Skip to contentSkip to conversation

block ioctl cleanups

PATCH In review Christoph Hellwig 6 participants 21 messages Oct 31, 2020 – Nov 3, 2020

Subsystems (8)
lore
Share & tools
lore
#0 lore raw

Hi Jens,

this series has a bunch of cleanups for the block layer ioctl code.

Diffstat:

 block/genhd.c                     |    7 ----
 block/ioctl.c                     |   62 ++++++--------------------------------
 drivers/block/loop.c              |    2 -
 drivers/block/mtip32xx/mtip32xx.c |   11 +-----
 drivers/block/pktcdvd.c           |    6 ++-
 drivers/block/rbd.c               |   41 ++-----------------------
 drivers/md/bcache/request.c       |    5 +--
 drivers/md/dm.c                   |    5 ++-
 drivers/md/md.c                   |   62 +++++++++++++++++++-------------------
 drivers/mtd/mtd_blkdevs.c         |   28 -----------------
 drivers/s390/block/dasd.c         |    1 
 drivers/s390/block/dasd_int.h     |    3 +
 drivers/s390/block/dasd_ioctl.c   |   27 +++++-----------
 include/linux/blkdev.h            |    3 -
 include/linux/genhd.h             |    1 
 15 files changed, 73 insertions(+), 191 deletions(-)
#1 lore raw

PATCH 1/11 mtd_blkdevs: don't override BLKFLSBUF

BLKFLSBUF does not actually send a flush command to the device, but teard down buffer cache structures. Remove the mtd_blkdevs implementation and just use the default semantics instead.

---
 drivers/mtd/mtd_blkdevs.c | 28 ----------------------------
 1 file changed, 28 deletions(-)
drivers/mtd/mtd_blkdevs.c +0 -28
index 0c05f77f9b216e..fb8e12d590a13a 100644
@@ -298,38 +298,10 @@ static int blktrans_getgeo(struct block_device *bdev, struct hd_geometry *geo)
 	return ret;
 }
 
-static int blktrans_ioctl(struct block_device *bdev, fmode_t mode,
-			      unsigned int cmd, unsigned long arg)
-{
-	struct mtd_blktrans_dev *dev = blktrans_dev_get(bdev->bd_disk);
-	int ret = -ENXIO;
-
-	if (!dev)
-		return ret;
-
-	mutex_lock(&dev->lock);
-
-	if (!dev->mtd)
-		goto unlock;
-
-	switch (cmd) {
-	case BLKFLSBUF:
-		ret = dev->tr->flush ? dev->tr->flush(dev) : 0;
-		break;
-	default:
-		ret = -ENOTTY;
-	}
-unlock:
-	mutex_unlock(&dev->lock);
-	blktrans_dev_put(dev);
-	return ret;
-}
-
 static const struct block_device_operations mtd_block_ops = {
 	.owner		= THIS_MODULE,
 	.open		= blktrans_open,
 	.release	= blktrans_release,
-	.ioctl		= blktrans_ioctl,
 	.getgeo		= blktrans_getgeo,
 };
 
Signature · 1 line
2.28.0
#3 lore raw

Re: [PATCH 01/11] mtd_blkdevs: don't override BLKFLSBUF

In reply to Christoph Hellwig
Quoted 2 lines · Christoph Hellwig

BLKFLSBUF does not actually send a flush command to the device, but teard down buffer cache structures. Remove the mtd_blkdevs

^^^^^ ?tears?

Quoted 4 lines

implementation and just use the default semantics instead.

---

Cheers,
Wol

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

#4 lore raw

PATCH 2/11 mtip32xx: return -ENOTTY for all unhanled ioctls

In reply to Christoph Hellwig

-ENOTTY is the convention for "driver does not support this ioctl". Use it properly in mtip32xx instead of the bogys -EINVAL.

---
 drivers/block/mtip32xx/mtip32xx.c | 11 ++---------
 1 file changed, 2 insertions(+), 9 deletions(-)
drivers/block/mtip32xx/mtip32xx.c +2 -9
index 153e2cdecb4d40..893be95eceb34e 100644
@@ -2029,7 +2029,7 @@ static int mtip_hw_ioctl(struct driver_data *dd, unsigned int cmd,
 	}
 
 	default:
-		return -EINVAL;
+		return -ENOTTY;
 	}
 	return 0;
 }
@@ -3215,12 +3215,7 @@ static int mtip_block_ioctl(struct block_device *dev,
 	if (unlikely(test_bit(MTIP_DDF_REMOVE_PENDING_BIT, &dd->dd_flag)))
 		return -ENOTTY;
 
