dev.c 8.5 KB
Newer Older
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
/*
* Filename: dev.c
*
*
* Authors: Joshua Morris <josh.h.morris@us.ibm.com>
*	Philip Kelleher <pjk1939@linux.vnet.ibm.com>
*
* (C) Copyright 2013 IBM 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
*/

#include <linux/kernel.h>
#include <linux/interrupt.h>
#include <linux/module.h>
#include <linux/pci.h>
#include <linux/slab.h>

#include <linux/hdreg.h>
#include <linux/genhd.h>
#include <linux/blkdev.h>
#include <linux/bio.h>

#include <linux/fs.h>

#include "rsxx_priv.h"

static unsigned int blkdev_minors = 64;
module_param(blkdev_minors, uint, 0444);
MODULE_PARM_DESC(blkdev_minors, "Number of minors(partitions)");

/*
 * For now I'm making this tweakable in case any applications hit this limit.
 * If you see a "bio too big" error in the log you will need to raise this
 * value.
 */
static unsigned int blkdev_max_hw_sectors = 1024;
module_param(blkdev_max_hw_sectors, uint, 0444);
MODULE_PARM_DESC(blkdev_max_hw_sectors, "Max hw sectors for a single BIO");

static unsigned int enable_blkdev = 1;
module_param(enable_blkdev , uint, 0444);
MODULE_PARM_DESC(enable_blkdev, "Enable block device interfaces");


struct rsxx_bio_meta {
	struct bio	*bio;
	atomic_t	pending_dmas;
	atomic_t	error;
	unsigned long	start_time;
};

static struct kmem_cache *bio_meta_pool;

/*----------------- Block Device Operations -----------------*/
static int rsxx_blkdev_ioctl(struct block_device *bdev,
				 fmode_t mode,
				 unsigned int cmd,
				 unsigned long arg)
{
	struct rsxx_cardinfo *card = bdev->bd_disk->private_data;

	switch (cmd) {
	case RSXX_GETREG:
		return rsxx_reg_access(card, (void __user *)arg, 1);
	case RSXX_SETREG:
		return rsxx_reg_access(card, (void __user *)arg, 0);
	}

	return -ENOTTY;
}

static int rsxx_getgeo(struct block_device *bdev, struct hd_geometry *geo)
{
	struct rsxx_cardinfo *card = bdev->bd_disk->private_data;
	u64 blocks = card->size8 >> 9;

	/*
	 * get geometry: Fake it. I haven't found any drivers that set
	 * geo->start, so we won't either.
	 */
	if (card->size8) {
		geo->heads = 64;
		geo->sectors = 16;
		do_div(blocks, (geo->heads * geo->sectors));
		geo->cylinders = blocks;
	} else {
		geo->heads = 0;
		geo->sectors = 0;
		geo->cylinders = 0;
	}
	return 0;
}

static const struct block_device_operations rsxx_fops = {
	.owner		= THIS_MODULE,
	.getgeo		= rsxx_getgeo,
	.ioctl		= rsxx_blkdev_ioctl,
};

static void disk_stats_start(struct rsxx_cardinfo *card, struct bio *bio)
{
	struct hd_struct *part0 = &card->gendisk->part0;
	int rw = bio_data_dir(bio);
	int cpu;

	cpu = part_stat_lock();

	part_round_stats(cpu, part0);
	part_inc_in_flight(part0, rw);

	part_stat_unlock();
}

static void disk_stats_complete(struct rsxx_cardinfo *card,
				struct bio *bio,
				unsigned long start_time)
{
	struct hd_struct *part0 = &card->gendisk->part0;
	unsigned long duration = jiffies - start_time;
	int rw = bio_data_dir(bio);
	int cpu;

	cpu = part_stat_lock();

	part_stat_add(cpu, part0, sectors[rw], bio_sectors(bio));
	part_stat_inc(cpu, part0, ios[rw]);
	part_stat_add(cpu, part0, ticks[rw], duration);

	part_round_stats(cpu, part0);
	part_dec_in_flight(part0, rw);

	part_stat_unlock();
}

static void bio_dma_done_cb(struct rsxx_cardinfo *card,
			    void *cb_data,
			    unsigned int error)
{
152
	struct rsxx_bio_meta *meta = cb_data;
153 154 155 156 157

	if (error)
		atomic_set(&meta->error, 1);

	if (atomic_dec_and_test(&meta->pending_dmas)) {
158 159
		if (!card->eeh_state && card->gendisk)
			disk_stats_complete(card, meta->bio, meta->start_time);
160 161 162 163 164 165 166 167 168 169 170 171 172 173

		bio_endio(meta->bio, atomic_read(&meta->error) ? -EIO : 0);
		kmem_cache_free(bio_meta_pool, meta);
	}
}

static void rsxx_make_request(struct request_queue *q, struct bio *bio)
{
	struct rsxx_cardinfo *card = q->queuedata;
	struct rsxx_bio_meta *bio_meta;
	int st = -EINVAL;

	might_sleep();

174 175 176
	if (!card)
		goto req_err;

177
	if (bio_end_sector(bio) > get_capacity(card->gendisk))
178 179
		goto req_err;

180 181 182 183 184 185 186 187 188 189
	if (unlikely(card->halt)) {
		st = -EFAULT;
		goto req_err;
	}

	if (unlikely(card->dma_fault)) {
		st = (-EFAULT);
		goto req_err;
	}

190
	if (bio->bi_iter.bi_size == 0) {
191 192 193 194 195 196 197 198 199 200 201 202 203 204 205
		dev_err(CARD_TO_DEV(card), "size zero BIO!\n");
		goto req_err;
	}

	bio_meta = kmem_cache_alloc(bio_meta_pool, GFP_KERNEL);
	if (!bio_meta) {
		st = -ENOMEM;
		goto req_err;
	}

	bio_meta->bio = bio;
	atomic_set(&bio_meta->error, 0);
	atomic_set(&bio_meta->pending_dmas, 0);
	bio_meta->start_time = jiffies;

