fix: use cargo workspace

This way bch_bindgen actually gets used by various cargo subcommands,
like clippy.

Signed-off-by: Thomas Mühlbacher <tmuehlbacher@posteo.net>
Signed-off-by: Kent Overstreet <kent.overstreet@linux.dev>

.bcachefs_revision

@@ -1 +1 @@
-dbe591cee299957e282eb7857edea35050b1d8b5
+7fdc3fa3cb5fb561f5945b4de418d48d1a726a8d

.github/workflows/nix-flake.yml

@@ -1,22 +1,31 @@
-name: "Nix-Tests"
+name: Nix Flake actions
+
 on:
   pull_request:
   push:
+
 jobs:
-  nix-flake-check:
+  nix-matrix:
     runs-on: ubuntu-latest
+    outputs:
+      matrix: ${{ steps.set-matrix.outputs.matrix }}
     steps:
-    - uses: actions/checkout@v4
-    - uses: cachix/install-nix-action@v27
-      with:
-        extra_nix_config: |
-          experimental-features = nix-command flakes
-          access-tokens = github.com=${{ secrets.GITHUB_TOKEN }}
-    - uses: cachix/cachix-action@v15
-      with:
-        name: bcachefs-tools
-        # If you chose API tokens for write access OR if you have a private cache
-        authToken: '${{ secrets.CACHIX_AUTH_TOKEN }}'
-    - run: nix flake show
-    - run: nix flake check --print-build-logs
-    - run: nix build --print-build-logs
+      - uses: actions/checkout@v4
+      - uses: cachix/install-nix-action@v30
+      - id: set-matrix
+        name: Generate Nix Matrix
+        run: |
+          set -Eeu
+          matrix="$(nix eval --json '.#githubActions.matrix')"
+          echo "matrix=$matrix" >> "$GITHUB_OUTPUT"
+
+  nix-build:
+    name: ${{ matrix.name }} (${{ matrix.system }})
+    needs: nix-matrix
+    runs-on: ${{ matrix.os }}
+    strategy:
+      matrix: ${{fromJSON(needs.nix-matrix.outputs.matrix)}}
+    steps:
+      - uses: actions/checkout@v4
+      - uses: cachix/install-nix-action@v30
+      - run: nix build -L '.#${{ matrix.attr }}'

Cargo.toml

@@ -1,9 +1,13 @@
+[workspace]
+resolver = "2"
+default-members = [".", "bch_bindgen"]
+
 [package]
 name = "bcachefs-tools"
 version = "1.25.0"
 authors = ["Yuxuan Shui <yshuiv7@gmail.com>", "Kayla Firestack <dev@kaylafire.me>", "Kent Overstreet <kent.overstreet@linux.dev>" ]
 edition = "2021"
-rust-version = "1.77"
+rust-version = "1.77.0"
 
 [[bin]]
 name = "bcachefs"

Makefile

@@ -22,6 +22,13 @@ else
   CARGO_CLEAN_ARGS = --quiet
 endif
 
+# when cross compiling, cargo places the built binary in a different location
+ifdef CARGO_BUILD_TARGET
+	BUILT_BIN = target/$(CARGO_BUILD_TARGET)/release/bcachefs
+else
+	BUILT_BIN = target/release/bcachefs
+endif
+
 # Prevent recursive expansions of $(CFLAGS) to avoid repeatedly performing
 # compile tests
 CFLAGS:=$(CFLAGS)
@@ -195,7 +202,7 @@ cmd_version.o : .version
 install: INITRAMFS_HOOK=$(INITRAMFS_DIR)/hooks/bcachefs
 install: INITRAMFS_SCRIPT=$(INITRAMFS_DIR)/scripts/local-premount/bcachefs
 install: bcachefs $(optional_install)
-	$(INSTALL) -m0755 -D target/release/bcachefs -t $(DESTDIR)$(ROOT_SBINDIR)
+	$(INSTALL) -m0755 -D $(BUILT_BIN)  -t $(DESTDIR)$(ROOT_SBINDIR)
 	$(INSTALL) -m0644 -D bcachefs.8    -t $(DESTDIR)$(PREFIX)/share/man/man8/
 	$(INSTALL) -m0755 -D initramfs/script $(DESTDIR)$(INITRAMFS_SCRIPT)
 	$(INSTALL) -m0755 -D initramfs/hook   $(DESTDIR)$(INITRAMFS_HOOK)

bch_bindgen/src/opts.rs

@@ -1,3 +1,7 @@
+use crate::c;
+use crate::fs::Fs;
+use std::ffi::{CString, c_char};
+
 #[macro_export]
 macro_rules! opt_set {
     ($opts:ident, $n:ident, $v:expr) => {
@@ -33,3 +37,29 @@ macro_rules! opt_get {
         }
     };
 }
+
+pub fn parse_mount_opts(fs: Option<&mut Fs>, optstr: Option<&str>, ignore_unknown: bool)
+        -> Result<c::bch_opts, c::bch_errcode> {
+    let mut opts: c::bch_opts = Default::default();
+
+    if let Some(optstr) = optstr {
+        let optstr = CString::new(optstr).unwrap();
+        let optstr_ptr = optstr.as_ptr();
+
+        let ret = unsafe {
+            c::bch2_parse_mount_opts(fs.map_or(std::ptr::null_mut(), |f| f.raw),
+                                     &mut opts as *mut c::bch_opts,
+                                     std::ptr::null_mut(),
+                                     optstr_ptr as *mut c_char,
+                                     ignore_unknown)
+        };
+
+        drop(optstr);
+
+        if ret != 0 {
+            let err: c::bch_errcode = unsafe { std::mem::transmute(-ret) };
+            return Err(err);
+        }
+    }
+    Ok(opts)
+}

c_src/bcachefs.c

@@ -93,9 +93,11 @@ void bcachefs_usage(void)
 	     "  list                     List filesystem metadata in textual form\n"
 	     "  list_journal             List contents of journal\n"
 	     "\n"
+#ifdef BCACHEFS_FUSE
 	     "FUSE:\n"
 	     "  fusemount                Mount a filesystem via FUSE\n"
 	     "\n"
+#endif
 	     "Miscellaneous:\n"
          "  completions              Generate shell completions\n"
 	     "  version                  Display the version of the invoked bcachefs tool\n");
@@ -116,16 +118,15 @@ int fs_cmds(int argc, char *argv[])
 {
 	char *cmd = pop_cmd(&argc, argv);
 
-	if (argc < 1) {
-		bcachefs_usage();
-		exit(EXIT_FAILURE);
-	}
+	if (argc < 1)
+		return fs_usage();
 	if (!strcmp(cmd, "usage"))
 		return cmd_fs_usage(argc, argv);
 	if (!strcmp(cmd, "top"))
 		return cmd_fs_top(argc, argv);
 
-	return 0;
+	fs_usage();
+	return -EINVAL;
 }
 
 int device_cmds(int argc, char *argv[])
@@ -151,7 +152,8 @@ int device_cmds(int argc, char *argv[])
 	if (!strcmp(cmd, "resize-journal"))
 		return cmd_device_resize_journal(argc, argv);
 
-	return 0;
+	device_usage();
+	return -EINVAL;
 }
 
 int data_cmds(int argc, char *argv[])
@@ -167,5 +169,6 @@ int data_cmds(int argc, char *argv[])
 	if (!strcmp(cmd, "job"))
 		return cmd_data_job(argc, argv);
 
-	return 0;
+	data_usage();
+	return -EINVAL;
 }

c_src/cmd_device.c

@@ -137,10 +137,11 @@ int cmd_device_add(int argc, char *argv[])
 	opt_set(fs_opts, btree_node_size,
 		read_file_u64(fs.sysfs_fd, "options/btree_node_size"));
 
-	struct bch_sb *sb = bch2_format(fs_opt_strs,
-					fs_opts,
-					format_opts,
-					&dev_opts, 1);
+	dev_opts_list devs = {};
+	darray_push(&devs, dev_opts);
+
+	struct bch_sb *sb = bch2_format(fs_opt_strs, fs_opts, format_opts, devs);
+	darray_exit(&devs);
 	free(sb);
 	bchu_disk_add(fs, dev_opts.path);
 	return 0;

c_src/cmd_format.c

@@ -121,13 +121,13 @@ void build_fs(struct bch_fs *c, const char *src_path)
 	if (!S_ISDIR(stat.st_mode))
 		die("%s is not a directory", src_path);
 
-	copy_fs(c, src_fd, src_path, &s);
+	copy_fs(c, src_fd, src_path, &s, 0);
 }
 
 int cmd_format(int argc, char *argv[])
 {
-	DARRAY(struct dev_opts) devices = { 0 };
-	DARRAY(char *) device_paths = { 0 };
+	dev_opts_list devices = {};
+	darray_str device_paths = {};
 	struct format_opts opts	= format_opts_default();
 	struct dev_opts dev_opts = dev_opts_default();
 	bool force = false, no_passphrase = false, quiet = false, initialize = true, verbose = false;
@@ -207,9 +207,8 @@ int cmd_format(int argc, char *argv[])
 			force = true;
 			break;
 		case O_fs_size:
-			if (bch2_strtoull_h(optarg, &dev_opts.opts.fs_size))
+			if (bch2_strtoull_h(optarg, &dev_opts.fs_size))
 				die("invalid filesystem size");
-			dev_opts.opts.fs_size_defined = true;
 			unconsumed_dev_option = true;
 			break;
 		case O_superblock_size:
@@ -233,8 +232,7 @@ int cmd_format(int argc, char *argv[])
 			darray_push(&device_paths, optarg);
 			dev_opts.path = optarg;
 			darray_push(&devices, dev_opts);
-			dev_opts.opts.fs_size = 0;
-			dev_opts.opts.fs_size_defined = 0;
+			dev_opts.fs_size = 0;
 			unconsumed_dev_option = false;
 			break;
 		case O_quiet:
@@ -277,11 +275,7 @@ int cmd_format(int argc, char *argv[])
 			die("Error opening %s: %s", dev_opts.path, strerror(-ret));
 	}
 
-	struct bch_sb *sb =
-		bch2_format(fs_opt_strs,
-			    fs_opts,
-			    opts,
-			    devices.data, devices.nr);
+	struct bch_sb *sb = bch2_format(fs_opt_strs, fs_opts, opts, devices);
 	bch2_opt_strs_free(&fs_opt_strs);
 
 	if (!quiet) {

c_src/cmd_fs.c

@@ -239,9 +239,10 @@ static void accounting_sort(darray_accounting_p *sorted,
 
 static void accounting_swab_if_old(struct bch_ioctl_query_accounting *in)
 {
-	u64 kernel_version = read_file_u64(AT_FDCWD, "/sys/module/bcachefs/parameters/version");
+	unsigned kernel_version = bcachefs_kernel_version();
 
-	if (kernel_version < bcachefs_metadata_version_disk_accounting_big_endian)
+	if (kernel_version &&
+	    kernel_version < bcachefs_metadata_version_disk_accounting_big_endian)
 		for (struct bkey_i_accounting *a = in->accounting;
 		     a < (struct bkey_i_accounting *) ((u64 *) in->accounting + in->accounting_u64s);
 		     a = bkey_i_to_accounting(bkey_next(&a->k_i)))
@@ -490,6 +491,19 @@ devs:
 	bcache_fs_close(fs);
 }
 
+int fs_usage(void)
+{
+	puts("bcachefs fs - manage a running filesystem\n"
+	     "Usage: bcachefs fs <CMD> [OPTIONS]\n"
+	     "\n"
+	     "Commands:\n"
+	     "  usage                   Display detailed filesystem usage\n"
+	     "  top                     Show runtime performance information\n"
+	     "\n"
+	     "Report bugs to <linux-bcachefs@vger.kernel.org>");
+	return 0;
+}
+
 static void fs_usage_usage(void)
 {
 	puts("bcachefs fs usage - display detailed filesystem usage\n"

c_src/cmd_fsck.c

@@ -21,7 +21,6 @@ static void fsck_usage(void)
 	     "  -y                      Assume \"yes\" to all questions\n"
 	     "  -f                      Force checking even if filesystem is marked clean\n"
 	     "  -r, --ratelimit_errors  Don't display more than 10 errors of a given type\n"
-	     "  -R, --reconstruct_alloc Reconstruct the alloc btree\n"
 	     "  -k, --kernel            Use the in-kernel fsck implementation\n"
 	     "  -v                      Be verbose\n"
 	     "  -h, --help              Display this help and exit\n"
@@ -117,9 +116,7 @@ static bool should_use_kernel_fsck(darray_str devs)
 {
 	system("modprobe bcachefs");
 
-	unsigned kernel_version = !access("/sys/module/bcachefs/parameters/version", R_OK)
-	    ? read_file_u64(AT_FDCWD, "/sys/module/bcachefs/parameters/version")
-	    : 0;
+	unsigned kernel_version = bcachefs_kernel_version();
 
 	if (!kernel_version)
 		return false;
@@ -205,7 +202,6 @@ int cmd_fsck(int argc, char *argv[])
 {
 	static const struct option longopts[] = {
 		{ "ratelimit_errors",	no_argument,		NULL, 'r' },
-		{ "reconstruct_alloc",	no_argument,		NULL, 'R' },
 		{ "kernel",		no_argument,		NULL, 'k' },
 		{ "no-kernel",		no_argument,		NULL, 'K' },
 		{ "help",		no_argument,		NULL, 'h' },
@@ -224,7 +220,7 @@ int cmd_fsck(int argc, char *argv[])
 	append_opt(&opts_str, "read_only");
 
 	while ((opt = getopt_long(argc, argv,
-				  "apynfo:rRkKvh",
+				  "apynfo:rkKvh",
 				  longopts, NULL)) != -1)
 		switch (opt) {
 		case 'a':
@@ -248,9 +244,6 @@ int cmd_fsck(int argc, char *argv[])
 		case 'r':
 			append_opt(&opts_str, "ratelimit_errors");
 			break;
-		case 'R':
-			append_opt(&opts_str, "reconstruct_alloc");
-			break;
 		case 'k':
 			kernel = true;
 			break;
@@ -326,7 +319,7 @@ kernel_fsck_err:
 	} else {
 userland_fsck:
 		printf("Running userspace offline fsck\n");
-		ret = bch2_parse_mount_opts(NULL, &opts, &parse_later, opts_str.buf);
+		ret = bch2_parse_mount_opts(NULL, &opts, &parse_later, opts_str.buf, false);
 		if (ret)
 			return ret;
 

c_src/cmd_fusemount.c

@@ -19,7 +19,7 @@
 #include "libbcachefs/dirent.h"
 #include "libbcachefs/errcode.h"
 #include "libbcachefs/error.h"
-#include "libbcachefs/fs-common.h"
+#include "libbcachefs/namei.h"
 #include "libbcachefs/inode.h"
 #include "libbcachefs/io_read.h"
 #include "libbcachefs/io_write.h"

c_src/cmd_list_journal.c

@@ -65,9 +65,21 @@ static bool bkey_matches_filter(d_bbpos_range filter, struct jset_entry *entry,
 		struct bbpos k_start	= BBPOS(entry->btree_id, bkey_start_pos(&k->k));
 		struct bbpos k_end	= BBPOS(entry->btree_id, k->k.p);
 
-		if (bbpos_cmp(k_start, i->start) >= 0 &&
-		    bbpos_cmp(k_end, i->end) <= 0)
-			return true;
+		if (!i->start.pos.snapshot &&
+		    !i->end.pos.snapshot) {
+			k_start.pos.snapshot = 0;
+			k_end.pos.snapshot = 0;
+		}
+
+		if (!k->k.size) {
+			if (bbpos_cmp(k_start, i->start) >= 0 &&
+			    bbpos_cmp(k_end, i->end) <= 0)
+				return true;
+		} else {
+			if (bbpos_cmp(i->start, k_end) <= 0 &&
+			    bbpos_cmp(i->end, k_start) >= 0)
+				return true;
+		}
 	}
 	return false;
 }
@@ -75,9 +87,9 @@ static bool bkey_matches_filter(d_bbpos_range filter, struct jset_entry *entry,
 static bool entry_matches_transaction_filter(struct jset_entry *entry,
 					     d_bbpos_range filter)
 {
-	if (entry->type == BCH_JSET_ENTRY_btree_root ||
-	    entry->type == BCH_JSET_ENTRY_btree_keys ||
-	    entry->type == BCH_JSET_ENTRY_overwrite)
+	if (!entry->level &&
+	    (entry->type == BCH_JSET_ENTRY_btree_keys ||
+	     entry->type == BCH_JSET_ENTRY_overwrite))
 		jset_entry_for_each_key(entry, k)
 			if (bkey_matches_filter(filter, entry, k))
 				return true;
@@ -90,6 +102,8 @@ static bool should_print_transaction(struct jset_entry *entry, struct jset_entry
 {
 	struct jset_entry_log *l = container_of(entry, struct jset_entry_log, entry);
 	unsigned b = jset_entry_log_msg_bytes(l);
+	bool have_log_messages = false;
+	bool have_non_log_messages = false;
 
 	darray_for_each(msg_filter, i)
 		if (!strncmp(*i, l->d, b))
@@ -100,11 +114,19 @@ static bool should_print_transaction(struct jset_entry *entry, struct jset_entry
 
 	for (entry = vstruct_next(entry);
 	     entry != end && !entry_is_transaction_start(entry);
-	     entry = vstruct_next(entry))
-		if (entry_is_log_msg(entry) ||
-		    entry_matches_transaction_filter(entry, key_filter))
+	     entry = vstruct_next(entry)) {
+		if (entry_matches_transaction_filter(entry, key_filter))
 			return true;
 
+		if (entry_is_log_msg(entry))
+			have_log_messages = true;
+		else
+			have_non_log_messages = true;
+	}
+
+	if (have_log_messages && !have_non_log_messages)
+		return true;
+
 	return false;
 }
 
@@ -134,6 +156,7 @@ static void journal_entry_header_to_text(struct printbuf *out,
 		prt_str(out, "blacklisted ");
 
 	prt_printf(out,
+		   "\n"
 		   "journal entry     %llu\n"
 		   "  version         %u\n"
 		   "  last seq        %llu\n"

c_src/cmd_migrate.c

@@ -209,38 +209,43 @@ static int migrate_fs(const char		*fs_path,
 	if (!S_ISDIR(stat.st_mode))
 		die("%s is not a directory", fs_path);
 
-	struct dev_opts dev = dev_opts_default();
+	dev_opts_list devs = {};
+	darray_push(&devs, dev_opts_default());
 
-	dev.path = dev_t_to_path(stat.st_dev);
-	dev.file = bdev_file_open_by_path(dev.path, BLK_OPEN_READ|BLK_OPEN_WRITE, &dev, NULL);
+	struct dev_opts *dev = &devs.data[0];
 
-	int ret = PTR_ERR_OR_ZERO(dev.file);
+	dev->path = dev_t_to_path(stat.st_dev);
+	dev->file = bdev_file_open_by_path(dev->path, BLK_OPEN_READ|BLK_OPEN_WRITE, dev, NULL);
+
+	int ret = PTR_ERR_OR_ZERO(dev->file);
 	if (ret < 0)
-		die("Error opening device to format %s: %s", dev.path, strerror(-ret));
-	dev.bdev = file_bdev(dev.file);
+		die("Error opening device to format %s: %s", dev->path, strerror(-ret));
+	dev->bdev = file_bdev(dev->file);
 
-	opt_set(fs_opts, block_size, get_blocksize(dev.bdev->bd_fd));
+	opt_set(fs_opts, block_size, get_blocksize(dev->bdev->bd_fd));
 
 	char *file_path = mprintf("%s/bcachefs", fs_path);
 	printf("Creating new filesystem on %s in space reserved at %s\n",
-	       dev.path, file_path);
+	       dev->path, file_path);
 
-	dev.opts.fs_size	= get_size(dev.bdev->bd_fd);
-	dev.opts.bucket_size	= bch2_pick_bucket_size(fs_opts, &dev);
-	dev.nbuckets		= dev.opts.fs_size / dev.opts.bucket_size;
+	dev->fs_size		= get_size(dev->bdev->bd_fd);
+	opt_set(dev->opts, bucket_size, bch2_pick_bucket_size(fs_opts, devs));
 
-	bch2_check_bucket_size(fs_opts, &dev);
+	dev->nbuckets		= dev->fs_size / dev->opts.bucket_size;
+
+	bch2_check_bucket_size(fs_opts, dev);
 
 	u64 bcachefs_inum;
 	ranges extents = reserve_new_fs_space(file_path,
 				fs_opts.block_size >> 9,
-				get_size(dev.bdev->bd_fd) / 5,
+				get_size(dev->bdev->bd_fd) / 5,
 				&bcachefs_inum, stat.st_dev, force);
 
-	find_superblock_space(extents, format_opts, &dev);
+	find_superblock_space(extents, format_opts, dev);
+
+	struct bch_sb *sb = bch2_format(fs_opt_strs, fs_opts, format_opts, devs);
+	darray_exit(&devs);
 
-	struct bch_sb *sb = bch2_format(fs_opt_strs,
-					fs_opts, format_opts, &dev, 1);
 	u64 sb_offset = le64_to_cpu(sb->layout.sb_offset[0]);
 
 	if (format_opts.passphrase)
@@ -248,16 +253,14 @@ static int migrate_fs(const char		*fs_path,
 
 	free(sb);
 
-	struct bch_opts opts = bch2_opts_empty();
-	struct bch_fs *c = NULL;
-	char *path[1] = { dev.path };
+	char *path[1] = { dev->path };
 
+	struct bch_opts opts = bch2_opts_empty();
 	opt_set(opts, sb,	sb_offset);
 	opt_set(opts, nostart,	true);
 	opt_set(opts, noexcl,	true);
-	opt_set(opts, nostart, true);
 
-	c = bch2_fs_open(path, 1, opts);
+	struct bch_fs *c = bch2_fs_open(path, 1, opts);
 	if (IS_ERR(c))
 		die("Error opening new filesystem: %s", bch2_err_str(PTR_ERR(c)));
 
@@ -265,10 +268,6 @@ static int migrate_fs(const char		*fs_path,
 	if (ret)
 		die("Error allocating buckets_nouse: %s", bch2_err_str(ret));
 
-	ret = bch2_fs_start(c);
-	if (IS_ERR(c))
-		die("Error starting new filesystem: %s", bch2_err_str(ret));
-
 	mark_unreserved_space(c, extents);
 
 	ret = bch2_fs_start(c);
@@ -282,7 +281,10 @@ static int migrate_fs(const char		*fs_path,
 		.type		= BCH_MIGRATE_migrate,
 	};
 
-	copy_fs(c, fs_fd, fs_path, &s);
+	u64 reserve_start = round_up((format_opts.superblock_size * 2 + 8) << 9,
+				     dev->opts.bucket_size);
+
+	copy_fs(c, fs_fd, fs_path, &s, reserve_start);
 
 	bch2_fs_stop(c);
 
@@ -310,7 +312,7 @@ static int migrate_fs(const char		*fs_path,
 	       "filesystem. That file can be deleted once the old filesystem is\n"
 	       "no longer needed (and should be deleted prior to running\n"
 	       "bcachefs migrate-superblock)\n",
-	       sb_offset, dev.path, dev.path, sb_offset);
+	       sb_offset, dev->path, dev->path, sb_offset);
 	return 0;
 }
 
@@ -374,7 +376,7 @@ static void migrate_superblock_usage(void)
 int cmd_migrate_superblock(int argc, char *argv[])
 {
 	char *dev = NULL;
-	u64 offset = 0;
+	u64 sb_offset = 0;
 	int opt, ret;
 
 	while ((opt = getopt(argc, argv, "d:o:h")) != -1)
@@ -383,7 +385,7 @@ int cmd_migrate_superblock(int argc, char *argv[])
 				dev = optarg;
 				break;
 			case 'o':
-				ret = kstrtou64(optarg, 10, &offset);
+				ret = kstrtou64(optarg, 10, &sb_offset);
 				if (ret)
 					die("Invalid offset");
 				break;
@@ -395,29 +397,72 @@ int cmd_migrate_superblock(int argc, char *argv[])
 	if (!dev)
 		die("Please specify a device");
 
-	if (!offset)
+	if (!sb_offset)
 		die("Please specify offset of existing superblock");
 
 	int fd = xopen(dev, O_RDWR);
-	struct bch_sb *sb = __bch2_super_read(fd, offset);
+	struct bch_sb *sb = __bch2_super_read(fd, sb_offset);
+	unsigned sb_size = 1U << sb->layout.sb_max_size_bits;
 
 	if (sb->layout.nr_superblocks >= ARRAY_SIZE(sb->layout.sb_offset))
 		die("Can't add superblock: no space left in superblock layout");
 
-	unsigned i;
-	for (i = 0; i < sb->layout.nr_superblocks; i++)
-		if (le64_to_cpu(sb->layout.sb_offset[i]) == BCH_SB_SECTOR)
-			die("Superblock layout already has default superblock");
+	for (unsigned i = 0; i < sb->layout.nr_superblocks; i++)
+		if (le64_to_cpu(sb->layout.sb_offset[i]) == BCH_SB_SECTOR ||
+		    le64_to_cpu(sb->layout.sb_offset[i]) == BCH_SB_SECTOR + sb_size)
+			die("Superblock layout already has default superblocks");
 
-	memmove(&sb->layout.sb_offset[1],
+	memmove(&sb->layout.sb_offset[2],
 		&sb->layout.sb_offset[0],
 		sb->layout.nr_superblocks * sizeof(u64));
-	sb->layout.nr_superblocks++;
-
+	sb->layout.nr_superblocks += 2;
 	sb->layout.sb_offset[0] = cpu_to_le64(BCH_SB_SECTOR);
+	sb->layout.sb_offset[1] = cpu_to_le64(BCH_SB_SECTOR + sb_size);
+
+	/* also write first 0-3.5k bytes with zeroes, ensure we blow away old
+	 * superblock */
+	static const char zeroes[BCH_SB_SECTOR << 9];
+	xpwrite(fd, zeroes, BCH_SB_SECTOR << 9, 0, "zeroing start of disk");
 
 	bch2_super_write(fd, sb);
 	close(fd);
 
+	/* mark new superblocks */
+
+	struct bch_opts opts = bch2_opts_empty();
+	opt_set(opts, nostart,	true);
+	opt_set(opts, sb,	sb_offset);
+
+	struct bch_fs *c = bch2_fs_open(&dev, 1, opts);
+	ret =   PTR_ERR_OR_ZERO(c) ?:
+		bch2_buckets_nouse_alloc(c);
+	if (ret)
+		die("error opening filesystem: %s", bch2_err_str(ret));
+
+	struct bch_dev *ca = c->devs[0];
+	for (u64 b = 0; bucket_to_sector(ca, b) < BCH_SB_SECTOR + sb_size * 2; b++)
+		set_bit(b, ca->buckets_nouse);
+
+	ret = bch2_fs_start(c);
+	if (ret)
+		die("Error starting filesystem: %s", bch2_err_str(ret));
+
+	bch2_fs_stop(c);
+
+	opts = bch2_opts_empty();
+	opt_set(opts, fsck, true);
+	opt_set(opts, fix_errors, true);
+
+	/*
+	 * Hack: the free space counters are coming out wrong after marking the
+	 * new superblock, but it's just the device counters so it's
+	 * inconsequential:
+	 */
+
+	c = bch2_fs_open(&dev, 1, opts);
+	ret =   PTR_ERR_OR_ZERO(c);
+	if (ret)
+		die("error opening filesystem: %s", bch2_err_str(ret));
+	bch2_fs_stop(c);
 	return 0;
 }

c_src/cmds.h

@@ -15,6 +15,7 @@ int cmd_recover_super(int argc, char *argv[]);
 int cmd_reset_counters(int argc, char *argv[]);
 int cmd_set_option(int argc, char *argv[]);
 
+int fs_usage(void);
 int cmd_fs_usage(int argc, char *argv[]);
 int cmd_fs_top(int argc, char *argv[]);
 

c_src/libbcachefs.c

@@ -16,6 +16,8 @@
 
 #include <uuid/uuid.h>
 
+#include <linux/mm.h>
+
 #include "libbcachefs.h"
 #include "crypto.h"
 #include "libbcachefs/bcachefs_format.h"
@@ -60,41 +62,54 @@ void bch2_sb_layout_init(struct bch_sb_layout *l,
 		    sb_start, sb_pos, sb_end, sb_size);
 }
 
-/* minimum size filesystem we can create, given a bucket size: */
-static u64 min_size(unsigned bucket_size)
+static u64 dev_max_bucket_size(u64 dev_size)
 {
-	return BCH_MIN_NR_NBUCKETS * bucket_size;
+	return rounddown_pow_of_two(dev_size / (BCH_MIN_NR_NBUCKETS * 4));
 }
 
-u64 bch2_pick_bucket_size(struct bch_opts opts, struct dev_opts *dev)
+u64 bch2_pick_bucket_size(struct bch_opts opts, dev_opts_list devs)
 {
-	if (dev->opts.fs_size < min_size(opts.block_size))
-		die("cannot format %s, too small (%llu bytes, min %llu)",
-		    dev->path, dev->opts.fs_size, min_size(opts.block_size));
-
 	/* Bucket size must be >= block size: */
 	u64 bucket_size = opts.block_size;
 