-	switch (cmd) {
-	case BLKFLSBUF:
-		return -ENOTTY;
-	default:
-		return mtip_hw_ioctl(dd, cmd, arg);
-	}
+	return mtip_hw_ioctl(dd, cmd, arg);
 }
 
 #ifdef CONFIG_COMPAT
@@ -3254,8 +3249,6 @@ static int mtip_block_compat_ioctl(struct block_device *dev,
 		return -ENOTTY;
 
 	switch (cmd) {
-	case BLKFLSBUF:
-		return -ENOTTY;
 	case HDIO_DRIVE_TASKFILE: {
 		struct mtip_compat_ide_task_request_s __user *compat_req_task;
 		ide_task_request_t req_task;
Signature · 1 line
2.28.0
maintainer
#5 lore raw

Re: [PATCH 02/11] mtip32xx: return -ENOTTY for all unhanled ioctls

Quoted 2 lines · Christoph Hellwig

-ENOTTY is the convention for "driver does not support this ioctl". Use it properly in mtip32xx instead of the bogys -EINVAL.

While that's certainly true, there is a risk in making a change like this years after the fact. Not that I expect there are any mtip32xx users left at this point, but...

Signature · 1 line
Jens Axboe
#6 lore raw

Re: [PATCH 02/11] mtip32xx: return -ENOTTY for all unhanled ioctls

Quoted 7 lines · Jens Axboe
Quoted 2 lines · Christoph Hellwig

-ENOTTY is the convention for "driver does not support this ioctl". Use it properly in mtip32xx instead of the bogys -EINVAL.

While that's certainly true, there is a risk in making a change like this years after the fact. Not that I expect there are any mtip32xx users left at this point, but...

-ENOTTY is what most drivers return. That being said we can keep the old behavior, so if you prepfer that I can respin to do that.

maintainer
#7 lore raw

Re: [PATCH 02/11] mtip32xx: return -ENOTTY for all unhanled ioctls

Quoted 11 lines · Christoph Hellwig
Quoted 7 lines · Jens Axboe
Quoted 2 lines · Christoph Hellwig

-ENOTTY is the convention for "driver does not support this ioctl". Use it properly in mtip32xx instead of the bogys -EINVAL.

While that's certainly true, there is a risk in making a change like this years after the fact. Not that I expect there are any mtip32xx users left at this point, but...

-ENOTTY is what most drivers return. That being said we can keep the old behavior, so if you prepfer that I can respin to do that.

Yeah I know that -ENOTTY is what they all should use (and most of course does), just saying that this change carries some risk. Given that mtip32xx can probably be retired in the not-too-distant future, I say we just keep the -EINVAL here.

Signature · 1 line
Jens Axboe
#8 lore raw

PATCH 3/11 block: don't call into the driver for BLKFLSBUF

In reply to Christoph Hellwig

BLKFLSBUF is entirely contained in the block core, and there is no good reason to give the driver a hook into processing it.

---
 block/ioctl.c | 7 -------
 1 file changed, 7 deletions(-)
block/ioctl.c +0 -7
index 3fbc382eb926d4..c6d8863f040945 100644
@@ -369,15 +369,8 @@ static inline int is_unrecognized_ioctl(int ret)
 static int blkdev_flushbuf(struct block_device *bdev, fmode_t mode,
 		unsigned cmd, unsigned long arg)
 {
-	int ret;
-
 	if (!capable(CAP_SYS_ADMIN))
 		return -EACCES;
-
-	ret = __blkdev_driver_ioctl(bdev, mode, cmd, arg);
-	if (!is_unrecognized_ioctl(ret))
-		return ret;
-
 	fsync_bdev(bdev);
 	invalidate_bdev(bdev);
 	return 0;
Signature · 1 line
2.28.0
#9 lore raw

PATCH 4/11 block: add a new set_read_only method

In reply to Christoph Hellwig

Add a new method to allow driver-specific callout when setting or clearing the block device read-only state. This allow to replace the cumbersome and error prone override of the whole ioctl implementation.

