diff options
author | James Morris <jmorris@namei.org> | 2009-02-06 01:01:45 +0100 |
---|---|---|
committer | James Morris <jmorris@namei.org> | 2009-02-06 01:01:45 +0100 |
commit | cb5629b10d64a8006622ce3a52bc887d91057d69 (patch) | |
tree | 7c06d8f30783115e3384721046258ce615b129c5 /drivers/mtd/ubi | |
parent | TPM: integrity fix (diff) | |
parent | seq_file: fix big-enough lseek() + read() (diff) | |
download | linux-cb5629b10d64a8006622ce3a52bc887d91057d69.tar.xz linux-cb5629b10d64a8006622ce3a52bc887d91057d69.zip |
Merge branch 'master' into next
Conflicts:
fs/namei.c
Manually merged per:
diff --cc fs/namei.c
index 734f2b5,bbc15c2..0000000
--- a/fs/namei.c
+++ b/fs/namei.c
@@@ -860,9 -848,8 +849,10 @@@ static int __link_path_walk(const char
nd->flags |= LOOKUP_CONTINUE;
err = exec_permission_lite(inode);
if (err == -EAGAIN)
- err = vfs_permission(nd, MAY_EXEC);
+ err = inode_permission(nd->path.dentry->d_inode,
+ MAY_EXEC);
+ if (!err)
+ err = ima_path_check(&nd->path, MAY_EXEC);
if (err)
break;
@@@ -1525,14 -1506,9 +1509,14 @@@ int may_open(struct path *path, int acc
flag &= ~O_TRUNC;
}
- error = vfs_permission(nd, acc_mode);
+ error = inode_permission(inode, acc_mode);
if (error)
return error;
+
- error = ima_path_check(&nd->path,
++ error = ima_path_check(path,
+ acc_mode & (MAY_READ | MAY_WRITE | MAY_EXEC));
+ if (error)
+ return error;
/*
* An append-only file must be opened in append mode for writing.
*/
Signed-off-by: James Morris <jmorris@namei.org>
Diffstat (limited to 'drivers/mtd/ubi')
-rw-r--r-- | drivers/mtd/ubi/Kconfig.debug | 10 | ||||
-rw-r--r-- | drivers/mtd/ubi/build.c | 30 | ||||
-rw-r--r-- | drivers/mtd/ubi/cdev.c | 187 | ||||
-rw-r--r-- | drivers/mtd/ubi/debug.h | 10 | ||||
-rw-r--r-- | drivers/mtd/ubi/eba.c | 53 | ||||
-rw-r--r-- | drivers/mtd/ubi/gluebi.c | 28 | ||||
-rw-r--r-- | drivers/mtd/ubi/io.c | 30 | ||||
-rw-r--r-- | drivers/mtd/ubi/kapi.c | 2 | ||||
-rw-r--r-- | drivers/mtd/ubi/scan.c | 10 | ||||
-rw-r--r-- | drivers/mtd/ubi/ubi-media.h | 4 | ||||
-rw-r--r-- | drivers/mtd/ubi/ubi.h | 56 | ||||
-rw-r--r-- | drivers/mtd/ubi/upd.c | 21 | ||||
-rw-r--r-- | drivers/mtd/ubi/vmt.c | 21 | ||||
-rw-r--r-- | drivers/mtd/ubi/vtbl.c | 2 | ||||
-rw-r--r-- | drivers/mtd/ubi/wl.c | 493 |
15 files changed, 497 insertions, 460 deletions
diff --git a/drivers/mtd/ubi/Kconfig.debug b/drivers/mtd/ubi/Kconfig.debug index 1e2ee22edeff..2246f154e2f7 100644 --- a/drivers/mtd/ubi/Kconfig.debug +++ b/drivers/mtd/ubi/Kconfig.debug @@ -33,16 +33,6 @@ config MTD_UBI_DEBUG_DISABLE_BGT This option switches the background thread off by default. The thread may be also be enabled/disabled via UBI sysfs. -config MTD_UBI_DEBUG_USERSPACE_IO - bool "Direct user-space write/erase support" - default n - depends on MTD_UBI_DEBUG - help - By default, users cannot directly write and erase individual - eraseblocks of dynamic volumes, and have to use update operation - instead. This option enables this capability - it is very useful for - debugging and testing. - config MTD_UBI_DEBUG_EMULATE_BITFLIPS bool "Emulate flash bit-flips" depends on MTD_UBI_DEBUG diff --git a/drivers/mtd/ubi/build.c b/drivers/mtd/ubi/build.c index c7630a228310..4048db83aef6 100644 --- a/drivers/mtd/ubi/build.c +++ b/drivers/mtd/ubi/build.c @@ -263,8 +263,12 @@ static ssize_t dev_attribute_show(struct device *dev, return ret; } -/* Fake "release" method for UBI devices */ -static void dev_release(struct device *dev) { } +static void dev_release(struct device *dev) +{ + struct ubi_device *ubi = container_of(dev, struct ubi_device, dev); + + kfree(ubi); +} /** * ubi_sysfs_init - initialize sysfs for an UBI device. @@ -280,7 +284,7 @@ static int ubi_sysfs_init(struct ubi_device *ubi) ubi->dev.release = dev_release; ubi->dev.devt = ubi->cdev.dev; ubi->dev.class = ubi_class; - sprintf(&ubi->dev.bus_id[0], UBI_NAME_STR"%d", ubi->ubi_num); + dev_set_name(&ubi->dev, UBI_NAME_STR"%d", ubi->ubi_num); err = device_register(&ubi->dev); if (err) return err; @@ -380,7 +384,7 @@ static void free_user_volumes(struct ubi_device *ubi) */ static int uif_init(struct ubi_device *ubi) { - int i, err, do_free = 0; + int i, err; dev_t dev; sprintf(ubi->ubi_name, UBI_NAME_STR "%d", ubi->ubi_num); @@ -427,13 +431,10 @@ static int uif_init(struct ubi_device *ubi) out_volumes: kill_volumes(ubi); - do_free = 0; out_sysfs: ubi_sysfs_close(ubi); cdev_del(&ubi->cdev); out_unreg: - if (do_free) - free_user_volumes(ubi); unregister_chrdev_region(ubi->cdev.dev, ubi->vtbl_slots + 1); ubi_err("cannot initialize UBI %s, error %d", ubi->ubi_name, err); return err; @@ -561,7 +562,7 @@ static int io_init(struct ubi_device *ubi) */ ubi->peb_size = ubi->mtd->erasesize; - ubi->peb_count = ubi->mtd->size / ubi->mtd->erasesize; + ubi->peb_count = mtd_div_by_eb(ubi->mtd->size, ubi->mtd); ubi->flash_size = ubi->mtd->size; if (ubi->mtd->block_isbad && ubi->mtd->block_markbad) @@ -815,19 +816,20 @@ int ubi_attach_mtd_dev(struct mtd_info *mtd, int ubi_num, int vid_hdr_offset) if (err) goto out_free; + err = -ENOMEM; ubi->peb_buf1 = vmalloc(ubi->peb_size); if (!ubi->peb_buf1) goto out_free; ubi->peb_buf2 = vmalloc(ubi->peb_size); if (!ubi->peb_buf2) - goto out_free; + goto out_free; #ifdef CONFIG_MTD_UBI_DEBUG mutex_init(&ubi->dbg_buf_mutex); ubi->dbg_peb_buf = vmalloc(ubi->peb_size); if (!ubi->dbg_peb_buf) - goto out_free; + goto out_free; #endif err = attach_by_scanning(ubi); @@ -946,6 +948,12 @@ int ubi_detach_mtd_dev(int ubi_num, int anyway) if (ubi->bgt_thread) kthread_stop(ubi->bgt_thread); + /* + * Get a reference to the device in order to prevent 'dev_release()' + * from freeing @ubi object. + */ + get_device(&ubi->dev); + uif_close(ubi); ubi_wl_close(ubi); free_internal_volumes(ubi); @@ -957,7 +965,7 @@ int ubi_detach_mtd_dev(int ubi_num, int anyway) vfree(ubi->dbg_peb_buf); #endif ubi_msg("mtd%d is detached from ubi%d", ubi->mtd->index, ubi->ubi_num); - kfree(ubi); + put_device(&ubi->dev); return 0; } diff --git a/drivers/mtd/ubi/cdev.c b/drivers/mtd/ubi/cdev.c index b30a0b83d7f1..e63c8fc3df3a 100644 --- a/drivers/mtd/ubi/cdev.c +++ b/drivers/mtd/ubi/cdev.c @@ -40,9 +40,9 @@ #include <linux/ioctl.h> #include <linux/capability.h> #include <linux/uaccess.h> -#include <linux/smp_lock.h> +#include <linux/compat.h> +#include <linux/math64.h> #include <mtd/ubi-user.h> -#include <asm/div64.h> #include "ubi.h" /** @@ -195,7 +195,6 @@ static ssize_t vol_cdev_read(struct file *file, __user char *buf, size_t count, int err, lnum, off, len, tbuf_size; size_t count_save = count; void *tbuf; - uint64_t tmp; dbg_gen("read %zd bytes from offset %lld of volume %d", count, *offp, vol->vol_id); @@ -225,10 +224,7 @@ static ssize_t vol_cdev_read(struct file *file, __user char *buf, size_t count, return -ENOMEM; len = count > tbuf_size ? tbuf_size : count; - - tmp = *offp; - off = do_div(tmp, vol->usable_leb_size); - lnum = tmp; + lnum = div_u64_rem(*offp, vol->usable_leb_size, &off); do { cond_resched(); @@ -263,12 +259,9 @@ static ssize_t vol_cdev_read(struct file *file, __user char *buf, size_t count, return err ? err : count_save - count; } -#ifdef CONFIG_MTD_UBI_DEBUG_USERSPACE_IO - /* * This function allows to directly write to dynamic UBI volumes, without - * issuing the volume update operation. Available only as a debugging feature. - * Very useful for testing UBI. + * issuing the volume update operation. */ static ssize_t vol_cdev_direct_write(struct file *file, const char __user *buf, size_t count, loff_t *offp) @@ -279,7 +272,9 @@ static ssize_t vol_cdev_direct_write(struct file *file, const char __user *buf, int lnum, off, len, tbuf_size, err = 0; size_t count_save = count; char *tbuf; - uint64_t tmp; + + if (!vol->direct_writes) + return -EPERM; dbg_gen("requested: write %zd bytes to offset %lld of volume %u", count, *offp, vol->vol_id); @@ -287,10 +282,7 @@ static ssize_t vol_cdev_direct_write(struct file *file, const char __user *buf, if (vol->vol_type == UBI_STATIC_VOLUME) return -EROFS; - tmp = *offp; - off = do_div(tmp, vol->usable_leb_size); - lnum = tmp; - + lnum = div_u64_rem(*offp, vol->usable_leb_size, &off); if (off & (ubi->min_io_size - 1)) { dbg_err("unaligned position"); return -EINVAL; @@ -347,10 +339,6 @@ static ssize_t vol_cdev_direct_write(struct file *file, const char __user *buf, return err ? err : count_save - count; } -#else -#define vol_cdev_direct_write(file, buf, count, offp) (-EPERM) -#endif /* CONFIG_MTD_UBI_DEBUG_USERSPACE_IO */ - static ssize_t vol_cdev_write(struct file *file, const char __user *buf, size_t count, loff_t *offp) { @@ -402,8 +390,8 @@ static ssize_t vol_cdev_write(struct file *file, const char __user *buf, return count; } -static int vol_cdev_ioctl(struct inode *inode, struct file *file, - unsigned int cmd, unsigned long arg) +static long vol_cdev_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) { int err = 0; struct ubi_volume_desc *desc = file->private_data; @@ -487,7 +475,6 @@ static int vol_cdev_ioctl(struct inode *inode, struct file *file, break; } -#ifdef CONFIG_MTD_UBI_DEBUG_USERSPACE_IO /* Logical eraseblock erasure command */ case UBI_IOCEBER: { @@ -518,13 +505,77 @@ static int vol_cdev_ioctl(struct inode *inode, struct file *file, err = ubi_wl_flush(ubi); break; } -#endif + + /* Logical eraseblock map command */ + case UBI_IOCEBMAP: + { + struct ubi_map_req req; + + err = copy_from_user(&req, argp, sizeof(struct ubi_map_req)); + if (err) { + err = -EFAULT; + break; + } + err = ubi_leb_map(desc, req.lnum, req.dtype); + break; + } + + /* Logical eraseblock un-map command */ + case UBI_IOCEBUNMAP: + { + int32_t lnum; + + err = get_user(lnum, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + err = ubi_leb_unmap(desc, lnum); + break; + } + + /* Check if logical eraseblock is mapped command */ + case UBI_IOCEBISMAP: + { + int32_t lnum; + + err = get_user(lnum, (__user int32_t *)argp); + if (err) { + err = -EFAULT; + break; + } + err = ubi_is_mapped(desc, lnum); + break; + } + + /* Set volume property command*/ + case UBI_IOCSETPROP: + { + struct ubi_set_prop_req req; + + err = copy_from_user(&req, argp, + sizeof(struct ubi_set_prop_req)); + if (err) { + err = -EFAULT; + break; + } + switch (req.property) { + case UBI_PROP_DIRECT_WRITE: + mutex_lock(&ubi->volumes_mutex); + desc->vol->direct_writes = !!req.value; + mutex_unlock(&ubi->volumes_mutex); + break; + default: + err = -EINVAL; + break; + } + break; + } default: err = -ENOTTY; break; } - return err; } @@ -721,7 +772,8 @@ static int rename_volumes(struct ubi_device *ubi, * It seems we need to remove volume with name @re->new_name, * if it exists. */ - desc = ubi_open_volume_nm(ubi->ubi_num, re->new_name, UBI_EXCLUSIVE); + desc = ubi_open_volume_nm(ubi->ubi_num, re->new_name, + UBI_EXCLUSIVE); if (IS_ERR(desc)) { err = PTR_ERR(desc); if (err == -ENODEV) @@ -761,8 +813,8 @@ out_free: return err; } -static int ubi_cdev_ioctl(struct inode *inode, struct file *file, - unsigned int cmd, unsigned long arg) +static long ubi_cdev_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) { int err = 0; struct ubi_device *ubi; @@ -772,7 +824,7 @@ static int ubi_cdev_ioctl(struct inode *inode, struct file *file, if (!capable(CAP_SYS_RESOURCE)) return -EPERM; - ubi = ubi_get_by_major(imajor(inode)); + ubi = ubi_get_by_major(imajor(file->f_mapping->host)); if (!ubi) return -ENODEV; @@ -842,7 +894,6 @@ static int ubi_cdev_ioctl(struct inode *inode, struct file *file, case UBI_IOCRSVOL: { int pebs; - uint64_t tmp; struct ubi_rsvol_req req; dbg_gen("re-size volume"); @@ -862,9 +913,8 @@ static int ubi_cdev_ioctl(struct inode *inode, struct file *file, break; } - tmp = req.bytes; - pebs = !!do_div(tmp, desc->vol->usable_leb_size); - pebs += tmp; + pebs = div_u64(req.bytes + desc->vol->usable_leb_size - 1, + desc->vol->usable_leb_size); mutex_lock(&ubi->volumes_mutex); err = ubi_resize_volume(desc, pebs); @@ -908,8 +958,8 @@ static int ubi_cdev_ioctl(struct inode *inode, struct file *file, return err; } -static int ctrl_cdev_ioctl(struct inode *inode, struct file *file, - unsigned int cmd, unsigned long arg) +static long ctrl_cdev_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) { int err = 0; void __user *argp = (void __user *)arg; @@ -985,26 +1035,59 @@ static int ctrl_cdev_ioctl(struct inode *inode, struct file *file, return err; } -/* UBI control character device operations */ -struct file_operations ubi_ctrl_cdev_operations = { - .ioctl = ctrl_cdev_ioctl, - .owner = THIS_MODULE, +#ifdef CONFIG_COMPAT +static long vol_cdev_compat_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + unsigned long translated_arg = (unsigned long)compat_ptr(arg); + + return vol_cdev_ioctl(file, cmd, translated_arg); +} + +static long ubi_cdev_compat_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + unsigned long translated_arg = (unsigned long)compat_ptr(arg); + + return ubi_cdev_ioctl(file, cmd, translated_arg); +} + +static long ctrl_cdev_compat_ioctl(struct file *file, unsigned int cmd, + unsigned long arg) +{ + unsigned long translated_arg = (unsigned long)compat_ptr(arg); + + return ctrl_cdev_ioctl(file, cmd, translated_arg); +} +#else +#define vol_cdev_compat_ioctl NULL +#define ubi_cdev_compat_ioctl NULL +#define ctrl_cdev_compat_ioctl NULL +#endif + +/* UBI volume character device operations */ +const struct file_operations ubi_vol_cdev_operations = { + .owner = THIS_MODULE, + .open = vol_cdev_open, + .release = vol_cdev_release, + .llseek = vol_cdev_llseek, + .read = vol_cdev_read, + .write = vol_cdev_write, + .unlocked_ioctl = vol_cdev_ioctl, + .compat_ioctl = vol_cdev_compat_ioctl, }; /* UBI character device operations */ -struct file_operations ubi_cdev_operations = { - .owner = THIS_MODULE, - .ioctl = ubi_cdev_ioctl, - .llseek = no_llseek, +const struct file_operations ubi_cdev_operations = { + .owner = THIS_MODULE, + .llseek = no_llseek, + .unlocked_ioctl = ubi_cdev_ioctl, + .compat_ioctl = ubi_cdev_compat_ioctl, }; -/* UBI volume character device operations */ -struct file_operations ubi_vol_cdev_operations = { - .owner = THIS_MODULE, - .open = vol_cdev_open, - .release = vol_cdev_release, - .llseek = vol_cdev_llseek, - .read = vol_cdev_read, - .write = vol_cdev_write, - .ioctl = vol_cdev_ioctl, +/* UBI control character device operations */ +const struct file_operations ubi_ctrl_cdev_operations = { + .owner = THIS_MODULE, + .unlocked_ioctl = ctrl_cdev_ioctl, + .compat_ioctl = ctrl_cdev_compat_ioctl, }; diff --git a/drivers/mtd/ubi/debug.h b/drivers/mtd/ubi/debug.h index 78e914d23ece..13777e5beac9 100644 --- a/drivers/mtd/ubi/debug.h +++ b/drivers/mtd/ubi/debug.h @@ -27,11 +27,11 @@ #define dbg_err(fmt, ...) ubi_err(fmt, ##__VA_ARGS__) #define ubi_assert(expr) do { \ - if (unlikely(!(expr))) { \ - printk(KERN_CRIT "UBI assert failed in %s at %u (pid %d)\n", \ - __func__, __LINE__, current->pid); \ - ubi_dbg_dump_stack(); \ - } \ + if (unlikely(!(expr))) { \ + printk(KERN_CRIT "UBI assert failed in %s at %u (pid %d)\n", \ + __func__, __LINE__, current->pid); \ + ubi_dbg_dump_stack(); \ + } \ } while (0) #define dbg_msg(fmt, ...) \ diff --git a/drivers/mtd/ubi/eba.c b/drivers/mtd/ubi/eba.c index d8966bae0e0b..25def348e5ba 100644 --- a/drivers/mtd/ubi/eba.c +++ b/drivers/mtd/ubi/eba.c @@ -504,12 +504,9 @@ static int recover_peb(struct ubi_device *ubi, int pnum, int vol_id, int lnum, if (!vid_hdr) return -ENOMEM; - mutex_lock(&ubi->buf_mutex); - retry: new_pnum = ubi_wl_get_peb(ubi, UBI_UNKNOWN); if (new_pnum < 0) { - mutex_unlock(&ubi->buf_mutex); ubi_free_vid_hdr(ubi, vid_hdr); return new_pnum; } @@ -529,20 +526,23 @@ retry: goto write_error; data_size = offset + len; + mutex_lock(&ubi->buf_mutex); memset(ubi->peb_buf1 + offset, 0xFF, len); /* Read everything before the area where the write failure happened */ if (offset > 0) { err = ubi_io_read_data(ubi, ubi->peb_buf1, pnum, 0, offset); if (err && err != UBI_IO_BITFLIPS) - goto out_put; + goto out_unlock; } memcpy(ubi->peb_buf1 + offset, buf, len); err = ubi_io_write_data(ubi, ubi->peb_buf1, new_pnum, 0, data_size); - if (err) + if (err) { + mutex_unlock(&ubi->buf_mutex); goto write_error; + } mutex_unlock(&ubi->buf_mutex); ubi_free_vid_hdr(ubi, vid_hdr); @@ -553,8 +553,9 @@ retry: ubi_msg("data was successfully recovered"); return 0; -out_put: +out_unlock: mutex_unlock(&ubi->buf_mutex); +out_put: ubi_wl_put_peb(ubi, new_pnum, 1); ubi_free_vid_hdr(ubi, vid_hdr); return err; @@ -567,7 +568,6 @@ write_error: ubi_warn("failed to write to PEB %d", new_pnum); ubi_wl_put_peb(ubi, new_pnum, 1); if (++tries > UBI_IO_RETRIES) { - mutex_unlock(&ubi->buf_mutex); ubi_free_vid_hdr(ubi, vid_hdr); return err; } @@ -717,7 +717,7 @@ write_error: * to the real data size, although the @buf buffer has to contain the * alignment. In all other cases, @len has to be aligned. * - * It is prohibited to write more then once to logical eraseblocks of static + * It is prohibited to write more than once to logical eraseblocks of static * volumes. This function returns zero in case of success and a negative error * code in case of failure. */ @@ -949,10 +949,14 @@ write_error: * This function copies logical eraseblock from physical eraseblock @from