 	/* Bucket size must be >= btree node size: */
 	if (opt_defined(opts, btree_node_size))
-		bucket_size = max_t(unsigned, bucket_size, opts.btree_node_size);
+		bucket_size = max_t(u64, bucket_size, opts.btree_node_size);
 
-	/* Want a bucket size of at least 128k, if possible: */
-	bucket_size = max(bucket_size, 128ULL << 10);
+	u64 min_dev_size = BCH_MIN_NR_NBUCKETS * bucket_size;
+	darray_for_each(devs, i)
+		if (i->fs_size < min_dev_size)
+			die("cannot format %s, too small (%llu bytes, min %llu)",
+			    i->path, i->fs_size, min_dev_size);
 
-	if (dev->opts.fs_size >= min_size(bucket_size)) {
-		unsigned scale = max(1,
-			ilog2(dev->opts.fs_size / min_size(bucket_size)) / 4);
+	u64 total_fs_size = 0;
+	darray_for_each(devs, i)
+		total_fs_size += i->fs_size;
 
-		scale = rounddown_pow_of_two(scale);
+	struct sysinfo info;
+	si_meminfo(&info);
 
-		/* max bucket size 1 mb */
-		bucket_size = min(bucket_size * scale, 1ULL << 20);
-	} else {
-		do {
-			bucket_size /= 2;
-		} while (dev->opts.fs_size < min_size(bucket_size));
-	}
+	/*
+	 * Large fudge factor to allow for other fsck processes and devices
+	 * being added after creation
+	 */
+	u64 mem_available_for_fsck = info.totalram / 8;
+	u64 buckets_can_fsck = mem_available_for_fsck / (sizeof(struct bucket) * 1.5);
+	u64 mem_lower_bound = roundup_pow_of_two(total_fs_size / buckets_can_fsck);
+
+	/*
+	 * Lower bound to avoid fragmenting encoded (checksummed, compressed)
+	 * extents too much as they're moved:
+	 */
+	bucket_size = max(bucket_size, opt_get(opts, encoded_extent_max) * 4);
+
+	/* Lower bound to ensure we can fsck: */
+	bucket_size = max(bucket_size, mem_lower_bound);
+
+	u64 perf_lower_bound = min(2ULL << 20, total_fs_size / (1ULL << 20));
+
+	/* We also prefer larger buckets for performance, up to 2MB at 2T */
+	bucket_size = max(bucket_size, perf_lower_bound);
 
 	return bucket_size;
 }
@@ -116,20 +131,17 @@ void bch2_check_bucket_size(struct bch_opts opts, struct dev_opts *dev)
 }
 
 static unsigned parse_target(struct bch_sb_handle *sb,
-			     struct dev_opts *devs, size_t nr_devs,
+			     dev_opts_list devs,
 			     const char *s)
 {
-	struct dev_opts *i;
-	int idx;
-
 	if (!s)
 		return 0;
 
-	for (i = devs; i < devs + nr_devs; i++)
+	darray_for_each(devs, i)
 		if (!strcmp(s, i->path))
-			return dev_to_target(i - devs);
+			return dev_to_target(i - devs.data);
 
-	idx = bch2_disk_path_find(sb, s);
+	int idx = bch2_disk_path_find(sb, s);
 	if (idx >= 0)
 		return group_to_target(idx);
 
@@ -151,15 +163,12 @@ static void bch2_opt_set_sb_all(struct bch_sb *sb, int dev_idx, struct bch_opts
 struct bch_sb *bch2_format(struct bch_opt_strs	fs_opt_strs,
 			   struct bch_opts	fs_opts,
 			   struct format_opts	opts,
-			   struct dev_opts	*devs,
-			   size_t		nr_devs)
+			   dev_opts_list devs)
 {
 	struct bch_sb_handle sb = { NULL };
-	struct dev_opts *i;
 	unsigned max_dev_block_size = 0;
-	u64 min_bucket_size = U64_MAX;
 
-	for (i = devs; i < devs + nr_devs; i++)
+	darray_for_each(devs, i)
 		max_dev_block_size = max(max_dev_block_size, get_blocksize(i->bdev->bd_fd));
 
 	/* calculate block size: */
@@ -171,21 +180,20 @@ struct bch_sb *bch2_format(struct bch_opt_strs	fs_opt_strs,
 		    fs_opts.block_size, max_dev_block_size);
 
 	/* get device size, if it wasn't specified: */
-	for (i = devs; i < devs + nr_devs; i++)
-		if (!opt_defined(i->opts, fs_size))
-			opt_set(i->opts, fs_size, get_size(i->bdev->bd_fd));
+	darray_for_each(devs, i)
+		if (!i->fs_size)
+			i->fs_size = get_size(i->bdev->bd_fd);
 
 	/* calculate bucket sizes: */
-	for (i = devs; i < devs + nr_devs; i++)
-		min_bucket_size = min(min_bucket_size,
-			i->opts.bucket_size ?: bch2_pick_bucket_size(fs_opts, i));
+	u64 fs_bucket_size = bch2_pick_bucket_size(fs_opts, devs);
 
-	for (i = devs; i < devs + nr_devs; i++)
+	darray_for_each(devs, i)
 		if (!opt_defined(i->opts, bucket_size))
-			opt_set(i->opts, bucket_size, min_bucket_size);
+			opt_set(i->opts, bucket_size,
+				min(fs_bucket_size, dev_max_bucket_size(i->fs_size)));
 
-	for (i = devs; i < devs + nr_devs; i++) {
-		i->nbuckets = i->opts.fs_size / i->opts.bucket_size;
+	darray_for_each(devs, i) {
+		i->nbuckets = i->fs_size / i->opts.bucket_size;
 		bch2_check_bucket_size(fs_opts, i);
 	}
 
@@ -193,7 +201,7 @@ struct bch_sb *bch2_format(struct bch_opt_strs	fs_opt_strs,
 	if (!opt_defined(fs_opts, btree_node_size)) {
 		unsigned s = bch2_opts_default.btree_node_size;
 
-		for (i = devs; i < devs + nr_devs; i++)
+		darray_for_each(devs, i)
 			s = min(s, i->opts.bucket_size);
 		opt_set(fs_opts, btree_node_size, s);
 	}
@@ -208,7 +216,7 @@ struct bch_sb *bch2_format(struct bch_opt_strs	fs_opt_strs,
 	sb.sb->version_min	= le16_to_cpu(opts.version);
 	sb.sb->magic		= BCHFS_MAGIC;
 	sb.sb->user_uuid	= opts.uuid;
-	sb.sb->nr_devices	= nr_devs;
+	sb.sb->nr_devices	= devs.nr;
 	SET_BCH_SB_VERSION_INCOMPAT_ALLOWED(sb.sb, opts.version);
 
 	if (opts.version == bcachefs_metadata_version_current)
@@ -233,11 +241,11 @@ struct bch_sb *bch2_format(struct bch_opt_strs	fs_opt_strs,
 	/* Member info: */
 	struct bch_sb_field_members_v2 *mi =
 		bch2_sb_field_resize(&sb, members_v2,
-			(sizeof(*mi) + sizeof(struct bch_member) *
-			nr_devs) / sizeof(u64));
+			(sizeof(*mi) + sizeof(struct bch_member) * devs.nr) / sizeof(u64));
+
 	mi->member_bytes = cpu_to_le16(sizeof(struct bch_member));
-	for (i = devs; i < devs + nr_devs; i++) {
-		unsigned idx = i - devs;
+	darray_for_each(devs, i) {
+		unsigned idx = i - devs.data;
 		struct bch_member *m = bch2_members_v2_get_mut(sb.sb, idx);
 
 		uuid_generate(m->uuid.b);
@@ -247,14 +255,11 @@ struct bch_sb *bch2_format(struct bch_opt_strs	fs_opt_strs,
 	}
 
 	/* Disk labels*/
-	for (i = devs; i < devs + nr_devs; i++) {
-		struct bch_member *m;
-		int idx;
-
+	darray_for_each(devs, i) {
 		if (!i->label)
 			continue;
 
-		idx = bch2_disk_path_find_or_create(&sb, i->label);
+		int idx = bch2_disk_path_find_or_create(&sb, i->label);
 		if (idx < 0)
 			die("error creating disk path: %s", strerror(-idx));
 
@@ -262,18 +267,18 @@ struct bch_sb *bch2_format(struct bch_opt_strs	fs_opt_strs,
 		 * Recompute mi and m after each sb modification: its location
 		 * in memory may have changed due to reallocation.
 		 */
-		m = bch2_members_v2_get_mut(sb.sb, (i - devs));
+		struct bch_member *m = bch2_members_v2_get_mut(sb.sb, (i - devs.data));
 		SET_BCH_MEMBER_GROUP(m,	idx + 1);
 	}
 
 	SET_BCH_SB_FOREGROUND_TARGET(sb.sb,
-		parse_target(&sb, devs, nr_devs, fs_opt_strs.foreground_target));
+		parse_target(&sb, devs, fs_opt_strs.foreground_target));
 	SET_BCH_SB_BACKGROUND_TARGET(sb.sb,
-		parse_target(&sb, devs, nr_devs, fs_opt_strs.background_target));
+		parse_target(&sb, devs, fs_opt_strs.background_target));
 	SET_BCH_SB_PROMOTE_TARGET(sb.sb,
-		parse_target(&sb, devs, nr_devs, fs_opt_strs.promote_target));
+		parse_target(&sb, devs, fs_opt_strs.promote_target));
 	SET_BCH_SB_METADATA_TARGET(sb.sb,
-		parse_target(&sb, devs, nr_devs, fs_opt_strs.metadata_target));
+		parse_target(&sb, devs, fs_opt_strs.metadata_target));
 
 	/* Crypt: */
 	if (opts.encrypted) {
@@ -286,10 +291,10 @@ struct bch_sb *bch2_format(struct bch_opt_strs	fs_opt_strs,
 
 	bch2_sb_members_cpy_v2_v1(&sb);
 
-	for (i = devs; i < devs + nr_devs; i++) {
-		u64 size_sectors = i->opts.fs_size >> 9;
+	darray_for_each(devs, i) {
+		u64 size_sectors = i->fs_size >> 9;
 
-		sb.sb->dev_idx = i - devs;
+		sb.sb->dev_idx = i - devs.data;
 
 		if (!i->sb_offset) {
 			i->sb_offset	= BCH_SB_SECTOR;

c_src/libbcachefs.h

@@ -45,14 +45,18 @@ struct format_opts {
 	char		*source;
 };
 
+static inline unsigned bcachefs_kernel_version(void)
+{
+	return !access("/sys/module/bcachefs/parameters/version", R_OK)
+	    ? read_file_u64(AT_FDCWD, "/sys/module/bcachefs/parameters/version")
+	    : 0;
+}
+
 static inline struct format_opts format_opts_default()
 {
-	unsigned version = !access(   "/sys/module/bcachefs/parameters/version", R_OK)
-	    ? read_file_u64(AT_FDCWD, "/sys/module/bcachefs/parameters/version")
-	    : bcachefs_metadata_version_current;
-
 	return (struct format_opts) {
-		.version		= version,
+		.version		= bcachefs_kernel_version() ?:
+			bcachefs_metadata_version_current,
 		.superblock_size	= SUPERBLOCK_SIZE_DEFAULT,
 	};
 }
@@ -66,12 +70,15 @@ struct dev_opts {
 	u64		sb_end;
 
 	u64		nbuckets;
+	u64		fs_size;
 
 	const char	*label; /* make this a bch_opt */
 
 	struct bch_opts	opts;
 };
 
+typedef DARRAY(struct dev_opts) dev_opts_list;
+
 static inline struct dev_opts dev_opts_default()
 {
 	return (struct dev_opts) { .opts = bch2_opts_empty() };
@@ -80,12 +87,13 @@ static inline struct dev_opts dev_opts_default()
 void bch2_sb_layout_init(struct bch_sb_layout *,
 			 unsigned, unsigned, u64, u64);
 
-u64 bch2_pick_bucket_size(struct bch_opts, struct dev_opts *);
+u64 bch2_pick_bucket_size(struct bch_opts, dev_opts_list);
 void bch2_check_bucket_size(struct bch_opts, struct dev_opts *);
 
 struct bch_sb *bch2_format(struct bch_opt_strs,
 			   struct bch_opts,
-			   struct format_opts, struct dev_opts *, size_t);
+			   struct format_opts,
+			   dev_opts_list devs);
 
 void bch2_super_write(int, struct bch_sb *);
 struct bch_sb *__bch2_super_read(int, u64);

c_src/posix_to_bcachefs.c

@@ -6,8 +6,8 @@
 #include "posix_to_bcachefs.h"
 #include "libbcachefs/alloc_foreground.h"
 #include "libbcachefs/buckets.h"
-#include "libbcachefs/fs-common.h"
 #include "libbcachefs/io_write.h"
+#include "libbcachefs/namei.h"
 #include "libbcachefs/str_hash.h"
 #include "libbcachefs/xattr.h"
 
@@ -264,7 +264,8 @@ void copy_link(struct bch_fs *c, struct bch_inode_unpacked *dst,
 
 static void copy_file(struct bch_fs *c, struct bch_inode_unpacked *dst,
 		      int src_fd, u64 src_size,
-		      char *src_path, struct copy_fs_state *s)
+		      char *src_path, struct copy_fs_state *s,
+		      u64 reserve_start)
 {
 	struct fiemap_iter iter;
 	struct fiemap_extent e;
@@ -295,11 +296,8 @@ static void copy_file(struct bch_fs *c, struct bch_inode_unpacked *dst,
 			continue;
 		}
 
-		/*
-		 * if the data is below 1 MB, copy it so it doesn't conflict
-		 * with bcachefs's potentially larger superblock:
-		 */
-		if (e.fe_physical < 1 << 20) {
+		/* If the data is in bcachefs's superblock region, copy it: */
+		if (e.fe_physical < reserve_start) {
 			copy_data(c, dst, src_fd, e.fe_logical,
 				  e.fe_logical + min(src_size - e.fe_logical,
 				      e.fe_length));
@@ -318,7 +316,8 @@ static void copy_file(struct bch_fs *c, struct bch_inode_unpacked *dst,
 static void copy_dir(struct copy_fs_state *s,
 		     struct bch_fs *c,
 		     struct bch_inode_unpacked *dst,
-		     int src_fd, const char *src_path)
+		     int src_fd, const char *src_path,
+		     u64 reserve_start)
 {
 	DIR *dir = fdopendir(src_fd);
 	struct dirent *d;
@@ -369,7 +368,7 @@ static void copy_dir(struct copy_fs_state *s,
 		switch (mode_to_type(stat.st_mode)) {
 		case DT_DIR:
 			fd = xopen(d->d_name, O_RDONLY|O_NOATIME);
-			copy_dir(s, c, &inode, fd, child_path);
+			copy_dir(s, c, &inode, fd, child_path, reserve_start);
 			close(fd);
 			break;
 		case DT_REG:
@@ -377,7 +376,7 @@ static void copy_dir(struct copy_fs_state *s,
 
 			fd = xopen(d->d_name, O_RDONLY|O_NOATIME);
 			copy_file(c, &inode, fd, stat.st_size,
-				  child_path, s);
+				  child_path, s, reserve_start);
 			close(fd);
 			break;
 		case DT_LNK:
@@ -409,7 +408,8 @@ next:
 
 static void reserve_old_fs_space(struct bch_fs *c,
 				 struct bch_inode_unpacked *root_inode,
-				 ranges *extents)
+				 ranges *extents,
+				 u64 reserve_start)
 {
 	struct bch_dev *ca = c->devs[0];
 	struct bch_inode_unpacked dst;
@@ -422,14 +422,20 @@ static void reserve_old_fs_space(struct bch_fs *c,
 
 	ranges_sort_merge(extents);
 
-	for_each_hole(iter, *extents, bucket_to_sector(ca, ca->mi.nbuckets) << 9, i)
-		link_data(c, &dst, i.start, i.start, i.end - i.start);
+	for_each_hole(iter, *extents, bucket_to_sector(ca, ca->mi.nbuckets) << 9, i) {
+		if (i.end <= reserve_start)
+			continue;
+
+		u64 start = max(i.start, reserve_start);
+
+		link_data(c, &dst, start, start, i.end - start);
+	}
 
 	update_inode(c, &dst);
 }
 
 void copy_fs(struct bch_fs *c, int src_fd, const char *src_path,
-		    struct copy_fs_state *s)
+	     struct copy_fs_state *s, u64 reserve_start)
 {
 	syncfs(src_fd);
 
@@ -448,10 +454,10 @@ void copy_fs(struct bch_fs *c, int src_fd, const char *src_path,
 
 
 	/* now, copy: */
-	copy_dir(s, c, &root_inode, src_fd, src_path);
+	copy_dir(s, c, &root_inode, src_fd, src_path, reserve_start);
 
 	if (BCH_MIGRATE_migrate == s->type)
-		reserve_old_fs_space(c, &root_inode, &s->extents);
+		reserve_old_fs_space(c, &root_inode, &s->extents, reserve_start);
 
 	update_inode(c, &root_inode);
 

c_src/posix_to_bcachefs.h

@@ -50,5 +50,5 @@ struct copy_fs_state {
  * initialized (`hardlinks` is initialized with zeroes).
  */
 void copy_fs(struct bch_fs *c, int src_fd, const char *src_path,
-		    struct copy_fs_state *s);
+		    struct copy_fs_state *s, u64);
 #endif /* _LIBBCACHE_H */

flake.lock

@@ -1,17 +1,12 @@
 {
   "nodes": {
     "crane": {
-      "inputs": {
-        "nixpkgs": [
-          "nixpkgs"
-        ]
-      },
       "locked": {
-        "lastModified": 1721842668,
-        "narHash": "sha256-k3oiD2z2AAwBFLa4+xfU+7G5fisRXfkvrMTCJrjZzXo=",
+        "lastModified": 1742394900,
+        "narHash": "sha256-vVOAp9ahvnU+fQoKd4SEXB2JG2wbENkpqcwlkIXgUC0=",
         "owner": "ipetkov",
         "repo": "crane",
-        "rev": "529c1a0b1f29f0d78fa3086b8f6a134c71ef3aaf",
+        "rev": "70947c1908108c0c551ddfd73d4f750ff2ea67cd",
         "type": "github"
       },
       "original": {
@@ -20,35 +15,14 @@
         "type": "github"
       }
     },
-    "fenix": {
-      "inputs": {
-        "nixpkgs": [
-          "nixpkgs"
-        ],
-        "rust-analyzer-src": "rust-analyzer-src"
-      },
-      "locked": {
-        "lastModified": 1722320953,
-        "narHash": "sha256-DfGaJtgrzcwPQYLTvjL1KaVIjpvi85b2MpM6yEGvJzM=",
-        "owner": "nix-community",
-        "repo": "fenix",
-        "rev": "483df76def3e5010d709aa3a0418ba2088503994",
-        "type": "github"
-      },
-      "original": {
-        "owner": "nix-community",
-        "repo": "fenix",
-        "type": "github"
-      }
-    },
     "flake-compat": {
       "flake": false,
       "locked": {
-        "lastModified": 1696426674,
-        "narHash": "sha256-kvjfFW7WAETZlt09AgDn1MrtKzP7t90Vf7vypd3OL1U=",
+        "lastModified": 1733328505,
+        "narHash": "sha256-NeCCThCEP3eCl2l/+27kNNK7QrwZB1IJCrXfrbv5oqU=",
         "owner": "edolstra",
         "repo": "flake-compat",
-        "rev": "0f9255e01c2351cc7d116c072cb317785dd33b33",
+        "rev": "ff81ac966bb2cae68946d5ed5fc4994f96d0ffec",
         "type": "github"
       },
       "original": {
@@ -62,11 +36,11 @@
         "nixpkgs-lib": "nixpkgs-lib"
       },
       "locked": {
-        "lastModified": 1719994518,
-        "narHash": "sha256-pQMhCCHyQGRzdfAkdJ4cIWiw+JNuWsTX7f0ZYSyz0VY=",
+        "lastModified": 1741352980,
+        "narHash": "sha256-+u2UunDA4Cl5Fci3m7S643HzKmIDAe+fiXrLqYsR2fs=",
         "owner": "hercules-ci",
         "repo": "flake-parts",
-        "rev": "9227223f6d922fee3c7b190b2cc238a99527bbb7",
+        "rev": "f4330d22f1c5d2ba72d3d22df5597d123fdb60a9",
         "type": "github"
       },
       "original": {
@@ -75,13 +49,33 @@
         "type": "github"
       }
     },
+    "nix-github-actions": {
+      "inputs": {
+        "nixpkgs": [
+          "nixpkgs"
+        ]
+      },
+      "locked": {
+        "lastModified": 1737420293,
+        "narHash": "sha256-F1G5ifvqTpJq7fdkT34e/Jy9VCyzd5XfJ9TO8fHhJWE=",
+        "owner": "nix-community",
+        "repo": "nix-github-actions",
+        "rev": "f4158fa080ef4503c8f4c820967d946c2af31ec9",
+        "type": "github"
+      },
+      "original": {
+        "owner": "nix-community",
+        "repo": "nix-github-actions",
+        "type": "github"
+      }
+    },
     "nixpkgs": {
       "locked": {
-        "lastModified": 1722185531,
-        "narHash": "sha256-veKR07psFoJjINLC8RK4DiLniGGMgF3QMlS4tb74S6k=",
+        "lastModified": 1742422364,
+        "narHash": "sha256-mNqIplmEohk5jRkqYqG19GA8MbQ/D4gQSK0Mu4LvfRQ=",
         "owner": "nixos",
         "repo": "nixpkgs",
-        "rev": "52ec9ac3b12395ad677e8b62106f0b98c1f8569d",
+        "rev": "a84ebe20c6bc2ecbcfb000a50776219f48d134cc",
         "type": "github"
       },
       "original": {
@@ -93,40 +87,47 @@
     },
     "nixpkgs-lib": {
       "locked": {
-        "lastModified": 1719876945,
-        "narHash": "sha256-Fm2rDDs86sHy0/1jxTOKB1118Q0O3Uc7EC0iXvXKpbI=",
-        "type": "tarball",
-        "url": "https://github.com/NixOS/nixpkgs/archive/5daf0514482af3f97abaefc78a6606365c9108e2.tar.gz"
+        "lastModified": 1740877520,
+        "narHash": "sha256-oiwv/ZK/2FhGxrCkQkB83i7GnWXPPLzoqFHpDD3uYpk=",
+        "owner": "nix-community",
+        "repo": "nixpkgs.lib",
+        "rev": "147dee35aab2193b174e4c0868bd80ead5ce755c",
+        "type": "github"
       },
       "original": {
-        "type": "tarball",
-        "url": "https://github.com/NixOS/nixpkgs/archive/5daf0514482af3f97abaefc78a6606365c9108e2.tar.gz"
+        "owner": "nix-community",
+        "repo": "nixpkgs.lib",
+        "type": "github"
       }
     },
     "root": {
       "inputs": {
         "crane": "crane",
-        "fenix": "fenix",
         "flake-compat": "flake-compat",
         "flake-parts": "flake-parts",
+        "nix-github-actions": "nix-github-actions",
         "nixpkgs": "nixpkgs",
+        "rust-overlay": "rust-overlay",
         "treefmt-nix": "treefmt-nix"
       }
     },
-    "rust-analyzer-src": {
-      "flake": false,
+    "rust-overlay": {
+      "inputs": {
+        "nixpkgs": [
+          "nixpkgs"
+        ]
+      },
       "locked": {
-        "lastModified": 1722262053,
-        "narHash": "sha256-KxjkPVn9rQqYam6DhiN/V2NcMXtYW25maxkJoiVMpmE=",
-        "owner": "rust-lang",
-        "repo": "rust-analyzer",
-        "rev": "a021b85be57d34b1eed687fcafd5d5ec64b2d853",
+        "lastModified": 1742524367,
+        "narHash": "sha256-KzTwk/5ETJavJZYV1DEWdCx05M4duFCxCpRbQSKWpng=",
+        "owner": "oxalica",
+        "repo": "rust-overlay",
+        "rev": "70bf752d176b2ce07417e346d85486acea9040ef",
         "type": "github"
       },
       "original": {
-        "owner": "rust-lang",
-        "ref": "nightly",
-        "repo": "rust-analyzer",
+        "owner": "oxalica",
+        "repo": "rust-overlay",
         "type": "github"
       }
     },
@@ -137,11 +138,11 @@
         ]
       },
       "locked": {
-        "lastModified": 1722330636,
-        "narHash": "sha256-uru7JzOa33YlSRwf9sfXpJG+UAV+bnBEYMjrzKrQZFw=",
+        "lastModified": 1742370146,
+        "narHash": "sha256-XRE8hL4vKIQyVMDXykFh4ceo3KSpuJF3ts8GKwh5bIU=",
         "owner": "numtide",
         "repo": "treefmt-nix",
-        "rev": "768acdb06968e53aa1ee8de207fd955335c754b7",
+        "rev": "adc195eef5da3606891cedf80c0d9ce2d3190808",
         "type": "github"
       },
       "original": {

flake.nix

@@ -11,13 +11,10 @@
       inputs.nixpkgs.follows = "nixpkgs";
     };
 
-    crane = {
-      url = "github:ipetkov/crane";
-      inputs.nixpkgs.follows = "nixpkgs";
-    };
+    crane.url = "github:ipetkov/crane";
 
-    fenix = {
-      url = "github:nix-community/fenix";
+    rust-overlay = {
+      url = "github:oxalica/rust-overlay";
       inputs.nixpkgs.follows = "nixpkgs";
     };
 
@@ -25,6 +22,11 @@
       url = "github:edolstra/flake-compat";
       flake = false;
     };
+
+    nix-github-actions = {
+      url = "github:nix-community/nix-github-actions";
+      inputs.nixpkgs.follows = "nixpkgs";
+    };
   };
 
   outputs =
@@ -33,27 +35,31 @@
       nixpkgs,
       flake-parts,
       treefmt-nix,
-      fenix,
       crane,
-      ...
+      rust-overlay,
+      flake-compat,
+      nix-github-actions,
     }:
+    let
+      systems = nixpkgs.lib.filter (s: nixpkgs.lib.hasSuffix "-linux" s) nixpkgs.lib.systems.flakeExposed;
+    in
     flake-parts.lib.mkFlake { inherit inputs; } {
       imports = [ inputs.treefmt-nix.flakeModule ];
 
-      # can be extended, but these have proper binary cache support in nixpkgs
-      # as of writing.
-      systems = [
-        "aarch64-linux"
-        "x86_64-linux"
-        "i686-linux"
-      ];
+      flake = {
+        githubActions = nix-github-actions.lib.mkGithubMatrix {
+          # github actions supports fewer architectures
+          checks = nixpkgs.lib.getAttrs [ "aarch64-linux" "x86_64-linux" ] self.checks;
+        };
+      };
+
+      inherit systems;
 
       perSystem =
         {
           self',
           config,
           lib,
-          pkgs,
           system,
           ...
         }:
@@ -62,119 +68,225 @@
           inherit (lib.lists) findFirst;
           inherit (lib.strings) hasPrefix removePrefix substring;
 
+          pkgs = import nixpkgs {
+            inherit system;
+            overlays = [ (import rust-overlay) ];
+          };
+
           cargoToml = builtins.fromTOML (builtins.readFile ./Cargo.toml);
           rustfmtToml = builtins.fromTOML (builtins.readFile ./rustfmt.toml);
 
-          craneLib = crane.mkLib pkgs;
-
           rev = self.shortRev or self.dirtyShortRev or (substring 0 8 self.lastModifiedDate);
           makefileVersion = removePrefix "VERSION=" (
             findFirst (line: hasPrefix "VERSION=" line) "VERSION=0.0.0" (split "\n" (readFile ./Makefile))
           );
           version = "${makefileVersion}+${rev}";
 
-          commonArgs = {
-            inherit version;
-            src = self;
+          mkCommon =
+            {
+              crane,
+              pkgs,
+              rustVersion ? "latest",
 
-            env = {
-              PKG_CONFIG_SYSTEMD_SYSTEMDSYSTEMUNITDIR = "${placeholder "out"}/lib/systemd/system";
-              PKG_CONFIG_UDEV_UDEVDIR = "${placeholder "out"}/lib/udev";
+              # build time
+              buildPackages,
+              pkg-config,
+              rustPlatform,
+              stdenv,
+
+              # run time
+              keyutils,
+              libaio,
+              libsodium,
+              liburcu,
+              libuuid,
+              lz4,
+              udev,
+              zlib,
+              zstd,
+            }:
+            let
+              inherit (stdenv) cc hostPlatform;
+
+              craneLib = (crane.mkLib pkgs).overrideToolchain (
+                p: p.rust-bin.stable."${rustVersion}".minimal.override { extensions = [ "clippy" ]; }
+              );
+
+              args = {
+                inherit version;
+                src = self;
+                strictDeps = true;
+
+                env = {
+                  PKG_CONFIG_SYSTEMD_SYSTEMDSYSTEMUNITDIR = "${placeholder "out"}/lib/systemd/system";
+                  PKG_CONFIG_UDEV_UDEVDIR = "${placeholder "out"}/lib/udev";
+
+                  CARGO_BUILD_TARGET = hostPlatform.rust.rustcTargetSpec;
+                  "CARGO_TARGET_${hostPlatform.rust.cargoEnvVarTarget}_LINKER" = "${cc.targetPrefix}cc";
+                  HOST_CC = "${cc.nativePrefix}cc";
+                  TARGET_CC = "${cc.targetPrefix}cc";
+                };
+
+                makeFlags = [
+                  "INITRAMFS_DIR=${placeholder "out"}/etc/initramfs-tools"
+                  "PREFIX=${placeholder "out"}"
+                  "VERSION=${version}"
+                ];
+
+                dontStrip = true;
+
+                depsBuildBuild = [
+                  buildPackages.stdenv.cc
+                ];
+
+                nativeBuildInputs = [
+                  pkg-config
+                  rustPlatform.bindgenHook
+                ];
+
+                buildInputs = [
+                  keyutils
+                  libaio
+                  libsodium
+                  liburcu
+                  libuuid
+                  lz4
+                  udev
+                  zlib
+                  zstd
+                ];
+
+                meta = {
+                  description = "Userspace tools for bcachefs";
+                  license = lib.licenses.gpl2Only;
+                  mainProgram = "bcachefs";
+                };
+              };
+
+              cargoArtifacts = craneLib.buildDepsOnly args;
+            in
+            {
+              inherit args cargoArtifacts craneLib;
             };
+          common = pkgs.callPackage mkCommon { inherit crane; };
 