---
 block/ioctl.c          | 5 +++++
 include/linux/blkdev.h | 1 +
 2 files changed, 6 insertions(+)
block/ioctl.c +5 -0
index c6d8863f040945..a6fa16b9770593 100644
@@ -389,6 +389,11 @@ static int blkdev_roset(struct block_device *bdev, fmode_t mode,
 		return ret;
 	if (get_user(n, (int __user *)arg))
 		return -EFAULT;
+	if (bdev->bd_disk->fops->set_read_only) {
+		ret = bdev->bd_disk->fops->set_read_only(bdev, n);
+		if (ret)
+			return ret;
+	}
 	set_device_ro(bdev, n);
 	return 0;
 }
include/linux/blkdev.h +1 -0
index 639cae2c158b59..5c1ba8a8d2bc7e 100644
@@ -1850,6 +1850,7 @@ struct block_device_operations {
 	void (*unlock_native_capacity) (struct gendisk *);
 	int (*revalidate_disk) (struct gendisk *);
 	int (*getgeo)(struct block_device *, struct hd_geometry *);
+	int (*set_read_only)(struct block_device *bdev, bool ro);
 	/* this callback is with swap_lock and sometimes page table lock held */
 	void (*swap_slot_free_notify) (struct block_device *, unsigned long);
 	int (*report_zones)(struct gendisk *, sector_t sector,
Signature · 1 line
2.28.0
#10 lore raw

PATCH 5/11 rbd: implement ->set_read_only to hook into BLKROSET processing

In reply to Christoph Hellwig

Implement the ->set_read_only method instead of parsing the actual ioctl command.

---
 drivers/block/rbd.c | 41 ++++-------------------------------------
 1 file changed, 4 insertions(+), 37 deletions(-)
drivers/block/rbd.c +4 -37
index f84128abade319..37f8fc28004acb 100644
@@ -692,12 +692,9 @@ static void rbd_release(struct gendisk *disk, fmode_t mode)
 	put_device(&rbd_dev->dev);
 }
 
-static int rbd_ioctl_set_ro(struct rbd_device *rbd_dev, unsigned long arg)
+static int rbd_set_read_only(struct block_device *bdev, bool ro)
 {
-	int ro;
-
-	if (get_user(ro, (int __user *)arg))
-		return -EFAULT;
+	struct rbd_device *rbd_dev = bdev->bd_disk->private_data;
 
 	/*
 	 * Both images mapped read-only and snapshots can't be marked
@@ -706,47 +703,17 @@ static int rbd_ioctl_set_ro(struct rbd_device *rbd_dev, unsigned long arg)
 	if (!ro) {
 		if (rbd_is_ro(rbd_dev))
 			return -EROFS;
-
 		rbd_assert(!rbd_is_snap(rbd_dev));
 	}
 
-	/* Let blkdev_roset() handle it */
-	return -ENOTTY;
-}
-
-static int rbd_ioctl(struct block_device *bdev, fmode_t mode,
-			unsigned int cmd, unsigned long arg)
-{
-	struct rbd_device *rbd_dev = bdev->bd_disk->private_data;
-	int ret;
-
-	switch (cmd) {
-	case BLKROSET:
-		ret = rbd_ioctl_set_ro(rbd_dev, arg);
-		break;
-	default:
-		ret = -ENOTTY;
-	}
-
-	return ret;
-}
-
-#ifdef CONFIG_COMPAT
-static int rbd_compat_ioctl(struct block_device *bdev, fmode_t mode,
-				unsigned int cmd, unsigned long arg)
-{
-	return rbd_ioctl(bdev, mode, cmd, arg);
+	return 0;
 }
-#endif /* CONFIG_COMPAT */
 
 static const struct block_device_operations rbd_bd_ops = {
 	.owner			= THIS_MODULE,
 	.open			= rbd_open,
 	.release		= rbd_release,
-	.ioctl			= rbd_ioctl,
-#ifdef CONFIG_COMPAT
-	.compat_ioctl		= rbd_compat_ioctl,
-#endif
+	.set_read_only		= rbd_set_read_only,
 };
 
 /*
Signature · 1 line
2.28.0
maintainer Acked-by on v1 5/11
#11 lore raw

Re: [PATCH 05/11] rbd: implement ->set_read_only to hook into BLKROSET processing

Quoted 29 lines · Christoph Hellwig

Implement the ->set_read_only method instead of parsing the actual ioctl command.