to * physical eraseblock @to. The @vid_hdr buffer may be changed by this * function. Returns: - * o %0 in case of success; - * o %1 if the operation was canceled and should be tried later (e.g., - * because a bit-flip was detected at the target PEB); - * o %2 if the volume is being deleted and this LEB should not be moved. + * o %0 in case of success; + * o %1 if the operation was canceled because the volume is being deleted + * or because the PEB was put meanwhile; + * o %2 if the operation was canceled because there was a write error to the + * target PEB; + * o %-EAGAIN if the operation was canceled because a bit-flip was detected + * in the target PEB; + * o a negative error code in case of failure. */ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, struct ubi_vid_hdr *vid_hdr) @@ -978,7 +982,7 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, /* * Note, we may race with volume deletion, which means that the volume * this logical eraseblock belongs to might be being deleted. Since the - * volume deletion unmaps all the volume's logical eraseblocks, it will + * volume deletion un-maps all the volume's logical eraseblocks, it will * be locked in 'ubi_wl_put_peb()' and wait for the WL worker to finish. */ vol = ubi->volumes[idx]; @@ -986,7 +990,7 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, /* No need to do further work, cancel */ dbg_eba("volume %d is being removed, cancel", vol_id); spin_unlock(&ubi->volumes_lock); - return 2; + return 1; } spin_unlock(&ubi->volumes_lock); @@ -1023,7 +1027,7 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, /* * OK, now the LEB is locked and we can safely start moving it. Since - * this function utilizes thie @ubi->peb1_buf buffer which is shared + * this function utilizes the @ubi->peb1_buf buffer which is shared * with some other functions, so lock the buffer by taking the * @ubi->buf_mutex. */ @@ -1068,8 +1072,11 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, vid_hdr->sqnum = cpu_to_be64(next_sqnum(ubi)); err = ubi_io_write_vid_hdr(ubi, to, vid_hdr); - if (err) + if (err) { + if (err == -EIO) + err = 2; goto out_unlock_buf; + } cond_resched(); @@ -1079,14 +1086,17 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, if (err != UBI_IO_BITFLIPS) ubi_warn("cannot read VID header back from PEB %d", to); else - err = 1; + err = -EAGAIN; goto out_unlock_buf; } if (data_size > 0) { err = ubi_io_write_data(ubi, ubi->peb_buf1, to, 0, aldata_size); - if (err) + if (err) { + if (err == -EIO) + err = 2; goto out_unlock_buf; + } cond_resched(); @@ -1101,15 +1111,16 @@ int ubi_eba_copy_leb(struct ubi_device *ubi, int from, int to, ubi_warn("cannot read data back from PEB %d", to); else - err = 1; + err = -EAGAIN; goto out_unlock_buf; } cond_resched(); if (memcmp(ubi->peb_buf1, ubi->peb_buf2, aldata_size)) { - ubi_warn("read data back from PEB %d - it is different", - to); + ubi_warn("read data back from PEB %d and it is " + "different", to); + err = -EINVAL; goto out_unlock_buf; } } diff --git a/drivers/mtd/ubi/gluebi.c b/drivers/mtd/ubi/gluebi.c index 605812bb0b1a..49cd55ade9c8 100644 --- a/drivers/mtd/ubi/gluebi.c +++ b/drivers/mtd/ubi/gluebi.c @@ -28,7 +28,7 @@ * eraseblock size is equivalent to the logical eraseblock size of the volume. */ -#include <asm/div64.h> +#include <linux/math64.h> #include "ubi.h" /** @@ -109,7 +109,6 @@ static int gluebi_read(struct mtd_info *mtd, loff_t from, size_t len, int err = 0, lnum, offs, total_read; struct ubi_volume *vol; struct ubi_device *ubi; - uint64_t tmp = from; dbg_gen("read %zd bytes from offset %lld", len, from); @@ -119,9 +118,7 @@ static int gluebi_read(struct mtd_info *mtd, loff_t from, size_t len, vol = container_of(mtd, struct ubi_volume, gluebi_mtd); ubi = vol->ubi; - offs = do_div(tmp, mtd->erasesize); - lnum = tmp; - + lnum = div_u64_rem(from, mtd->erasesize, &offs); total_read = len; while (total_read) { size_t to_read = mtd->erasesize - offs; @@ -160,7 +157,6 @@ static int gluebi_write(struct mtd_info *mtd, loff_t to, size_t len, int err = 0, lnum, offs, total_written; struct ubi_volume *vol; struct ubi_device *ubi; - uint64_t tmp = to; dbg_gen("write %zd bytes to offset %lld", len, to); @@ -173,8 +169,7 @@ static int gluebi_write(struct mtd_info *mtd, loff_t to, size_t len, if (ubi->ro_mode) return -EROFS; - offs = do_div(tmp, mtd->erasesize); - lnum = tmp; + lnum = div_u64_rem(to, mtd->erasesize, &offs); if (len % mtd->writesize || offs % mtd->writesize) return -EINVAL; @@ -215,7 +210,8 @@ static int gluebi_erase(struct mtd_info *mtd, struct erase_info *instr) struct ubi_volume *vol; struct ubi_device *ubi; - dbg_gen("erase %u bytes at offset %u", instr->len, instr->addr); + dbg_gen("erase %llu bytes at offset %llu", (unsigned long long)instr->len, + (unsigned long long)instr->addr); if (instr->addr < 0 || instr->addr > mtd->size - mtd->erasesize) return -EINVAL; @@ -223,11 +219,11 @@ static int gluebi_erase(struct mtd_info *mtd, struct erase_info *instr) if (instr->len < 0 || instr->addr + instr->len > mtd->size) return -EINVAL; - if (instr->addr % mtd->writesize || instr->len % mtd->writesize) + if (mtd_mod_by_ws(instr->addr, mtd) || mtd_mod_by_ws(instr->len, mtd)) return -EINVAL; - lnum = instr->addr / mtd->erasesize; - count = instr->len / mtd->erasesize; + lnum = mtd_div_by_eb(instr->addr, mtd); + count = mtd_div_by_eb(instr->len, mtd); vol = container_of(mtd, struct ubi_volume, gluebi_mtd); ubi = vol->ubi; @@ -255,7 +251,7 @@ static int gluebi_erase(struct mtd_info *mtd, struct erase_info *instr) out_err: instr->state = MTD_ERASE_FAILED; - instr->fail_addr = lnum * mtd->erasesize; + instr->fail_addr = (long long)lnum * mtd->erasesize; return err; } @@ -294,7 +290,7 @@ int ubi_create_gluebi(struct ubi_device *ubi, struct ubi_volume *vol) * bytes. */ if (vol->vol_type == UBI_DYNAMIC_VOLUME) - mtd->size = vol->usable_leb_size * vol->reserved_pebs; + mtd->size = (long long)vol->usable_leb_size * vol->reserved_pebs; else mtd->size = vol->used_bytes; @@ -304,8 +300,8 @@ int ubi_create_gluebi(struct ubi_device *ubi, struct ubi_volume *vol) return -ENFILE; } - dbg_gen("added mtd%d (\"%s\"), size %u, EB size %u", - mtd->index, mtd->name, mtd->size, mtd->erasesize); + dbg_gen("added mtd%d (\"%s\"), size %llu, EB size %u", + mtd->index, mtd->name, (unsigned long long)mtd->size, mtd->erasesize); return 0; } diff --git a/drivers/mtd/ubi/io.c b/drivers/mtd/ubi/io.c index 2fb64be44f1b..fe81039f2a7c 100644 --- a/drivers/mtd/ubi/io.c +++ b/drivers/mtd/ubi/io.c @@ -465,7 +465,7 @@ out: * This function synchronously erases physical eraseblock @pnum. If @torture * flag is not zero, the physical eraseblock is checked by means of writing * different patterns to it and reading them back. If the torturing is enabled, - * the physical eraseblock is erased more then once. + * the physical eraseblock is erased more than once. * * This function returns the number of erasures made in case of success, %-EIO * if the erasure failed or the torturing test failed, and other negative error @@ -637,8 +637,6 @@ int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum, dbg_io("read EC header from PEB %d", pnum); ubi_assert(pnum >= 0 && pnum < ubi->peb_count); - if (UBI_IO_DEBUG) - verbose = 1; err = ubi_io_read(ubi, ec_hdr, pnum, 0, UBI_EC_HDR_SIZE); if (err) { @@ -685,6 +683,9 @@ int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum, if (verbose) ubi_warn("no EC header found at PEB %d, " "only 0xFF bytes", pnum); + else if (UBI_IO_DEBUG) + dbg_msg("no EC header found at PEB %d, " + "only 0xFF bytes", pnum); return UBI_IO_PEB_EMPTY; } @@ -696,7 +697,9 @@ int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum, ubi_warn("bad magic number at PEB %d: %08x instead of " "%08x", pnum, magic, UBI_EC_HDR_MAGIC); ubi_dbg_dump_ec_hdr(ec_hdr); - } + } else if (UBI_IO_DEBUG) + dbg_msg("bad magic number at PEB %d: %08x instead of " + "%08x", pnum, magic, UBI_EC_HDR_MAGIC); return UBI_IO_BAD_EC_HDR; } @@ -708,7 +711,9 @@ int ubi_io_read_ec_hdr(struct ubi_device *ubi, int pnum, ubi_warn("bad EC header CRC at PEB %d, calculated " "%#08x, read %#08x", pnum, crc, hdr_crc); ubi_dbg_dump_ec_hdr(ec_hdr); - } + } else if (UBI_IO_DEBUG) + dbg_msg("bad EC header CRC at PEB %d, calculated " + "%#08x, read %#08x", pnum, crc, hdr_crc); return UBI_IO_BAD_EC_HDR; } @@ -912,8 +917,6 @@ int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum, dbg_io("read VID header from PEB %d", pnum); ubi_assert(pnum >= 0 && pnum < ubi->peb_count); - if (UBI_IO_DEBUG) - verbose = 1; p = (char *)vid_hdr - ubi->vid_hdr_shift; err = ubi_io_read(ubi, p, pnum, ubi->vid_hdr_aloffset, @@ -960,6 +963,9 @@ int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum, if (verbose) ubi_warn("no VID header found at PEB %d, " "only 0xFF bytes", pnum); + else if (UBI_IO_DEBUG) + dbg_msg("no VID header found at PEB %d, " + "only 0xFF bytes", pnum); return UBI_IO_PEB_FREE; } @@ -971,7 +977,9 @@ int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum, ubi_warn("bad magic number at PEB %d: %08x instead of " "%08x", pnum, magic, UBI_VID_HDR_MAGIC); ubi_dbg_dump_vid_hdr(vid_hdr); - } + } else if (UBI_IO_DEBUG) + dbg_msg("bad magic number at PEB %d: %08x instead of " + "%08x", pnum, magic, UBI_VID_HDR_MAGIC); return UBI_IO_BAD_VID_HDR; } @@ -983,7 +991,9 @@ int ubi_io_read_vid_hdr(struct ubi_device *ubi, int pnum, ubi_warn("bad CRC at PEB %d, calculated %#08x, " "read %#08x", pnum, crc, hdr_crc); ubi_dbg_dump_vid_hdr(vid_hdr); - } + } else if (UBI_IO_DEBUG) + dbg_msg("bad CRC at PEB %d, calculated %#08x, " + "read %#08x", pnum, crc, hdr_crc); return UBI_IO_BAD_VID_HDR; } @@ -1024,7 +1034,7 @@ int ubi_io_write_vid_hdr(struct ubi_device *ubi, int pnum, err = paranoid_check_peb_ec_hdr(ubi, pnum); if (err) - return err > 0 ? -EINVAL: err; + return err > 0 ? -EINVAL : err; vid_hdr->magic = cpu_to_be32(UBI_VID_HDR_MAGIC); vid_hdr->version = UBI_VERSION; diff --git a/drivers/mtd/ubi/kapi.c b/drivers/mtd/ubi/kapi.c index 5d9bcf109c13..4abbe573fa40 100644 --- a/drivers/mtd/ubi/kapi.c +++ b/drivers/mtd/ubi/kapi.c @@ -564,7 +564,7 @@ EXPORT_SYMBOL_GPL(ubi_leb_unmap); * @dtype: expected data type * * This function maps an un-mapped logical eraseblock @lnum to a physical - * eraseblock. This means, that after a successfull invocation of this + * eraseblock. This means, that after a successful invocation of this * function the logical eraseblock @lnum will be empty (contain only %0xFF * bytes) and be mapped to a physical eraseblock, even if an unclean reboot * happens. diff --git a/drivers/mtd/ubi/scan.c b/drivers/mtd/ubi/scan.c index 41d47e1cf15c..c3d653ba5ca0 100644 --- a/drivers/mtd/ubi/scan.c +++ b/drivers/mtd/ubi/scan.c @@ -42,7 +42,7 @@ #include <linux/err.h> #include <linux/crc32.h> -#include <asm/div64.h> +#include <linux/math64.h> #include "ubi.h" #ifdef CONFIG_MTD_UBI_DEBUG_PARANOID @@ -478,7 +478,7 @@ int ubi_scan_add_used(struct ubi_device *ubi, struct ubi_scan_info *si, return 0; } else { /* - * This logical eraseblock is older then the one found + * This logical eraseblock is older than the one found * previously. */ if (cmp_res & 4) @@ -904,10 +904,8 @@ struct ubi_scan_info *ubi_scan(struct ubi_device *ubi) dbg_msg("scanning is finished"); /* Calculate mean erase counter */ - if (si->ec_count) { - do_div(si->ec_sum, si->ec_count); - si->mean_ec = si->ec_sum; - } + if (si->ec_count) + si->mean_ec = div_u64(si->ec_sum, si->ec_count); if (si->is_empty) ubi_msg("empty MTD device detected"); diff --git a/drivers/mtd/ubi/ubi-media.h b/drivers/mtd/ubi/ubi-media.h index 2ad940409053..8419fdccc79c 100644 --- a/drivers/mtd/ubi/ubi-media.h +++ b/drivers/mtd/ubi/ubi-media.h @@ -135,7 +135,7 @@ enum { * The erase counter header takes 64 bytes and has a plenty of unused space for * future usage. The unused fields are zeroed. The @version field is used to * indicate the version of UBI implementation which is supposed to be able to - * work with this UBI image. If @version is greater then the current UBI + * work with this UBI image. If @version is greater than the current UBI * version, the image is rejected. This may be useful in future if something * is changed radically. This field is duplicated in the volume identifier * header. @@ -187,7 +187,7 @@ struct ubi_ec_hdr { * (sequence number) is used to distinguish between older and newer versions of * logical eraseblocks. * - * There are 2 situations when there may be more then one physical eraseblock + * There are 2 situations when there may be more than one physical eraseblock * corresponding to the same logical eraseblock, i.e., having the same @vol_id * and @lnum values in the volume identifier header. Suppose we have a logical * eraseblock L and it is mapped to the physical eraseblock P. diff --git a/drivers/mtd/ubi/ubi.h b/drivers/mtd/ubi/ubi.h index 1c3fa18c26a7..c055511bb1b2 100644 --- a/drivers/mtd/ubi/ubi.h +++ b/drivers/mtd/ubi/ubi.h @@ -74,6 +74,13 @@ #define UBI_IO_RETRIES 3 /* + * Length of the protection queue. The length is effectively equivalent to the + * number of (global) erase cycles PEBs are protected from the wear-leveling + * worker. + */ +#define UBI_PROT_QUEUE_LEN 10 + +/* * Error codes returned by the I/O sub-system. * * UBI_IO_PEB_EMPTY: the physical eraseblock is empty, i.e. it contains only @@ -95,7 +102,8 @@ enum { /** * struct ubi_wl_entry - wear-leveling entry. - * @rb: link in the corresponding RB-tree + * @u.rb: link in the corresponding (free/used) RB-tree + * @u.list: link in the protection queue * @ec: erase counter * @pnum: physical eraseblock number * @@ -104,7 +112,10 @@ enum { * RB-trees. See WL sub-system for details. */ struct ubi_wl_entry { - struct rb_node rb; + union { + struct rb_node rb; + struct list_head list; + } u; int ec; int pnum; }; @@ -195,6 +206,7 @@ struct ubi_volume_desc; * @upd_marker: %1 if the update marker is set for this volume * @updating: %1 if the volume is being updated * @changing_leb: %1 if the atomic LEB change ioctl command is in progress + * @direct_writes: %1 if direct writes are enabled for this volume * * @gluebi_desc: gluebi UBI volume descriptor * @gluebi_refcount: reference count of the gluebi MTD device @@ -242,6 +254,7 @@ struct ubi_volume { unsigned int upd_marker:1; unsigned int updating:1; unsigned int changing_leb:1; + unsigned int direct_writes:1; #ifdef CONFIG_MTD_UBI_GLUEBI /* @@ -288,12 +301,13 @@ struct ubi_wl_entry; * @beb_rsvd_level: normal level of PEBs reserved for bad PEB handling * * @autoresize_vol_id: ID of the volume which has to be auto-resized at the end - * of UBI ititializetion + * of UBI initialization * @vtbl_slots: how many slots are available in the volume table * @vtbl_size: size of the volume table in bytes * @vtbl: in-RAM volume table copy * @volumes_mutex: protects on-flash volume table and serializes volume - * changes, like creation, deletion, update, re-size and re-name + * changes, like creation, deletion, update, re-size, + * re-name and set property * * @max_ec: current highest erase counter value * @mean_ec: current mean erase counter value @@ -306,18 +320,17 @@ struct ubi_wl_entry; * @used: RB-tree of used physical eraseblocks * @free: RB-tree of free physical eraseblocks * @scrub: RB-tree of physical eraseblocks which need scrubbing - * @prot: protection trees - * @prot.pnum: protection tree indexed by physical eraseblock numbers - * @prot.aec: protection tree indexed by absolute erase counter value - * @wl_lock: protects the @used, @free, @prot, @lookuptbl, @abs_ec, @move_from, - * @move_to, @move_to_put @erase_pending, @wl_scheduled, and @works - * fields + * @pq: protection queue (contain physical eraseblocks which are temporarily + * protected from the wear-leveling worker) + * @pq_head: protection queue head + * @wl_lock: protects the @used, @free, @pq, @pq_head, @lookuptbl, @move_from, + * @move_to, @move_to_put @erase_pending, @wl_scheduled and @works + * fields * @move_mutex: serializes eraseblock moves - * @work_sem: sycnhronizes the WL worker with use tasks + * @work_sem: synchronizes the WL worker with use tasks * @wl_scheduled: non-zero if the wear-leveling was scheduled * @lookuptbl: a table to quickly find a &struct ubi_wl_entry object for any * physical eraseblock - * @abs_ec: absolute erase counter * @move_from: physical eraseblock from where the data is being moved * @move_to: physical eraseblock where the data is being moved to * @move_to_put: if the "to" PEB was put @@ -351,11 +364,11 @@ struct ubi_wl_entry; * * @peb_buf1: a buffer of PEB size used for different purposes * @peb_buf2: another buffer of PEB size used for different purposes - * @buf_mutex: proptects @peb_buf1 and @peb_buf2 + * @buf_mutex: protects @peb_buf1 and @peb_buf2 * @ckvol_mutex: serializes static volume checking when opening - * @mult_mutex: serializes operations on multiple volumes, like re-nameing + * @mult_mutex: serializes operations on multiple volumes, like re-naming * @dbg_peb_buf: buffer of PEB size used for debugging - * @dbg_buf_mutex: proptects @dbg_peb_buf + * @dbg_buf_mutex: protects @dbg_peb_buf */ struct ubi_device { struct cdev cdev; @@ -392,16 +405,13 @@ struct ubi_device { struct