-            makeFlags = [
-              "INITRAMFS_DIR=${placeholder "out"}/etc/initramfs-tools"
-              "PREFIX=${placeholder "out"}"
-              "VERSION=${version}"
-            ];
+          mkPackage =
+            { common, name }:
+            common.craneLib.buildPackage (
+              common.args
+              // {
+                inherit (common) cargoArtifacts;
+                pname = name;
 
-            dontStrip = true;
+                enableParallelBuilding = true;
+                buildPhaseCargoCommand = ''
+                  make ''${enableParallelBuilding:+-j''${NIX_BUILD_CORES}} $makeFlags
+                '';
+                doNotPostBuildInstallCargoBinaries = true;
+                installPhaseCommand = ''
+                  make ''${enableParallelBuilding:+-j''${NIX_BUILD_CORES}} $makeFlags install
+                '';
 
-            nativeBuildInputs = with pkgs; [
-              pkg-config
-              rustPlatform.bindgenHook
-            ];
+                doInstallCheck = true;
+                installCheckPhase = ''
+                  runHook preInstallCheck
 
-            buildInputs = with pkgs; [
-              attr
-              keyutils
-              libaio
-              libsodium
-              liburcu
-              libuuid
-              lz4
-              udev
-              zlib
-              zstd
-            ];
+                  test "$($out/bin/bcachefs version)" = "${version}"
 
-            meta = {
-              description = "Userspace tools for bcachefs";
-              license = lib.licenses.gpl2Only;
-              mainProgram = "bcachefs";
-            };
-          };
+                  runHook postInstallCheck
+                '';
+              }
+            );
 
-          cargoArtifacts = craneLib.buildDepsOnly (commonArgs // { pname = cargoToml.package.name; });
+          mkPackages =
+            name: systems:
+            let
+              packagesForSystem =
+                crossSystem:
+                let
+                  localSystem = system;
+                  pkgs' = import nixpkgs {
+                    inherit crossSystem localSystem;
+                    overlays = [ (import rust-overlay) ];
+                  };
+
+                  common = pkgs'.callPackage mkCommon { inherit crane; };
+                  package = pkgs'.callPackage mkPackage { inherit common name; };
+                  packageFuse = package.overrideAttrs (
+                    final: prev: {
+                      makeFlags = prev.makeFlags ++ [ "BCACHEFS_FUSE=1" ];
+                      buildInputs = prev.buildInputs ++ [ pkgs'.fuse3 ];
+                    }
+                  );
+                in
+                [
+                  (lib.nameValuePair "${name}-${crossSystem}" package)
+                  (lib.nameValuePair "${name}-fuse-${crossSystem}" packageFuse)
+                ];
+            in
+            lib.listToAttrs (lib.flatten (map packagesForSystem systems));
         in
         {
-          packages.default = config.packages.bcachefs-tools;
-          packages.bcachefs-tools = craneLib.buildPackage (
-            commonArgs
+          packages =
+            let
+              inherit (cargoToml.package) name;
+            in
+            (mkPackages name systems)
             // {
-              inherit cargoArtifacts;
+              ${name} = config.packages."${name}-${system}";
+              "${name}-fuse" = config.packages."${name}-fuse-${system}";
+              default = config.packages.${name};
+            };
 
-              enableParallelBuilding = true;
-              buildPhaseCargoCommand = ''
-                make ''${enableParallelBuilding:+-j''${NIX_BUILD_CORES}} $makeFlags
-              '';
-              installPhaseCommand = ''
-                make ''${enableParallelBuilding:+-j''${NIX_BUILD_CORES}} $makeFlags install
-              '';
+          checks = {
+            inherit (config.packages)
+              bcachefs-tools
+              bcachefs-tools-fuse
+              bcachefs-tools-fuse-i686-linux
+              ;
 
-              doInstallCheck = true;
-              installCheckPhase = ''
-                runHook preInstallCheck
+            cargo-clippy = common.craneLib.cargoClippy (
+              common.args
+              // {
+                inherit (common) cargoArtifacts;
+                cargoClippyExtraArgs = "--all-targets --all-features -- --deny warnings";
+              }
+            );
 
-                test "$($out/bin/bcachefs version)" = "${version}"
+            # we have to build our own `craneLib.cargoTest`
+            cargo-test = common.craneLib.mkCargoDerivation (
+              common.args
+              // {
+                inherit (common) cargoArtifacts;
+                doCheck = true;
 
-                runHook postInstallCheck
-              '';
-            }
-          );
+                enableParallelChecking = true;
 
-          packages.bcachefs-tools-fuse = config.packages.bcachefs-tools.overrideAttrs (
-            final: prev: {
-              makeFlags = prev.makeFlags ++ [ "BCACHEFS_FUSE=1" ];
-              buildInputs = prev.buildInputs ++ [ pkgs.fuse3 ];
-            }
-          );
+                pnameSuffix = "-test";
+                buildPhaseCargoCommand = "";
+                checkPhaseCargoCommand = ''
+                  make ''${enableParallelChecking:+-j''${NIX_BUILD_CORES}} $makeFlags libbcachefs.a
+                  cargo test --profile release -- --nocapture
+                '';
+              }
+            );
 
-          checks.cargo-clippy = craneLib.cargoClippy (
-            commonArgs
-            // {
-              inherit cargoArtifacts;
-              cargoClippyExtraArgs = "--all-targets -- --deny warnings";
-            }
-          );
-
-          # we have to build our own `craneLib.cargoTest`
-          checks.cargo-test = craneLib.mkCargoDerivation (
-            commonArgs
-            // {
-              inherit cargoArtifacts;
-              doCheck = true;
-
-              enableParallelChecking = true;
-
-              pnameSuffix = "-test";
-              buildPhaseCargoCommand = "";
-              checkPhaseCargoCommand = ''
-                make ''${enableParallelChecking:+-j''${NIX_BUILD_CORES}} $makeFlags libbcachefs.a
-                cargo test --profile release -- --nocapture
-              '';
-            }
-          );
+            # cargo clippy with the current minimum supported rust version
+            # according to Cargo.toml
+            msrv =
+              let
+                rustVersion = cargoToml.package.rust-version;
+                common = pkgs.callPackage mkCommon { inherit crane rustVersion; };
+              in
+              common.craneLib.cargoClippy (
+                common.args
+                // {
+                  pname = "msrv";
+                  inherit (common) cargoArtifacts;
+                  cargoClippyExtraArgs = "--all-targets --all-features -- --deny warnings";
+                }
+              );
+          };
 
           devShells.default = pkgs.mkShell {
             inputsFrom = [
@@ -190,9 +302,12 @@
               cargo-audit
               cargo-outdated
               clang-tools
-              clippy
-              rust-analyzer
-              rustc
+              (rust-bin.stable.latest.minimal.override {
+                extensions = [
+                  "rust-analyzer"
+                  "rust-src"
+                ];
+              })
             ];
           };
 
@@ -204,7 +319,7 @@
               nixfmt.enable = true;
               rustfmt.edition = rustfmtToml.edition;
               rustfmt.enable = true;
-              rustfmt.package = fenix.packages.${system}.default.rustfmt;
+              rustfmt.package = pkgs.rust-bin.selectLatestNightlyWith (toolchain: toolchain.rustfmt);
             };
           };
         };

include/linux/compiler.h

@@ -67,6 +67,7 @@
 #define __same_type(a, b)	__builtin_types_compatible_p(typeof(a), typeof(b))
 #define fallthrough		__attribute__((__fallthrough__))
 #define __noreturn		__attribute__((__noreturn__))
+#define __no_kmsan_checks
 
 #ifndef __counted_by
 #define __counted_by(nr)

include/linux/kernel.h

@@ -12,6 +12,7 @@
 #include <linux/byteorder.h>
 #include <linux/compiler.h>
 #include <linux/dcache.h>
+#include <linux/kmsan-checks.h>
 #include <linux/math.h>
 #include <linux/minmax.h>
 

include/linux/kmsan-checks.h

@@ -0,0 +1,98 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * KMSAN checks to be used for one-off annotations in subsystems.
+ *
+ * Copyright (C) 2017-2022 Google LLC
+ * Author: Alexander Potapenko <glider@google.com>
+ *
+ */
+
+#ifndef _LINUX_KMSAN_CHECKS_H
+#define _LINUX_KMSAN_CHECKS_H
+
+#include <linux/types.h>
+
+#ifdef CONFIG_KMSAN
+
+/**
+ * kmsan_poison_memory() - Mark the memory range as uninitialized.
+ * @address: address to start with.
+ * @size:    size of buffer to poison.
+ * @flags:   GFP flags for allocations done by this function.
+ *
+ * Until other data is written to this range, KMSAN will treat it as
+ * uninitialized. Error reports for this memory will reference the call site of
+ * kmsan_poison_memory() as origin.
+ */
+void kmsan_poison_memory(const void *address, size_t size, gfp_t flags);
+
+/**
+ * kmsan_unpoison_memory() -  Mark the memory range as initialized.
+ * @address: address to start with.
+ * @size:    size of buffer to unpoison.
+ *
+ * Until other data is written to this range, KMSAN will treat it as
+ * initialized.
+ */
+void kmsan_unpoison_memory(const void *address, size_t size);
+
+/**
+ * kmsan_check_memory() - Check the memory range for being initialized.
+ * @address: address to start with.
+ * @size:    size of buffer to check.
+ *
+ * If any piece of the given range is marked as uninitialized, KMSAN will report
+ * an error.
+ */
+void kmsan_check_memory(const void *address, size_t size);
+
+/**
+ * kmsan_copy_to_user() - Notify KMSAN about a data transfer to userspace.
+ * @to:      destination address in the userspace.
+ * @from:    source address in the kernel.
+ * @to_copy: number of bytes to copy.
+ * @left:    number of bytes not copied.
+ *
+ * If this is a real userspace data transfer, KMSAN checks the bytes that were
+ * actually copied to ensure there was no information leak. If @to belongs to
+ * the kernel space (which is possible for compat syscalls), KMSAN just copies
+ * the metadata.
+ */
+void kmsan_copy_to_user(void __user *to, const void *from, size_t to_copy,
+			size_t left);
+
+/**
+ * kmsan_memmove() - Notify KMSAN about a data copy within kernel.
+ * @to:   destination address in the kernel.
+ * @from: source address in the kernel.
+ * @size: number of bytes to copy.
+ *
+ * Invoked after non-instrumented version (e.g. implemented using assembly
+ * code) of memmove()/memcpy() is called, in order to copy KMSAN's metadata.
+ */
+void kmsan_memmove(void *to, const void *from, size_t to_copy);
+
+#else
+
+static inline void kmsan_poison_memory(const void *address, size_t size,
+				       gfp_t flags)
+{
+}
+static inline void kmsan_unpoison_memory(const void *address, size_t size)
+{
+}
+static inline void kmsan_check_memory(const void *address, size_t size)
+{
+}
+static inline void kmsan_copy_to_user(void __user *to, const void *from,
+				      size_t to_copy, size_t left)
+{
+}
+
+static inline void kmsan_memmove(void *to, const void *from, size_t to_copy)
+{
+}
+
+#endif
+
+#endif /* _LINUX_KMSAN_CHECKS_H */

include/linux/mm.h

@@ -4,6 +4,7 @@
 
 #include <sys/syscall.h>
 #include <unistd.h>
+#include <linux/bug.h>
 #include <linux/types.h>
 
 struct sysinfo {

include/linux/sched/sysctl.h

@@ -0,0 +1,6 @@
+#ifndef __TOOLS_LINUX_SCHED_SYSCTL_H
+#define __TOOLS_LINUX_SCHED_SYSCTL_H
+
+#define sysctl_hung_task_timeout_secs		(HZ * 10)
+
+#endif /* __TOOLS_LINUX_SCHED_SYSCTL_H */

libbcachefs/alloc_background.c

@@ -589,6 +589,8 @@ iter_err:
 
 int bch2_alloc_read(struct bch_fs *c)
 {
+	down_read(&c->state_lock);
+
 	struct btree_trans *trans = bch2_trans_get(c);
 	struct bch_dev *ca = NULL;
 	int ret;
@@ -652,6 +654,7 @@ int bch2_alloc_read(struct bch_fs *c)
 	bch2_dev_put(ca);
 	bch2_trans_put(trans);
 
+	up_read(&c->state_lock);
 	bch_err_fn(c, ret);
 	return ret;
 }
@@ -673,8 +676,7 @@ static int __need_discard_or_freespace_err(struct btree_trans *trans,
 	bch2_bkey_val_to_text(&buf, c, alloc_k);
 
 	int ret = __bch2_fsck_err(NULL, trans, flags, err_id,
-				  "bucket incorrectly %sset in %s btree\n"
-				  "  %s",
+				  "bucket incorrectly %sset in %s btree\n%s",
 				  set ? "" : "un",
 				  bch2_btree_id_str(btree),
 				  buf.buf);
@@ -777,14 +779,12 @@ static inline int bch2_dev_data_type_accounting_mod(struct btree_trans *trans, s
 						    s64 delta_sectors,
 						    s64 delta_fragmented, unsigned flags)
 {
-	struct disk_accounting_pos acc = {
-		.type = BCH_DISK_ACCOUNTING_dev_data_type,
-		.dev_data_type.dev		= ca->dev_idx,
-		.dev_data_type.data_type	= data_type,
-	};
 	s64 d[3] = { delta_buckets, delta_sectors, delta_fragmented };
 
-	return bch2_disk_accounting_mod(trans, &acc, d, 3, flags & BTREE_TRIGGER_gc);
+	return bch2_disk_accounting_mod2(trans, flags & BTREE_TRIGGER_gc,
+					 d, dev_data_type,
+					 .dev		= ca->dev_idx,
+					 .data_type	= data_type);
 }
 
 int bch2_alloc_key_to_dev_counters(struct btree_trans *trans, struct bch_dev *ca,
@@ -837,7 +837,7 @@ int bch2_trigger_alloc(struct btree_trans *trans,
 
 	struct bch_dev *ca = bch2_dev_bucket_tryget(c, new.k->p);
 	if (!ca)
-		return -EIO;
+		return -BCH_ERR_trigger_alloc;
 
 	struct bch_alloc_v4 old_a_convert;
 	const struct bch_alloc_v4 *old_a = bch2_alloc_to_v4(old, &old_a_convert);
@@ -1029,9 +1029,9 @@ fsck_err:
 	bch2_dev_put(ca);
 	return ret;
 invalid_bucket:
-	bch2_fs_inconsistent(c, "reference to invalid bucket\n  %s",
+	bch2_fs_inconsistent(c, "reference to invalid bucket\n%s",
 			     (bch2_bkey_val_to_text(&buf, c, new.s_c), buf.buf));
-	ret = -EIO;
+	ret = -BCH_ERR_trigger_alloc;
 	goto err;
 }
 
@@ -1203,8 +1203,7 @@ int bch2_check_alloc_key(struct btree_trans *trans,
 
 	if (fsck_err_on(a->gen != alloc_gen(k, gens_offset),
 			trans, bucket_gens_key_wrong,
-			"incorrect gen in bucket_gens btree (got %u should be %u)\n"
-			"  %s",
+			"incorrect gen in bucket_gens btree (got %u should be %u)\n%s",
 			alloc_gen(k, gens_offset), a->gen,
 			(printbuf_reset(&buf),
 			 bch2_bkey_val_to_text(&buf, c, alloc_k), buf.buf))) {
@@ -1262,7 +1261,7 @@ int bch2_check_alloc_hole_freespace(struct btree_trans *trans,
 	if (fsck_err_on(k.k->type != KEY_TYPE_set,
 			trans, freespace_hole_missing,
 			"hole in alloc btree missing in freespace btree\n"
-			"  device %llu buckets %llu-%llu",
+			"device %llu buckets %llu-%llu",
 			freespace_iter->pos.inode,
 			freespace_iter->pos.offset,
 			end->offset)) {
@@ -1421,7 +1420,7 @@ int bch2_check_discard_freespace_key(struct btree_trans *trans, struct btree_ite
 	    (state == BCH_DATA_free &&
 	     genbits != alloc_freespace_genbits(*a))) {
 		if (fsck_err(trans, need_discard_freespace_key_bad,
-			     "%s\n  incorrectly set at %s:%llu:%llu:0 (free %u, genbits %llu should be %llu)",
+			     "%s\nincorrectly set at %s:%llu:%llu:0 (free %u, genbits %llu should be %llu)",
 			     (bch2_bkey_val_to_text(&buf, c, alloc_k), buf.buf),
 			     bch2_btree_id_str(iter->btree_id),
 			     iter->pos.inode,
@@ -1502,7 +1501,7 @@ int bch2_check_bucket_gens_key(struct btree_trans *trans,
 	struct bch_dev *ca = bch2_dev_tryget_noerror(c, k.k->p.inode);
 	if (!ca) {
 		if (fsck_err(trans, bucket_gens_to_invalid_dev,
-			     "bucket_gens key for invalid device:\n  %s",
+			     "bucket_gens key for invalid device:\n%s",
 			     (bch2_bkey_val_to_text(&buf, c, k), buf.buf)))
 			ret = bch2_btree_delete_at(trans, iter, 0);
 		goto out;
@@ -1511,7 +1510,7 @@ int bch2_check_bucket_gens_key(struct btree_trans *trans,
 	if (fsck_err_on(end <= ca->mi.first_bucket ||
 			start >= ca->mi.nbuckets,
 			trans, bucket_gens_to_invalid_buckets,
-			"bucket_gens key for invalid buckets:\n  %s",
+			"bucket_gens key for invalid buckets:\n%s",
 			(bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
 		ret = bch2_btree_delete_at(trans, iter, 0);
 		goto out;
@@ -1714,8 +1713,7 @@ static int bch2_check_alloc_to_lru_ref(struct btree_trans *trans,
 
 	if (fsck_err_on(!a->io_time[READ],
 			trans, alloc_key_cached_but_read_time_zero,
-			"cached bucket with read_time 0\n"
-			"  %s",
+			"cached bucket with read_time 0\n%s",
 		(printbuf_reset(&buf),
 		 bch2_bkey_val_to_text(&buf, c, alloc_k), buf.buf))) {
 		struct bkey_i_alloc_v4 *a_mut =

libbcachefs/alloc_foreground.c

@@ -127,14 +127,14 @@ void __bch2_open_bucket_put(struct bch_fs *c, struct open_bucket *ob)
 
 void bch2_open_bucket_write_error(struct bch_fs *c,
 				  struct open_buckets *obs,
-				  unsigned dev)
+				  unsigned dev, int err)
 {
 	struct open_bucket *ob;
 	unsigned i;
 
 	open_bucket_for_each(c, obs, ob, i)
 		if (ob->dev == dev && ob->ec)
-			bch2_ec_bucket_cancel(c, ob);
+			bch2_ec_bucket_cancel(c, ob, err);
 }
 
 static struct open_bucket *bch2_open_bucket_alloc(struct bch_fs *c)
@@ -631,7 +631,7 @@ static inline void bch2_dev_stripe_increment_inlined(struct bch_dev *ca,
 			       struct bch_dev_usage *usage)
 {
 	u64 *v = stripe->next_alloc + ca->dev_idx;
-	u64 free_space = dev_buckets_available(ca, BCH_WATERMARK_normal);
+	u64 free_space = __dev_buckets_available(ca, *usage, BCH_WATERMARK_normal);
 	u64 free_space_inv = free_space
 		? div64_u64(1ULL << 48, free_space)
 		: 1ULL << 48;

libbcachefs/alloc_foreground.h

@@ -82,7 +82,7 @@ static inline struct open_bucket *ec_open_bucket(struct bch_fs *c,
 }
 
 void bch2_open_bucket_write_error(struct bch_fs *,
-			struct open_buckets *, unsigned);
+			struct open_buckets *, unsigned, int);
 
 void __bch2_open_bucket_put(struct bch_fs *, struct open_bucket *);
 

libbcachefs/backpointers.c

@@ -50,6 +50,8 @@ void bch2_backpointer_to_text(struct printbuf *out, struct bch_fs *c, struct bke
 	}
 
 	bch2_btree_id_level_to_text(out, bp.v->btree_id, bp.v->level);
+	prt_str(out, " data_type=");
+	bch2_prt_data_type(out, bp.v->data_type);
 	prt_printf(out, " suboffset=%u len=%u gen=%u pos=",
 		   (u32) bp.k->p.offset & ~(~0U << MAX_EXTENT_COMPRESS_RATIO_SHIFT),
 		   bp.v->bucket_len,
@@ -94,6 +96,7 @@ static noinline int backpointer_mod_err(struct btree_trans *trans,
 {
 	struct bch_fs *c = trans->c;
 	struct printbuf buf = PRINTBUF;
+	int ret = 0;
 
 	if (insert) {
 		prt_printf(&buf, "existing backpointer found when inserting ");
@@ -123,17 +126,15 @@ static noinline int backpointer_mod_err(struct btree_trans *trans,
 
 		prt_printf(&buf, "for ");
 		bch2_bkey_val_to_text(&buf, c, orig_k);
-
-		bch_err(c, "%s", buf.buf);
 	}
 
+	if (c->curr_recovery_pass > BCH_RECOVERY_PASS_check_extents_to_backpointers &&
+	    __bch2_inconsistent_error(c, &buf))
+		ret = -BCH_ERR_erofs_unfixed_errors;
+
+	bch_err(c, "%s", buf.buf);
 	printbuf_exit(&buf);
-
-	if (c->curr_recovery_pass > BCH_RECOVERY_PASS_check_extents_to_backpointers) {
-		return bch2_inconsistent_error(c) ? BCH_ERR_erofs_unfixed_errors : 0;
-	} else {
-		return 0;
-	}
+	return ret;
 }
 
 int bch2_bucket_backpointer_mod_nowritebuffer(struct btree_trans *trans,
@@ -208,11 +209,11 @@ static int backpointer_target_not_found(struct btree_trans *trans,
 	if (ret)
 		return ret;
 
-	prt_printf(&buf, "backpointer doesn't match %s it points to:\n  ",
+	prt_printf(&buf, "backpointer doesn't match %s it points to:\n",
 		   bp.v->level ? "btree node" : "extent");
 	bch2_bkey_val_to_text(&buf, c, bp.s_c);
 
-	prt_printf(&buf, "\n  ");
+	prt_newline(&buf);
 	bch2_bkey_val_to_text(&buf, c, target_k);
 
 	struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(target_k);
@@ -220,7 +221,7 @@ static int backpointer_target_not_found(struct btree_trans *trans,
 	struct extent_ptr_decoded p;
 	bkey_for_each_ptr_decode(target_k.k, ptrs, p, entry)
 		if (p.ptr.dev == bp.k->p.inode) {
-			prt_printf(&buf, "\n  ");
+			prt_newline(&buf);
 			struct bkey_i_backpointer bp2;
 			bch2_extent_ptr_to_bp(c, bp.v->btree_id, bp.v->level, target_k, p, entry, &bp2);
 			bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&bp2.k_i));
@@ -441,12 +442,11 @@ found:
 	if (ret)
 		goto err;
 
-	prt_str(&buf, "extents pointing to same space, but first extent checksum bad:");
-	prt_printf(&buf, "\n  ");
+	prt_printf(&buf, "extents pointing to same space, but first extent checksum bad:\n");
 	bch2_btree_id_to_text(&buf, btree);
 	prt_str(&buf, " ");
 	bch2_bkey_val_to_text(&buf, c, extent);
-	prt_printf(&buf, "\n  ");
+	prt_newline(&buf);
 	bch2_btree_id_to_text(&buf, o_btree);
 	prt_str(&buf, " ");
 	bch2_bkey_val_to_text(&buf, c, extent2);
@@ -537,9 +537,9 @@ check_existing_bp:
 
 	if (bch2_extents_match(orig_k, other_extent)) {
 		printbuf_reset(&buf);
-		prt_printf(&buf, "duplicate versions of same extent, deleting smaller\n  ");
+		prt_printf(&buf, "duplicate versions of same extent, deleting smaller\n");
 		bch2_bkey_val_to_text(&buf, c, orig_k);
-		prt_str(&buf, "\n  ");
+		prt_newline(&buf);
 		bch2_bkey_val_to_text(&buf, c, other_extent);
 		bch_err(c, "%s", buf.buf);
 
@@ -578,20 +578,20 @@ check_existing_bp:
 	}
 
 	printbuf_reset(&buf);
-	prt_printf(&buf, "duplicate extents pointing to same space on dev %llu\n  ", bp->k.p.inode);
+	prt_printf(&buf, "duplicate extents pointing to same space on dev %llu\n", bp->k.p.inode);
 	bch2_bkey_val_to_text(&buf, c, orig_k);
-	prt_str(&buf, "\n  ");
+	prt_newline(&buf);
 	bch2_bkey_val_to_text(&buf, c, other_extent);
 	bch_err(c, "%s", buf.buf);
 	ret = -BCH_ERR_fsck_repair_unimplemented;
 	goto err;
 missing:
 	printbuf_reset(&buf);
-	prt_str(&buf, "missing backpointer\n  for:  ");
+	prt_str(&buf, "missing backpointer\nfor:  ");
 	bch2_bkey_val_to_text(&buf, c, orig_k);
-	prt_printf(&buf, "\n  want: ");
+	prt_printf(&buf, "\nwant: ");
 	bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&bp->k_i));
-	prt_printf(&buf, "\n  got:  ");
+	prt_printf(&buf, "\ngot:  ");
 	bch2_bkey_val_to_text(&buf, c, bp_k);
 
 	if (fsck_err(trans, ptr_to_missing_backpointer, "%s", buf.buf))
@@ -782,7 +782,7 @@ enum alloc_sector_counter {
 	ALLOC_SECTORS_NR
 };
 
-static enum alloc_sector_counter data_type_to_alloc_counter(enum bch_data_type t)
+static int data_type_to_alloc_counter(enum bch_data_type t)
 {
 	switch (t) {
 	case BCH_DATA_btree:
@@ -791,9 +791,10 @@ static enum alloc_sector_counter data_type_to_alloc_counter(enum bch_data_type t
 	case BCH_DATA_cached:
 		return ALLOC_cached;
 	case BCH_DATA_stripe:
+	case BCH_DATA_parity:
 		return ALLOC_stripe;
 	default:
-		BUG();
+		return -1;
 	}
 }
 
@@ -844,7 +845,11 @@ static int check_bucket_backpointer_mismatch(struct btree_trans *trans, struct b
 		if (bp.v->bucket_gen != a->gen)
 			continue;
 
-		sectors[data_type_to_alloc_counter(bp.v->data_type)] += bp.v->bucket_len;
+		int alloc_counter = data_type_to_alloc_counter(bp.v->data_type);
+		if (alloc_counter < 0)
+			continue;
+
+		sectors[alloc_counter] += bp.v->bucket_len;
 	};
 	bch2_trans_iter_exit(trans, &iter);
 	if (ret)
@@ -1016,7 +1021,7 @@ int bch2_check_extents_to_backpointers(struct bch_fs *c)
 	 * Can't allow devices to come/go/resize while we have bucket bitmaps
 	 * allocated
 	 */
-	lockdep_assert_held(&c->state_lock);
+	down_read(&c->state_lock);
 
 	for_each_member_device(c, ca) {
 		BUG_ON(ca->bucket_backpointer_mismatches);
@@ -1101,6 +1106,7 @@ err_free_bitmaps:
 		ca->bucket_backpointer_mismatches = NULL;
 	}
 