---
 drivers/block/rbd.c | 41 ++++-------------------------------------
 1 file changed, 4 insertions(+), 37 deletions(-)
drivers/block/rbd.c +2 -5
index f84128abade319..37f8fc28004acb 100644
@@ -692,12 +692,9 @@ static void rbd_release(struct gendisk *disk, fmode_t mode)
        put_device(&rbd_dev->dev);
 }
 
-static int rbd_ioctl_set_ro(struct rbd_device *rbd_dev, unsigned long arg)
+static int rbd_set_read_only(struct block_device *bdev, bool ro)
 {
-       int ro;
-
-       if (get_user(ro, (int __user *)arg))
-               return -EFAULT;
+       struct rbd_device *rbd_dev = bdev->bd_disk->private_data;
 
        /*
         * Both images mapped read-only and snapshots can't be marked

drivers/block/rbd.c static int rbd_ioctl_set_ro(struct rbd_device *rbd_dev, unsigned long arg) · line 706

Quoted 5 lines
drivers/block/rbd.c +0 -1
@@ -706,47 +703,17 @@ static int rbd_ioctl_set_ro(struct rbd_device *rbd_dev, unsigned long arg)
        if (!ro) {
                if (rbd_is_ro(rbd_dev))
                        return -EROFS;
-
                rbd_assert(!rbd_is_snap(rbd_dev));

If you repost, please leave this empty line.

Quoted 45 lines
drivers/block/rbd.c +2 -32
        }
 
-       /* Let blkdev_roset() handle it */
-       return -ENOTTY;
-}
-
-static int rbd_ioctl(struct block_device *bdev, fmode_t mode,
-                       unsigned int cmd, unsigned long arg)
-{
-       struct rbd_device *rbd_dev = bdev->bd_disk->private_data;
-       int ret;
-
-       switch (cmd) {
-       case BLKROSET:
-               ret = rbd_ioctl_set_ro(rbd_dev, arg);
-               break;
-       default:
-               ret = -ENOTTY;
-       }
-
-       return ret;
-}
-
-#ifdef CONFIG_COMPAT
-static int rbd_compat_ioctl(struct block_device *bdev, fmode_t mode,
-                               unsigned int cmd, unsigned long arg)
-{
-       return rbd_ioctl(bdev, mode, cmd, arg);
+       return 0;
 }
-#endif /* CONFIG_COMPAT */
 
 static const struct block_device_operations rbd_bd_ops = {
        .owner                  = THIS_MODULE,
        .open                   = rbd_open,
        .release                = rbd_release,
-       .ioctl                  = rbd_ioctl,
-#ifdef CONFIG_COMPAT
-       .compat_ioctl           = rbd_compat_ioctl,
-#endif
+       .set_read_only          = rbd_set_read_only,
 };
 
 /*
--
Quoted 2 lines

2.28.0

With that nit,

Thanks,

Ilya

#12 lore raw

PATCH 6/11 md: implement ->set_read_only to hook into BLKROSET processing

In reply to Christoph Hellwig

Implement the ->set_read_only method instead of parsing the actual ioctl command.

---
 drivers/md/md.c | 62 ++++++++++++++++++++++++-------------------------
 1 file changed, 31 insertions(+), 31 deletions(-)
drivers/md/md.c +31 -31
index 98bac4f304ae26..8c027b81b78c0d 100644
@@ -7480,7 +7480,6 @@ static inline bool md_ioctl_valid(unsigned int cmd)
 {
 	switch (cmd) {
 	case ADD_NEW_DISK:
-	case BLKROSET:
 	case GET_ARRAY_INFO:
 	case GET_BITMAP_FILE:
 	case GET_DISK_INFO:
@@ -7507,7 +7506,6 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 	int err = 0;
 	void __user *argp = (void __user *)arg;
 	struct mddev *mddev = NULL;
-	int ro;
 	bool did_set_md_closing = false;
 
 	if (!md_ioctl_valid(cmd))
@@ -7687,35 +7685,6 @@ static int md_ioctl(struct block_device *bdev, fmode_t mode,
 			goto unlock;
 		}
 		break;
-
-	case BLKROSET:
-		if (get_user(ro, (int __user *)(arg))) {
-			err = -EFAULT;
-			goto unlock;
-		}
-		err = -EINVAL;
-
-		/* if the bdev is going readonly the value of mddev->ro
-		 * does not matter, no writes are coming
-		 */
-		if (ro)
-			goto unlock;
-
-		/* are we are already prepared for writes? */
-		if (mddev->ro != 1)
-			goto unlock;
-
-		/* transitioning to readauto need only happen for
-		 * arrays that call md_write_start
-		 */
-		if (mddev->pers) {
-			err = restart_array(mddev);
-			if (err == 0) {
-				mddev->ro = 2;
-				set_disk_ro(mddev->gendisk, 0);
-			}
-		}
-		goto unlock;
 	}
 
