldCount++) {
ld = MR_TargetIdToLdGet(ldCount, map);
if (ld >= MAX_LOGICAL_DRIVES) {
lbInfo[ldCount].loadBalanceFlag = 0;
continue;
}
raid = MR_LdRaidGet(ld, map);
/* Two drive Optimal RAID 1 */
if ((raid->level == 1) && (raid->rowSize == 2) &&
(raid->spanDepth == 1) && raid->ldState ==
MR_LD_STATE_OPTIMAL) {
u32 pd, arRef;
lbInfo[ldCount].loadBalanceFlag = 1;
/* Get the array on which this span is present */
arRef = MR_LdSpanArrayGet(ld, 0, map);
/* Get the Pd */
pd = MR_ArPdGet(arRef, 0, map);
/* Get dev handle from Pd */
lbInfo[ldCount].raid1DevHandle[0] =
MR_PdDevHandleGet(pd, map);
/* Get the Pd */
pd = MR_ArPdGet(arRef, 1, map);
/* Get the dev handle from Pd */
lbInfo[ldCount].raid1DevHandle[1] =
MR_PdDevHandleGet(pd, map);
} else
lbInfo[ldCount].loadBalanceFlag = 0;
}
}
u8 megasas_get_best_arm(struct LD_LOAD_BALANCE_INFO *lbInfo, u8 arm, u64 block,
u32 count)
{
u16 pend0, pend1;
u64 diff0, diff1;
u8 bestArm;
/* get the pending cmds for the data and mirror arms */
pend0 = atomic_read(&lbInfo->scsi_pending_cmds[0]);
pend1 = atomic_read(&lbInfo->scsi_pending_cmds[1]);
/* Determine the disk whose head is nearer to the req. block */
diff0 = ABS_DIFF(block, lbInfo->last_accessed_block[0]);
diff1 = ABS_DIFF(block, lbInfo->last_accessed_block[1]);
bestArm = (diff0 <= diff1 ? 0 : 1);
if ((bestArm == arm && pend0 > pend1 + 16) ||
(bestArm != arm && pend1 > pend0 + 16))
bestArm ^= 1;
/* Update the last accessed block on the correct pd */
lbInfo->last_accessed_block[bestArm] = block + count - 1;
return bestArm;
}
u16 get_updated_dev_handle(struct LD_LOAD_BALANCE_INFO *lbInfo,
struct IO_REQUEST_INFO *io_info)
{
u8 arm, old_arm;
u16 devHandle;
old_arm = lbInfo->raid1DevHandle[0] == io_info->devHandle ? 0 : 1;
/* get best new arm */
arm = megasas_get_best_arm(lbInfo, old_arm, io_info->ldStartBlock,
io_info->numBlocks);
devHandle = lbInfo->raid1DevHandle[arm];
atomic_inc(&lbInfo->scsi_pending_cmds[arm]);
return devHandle;
}
|