+	up_read(&c->state_lock);
 	bch_err_fn(c, ret);
 	return ret;
 }

libbcachefs/bcachefs_format.h

@@ -1143,7 +1143,8 @@ static inline __u64 __bset_magic(struct bch_sb *sb)
 	x(log,			9)		\
 	x(overwrite,		10)		\
 	x(write_buffer_keys,	11)		\
-	x(datetime,		12)
+	x(datetime,		12)		\
+	x(log_bkey,		13)
 
 enum bch_jset_entry_type {
 #define x(f, nr)	BCH_JSET_ENTRY_##f	= nr,

libbcachefs/btree_cache.c

@@ -1417,7 +1417,7 @@ void __bch2_btree_pos_to_text(struct printbuf *out, struct bch_fs *c,
 		prt_printf(out, "%u", r->level);
 	else
 		prt_printf(out, "(unknown)");
-	prt_printf(out, "\n  ");
+	prt_newline(out);
 
 	bch2_bkey_val_to_text(out, c, k);
 }

libbcachefs/btree_gc.c

@@ -213,15 +213,15 @@ static int btree_check_node_boundaries(struct btree_trans *trans, struct btree *
 
 	prt_printf(&buf, "  at ");
 	bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level);
-	prt_printf(&buf, ":\n  parent: ");
+	prt_printf(&buf, ":\nparent: ");
 	bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key));
 
 	if (prev) {
-		prt_printf(&buf, "\n  prev: ");
+		prt_printf(&buf, "\nprev: ");
 		bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&prev->key));
 	}
 
-	prt_str(&buf, "\n  next: ");
+	prt_str(&buf, "\nnext: ");
 	bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&cur->key));
 
 	if (bpos_lt(expected_start, cur->data->min_key)) {				/* gap */
@@ -280,12 +280,12 @@ static int btree_repair_node_end(struct btree_trans *trans, struct btree *b,
 	if (bpos_eq(child->key.k.p, b->key.k.p))
 		return 0;
 
-	prt_printf(&buf, "  at ");
+	prt_printf(&buf, "\nat: ");
 	bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level);
-	prt_printf(&buf, ":\n  parent: ");
+	prt_printf(&buf, "\nparent: ");
 	bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key));
 
-	prt_str(&buf, "\n  child: ");
+	prt_str(&buf, "\nchild: ");
 	bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&child->key));
 
 	if (mustfix_fsck_err(trans, btree_node_topology_bad_max_key,
@@ -351,8 +351,7 @@ again:
 
 		if (mustfix_fsck_err_on(bch2_err_matches(ret, EIO),
 				trans, btree_node_read_error,
-				"Topology repair: unreadable btree node at\n"
-				"  %s",
+				"Topology repair: unreadable btree node at\n%s",
 				buf.buf)) {
 			bch2_btree_node_evict(trans, cur_k.k);
 			cur = NULL;
@@ -612,7 +611,7 @@ static int bch2_gc_mark_key(struct btree_trans *trans, enum btree_id btree_id,
 		if (fsck_err_on(btree_id != BTREE_ID_accounting &&
 				k.k->bversion.lo > atomic64_read(&c->key_version),
 				trans, bkey_version_in_future,
-				"key version number higher than recorded %llu\n  %s",
+				"key version number higher than recorded %llu\n%s",
 				atomic64_read(&c->key_version),
 				(bch2_bkey_val_to_text(&buf, c, k), buf.buf)))
 			atomic64_set(&c->key_version, k.k->bversion.lo);
@@ -620,7 +619,7 @@ static int bch2_gc_mark_key(struct btree_trans *trans, enum btree_id btree_id,
 
 	if (mustfix_fsck_err_on(level && !bch2_dev_btree_bitmap_marked(c, k),
 				trans, btree_bitmap_not_marked,
-				"btree ptr not marked in member info btree allocated bitmap\n  %s",
+				"btree ptr not marked in member info btree allocated bitmap\n%s",
 				(printbuf_reset(&buf),
 				 bch2_bkey_val_to_text(&buf, c, k),
 				 buf.buf))) {
@@ -1021,8 +1020,7 @@ int bch2_check_allocations(struct bch_fs *c)
 {
 	int ret;
 
-	lockdep_assert_held(&c->state_lock);
-
+	down_read(&c->state_lock);
 	down_write(&c->gc_lock);
 
 	bch2_btree_interior_updates_flush(c);
@@ -1060,6 +1058,7 @@ out:
 	percpu_up_write(&c->mark_lock);
 
 	up_write(&c->gc_lock);
+	up_read(&c->state_lock);
 
 	/*
 	 * At startup, allocations can happen directly instead of via the

libbcachefs/btree_io.c

@@ -525,8 +525,6 @@ static void btree_err_msg(struct printbuf *out, struct bch_fs *c,
 	prt_printf(out, "at btree ");
 	bch2_btree_pos_to_text(out, c, b);
 
-	printbuf_indent_add(out, 2);
-
 	prt_printf(out, "\nnode offset %u/%u",
 		   b->written, btree_ptr_sectors_written(bkey_i_to_s_c(&b->key)));
 	if (i)
@@ -550,23 +548,7 @@ static int __btree_err(int ret,
 		       enum bch_sb_error_id err_type,
 		       const char *fmt, ...)
 {
-	struct printbuf out = PRINTBUF;
 	bool silent = c->curr_recovery_pass == BCH_RECOVERY_PASS_scan_for_btree_nodes;
-	va_list args;
-
-	btree_err_msg(&out, c, ca, b, i, k, b->written, write);
-
-	va_start(args, fmt);
-	prt_vprintf(&out, fmt, args);
-	va_end(args);
-
-	if (write == WRITE) {
-		bch2_print_string_as_lines(KERN_ERR, out.buf);
-		ret = c->opts.errors == BCH_ON_ERROR_continue
-			? 0
-			: -BCH_ERR_fsck_errors_not_fixed;
-		goto out;
-	}
 
 	if (!have_retry && ret == -BCH_ERR_btree_node_read_err_want_retry)
 		ret = -BCH_ERR_btree_node_read_err_fixable;
@@ -576,6 +558,29 @@ static int __btree_err(int ret,
 	if (!silent && ret != -BCH_ERR_btree_node_read_err_fixable)
 		bch2_sb_error_count(c, err_type);
 
+	struct printbuf out = PRINTBUF;
+	if (write != WRITE && ret != -BCH_ERR_btree_node_read_err_fixable) {
+		printbuf_indent_add_nextline(&out, 2);
+#ifdef BCACHEFS_LOG_PREFIX
+		prt_printf(&out, bch2_log_msg(c, ""));
+#endif
+	}
+
+	btree_err_msg(&out, c, ca, b, i, k, b->written, write);
+
+	va_list args;
+	va_start(args, fmt);
+	prt_vprintf(&out, fmt, args);
+	va_end(args);
+
+	if (write == WRITE) {
+		prt_str(&out, ", ");
+		ret = __bch2_inconsistent_error(c, &out)
+			? -BCH_ERR_fsck_errors_not_fixed
+			: 0;
+		silent = false;
+	}
+
 	switch (ret) {
 	case -BCH_ERR_btree_node_read_err_fixable:
 		ret = !silent
@@ -585,25 +590,21 @@ static int __btree_err(int ret,
 		    ret != -BCH_ERR_fsck_ignore)
 			goto fsck_err;
 		ret = -BCH_ERR_fsck_fix;
-		break;
-	case -BCH_ERR_btree_node_read_err_want_retry:
-	case -BCH_ERR_btree_node_read_err_must_retry:
-		if (!silent)
-			bch2_print_string_as_lines(KERN_ERR, out.buf);
-		break;
+		goto out;
 	case -BCH_ERR_btree_node_read_err_bad_node:
-		if (!silent)
-			bch2_print_string_as_lines(KERN_ERR, out.buf);
-		ret = bch2_topology_error(c);
+		prt_str(&out, ", ");
+		ret = __bch2_topology_error(c, &out);
+		if (ret)
+			silent = false;
 		break;
 	case -BCH_ERR_btree_node_read_err_incompatible:
-		if (!silent)
-			bch2_print_string_as_lines(KERN_ERR, out.buf);
 		ret = -BCH_ERR_fsck_errors_not_fixed;
+		silent = false;
 		break;
-	default:
-		BUG();
 	}
+
+	if (!silent)
+		bch2_print_string_as_lines(KERN_ERR, out.buf);
 out:
 fsck_err:
 	printbuf_exit(&out);
@@ -817,7 +818,7 @@ static int validate_bset(struct bch_fs *c, struct bch_dev *ca,
 			     -BCH_ERR_btree_node_read_err_bad_node,
 			     c, ca, b, i, NULL,
 			     btree_node_bad_format,
-			     "invalid bkey format: %s\n  %s", buf1.buf,
+			     "invalid bkey format: %s\n%s", buf1.buf,
 			     (printbuf_reset(&buf2),
 			      bch2_bkey_format_to_text(&buf2, &bn->format), buf2.buf));
 		printbuf_reset(&buf1);
@@ -2117,8 +2118,14 @@ out:
 	return;
 err:
 	set_btree_node_noevict(b);
-	bch2_fs_fatal_err_on(!bch2_err_matches(ret, EROFS), c,
-			     "writing btree node: %s", bch2_err_str(ret));
+
+	if (!bch2_err_matches(ret, EROFS)) {
+		struct printbuf buf = PRINTBUF;
+		prt_printf(&buf, "writing btree node: %s\n  ", bch2_err_str(ret));
+		bch2_btree_pos_to_text(&buf, c, b);
+		bch2_fs_fatal_error(c, "%s", buf.buf);
+		printbuf_exit(&buf);
+	}
 	goto out;
 }
 
@@ -2135,10 +2142,14 @@ static void btree_node_write_endio(struct bio *bio)
 	bch2_account_io_completion(ca, BCH_MEMBER_ERROR_write,
 				   wbio->submit_time, !bio->bi_status);
 
-	if (ca && bio->bi_status)
-		bch_err_dev_ratelimited(ca,
-				   "btree write error: %s",
-				   bch2_blk_status_to_str(bio->bi_status));
+	if (ca && bio->bi_status) {
+		struct printbuf buf = PRINTBUF;
+		prt_printf(&buf, "btree write error: %s\n  ",
+			   bch2_blk_status_to_str(bio->bi_status));
+		bch2_btree_pos_to_text(&buf, c, b);
+		bch_err_dev_ratelimited(ca, "%s", buf.buf);
+		printbuf_exit(&buf);
+	}
 
 	if (bio->bi_status) {
 		unsigned long flags;

libbcachefs/btree_iter.c

@@ -1487,22 +1487,14 @@ void bch2_trans_updates_to_text(struct printbuf *buf, struct btree_trans *trans)
 
 	for (struct jset_entry *e = trans->journal_entries;
 	     e != btree_trans_journal_entries_top(trans);
-	     e = vstruct_next(e))
+	     e = vstruct_next(e)) {
 		bch2_journal_entry_to_text(buf, trans->c, e);
+		prt_newline(buf);
+	}
 
 	printbuf_indent_sub(buf, 2);
 }
 
-noinline __cold
-void bch2_dump_trans_updates(struct btree_trans *trans)
-{
-	struct printbuf buf = PRINTBUF;
-
-	bch2_trans_updates_to_text(&buf, trans);
-	bch2_print_str(trans->c, buf.buf);
-	printbuf_exit(&buf);
-}
-
 static void bch2_btree_path_to_text_short(struct printbuf *out, struct btree_trans *trans, btree_path_idx_t path_idx)
 {
 	struct btree_path *path = trans->paths + path_idx;

libbcachefs/btree_iter.h

@@ -9,7 +9,6 @@
 void bch2_trans_updates_to_text(struct printbuf *, struct btree_trans *);
 void bch2_btree_path_to_text(struct printbuf *, struct btree_trans *, btree_path_idx_t);
 void bch2_trans_paths_to_text(struct printbuf *, struct btree_trans *);
-void bch2_dump_trans_updates(struct btree_trans *);
 void bch2_dump_trans_paths_updates(struct btree_trans *);
 
 static inline int __bkey_err(const struct bkey *k)
@@ -335,13 +334,20 @@ static inline void bch2_trans_verify_not_unlocked_or_in_restart(struct btree_tra
 }
 
 __always_inline
-static int btree_trans_restart_ip(struct btree_trans *trans, int err, unsigned long ip)
+static int btree_trans_restart_foreign_task(struct btree_trans *trans, int err, unsigned long ip)
 {
 	BUG_ON(err <= 0);
 	BUG_ON(!bch2_err_matches(-err, BCH_ERR_transaction_restart));
 
 	trans->restarted = err;
 	trans->last_restarted_ip = ip;
+	return -err;
+}
+
+__always_inline
+static int btree_trans_restart_ip(struct btree_trans *trans, int err, unsigned long ip)
+{
+	btree_trans_restart_foreign_task(trans, err, ip);
 #ifdef CONFIG_BCACHEFS_DEBUG
 	darray_exit(&trans->last_restarted_trace);
 	bch2_save_backtrace(&trans->last_restarted_trace, current, 0, GFP_NOWAIT);

libbcachefs/btree_journal_iter.c

@@ -644,6 +644,8 @@ void bch2_btree_and_journal_iter_init_node_iter(struct btree_trans *trans,
  */
 static int journal_sort_key_cmp(const void *_l, const void *_r)
 {
+	cond_resched();
+
 	const struct journal_key *l = _l;
 	const struct journal_key *r = _r;
 

libbcachefs/btree_locking.c

@@ -91,10 +91,10 @@ static noinline void print_chain(struct printbuf *out, struct lock_graph *g)
 	struct trans_waiting_for_lock *i;
 
 	for (i = g->g; i != g->g + g->nr; i++) {
-		struct task_struct *task = i->trans->locking_wait.task;
+		struct task_struct *task = READ_ONCE(i->trans->locking_wait.task);
 		if (i != g->g)
 			prt_str(out, "<- ");
-		prt_printf(out, "%u ", task ?task->pid : 0);
+		prt_printf(out, "%u ", task ? task->pid : 0);
 	}
 	prt_newline(out);
 }
@@ -172,7 +172,9 @@ static int abort_lock(struct lock_graph *g, struct trans_waiting_for_lock *i)
 {
 	if (i == g->g) {
 		trace_would_deadlock(g, i->trans);
-		return btree_trans_restart(i->trans, BCH_ERR_transaction_restart_would_deadlock);
+		return btree_trans_restart_foreign_task(i->trans,
+					BCH_ERR_transaction_restart_would_deadlock,
+					_THIS_IP_);
 	} else {
 		i->trans->lock_must_abort = true;
 		wake_up_process(i->trans->locking_wait.task);

libbcachefs/btree_node_scan.c

@@ -13,6 +13,7 @@
 
 #include <linux/kthread.h>
 #include <linux/min_heap.h>
+#include <linux/sched/sysctl.h>
 #include <linux/sort.h>
 
 struct find_btree_nodes_worker {
@@ -313,7 +314,8 @@ static int read_btree_nodes(struct find_btree_nodes *f)
 		wake_up_process(t);
 	}
 err:
-	closure_sync(&cl);
+	while (closure_sync_timeout(&cl, sysctl_hung_task_timeout_secs * HZ / 2))
+		;
 	return f->ret ?: ret;
 }
 
@@ -577,10 +579,12 @@ int bch2_get_scanned_nodes(struct bch_fs *c, enum btree_id btree,
 
 		found_btree_node_to_key(&tmp.k, &n);
 
-		struct printbuf buf = PRINTBUF;
-		bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&tmp.k));
-		bch_verbose(c, "%s(): recovering %s", __func__, buf.buf);
-		printbuf_exit(&buf);
+		if (c->opts.verbose) {
+			struct printbuf buf = PRINTBUF;
+			bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&tmp.k));
+			bch_verbose(c, "%s(): recovering %s", __func__, buf.buf);
+			printbuf_exit(&buf);
+		}
 
 		BUG_ON(bch2_bkey_validate(c, bkey_i_to_s_c(&tmp.k),
 					  (struct bkey_validate_context) {

libbcachefs/btree_trans_commit.c

@@ -164,6 +164,7 @@ bool bch2_btree_bset_insert_key(struct btree_trans *trans,
 	EBUG_ON(bpos_gt(insert->k.p, b->data->max_key));
 	EBUG_ON(insert->k.u64s > bch2_btree_keys_u64s_remaining(b));
 	EBUG_ON(!b->c.level && !bpos_eq(insert->k.p, path->pos));
+	kmsan_check_memory(insert, bkey_bytes(&insert->k));
 
 	k = bch2_btree_node_iter_peek_all(node_iter, b);
 	if (k && bkey_cmp_left_packed(b, k, &insert->k.p))

libbcachefs/btree_update.c

@@ -512,6 +512,8 @@ static noinline int bch2_trans_update_get_key_cache(struct btree_trans *trans,
 int __must_check bch2_trans_update(struct btree_trans *trans, struct btree_iter *iter,
 				   struct bkey_i *k, enum btree_iter_update_trigger_flags flags)
 {
+	kmsan_check_memory(k, bkey_bytes(&k->k));
+
 	btree_path_idx_t path_idx = iter->update_path ?: iter->path;
 	int ret;
 
@@ -844,6 +846,19 @@ int bch2_trans_log_msg(struct btree_trans *trans, struct printbuf *buf)
 	return 0;
 }
 
+int bch2_trans_log_bkey(struct btree_trans *trans, enum btree_id btree,
+			unsigned level, struct bkey_i *k)
+{
+	struct jset_entry *e = bch2_trans_jset_entry_alloc(trans, jset_u64s(k->k.u64s));
+	int ret = PTR_ERR_OR_ZERO(e);
+	if (ret)
+		return ret;
+
+	journal_entry_init(e, BCH_JSET_ENTRY_log_bkey, btree, level, k->k.u64s);
+	bkey_copy(e->start, k);
+	return 0;
+}
+
 __printf(3, 0)
 static int
 __bch2_fs_log_msg(struct bch_fs *c, unsigned commit_flags, const char *fmt,

libbcachefs/btree_update.h

@@ -133,6 +133,8 @@ static inline int __must_check bch2_trans_update_buffered(struct btree_trans *tr
 					    enum btree_id btree,
 					    struct bkey_i *k)
 {
+	kmsan_check_memory(k, bkey_bytes(&k->k));
+
 	if (unlikely(!btree_type_uses_write_buffer(btree))) {
 		int ret = bch2_btree_write_buffer_insert_err(trans, btree, k);
 		dump_stack();
@@ -168,6 +170,8 @@ void bch2_trans_commit_hook(struct btree_trans *,
 int __bch2_trans_commit(struct btree_trans *, unsigned);
 
 int bch2_trans_log_msg(struct btree_trans *, struct printbuf *);
+int bch2_trans_log_bkey(struct btree_trans *, enum btree_id, unsigned, struct bkey_i *);
+
 __printf(2, 3) int bch2_fs_log_msg(struct bch_fs *, const char *, ...);
 __printf(2, 3) int bch2_journal_log_msg(struct bch_fs *, const char *, ...);
 

libbcachefs/btree_update_interior.c

@@ -35,6 +35,8 @@ static const char * const bch2_btree_update_modes[] = {
 	NULL
 };
 
+static void bch2_btree_update_to_text(struct printbuf *, struct btree_update *);
+
 static int bch2_btree_insert_node(struct btree_update *, struct btree_trans *,
 				  btree_path_idx_t, struct btree *, struct keylist *);
 static void bch2_btree_update_add_new_node(struct btree_update *, struct btree *);
@@ -54,6 +56,8 @@ int bch2_btree_node_check_topology(struct btree_trans *trans, struct btree *b)
 	struct bkey_buf prev;
 	int ret = 0;
 
+	printbuf_indent_add_nextline(&buf, 2);
+
 	BUG_ON(b->key.k.type == KEY_TYPE_btree_ptr_v2 &&
 	       !bpos_eq(bkey_i_to_btree_ptr_v2(&b->key)->v.min_key,
 			b->data->min_key));
@@ -64,19 +68,20 @@ int bch2_btree_node_check_topology(struct btree_trans *trans, struct btree *b)
 
 	if (b == btree_node_root(c, b)) {
 		if (!bpos_eq(b->data->min_key, POS_MIN)) {
-			printbuf_reset(&buf);
+			ret = __bch2_topology_error(c, &buf);
+
 			bch2_bpos_to_text(&buf, b->data->min_key);
 			log_fsck_err(trans, btree_root_bad_min_key,
 				      "btree root with incorrect min_key: %s", buf.buf);
-			goto topology_repair;
+			goto out;
 		}
 
 		if (!bpos_eq(b->data->max_key, SPOS_MAX)) {
-			printbuf_reset(&buf);
+			ret = __bch2_topology_error(c, &buf);
 			bch2_bpos_to_text(&buf, b->data->max_key);
 			log_fsck_err(trans, btree_root_bad_max_key,
 				      "btree root with incorrect max_key: %s", buf.buf);
-			goto topology_repair;
+			goto out;
 		}
 	}
 
@@ -94,20 +99,19 @@ int bch2_btree_node_check_topology(struct btree_trans *trans, struct btree *b)
 			: bpos_successor(prev.k->k.p);
 
 		if (!bpos_eq(expected_min, bp.v->min_key)) {
-			bch2_topology_error(c);
+			ret = __bch2_topology_error(c, &buf);
 
-			printbuf_reset(&buf);
-			prt_str(&buf, "end of prev node doesn't match start of next node\n  in ");
+			prt_str(&buf, "end of prev node doesn't match start of next node\nin ");
 			bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level);
 			prt_str(&buf, " node ");
 			bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key));
-			prt_str(&buf, "\n  prev ");
+			prt_str(&buf, "\nprev ");
 			bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(prev.k));
-			prt_str(&buf, "\n  next ");
+			prt_str(&buf, "\nnext ");
 			bch2_bkey_val_to_text(&buf, c, k);
 
 			log_fsck_err(trans, btree_node_topology_bad_min_key, "%s", buf.buf);
-			goto topology_repair;
+			goto out;
 		}
 
 		bch2_bkey_buf_reassemble(&prev, c, k);
@@ -115,29 +119,25 @@ int bch2_btree_node_check_topology(struct btree_trans *trans, struct btree *b)
 	}
 
 	if (bkey_deleted(&prev.k->k)) {
-		bch2_topology_error(c);
+		ret = __bch2_topology_error(c, &buf);
 
-		printbuf_reset(&buf);
-		prt_str(&buf, "empty interior node\n  in ");
+		prt_str(&buf, "empty interior node\nin ");
 		bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level);
 		prt_str(&buf, " node ");
 		bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key));
 
 		log_fsck_err(trans, btree_node_topology_empty_interior_node, "%s", buf.buf);
-		goto topology_repair;
 	} else if (!bpos_eq(prev.k->k.p, b->key.k.p)) {
-		bch2_topology_error(c);
+		ret = __bch2_topology_error(c, &buf);
 
-		printbuf_reset(&buf);
-		prt_str(&buf, "last child node doesn't end at end of parent node\n  in ");
+		prt_str(&buf, "last child node doesn't end at end of parent node\nin ");
 		bch2_btree_id_level_to_text(&buf, b->c.btree_id, b->c.level);
 		prt_str(&buf, " node ");
 		bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(&b->key));
-		prt_str(&buf, "\n  last key ");
+		prt_str(&buf, "\nlast key ");
 		bch2_bkey_val_to_text(&buf, c, bkey_i_to_s_c(prev.k));
 
 		log_fsck_err(trans, btree_node_topology_bad_max_key, "%s", buf.buf);
-		goto topology_repair;
 	}
 out:
 fsck_err:
@@ -145,9 +145,6 @@ fsck_err:
 	bch2_bkey_buf_exit(&prev, c);
 	printbuf_exit(&buf);
 	return ret;
-topology_repair:
-	ret = bch2_topology_error(c);
-	goto out;
 }
 
 /* Calculate ideal packed bkey format for new btree nodes: */
@@ -649,6 +646,14 @@ static int btree_update_nodes_written_trans(struct btree_trans *trans,
 	return 0;
 }
 
+/* If the node has been reused, we might be reading uninitialized memory - that's fine: */
+static noinline __no_kmsan_checks bool btree_node_seq_matches(struct btree *b, __le64 seq)
+{
+	struct btree_node *b_data = READ_ONCE(b->data);
+
+	return (b_data ? b_data->keys.seq : 0) == seq;
+}
+
 static void btree_update_nodes_written(struct btree_update *as)
 {
 	struct bch_fs *c = as->c;
@@ -677,17 +682,9 @@ static void btree_update_nodes_written(struct btree_update *as)
 	 * on disk:
 	 */
 	for (i = 0; i < as->nr_old_nodes; i++) {
-		__le64 seq;
-
 		b = as->old_nodes[i];
 
-		bch2_trans_begin(trans);
-		btree_node_lock_nopath_nofail(trans, &b->c, SIX_LOCK_read);
-		seq = b->data ? b->data->keys.seq : 0;
-		six_unlock_read(&b->c.lock);
-		bch2_trans_unlock_long(trans);
-
-		if (seq == as->old_nodes_seq[i])
+		if (btree_node_seq_matches(b, as->old_nodes_seq[i]))
 			wait_on_bit_io(&b->flags, BTREE_NODE_write_in_flight_inner,
 				       TASK_UNINTERRUPTIBLE);
 	}
@@ -1271,7 +1268,8 @@ err:
 	bch2_btree_update_free(as, trans);
 	if (!bch2_err_matches(ret, ENOSPC) &&
 	    !bch2_err_matches(ret, EROFS) &&
-	    ret != -BCH_ERR_journal_reclaim_would_deadlock)
+	    ret != -BCH_ERR_journal_reclaim_would_deadlock &&
+	    ret != -BCH_ERR_journal_shutdown)
 		bch_err_fn_ratelimited(c, ret);
 	return ERR_PTR(ret);
 }
@@ -1782,11 +1780,24 @@ static int bch2_btree_insert_node(struct btree_update *as, struct btree_trans *t
 	int ret;
 
 	lockdep_assert_held(&c->gc_lock);
-	BUG_ON(!btree_node_intent_locked(path, b->c.level));
 	BUG_ON(!b->c.level);
 	BUG_ON(!as || as->b);
 	bch2_verify_keylist_sorted(keys);
 
+	if (!btree_node_intent_locked(path, b->c.level)) {
+		struct printbuf buf = PRINTBUF;
+		bch2_log_msg_start(c, &buf);
+		prt_printf(&buf, "%s(): node not locked at level %u\n",
+			   __func__, b->c.level);
+		bch2_btree_update_to_text(&buf, as);
+		bch2_btree_path_to_text(&buf, trans, path_idx);
+
+		bch2_print_string_as_lines(KERN_ERR, buf.buf);
+		printbuf_exit(&buf);
+		bch2_fs_emergency_read_only(c);
+		return -EIO;
+	}
+
 	ret = bch2_btree_node_lock_write(trans, path, &b->c);
 	if (ret)
 		return ret;
@@ -2007,18 +2018,22 @@ int __bch2_foreground_maybe_merge(struct btree_trans *trans,
 	}
 
 	if (!bpos_eq(bpos_successor(prev->data->max_key), next->data->min_key)) {
-		struct printbuf buf1 = PRINTBUF, buf2 = PRINTBUF;
+		struct printbuf buf = PRINTBUF;
 
-		bch2_bpos_to_text(&buf1, prev->data->max_key);
-		bch2_bpos_to_text(&buf2, next->data->min_key);
-		bch_err(c,
-			"%s(): btree topology error:\n"
-			"  prev ends at   %s\n"
-			"  next starts at %s",
-			__func__, buf1.buf, buf2.buf);
-		printbuf_exit(&buf1);
-		printbuf_exit(&buf2);
-		ret = bch2_topology_error(c);
+		printbuf_indent_add_nextline(&buf, 2);
+		prt_printf(&buf, "%s(): ", __func__);
+		ret = __bch2_topology_error(c, &buf);
+		prt_newline(&buf);
+
+		prt_printf(&buf, "prev ends at   ");
+		bch2_bpos_to_text(&buf, prev->data->max_key);
+		prt_newline(&buf);
+
+		prt_printf(&buf, "next starts at ");
+		bch2_bpos_to_text(&buf, next->data->min_key);
+
+		bch_err(c, "%s", buf.buf);
+		printbuf_exit(&buf);
 		goto err;
 	}
 
@@ -2288,7 +2303,9 @@ static void async_btree_node_rewrite_work(struct work_struct *work)
 
 	int ret = bch2_trans_do(c, bch2_btree_node_rewrite_key(trans,
 						a->btree_id, a->level, a->key.k, 0));
-	if (ret != -ENOENT)
+	if (ret != -ENOENT &&
+	    !bch2_err_matches(ret, EROFS) &&
+	    ret != -BCH_ERR_journal_shutdown)
 		bch_err_fn_ratelimited(c, ret);
 
 	spin_lock(&c->btree_node_rewrites_lock);

libbcachefs/buckets.c

@@ -381,6 +381,36 @@ err:
 	return ret;
 }
 
+static int bucket_ref_update_err(struct btree_trans *trans, struct printbuf *buf,
+				 struct bkey_s_c k, bool insert, enum bch_sb_error_id id)
+{
+	struct bch_fs *c = trans->c;
+	bool repeat = false, print = true, suppress = false;
+
+	prt_printf(buf, "\nwhile marking ");
+	bch2_bkey_val_to_text(buf, c, k);
+	prt_newline(buf);
+
+	__bch2_count_fsck_err(c, id, buf->buf, &repeat, &print, &suppress);
+
+	int ret = bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_allocations);
+
+	if (insert) {
+		print = true;
+		suppress = false;
+
+		bch2_trans_updates_to_text(buf, trans);
+		__bch2_inconsistent_error(c, buf);
+		ret = -BCH_ERR_bucket_ref_update;
+	}
+
+	if (suppress)
+		prt_printf(buf, "Ratelimiting new instances of previous error\n");
+	if (print)
+		bch2_print_string_as_lines(KERN_ERR, buf->buf);
+	return ret;
+}
+
 int bch2_bucket_ref_update(struct btree_trans *trans, struct bch_dev *ca,
 			   struct bkey_s_c k,
 			   const struct bch_extent_ptr *ptr,
@@ -396,32 +426,29 @@ int bch2_bucket_ref_update(struct btree_trans *trans, struct bch_dev *ca,
 
