[PATCH 4/8] LPDDR Command set driver

Alexey Korolev akorolev at infradead.org
Tue Dec 16 13:20:03 EST 2008


Driver which handles device command operation.
Details on device operations are available here:
http://www.numonyx.com/Documents/Datasheets/DS-315768_Velocity-Discrete.pdf

Signed-off-by: Alexey Korolev <akorolev at infradead.org>
Acked-by: Jared Hulbert <jaredeh at gmail.com>
---

diff --git a/drivers/mtd/lpddr/lpddr_cmds.c b/drivers/mtd/lpddr/lpddr_cmds.c
new file mode 100644
index 0000000..6236102
--- /dev/null
+++ b/drivers/mtd/lpddr/lpddr_cmds.c
@@ -0,0 +1,796 @@
+/*
+ * LPDDR flash memory device operations. This module provides read, write,
+ * erase, lock/unlock support for LPDDR flash memories
+ * (C) 2008 Korolev Alexey <akorolev at infradead.org>
+ * (C) 2008 Vasiliy Leonenko <vasiliy.leonenko at gmail.com>
+ * Many thanks to Roman Borisov for intial enabling
+ * 
+ * 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., 51 Franklin Street, Fifth Floor, Boston, MA 
+ * 02110-1301, USA.
+ * TODO:
+ * Implement VPP management
+ * Implement XIP support
+ * Implement OTP support
+ */
+#include <linux/mtd/pfow.h>
+#include <linux/mtd/qinfo.h>
+
+static int lpddr_read(struct mtd_info *mtd, loff_t adr, size_t len,
+					size_t *retlen, u_char *buf);
+static int lpddr_write_buffers(struct mtd_info *mtd, loff_t to,
+				size_t len, size_t *retlen, const u_char *buf);
+static int lpddr_writev(struct mtd_info *mtd, const struct kvec *vecs,
+				unsigned long count, loff_t to, size_t *retlen);
+static int lpddr_erase(struct mtd_info *mtd, struct erase_info *instr);
+static int lpddr_lock(struct mtd_info *mtd, loff_t ofs, uint64_t len);
+static int lpddr_unlock(struct mtd_info *mtd, loff_t ofs, uint64_t len);
+static int lpddr_point(struct mtd_info *mtd, loff_t adr, size_t len,
+			size_t *retlen, void **mtdbuf, resource_size_t *phys);
+static void lpddr_unpoint(struct mtd_info *mtd, loff_t adr, size_t len);
+static int get_chip(struct map_info *map, struct flchip *chip, int mode);
+static int chip_ready(struct map_info *map, struct flchip *chip, int mode);
+static void put_chip(struct map_info *map, struct flchip *chip);
+
+struct mtd_info *lpddr_cmdset(struct map_info *map)
+{
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	struct flchip_shared *shared;
+	struct flchip *chip;
+	struct mtd_info *mtd;
+	int numchips;
+	int i, j;
+
+	mtd = kzalloc(sizeof(*mtd), GFP_KERNEL);
+	if (!mtd) {
+		printk(KERN_ERR "Failed to allocate memory for MTD device\n");
+		return NULL;
+	}
+	mtd->priv = map;
+	mtd->type = MTD_NORFLASH;
+
+	/* Fill in the default mtd operations */
+	mtd->read = lpddr_read;
+	mtd->type = MTD_NORFLASH;
+	mtd->flags = MTD_CAP_NORFLASH;
+	mtd->flags &= ~MTD_BIT_WRITEABLE;
+	mtd->erase = lpddr_erase;
+	mtd->write = lpddr_write_buffers;
+	mtd->writev = lpddr_writev;
+	mtd->read_oob = NULL;
+	mtd->write_oob = NULL;
+	mtd->sync = NULL;
+	mtd->lock = lpddr_lock;
+	mtd->unlock = lpddr_unlock;
+	mtd->suspend = NULL;
+	mtd->resume = NULL;
+	if (map_is_linear(map)) {
+		mtd->point = lpddr_point;
+		mtd->unpoint = lpddr_unpoint;
+	}
+	mtd->block_isbad = NULL;
+	mtd->block_markbad = NULL;
+	mtd->size = 1 << lpddr->qinfo->DevSizeShift;
+	mtd->erasesize = 1 << lpddr->qinfo->UniformBlockSizeShift;