rb_root used; struct rb_root free; struct rb_root scrub; - struct { - struct rb_root pnum; - struct rb_root aec; - } prot; + struct list_head pq[UBI_PROT_QUEUE_LEN]; + int pq_head; spinlock_t wl_lock; struct mutex move_mutex; struct rw_semaphore work_sem; int wl_scheduled; struct ubi_wl_entry **lookuptbl; - unsigned long long abs_ec; struct ubi_wl_entry *move_from; struct ubi_wl_entry *move_to; int move_to_put; @@ -442,9 +452,9 @@ struct ubi_device { }; extern struct kmem_cache *ubi_wl_entry_slab; -extern struct file_operations ubi_ctrl_cdev_operations; -extern struct file_operations ubi_cdev_operations; -extern struct file_operations ubi_vol_cdev_operations; +extern const struct file_operations ubi_ctrl_cdev_operations; +extern const struct file_operations ubi_cdev_operations; +extern const struct file_operations ubi_vol_cdev_operations; extern struct class *ubi_class; extern struct mutex ubi_devices_mutex; diff --git a/drivers/mtd/ubi/upd.c b/drivers/mtd/ubi/upd.c index 8b89cc18ff0b..6b4d1ae891ae 100644 --- a/drivers/mtd/ubi/upd.c +++ b/drivers/mtd/ubi/upd.c @@ -40,7 +40,7 @@ #include <linux/err.h> #include <linux/uaccess.h> -#include <asm/div64.h> +#include <linux/math64.h> #include "ubi.h" /** @@ -89,7 +89,6 @@ static int clear_update_marker(struct ubi_device *ubi, struct ubi_volume *vol, long long bytes) { int err; - uint64_t tmp; struct ubi_vtbl_record vtbl_rec; dbg_gen("clear update marker for volume %d", vol->vol_id); @@ -101,9 +100,9 @@ static int clear_update_marker(struct ubi_device *ubi, struct ubi_volume *vol, if (vol->vol_type == UBI_STATIC_VOLUME) { vol->corrupted = 0; - vol->used_bytes = tmp = bytes; - vol->last_eb_bytes = do_div(tmp, vol->usable_leb_size); - vol->used_ebs = tmp; + vol->used_bytes = bytes; + vol->used_ebs = div_u64_rem(bytes, vol->usable_leb_size, + &vol->last_eb_bytes); if (vol->last_eb_bytes) vol->used_ebs += 1; else @@ -131,7 +130,6 @@ int ubi_start_update(struct ubi_device *ubi, struct ubi_volume *vol, long long bytes) { int i, err; - uint64_t tmp; dbg_gen("start update of volume %d, %llu bytes", vol->vol_id, bytes); ubi_assert(!vol->updating && !vol->changing_leb); @@ -161,9 +159,8 @@ int ubi_start_update(struct ubi_device *ubi, struct ubi_volume *vol, if (!vol->upd_buf) return -ENOMEM; - tmp = bytes; - vol->upd_ebs = !!do_div(tmp, vol->usable_leb_size); - vol->upd_ebs += tmp; + vol->upd_ebs = div_u64(bytes + vol->usable_leb_size - 1, + vol->usable_leb_size); vol->upd_bytes = bytes; vol->upd_received = 0; return 0; @@ -282,7 +279,6 @@ static int write_leb(struct ubi_device *ubi, struct ubi_volume *vol, int lnum, int ubi_more_update_data(struct ubi_device *ubi, struct ubi_volume *vol, const void __user *buf, int count) { - uint64_t tmp; int lnum, offs, err = 0, len, to_write = count; dbg_gen("write %d of %lld bytes, %lld already passed", @@ -291,10 +287,7 @@ int ubi_more_update_data(struct ubi_device *ubi, struct ubi_volume *vol, if (ubi->ro_mode) return -EROFS; - tmp = vol->upd_received; - offs = do_div(tmp, vol->usable_leb_size); - lnum = tmp; - + lnum = div_u64_rem(vol->upd_received, vol->usable_leb_size, &offs); if (vol->upd_received + count > vol->upd_bytes) to_write = count = vol->upd_bytes - vol->upd_received; diff --git a/drivers/mtd/ubi/vmt.c b/drivers/mtd/ubi/vmt.c index 3531ca9a1e24..df5483562b7a 100644 --- a/drivers/mtd/ubi/vmt.c +++ b/drivers/mtd/ubi/vmt.c @@ -24,7 +24,7 @@ */ #include <linux/err.h> -#include <asm/div64.h> +#include <linux/math64.h> #include "ubi.h" #ifdef CONFIG_MTD_UBI_DEBUG_PARANOID @@ -205,7 +205,6 @@ int ubi_create_volume(struct ubi_device *ubi, struct ubi_mkvol_req *req) int i, err, vol_id = req->vol_id, do_free = 1; struct ubi_volume *vol; struct ubi_vtbl_record vtbl_rec; - uint64_t bytes; dev_t dev; if (ubi->ro_mode) @@ -255,10 +254,8 @@ int ubi_create_volume(struct ubi_device *ubi, struct ubi_mkvol_req *req) /* Calculate how many eraseblocks are requested */ vol->usable_leb_size = ubi->leb_size - ubi->leb_size % req->alignment; - bytes = req->bytes; - if (do_div(bytes, vol->usable_leb_size)) - vol->reserved_pebs = 1; - vol->reserved_pebs += bytes; + vol->reserved_pebs += div_u64(req->bytes + vol->usable_leb_size - 1, + vol->usable_leb_size); /* Reserve physical eraseblocks */ if (vol->reserved_pebs > ubi->avail_pebs) { @@ -301,10 +298,10 @@ int ubi_create_volume(struct ubi_device *ubi, struct ubi_mkvol_req *req) vol->used_bytes = (long long)vol->used_ebs * vol->usable_leb_size; } else { - bytes = vol->used_bytes; - vol->last_eb_bytes = do_div(bytes, vol->usable_leb_size); - vol->used_ebs = bytes; - if (vol->last_eb_bytes) + vol->used_ebs = div_u64_rem(vol->used_bytes, + vol->usable_leb_size, + &vol->last_eb_bytes); + if (vol->last_eb_bytes != 0) vol->used_ebs += 1; else vol->last_eb_bytes = vol->usable_leb_size; @@ -329,7 +326,7 @@ int ubi_create_volume(struct ubi_device *ubi, struct ubi_mkvol_req *req) vol->dev.devt = dev; vol->dev.class = ubi_class; - sprintf(&vol->dev.bus_id[0], "%s_%d", ubi->ubi_name, vol->vol_id); + dev_set_name(&vol->dev, "%s_%d", ubi->ubi_name, vol->vol_id); err = device_register(&vol->dev); if (err) { ubi_err("cannot register device"); @@ -678,7 +675,7 @@ int ubi_add_volume(struct ubi_device *ubi, struct ubi_volume *vol) vol->dev.parent = &ubi->dev; vol->dev.devt = dev; vol->dev.class = ubi_class; - sprintf(&vol->dev.bus_id[0], "%s_%d", ubi->ubi_name, vol->vol_id); + dev_set_name(&vol->dev, "%s_%d", ubi->ubi_name, vol->vol_id); err = device_register(&vol->dev); if (err) goto out_gluebi; diff --git a/drivers/mtd/ubi/vtbl.c b/drivers/mtd/ubi/vtbl.c index 333c8941552f..1afc61e7455d 100644 --- a/drivers/mtd/ubi/vtbl.c +++ b/drivers/mtd/ubi/vtbl.c @@ -577,7 +577,7 @@ static int init_volumes(struct ubi_device *ubi, const struct ubi_scan_info *si, if (vtbl[i].flags & UBI_VTBL_AUTORESIZE_FLG) { /* Auto re-size flag may be set only for one volume */ if (ubi->autoresize_vol_id != -1) { - ubi_err("more then one auto-resize volume (%d " + ubi_err("more than one auto-resize volume (%d " "and %d)", ubi->autoresize_vol_id, i); kfree(vol); return -EINVAL; diff --git a/drivers/mtd/ubi/wl.c b/drivers/mtd/ubi/wl.c index dcb6dac1dc54..891534f8210d 100644 --- a/drivers/mtd/ubi/wl.c +++ b/drivers/mtd/ubi/wl.c @@ -22,7 +22,7 @@ * UBI wear-leveling sub-system. * * This sub-system is responsible for wear-leveling. It works in terms of - * physical* eraseblocks and erase counters and knows nothing about logical + * physical eraseblocks and erase counters and knows nothing about logical * eraseblocks, volumes, etc. From this sub-system's perspective all physical * eraseblocks are of two types - used and free. Used physical eraseblocks are * those that were "get" by the 'ubi_wl_get_peb()' function, and free physical @@ -55,8 +55,39 @@ * * As it was said, for the UBI sub-system all physical eraseblocks are either * "free" or "used". Free eraseblock are kept in the @wl->free RB-tree, while - * used eraseblocks are kept in a set of different RB-trees: @wl->used, - * @wl->prot.pnum, @wl->prot.aec, and @wl->scrub. + * used eraseblocks are kept in @wl->used or @wl->scrub RB-trees, or + * (temporarily) in the @wl->pq queue. + * + * When the WL sub-system returns a physical eraseblock, the physical + * eraseblock is protected from being moved for some "time". For this reason, + * the physical eraseblock is not directly moved from the @wl->free tree to the + * @wl->used tree. There is a protection queue in between where this + * physical eraseblock is temporarily stored (@wl->pq). + * + * All this protection stuff is needed because: + * o we don't want to move physical eraseblocks just after we have given them + * to the user; instead, we first want to let users fill them up with data; + * + * o there is a chance that the user will put the physical eraseblock very + * soon, so it makes sense not to move it for some time, but wait; this is + * especially important in case of "short term" physical eraseblocks. + * + * Physical eraseblocks stay protected only for limited time. But the "time" is + * measured in erase cycles in this case. This is implemented with help of the + * protection queue. Eraseblocks are put to the tail of this queue when they + * are returned by the 'ubi_wl_get_peb()', and eraseblocks are removed from the + * head of the queue on each erase operation (for any eraseblock). So the + * length of the queue defines how may (global) erase cycles PEBs are protected. + * + * To put it differently, each physical eraseblock has 2 main states: free and + * used. The former state corresponds to the @wl->free tree. The latter