 	BUG_ON(!sectors);
 
-	if (gen_after(ptr->gen, b_gen)) {
-		bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_allocations);
-		log_fsck_err(trans, ptr_gen_newer_than_bucket_gen,
-			"bucket %u:%zu gen %u data type %s: ptr gen %u newer than bucket gen\n"
-			"while marking %s",
+	if (unlikely(gen_after(ptr->gen, b_gen))) {
+		bch2_log_msg_start(c, &buf);
+		prt_printf(&buf,
+			"bucket %u:%zu gen %u data type %s: ptr gen %u newer than bucket gen",
 			ptr->dev, bucket_nr, b_gen,
 			bch2_data_type_str(bucket_data_type ?: ptr_data_type),
-			ptr->gen,
-			(bch2_bkey_val_to_text(&buf, c, k), buf.buf));
-		if (inserting)
-			goto err;
+			ptr->gen);
+
+		ret = bucket_ref_update_err(trans, &buf, k, inserting,
+					    BCH_FSCK_ERR_ptr_gen_newer_than_bucket_gen);
 		goto out;
 	}
 
-	if (gen_cmp(b_gen, ptr->gen) > BUCKET_GC_GEN_MAX) {
-		bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_allocations);
-		log_fsck_err(trans, ptr_too_stale,
-			"bucket %u:%zu gen %u data type %s: ptr gen %u too stale\n"
-			"while marking %s",
+	if (unlikely(gen_cmp(b_gen, ptr->gen) > BUCKET_GC_GEN_MAX)) {
+		bch2_log_msg_start(c, &buf);
+		prt_printf(&buf,
+			"bucket %u:%zu gen %u data type %s: ptr gen %u too stale",
 			ptr->dev, bucket_nr, b_gen,
 			bch2_data_type_str(bucket_data_type ?: ptr_data_type),
-			ptr->gen,
-			(printbuf_reset(&buf),
-			 bch2_bkey_val_to_text(&buf, c, k), buf.buf));
-		if (inserting)
-			goto err;
+			ptr->gen);
+
+		ret = bucket_ref_update_err(trans, &buf, k, inserting,
+					    BCH_FSCK_ERR_ptr_too_stale);
 		goto out;
 	}
 
@@ -430,62 +457,50 @@ int bch2_bucket_ref_update(struct btree_trans *trans, struct bch_dev *ca,
 		goto out;
 	}
 
-	if (b_gen != ptr->gen) {
-		bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_allocations);
-		log_fsck_err(trans, stale_dirty_ptr,
-			"bucket %u:%zu gen %u (mem gen %u) data type %s: stale dirty ptr (gen %u)\n"
-			"while marking %s",
+	if (unlikely(b_gen != ptr->gen)) {
+		bch2_log_msg_start(c, &buf);
+		prt_printf(&buf,
+			"bucket %u:%zu gen %u (mem gen %u) data type %s: stale dirty ptr (gen %u)",
 			ptr->dev, bucket_nr, b_gen,
 			bucket_gen_get(ca, bucket_nr),
 			bch2_data_type_str(bucket_data_type ?: ptr_data_type),
-			ptr->gen,
-			(printbuf_reset(&buf),
-			 bch2_bkey_val_to_text(&buf, c, k), buf.buf));
-		if (inserting)
-			goto err;
+			ptr->gen);
+
+		ret = bucket_ref_update_err(trans, &buf, k, inserting,
+					    BCH_FSCK_ERR_stale_dirty_ptr);
 		goto out;
 	}
 
-	if (bucket_data_type_mismatch(bucket_data_type, ptr_data_type)) {
-		bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_allocations);
-		log_fsck_err(trans, ptr_bucket_data_type_mismatch,
-			"bucket %u:%zu gen %u different types of data in same bucket: %s, %s\n"
-			"while marking %s",
-			ptr->dev, bucket_nr, b_gen,
-			bch2_data_type_str(bucket_data_type),
-			bch2_data_type_str(ptr_data_type),
-			(printbuf_reset(&buf),
-			 bch2_bkey_val_to_text(&buf, c, k), buf.buf));
-		if (inserting)
-			goto err;
+	if (unlikely(bucket_data_type_mismatch(bucket_data_type, ptr_data_type))) {
+		bch2_log_msg_start(c, &buf);
+		prt_printf(&buf, "bucket %u:%zu gen %u different types of data in same bucket: %s, %s",
+			   ptr->dev, bucket_nr, b_gen,
+			   bch2_data_type_str(bucket_data_type),
+			   bch2_data_type_str(ptr_data_type));
+
+		ret = bucket_ref_update_err(trans, &buf, k, inserting,
+					    BCH_FSCK_ERR_ptr_bucket_data_type_mismatch);
 		goto out;
 	}
 
-	if ((u64) *bucket_sectors + sectors > U32_MAX) {
-		bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_allocations);
-		log_fsck_err(trans, bucket_sector_count_overflow,
-			"bucket %u:%zu gen %u data type %s sector count overflow: %u + %lli > U32_MAX\n"
-			"while marking %s",
+	if (unlikely((u64) *bucket_sectors + sectors > U32_MAX)) {
+		bch2_log_msg_start(c, &buf);
+		prt_printf(&buf,
+			"bucket %u:%zu gen %u data type %s sector count overflow: %u + %lli > U32_MAX",
 			ptr->dev, bucket_nr, b_gen,
 			bch2_data_type_str(bucket_data_type ?: ptr_data_type),
-			*bucket_sectors, sectors,
-			(printbuf_reset(&buf),
-			 bch2_bkey_val_to_text(&buf, c, k), buf.buf));
-		if (inserting)
-			goto err;
+			*bucket_sectors, sectors);
+
+		ret = bucket_ref_update_err(trans, &buf, k, inserting,
+					    BCH_FSCK_ERR_bucket_sector_count_overflow);
 		sectors = -*bucket_sectors;
+		goto out;
 	}
 
 	*bucket_sectors += sectors;
 out:
 	printbuf_exit(&buf);
 	return ret;
-err:
-fsck_err:
-	bch2_dump_trans_updates(trans);
-	bch2_inconsistent_error(c);
-	ret = -BCH_ERR_bucket_ref_update;
-	goto out;
 }
 
 void bch2_trans_account_disk_usage_change(struct btree_trans *trans)
@@ -651,9 +666,9 @@ static int bch2_trigger_stripe_ptr(struct btree_trans *trans,
 			stripe_blockcount_get(&s->v, p.ec.block) +
 			sectors);
 
-		struct disk_accounting_pos acc = {
-			.type = BCH_DISK_ACCOUNTING_replicas,
-		};
+		struct disk_accounting_pos acc;
+		memset(&acc, 0, sizeof(acc));
+		acc.type = BCH_DISK_ACCOUNTING_replicas;
 		bch2_bkey_to_replicas(&acc.replicas, bkey_i_to_s_c(&s->k_i));
 		acc.replicas.data_type = data_type;
 		ret = bch2_disk_accounting_mod(trans, &acc, &sectors, 1, false);
@@ -677,19 +692,21 @@ err:
 		if (!m || !m->alive) {
 			gc_stripe_unlock(m);
 			struct printbuf buf = PRINTBUF;
+			bch2_log_msg_start(c, &buf);
+			prt_printf(&buf, "pointer to nonexistent stripe %llu\n  while marking ",
+				   (u64) p.ec.idx);
 			bch2_bkey_val_to_text(&buf, c, k);
-			bch_err_ratelimited(c, "pointer to nonexistent stripe %llu\n  while marking %s",
-					    (u64) p.ec.idx, buf.buf);
+			__bch2_inconsistent_error(c, &buf);
+			bch2_print_string_as_lines(KERN_ERR, buf.buf);
 			printbuf_exit(&buf);
-			bch2_inconsistent_error(c);
 			return -BCH_ERR_trigger_stripe_pointer;
 		}
 
 		m->block_sectors[p.ec.block] += sectors;
 
-		struct disk_accounting_pos acc = {
-			.type = BCH_DISK_ACCOUNTING_replicas,
-		};
+		struct disk_accounting_pos acc;
+		memset(&acc, 0, sizeof(acc));
+		acc.type = BCH_DISK_ACCOUNTING_replicas;
 		memcpy(&acc.replicas, &m->r.e, replicas_entry_bytes(&m->r.e));
 		gc_stripe_unlock(m);
 
@@ -717,16 +734,14 @@ static int __trigger_extent(struct btree_trans *trans,
 		: BCH_DATA_user;
 	int ret = 0;
 
-	struct disk_accounting_pos acc_replicas_key = {
-		.type			= BCH_DISK_ACCOUNTING_replicas,
-		.replicas.data_type	= data_type,
-		.replicas.nr_devs	= 0,
-		.replicas.nr_required	= 1,
-	};
+	struct disk_accounting_pos acc_replicas_key;
+	memset(&acc_replicas_key, 0, sizeof(acc_replicas_key));
+	acc_replicas_key.type = BCH_DISK_ACCOUNTING_replicas;
+	acc_replicas_key.replicas.data_type	= data_type;
+	acc_replicas_key.replicas.nr_devs	= 0;
+	acc_replicas_key.replicas.nr_required	= 1;
 
-	struct disk_accounting_pos acct_compression_key = {
-		.type			= BCH_DISK_ACCOUNTING_compression,
-	};
+	unsigned cur_compression_type = 0;
 	u64 compression_acct[3] = { 1, 0, 0 };
 
 	bkey_for_each_ptr_decode(k.k, ptrs, p, entry) {
@@ -760,13 +775,13 @@ static int __trigger_extent(struct btree_trans *trans,
 			acc_replicas_key.replicas.nr_required = 0;
 		}
 
-		if (acct_compression_key.compression.type &&
-		    acct_compression_key.compression.type != p.crc.compression_type) {
+		if (cur_compression_type &&
+		    cur_compression_type != p.crc.compression_type) {
 			if (flags & BTREE_TRIGGER_overwrite)
 				bch2_u64s_neg(compression_acct, ARRAY_SIZE(compression_acct));
 
-			ret = bch2_disk_accounting_mod(trans, &acct_compression_key, compression_acct,
-						       ARRAY_SIZE(compression_acct), gc);
+			ret = bch2_disk_accounting_mod2(trans, gc, compression_acct,
+							compression, cur_compression_type);
 			if (ret)
 				return ret;
 
@@ -775,7 +790,7 @@ static int __trigger_extent(struct btree_trans *trans,
 			compression_acct[2] = 0;
 		}
 
-		acct_compression_key.compression.type = p.crc.compression_type;
+		cur_compression_type = p.crc.compression_type;
 		if (p.crc.compression_type) {
 			compression_acct[1] += p.crc.uncompressed_size;
 			compression_acct[2] += p.crc.compressed_size;
@@ -789,45 +804,34 @@ static int __trigger_extent(struct btree_trans *trans,
 	}
 
 	if (acc_replicas_key.replicas.nr_devs && !level && k.k->p.snapshot) {
-		struct disk_accounting_pos acc_snapshot_key = {
-			.type			= BCH_DISK_ACCOUNTING_snapshot,
-			.snapshot.id		= k.k->p.snapshot,
-		};
-		ret = bch2_disk_accounting_mod(trans, &acc_snapshot_key, replicas_sectors, 1, gc);
+		ret = bch2_disk_accounting_mod2_nr(trans, gc, replicas_sectors, 1, snapshot, k.k->p.snapshot);
 		if (ret)
 			return ret;
 	}
 
-	if (acct_compression_key.compression.type) {
+	if (cur_compression_type) {
 		if (flags & BTREE_TRIGGER_overwrite)
 			bch2_u64s_neg(compression_acct, ARRAY_SIZE(compression_acct));
 
-		ret = bch2_disk_accounting_mod(trans, &acct_compression_key, compression_acct,
-					       ARRAY_SIZE(compression_acct), gc);
+		ret = bch2_disk_accounting_mod2(trans, gc, compression_acct,
+						compression, cur_compression_type);
 		if (ret)
 			return ret;
 	}
 
 	if (level) {
-		struct disk_accounting_pos acc_btree_key = {
-			.type		= BCH_DISK_ACCOUNTING_btree,
-			.btree.id	= btree_id,
-		};
-		ret = bch2_disk_accounting_mod(trans, &acc_btree_key, replicas_sectors, 1, gc);
+		ret = bch2_disk_accounting_mod2_nr(trans, gc, replicas_sectors, 1, btree, btree_id);
 		if (ret)
 			return ret;
 	} else {
 		bool insert = !(flags & BTREE_TRIGGER_overwrite);
-		struct disk_accounting_pos acc_inum_key = {
-			.type		= BCH_DISK_ACCOUNTING_inum,
-			.inum.inum	= k.k->p.inode,
-		};
+
 		s64 v[3] = {
 			insert ? 1 : -1,
 			insert ? k.k->size : -((s64) k.k->size),
 			*replicas_sectors,
 		};
-		ret = bch2_disk_accounting_mod(trans, &acc_inum_key, v, ARRAY_SIZE(v), gc);
+		ret = bch2_disk_accounting_mod2(trans, gc, v, inum, k.k->p.inode);
 		if (ret)
 			return ret;
 	}
@@ -876,15 +880,15 @@ int bch2_trigger_extent(struct btree_trans *trans,
 		}
 
 		int need_rebalance_delta = 0;
-		s64 need_rebalance_sectors_delta = 0;
+		s64 need_rebalance_sectors_delta[1] = { 0 };
 
 		s64 s = bch2_bkey_sectors_need_rebalance(c, old);
 		need_rebalance_delta -= s != 0;
-		need_rebalance_sectors_delta -= s;
+		need_rebalance_sectors_delta[0] -= s;
 
 		s = bch2_bkey_sectors_need_rebalance(c, new.s_c);
 		need_rebalance_delta += s != 0;
-		need_rebalance_sectors_delta += s;
+		need_rebalance_sectors_delta[0] += s;
 
 		if ((flags & BTREE_TRIGGER_transactional) && need_rebalance_delta) {
 			int ret = bch2_btree_bit_mod_buffered(trans, BTREE_ID_rebalance_work,
@@ -893,12 +897,9 @@ int bch2_trigger_extent(struct btree_trans *trans,
 				return ret;
 		}
 
-		if (need_rebalance_sectors_delta) {
-			struct disk_accounting_pos acc = {
-				.type		= BCH_DISK_ACCOUNTING_rebalance_work,
-			};
-			int ret = bch2_disk_accounting_mod(trans, &acc, &need_rebalance_sectors_delta, 1,
-							   flags & BTREE_TRIGGER_gc);
+		if (need_rebalance_sectors_delta[0]) {
+			int ret = bch2_disk_accounting_mod2(trans, flags & BTREE_TRIGGER_gc,
+							    need_rebalance_sectors_delta, rebalance_work);
 			if (ret)
 				return ret;
 		}
@@ -914,17 +915,13 @@ static int __trigger_reservation(struct btree_trans *trans,
 			enum btree_iter_update_trigger_flags flags)
 {
 	if (flags & (BTREE_TRIGGER_transactional|BTREE_TRIGGER_gc)) {
-		s64 sectors = k.k->size;
+		s64 sectors[1] = { k.k->size };
 
 		if (flags & BTREE_TRIGGER_overwrite)
-			sectors = -sectors;
+			sectors[0] = -sectors[0];
 
-		struct disk_accounting_pos acc = {
-			.type = BCH_DISK_ACCOUNTING_persistent_reserved,
-			.persistent_reserved.nr_replicas = bkey_s_c_to_reservation(k).v->nr_replicas,
-		};
-
-		return bch2_disk_accounting_mod(trans, &acc, &sectors, 1, flags & BTREE_TRIGGER_gc);
+		return bch2_disk_accounting_mod2(trans, flags & BTREE_TRIGGER_gc, sectors,
+				persistent_reserved, bkey_s_c_to_reservation(k).v->nr_replicas);
 	}
 
 	return 0;

libbcachefs/chardev.c

@@ -426,10 +426,8 @@ static long bch2_ioctl_fs_usage(struct bch_fs *c,
 	arg.replica_entries_bytes = replicas.nr;
 
 	for (unsigned i = 0; i < BCH_REPLICAS_MAX; i++) {
-		struct disk_accounting_pos k = {
-			.type = BCH_DISK_ACCOUNTING_persistent_reserved,
-			.persistent_reserved.nr_replicas = i,
-		};
+		struct disk_accounting_pos k;
+		disk_accounting_key_init(k, persistent_reserved, .nr_replicas = i);
 
 		bch2_accounting_mem_read(c,
 					 disk_accounting_pos_to_bpos(&k),

libbcachefs/checksum.c

@@ -466,7 +466,7 @@ int bch2_rechecksum_bio(struct bch_fs *c, struct bio *bio,
 		prt_str(&buf, ")");
 		WARN_RATELIMIT(1, "%s", buf.buf);
 		printbuf_exit(&buf);
-		return -EIO;
+		return -BCH_ERR_recompute_checksum;
 	}
 
 	for (i = splits; i < splits + ARRAY_SIZE(splits); i++) {

libbcachefs/compress.c

@@ -177,7 +177,7 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src,
 	size_t src_len = src->bi_iter.bi_size;
 	size_t dst_len = crc.uncompressed_size << 9;
 	void *workspace;
-	int ret;
+	int ret = 0, ret2;
 
 	enum bch_compression_opts opt = bch2_compression_type_to_opt(crc.compression_type);
 	mempool_t *workspace_pool = &c->compress_workspace[opt];
@@ -189,7 +189,7 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src,
 		else
 			ret = -BCH_ERR_compression_workspace_not_initialized;
 		if (ret)
-			goto out;
+			goto err;
 	}
 
 	src_data = bio_map_or_bounce(c, src, READ);
@@ -197,10 +197,10 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src,
 	switch (crc.compression_type) {
 	case BCH_COMPRESSION_TYPE_lz4_old:
 	case BCH_COMPRESSION_TYPE_lz4:
-		ret = LZ4_decompress_safe_partial(src_data.b, dst_data,
-						  src_len, dst_len, dst_len);
-		if (ret != dst_len)
-			goto err;
+		ret2 = LZ4_decompress_safe_partial(src_data.b, dst_data,
+						   src_len, dst_len, dst_len);
+		if (ret2 != dst_len)
+			ret = -BCH_ERR_decompress_lz4;
 		break;
 	case BCH_COMPRESSION_TYPE_gzip: {
 		z_stream strm = {
@@ -214,45 +214,43 @@ static int __bio_uncompress(struct bch_fs *c, struct bio *src,
 
 		zlib_set_workspace(&strm, workspace);
 		zlib_inflateInit2(&strm, -MAX_WBITS);
-		ret = zlib_inflate(&strm, Z_FINISH);
+		ret2 = zlib_inflate(&strm, Z_FINISH);
 
 		mempool_free(workspace, workspace_pool);
 
-		if (ret != Z_STREAM_END)
-			goto err;
+		if (ret2 != Z_STREAM_END)
+			ret = -BCH_ERR_decompress_gzip;
 		break;
 	}
 	case BCH_COMPRESSION_TYPE_zstd: {
 		ZSTD_DCtx *ctx;
 		size_t real_src_len = le32_to_cpup(src_data.b);
 
-		if (real_src_len > src_len - 4)
+		if (real_src_len > src_len - 4) {
+			ret = -BCH_ERR_decompress_zstd_src_len_bad;
 			goto err;
+		}
 
 		workspace = mempool_alloc(workspace_pool, GFP_NOFS);
 		ctx = zstd_init_dctx(workspace, zstd_dctx_workspace_bound());
 
-		ret = zstd_decompress_dctx(ctx,
+		ret2 = zstd_decompress_dctx(ctx,
 				dst_data,	dst_len,
 				src_data.b + 4, real_src_len);
 
 		mempool_free(workspace, workspace_pool);
 
-		if (ret != dst_len)
-			goto err;
+		if (ret2 != dst_len)
+			ret = -BCH_ERR_decompress_zstd;
 		break;
 	}
 	default:
 		BUG();
 	}
-	ret = 0;
+err:
 fsck_err:
-out:
 	bio_unmap_or_unbounce(c, src_data);
 	return ret;
-err:
-	ret = -EIO;
-	goto out;
 }
 
 int bch2_bio_uncompress_inplace(struct bch_write_op *op,
@@ -268,27 +266,22 @@ int bch2_bio_uncompress_inplace(struct bch_write_op *op,
 	BUG_ON(!bio->bi_vcnt);
 	BUG_ON(DIV_ROUND_UP(crc->live_size, PAGE_SECTORS) > bio->bi_max_vecs);
 
-	if (crc->uncompressed_size << 9	> c->opts.encoded_extent_max ||
-	    crc->compressed_size << 9	> c->opts.encoded_extent_max) {
-		struct printbuf buf = PRINTBUF;
-		bch2_write_op_error(&buf, op, op->pos.offset,
-				    "extent too big to decompress");
-		bch_err_ratelimited(c, "%s", buf.buf);
-		printbuf_exit(&buf);
-		return -EIO;
+	if (crc->uncompressed_size << 9	> c->opts.encoded_extent_max) {
+		bch2_write_op_error(op, op->pos.offset,
+				    "extent too big to decompress (%u > %u)",
+				    crc->uncompressed_size << 9, c->opts.encoded_extent_max);
+		return -BCH_ERR_decompress_exceeded_max_encoded_extent;
 	}
 
 	data = __bounce_alloc(c, dst_len, WRITE);
 
-	if (__bio_uncompress(c, bio, data.b, *crc)) {
-		if (!c->opts.no_data_io) {
-			struct printbuf buf = PRINTBUF;
-			bch2_write_op_error(&buf, op, op->pos.offset,
-					    "decompression error");
-			bch_err_ratelimited(c, "%s", buf.buf);
-			printbuf_exit(&buf);
-		}
-		ret = -EIO;
+	ret = __bio_uncompress(c, bio, data.b, *crc);
+
+	if (c->opts.no_data_io)
+		ret = 0;
+
+	if (ret) {
+		bch2_write_op_error(op, op->pos.offset, "%s", bch2_err_str(ret));
 		goto err;
 	}
 
@@ -321,7 +314,7 @@ int bch2_bio_uncompress(struct bch_fs *c, struct bio *src,
 
 	if (crc.uncompressed_size << 9	> c->opts.encoded_extent_max ||
 	    crc.compressed_size << 9	> c->opts.encoded_extent_max)
-		return -EIO;
+		return -BCH_ERR_decompress_exceeded_max_encoded_extent;
 
 	dst_data = dst_len == dst_iter.bi_size
 		? __bio_map_or_bounce(c, dst, dst_iter, WRITE)

libbcachefs/data_update.c

@@ -22,6 +22,13 @@
 
 #include <linux/ioprio.h>
 
+static const char * const bch2_data_update_type_strs[] = {
+#define x(t, n, ...) [n] = #t,
+	BCH_DATA_UPDATE_TYPES()
+#undef x
+	NULL
+};
+
 static void bkey_put_dev_refs(struct bch_fs *c, struct bkey_s_c k)
 {
 	struct bkey_ptrs_c ptrs = bch2_bkey_ptrs_c(k);
@@ -181,6 +188,7 @@ static int __bch2_data_update_index_update(struct btree_trans *trans,
 		container_of(op, struct data_update, op);
 	struct keylist *keys = &op->insert_keys;
 	struct bkey_buf _new, _insert;
+	struct printbuf journal_msg = PRINTBUF;
 	int ret = 0;
 
 	bch2_bkey_buf_init(&_new);
@@ -354,7 +362,7 @@ restart_drop_extra_replicas:
 			printbuf_exit(&buf);
 
 			bch2_fatal_error(c);
-			ret = -EIO;
+			ret = -BCH_ERR_invalid_bkey;
 			goto out;
 		}
 
@@ -372,7 +380,12 @@ restart_drop_extra_replicas:
 			printbuf_exit(&buf);
 		}
 
-		ret =   bch2_insert_snapshot_whiteouts(trans, m->btree_id,
+		printbuf_reset(&journal_msg);
+		prt_str(&journal_msg, bch2_data_update_type_strs[m->type]);
+
+		ret =   bch2_trans_log_msg(trans, &journal_msg) ?:
+			bch2_trans_log_bkey(trans, m->btree_id, 0, m->k.k) ?:
+			bch2_insert_snapshot_whiteouts(trans, m->btree_id,
 						k.k->p, bkey_start_pos(&insert->k)) ?:
 			bch2_insert_snapshot_whiteouts(trans, m->btree_id,
 						k.k->p, insert->k.p) ?:
@@ -417,6 +430,7 @@ nowork:
 		goto next;
 	}
 out:
+	printbuf_exit(&journal_msg);
 	bch2_trans_iter_exit(trans, &iter);
 	bch2_bkey_buf_exit(&_insert, c);
 	bch2_bkey_buf_exit(&_new, c);
@@ -577,6 +591,9 @@ void bch2_data_update_opts_to_text(struct printbuf *out, struct bch_fs *c,
 
 void bch2_data_update_to_text(struct printbuf *out, struct data_update *m)
 {
+	prt_str(out, bch2_data_update_type_strs[m->type]);
+	prt_newline(out);
+
 	bch2_data_update_opts_to_text(out, m->op.c, &m->op.opts, &m->data_opts);
 	prt_newline(out);
 
@@ -738,6 +755,9 @@ int bch2_data_update_init(struct btree_trans *trans,
 
 	bch2_bkey_buf_init(&m->k);
 	bch2_bkey_buf_reassemble(&m->k, c, k);
+	m->type		= data_opts.btree_insert_flags & BCH_WATERMARK_copygc
+		? BCH_DATA_UPDATE_copygc
+		: BCH_DATA_UPDATE_rebalance;
 	m->btree_id	= btree_id;
 	m->data_opts	= data_opts;
 	m->ctxt		= ctxt;

libbcachefs/data_update.h

@@ -24,7 +24,19 @@ struct data_update_opts {
 void bch2_data_update_opts_to_text(struct printbuf *, struct bch_fs *,
 				   struct bch_io_opts *, struct data_update_opts *);
 
+#define BCH_DATA_UPDATE_TYPES()		\
+	x(copygc,	0)		\
+	x(rebalance,	1)		\
+	x(promote,	2)
+
+enum bch_data_update_types {
+#define x(n, id)	BCH_DATA_UPDATE_##n = id,
+	BCH_DATA_UPDATE_TYPES()
+#undef x
+};
+
 struct data_update {
+	enum bch_data_update_types type;
 	/* extent being updated: */
 	bool			read_done;
 	enum btree_id		btree_id;

libbcachefs/dirent.c

@@ -729,3 +729,54 @@ int bch2_readdir(struct bch_fs *c, subvol_inum inum, struct dir_context *ctx)
 
 	return ret < 0 ? ret : 0;
 }
+
+/* fsck */
+
+static int lookup_first_inode(struct btree_trans *trans, u64 inode_nr,
+			      struct bch_inode_unpacked *inode)
+{
+	struct btree_iter iter;
+	struct bkey_s_c k;
+	int ret;
+
+	for_each_btree_key_norestart(trans, iter, BTREE_ID_inodes, POS(0, inode_nr),
+				     BTREE_ITER_all_snapshots, k, ret) {
+		if (k.k->p.offset != inode_nr)
+			break;
+		if (!bkey_is_inode(k.k))
+			continue;
+		ret = bch2_inode_unpack(k, inode);
+		goto found;
+	}
+	ret = -BCH_ERR_ENOENT_inode;
+found:
+	bch_err_msg(trans->c, ret, "fetching inode %llu", inode_nr);
+	bch2_trans_iter_exit(trans, &iter);
+	return ret;
+}
+
+int bch2_fsck_remove_dirent(struct btree_trans *trans, struct bpos pos)
+{
+	struct bch_fs *c = trans->c;
+	struct btree_iter iter;
+	struct bch_inode_unpacked dir_inode;
+	struct bch_hash_info dir_hash_info;
+	int ret;
+
+	ret = lookup_first_inode(trans, pos.inode, &dir_inode);
+	if (ret)
+		goto err;
+
+	dir_hash_info = bch2_hash_info_init(c, &dir_inode);
+
+	bch2_trans_iter_init(trans, &iter, BTREE_ID_dirents, pos, BTREE_ITER_intent);
+
+	ret =   bch2_btree_iter_traverse(&iter) ?:
+		bch2_hash_delete_at(trans, bch2_dirent_hash_desc,
+				    &dir_hash_info, &iter,
+				    BTREE_UPDATE_internal_snapshot_node);
+	bch2_trans_iter_exit(trans, &iter);
+err:
+	bch_err_fn(c, ret);
+	return ret;
+}

libbcachefs/dirent.h

@@ -82,4 +82,6 @@ int bch2_empty_dir_snapshot(struct btree_trans *, u64, u32, u32);
 int bch2_empty_dir_trans(struct btree_trans *, subvol_inum);
 int bch2_readdir(struct bch_fs *, subvol_inum, struct dir_context *);
 