+	mtd->writesize = 1 << lpddr->qinfo->BufSizeShift;
+
+	shared = kmalloc(sizeof(struct flchip_shared) * lpddr->numchips,
+						GFP_KERNEL);
+	if (!shared) {
+		kfree(lpddr);
+		kfree(mtd);
+		return NULL;
+	}
+
+	chip = &lpddr->chips[0];
+	numchips = lpddr->numchips / lpddr->qinfo->HWPartsNum;
+	for (i = 0; i < numchips; i++) {
+		shared[i].writing = shared[i].erasing = NULL;
+		spin_lock_init(&shared[i].lock);
+		for (j = 0; j < lpddr->qinfo->HWPartsNum; j++) {
+			*chip = lpddr->chips[i];
+			chip->start += j << lpddr->chipshift;
+			chip->oldstate = chip->state = FL_READY;
+			chip->priv = &shared[i];
+			/* those should be reset too since
+			   they create memory references. */
+			init_waitqueue_head(&chip->wq);
+			spin_lock_init(&chip->_spinlock);
+			chip->mutex = &chip->_spinlock;
+			chip++;
+		}
+	}
+
+	return mtd;
+}
+EXPORT_SYMBOL(lpddr_cmdset);
+
+static int wait_for_ready(struct map_info *map, struct flchip *chip,
+		unsigned int chip_op_time)
+{
+	unsigned int timeo, reset_timeo, sleep_time;
+	unsigned int dsr;
+	flstate_t chip_state = chip->state;
+	int ret = 0;
+
+	/* set our timeout to 8 times the expected delay */
+	timeo = chip_op_time * 8;
+	if (!timeo)
+		timeo = 500000;
+	reset_timeo = timeo;
+	sleep_time = chip_op_time / 2;
+
+	for (;;) {
+		dsr = CMDVAL(map_read(map, map->pfow_base + PFOW_DSR));
+		if (dsr & DSR_READY_STATUS)
+			break;
+		if (!timeo) {
+			printk(KERN_ERR "%s: Flash timeout error state %d \n",
+							map->name, chip_state);
+			ret = -ETIME;
+			break;
+		}
+
+		/* OK Still waiting. Drop the lock, wait a while and retry. */
+		spin_unlock(chip->mutex);
+		if (sleep_time >= 1000000/HZ) {
+			/*
+			 * Half of the normal delay still remaining
+			 * can be performed with a sleeping delay instead
+			 * of busy waiting.
+			 */
+			msleep(sleep_time/1000);
+			timeo -= sleep_time;
+			sleep_time = 1000000/HZ;
+		} else {
+			udelay(1);
+			cond_resched();
+			timeo--;
+		}
+		spin_lock(chip->mutex);
+
+		while (chip->state != chip_state) {
+			/* Someone's suspended the operation: sleep */
+			DECLARE_WAITQUEUE(wait, current);
+			set_current_state(TASK_UNINTERRUPTIBLE);
+			add_wait_queue(&chip->wq, &wait);
+			spin_unlock(chip->mutex);
+			schedule();
+			remove_wait_queue(&chip->wq, &wait);
+			spin_lock(chip->mutex);
+		}
+		if (chip->erase_suspended || chip->write_suspended)  {
+			/* Suspend has occured while sleep: reset timeout */
+			timeo = reset_timeo;
+			chip->erase_suspended = chip->write_suspended = 0;
+		}
+	}
+	/* check status for errors */
+	if (dsr & DSR_ERR) {
+		/* Clear DSR*/
+		map_write(map, CMD(~(DSR_ERR)), map->pfow_base + PFOW_DSR);
+		printk(KERN_WARNING"%s: Bad status on wait: 0x%x \n",
+				map->name, dsr);
+		print_drs_error(dsr);
+		ret = -EIO;
+	}
+	chip->state = FL_READY;
+	return ret;
+}
+
+static int get_chip(struct map_info *map, struct flchip *chip, int mode)
+{
+	int ret;
+	DECLARE_WAITQUEUE(wait, current);
+
+ retry:
+	if (chip->priv && (mode == FL_WRITING || mode == FL_ERASING)
+		&& chip->state != FL_SYNCING) {
+		/*
+		 * OK. We have possibility for contension on the write/erase
+		 * operations which are global to the real chip and not per
+		 * partition.  So let's fight it over in the partition which
+		 * currently has authority on the operation.
