mirror of
https://github.com/morgan9e/systemd
synced 2026-04-14 08:25:20 +09:00
e582484789
udev's block device locking protocol has one pitfall not even the example in the documentation got right so far (even though this is explained in all detail above): udev's rescanning is only triggered when an fd that is opened for writing is closed. This means that if a separate locking fd is opened on a block device – one that is maintained independently of the fd actually used for writing – it must be opened for writing too, so that closing the lock definitely triggers a rescan. This matters in cases where the lock fd is kept for longer than the fd used for writing to disk. (Because otherwise udev might get the IN_CLOSE_WRITE event, but when it tries to rescan will find the device locked, and never retry because no IN_CLOSE_WRITE is triggred anymore.) Let's fix that across the codebase, at 4 places: 1. in makefs (a lock fd is kept, and mkfs then invoked as child, which uses a different fd, and the lock fd is closed only once the child died) 2. in udevadm lock (embarassing!): which is intended to be used to wrap tools that modify disk contents, very similar to the makefs case. The lock is also kept until after the tool exited. 3. In storagetm: the kernel nvme-tcp layer writes to the device directly, we just keep a lock fd. 4. the example in BLOCK_DEVICE_LOCKING.md
61 lines
2.6 KiB
C
61 lines
2.6 KiB
C
/* SPDX-License-Identifier: LGPL-2.1-or-later */
|
|
#pragma once
|
|
|
|
#include <sys/sysmacros.h>
|
|
|
|
#include "shared-forward.h"
|
|
#include "stdio-util.h"
|
|
|
|
#define SYS_BLOCK_PATH_MAX(suffix) \
|
|
(STRLEN("/sys/dev/block/") + DECIMAL_STR_MAX(dev_t) + 1 + DECIMAL_STR_MAX(dev_t) + STRLEN(suffix))
|
|
#define xsprintf_sys_block_path(buf, suffix, devno) \
|
|
xsprintf(buf, "/sys/dev/block/%u:%u%s", major(devno), minor(devno), suffix ?: "")
|
|
|
|
typedef enum BlockDeviceLookupFlags {
|
|
BLOCK_DEVICE_LOOKUP_WHOLE_DISK = 1 << 0, /* whole block device, e.g. sda, nvme0n1, or loop0. */
|
|
BLOCK_DEVICE_LOOKUP_BACKING = 1 << 1, /* fd may be regular file or directory on file system, in
|
|
* which case backing block device is determined. */
|
|
BLOCK_DEVICE_LOOKUP_ORIGINATING = 1 << 2, /* Try to find the underlying layer device for stacked
|
|
* block device, e.g. LUKS-style DM. */
|
|
} BlockDeviceLookupFlags;
|
|
|
|
int block_device_new_from_fd(int fd, BlockDeviceLookupFlags flags, sd_device **ret);
|
|
int block_device_new_from_path(const char *path, BlockDeviceLookupFlags flags, sd_device **ret);
|
|
|
|
int block_device_is_whole_disk(sd_device *dev);
|
|
int block_device_get_whole_disk(sd_device *dev, sd_device **ret);
|
|
int block_device_get_originating(sd_device *dev, sd_device **ret);
|
|
|
|
int block_get_whole_disk(dev_t d, dev_t *ret);
|
|
int block_get_originating(dev_t d, dev_t *ret);
|
|
|
|
int get_block_device_fd(int fd, dev_t *ret);
|
|
int get_block_device(const char *path, dev_t *dev);
|
|
|
|
int get_block_device_harder_fd(int fd, dev_t *dev);
|
|
int get_block_device_harder(const char *path, dev_t *dev);
|
|
|
|
int lock_whole_block_device(dev_t devt, int open_flags, int operation);
|
|
|
|
int blockdev_partscan_enabled(sd_device *d);
|
|
int blockdev_partscan_enabled_fd(int fd);
|
|
|
|
int fd_is_encrypted(int fd);
|
|
int path_is_encrypted(const char *path);
|
|
|
|
int fd_get_whole_disk(int fd, bool backing, dev_t *ret);
|
|
int path_get_whole_disk(const char *path, bool backing, dev_t *ret);
|
|
|
|
int block_device_add_partition(int fd, const char *name, int nr, uint64_t start, uint64_t size);
|
|
int block_device_remove_partition(int fd, const char *name, int nr);
|
|
int block_device_resize_partition(int fd, int nr, uint64_t start, uint64_t size);
|
|
int partition_enumerator_new(sd_device *dev, sd_device_enumerator **ret);
|
|
int block_device_remove_all_partitions(sd_device *dev, int fd);
|
|
|
|
int blockdev_get_sector_size(int fd, uint32_t *ret);
|
|
int blockdev_get_device_size(int fd, uint64_t *ret);
|
|
|
|
int blockdev_get_root(int level, dev_t *ret);
|
|
|
|
int partition_node_of(const char *node, unsigned nr, char **ret);
|