+int bch2_fsck_remove_dirent(struct btree_trans *, struct bpos);
+
 #endif /* _BCACHEFS_DIRENT_H */

libbcachefs/disk_accounting.c

@@ -114,10 +114,9 @@ int bch2_mod_dev_cached_sectors(struct btree_trans *trans,
 				unsigned dev, s64 sectors,
 				bool gc)
 {
-	struct disk_accounting_pos acc = {
-		.type = BCH_DISK_ACCOUNTING_replicas,
-	};
-
+	struct disk_accounting_pos acc;
+	memset(&acc, 0, sizeof(acc));
+	acc.type = BCH_DISK_ACCOUNTING_replicas;
 	bch2_replicas_entry_cached(&acc.replicas, dev);
 
 	return bch2_disk_accounting_mod(trans, &acc, &sectors, 1, gc);
@@ -135,6 +134,12 @@ static inline bool is_zero(char *start, char *end)
 
 #define field_end(p, member)	(((void *) (&p.member)) + sizeof(p.member))
 
+static const unsigned bch2_accounting_type_nr_counters[] = {
+#define x(f, id, nr)	[BCH_DISK_ACCOUNTING_##f]	= nr,
+	BCH_DISK_ACCOUNTING_TYPES()
+#undef x
+};
+
 int bch2_accounting_validate(struct bch_fs *c, struct bkey_s_c k,
 			     struct bkey_validate_context from)
 {
@@ -193,6 +198,11 @@ int bch2_accounting_validate(struct bch_fs *c, struct bkey_s_c k,
 	bkey_fsck_err_on(!is_zero(end, (void *) (&acc_k + 1)),
 			 c, accounting_key_junk_at_end,
 			 "junk at end of accounting key");
+
+	bkey_fsck_err_on(bch2_accounting_counters(k.k) != bch2_accounting_type_nr_counters[acc_k.type],
+			 c, accounting_key_nr_counters_wrong,
+			 "accounting key with %u counters, should be %u",
+			 bch2_accounting_counters(k.k), bch2_accounting_type_nr_counters[acc_k.type]);
 fsck_err:
 	return ret;
 }
@@ -635,7 +645,7 @@ static int bch2_disk_accounting_validate_late(struct btree_trans *trans,
 
 		if (fsck_err_on(!bch2_replicas_marked_locked(c, &r.e),
 				trans, accounting_replicas_not_marked,
-				"accounting not marked in superblock replicas\n  %s",
+				"accounting not marked in superblock replicas\n%s",
 				(printbuf_reset(&buf),
 				 bch2_accounting_key_to_text(&buf, &acc),
 				 buf.buf))) {
@@ -665,7 +675,7 @@ fsck_err:
 	return ret;
 invalid_device:
 	if (fsck_err(trans, accounting_to_invalid_device,
-		     "accounting entry points to invalid device %i\n  %s",
+		     "accounting entry points to invalid device %i\n%s",
 		     invalid_dev,
 		     (printbuf_reset(&buf),
 		      bch2_accounting_key_to_text(&buf, &acc),
@@ -726,7 +736,9 @@ int bch2_accounting_read(struct bch_fs *c)
 				break;
 
 			if (!bch2_accounting_is_mem(acc_k)) {
-				struct disk_accounting_pos next = { .type = acc_k.type + 1 };
+				struct disk_accounting_pos next;
+				memset(&next, 0, sizeof(next));
+				next.type = acc_k.type + 1;
 				bch2_btree_iter_set_pos(&iter, disk_accounting_pos_to_bpos(&next));
 				continue;
 			}
@@ -882,15 +894,13 @@ int bch2_dev_usage_remove(struct bch_fs *c, unsigned dev)
 int bch2_dev_usage_init(struct bch_dev *ca, bool gc)
 {
 	struct bch_fs *c = ca->fs;
-	struct disk_accounting_pos acc = {
-		.type = BCH_DISK_ACCOUNTING_dev_data_type,
-		.dev_data_type.dev = ca->dev_idx,
-		.dev_data_type.data_type = BCH_DATA_free,
-	};
 	u64 v[3] = { ca->mi.nbuckets - ca->mi.first_bucket, 0, 0 };
 
 	int ret = bch2_trans_do(c, ({
-		bch2_disk_accounting_mod(trans, &acc, v, ARRAY_SIZE(v), gc) ?:
+		bch2_disk_accounting_mod2(trans, gc,
+					  v, dev_data_type,
+					  .dev = ca->dev_idx,
+					  .data_type = BCH_DATA_free) ?:
 		(!gc ? bch2_trans_commit(trans, NULL, NULL, 0) : 0);
 	}));
 	bch_err_fn(c, ret);
@@ -917,7 +927,9 @@ void bch2_verify_accounting_clean(struct bch_fs *c)
 				break;
 
 			if (!bch2_accounting_is_mem(acc_k)) {
-				struct disk_accounting_pos next = { .type = acc_k.type + 1 };
+				struct disk_accounting_pos next;
+				memset(&next, 0, sizeof(next));
+				next.type = acc_k.type + 1;
 				bch2_btree_iter_set_pos(&iter, disk_accounting_pos_to_bpos(&next));
 				continue;
 			}

libbcachefs/disk_accounting.h

@@ -33,10 +33,12 @@ static inline bool bch2_accounting_key_is_zero(struct bkey_s_c_accounting a)
 static inline void bch2_accounting_accumulate(struct bkey_i_accounting *dst,
 					      struct bkey_s_c_accounting src)
 {
-	EBUG_ON(dst->k.u64s != src.k->u64s);
-
-	for (unsigned i = 0; i < bch2_accounting_counters(&dst->k); i++)
+	for (unsigned i = 0;
+	     i < min(bch2_accounting_counters(&dst->k),
+		     bch2_accounting_counters(src.k));
+	     i++)
 		dst->v.d[i] += src.v->d[i];
+
 	if (bversion_cmp(dst->k.bversion, src.k->bversion) < 0)
 		dst->k.bversion = src.k->bversion;
 }
@@ -85,6 +87,24 @@ static inline struct bpos disk_accounting_pos_to_bpos(struct disk_accounting_pos
 
 int bch2_disk_accounting_mod(struct btree_trans *, struct disk_accounting_pos *,
 			     s64 *, unsigned, bool);
+
+#define disk_accounting_key_init(_k, _type, ...)			\
+do {									\
+	memset(&(_k), 0, sizeof(_k));					\
+	(_k).type	= BCH_DISK_ACCOUNTING_##_type;			\
+	(_k)._type	= (struct bch_acct_##_type) { __VA_ARGS__ };	\
+} while (0)
+
+#define bch2_disk_accounting_mod2_nr(_trans, _gc, _v, _nr, ...)		\
+({									\
+	struct disk_accounting_pos pos;					\
+	disk_accounting_key_init(pos, __VA_ARGS__);			\
+	bch2_disk_accounting_mod(trans, &pos, _v, _nr, _gc);		\
+})
+
+#define bch2_disk_accounting_mod2(_trans, _gc, _v, ...)			\
+	bch2_disk_accounting_mod2_nr(_trans, _gc, _v, ARRAY_SIZE(_v), __VA_ARGS__)
+
 int bch2_mod_dev_cached_sectors(struct btree_trans *, unsigned, s64, bool);
 
 int bch2_accounting_validate(struct bch_fs *, struct bkey_s_c,

libbcachefs/disk_accounting_format.h

@@ -95,40 +95,81 @@ static inline bool data_type_is_hidden(enum bch_data_type type)
 	}
 }
 
+/*
+ * field 1: name
+ * field 2: id
+ * field 3: number of counters (max 3)
+ */
+
 #define BCH_DISK_ACCOUNTING_TYPES()		\
-	x(nr_inodes,		0)		\
-	x(persistent_reserved,	1)		\
-	x(replicas,		2)		\
-	x(dev_data_type,	3)		\
-	x(compression,		4)		\
-	x(snapshot,		5)		\
-	x(btree,		6)		\
-	x(rebalance_work,	7)		\
-	x(inum,			8)
+	x(nr_inodes,		0,	1)	\
+	x(persistent_reserved,	1,	1)	\
+	x(replicas,		2,	1)	\
+	x(dev_data_type,	3,	3)	\
+	x(compression,		4,	3)	\
+	x(snapshot,		5,	1)	\
+	x(btree,		6,	1)	\
+	x(rebalance_work,	7,	1)	\
+	x(inum,			8,	3)
 
 enum disk_accounting_type {
-#define x(f, nr)	BCH_DISK_ACCOUNTING_##f	= nr,
+#define x(f, nr, ...)	BCH_DISK_ACCOUNTING_##f	= nr,
 	BCH_DISK_ACCOUNTING_TYPES()
 #undef x
 	BCH_DISK_ACCOUNTING_TYPE_NR,
 };
 
-struct bch_nr_inodes {
+/*
+ * No subtypes - number of inodes in the entire filesystem
+ *
+ * XXX: perhaps we could add a per-subvolume counter?
+ */
+struct bch_acct_nr_inodes {
 };
 
-struct bch_persistent_reserved {
+/*
+ * Tracks KEY_TYPE_reservation sectors, broken out by number of replicas for the
+ * reservation:
+ */
+struct bch_acct_persistent_reserved {
 	__u8			nr_replicas;
 };
 
-struct bch_dev_data_type {
+/*
+ * device, data type counter fields:
+ * [
+ *   nr_buckets
+ *   live sectors (in buckets of that data type)
+ *   sectors of internal fragmentation
+ * ]
+ *
+ * XXX: live sectors should've been done differently, you can have multiple data
+ * types in the same bucket (user, stripe, cached) and this collapses them to
+ * the bucket data type, and makes the internal fragmentation counter redundant
+ */
+struct bch_acct_dev_data_type {
 	__u8			dev;
 	__u8			data_type;
 };
 
+/*
+ * Compression type fields:
+ * [
+ *   number of extents
+ *   uncompressed size
+ *   compressed size
+ * ]
+ *
+ * Compression ratio, average extent size (fragmentation).
+ */
 struct bch_acct_compression {
 	__u8			type;
 };
 
+/*
+ * On disk usage by snapshot id; counts same values as replicas counter, but
+ * aggregated differently
+ */
 struct bch_acct_snapshot {
 	__u32			id;
 } __packed;
@@ -137,10 +178,27 @@ struct bch_acct_btree {
 	__u32			id;
 } __packed;
 
+/*
+ * inum counter fields:
+ * [
+ *   number of extents
+ *   sum of extent sizes - bkey size
+ *     this field is similar to inode.bi_sectors, except here extents in
+ *     different snapshots but the same inode number are all collapsed to the
+ *     same counter
+ *   sum of on disk size - same values tracked by replicas counters
+ * ]
+ *
+ * This tracks on disk fragmentation.
+ */
 struct bch_acct_inum {
 	__u64			inum;
 } __packed;
 
+/*
+ * Simple counter of the amount of data (on disk sectors) rebalance needs to
+ * move, extents counted here are also in the rebalance_work btree.
+ */
 struct bch_acct_rebalance_work {
 };
 
@@ -149,10 +207,10 @@ struct disk_accounting_pos {
 	struct {
 		__u8				type;
 		union {
-		struct bch_nr_inodes		nr_inodes;
-		struct bch_persistent_reserved	persistent_reserved;
+		struct bch_acct_nr_inodes	nr_inodes;
+		struct bch_acct_persistent_reserved	persistent_reserved;
 		struct bch_replicas_entry_v1	replicas;
-		struct bch_dev_data_type	dev_data_type;
+		struct bch_acct_dev_data_type	dev_data_type;
 		struct bch_acct_compression	compression;
 		struct bch_acct_snapshot	snapshot;
 		struct bch_acct_btree		btree;

libbcachefs/ec.c

@@ -320,7 +320,7 @@ static int mark_stripe_bucket(struct btree_trans *trans,
 
 	if (flags & BTREE_TRIGGER_gc) {
 		struct bucket *g = gc_bucket(ca, bucket.offset);
-		if (bch2_fs_inconsistent_on(!g, c, "reference to invalid bucket on device %u\n  %s",
+		if (bch2_fs_inconsistent_on(!g, c, "reference to invalid bucket on device %u\n%s",
 					    ptr->dev,
 					    (bch2_bkey_val_to_text(&buf, c, s.s_c), buf.buf))) {
 			ret = -BCH_ERR_mark_stripe;
@@ -453,9 +453,9 @@ int bch2_trigger_stripe(struct btree_trans *trans,
 		if (new_s) {
 			s64 sectors = (u64) le16_to_cpu(new_s->sectors) * new_s->nr_redundant;
 
-			struct disk_accounting_pos acc = {
-				.type = BCH_DISK_ACCOUNTING_replicas,
-			};
+			struct disk_accounting_pos acc;
+			memset(&acc, 0, sizeof(acc));
+			acc.type = BCH_DISK_ACCOUNTING_replicas;
 			bch2_bkey_to_replicas(&acc.replicas, new);
 			int ret = bch2_disk_accounting_mod(trans, &acc, &sectors, 1, gc);
 			if (ret)
@@ -468,9 +468,9 @@ int bch2_trigger_stripe(struct btree_trans *trans,
 		if (old_s) {
 			s64 sectors = -((s64) le16_to_cpu(old_s->sectors)) * old_s->nr_redundant;
 
-			struct disk_accounting_pos acc = {
-				.type = BCH_DISK_ACCOUNTING_replicas,
-			};
+			struct disk_accounting_pos acc;
+			memset(&acc, 0, sizeof(acc));
+			acc.type = BCH_DISK_ACCOUNTING_replicas;
 			bch2_bkey_to_replicas(&acc.replicas, old);
 			int ret = bch2_disk_accounting_mod(trans, &acc, &sectors, 1, gc);
 			if (ret)
@@ -1124,7 +1124,7 @@ static int ec_stripe_update_extent(struct btree_trans *trans,
 
 		bch2_fs_inconsistent(c, "%s", buf.buf);
 		printbuf_exit(&buf);
-		return -EIO;
+		return -BCH_ERR_erasure_coding_found_btree_node;
 	}
 
 	k = bch2_backpointer_get_key(trans, bp, &iter, BTREE_ITER_intent, last_flushed);
@@ -1190,7 +1190,7 @@ static int ec_stripe_update_bucket(struct btree_trans *trans, struct ec_stripe_b
 
 	struct bch_dev *ca = bch2_dev_tryget(c, ptr.dev);
 	if (!ca)
-		return -EIO;
+		return -BCH_ERR_ENOENT_dev_not_found;
 
 	struct bpos bucket_pos = PTR_BUCKET_POS(ca, &ptr);
 
@@ -1227,21 +1227,19 @@ static int ec_stripe_update_extents(struct bch_fs *c, struct ec_stripe_buf *s)
 {
 	struct btree_trans *trans = bch2_trans_get(c);
 	struct bch_stripe *v = &bkey_i_to_stripe(&s->key)->v;
-	unsigned i, nr_data = v->nr_blocks - v->nr_redundant;
-	int ret = 0;
+	unsigned nr_data = v->nr_blocks - v->nr_redundant;
 
-	ret = bch2_btree_write_buffer_flush_sync(trans);
+	int ret = bch2_btree_write_buffer_flush_sync(trans);
 	if (ret)
 		goto err;
 
-	for (i = 0; i < nr_data; i++) {
+	for (unsigned i = 0; i < nr_data; i++) {
 		ret = ec_stripe_update_bucket(trans, s, i);
 		if (ret)
 			break;
 	}
 err:
 	bch2_trans_put(trans);
-
 	return ret;
 }
 
@@ -1451,11 +1449,11 @@ static void ec_stripe_new_cancel(struct bch_fs *c, struct ec_stripe_head *h, int
 	ec_stripe_new_set_pending(c, h);
 }
 
-void bch2_ec_bucket_cancel(struct bch_fs *c, struct open_bucket *ob)
+void bch2_ec_bucket_cancel(struct bch_fs *c, struct open_bucket *ob, int err)
 {
 	struct ec_stripe_new *s = ob->ec;
 
-	s->err = -EIO;
+	s->err = err;
 }
 
 void *bch2_writepoint_ec_buf(struct bch_fs *c, struct write_point *wp)
@@ -2112,14 +2110,14 @@ static int bch2_invalidate_stripe_to_dev(struct btree_trans *trans, struct bkey_
 	if (ret)
 		return ret;
 
-	struct disk_accounting_pos acc = {
-		.type = BCH_DISK_ACCOUNTING_replicas,
-	};
+	struct disk_accounting_pos acc;
 
 	s64 sectors = 0;
 	for (unsigned i = 0; i < s->v.nr_blocks; i++)
 		sectors -= stripe_blockcount_get(&s->v, i);
 
+	memset(&acc, 0, sizeof(acc));
+	acc.type = BCH_DISK_ACCOUNTING_replicas;
 	bch2_bkey_to_replicas(&acc.replicas, bkey_i_to_s_c(&s->k_i));
 	acc.replicas.data_type = BCH_DATA_user;
 	ret = bch2_disk_accounting_mod(trans, &acc, &sectors, 1, false);
@@ -2133,6 +2131,8 @@ static int bch2_invalidate_stripe_to_dev(struct btree_trans *trans, struct bkey_
 
 	sectors = -sectors;
 
+	memset(&acc, 0, sizeof(acc));
+	acc.type = BCH_DISK_ACCOUNTING_replicas;
 	bch2_bkey_to_replicas(&acc.replicas, bkey_i_to_s_c(&s->k_i));
 	acc.replicas.data_type = BCH_DATA_user;
 	ret = bch2_disk_accounting_mod(trans, &acc, &sectors, 1, false);

libbcachefs/ec.h

@@ -249,7 +249,7 @@ int bch2_ec_read_extent(struct btree_trans *, struct bch_read_bio *, struct bkey
 
 void *bch2_writepoint_ec_buf(struct bch_fs *, struct write_point *);
 
-void bch2_ec_bucket_cancel(struct bch_fs *, struct open_bucket *);
+void bch2_ec_bucket_cancel(struct bch_fs *, struct open_bucket *, int);
 
 int bch2_ec_stripe_new_alloc(struct bch_fs *, struct ec_stripe_head *);
 

libbcachefs/errcode.h

@@ -5,6 +5,8 @@
 #define BCH_ERRCODES()								\
 	x(ERANGE,			ERANGE_option_too_small)		\
 	x(ERANGE,			ERANGE_option_too_big)			\
+	x(EINVAL,			injected)				\
+	x(BCH_ERR_injected,		injected_fs_start)			\
 	x(EINVAL,			mount_option)				\
 	x(BCH_ERR_mount_option,		option_name)				\
 	x(BCH_ERR_mount_option,		option_value)				\
@@ -116,6 +118,7 @@
 	x(ENOENT,			ENOENT_snapshot_tree)			\
 	x(ENOENT,			ENOENT_dirent_doesnt_match_inode)	\
 	x(ENOENT,			ENOENT_dev_not_found)			\
+	x(ENOENT,			ENOENT_dev_bucket_not_found)		\
 	x(ENOENT,			ENOENT_dev_idx_not_found)		\
 	x(ENOENT,			ENOENT_inode_no_backpointer)		\
 	x(ENOENT,			ENOENT_no_snapshot_tree_subvol)		\
@@ -207,6 +210,7 @@
 	x(EINVAL,			no_resize_with_buckets_nouse)		\
 	x(EINVAL,			inode_unpack_error)			\
 	x(EINVAL,			varint_decode_error)			\
+	x(EINVAL,			erasure_coding_found_btree_node)	\
 	x(EOPNOTSUPP,			may_not_use_incompat_feature)		\
 	x(EROFS,			erofs_trans_commit)			\
 	x(EROFS,			erofs_no_writes)			\
@@ -267,6 +271,7 @@
 	x(BCH_ERR_operation_blocked,    nocow_lock_blocked)			\
 	x(EIO,				journal_shutdown)			\
 	x(EIO,				journal_flush_err)			\
+	x(EIO,				journal_write_err)			\
 	x(EIO,				btree_node_read_err)			\
 	x(BCH_ERR_btree_node_read_err,	btree_node_read_err_cached)		\
 	x(EIO,				sb_not_downgraded)			\
@@ -275,6 +280,7 @@
 	x(EIO,				btree_node_read_validate_error)		\
 	x(EIO,				btree_need_topology_repair)		\
 	x(EIO,				bucket_ref_update)			\
+	x(EIO,				trigger_alloc)				\
 	x(EIO,				trigger_pointer)			\
 	x(EIO,				trigger_stripe_pointer)			\
 	x(EIO,				metadata_bucket_inconsistency)		\
@@ -290,8 +296,21 @@
 	x(EIO,				EIO_fault_injected)			\
 	x(EIO,				ec_block_read)				\
 	x(EIO,				ec_block_write)				\
-	x(EIO,				data_read)				\
+	x(EIO,				recompute_checksum)			\
+	x(EIO,				decompress)				\
+	x(BCH_ERR_decompress,		decompress_exceeded_max_encoded_extent)	\
+	x(BCH_ERR_decompress,		decompress_lz4)				\
+	x(BCH_ERR_decompress,		decompress_gzip)			\
+	x(BCH_ERR_decompress,		decompress_zstd_src_len_bad)		\
+	x(BCH_ERR_decompress,		decompress_zstd)			\
+	x(EIO,				data_write)				\
+	x(BCH_ERR_data_write,		data_write_io)				\
+	x(BCH_ERR_data_write,		data_write_csum)			\
+	x(BCH_ERR_data_write,		data_write_invalid_ptr)			\
+	x(BCH_ERR_data_write,		data_write_misaligned)			\
+	x(BCH_ERR_decompress,		data_read)				\
 	x(BCH_ERR_data_read,		no_device_to_read_from)			\
+	x(BCH_ERR_data_read,		no_devices_valid)			\
 	x(BCH_ERR_data_read,		data_read_io_err)			\
 	x(BCH_ERR_data_read,		data_read_csum_err)			\
 	x(BCH_ERR_data_read,		data_read_retry)			\

libbcachefs/error.c

@@ -3,15 +3,24 @@
 #include "btree_cache.h"
 #include "btree_iter.h"
 #include "error.h"
-#include "fs-common.h"
 #include "journal.h"
+#include "namei.h"
 #include "recovery_passes.h"
 #include "super.h"
 #include "thread_with_file.h"
 
 #define FSCK_ERR_RATELIMIT_NR	10
 
-bool bch2_inconsistent_error(struct bch_fs *c)
+void bch2_log_msg_start(struct bch_fs *c, struct printbuf *out)
+{
+	printbuf_indent_add_nextline(out, 2);
+
+#ifdef BCACHEFS_LOG_PREFIX
+	prt_printf(out, bch2_log_msg(c, ""));
+#endif
+}
+
+bool __bch2_inconsistent_error(struct bch_fs *c, struct printbuf *out)
 {
 	set_bit(BCH_FS_error, &c->flags);
 
@@ -21,10 +30,11 @@ bool bch2_inconsistent_error(struct bch_fs *c)
 	case BCH_ON_ERROR_fix_safe:
 	case BCH_ON_ERROR_ro:
 		if (bch2_fs_emergency_read_only(c))
-			bch_err(c, "inconsistency detected - emergency read only at journal seq %llu",
-				journal_cur_seq(&c->journal));
+			prt_printf(out, "inconsistency detected - emergency read only at journal seq %llu\n",
+				   journal_cur_seq(&c->journal));
 		return true;
 	case BCH_ON_ERROR_panic:
+		bch2_print_string_as_lines(KERN_ERR, out->buf);
 		panic(bch2_fmt(c, "panic after error"));
 		return true;
 	default:
@@ -32,11 +42,63 @@ bool bch2_inconsistent_error(struct bch_fs *c)
 	}
 }
 
-int bch2_topology_error(struct bch_fs *c)
+bool bch2_inconsistent_error(struct bch_fs *c)
 {
+	struct printbuf buf = PRINTBUF;
+	printbuf_indent_add_nextline(&buf, 2);
+
+	bool ret = __bch2_inconsistent_error(c, &buf);
+	if (ret)
+		bch_err(c, "%s", buf.buf);
+	printbuf_exit(&buf);
+	return ret;
+}
+
+__printf(3, 0)
+static bool bch2_fs_trans_inconsistent(struct bch_fs *c, struct btree_trans *trans,
+				       const char *fmt, va_list args)
+{
+	struct printbuf buf = PRINTBUF;
+
+	bch2_log_msg_start(c, &buf);
+
+	prt_vprintf(&buf, fmt, args);
+	prt_newline(&buf);
+
+	if (trans)
+		bch2_trans_updates_to_text(&buf, trans);
+	bool ret = __bch2_inconsistent_error(c, &buf);
+	bch2_print_string_as_lines(KERN_ERR, buf.buf);
+
+	printbuf_exit(&buf);
+	return ret;
+}
+
+bool bch2_fs_inconsistent(struct bch_fs *c, const char *fmt, ...)
+{
+	va_list args;
+	va_start(args, fmt);
+	bool ret = bch2_fs_trans_inconsistent(c, NULL, fmt, args);
+	va_end(args);
+	return ret;
+}
+
+bool bch2_trans_inconsistent(struct btree_trans *trans, const char *fmt, ...)
+{
+	va_list args;
+	va_start(args, fmt);
+	bool ret = bch2_fs_trans_inconsistent(trans->c, trans, fmt, args);
+	va_end(args);
+	return ret;
+}
+
+int __bch2_topology_error(struct bch_fs *c, struct printbuf *out)
+{
+	prt_printf(out, "btree topology error: ");
+
 	set_bit(BCH_FS_topology_error, &c->flags);
 	if (!test_bit(BCH_FS_recovery_running, &c->flags)) {
-		bch2_inconsistent_error(c);
+		__bch2_inconsistent_error(c, out);
 		return -BCH_ERR_btree_need_topology_repair;
 	} else {
 		return bch2_run_explicit_recovery_pass(c, BCH_RECOVERY_PASS_check_topology) ?:
@@ -44,6 +106,24 @@ int bch2_topology_error(struct bch_fs *c)
 	}
 }
 
+int bch2_fs_topology_error(struct bch_fs *c, const char *fmt, ...)
+{
+	struct printbuf buf = PRINTBUF;
+
+	bch2_log_msg_start(c, &buf);
+
+	va_list args;
+	va_start(args, fmt);
+	prt_vprintf(&buf, fmt, args);
+	va_end(args);
+
+	int ret = __bch2_topology_error(c, &buf);
+	bch2_print_string_as_lines(KERN_ERR, buf.buf);
+
+	printbuf_exit(&buf);
+	return ret;
+}
+
 void bch2_fatal_error(struct bch_fs *c)
 {
 	if (bch2_fs_emergency_read_only(c))
@@ -184,7 +264,8 @@ static enum ask_yn bch2_fsck_ask_yn(struct bch_fs *c, struct btree_trans *trans)
 
 #endif
 
-static struct fsck_err_state *fsck_err_get(struct bch_fs *c, const char *fmt)
+static struct fsck_err_state *fsck_err_get(struct bch_fs *c,
+					   enum bch_sb_error_id id)
 {
 	struct fsck_err_state *s;
 
@@ -192,7 +273,7 @@ static struct fsck_err_state *fsck_err_get(struct bch_fs *c, const char *fmt)
 		return NULL;
 
 	list_for_each_entry(s, &c->fsck_error_msgs, list)
-		if (s->fmt == fmt) {
+		if (s->id == id) {
 			/*
 			 * move it to the head of the list: repeated fsck errors
 			 * are common
@@ -210,7 +291,7 @@ static struct fsck_err_state *fsck_err_get(struct bch_fs *c, const char *fmt)
 	}
 
 	INIT_LIST_HEAD(&s->list);
-	s->fmt = fmt;
+	s->id = id;
 	list_add(&s->list, &c->fsck_error_msgs);
 	return s;
 }
@@ -260,15 +341,59 @@ static int do_fsck_ask_yn(struct bch_fs *c,
 	return ask;
 }
 
+static struct fsck_err_state *count_fsck_err_locked(struct bch_fs *c,
+			  enum bch_sb_error_id id, const char *msg,
+			  bool *repeat, bool *print, bool *suppress)
+{
+	bch2_sb_error_count(c, id);
+
+	struct fsck_err_state *s = fsck_err_get(c, id);
+	if (s) {
+		/*
+		 * We may be called multiple times for the same error on
+		 * transaction restart - this memoizes instead of asking the user
+		 * multiple times for the same error:
+		 */
+		if (s->last_msg && !strcmp(msg, s->last_msg)) {
+			*repeat = true;
+			*print = false;
+			return s;
+		}
+
+		kfree(s->last_msg);
+		s->last_msg = kstrdup(msg, GFP_KERNEL);
+
+		if (c->opts.ratelimit_errors &&
+		    s->nr >= FSCK_ERR_RATELIMIT_NR) {
+			if (s->nr == FSCK_ERR_RATELIMIT_NR)
+				*suppress = true;
+			else
+				*print = false;
+		}
+
+		s->nr++;
+	}
+	return s;
+}
+
+void __bch2_count_fsck_err(struct bch_fs *c,
+			   enum bch_sb_error_id id, const char *msg,
+			   bool *repeat, bool *print, bool *suppress)
+{
+	bch2_sb_error_count(c, id);
+
+	mutex_lock(&c->fsck_error_msgs_lock);
+	count_fsck_err_locked(c, id, msg, repeat, print, suppress);
+	mutex_unlock(&c->fsck_error_msgs_lock);
+}
+
 int __bch2_fsck_err(struct bch_fs *c,
 		  struct btree_trans *trans,
 		  enum bch_fsck_flags flags,
 		  enum bch_sb_error_id err,
 		  const char *fmt, ...)
 {
-	struct fsck_err_state *s = NULL;
 	va_list args;
-	bool print = true, suppressing = false, inconsistent = false, exiting = false;
 	struct printbuf buf = PRINTBUF, *out = &buf;
 	int ret = -BCH_ERR_fsck_ignore;
 	const char *action_orig = "fix?", *action = action_orig;
@@ -303,7 +428,12 @@ int __bch2_fsck_err(struct bch_fs *c,
 			? -BCH_ERR_fsck_fix
 			: -BCH_ERR_fsck_ignore;
 