+		 *
+		 * The rules are as follows:
+		 *
+		 * - any write operation must own shared->writing.
+		 *
+		 * - any erase operation must own _both_ shared->writing and
+		 *   shared->erasing.
+		 *
+		 * - contension arbitration is handled in the owner's context.
+		 *
+		 * The 'shared' struct can be read and/or written only when
+		 * its lock is taken.
+		 */
+		struct flchip_shared *shared = chip->priv;
+		struct flchip *contender;
+		spin_lock(&shared->lock);
+		contender = shared->writing;
+		if (contender && contender != chip) {
+			/*
+			 * The engine to perform desired operation on this
+			 * partition is already in use by someone else.
+			 * Let's fight over it in the context of the chip
+			 * currently using it.  If it is possible to suspend,
+			 * that other partition will do just that, otherwise
+			 * it'll happily send us to sleep.  In any case, when
+			 * get_chip returns success we're clear to go ahead.
+			 */
+			ret = spin_trylock(contender->mutex);
+			spin_unlock(&shared->lock);
+			if (!ret)
+				goto retry;
+			spin_unlock(chip->mutex);
+			ret = chip_ready(map, contender, mode);
+			spin_lock(chip->mutex);
+
+			if (ret == -EAGAIN) {
+				spin_unlock(contender->mutex);
+				goto retry;
+			}
+			if (ret) {
+				spin_unlock(contender->mutex);
+				return ret;
+			}
+			spin_lock(&shared->lock);
+
+			/* We should not own chip if it is already in FL_SYNCING
+			 * state. Put contender and retry. */
+			if (chip->state == FL_SYNCING) {
+				put_chip(map, contender);
+				spin_unlock(contender->mutex);
+				goto retry;
+			}
+			spin_unlock(contender->mutex);
+		}
+
+		/* Check if we have suspended erase on this chip.
+		   Must sleep in such a case. */
+		if (mode == FL_ERASING && shared->erasing
+		    && shared->erasing->oldstate == FL_ERASING) {
+			spin_unlock(&shared->lock);
+			set_current_state(TASK_UNINTERRUPTIBLE);
+			add_wait_queue(&chip->wq, &wait);
+			spin_unlock(chip->mutex);
+			schedule();
+			remove_wait_queue(&chip->wq, &wait);
+			spin_lock(chip->mutex);
+			goto retry;
+		}
+
+		/* We now own it */
+		shared->writing = chip;
+		if (mode == FL_ERASING)
+			shared->erasing = chip;
+		spin_unlock(&shared->lock);
+	}
+
+	ret = chip_ready(map, chip, mode);
+	if (ret == -EAGAIN)
+		goto retry;
+
+	return ret;
+}
+
+static int chip_ready(struct map_info *map, struct flchip *chip, int mode)
+{
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int ret = 0;
+	DECLARE_WAITQUEUE(wait, current);
+
+	/* Prevent setting state FL_SYNCING for chip in suspended state. */
+	if (FL_SYNCING == mode && FL_READY != chip->oldstate)
+		goto sleep;
+
+	switch (chip->state) {
+	case FL_READY:
+	case FL_JEDEC_QUERY:
+		return 0;
+
+	case FL_ERASING:
+		if (!lpddr->qinfo->SuspEraseSupp ||
+			!(mode == FL_READY || mode == FL_POINT))
+			goto sleep;
+
+		map_write(map, CMD(LPDDR_SUSPEND),
+			map->pfow_base + PFOW_PROGRAM_ERASE_SUSPEND);
+		chip->oldstate = FL_ERASING;
+		chip->state = FL_ERASE_SUSPENDING;
+		ret = wait_for_ready(map, chip, 0);
+		if (ret) {
+			/* Oops. something got wrong. */
+			/* Resume and pretend we weren't here.  */
+			map_write(map, CMD(LPDDR_RESUME),
+				map->pfow_base + PFOW_COMMAND_CODE);
+			map_write(map, CMD(LPDDR_START_EXECUTION),
+				map->pfow_base + PFOW_COMMAND_EXECUTE);
+			chip->state = FL_ERASING;
+			chip->oldstate = FL_READY;
+			printk(KERN_ERR "%s: suspend operation failed."