state + * is split up on several sub-states: + * o the WL movement is allowed (@wl->used tree); + * o the WL movement is temporarily prohibited (@wl->pq queue); + * o scrubbing is needed (@wl->scrub tree). + * + * Depending on the sub-state, wear-leveling entries of the used physical + * eraseblocks may be kept in one of those structures. * * Note, in this implementation, we keep a small in-RAM object for each physical * eraseblock. This is surely not a scalable solution. But it appears to be good @@ -70,9 +101,6 @@ * target PEB, we pick a PEB with the highest EC if our PEB is "old" and we * pick target PEB with an average EC if our PEB is not very "old". This is a * room for future re-works of the WL sub-system. - * - * Note: the stuff with protection trees looks too complex and is difficult to - * understand. Should be fixed. */ #include <linux/slab.h> @@ -85,14 +113,6 @@ #define WL_RESERVED_PEBS 1 /* - * How many erase cycles are short term, unknown, and long term physical - * eraseblocks protected. - */ -#define ST_PROTECTION 16 -#define U_PROTECTION 10 -#define LT_PROTECTION 4 - -/* * Maximum difference between two erase counters. If this threshold is * exceeded, the WL sub-system starts moving data from used physical * eraseblocks with low erase counter to free physical eraseblocks with high @@ -108,7 +128,7 @@ * situation when the picked physical eraseblock is constantly erased after the * data is written to it. So, we have a constant which limits the highest erase * counter of the free physical eraseblock to pick. Namely, the WL sub-system - * does not pick eraseblocks with erase counter greater then the lowest erase + * does not pick eraseblocks with erase counter greater than the lowest erase * counter plus %WL_FREE_MAX_DIFF. */ #define WL_FREE_MAX_DIFF (2*UBI_WL_THRESHOLD) @@ -120,64 +140,9 @@ #define WL_MAX_FAILURES 32 /** - * struct ubi_wl_prot_entry - PEB protection entry. - * @rb_pnum: link in the @wl->prot.pnum RB-tree - * @rb_aec: link in the @wl->prot.aec RB-tree - * @abs_ec: the absolute erase counter value when the protection ends - * @e: the wear-leveling entry of the physical eraseblock under protection - * - * When the WL sub-system returns a physical eraseblock, the physical - * eraseblock is protected from being moved for some "time". For this reason, - * the physical eraseblock is not directly moved from the @wl->free tree to the - * @wl->used tree. There is one more tree in between where this physical - * eraseblock is temporarily stored (@wl->prot). - * - * All this protection stuff is needed because: - * o we don't want to move physical eraseblocks just after we have given them - * to the user; instead, we first want to let users fill them up with data; - * - * o there is a chance that the user will put the physical eraseblock very - * soon, so it makes sense not to move it for some time, but wait; this is - * especially important in case of "short term" physical eraseblocks. - * - * Physical eraseblocks stay protected only for limited time. But the "time" is - * measured in erase cycles in this case. This is implemented with help of the - * absolute erase counter (@wl->abs_ec). When it reaches certain value, the - * physical eraseblocks are moved from the protection trees (@wl->prot.*) to - * the @wl->used tree. - * - * Protected physical eraseblocks are searched by physical eraseblock number - * (when they are put) and by the absolute erase counter (to check if it is - * time to move them to the @wl->used tree). So there are actually 2 RB-trees - * storing the protected physical eraseblocks: @wl->prot.pnum and - * @wl->prot.aec. They are referred to as the "protection" trees. The - * first one is indexed by the physical eraseblock number. The second one is - * indexed by the absolute erase counter. Both trees store - * &struct ubi_wl_prot_entry objects. - * - * Each physical eraseblock has 2 main states: free and used. The former state - * corresponds to the @wl->free tree. The latter state is split up on several - * sub-states: - * o the WL movement is allowed (@wl->used tree); - * o the WL movement is temporarily prohibited (@wl->prot.pnum and - * @wl->prot.aec trees); - * o scrubbing is needed (@wl->scrub tree). - * - * Depending on the sub-state, wear-leveling entries of the used physical - * eraseblocks may be kept in one of those trees. - */ -struct ubi_wl_prot_entry { - struct rb_node rb_pnum; - struct rb_node rb_aec; - unsigned long long abs_ec; - struct ubi_wl_entry *e; -}; - -/** * struct ubi_work - UBI work description data structure. * @list: a link in the list of pending works * @func: worker function - * @priv: private data of the worker function * @e: physical eraseblock to erase * @torture: if the physical eraseblock has to be tortured * @@ -198,9 +163,11 @@ struct ubi_work { static int paranoid_check_ec(struct ubi_device *ubi, int pnum, int ec); static int paranoid_check_in_wl_tree(struct ubi_wl_entry *e, struct rb_root *root); +static int paranoid_check_in_pq(struct ubi_device *ubi, struct ubi_wl_entry *e); #else #define paranoid_check_ec(ubi, pnum, ec) 0 #define paranoid_check_in_wl_tree(e, root) +#define paranoid_check_in_pq(ubi, e) 0 #endif /** @@ -220,7 +187,7 @@ static void wl_tree_add(struct ubi_wl_entry *e, struct rb_root *root) struct ubi_wl_entry *e1; parent = *p; - e1 = rb_entry(parent, struct ubi_wl_entry, rb); + e1 = rb_entry(parent, struct ubi_wl_entry, u.rb); if (e->ec < e1->ec) p = &(*p)->rb_left; @@ -235,8 +202,8 @@ static void wl_tree_add(struct ubi_wl_entry *e, struct rb_root *root) } } - rb_link_node(&e->rb, parent, p); - rb_insert_color(&e->rb, root); + rb_link_node(&e->u.rb, parent, p); + rb_insert_color(&e->u.rb, root); } /** @@ -331,7 +298,7 @@ static int in_wl_tree(struct ubi_wl_entry *e, struct rb_root *root) while (p) { struct ubi_wl_entry *e1; - e1 = rb_entry(p, struct ubi_wl_entry, rb); + e1 = rb_entry(p, struct ubi_wl_entry, u.rb); if (e->pnum == e1->pnum) { ubi_assert(e == e1); @@ -355,50 +322,24 @@ static int in_wl_tree(struct ubi_wl_entry *e, struct rb_root *root) } /** - * prot_tree_add - add physical eraseblock to protection trees. + * prot_queue_add - add physical eraseblock to the protection queue. * @ubi: UBI device description object * @e: the physical eraseblock to add - * @pe: protection entry object to use - * @abs_ec: absolute erase counter value when this physical eraseblock has - * to be removed from the protection trees. * - * @wl->lock has to be locked. + * This function adds @e to the tail of the protection queue @ubi->pq, where + * @e will stay for %UBI_PROT_QUEUE_LEN erase operations and will be + * temporarily protected from the wear-leveling worker. Note, @wl->lock has to + * be locked. */ -static void prot_tree_add(struct ubi_device *ubi, struct ubi_wl_entry *e, - struct ubi_wl_prot_entry *pe, int abs_ec) +static void prot_queue_add(struct ubi_device *ubi, struct ubi_wl_entry *e) { - struct rb_node **p, *parent = NULL; - struct ubi_wl_prot_entry *pe1; - - pe->e = e; - pe->abs_ec = ubi->abs_ec + abs_ec; - - p = &ubi->prot.pnum.rb_node; - while (*p) { - parent = *p; - pe1 = rb_entry(parent, struct ubi_wl_prot_entry, rb_pnum); - - if (e->pnum < pe1->e->pnum) - p = &(*p)->rb_left; - else - p = &(*p)->rb_right; - } - rb_link_node(&pe->rb_pnum, parent, p); - rb_insert_color(&pe->rb_pnum, &ubi->prot.pnum); - - p = &ubi->prot.aec.rb_node; - parent = NULL; - while (*p) { - parent = *p; - pe1 = rb_entry(parent, struct ubi_wl_prot_entry, rb_aec); + int pq_tail = ubi->pq_head - 1; - if (pe->abs_ec < pe1->abs_ec) - p = &(*p)->rb_left; - else - p = &(*p)->rb_right; - } - rb_link_node(&pe->rb_aec, parent, p); - rb_insert_color(&pe->rb_aec, &ubi->prot.aec); + if (pq_tail < 0) + pq_tail = UBI_PROT_QUEUE_LEN - 1; + ubi_assert(pq_tail >= 0 && pq_tail < UBI_PROT_QUEUE_LEN); + list_add_tail(&e->u.list, &ubi->pq[pq_tail]); + dbg_wl("added PEB %d EC %d to the protection queue", e->pnum, e->ec); } /** @@ -414,14 +355,14 @@ static struct ubi_wl_entry *find_wl_entry(struct rb_root *root, int max) struct rb_node *p; struct ubi_wl_entry *e; - e = rb_entry(rb_first(root), struct ubi_wl_entry, rb); + e = rb_entry(rb_first(root), struct ubi_wl_entry, u.rb); max += e->ec; p = root->rb_node; while (p) { struct ubi_wl_entry *e1; - e1 = rb_entry(p, struct ubi_wl_entry, rb); + e1 = rb_entry(p, struct ubi_wl_entry, u.rb); if (e1->ec >= max) p = p->rb_left; else { @@ -443,17 +384,12 @@ static struct ubi_wl_entry *find_wl_entry(struct rb_root *root, int max) */ int ubi_wl_get_peb(struct ubi_device *ubi, int dtype) { - int