 	/*
@@ -7809,6 +7778,36 @@ static int md_compat_ioctl(struct block_device *bdev, fmode_t mode,
 }
 #endif /* CONFIG_COMPAT */
 
+static int md_set_read_only(struct block_device *bdev, bool ro)
+{
+	struct mddev *mddev = bdev->bd_disk->private_data;
+	int err;
+
+	err = mddev_lock(mddev);
+	if (err)
+		return err;
+
+	if (!mddev->raid_disks && !mddev->external) {
+		err = -ENODEV;
+		goto out_unlock;
+	}
+
+	/*
+	 * Transitioning to readauto need only happen for arrays that call
+	 * md_write_start and which are not ready for writes yet.
+	 */
+	if (!ro && mddev->ro == 1 && mddev->pers) {
+		err = restart_array(mddev);
+		if (err)
+			goto out_unlock;
+		mddev->ro = 2;
+	}
+
+out_unlock:
+	mddev_unlock(mddev);
+	return err;
+}
+
 static int md_open(struct block_device *bdev, fmode_t mode)
 {
 	/*
@@ -7886,6 +7885,7 @@ const struct block_device_operations md_fops =
 #endif
 	.getgeo		= md_getgeo,
 	.check_events	= md_check_events,
+	.set_read_only	= md_set_read_only,
 };
 
 static int md_thread(void *arg)
Signature · 1 line
2.28.0
#13 lore raw

Re: [PATCH 06/11] md: implement ->set_read_only to hook into BLKROSET processing

Quoted 1 line · Christoph Hellwig
(quoted diff) +0 -0
@@ -7809,6 +7778,36 @@ static int md_compat_ioctl(struct block_device *bdev, fmode_t mode,

[...]

drivers/md/md.c static int md_compat_ioctl(struct block_device *bdev, fmode_t mode, · line 7797

Quoted 2 lines
drivers/md/md.c +2 -0
+	 * Transitioning to readauto need only happen for arrays that call
+	 * md_write_start and which are not ready for writes yet.

I realise you're just moving the comment around but perhaps you could s/readauto/readonly/ while you're doing so?

Signature · 1 line
James
#14 lore raw

Re: [PATCH 06/11] md: implement ->set_read_only to hook into BLKROSET processing

Quoted 11 lines · James Troup
Quoted 1 line · Christoph Hellwig
(quoted diff) +0 -0
@@ -7809,6 +7778,36 @@ static int md_compat_ioctl(struct block_device *bdev, fmode_t mode,

[...]

drivers/md/md.c static int md_compat_ioctl(struct block_device *bdev, fmode_t mode, · line 7797

Quoted 2 lines
drivers/md/md.c +2 -0
+	 * Transitioning to readauto need only happen for arrays that call
+	 * md_write_start and which are not ready for writes yet.

I realise you're just moving the comment around but perhaps you could s/readauto/readonly/ while you're doing so?

readonly doesn't make sense here as we transition away from read-only. MD seems to have a read-auto mode, although it usually is spelled with the dash, so I'll switch this comment to use the more common spelling.

#15 lore raw

PATCH 7/11 dasd: implement ->set_read_only to hook into BLKROSET processing

In reply to Christoph Hellwig

Implement the ->set_read_only method instead of parsing the actual ioctl command.

---
 drivers/s390/block/dasd.c       |  1 +
 drivers/s390/block/dasd_int.h   |  3 ++-
 drivers/s390/block/dasd_ioctl.c | 27 +++++++++------------------
 3 files changed, 12 insertions(+), 19 deletions(-)
drivers/s390/block/dasd.c +1 -0
index eb17fea8075c6f..db24e04ee9781e 100644
@@ -3394,6 +3394,7 @@ dasd_device_operations = {
 	.ioctl		= dasd_ioctl,
 	.compat_ioctl	= dasd_ioctl,
 	.getgeo		= dasd_getgeo,
+	.set_read_only	= dasd_set_read_only,
 };
 
 /*******************************************************************************
drivers/s390/block/dasd_int.h +2 -1
index fa552f9f166671..c59a0d63b506e6 100644
@@ -844,7 +844,8 @@ int dasd_scan_partitions(struct dasd_block *);
 void dasd_destroy_partitions(struct dasd_block *);
 
 /* externals in dasd_ioctl.c */
-int  dasd_ioctl(struct block_device *, fmode_t, unsigned int, unsigned long);
+int dasd_ioctl(struct block_device *, fmode_t, unsigned int, unsigned long);
+int dasd_set_read_only(struct block_device *bdev, bool ro);
 