+					"State may be wrong \n", map->name);
+			return -EIO;
+		}
+		chip->erase_suspended = 1;
+		chip->state = FL_READY;
+		return 0;
+		/* Erase suspend */
+	case FL_POINT:
+		/* Only if there's no operation suspended... */
+		if (mode == FL_READY && chip->oldstate == FL_READY)
+			return 0;
+
+	default:
+sleep:
+		set_current_state(TASK_UNINTERRUPTIBLE);
+		add_wait_queue(&chip->wq, &wait);
+		spin_unlock(chip->mutex);
+		schedule();
+		remove_wait_queue(&chip->wq, &wait);
+		spin_lock(chip->mutex);
+		return -EAGAIN;
+	}
+}
+
+static void put_chip(struct map_info *map, struct flchip *chip)
+{
+	if (chip->priv) {
+		struct flchip_shared *shared = chip->priv;
+		spin_lock(&shared->lock);
+		if (shared->writing == chip && chip->oldstate == FL_READY) {
+			/* We own the ability to write, but we're done */
+			shared->writing = shared->erasing;
+			if (shared->writing && shared->writing != chip) {
+				/* give back the ownership */
+				struct flchip *loaner = shared->writing;
+				spin_lock(loaner->mutex);
+				spin_unlock(&shared->lock);
+				spin_unlock(chip->mutex);
+				put_chip(map, loaner);
+				spin_lock(chip->mutex);
+				spin_unlock(loaner->mutex);
+				wake_up(&chip->wq);
+				return;
+			}
+			shared->erasing = NULL;
+			shared->writing = NULL;
+		} else if (shared->erasing == chip && shared->writing != chip) {
+			/*
+			 * We own the ability to erase without the ability
+			 * to write, which means the erase was suspended
+			 * and some other partition is currently writing.
+			 * Don't let the switch below mess things up since
+			 * we don't have ownership to resume anything.
+			 */
+			spin_unlock(&shared->lock);
+			wake_up(&chip->wq);
+			return;
+		}
+		spin_unlock(&shared->lock);
+	}
+
+	switch (chip->oldstate) {
+	case FL_ERASING:
+		chip->state = chip->oldstate;
+		map_write(map, CMD(LPDDR_RESUME),
+				map->pfow_base + PFOW_COMMAND_CODE);
+		map_write(map, CMD(LPDDR_START_EXECUTION),
+				map->pfow_base + PFOW_COMMAND_EXECUTE);
+		chip->oldstate = FL_READY;
+		chip->state = FL_ERASING;
+		break;
+	case FL_READY:
+		break;
+	default:
+		printk(KERN_ERR "%s: put_chip() called with oldstate %d!\n",
+				map->name, chip->oldstate);
+	}
+	wake_up(&chip->wq);
+}
+
+int do_write_buffer(struct map_info *map, struct flchip *chip,
+			unsigned long adr, const struct kvec **pvec,
+			unsigned long *pvec_seek, int len)
+{
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	map_word datum;
+	int ret, wbufsize, word_gap, words;
+	const struct kvec *vec;
+	unsigned long vec_seek;
+	unsigned long prog_buf_ofs;
+
+	wbufsize = 1 << lpddr->qinfo->BufSizeShift;
+
+	spin_lock(chip->mutex);
+	ret = get_chip(map, chip, FL_WRITING);
+	if (ret) {
+		spin_unlock(chip->mutex);
+		return ret;
+	}
+	/* Figure out the number of words to write */
+	word_gap = (-adr & (map_bankwidth(map)-1));
+	words = (len - word_gap + map_bankwidth(map) - 1) / map_bankwidth(map);
+	if (!word_gap) {
+		words--;
+	} else {