err, protect, medium_ec; + int err, medium_ec; struct ubi_wl_entry *e, *first, *last; - struct ubi_wl_prot_entry *pe; ubi_assert(dtype == UBI_LONGTERM || dtype == UBI_SHORTTERM || dtype == UBI_UNKNOWN); - pe = kmalloc(sizeof(struct ubi_wl_prot_entry), GFP_NOFS); - if (!pe) - return -ENOMEM; - retry: spin_lock(&ubi->wl_lock); if (!ubi->free.rb_node) { @@ -461,16 +397,13 @@ retry: ubi_assert(list_empty(&ubi->works)); ubi_err("no free eraseblocks"); spin_unlock(&ubi->wl_lock); - kfree(pe); return -ENOSPC; } spin_unlock(&ubi->wl_lock); err = produce_free_peb(ubi); - if (err < 0) { - kfree(pe); + if (err < 0) return err; - } goto retry; } @@ -483,7 +416,6 @@ retry: * %WL_FREE_MAX_DIFF. */ e = find_wl_entry(&ubi->free, WL_FREE_MAX_DIFF); - protect = LT_PROTECTION; break; case UBI_UNKNOWN: /* @@ -492,81 +424,63 @@ retry: * eraseblock with erase counter greater or equivalent than the * lowest erase counter plus %WL_FREE_MAX_DIFF. */ - first = rb_entry(rb_first(&ubi->free), struct ubi_wl_entry, rb); - last = rb_entry(rb_last(&ubi->free), struct ubi_wl_entry, rb); + first = rb_entry(rb_first(&ubi->free), struct ubi_wl_entry, + u.rb); + last = rb_entry(rb_last(&ubi->free), struct ubi_wl_entry, u.rb); if (last->ec - first->ec < WL_FREE_MAX_DIFF) e = rb_entry(ubi->free.rb_node, - struct ubi_wl_entry, rb); + struct ubi_wl_entry, u.rb); else { medium_ec = (first->ec + WL_FREE_MAX_DIFF)/2; e = find_wl_entry(&ubi->free, medium_ec); } - protect = U_PROTECTION; break; case UBI_SHORTTERM: /* * For short term data we pick a physical eraseblock with the * lowest erase counter as we expect it will be erased soon. */ - e = rb_entry(rb_first(&ubi->free), struct ubi_wl_entry, rb); - protect = ST_PROTECTION; + e = rb_entry(rb_first(&ubi->free), struct ubi_wl_entry, u.rb); break; default: - protect = 0; - e = NULL; BUG(); } + paranoid_check_in_wl_tree(e, &ubi->free); + /* - * Move the physical eraseblock to the protection trees where it will + * Move the physical eraseblock to the protection queue where it will * be protected from being moved for some time. */ - paranoid_check_in_wl_tree(e, &ubi->free); - rb_erase(&e->rb, &ubi->free); - prot_tree_add(ubi, e, pe, protect); - - dbg_wl("PEB %d EC %d, protection %d", e->pnum, e->ec, protect); + rb_erase(&e->u.rb, &ubi->free); + dbg_wl("PEB %d EC %d", e->pnum, e->ec); + prot_queue_add(ubi, e); spin_unlock(&ubi->wl_lock); - return e->pnum; } /** - * prot_tree_del - remove a physical eraseblock from the protection trees + * prot_queue_del - remove a physical eraseblock from the protection queue. * @ubi: UBI device description object * @pnum: the physical eraseblock to remove * - * This function returns PEB @pnum from the protection trees and returns zero - * in case of success and %-ENODEV if the PEB was not found in the protection - * trees. + * This function deletes PEB @pnum from the protection queue and returns zero + * in case of success and %-ENODEV if the PEB was not found. */ -static int prot_tree_del(struct ubi_device *ubi, int pnum) +static int prot_queue_del(struct ubi_device *ubi, int pnum) { - struct rb_node *p; - struct ubi_wl_prot_entry *pe = NULL; - - p = ubi->prot.pnum.rb_node; - while (p) { - - pe = rb_entry(p, struct ubi_wl_prot_entry, rb_pnum); - - if (pnum == pe->e->pnum) - goto found; + struct ubi_wl_entry *e; - if (pnum < pe->e->pnum) - p = p->rb_left; - else - p = p->rb_right; - } + e = ubi->lookuptbl[pnum]; + if (!e) + return -ENODEV; - return -ENODEV; + if (paranoid_check_in_pq(ubi, e)) + return -ENODEV; -found: - ubi_assert(pe->e->pnum == pnum); - rb_erase(&pe->rb_aec, &ubi->prot.aec); - rb_erase(&pe->rb_pnum, &ubi->prot.pnum); - kfree(pe); + list_del(&e->u.list); + dbg_wl("deleted PEB %d from the protection queue", e->pnum); return 0; } @@ -632,47 +546,47 @@ out_free: } /** - * check_protection_over - check if it is time to stop protecting some PEBs. + * serve_prot_queue - check if it is time to stop protecting PEBs. * @ubi: UBI device description object * - * This function is called after each erase operation, when the absolute erase - * counter is incremented, to check if some physical eraseblock have not to be - * protected any longer. These physical eraseblocks are moved from the - * protection trees to the used tree. + * This function is called after each erase operation and removes PEBs from the + * tail of the protection queue. These PEBs have been protected for long enough + * and should be moved to the used tree. */ -static void check_protection_over(struct ubi_device *ubi) +static void serve_prot_queue(struct ubi_device *ubi) { - struct ubi_wl_prot_entry *pe; + struct ubi_wl_entry *e, *tmp; + int count; /* * There may be several protected physical eraseblock to remove, * process them all. */ - while (1) { - spin_lock(&ubi->wl_lock); - if (!ubi->prot.aec.rb_node) { - spin_unlock(&ubi->wl_lock); - break; - } - - pe = rb_entry(rb_first(&ubi->prot.aec), - struct ubi_wl_prot_entry, rb_aec); +repeat: + count = 0; + spin_lock(&ubi->wl_lock); + list_for_each_entry_safe(e, tmp, &ubi->pq[ubi->pq_head], u.list) { + dbg_wl("PEB %d EC %d protection over, move to used tree", + e->pnum, e->ec); - if (pe->abs_ec > ubi->abs_ec) { + list_del(&e->u.list); + wl_tree_add(e, &ubi->used); + if (count++ > 32) { + /* + * Let's be nice and avoid holding the spinlock for + * too long. + */ spin_unlock(&ubi->wl_lock); - break; + cond_resched(); + goto repeat; } - - dbg_wl("PEB %d protection over, abs_ec %llu, PEB abs_ec %llu", - pe->e->pnum, ubi->abs_ec, pe->abs_ec); - rb_erase(&pe->rb_aec, &ubi->prot.aec); - rb_erase(&pe->rb_pnum, &ubi->prot.pnum); - wl_tree_add(pe->e, &ubi->used); - spin_unlock(&ubi->wl_lock); - - kfree(pe); - cond_resched(); } + + ubi->pq_head += 1; + if (ubi->pq_head == UBI_PROT_QUEUE_LEN) + ubi->pq_head = 0; + ubi_assert(ubi->pq_head >= 0 && ubi->pq_head < UBI_PROT_QUEUE_LEN); + spin_unlock(&ubi->wl_lock); } /** @@ -680,8 +594,8 @@ static void check_protection_over(struct ubi_device *ubi) * @ubi: UBI device description object * @wrk: the work to schedule * - * This function enqueues a work defined by @wrk to the tail of the pending - * works list. + * This function adds a work defined by @wrk to the tail of the pending works + * list. */ static void schedule_ubi_work(struct ubi_device *ubi, struct ubi_work *wrk) { @@ -739,13 +653,11 @@ static int schedule_erase(struct ubi_device *ubi, struct ubi_wl_entry *e, static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, int cancel) { - int err, put = 0, scrubbing = 0, protect = 0; - struct ubi_wl_prot_entry *uninitialized_var(pe); + int err, scrubbing = 0, torture = 0; struct ubi_wl_entry *e1, *e2; struct ubi_vid_hdr *vid_hdr; kfree(wrk); - if (cancel) return 0; @@ -781,7 +693,7 @@ static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, * highly worn-out free physical eraseblock. If the erase * counters differ much enough, start wear-leveling. */ - e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, rb); + e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, u.rb); e2 = find_wl_entry(&ubi->free, WL_FREE_MAX_DIFF); if (!(e2->ec - e1->ec >= UBI_WL_THRESHOLD)) { @@ -790,21 +702,21 @@ static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, goto out_cancel; } paranoid_check_in_wl_tree(e1, &ubi->used); - rb_erase(&e1->rb, &ubi->used); + rb_erase(&e1->u.rb, &ubi->used); dbg_wl("move PEB %d EC %d to PEB %d EC %d", e1->pnum, e1->ec, e2->pnum, e2->ec); } else { /* Perform scrubbing */ scrubbing = 1; - e1 = rb_entry(rb_first(&ubi->scrub), struct ubi_wl_entry, rb); + e1 = rb_entry(rb_first(&ubi->scrub), struct ubi_wl_entry, u.rb); e2 = find_wl_entry(&ubi->free, WL_FREE_MAX_DIFF); paranoid_check_in_wl_tree(e1, &ubi->scrub); - rb_erase(&e1->rb, &ubi->scrub); + rb_erase(&e1->u.rb, &ubi->scrub); dbg_wl("scrub PEB %d to PEB %d", e1->pnum, e2->pnum); } paranoid_check_in_wl_tree(e2, &ubi->free); - rb_erase(&e2->rb, &ubi->free); + rb_erase(&e2->u.rb, &ubi->free); ubi->move_from = e1; ubi->move_to = e2; spin_unlock(&ubi->wl_lock); @@ -844,46 +756,67 @@ static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, err = ubi_eba_copy_leb(ubi, e1->pnum, e2->pnum, vid_hdr); if (err) { - + if (err == -EAGAIN) + goto out_not_moved; if (err < 0) goto out_error; - if (err == 1) + if (err == 2) { + /* Target PEB write error, torture it */ + torture = 1; goto out_not_moved; + } /* - * For some reason the LEB was not moved - it might be because - * the volume is being deleted. We should prevent this PEB from - * being selected for wear-levelling movement for some "time", - * so put it to the protection tree. + * The LEB has not