-	bch2_sb_error_count(c, err);
+	printbuf_indent_add_nextline(out, 2);
+
+#ifdef BCACHEFS_LOG_PREFIX
+	if (strncmp(fmt, "bcachefs", 8))
+		prt_printf(out, bch2_log_msg(c, ""));
+#endif
 
 	va_start(args, fmt);
 	prt_vprintf(out, fmt, args);
@@ -323,42 +453,15 @@ int __bch2_fsck_err(struct bch_fs *c,
 	}
 
 	mutex_lock(&c->fsck_error_msgs_lock);
-	s = fsck_err_get(c, fmt);
-	if (s) {
-		/*
-		 * We may be called multiple times for the same error on
-		 * transaction restart - this memoizes instead of asking the user
-		 * multiple times for the same error:
-		 */
-		if (s->last_msg && !strcmp(buf.buf, s->last_msg)) {
-			ret = s->ret;
-			goto err_unlock;
-		}
-
-		kfree(s->last_msg);
-		s->last_msg = kstrdup(buf.buf, GFP_KERNEL);
-		if (!s->last_msg) {
-			ret = -ENOMEM;
-			goto err_unlock;
-		}
-
-		if (c->opts.ratelimit_errors &&
-		    !(flags & FSCK_NO_RATELIMIT) &&
-		    s->nr >= FSCK_ERR_RATELIMIT_NR) {
-			if (s->nr == FSCK_ERR_RATELIMIT_NR)
-				suppressing = true;
-			else
-				print = false;
-		}
-
-		s->nr++;
+	bool repeat = false, print = true, suppress = false;
+	bool inconsistent = false, exiting = false;
+	struct fsck_err_state *s =
+		count_fsck_err_locked(c, err, buf.buf, &repeat, &print, &suppress);
+	if (repeat) {
+		ret = s->ret;
+		goto err_unlock;
 	}
 
-#ifdef BCACHEFS_LOG_PREFIX
-	if (!strncmp(fmt, "bcachefs:", 9))
-		prt_printf(out, bch2_log_msg(c, ""));
-#endif
-
 	if ((flags & FSCK_AUTOFIX) &&
 	    (c->opts.errors == BCH_ON_ERROR_continue ||
 	     c->opts.errors == BCH_ON_ERROR_fix_safe)) {
@@ -377,6 +480,7 @@ int __bch2_fsck_err(struct bch_fs *c,
 		    !(flags & (FSCK_CAN_FIX|FSCK_CAN_IGNORE))) {
 			prt_str(out, ", shutting down");
 			inconsistent = true;
+			print = true;
 			ret = -BCH_ERR_fsck_errors_not_fixed;
 		} else if (flags & FSCK_CAN_FIX) {
 			prt_str(out, ", ");
@@ -435,24 +539,30 @@ int __bch2_fsck_err(struct bch_fs *c,
 		print = true;
 	}
 print:
+	prt_newline(out);
+
+	if (inconsistent)
+		__bch2_inconsistent_error(c, out);
+	else if (exiting)
+		prt_printf(out, "Unable to continue, halting\n");
+	else if (suppress)
+		prt_printf(out, "Ratelimiting new instances of previous error\n");
+
 	if (print) {
+		/* possibly strip an empty line, from printbuf_indent_add */
+		while (out->pos && out->buf[out->pos - 1] == ' ')
+			--out->pos;
+		printbuf_nul_terminate(out);
+
 		if (bch2_fs_stdio_redirect(c))
-			bch2_print(c, "%s\n", out->buf);
+			bch2_print(c, "%s", out->buf);
 		else
 			bch2_print_string_as_lines(KERN_ERR, out->buf);
 	}
 
-	if (exiting)
-		bch_err(c, "Unable to continue, halting");
-	else if (suppressing)
-		bch_err(c, "Ratelimiting new instances of previous error");
-
 	if (s)
 		s->ret = ret;
 
-	if (inconsistent)
-		bch2_inconsistent_error(c);
-
 	/*
 	 * We don't yet track whether the filesystem currently has errors, for
 	 * log_fsck_err()s: that would require us to track for every error type
@@ -514,16 +624,14 @@ int __bch2_bkey_fsck_err(struct bch_fs *c,
 	prt_printf(&buf, " level=%u: ", from.level);
 
 	bch2_bkey_val_to_text(&buf, c, k);
-	prt_str(&buf, "\n  ");
+	prt_newline(&buf);
 
 	va_list args;
 	va_start(args, fmt);
 	prt_vprintf(&buf, fmt, args);
 	va_end(args);
 
-	prt_str(&buf, ": delete?");
-
-	int ret = __bch2_fsck_err(c, NULL, fsck_flags, err, "%s", buf.buf);
+	int ret = __bch2_fsck_err(c, NULL, fsck_flags, err, "%s, delete?", buf.buf);
 	printbuf_exit(&buf);
 	return ret;
 }
@@ -536,7 +644,7 @@ void bch2_flush_fsck_errs(struct bch_fs *c)
 
 	list_for_each_entry_safe(s, n, &c->fsck_error_msgs, list) {
 		if (s->ratelimited && s->last_msg)
-			bch_err(c, "Saw %llu errors like:\n    %s", s->nr, s->last_msg);
+			bch_err(c, "Saw %llu errors like:\n  %s", s->nr, s->last_msg);
 
 		list_del(&s->list);
 		kfree(s->last_msg);

libbcachefs/error.h

@@ -18,6 +18,8 @@ struct work_struct;
 
 /* Error messages: */
 
+void bch2_log_msg_start(struct bch_fs *, struct printbuf *);
+
 /*
  * Inconsistency errors: The on disk data is inconsistent. If these occur during
  * initial recovery, they don't indicate a bug in the running code - we walk all
@@ -29,21 +31,10 @@ struct work_struct;
  * BCH_ON_ERROR_CONTINUE mode
  */
 
+bool __bch2_inconsistent_error(struct bch_fs *, struct printbuf *);
 bool bch2_inconsistent_error(struct bch_fs *);
-
-int bch2_topology_error(struct bch_fs *);
-
-#define bch2_fs_topology_error(c, ...)					\
-({									\
-	bch_err(c, "btree topology error: " __VA_ARGS__);		\
-	bch2_topology_error(c);						\
-})
-
-#define bch2_fs_inconsistent(c, ...)					\
-({									\
-	bch_err(c, __VA_ARGS__);					\
-	bch2_inconsistent_error(c);					\
-})
+__printf(2, 3)
+bool bch2_fs_inconsistent(struct bch_fs *, const char *, ...);
 
 #define bch2_fs_inconsistent_on(cond, ...)				\
 ({									\
@@ -53,26 +44,21 @@ int bch2_topology_error(struct bch_fs *);
 	_ret;								\
 })
 
-/*
- * When a transaction update discovers or is causing a fs inconsistency, it's
- * helpful to also dump the pending updates:
- */
-#define bch2_trans_inconsistent(trans, ...)				\
-({									\
-	bch_err(trans->c, __VA_ARGS__);					\
-	bch2_dump_trans_updates(trans);					\
-	bch2_inconsistent_error(trans->c);				\
-})
+__printf(2, 3)
+bool bch2_trans_inconsistent(struct btree_trans *, const char *, ...);
 
-#define bch2_trans_inconsistent_on(cond, trans, ...)			\
+#define bch2_trans_inconsistent_on(cond, ...)				\
 ({									\
 	bool _ret = unlikely(!!(cond));					\
-									\
 	if (_ret)							\
-		bch2_trans_inconsistent(trans, __VA_ARGS__);		\
+		bch2_trans_inconsistent(__VA_ARGS__);			\
 	_ret;								\
 })
 
+int __bch2_topology_error(struct bch_fs *, struct printbuf *);
+__printf(2, 3)
+int bch2_fs_topology_error(struct bch_fs *, const char *, ...);
+
 /*
  * Fsck errors: inconsistency errors we detect at mount time, and should ideally
  * be able to repair:
@@ -80,7 +66,7 @@ int bch2_topology_error(struct bch_fs *);
 
 struct fsck_err_state {
 	struct list_head	list;
-	const char		*fmt;
+	enum bch_sb_error_id	id;
 	u64			nr;
 	bool			ratelimited;
 	int			ret;
@@ -90,6 +76,12 @@ struct fsck_err_state {
 
 #define fsck_err_count(_c, _err)	bch2_sb_err_count(_c, BCH_FSCK_ERR_##_err)
 
+void __bch2_count_fsck_err(struct bch_fs *,
+			   enum bch_sb_error_id, const char *,
+			   bool *, bool *, bool *);
+#define bch2_count_fsck_err(_c, _err, ...)				\
+	__bch2_count_fsck_err(_c, BCH_FSCK_ERR_##_err, __VA_ARGS__)
+
 __printf(5, 6) __cold
 int __bch2_fsck_err(struct bch_fs *, struct btree_trans *,
 		  enum bch_fsck_flags,

libbcachefs/extents.c

@@ -227,8 +227,11 @@ int bch2_bkey_pick_read_device(struct bch_fs *c, struct bkey_s_c k,
 	if (have_io_errors)
 		return -BCH_ERR_data_read_io_err;
 
-	WARN_ONCE(1, "unhandled error case in %s\n", __func__);
-	return -EINVAL;
+	/*
+	 * If we get here, we have pointers (bkey_ptrs_validate() ensures that),
+	 * but they don't point to valid devices:
+	 */
+	return -BCH_ERR_no_devices_valid;
 }
 
 /* KEY_TYPE_btree_ptr: */
@@ -592,29 +595,35 @@ static void bch2_extent_crc_pack(union bch_extent_crc *dst,
 				 struct bch_extent_crc_unpacked src,
 				 enum bch_extent_entry_type type)
 {
-#define set_common_fields(_dst, _src)					\
-		_dst.type		= 1 << type;			\
-		_dst.csum_type		= _src.csum_type,		\
-		_dst.compression_type	= _src.compression_type,	\
-		_dst._compressed_size	= _src.compressed_size - 1,	\
-		_dst._uncompressed_size	= _src.uncompressed_size - 1,	\
-		_dst.offset		= _src.offset
+#define common_fields(_src)						\
+		.type			= BIT(type),			\
+		.csum_type		= _src.csum_type,		\
+		.compression_type	= _src.compression_type,	\
+		._compressed_size	= _src.compressed_size - 1,	\
+		._uncompressed_size	= _src.uncompressed_size - 1,	\
+		.offset			= _src.offset
 
 	switch (type) {
 	case BCH_EXTENT_ENTRY_crc32:
-		set_common_fields(dst->crc32, src);
-		dst->crc32.csum		= (u32 __force) *((__le32 *) &src.csum.lo);
+		dst->crc32		= (struct bch_extent_crc32) {
+			common_fields(src),
+			.csum		= (u32 __force) *((__le32 *) &src.csum.lo),
+		};
 		break;
 	case BCH_EXTENT_ENTRY_crc64:
-		set_common_fields(dst->crc64, src);
-		dst->crc64.nonce	= src.nonce;
-		dst->crc64.csum_lo	= (u64 __force) src.csum.lo;
-		dst->crc64.csum_hi	= (u64 __force) *((__le16 *) &src.csum.hi);
+		dst->crc64		= (struct bch_extent_crc64) {
+			common_fields(src),
+			.nonce		= src.nonce,
+			.csum_lo	= (u64 __force) src.csum.lo,
+			.csum_hi	= (u64 __force) *((__le16 *) &src.csum.hi),
+		};
 		break;
 	case BCH_EXTENT_ENTRY_crc128:
-		set_common_fields(dst->crc128, src);
-		dst->crc128.nonce	= src.nonce;
-		dst->crc128.csum	= src.csum;
+		dst->crc128		= (struct bch_extent_crc128) {
+			common_fields(src),
+			.nonce		= src.nonce,
+			.csum		= src.csum,
+		};
 		break;
 	default:
 		BUG();

libbcachefs/fs-io-buffered.c

@@ -225,11 +225,11 @@ static void bchfs_read(struct btree_trans *trans,
 
 		bch2_read_extent(trans, rbio, iter.pos,
 				 data_btree, k, offset_into_extent, flags);
+		swap(rbio->bio.bi_iter.bi_size, bytes);
 
 		if (flags & BCH_READ_last_fragment)
 			break;
 
-		swap(rbio->bio.bi_iter.bi_size, bytes);
 		bio_advance(&rbio->bio, bytes);
 err:
 		if (ret &&

libbcachefs/fs-io.c

@@ -999,17 +999,28 @@ static loff_t bch2_seek_hole(struct file *file, u64 offset)
 				   POS(inode->v.i_ino, offset >> 9),
 				   POS(inode->v.i_ino, U64_MAX),
 				   inum.subvol, BTREE_ITER_slots, k, ({
-			if (k.k->p.inode != inode->v.i_ino) {
-				next_hole = bch2_seek_pagecache_hole(&inode->v,
-						offset, MAX_LFS_FILESIZE, 0, false);
-				break;
-			} else if (!bkey_extent_is_data(k.k)) {
-				next_hole = bch2_seek_pagecache_hole(&inode->v,
-						max(offset, bkey_start_offset(k.k) << 9),
-						k.k->p.offset << 9, 0, false);
+			if (k.k->p.inode != inode->v.i_ino ||
+			    !bkey_extent_is_data(k.k)) {
+				loff_t start_offset = k.k->p.inode == inode->v.i_ino
+					? max(offset, bkey_start_offset(k.k) << 9)
+					: offset;
+				loff_t end_offset = k.k->p.inode == inode->v.i_ino
+					? MAX_LFS_FILESIZE
+					: k.k->p.offset << 9;
 
-				if (next_hole < k.k->p.offset << 9)
+				/*
+				 * Found a hole in the btree, now make sure it's
+				 * a hole in the pagecache. We might have to
+				 * keep searching if this hole is entirely dirty
+				 * in the page cache:
+				 */
+				bch2_trans_unlock(trans);
+				loff_t pagecache_hole = bch2_seek_pagecache_hole(&inode->v,
+								start_offset, end_offset, 0, false);
+				if (pagecache_hole < end_offset) {
+					next_hole = pagecache_hole;
 					break;
+				}
 			} else {
 				offset = max(offset, bkey_start_offset(k.k) << 9);
 			}

libbcachefs/fs-ioctl.c

@@ -5,8 +5,8 @@
 #include "chardev.h"
 #include "dirent.h"
 #include "fs.h"
-#include "fs-common.h"
 #include "fs-ioctl.h"
+#include "namei.h"
 #include "quota.h"
 
 #include <linux/compat.h>
@@ -541,10 +541,12 @@ static long bch2_ioctl_subvolume_destroy(struct bch_fs *c, struct file *filp,
 		ret = -ENOENT;
 		goto err;
 	}
-	ret = __bch2_unlink(dir, victim, true);
+
+	ret =   inode_permission(file_mnt_idmap(filp), d_inode(victim), MAY_WRITE) ?:
+		__bch2_unlink(dir, victim, true);
 	if (!ret) {
 		fsnotify_rmdir(dir, victim);
-		d_delete(victim);
+		d_invalidate(victim);
 	}
 err:
 	inode_unlock(dir);

libbcachefs/fs.c

@@ -11,7 +11,6 @@
 #include "errcode.h"
 #include "extents.h"
 #include "fs.h"
-#include "fs-common.h"
 #include "fs-io.h"
 #include "fs-ioctl.h"
 #include "fs-io-buffered.h"
@@ -22,6 +21,7 @@
 #include "io_read.h"
 #include "journal.h"
 #include "keylist.h"
+#include "namei.h"
 #include "quota.h"
 #include "rebalance.h"
 #include "snapshot.h"
@@ -641,7 +641,9 @@ static struct bch_inode_info *bch2_lookup_trans(struct btree_trans *trans,
 	if (ret)
 		return ERR_PTR(ret);
 
-	ret = bch2_dirent_read_target(trans, dir, bkey_s_c_to_dirent(k), &inum);
+	struct bkey_s_c_dirent d = bkey_s_c_to_dirent(k);
+
+	ret = bch2_dirent_read_target(trans, dir, d, &inum);
 	if (ret > 0)
 		ret = -ENOENT;
 	if (ret)
@@ -651,30 +653,30 @@ static struct bch_inode_info *bch2_lookup_trans(struct btree_trans *trans,
 	if (inode)
 		goto out;
 
+	/*
+	 * Note: if check/repair needs it, we commit before
+	 * bch2_inode_hash_init_insert(), as after that point we can't take a
+	 * restart - not in the top level loop with a commit_do(), like we
+	 * usually do:
+	 */
+
 	struct bch_subvolume subvol;
 	struct bch_inode_unpacked inode_u;
 	ret =   bch2_subvolume_get(trans, inum.subvol, true, &subvol) ?:
 		bch2_inode_find_by_inum_nowarn_trans(trans, inum, &inode_u) ?:
+		bch2_check_dirent_target(trans, &dirent_iter, d, &inode_u, false) ?:
+		bch2_trans_commit(trans, NULL, NULL, BCH_TRANS_COMMIT_no_enospc) ?:
 		PTR_ERR_OR_ZERO(inode = bch2_inode_hash_init_insert(trans, inum, &inode_u, &subvol));
 
+	/*
+	 * don't remove it: check_inodes might find another inode that points
+	 * back to this dirent
+	 */
 	bch2_fs_inconsistent_on(bch2_err_matches(ret, ENOENT),
-				c, "dirent to missing inode:\n  %s",
-				(bch2_bkey_val_to_text(&buf, c, k), buf.buf));
+				c, "dirent to missing inode:\n%s",
+				(bch2_bkey_val_to_text(&buf, c, d.s_c), buf.buf));
 	if (ret)
 		goto err;
-
-	/* regular files may have hardlinks: */
-	if (bch2_fs_inconsistent_on(bch2_inode_should_have_single_bp(&inode_u) &&
-				    !bkey_eq(k.k->p, POS(inode_u.bi_dir, inode_u.bi_dir_offset)),
-				    c,
-				    "dirent points to inode that does not point back:\n  %s",
-				    (bch2_bkey_val_to_text(&buf, c, k),
-				     prt_printf(&buf, "\n  "),
-				     bch2_inode_unpacked_to_text(&buf, &inode_u),
-				     buf.buf))) {
-		ret = -ENOENT;
-		goto err;
-	}
 out:
 	bch2_trans_iter_exit(trans, &dirent_iter);
 	printbuf_exit(&buf);
@@ -2177,7 +2179,7 @@ static int bch2_fs_get_tree(struct fs_context *fc)
 
 	/* Some options can't be parsed until after the fs is started: */
 	opts = bch2_opts_empty();
-	ret = bch2_parse_mount_opts(c, &opts, NULL, opts_parse->parse_later.buf);
+	ret = bch2_parse_mount_opts(c, &opts, NULL, opts_parse->parse_later.buf, false);
 	if (ret)
 		goto err_stop_fs;
 
@@ -2288,7 +2290,8 @@ err_stop_fs:
 	goto err;
 
 err_put_super:
-	__bch2_fs_stop(c);
+	if (!sb->s_root)
+		__bch2_fs_stop(c);
 	deactivate_locked_super(sb);
 	goto err;
 }
@@ -2331,6 +2334,8 @@ static int bch2_fs_parse_param(struct fs_context *fc,
 	int ret = bch2_parse_one_mount_opt(c, &opts->opts,
 					   &opts->parse_later, param->key,
 					   param->string);
+	if (ret)
+		pr_err("Error parsing option %s: %s", param->key, bch2_err_str(ret));
 
 	return bch2_err_class(ret);
 }

libbcachefs/fsck.c

@@ -10,10 +10,10 @@
 #include "dirent.h"
 #include "error.h"
 #include "fs.h"
-#include "fs-common.h"
 #include "fsck.h"
 #include "inode.h"
 #include "keylist.h"
+#include "namei.h"
 #include "recovery_passes.h"
 #include "snapshot.h"
 #include "super.h"
@@ -23,13 +23,6 @@
 #include <linux/bsearch.h>
 #include <linux/dcache.h> /* struct qstr */
 
-static bool inode_points_to_dirent(struct bch_inode_unpacked *inode,
-				   struct bkey_s_c_dirent d)
-{
-	return  inode->bi_dir		== d.k->p.inode &&
-		inode->bi_dir_offset	== d.k->p.offset;
-}
-
 static int dirent_points_to_inode_nowarn(struct bkey_s_c_dirent d,
 					 struct bch_inode_unpacked *inode)
 {
@@ -116,29 +109,6 @@ static int subvol_lookup(struct btree_trans *trans, u32 subvol,
 	return ret;
 }
 
-static int lookup_first_inode(struct btree_trans *trans, u64 inode_nr,
-			      struct bch_inode_unpacked *inode)
-{
-	struct btree_iter iter;
-	struct bkey_s_c k;
-	int ret;
-
-	for_each_btree_key_norestart(trans, iter, BTREE_ID_inodes, POS(0, inode_nr),
-				     BTREE_ITER_all_snapshots, k, ret) {
-		if (k.k->p.offset != inode_nr)
-			break;
-		if (!bkey_is_inode(k.k))
-			continue;
-		ret = bch2_inode_unpack(k, inode);
-		goto found;
-	}
-	ret = -BCH_ERR_ENOENT_inode;
-found:
-	bch_err_msg(trans->c, ret, "fetching inode %llu", inode_nr);
-	bch2_trans_iter_exit(trans, &iter);
-	return ret;
-}
-
 static int lookup_inode(struct btree_trans *trans, u64 inode_nr, u32 snapshot,
 			struct bch_inode_unpacked *inode)
 {
@@ -179,32 +149,6 @@ static int lookup_dirent_in_snapshot(struct btree_trans *trans,
 	return 0;
 }
 
-static int __remove_dirent(struct btree_trans *trans, struct bpos pos)
-{
-	struct bch_fs *c = trans->c;
-	struct btree_iter iter;
-	struct bch_inode_unpacked dir_inode;
-	struct bch_hash_info dir_hash_info;
-	int ret;
-
-	ret = lookup_first_inode(trans, pos.inode, &dir_inode);
-	if (ret)
-		goto err;
-
-	dir_hash_info = bch2_hash_info_init(c, &dir_inode);
-
-	bch2_trans_iter_init(trans, &iter, BTREE_ID_dirents, pos, BTREE_ITER_intent);
-
-	ret =   bch2_btree_iter_traverse(&iter) ?:
-		bch2_hash_delete_at(trans, bch2_dirent_hash_desc,
-				    &dir_hash_info, &iter,
-				    BTREE_UPDATE_internal_snapshot_node);
-	bch2_trans_iter_exit(trans, &iter);
-err:
-	bch_err_fn(c, ret);
-	return ret;
-}
-
 /*
  * Find any subvolume associated with a tree of snapshots
  * We can't rely on master_subvol - it might have been deleted.
@@ -548,7 +492,7 @@ static int remove_backpointer(struct btree_trans *trans,
 				     SPOS(inode->bi_dir, inode->bi_dir_offset, inode->bi_snapshot));
 	int ret = bkey_err(d) ?:
 		  dirent_points_to_inode(c, d, inode) ?:
-		  __remove_dirent(trans, d.k->p);
+		  bch2_fsck_remove_dirent(trans, d.k->p);
 	bch2_trans_iter_exit(trans, &iter);
 	return ret;
 }
@@ -1477,14 +1421,14 @@ static int check_key_has_inode(struct btree_trans *trans,
 
 	if (fsck_err_on(!i,
 			trans, key_in_missing_inode,
-			"key in missing inode:\n  %s",
+			"key in missing inode:\n%s",
 			(printbuf_reset(&buf),
 			 bch2_bkey_val_to_text(&buf, c, k), buf.buf)))
 		goto delete;
 
 	if (fsck_err_on(i && !btree_matches_i_mode(iter->btree_id, i->inode.bi_mode),
 			trans, key_in_wrong_inode_type,
-			"key for wrong inode mode %o:\n  %s",
+			"key for wrong inode mode %o:\n%s",
 			i->inode.bi_mode,
 			(printbuf_reset(&buf),
 			 bch2_bkey_val_to_text(&buf, c, k), buf.buf)))
@@ -1627,13 +1571,13 @@ static int overlapping_extents_found(struct btree_trans *trans,
 	if (ret)
 		goto err;
 
-	prt_str(&buf, "\n  ");
+	prt_newline(&buf);
 	bch2_bkey_val_to_text(&buf, c, k1);
 
 	if (!bpos_eq(pos1, k1.k->p)) {
-		prt_str(&buf, "\n  wanted\n  ");
+		prt_str(&buf, "\nwanted\n  ");
 		bch2_bpos_to_text(&buf, pos1);
-		prt_str(&buf, "\n  ");
+		prt_str(&buf, "\n");
 		bch2_bkey_to_text(&buf, &pos2);
 
 		bch_err(c, "%s: error finding first overlapping extent when repairing, got%s",
@@ -1656,7 +1600,7 @@ static int overlapping_extents_found(struct btree_trans *trans,
 			break;
 	}
 
-	prt_str(&buf, "\n  ");
+	prt_newline(&buf);
 	bch2_bkey_val_to_text(&buf, c, k2);
 
 	if (bpos_gt(k2.k->p, pos2.p) ||
@@ -1667,7 +1611,7 @@ static int overlapping_extents_found(struct btree_trans *trans,
 		goto err;
 	}
 
-	prt_printf(&buf, "\n  overwriting %s extent",
+	prt_printf(&buf, "\noverwriting %s extent",
 		   pos1.snapshot >= pos2.p.snapshot ? "first" : "second");
 
 	if (fsck_err(trans, extent_overlapping,
@@ -1688,6 +1632,8 @@ static int overlapping_extents_found(struct btree_trans *trans,
 			bch2_trans_commit(trans, &res, NULL, BCH_TRANS_COMMIT_no_enospc);
 		bch2_disk_reservation_put(c, &res);
 
+		bch_info(c, "repair ret %s", bch2_err_str(ret));
+
 		if (ret)
 			goto err;
 
@@ -1840,7 +1786,7 @@ static int check_extent(struct btree_trans *trans, struct btree_iter *iter,
 			if (fsck_err_on(k.k->p.offset > round_up(i->inode.bi_size, block_bytes(c)) >> 9 &&
 					!bkey_extent_is_reservation(k),
 					trans, extent_past_end_of_inode,
-					"extent type past end of inode %llu:%u, i_size %llu\n  %s",
+					"extent type past end of inode %llu:%u, i_size %llu\n%s",
 					i->inode.bi_inum, i->snapshot, i->inode.bi_size,
 					(bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
 				struct btree_iter iter2;
@@ -1985,169 +1931,6 @@ static int check_subdir_dirents_count(struct btree_trans *trans, struct inode_wa
 		trans_was_restarted(trans, restart_count);
 }
 