+		word_gap = map_bankwidth(map) - word_gap;
+		adr -= word_gap;
+		datum = map_word_ff(map);
+	}
+	/* Write data */
+	/* Get the program buffer offset from PFOW register data first*/
+	prog_buf_ofs = map->pfow_base + CMDVAL(map_read(map,
+				map->pfow_base + PFOW_PROGRAM_BUFFER_OFFSET));
+	vec = *pvec;
+	vec_seek = *pvec_seek;
+	do {
+		int n = map_bankwidth(map) - word_gap;
+
+		if (n > vec->iov_len - vec_seek)
+			n = vec->iov_len - vec_seek;
+		if (n > len)
+			n = len;
+
+		if (!word_gap && (len < map_bankwidth(map)))
+			datum = map_word_ff(map);
+
+		datum = map_word_load_partial(map, datum,
+				vec->iov_base + vec_seek, word_gap, n);
+
+		len -= n;
+		word_gap += n;
+		if (!len || word_gap == map_bankwidth(map)) {
+			map_write(map, datum, prog_buf_ofs);
+			prog_buf_ofs += map_bankwidth(map);
+			word_gap = 0;
+		}
+
+		vec_seek += n;
+		if (vec_seek == vec->iov_len) {
+			vec++;
+			vec_seek = 0;
+		}
+	} while (len);
+	*pvec = vec;
+	*pvec_seek = vec_seek;
+
+	/* GO GO GO */
+	send_pfow_command(map, LPDDR_BUFF_PROGRAM, adr, wbufsize, NULL);
+	chip->state = FL_WRITING;
+	ret = wait_for_ready(map, chip, (1<<lpddr->qinfo->ProgBufferTime));
+	if (ret)	{
+		printk(KERN_WARNING"%s Buffer program error: %d at %lx; \n",
+			map->name, ret, adr);
+		goto out;
+	}
+
+ out:	put_chip(map, chip);
+	spin_unlock(chip->mutex);
+	return ret;
+}
+
+int do_erase_oneblock(struct mtd_info *mtd, loff_t adr)
+{
+	struct map_info *map = mtd->priv;
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int chipnum = adr >> lpddr->chipshift;
+	struct flchip *chip = &lpddr->chips[chipnum];
+	int ret;
+
+	spin_lock(chip->mutex);
+	ret = get_chip(map, chip, FL_ERASING);
+	if (ret) {
+		spin_unlock(chip->mutex);
+		return ret;
+	}
+	send_pfow_command(map, LPDDR_BLOCK_ERASE, adr, 0, NULL);
+	chip->state = FL_ERASING;
+	ret = wait_for_ready(map, chip, (1<<lpddr->qinfo->BlockEraseTime)*1000);
+	if (ret) {
+		printk(KERN_WARNING"%s Erase block error %d at : %llx\n",
+			map->name, ret, adr);
+		goto out;
+	}
+ out:	put_chip(map, chip);
+	spin_unlock(chip->mutex);
+	return ret;
+}
+
+static int lpddr_read(struct mtd_info *mtd, loff_t adr, size_t len,
+			size_t *retlen, u_char *buf)
+{
+	struct map_info *map = mtd->priv;
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int chipnum = adr >> lpddr->chipshift;
+	struct flchip *chip = &lpddr->chips[chipnum];
+	int ret = 0;
+
+	spin_lock(chip->mutex);
+	ret = get_chip(map, chip, FL_READY);
+	if (ret) {
+		spin_unlock(chip->mutex);
+		return ret;
+	}
+
+	map_copy_from(map, buf, adr, len);
+	*retlen = len;
+
+	put_chip(map, chip);
+	spin_unlock(chip->mutex);
+	return ret;
+}
+
+static int lpddr_point(struct mtd_info *mtd, loff_t adr, size_t len,
+			size_t *retlen, void **mtdbuf, resource_size_t *phys)
+{
+	struct map_info *map = mtd->priv;
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int chipnum = adr >> lpddr->chipshift;
+	unsigned long ofs, last_end = 0;
+	struct flchip *chip = &lpddr->chips[chipnum];