 /* externals in dasd_proc.c */
 int dasd_proc_init(void);
drivers/s390/block/dasd_ioctl.c +9 -18
index cb6427fb9f3d16..3359559517bfcf 100644
@@ -532,28 +532,22 @@ static int dasd_ioctl_information(struct dasd_block *block, void __user *argp,
 /*
  * Set read only
  */
-static int
-dasd_ioctl_set_ro(struct block_device *bdev, void __user *argp)
+int dasd_set_read_only(struct block_device *bdev, bool ro)
 {
 	struct dasd_device *base;
-	int intval, rc;
+	int rc;
 
-	if (!capable(CAP_SYS_ADMIN))
-		return -EACCES;
+	/* do not manipulate hardware state for partitions */
 	if (bdev_is_partition(bdev))
-		// ro setting is not allowed for partitions
-		return -EINVAL;
-	if (get_user(intval, (int __user *)argp))
-		return -EFAULT;
+		return 0;
+
 	base = dasd_device_from_gendisk(bdev->bd_disk);
 	if (!base)
 		return -ENODEV;
-	if (!intval && test_bit(DASD_FLAG_DEVICE_RO, &base->flags)) {
-		dasd_put_device(base);
-		return -EROFS;
-	}
-	set_disk_ro(bdev->bd_disk, intval);
-	rc = dasd_set_feature(base->cdev, DASD_FEATURE_READONLY, intval);
+	if (!ro && test_bit(DASD_FLAG_DEVICE_RO, &base->flags))
+		rc = -EROFS;
+	else
+		rc = dasd_set_feature(base->cdev, DASD_FEATURE_READONLY, ro);
 	dasd_put_device(base);
 	return rc;
 }
@@ -633,9 +627,6 @@ int dasd_ioctl(struct block_device *bdev, fmode_t mode,
 	case BIODASDPRRST:
 		rc = dasd_ioctl_reset_profile(block);
 		break;
-	case BLKROSET:
-		rc = dasd_ioctl_set_ro(bdev, argp);
-		break;
 	case DASDAPIVER:
 		rc = dasd_ioctl_api_version(argp);
 		break;
Signature · 1 line
2.28.0
#16 lore raw

PATCH 8/11 block: don't call into the driver for BLKROSET

In reply to Christoph Hellwig

Now that all drivers that want to hook into setting or clearing the read-only flag use the set_read_only method this code can be removed.

---
 block/ioctl.c | 23 -----------------------
 1 file changed, 23 deletions(-)
block/ioctl.c +0 -23
index a6fa16b9770593..96cb4544736468 100644
@@ -346,26 +346,6 @@ static int blkdev_pr_clear(struct block_device *bdev,
 	return ops->pr_clear(bdev, c.key);
 }
 
-/*
- * Is it an unrecognized ioctl? The correct returns are either
- * ENOTTY (final) or ENOIOCTLCMD ("I don't know this one, try a
- * fallback"). ENOIOCTLCMD gets turned into ENOTTY by the ioctl
- * code before returning.
- *
- * Confused drivers sometimes return EINVAL, which is wrong. It
- * means "I understood the ioctl command, but the parameters to
- * it were wrong".
- *
- * We should aim to just fix the broken drivers, the EINVAL case
- * should go away.
- */
-static inline int is_unrecognized_ioctl(int ret)
-{
-	return	ret == -EINVAL ||
-		ret == -ENOTTY ||
-		ret == -ENOIOCTLCMD;
-}
-
 static int blkdev_flushbuf(struct block_device *bdev, fmode_t mode,
 		unsigned cmd, unsigned long arg)
 {
@@ -384,9 +364,6 @@ static int blkdev_roset(struct block_device *bdev, fmode_t mode,
 	if (!capable(CAP_SYS_ADMIN))
 		return -EACCES;
 
-	ret = __blkdev_driver_ioctl(bdev, mode, cmd, arg);
-	if (!is_unrecognized_ioctl(ret))
-		return ret;
 	if (get_user(n, (int __user *)arg))
 		return -EFAULT;
 	if (bdev->bd_disk->fops->set_read_only) {
Signature · 1 line
2.28.0
#17 lore raw

PATCH 9/11 loop: use set_disk_ro

In reply to Christoph Hellwig

Use set_disk_ro instead of set_device_ro to match all other block drivers and to ensure all partitions mirror the read-only flag.