206 207
	if (!unlikely(card->halt))
		disk_stats_start(card, bio);
208 209 210

	dev_dbg(CARD_TO_DEV(card), "BIO[%c]: meta: %p addr8: x%llx size: %d\n",
		 bio_data_dir(bio) ? 'W' : 'R', bio_meta,
211
		 (u64)bio->bi_iter.bi_sector << 9, bio->bi_iter.bi_size);
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

	st = rsxx_dma_queue_bio(card, bio, &bio_meta->pending_dmas,
				    bio_dma_done_cb, bio_meta);
	if (st)
		goto queue_err;

	return;

queue_err:
	kmem_cache_free(bio_meta_pool, bio_meta);
req_err:
	bio_endio(bio, st);
}

/*----------------- Device Setup -------------------*/
static bool rsxx_discard_supported(struct rsxx_cardinfo *card)
{
	unsigned char pci_rev;

	pci_read_config_byte(card->dev, PCI_REVISION_ID, &pci_rev);

	return (pci_rev >= RSXX_DISCARD_SUPPORT);
}

int rsxx_attach_dev(struct rsxx_cardinfo *card)
{
	mutex_lock(&card->dev_lock);

	/* The block device requires the stripe size from the config. */
	if (enable_blkdev) {
		if (card->config_valid)
			set_capacity(card->gendisk, card->size8 >> 9);
		else
			set_capacity(card->gendisk, 0);
		add_disk(card->gendisk);

		card->bdev_attached = 1;
	}

	mutex_unlock(&card->dev_lock);

	return 0;
}

void rsxx_detach_dev(struct rsxx_cardinfo *card)
{
	mutex_lock(&card->dev_lock);

	if (card->bdev_attached) {
		del_gendisk(card->gendisk);
		card->bdev_attached = 0;
	}

	mutex_unlock(&card->dev_lock);
}

int rsxx_setup_dev(struct rsxx_cardinfo *card)
{
	unsigned short blk_size;

	mutex_init(&card->dev_lock);

	if (!enable_blkdev)
		return 0;

	card->major = register_blkdev(0, DRIVER_NAME);
	if (card->major < 0) {
		dev_err(CARD_TO_DEV(card), "Failed to get major number\n");
		return -ENOMEM;
	}

	card->queue = blk_alloc_queue(GFP_KERNEL);
	if (!card->queue) {
		dev_err(CARD_TO_DEV(card), "Failed queue alloc\n");
		unregister_blkdev(card->major, DRIVER_NAME);
		return -ENOMEM;
	}

	card->gendisk = alloc_disk(blkdev_minors);
	if (!card->gendisk) {
		dev_err(CARD_TO_DEV(card), "Failed disk alloc\n");
		blk_cleanup_queue(card->queue);
		unregister_blkdev(card->major, DRIVER_NAME);
		return -ENOMEM;
	}

298 299 300 301 302
	if (card->config_valid) {
		blk_size = card->config.data.block_size;
		blk_queue_dma_alignment(card->queue, blk_size - 1);
		blk_queue_logical_block_size(card->queue, blk_size);
	}
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

	blk_queue_make_request(card->queue, rsxx_make_request);
	blk_queue_bounce_limit(card->queue, BLK_BOUNCE_ANY);
	blk_queue_max_hw_sectors(card->queue, blkdev_max_hw_sectors);
	blk_queue_physical_block_size(card->queue, RSXX_HW_BLK_SIZE);

	queue_flag_set_unlocked(QUEUE_FLAG_NONROT, card->queue);
	if (rsxx_discard_supported(card)) {
		queue_flag_set_unlocked(QUEUE_FLAG_DISCARD, card->queue);
		blk_queue_max_discard_sectors(card->queue,
						RSXX_HW_BLK_SIZE >> 9);
		card->queue->limits.discard_granularity = RSXX_HW_BLK_SIZE;
		card->queue->limits.discard_alignment   = RSXX_HW_BLK_SIZE;
		card->queue->limits.discard_zeroes_data = 1;
	}

	card->queue->queuedata = card;

	snprintf(card->gendisk->disk_name, sizeof(card->gendisk->disk_name),
		 "rsxx%d", card->disk_id);
	card->gendisk->driverfs_dev = &card->dev->dev;
	card->gendisk->major = card->major;
	card->gendisk->first_minor = 0;
	card->gendisk->fops = &rsxx_fops;
	card->gendisk->private_data = card;
	card->gendisk->queue = card->queue;

	return 0;
}

void rsxx_destroy_dev(struct rsxx_cardinfo *card)
{
	if (!enable_blkdev)
		return;

	put_disk(card->gendisk);
	card->gendisk = NULL;

	blk_cleanup_queue(card->queue);
342
	card->queue->queuedata = NULL;
343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360
	unregister_blkdev(card->major, DRIVER_NAME);
}

int rsxx_dev_init(void)
{
	bio_meta_pool = KMEM_CACHE(rsxx_bio_meta, SLAB_HWCACHE_ALIGN);
	if (!bio_meta_pool)
		return -ENOMEM;

	return 0;
}

void rsxx_dev_cleanup(void)
{
	kmem_cache_destroy(bio_meta_pool);
}