+	int ret = 0;
+
+	if (!map->virt || (adr + len > mtd->size))
+		return -EINVAL;
+
+	/* ofs: offset within the first chip that the first read should start */
+	ofs = adr - (chipnum << lpddr->chipshift);
+
+	*mtdbuf = (void *)map->virt + chip->start + ofs;
+	*retlen = 0;
+
+	while (len) {
+		unsigned long thislen;
+
+		if (chipnum >= lpddr->numchips)
+			break;
+
+		/* We cannot point across chips that are virtually disjoint */
+		if (!last_end)
+			last_end = chip->start;
+		else if (chip->start != last_end)
+			break;
+
+		if ((len + ofs - 1) >> lpddr->chipshift)
+			thislen = (1<<lpddr->chipshift) - ofs;
+		else
+			thislen = len;
+		/* get the chip */
+		spin_lock(chip->mutex);
+		ret = get_chip(map, chip, FL_POINT);
+		spin_unlock(chip->mutex);
+		if (ret)
+			break;
+
+		chip->state = FL_POINT;
+		chip->ref_point_counter++;
+		*retlen += thislen;
+		len -= thislen;
+
+		ofs = 0;
+		last_end += 1 << lpddr->chipshift;
+		chipnum++;
+		chip = &lpddr->chips[chipnum];
+	}
+	return 0;
+}
+
+static void lpddr_unpoint (struct mtd_info *mtd, loff_t adr, size_t len)
+{
+	struct map_info *map = mtd->priv;
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int chipnum = adr >> lpddr->chipshift;
+	unsigned long ofs;
+
+	/* ofs: offset within the first chip that the first read should start */
+	ofs = adr - (chipnum << lpddr->chipshift);
+
+	while (len) {
+		unsigned long thislen;
+		struct flchip *chip;
+
+		chip = &lpddr->chips[chipnum];
+		if (chipnum >= lpddr->numchips)
+			break;
+
+		if ((len + ofs - 1) >> lpddr->chipshift)
+			thislen = (1<<lpddr->chipshift) - ofs;
+		else
+			thislen = len;
+
+		spin_lock(chip->mutex);
+		if (chip->state == FL_POINT) {
+			chip->ref_point_counter--;
+			if (chip->ref_point_counter == 0)
+				chip->state = FL_READY;
+		} else
+			printk(KERN_WARNING "%s: Warning: unpoint called on non"
+					"pointed region\n", map->name);
+
+		put_chip(map, chip);
+		spin_unlock(chip->mutex);
+
+		len -= thislen;
+		ofs = 0;
+		chipnum++;
+	}
+}
+
+static int lpddr_write_buffers(struct mtd_info *mtd, loff_t to, size_t len,
+				size_t *retlen, const u_char *buf)
+{
+	struct kvec vec;
+
+	vec.iov_base = (void *) buf;
+	vec.iov_len = len;
+
+	return lpddr_writev(mtd, &vec, 1, to, retlen);
+}
+
+
+static int lpddr_writev(struct mtd_info *mtd, const struct kvec *vecs,
+				unsigned long count, loff_t to, size_t *retlen)
+{
+	struct map_info *map = mtd->priv;
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int ret = 0;
+	int chipnum;
+	unsigned long ofs, vec_seek, i;
+	int wbufsize = 1 << lpddr->qinfo->BufSizeShift;
+
+	size_t len = 0;
+
+	for (i = 0; i < count; i++)
+		len += vecs[i].iov_len;
+
+	*retlen = 0;
+	if (!len)
+		return 0;
+
+	chipnum = to >> lpddr->chipshift;
+
+	ofs = to;
+	vec_seek = 0;
+
+	do {
+		/* We must not cross write block boundaries */
+		int size = wbufsize - (ofs & (wbufsize-1));
+
+		if (size > len)
+			size = len;
+
+		ret = do_write_buffer(map, &lpddr->chips[chipnum],