---
 drivers/block/loop.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/block/loop.c +1 -1
index cb1191d6e945f2..3e1ea45bb315d8 100644
@@ -1137,7 +1137,7 @@ static int loop_configure(struct loop_device *lo, fmode_t mode,
 	if (error)
 		goto out_unlock;
 
-	set_device_ro(bdev, (lo->lo_flags & LO_FLAGS_READ_ONLY) != 0);
+	set_disk_ro(lo->lo_disk, (lo->lo_flags & LO_FLAGS_READ_ONLY) != 0);
 
 	lo->use_dio = lo->lo_flags & LO_FLAGS_DIRECT_IO;
 	lo->lo_device = bdev;
Signature · 1 line
2.28.0
#18 lore raw

PATCH 10/11 block: remove set_device_ro

In reply to Christoph Hellwig

Fold set_device_ro into its only remaining caller.

---
 block/genhd.c         | 7 -------
 block/ioctl.c         | 2 +-
 include/linux/genhd.h | 1 -
 3 files changed, 1 insertion(+), 9 deletions(-)
block/genhd.c +0 -7
index 0a273211fec283..a487590b365e89 100644
@@ -1843,13 +1843,6 @@ static void set_disk_ro_uevent(struct gendisk *gd, int ro)
 	kobject_uevent_env(&disk_to_dev(gd)->kobj, KOBJ_CHANGE, envp);
 }
 
-void set_device_ro(struct block_device *bdev, int flag)
-{
-	bdev->bd_part->policy = flag;
-}
-
-EXPORT_SYMBOL(set_device_ro);
-
 void set_disk_ro(struct gendisk *disk, int flag)
 {
 	struct disk_part_iter piter;
block/ioctl.c +1 -1
index 96cb4544736468..04255dc5f3bff3 100644
@@ -371,7 +371,7 @@ static int blkdev_roset(struct block_device *bdev, fmode_t mode,
 		if (ret)
 			return ret;
 	}
-	set_device_ro(bdev, n);
+	bdev->bd_part->policy = n;
 	return 0;
 }
 
include/linux/genhd.h +0 -1
index 38f23d75701379..f232f26d7c777b 100644
@@ -304,7 +304,6 @@ extern void del_gendisk(struct gendisk *gp);
 extern struct gendisk *get_gendisk(dev_t dev, int *partno);
 extern struct block_device *bdget_disk(struct gendisk *disk, int partno);
 
-extern void set_device_ro(struct block_device *bdev, int flag);
 extern void set_disk_ro(struct gendisk *disk, int flag);
 
 static inline int get_disk_ro(struct gendisk *disk)
Signature · 1 line
2.28.0
#19 lore raw

PATCH 11/11 block: remove __blkdev_driver_ioctl

In reply to Christoph Hellwig

Just open code it in the few callers.

---
 block/ioctl.c               | 25 +++++--------------------
 drivers/block/pktcdvd.c     |  6 ++++--
 drivers/md/bcache/request.c |  5 +++--
 drivers/md/dm.c             |  5 ++++-
 include/linux/blkdev.h      |  2 --
 5 files changed, 16 insertions(+), 27 deletions(-)
block/ioctl.c +5 -20
index 04255dc5f3bff3..6b785181344fe1 100644
@@ -219,23 +219,6 @@ static int compat_put_ulong(compat_ulong_t __user *argp, compat_ulong_t val)
 }
 #endif
 
-int __blkdev_driver_ioctl(struct block_device *bdev, fmode_t mode,
-			unsigned cmd, unsigned long arg)
-{
-	struct gendisk *disk = bdev->bd_disk;
-
-	if (disk->fops->ioctl)
-		return disk->fops->ioctl(bdev, mode, cmd, arg);
-
-	return -ENOTTY;
-}
-/*
- * For the record: _GPL here is only because somebody decided to slap it
- * on the previous export.  Sheer idiocy, since it wasn't copyrightable
- * at all and could be open-coded without any exports by anybody who cares.
- */
-EXPORT_SYMBOL_GPL(__blkdev_driver_ioctl);
-
 #ifdef CONFIG_COMPAT
 /*
  * This is the equivalent of compat_ptr_ioctl(), to be used by block
@@ -594,10 +577,12 @@ int blkdev_ioctl(struct block_device *bdev, fmode_t mode, unsigned cmd,
 	}
 
 	ret = blkdev_common_ioctl(bdev, mode, cmd, arg, argp);
-	if (ret == -ENOIOCTLCMD)
-		return __blkdev_driver_ioctl(bdev, mode, cmd, arg);
+	if (ret != -ENOIOCTLCMD)
+		return ret;
 