been moved because the volume is being + * deleted or the PEB has been put meanwhile. We should prevent + * this PEB from being selected for wear-leveling movement + * again, so put it to the protection queue. */ - dbg_wl("cancelled moving PEB %d", e1->pnum); - pe = kmalloc(sizeof(struct ubi_wl_prot_entry), GFP_NOFS); - if (!pe) { - err = -ENOMEM; - goto out_error; - } + dbg_wl("canceled moving PEB %d", e1->pnum); + ubi_assert(err == 1); + + ubi_free_vid_hdr(ubi, vid_hdr); + vid_hdr = NULL; + + spin_lock(&ubi->wl_lock); + prot_queue_add(ubi, e1); + ubi_assert(!ubi->move_to_put); + ubi->move_from = ubi->move_to = NULL; + ubi->wl_scheduled = 0; + spin_unlock(&ubi->wl_lock); - protect = 1; + e1 = NULL; + err = schedule_erase(ubi, e2, 0); + if (err) + goto out_error; + mutex_unlock(&ubi->move_mutex); + return 0; } + /* The PEB has been successfully moved */ ubi_free_vid_hdr(ubi, vid_hdr); - if (scrubbing && !protect) + vid_hdr = NULL; + if (scrubbing) ubi_msg("scrubbed PEB %d, data moved to PEB %d", e1->pnum, e2->pnum); spin_lock(&ubi->wl_lock); - if (protect) - prot_tree_add(ubi, e1, pe, protect); - if (!ubi->move_to_put) + if (!ubi->move_to_put) { wl_tree_add(e2, &ubi->used); - else - put = 1; + e2 = NULL; + } ubi->move_from = ubi->move_to = NULL; ubi->move_to_put = ubi->wl_scheduled = 0; spin_unlock(&ubi->wl_lock); - if (put) { + err = schedule_erase(ubi, e1, 0); + if (err) { + e1 = NULL; + goto out_error; + } + + if (e2) { /* * Well, the target PEB was put meanwhile, schedule it for * erasure. @@ -894,13 +827,6 @@ static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, goto out_error; } - if (!protect) { - err = schedule_erase(ubi, e1, 0); - if (err) - goto out_error; - } - - dbg_wl("done"); mutex_unlock(&ubi->move_mutex); return 0; @@ -908,20 +834,24 @@ static int wear_leveling_worker(struct ubi_device *ubi, struct ubi_work *wrk, /* * For some reasons the LEB was not moved, might be an error, might be * something else. @e1 was not changed, so return it back. @e2 might - * be changed, schedule it for erasure. + * have been changed, schedule it for erasure. */ out_not_moved: + dbg_wl("canceled moving PEB %d", e1->pnum); ubi_free_vid_hdr(ubi, vid_hdr); + vid_hdr = NULL; spin_lock(&ubi->wl_lock); if (scrubbing) wl_tree_add(e1, &ubi->scrub); else wl_tree_add(e1, &ubi->used); + ubi_assert(!ubi->move_to_put); ubi->move_from = ubi->move_to = NULL; - ubi->move_to_put = ubi->wl_scheduled = 0; + ubi->wl_scheduled = 0; spin_unlock(&ubi->wl_lock); - err = schedule_erase(ubi, e2, 0); + e1 = NULL; + err = schedule_erase(ubi, e2, torture); if (err) goto out_error; @@ -938,8 +868,10 @@ out_error: ubi->move_to_put = ubi->wl_scheduled = 0; spin_unlock(&ubi->wl_lock); - kmem_cache_free(ubi_wl_entry_slab, e1); - kmem_cache_free(ubi_wl_entry_slab, e2); + if (e1) + kmem_cache_free(ubi_wl_entry_slab, e1); + if (e2) + kmem_cache_free(ubi_wl_entry_slab, e2); ubi_ro_mode(ubi); mutex_unlock(&ubi->move_mutex); @@ -985,10 +917,10 @@ static int ensure_wear_leveling(struct ubi_device *ubi) /* * We schedule wear-leveling only if the difference between the * lowest erase counter of used physical eraseblocks and a high - * erase counter of free physical eraseblocks is greater then + * erase counter of free physical eraseblocks is greater than * %UBI_WL_THRESHOLD. */ - e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, rb); + e1 = rb_entry(rb_first(&ubi->used), struct ubi_wl_entry, u.rb); e2 = find_wl_entry(&ubi->free, WL_FREE_MAX_DIFF); if (!(e2->ec - e1->ec >= UBI_WL_THRESHOLD)) @@ -1050,7 +982,6 @@ static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, kfree(wl_wrk); spin_lock(&ubi->wl_lock); - ubi->abs_ec += 1; wl_tree_add(e, &ubi->free); spin_unlock(&ubi->wl_lock); @@ -1058,7 +989,7 @@ static int erase_worker(struct ubi_device *ubi, struct ubi_work *wl_wrk, * One more erase operation has happened, take care about * protected physical eraseblocks. */ - check_protection_over(ubi); + serve_prot_queue(ubi); /* And take care about wear-leveling */ err = ensure_wear_leveling(ubi); @@ -1190,12 +1121,12 @@ retry: } else { if (in_wl_tree(e, &ubi->used)) { paranoid_check_in_wl_tree(e, &ubi->used); - rb_erase(&e->rb, &ubi->used); + rb_erase(&e->u.rb, &ubi->used); } else if (in_wl_tree(e, &ubi->scrub)) { paranoid_check_in_wl_tree(e, &ubi->scrub); - rb_erase(&e->rb, &ubi->scrub); + rb_erase(&e->u.rb, &ubi->scrub); } else { - err = prot_tree_del(ubi, e->pnum); + err = prot_queue_del(ubi, e->pnum); if (err) { ubi_err("PEB %d not found", pnum); ubi_ro_mode(ubi); @@ -1255,11 +1186,11 @@ retry: if (in_wl_tree(e, &ubi->used)) { paranoid_check_in_wl_tree(e, &ubi->used); - rb_erase(&e->rb, &ubi->used); + rb_erase(&e->u.rb, &ubi->used); } else { int err; - err = prot_tree_del(ubi, e->pnum); + err = prot_queue_del(ubi, e->pnum); if (err) { ubi_err("PEB %d not found", pnum); ubi_ro_mode(ubi); @@ -1290,7 +1221,7 @@ int ubi_wl_flush(struct ubi_device *ubi) int err; /* - * Erase while the pending works queue is not empty, but not more then + * Erase while the pending works queue is not empty, but not more than * the number of currently pending works. */ dbg_wl("flush (%d pending works)", ubi->works_count); @@ -1308,7 +1239,7 @@ int ubi_wl_flush(struct ubi_device *ubi) up_write(&ubi->work_sem); /* - * And in case last was the WL worker and it cancelled the LEB + * And in case last was the WL worker and it canceled the LEB * movement, flush again. */ while (ubi->works_count) { @@ -1337,11 +1268,11 @@ static void tree_destroy(struct rb_root *root) else if (rb->rb_right) rb = rb->rb_right; else { - e = rb_entry(rb, struct ubi_wl_entry, rb); + e = rb_entry(rb, struct ubi_wl_entry, u.rb); rb = rb_parent(rb); if (rb) { - if (rb->rb_left == &e->rb) + if (rb->rb_left == &e->u.rb) rb->rb_left = NULL; else rb->rb_right = NULL; @@ -1436,15 +1367,13 @@ static void cancel_pending(struct ubi_device *ubi) */ int ubi_wl_init_scan(struct ubi_device *ubi, struct ubi_scan_info *si) { - int err; + int err, i; struct rb_node *rb1, *rb2; struct ubi_scan_volume *sv; struct ubi_scan_leb *seb, *tmp; struct ubi_wl_entry *e; - ubi->used = ubi->free = ubi->scrub = RB_ROOT; - ubi->prot.pnum = ubi->prot.aec = RB_ROOT; spin_lock_init(&ubi->wl_lock); mutex_init(&ubi->move_mutex); init_rwsem(&ubi->work_sem); @@ -1458,6 +1387,10 @@ int ubi_wl_init_scan(struct ubi_device *ubi, struct ubi_scan_info *si) if (!ubi->lookuptbl) return err; + for (i = 0; i < UBI_PROT_QUEUE_LEN; i++) + INIT_LIST_HEAD(&ubi->pq[i]); + ubi->pq_head = 0; + list_for_each_entry_safe(seb, tmp, &si->erase, u.list) { cond_resched(); @@ -1552,33 +1485,18 @@ out_free: } /** - * protection_trees_destroy - destroy the protection RB-trees. + * protection_queue_destroy - destroy the protection queue. * @ubi: UBI device description object */ -static void protection_trees_destroy(struct ubi_device *ubi) +static void protection_queue_destroy(struct ubi_device *ubi) { - struct rb_node *rb; - struct ubi_wl_prot_entry *pe; - - rb = ubi->prot.aec.rb_node; - while (rb) { - if (rb->rb_left) - rb = rb->rb_left; - else if (rb->rb_right) - rb = rb->rb_right; - else { - pe = rb_entry(rb, struct ubi_wl_prot_entry, rb_aec); - - rb = rb_parent(rb); - if (rb) { - if (rb->rb_left == &pe->rb_aec) - rb->rb_left = NULL; - else - rb->rb_right = NULL; - } + int i; + struct ubi_wl_entry *e, *tmp; - kmem_cache_free(ubi_wl_entry_slab, pe->e); - kfree(pe); + for (i = 0; i < UBI_PROT_QUEUE_LEN; ++i) { + list_for_each_entry_safe(e, tmp, &ubi->pq[i], u.list) { + list_del(&e->u.list); + kmem_cache_free(ubi_wl_entry_slab, e); } } } @@ -1591,7 +1509,7 @@ void ubi_wl_close(struct ubi_device *ubi) { dbg_wl("close the WL sub-system"); cancel_pending(ubi); - protection_trees_destroy(ubi); + protection_queue_destroy(ubi); tree_destroy(&ubi->used); tree_destroy(&ubi->free); tree_destroy(&ubi->scrub); @@ -1661,4 +1579,27 @@ static int paranoid_check_in_wl_tree(struct ubi_wl_entry *e, return 1; } +/** + * paranoid_check_in_pq - check if wear-leveling entry is in the protection + * queue. + * @ubi: UBI device description object + * @e: the wear-leveling entry to check + * + * This function returns zero if @e is in @ubi->pq and %1 if it is not. + */ +static int paranoid_check_in_pq(struct ubi_device *ubi, struct ubi_wl_entry *e) +{ + struct ubi_wl_entry *p; + int i; + + for (i = 0; i < UBI_PROT_QUEUE_LEN; ++i) + list_for_each_entry(p, &ubi->pq[i], u.list) + if (p == e) + return 0; + + ubi_err("paranoid check failed for PEB %d, EC %d, Protect queue", + e->pnum, e->ec); + ubi_dbg_dump_stack(); + return 1; +} #endif /* CONFIG_MTD_UBI_DEBUG_PARANOID */ |