+					  ofs, &vecs, &vec_seek, size);
+		if (ret)
+			return ret;
+
+		ofs += size;
+		(*retlen) += size;
+		len -= size;
+
+		/* Be nice and reschedule with the chip in a usable
+		 * state for other processes */
+		cond_resched();
+
+	} while (len);
+
+	return 0;
+}
+
+static int lpddr_erase(struct mtd_info *mtd, struct erase_info *instr)
+{
+	unsigned long ofs, len;
+	int ret;
+	struct map_info *map = mtd->priv;
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int size = 1 << lpddr->qinfo->UniformBlockSizeShift;
+
+	ofs = instr->addr;
+	len = instr->len;
+
+	if (ofs > mtd->size || (len + ofs) > mtd->size)
+		return -EINVAL;
+
+	while (len > 0) {
+		ret = do_erase_oneblock(mtd, ofs);
+		if (ret)
+			return ret;
+		ofs += size;
+		len -= size;
+	}
+	instr->state = MTD_ERASE_DONE;
+	mtd_erase_callback(instr);
+
+	return 0;
+}
+
+#define DO_XXLOCK_LOCK		1
+#define DO_XXLOCK_UNLOCK	2
+int do_xxlock(struct mtd_info *mtd, loff_t adr, uint32_t len, int thunk)
+{
+	int ret = 0;
+	struct map_info *map = mtd->priv;
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int chipnum = adr >> lpddr->chipshift;
+	struct flchip *chip = &lpddr->chips[chipnum];
+
+	spin_lock(chip->mutex);
+	ret = get_chip(map, chip, FL_LOCKING);
+	if (ret) {
+		spin_unlock(chip->mutex);
+		return ret;
+	}
+
+	if (thunk == DO_XXLOCK_LOCK) {
+		send_pfow_command(map, LPDDR_LOCK_BLOCK, adr, adr + len, NULL);
+		chip->state = FL_LOCKING;
+	} else if (thunk == DO_XXLOCK_UNLOCK) {
+		send_pfow_command(map, LPDDR_UNLOCK_BLOCK, adr, adr + len, NULL);
+		chip->state = FL_UNLOCKING;
+	} else
+		BUG();
+
+	ret = wait_for_ready(map, chip, 1);
+	if (ret)	{
+		printk(KERN_ERR "%s: block unlock error status %d \n",
+				map->name, ret);
+		goto out;
+	}
+out:	put_chip(map, chip);
+	spin_unlock(chip->mutex);
+	return ret;
+}
+
+static int lpddr_lock(struct mtd_info *mtd, loff_t ofs, uint64_t len)
+{
+	return do_xxlock(mtd, ofs, len, DO_XXLOCK_LOCK);
+}
+
+static int lpddr_unlock(struct mtd_info *mtd, loff_t ofs, uint64_t len)
+{
+	return do_xxlock(mtd, ofs, len, DO_XXLOCK_UNLOCK);
+}
+
+int word_program(struct map_info *map, loff_t adr, uint32_t curval)
+{
+    int ret;
+	struct lpddr_private *lpddr = map->fldrv_priv;
+	int chipnum = adr >> lpddr->chipshift;
+	struct flchip *chip = &lpddr->chips[chipnum];
+
+	spin_lock(chip->mutex);
+	ret = get_chip(map, chip, FL_WRITING);
+	if (ret) {
+		spin_unlock(chip->mutex);
+		return ret;
+	}
+
+	send_pfow_command(map, LPDDR_WORD_PROGRAM, adr, 0x00, (map_word *)&curval);
+
+	ret = wait_for_ready(map, chip, (1<<lpddr->qinfo->SingleWordProgTime));
+	if (ret)	{
+		printk(KERN_WARNING"%s word_program error at: %llx; val: %x\n",
+			map->name, adr, curval);
+		goto out;
+	}
+
+out:	put_chip(map, chip);
+	spin_unlock(chip->mutex);
+	return ret;
+}
+
+MODULE_LICENSE("GPL");
+MODULE_AUTHOR("Alexey Korolev <akorolev at infradead.org>");
+MODULE_DESCRIPTION("MTD driver for LPDDR flash chips");




More information about the linux-mtd mailing list