-	return ret;
+	if (!bdev->bd_disk->fops->ioctl)
+		return -ENOTTY;
+	return bdev->bd_disk->fops->ioctl(bdev, mode, cmd, arg);
 }
 EXPORT_SYMBOL_GPL(blkdev_ioctl); /* for /dev/raw */
 
drivers/block/pktcdvd.c +4 -2
index 467dbd06b7cdb1..ef1c1f094ea4fc 100644
@@ -2584,9 +2584,11 @@ static int pkt_ioctl(struct block_device *bdev, fmode_t mode, unsigned int cmd,
 	case CDROM_LAST_WRITTEN:
 	case CDROM_SEND_PACKET:
 	case SCSI_IOCTL_SEND_COMMAND:
-		ret = __blkdev_driver_ioctl(pd->bdev, mode, cmd, arg);
+		if (!bdev->bd_disk->fops->ioctl)
+			ret = -ENOTTY;
+		else
+			ret = bdev->bd_disk->fops->ioctl(bdev, mode, cmd, arg);
 		break;
-
 	default:
 		pkt_dbg(2, pd, "Unknown ioctl (%x)\n", cmd);
 		ret = -ENOTTY;
drivers/md/bcache/request.c +3 -2
index 21432638314562..afac8d07c1bd00 100644
@@ -1230,8 +1230,9 @@ static int cached_dev_ioctl(struct bcache_device *d, fmode_t mode,
 
 	if (dc->io_disable)
 		return -EIO;
-
-	return __blkdev_driver_ioctl(dc->bdev, mode, cmd, arg);
+	if (!dc->bdev->bd_disk->fops->ioctl)
+		return -ENOTTY;
+	return dc->bdev->bd_disk->fops->ioctl(dc->bdev, mode, cmd, arg);
 }
 
 void bch_cached_dev_request_init(struct cached_dev *dc)
drivers/md/dm.c +4 -1
index c18fc25485186d..6db395c3d28be8 100644
@@ -570,7 +570,10 @@ static int dm_blk_ioctl(struct block_device *bdev, fmode_t mode,
 		}
 	}
 
-	r =  __blkdev_driver_ioctl(bdev, mode, cmd, arg);
+	if (!bdev->bd_disk->fops->ioctl)
+		r = -ENOTTY;
+	else
+		r = bdev->bd_disk->fops->ioctl(bdev, mode, cmd, arg);
 out:
 	dm_unprepare_ioctl(md, srcu_idx);
 	return r;
include/linux/blkdev.h +0 -2
index 5c1ba8a8d2bc7e..05b346a68c2eee 100644
@@ -1867,8 +1867,6 @@ extern int blkdev_compat_ptr_ioctl(struct block_device *, fmode_t,
 #define blkdev_compat_ptr_ioctl NULL
 #endif
 
-extern int __blkdev_driver_ioctl(struct block_device *, fmode_t, unsigned int,
-				 unsigned long);
 extern int bdev_read_page(struct block_device *, sector_t, struct page *);
 extern int bdev_write_page(struct block_device *, sector_t, struct page *,
 						struct writeback_control *);
Signature · 1 line
2.28.0
maintainer
#20 lore raw
In reply to Christoph Hellwig
Quoted 5 lines · Christoph Hellwig

Hi Jens,

this series has a bunch of cleanups for the block layer ioctl code.

Diffstat:

Quoted 16 lines
 block/genhd.c                     |    7 ----
 block/ioctl.c                     |   62 ++++++--------------------------------
 drivers/block/loop.c              |    2 -
 drivers/block/mtip32xx/mtip32xx.c |   11 +-----
 drivers/block/pktcdvd.c           |    6 ++-
 drivers/block/rbd.c               |   41 ++-----------------------
 drivers/md/bcache/request.c       |    5 +--
 drivers/md/dm.c                   |    5 ++-
 drivers/md/md.c                   |   62 +++++++++++++++++++-------------------
 drivers/mtd/mtd_blkdevs.c         |   28 -----------------
 drivers/s390/block/dasd.c         |    1 
 drivers/s390/block/dasd_int.h     |    3 +
 drivers/s390/block/dasd_ioctl.c   |   27 +++++-----------
 include/linux/blkdev.h            |    3 -
 include/linux/genhd.h             |    1 
 15 files changed, 73 insertions(+), 191 deletions(-)

Series looks good to me, apart from the mentioned mtip32xx change. If you repost this, can you look through commit messages and titles for typos, I spotted quite a few. If not I'll do it when applying.

Signature · 1 line
Jens Axboe