-noinline_for_stack
-static int check_dirent_inode_dirent(struct btree_trans *trans,
-				   struct btree_iter *iter,
-				   struct bkey_s_c_dirent d,
-				   struct bch_inode_unpacked *target)
-{
-	struct bch_fs *c = trans->c;
-	struct printbuf buf = PRINTBUF;
-	struct btree_iter bp_iter = { NULL };
-	int ret = 0;
-
-	if (inode_points_to_dirent(target, d))
-		return 0;
-
-	if (!target->bi_dir &&
-	    !target->bi_dir_offset) {
-		fsck_err_on(S_ISDIR(target->bi_mode),
-			    trans, inode_dir_missing_backpointer,
-			    "directory with missing backpointer\n%s",
-			    (printbuf_reset(&buf),
-			     bch2_bkey_val_to_text(&buf, c, d.s_c),
-			     prt_printf(&buf, "\n"),
-			     bch2_inode_unpacked_to_text(&buf, target),
-			     buf.buf));
-
-		fsck_err_on(target->bi_flags & BCH_INODE_unlinked,
-			    trans, inode_unlinked_but_has_dirent,
-			    "inode unlinked but has dirent\n%s",
-			    (printbuf_reset(&buf),
-			     bch2_bkey_val_to_text(&buf, c, d.s_c),
-			     prt_printf(&buf, "\n"),
-			     bch2_inode_unpacked_to_text(&buf, target),
-			     buf.buf));
-
-		target->bi_flags &= ~BCH_INODE_unlinked;
-		target->bi_dir		= d.k->p.inode;
-		target->bi_dir_offset	= d.k->p.offset;
-		return __bch2_fsck_write_inode(trans, target);
-	}
-
-	if (bch2_inode_should_have_single_bp(target) &&
-	    !fsck_err(trans, inode_wrong_backpointer,
-		      "dirent points to inode that does not point back:\n  %s",
-		      (bch2_bkey_val_to_text(&buf, c, d.s_c),
-		       prt_printf(&buf, "\n  "),
-		       bch2_inode_unpacked_to_text(&buf, target),
-		       buf.buf)))
-		goto err;
-
-	struct bkey_s_c_dirent bp_dirent = dirent_get_by_pos(trans, &bp_iter,
-			      SPOS(target->bi_dir, target->bi_dir_offset, target->bi_snapshot));
-	ret = bkey_err(bp_dirent);
-	if (ret && !bch2_err_matches(ret, ENOENT))
-		goto err;
-
-	bool backpointer_exists = !ret;
-	ret = 0;
-
-	if (fsck_err_on(!backpointer_exists,
-			trans, inode_wrong_backpointer,
-			"inode %llu:%u has wrong backpointer:\n"
-			"got       %llu:%llu\n"
-			"should be %llu:%llu",
-			target->bi_inum, target->bi_snapshot,
-			target->bi_dir,
-			target->bi_dir_offset,
-			d.k->p.inode,
-			d.k->p.offset)) {
-		target->bi_dir		= d.k->p.inode;
-		target->bi_dir_offset	= d.k->p.offset;
-		ret = __bch2_fsck_write_inode(trans, target);
-		goto out;
-	}
-
-	bch2_bkey_val_to_text(&buf, c, d.s_c);
-	prt_newline(&buf);
-	if (backpointer_exists)
-		bch2_bkey_val_to_text(&buf, c, bp_dirent.s_c);
-
-	if (fsck_err_on(backpointer_exists &&
-			(S_ISDIR(target->bi_mode) ||
-			 target->bi_subvol),
-			trans, inode_dir_multiple_links,
-			"%s %llu:%u with multiple links\n%s",
-			S_ISDIR(target->bi_mode) ? "directory" : "subvolume",
-			target->bi_inum, target->bi_snapshot, buf.buf)) {
-		ret = __remove_dirent(trans, d.k->p);
-		goto out;
-	}
-
-	/*
-	 * hardlinked file with nlink 0:
-	 * We're just adjusting nlink here so check_nlinks() will pick
-	 * it up, it ignores inodes with nlink 0
-	 */
-	if (fsck_err_on(backpointer_exists && !target->bi_nlink,
-			trans, inode_multiple_links_but_nlink_0,
-			"inode %llu:%u type %s has multiple links but i_nlink 0\n%s",
-			target->bi_inum, target->bi_snapshot, bch2_d_types[d.v->d_type], buf.buf)) {
-		target->bi_nlink++;
-		target->bi_flags &= ~BCH_INODE_unlinked;
-		ret = __bch2_fsck_write_inode(trans, target);
-		if (ret)
-			goto err;
-	}
-out:
-err:
-fsck_err:
-	bch2_trans_iter_exit(trans, &bp_iter);
-	printbuf_exit(&buf);
-	bch_err_fn(c, ret);
-	return ret;
-}
-
-noinline_for_stack
-static int check_dirent_target(struct btree_trans *trans,
-			       struct btree_iter *iter,
-			       struct bkey_s_c_dirent d,
-			       struct bch_inode_unpacked *target)
-{
-	struct bch_fs *c = trans->c;
-	struct bkey_i_dirent *n;
-	struct printbuf buf = PRINTBUF;
-	int ret = 0;
-
-	ret = check_dirent_inode_dirent(trans, iter, d, target);
-	if (ret)
-		goto err;
-
-	if (fsck_err_on(d.v->d_type != inode_d_type(target),
-			trans, dirent_d_type_wrong,
-			"incorrect d_type: got %s, should be %s:\n%s",
-			bch2_d_type_str(d.v->d_type),
-			bch2_d_type_str(inode_d_type(target)),
-			(printbuf_reset(&buf),
-			 bch2_bkey_val_to_text(&buf, c, d.s_c), buf.buf))) {
-		n = bch2_trans_kmalloc(trans, bkey_bytes(d.k));
-		ret = PTR_ERR_OR_ZERO(n);
-		if (ret)
-			goto err;
-
-		bkey_reassemble(&n->k_i, d.s_c);
-		n->v.d_type = inode_d_type(target);
-		if (n->v.d_type == DT_SUBVOL) {
-			n->v.d_parent_subvol = cpu_to_le32(target->bi_parent_subvol);
-			n->v.d_child_subvol = cpu_to_le32(target->bi_subvol);
-		} else {
-			n->v.d_inum = cpu_to_le64(target->bi_inum);
-		}
-
-		ret = bch2_trans_update(trans, iter, &n->k_i, 0);
-		if (ret)
-			goto err;
-
-		d = dirent_i_to_s_c(n);
-	}
-err:
-fsck_err:
-	printbuf_exit(&buf);
-	bch_err_fn(c, ret);
-	return ret;
-}
-
 /* find a subvolume that's a descendent of @snapshot: */
 static int find_snapshot_subvol(struct btree_trans *trans, u32 snapshot, u32 *subvolid)
 {
@@ -2247,7 +2030,7 @@ static int check_dirent_to_subvol(struct btree_trans *trans, struct btree_iter *
 		if (fsck_err(trans, dirent_to_missing_subvol,
 			     "dirent points to missing subvolume\n%s",
 			     (bch2_bkey_val_to_text(&buf, c, d.s_c), buf.buf)))
-			return __remove_dirent(trans, d.k->p);
+			return bch2_fsck_remove_dirent(trans, d.k->p);
 		ret = 0;
 		goto out;
 	}
@@ -2291,7 +2074,7 @@ static int check_dirent_to_subvol(struct btree_trans *trans, struct btree_iter *
 			goto err;
 	}
 
-	ret = check_dirent_target(trans, iter, d, &subvol_root);
+	ret = bch2_check_dirent_target(trans, iter, d, &subvol_root, true);
 	if (ret)
 		goto err;
 out:
@@ -2378,13 +2161,13 @@ static int check_dirent(struct btree_trans *trans, struct btree_iter *iter,
 				(printbuf_reset(&buf),
 				 bch2_bkey_val_to_text(&buf, c, k),
 				 buf.buf))) {
-			ret = __remove_dirent(trans, d.k->p);
+			ret = bch2_fsck_remove_dirent(trans, d.k->p);
 			if (ret)
 				goto err;
 		}
 
 		darray_for_each(target->inodes, i) {
-			ret = check_dirent_target(trans, iter, d, &i->inode);
+			ret = bch2_check_dirent_target(trans, iter, d, &i->inode, true);
 			if (ret)
 				goto err;
 		}
@@ -3240,7 +3023,7 @@ long bch2_ioctl_fsck_offline(struct bch_ioctl_fsck_offline __user *user_arg)
 	if (arg.opts) {
 		char *optstr = strndup_user((char __user *)(unsigned long) arg.opts, 1 << 16);
 		ret =   PTR_ERR_OR_ZERO(optstr) ?:
-			bch2_parse_mount_opts(NULL, &thr->opts, NULL, optstr);
+			bch2_parse_mount_opts(NULL, &thr->opts, NULL, optstr, false);
 		if (!IS_ERR(optstr))
 			kfree(optstr);
 
@@ -3348,7 +3131,7 @@ long bch2_ioctl_fsck_online(struct bch_fs *c, struct bch_ioctl_fsck_online arg)
 		char *optstr = strndup_user((char __user *)(unsigned long) arg.opts, 1 << 16);
 
 		ret =   PTR_ERR_OR_ZERO(optstr) ?:
-			bch2_parse_mount_opts(c, &thr->opts, NULL, optstr);
+			bch2_parse_mount_opts(c, &thr->opts, NULL, optstr, false);
 		if (!IS_ERR(optstr))
 			kfree(optstr);
 

libbcachefs/inode.c

@@ -731,10 +731,9 @@ int bch2_trigger_inode(struct btree_trans *trans,
 		bkey_s_to_inode_v3(new).v->bi_journal_seq = cpu_to_le64(trans->journal_res.seq);
 	}
 
-	s64 nr = bkey_is_inode(new.k) - bkey_is_inode(old.k);
-	if ((flags & (BTREE_TRIGGER_transactional|BTREE_TRIGGER_gc)) && nr) {
-		struct disk_accounting_pos acc = { .type = BCH_DISK_ACCOUNTING_nr_inodes };
-		int ret = bch2_disk_accounting_mod(trans, &acc, &nr, 1, flags & BTREE_TRIGGER_gc);
+	s64 nr[1] = { bkey_is_inode(new.k) - bkey_is_inode(old.k) };
+	if ((flags & (BTREE_TRIGGER_transactional|BTREE_TRIGGER_gc)) && nr[0]) {
+		int ret = bch2_disk_accounting_mod2(trans, flags & BTREE_TRIGGER_gc, nr, nr_inodes);
 		if (ret)
 			return ret;
 	}
@@ -1079,7 +1078,7 @@ retry:
 		bch2_fs_inconsistent(c,
 				     "inode %llu:%u not found when deleting",
 				     inum.inum, snapshot);
-		ret = -EIO;
+		ret = -BCH_ERR_ENOENT_inode;
 		goto err;
 	}
 
@@ -1243,7 +1242,7 @@ retry:
 		bch2_fs_inconsistent(c,
 				     "inode %llu:%u not found when deleting",
 				     inum, snapshot);
-		ret = -EIO;
+		ret = -BCH_ERR_ENOENT_inode;
 		goto err;
 	}
 

libbcachefs/inode.h

@@ -277,6 +277,7 @@ static inline bool bch2_inode_should_have_single_bp(struct bch_inode_unpacked *i
 	bool inode_has_bp = inode->bi_dir || inode->bi_dir_offset;
 
 	return S_ISDIR(inode->bi_mode) ||
+		inode->bi_subvol ||
 		(!inode->bi_nlink && inode_has_bp);
 }
 

libbcachefs/io_read.c

@@ -259,6 +259,7 @@ static struct bch_read_bio *__promote_alloc(struct btree_trans *trans,
 			&orig->opts,
 			update_opts,
 			btree_id, k);
+	op->write.type = BCH_DATA_UPDATE_promote;
 	/*
 	 * possible errors: -BCH_ERR_nocow_lock_blocked,
 	 * -BCH_ERR_ENOSPC_disk_reservation:
@@ -1132,10 +1133,6 @@ retry_pick:
 	rbio->bio.bi_iter.bi_sector = pick.ptr.offset;
 	rbio->bio.bi_end_io	= bch2_read_endio;
 
-	/* XXX: also nvme read recovery level */
-	if (unlikely(failed && bch2_dev_io_failures(failed, pick.ptr.dev)))
-		rbio->bio.bi_opf |= REQ_FUA;
-
 	if (rbio->bounce)
 		trace_and_count(c, io_read_bounce, &rbio->bio);
 
@@ -1326,13 +1323,14 @@ int __bch2_read(struct btree_trans *trans, struct bch_read_bio *rbio,
 		ret = __bch2_read_extent(trans, rbio, bvec_iter, iter.pos,
 					 data_btree, k,
 					 offset_into_extent, failed, flags, -1);
+		swap(bvec_iter.bi_size, bytes);
+
 		if (ret)
 			goto err;
 
 		if (flags & BCH_READ_last_fragment)
 			break;
 
-		swap(bvec_iter.bi_size, bytes);
 		bio_advance_iter(&rbio->bio, &bvec_iter, bytes);
 err:
 		if (ret == -BCH_ERR_data_read_retry_csum_err_maybe_userspace)

libbcachefs/io_read.h

@@ -137,8 +137,10 @@ static inline void bch2_read_extent(struct btree_trans *trans,
 			enum btree_id data_btree, struct bkey_s_c k,
 			unsigned offset_into_extent, unsigned flags)
 {
-	__bch2_read_extent(trans, rbio, rbio->bio.bi_iter, read_pos,
-			   data_btree, k, offset_into_extent, NULL, flags, -1);
+	int ret = __bch2_read_extent(trans, rbio, rbio->bio.bi_iter, read_pos,
+				     data_btree, k, offset_into_extent, NULL, flags, -1);
+	/* __bch2_read_extent only returns errors if BCH_READ_in_retry is set */
+	WARN(ret, "unhandled error from __bch2_read_extent()");
 }
 
 int __bch2_read(struct btree_trans *, struct bch_read_bio *, struct bvec_iter,

libbcachefs/io_write.c

@@ -402,61 +402,36 @@ static int bch2_write_index_default(struct bch_write_op *op)
 
 /* Writes */
 
-void bch2_write_op_error_trans(struct btree_trans *trans, struct printbuf *out,
-			       struct bch_write_op *op, u64 offset, const char *fmt, ...)
+void bch2_write_op_error(struct bch_write_op *op, u64 offset, const char *fmt, ...)
 {
-	if (op->subvol)
-		lockrestart_do(trans,
-			bch2_inum_offset_err_msg_trans(trans, out,
-						       (subvol_inum) { op->subvol, op->pos.inode, },
-						       offset << 9));
-	else {
-		struct bpos pos = op->pos;
-		pos.offset = offset;
-		lockrestart_do(trans, bch2_inum_snap_offset_err_msg_trans(trans, out, pos));
-	}
+	struct printbuf buf = PRINTBUF;
 
-	prt_str(out, "write error: ");
-
-	va_list args;
-	va_start(args, fmt);
-	prt_vprintf(out, fmt, args);
-	va_end(args);
-
-	if (op->flags & BCH_WRITE_move) {
-		struct data_update *u = container_of(op, struct data_update, op);
-
-		prt_printf(out, "\n  from internal move ");
-		bch2_bkey_val_to_text(out, op->c, bkey_i_to_s_c(u->k.k));
-	}
-}
-
-void bch2_write_op_error(struct printbuf *out, struct bch_write_op *op, u64 offset,
-			 const char *fmt, ...)
-{
-	if (op->subvol)
-		bch2_inum_offset_err_msg(op->c, out,
+	if (op->subvol) {
+		bch2_inum_offset_err_msg(op->c, &buf,
 					 (subvol_inum) { op->subvol, op->pos.inode, },
 					 offset << 9);
-	else {
+	} else {
 		struct bpos pos = op->pos;
 		pos.offset = offset;
-		bch2_inum_snap_offset_err_msg(op->c, out, pos);
+		bch2_inum_snap_offset_err_msg(op->c, &buf, pos);
 	}
 
-	prt_str(out, "write error: ");
+	prt_str(&buf, "write error: ");
 
 	va_list args;
 	va_start(args, fmt);
-	prt_vprintf(out, fmt, args);
+	prt_vprintf(&buf, fmt, args);
 	va_end(args);
 
 	if (op->flags & BCH_WRITE_move) {
 		struct data_update *u = container_of(op, struct data_update, op);
 
-		prt_printf(out, "\n  from internal move ");
-		bch2_bkey_val_to_text(out, op->c, bkey_i_to_s_c(u->k.k));
+		prt_printf(&buf, "\n  from internal move ");
+		bch2_bkey_val_to_text(&buf, op->c, bkey_i_to_s_c(u->k.k));
 	}
+
+	bch_err_ratelimited(op->c, "%s", buf.buf);
+	printbuf_exit(&buf);
 }
 
 void bch2_submit_wbio_replicas(struct bch_write_bio *wbio, struct bch_fs *c,
@@ -554,7 +529,7 @@ static noinline int bch2_write_drop_io_error_ptrs(struct bch_write_op *op)
 					    test_bit(ptr->dev, op->failed.d));
 
 			if (!bch2_bkey_nr_ptrs(bkey_i_to_s_c(src)))
-				return -EIO;
+				return -BCH_ERR_data_write_io;
 		}
 
 		if (dst != src)
@@ -598,11 +573,8 @@ static void __bch2_write_index(struct bch_write_op *op)
 		if (unlikely(ret && !bch2_err_matches(ret, EROFS))) {
 			struct bkey_i *insert = bch2_keylist_front(&op->insert_keys);
 
-			struct printbuf buf = PRINTBUF;
-			bch2_write_op_error(&buf, op, bkey_start_offset(&insert->k),
+			bch2_write_op_error(op, bkey_start_offset(&insert->k),
 					    "btree update error: %s", bch2_err_str(ret));
-			bch_err_ratelimited(c, "%s", buf.buf);
-			printbuf_exit(&buf);
 		}
 
 		if (ret)
@@ -611,7 +583,7 @@ static void __bch2_write_index(struct bch_write_op *op)
 out:
 	/* If some a bucket wasn't written, we can't erasure code it: */
 	for_each_set_bit(dev, op->failed.d, BCH_SB_MEMBERS_MAX)
-		bch2_open_bucket_write_error(c, &op->open_buckets, dev);
+		bch2_open_bucket_write_error(c, &op->open_buckets, dev, -BCH_ERR_data_write_io);
 
 	bch2_open_buckets_put(c, &op->open_buckets);
 	return;
@@ -837,7 +809,6 @@ static int bch2_write_rechecksum(struct bch_fs *c,
 {
 	struct bio *bio = &op->wbio.bio;
 	struct bch_extent_crc_unpacked new_crc;
-	int ret;
 
 	/* bch2_rechecksum_bio() can't encrypt or decrypt data: */
 
@@ -845,10 +816,10 @@ static int bch2_write_rechecksum(struct bch_fs *c,
 	    bch2_csum_type_is_encryption(new_csum_type))
 		new_csum_type = op->crc.csum_type;
 
-	ret = bch2_rechecksum_bio(c, bio, op->version, op->crc,
-				  NULL, &new_crc,
-				  op->crc.offset, op->crc.live_size,
-				  new_csum_type);
+	int ret = bch2_rechecksum_bio(c, bio, op->version, op->crc,
+				      NULL, &new_crc,
+				      op->crc.offset, op->crc.live_size,
+				      new_csum_type);
 	if (ret)
 		return ret;
 
@@ -858,44 +829,12 @@ static int bch2_write_rechecksum(struct bch_fs *c,
 	return 0;
 }
 
-static int bch2_write_decrypt(struct bch_write_op *op)
-{
-	struct bch_fs *c = op->c;
-	struct nonce nonce = extent_nonce(op->version, op->crc);
-	struct bch_csum csum;
-	int ret;
-
-	if (!bch2_csum_type_is_encryption(op->crc.csum_type))
-		return 0;
-
-	/*
-	 * If we need to decrypt data in the write path, we'll no longer be able
-	 * to verify the existing checksum (poly1305 mac, in this case) after
-	 * it's decrypted - this is the last point we'll be able to reverify the
-	 * checksum:
-	 */
-	csum = bch2_checksum_bio(c, op->crc.csum_type, nonce, &op->wbio.bio);
-	if (bch2_crc_cmp(op->crc.csum, csum) && !c->opts.no_data_io)
-		return -EIO;
-
-	ret = bch2_encrypt_bio(c, op->crc.csum_type, nonce, &op->wbio.bio);
-	op->crc.csum_type = 0;
-	op->crc.csum = (struct bch_csum) { 0, 0 };
-	return ret;
-}
-
-static enum prep_encoded_ret {
-	PREP_ENCODED_OK,
-	PREP_ENCODED_ERR,
-	PREP_ENCODED_CHECKSUM_ERR,
-	PREP_ENCODED_DO_WRITE,
-} bch2_write_prep_encoded_data(struct bch_write_op *op, struct write_point *wp)
+static noinline int bch2_write_prep_encoded_data(struct bch_write_op *op, struct write_point *wp)
 {
 	struct bch_fs *c = op->c;
 	struct bio *bio = &op->wbio.bio;
-
-	if (!(op->flags & BCH_WRITE_data_encoded))
-		return PREP_ENCODED_OK;
+	struct bch_csum csum;
+	int ret = 0;
 
 	BUG_ON(bio_sectors(bio) != op->crc.compressed_size);
 
@@ -906,12 +845,13 @@ static enum prep_encoded_ret {
 	    (op->crc.compression_type == bch2_compression_opt_to_type(op->compression_opt) ||
 	     op->incompressible)) {
 		if (!crc_is_compressed(op->crc) &&
-		    op->csum_type != op->crc.csum_type &&
-		    bch2_write_rechecksum(c, op, op->csum_type) &&
-		    !c->opts.no_data_io)
-			return PREP_ENCODED_CHECKSUM_ERR;
+		    op->csum_type != op->crc.csum_type) {
+			ret = bch2_write_rechecksum(c, op, op->csum_type);
+			if (ret)
+				return ret;
+		}
 
-		return PREP_ENCODED_DO_WRITE;
+		return 1;
 	}
 
 	/*
@@ -919,20 +859,24 @@ static enum prep_encoded_ret {
 	 * is, we have to decompress it:
 	 */
 	if (crc_is_compressed(op->crc)) {
-		struct bch_csum csum;
-
-		if (bch2_write_decrypt(op))
-			return PREP_ENCODED_CHECKSUM_ERR;
-
 		/* Last point we can still verify checksum: */
-		csum = bch2_checksum_bio(c, op->crc.csum_type,
-					 extent_nonce(op->version, op->crc),
-					 bio);
+		struct nonce nonce = extent_nonce(op->version, op->crc);
+		csum = bch2_checksum_bio(c, op->crc.csum_type, nonce, bio);
 		if (bch2_crc_cmp(op->crc.csum, csum) && !c->opts.no_data_io)
-			return PREP_ENCODED_CHECKSUM_ERR;
+			goto csum_err;
 
-		if (bch2_bio_uncompress_inplace(op, bio))
-			return PREP_ENCODED_ERR;
+		if (bch2_csum_type_is_encryption(op->crc.csum_type)) {
+			ret = bch2_encrypt_bio(c, op->crc.csum_type, nonce, bio);
+			if (ret)
+				return ret;
+
+			op->crc.csum_type = 0;
+			op->crc.csum = (struct bch_csum) { 0, 0 };
+		}
+
+		ret = bch2_bio_uncompress_inplace(op, bio);
+		if (ret)
+			return ret;
 	}
 
 	/*
@@ -944,22 +888,44 @@ static enum prep_encoded_ret {
 	 * If the data is checksummed and we're only writing a subset,
 	 * rechecksum and adjust bio to point to currently live data:
 	 */
-	if ((op->crc.live_size != op->crc.uncompressed_size ||
-	     op->crc.csum_type != op->csum_type) &&
-	    bch2_write_rechecksum(c, op, op->csum_type) &&
-	    !c->opts.no_data_io)
-		return PREP_ENCODED_CHECKSUM_ERR;
+	if (op->crc.live_size != op->crc.uncompressed_size ||
+	    op->crc.csum_type != op->csum_type) {
+		ret = bch2_write_rechecksum(c, op, op->csum_type);
+		if (ret)
+			return ret;
+	}
 
 	/*
 	 * If we want to compress the data, it has to be decrypted:
 	 */
-	if ((op->compression_opt ||
-	     bch2_csum_type_is_encryption(op->crc.csum_type) !=
-	     bch2_csum_type_is_encryption(op->csum_type)) &&
-	    bch2_write_decrypt(op))
-		return PREP_ENCODED_CHECKSUM_ERR;
+	if (bch2_csum_type_is_encryption(op->crc.csum_type) &&
+	    (op->compression_opt || op->crc.csum_type != op->csum_type)) {
+		struct nonce nonce = extent_nonce(op->version, op->crc);
+		csum = bch2_checksum_bio(c, op->crc.csum_type, nonce, bio);
+		if (bch2_crc_cmp(op->crc.csum, csum) && !c->opts.no_data_io)
+			goto csum_err;
 
-	return PREP_ENCODED_OK;
+		ret = bch2_encrypt_bio(c, op->crc.csum_type, nonce, bio);
+		if (ret)
+			return ret;
+
+		op->crc.csum_type = 0;
+		op->crc.csum = (struct bch_csum) { 0, 0 };
+	}
+
+	return 0;
+csum_err:
+	bch2_write_op_error(op, op->pos.offset,
+		"error verifying existing checksum while moving existing data (memory corruption?)\n"
+		"  expected %0llx:%0llx got %0llx:%0llx type %s",
+		op->crc.csum.hi,
+		op->crc.csum.lo,
+		csum.hi,
+		csum.lo,
+		op->crc.csum_type < BCH_CSUM_NR
+		? __bch2_csum_types[op->crc.csum_type]
+		: "(unknown)");
+	return -BCH_ERR_data_write_csum;
 }
 
 static int bch2_write_extent(struct bch_write_op *op, struct write_point *wp,
@@ -974,29 +940,28 @@ static int bch2_write_extent(struct bch_write_op *op, struct write_point *wp,
 	bool page_alloc_failed = false;
 	int ret, more = 0;
 
+	if (op->incompressible)
+		op->compression_opt = 0;
+
 	BUG_ON(!bio_sectors(src));
 
 	ec_buf = bch2_writepoint_ec_buf(c, wp);
 
-	switch (bch2_write_prep_encoded_data(op, wp)) {
-	case PREP_ENCODED_OK:
-		break;
-	case PREP_ENCODED_ERR:
-		ret = -EIO;
-		goto err;
-	case PREP_ENCODED_CHECKSUM_ERR:
-		goto csum_err;
-	case PREP_ENCODED_DO_WRITE:
-		/* XXX look for bug here */
-		if (ec_buf) {
-			dst = bch2_write_bio_alloc(c, wp, src,
-						   &page_alloc_failed,
-						   ec_buf);
-			bio_copy_data(dst, src);
-			bounce = true;
+	if (unlikely(op->flags & BCH_WRITE_data_encoded)) {
+		ret = bch2_write_prep_encoded_data(op, wp);
+		if (ret < 0)
+			goto err;
+		if (ret) {
+			if (ec_buf) {
+				dst = bch2_write_bio_alloc(c, wp, src,
+							   &page_alloc_failed,
+							   ec_buf);
+				bio_copy_data(dst, src);
+				bounce = true;
+			}
+			init_append_extent(op, wp, op->version, op->crc);
+			goto do_write;
 		}
-		init_append_extent(op, wp, op->version, op->crc);
-		goto do_write;
 	}
 
 	if (ec_buf ||
@@ -1089,12 +1054,13 @@ static int bch2_write_extent(struct bch_write_op *op, struct write_point *wp,
 			 * data can't be modified (by userspace) while it's in
 			 * flight.
 			 */
-			if (bch2_rechecksum_bio(c, src, version, op->crc,
+			ret = bch2_rechecksum_bio(c, src, version, op->crc,
 					&crc, &op->crc,
 					src_len >> 9,
 					bio_sectors(src) - (src_len >> 9),
-					op->csum_type))
-				goto csum_err;
+					op->csum_type);
+			if (ret)
+				goto err;
 			/*
 			 * rchecksum_bio sets compression_type on crc from op->crc,
 			 * this isn't always correct as sometimes we're changing
@@ -1104,12 +1070,12 @@ static int bch2_write_extent(struct bch_write_op *op, struct write_point *wp,
 			crc.nonce = nonce;
 		} else {
 			if ((op->flags & BCH_WRITE_data_encoded) &&
-			    bch2_rechecksum_bio(c, src, version, op->crc,
+			    (ret = bch2_rechecksum_bio(c, src, version, op->crc,
 					NULL, &op->crc,
 					src_len >> 9,
 					bio_sectors(src) - (src_len >> 9),
-					op->crc.csum_type))
-				goto csum_err;
+					op->crc.csum_type)))
+				goto err;
 
 			crc.compressed_size	= dst_len >> 9;
 			crc.uncompressed_size	= src_len >> 9;
@@ -1168,16 +1134,6 @@ static int bch2_write_extent(struct bch_write_op *op, struct write_point *wp,
 do_write:
 	*_dst = dst;
 	return more;
-csum_err:
-	{
-		struct printbuf buf = PRINTBUF;
-		bch2_write_op_error(&buf, op, op->pos.offset,
-				    "error verifying existing checksum while rewriting existing data (memory corruption?)");
-		bch_err_ratelimited(c, "%s", buf.buf);
-		printbuf_exit(&buf);
-	}
-
-	ret = -EIO;
 err:
 	if (to_wbio(dst)->bounce)
 		bch2_bio_free_pages_pool(c, dst);
@@ -1255,38 +1211,35 @@ static void bch2_nocow_write_convert_unwritten(struct bch_write_op *op)
 {
 	struct bch_fs *c = op->c;
 	struct btree_trans *trans = bch2_trans_get(c);
+	int ret = 0;
 
 	for_each_keylist_key(&op->insert_keys, orig) {
-		int ret = for_each_btree_key_max_commit(trans, iter, BTREE_ID_extents,
+		ret = for_each_btree_key_max_commit(trans, iter, BTREE_ID_extents,
 				     bkey_start_pos(&orig->k), orig->k.p,
 				     BTREE_ITER_intent, k,
 				     NULL, NULL, BCH_TRANS_COMMIT_no_enospc, ({
 			bch2_nocow_write_convert_one_unwritten(trans, &iter, orig, k, op->new_i_size);
 		}));
-
-		if (ret && !bch2_err_matches(ret, EROFS)) {
-			struct bkey_i *insert = bch2_keylist_front(&op->insert_keys);
-
-			struct printbuf buf = PRINTBUF;
-			bch2_write_op_error_trans(trans, &buf, op, bkey_start_offset(&insert->k),
-						  "btree update error: %s", bch2_err_str(ret));
-			bch_err_ratelimited(c, "%s", buf.buf);
-			printbuf_exit(&buf);
-		}
-
-		if (ret) {
-			op->error = ret;
+		if (ret)
 			break;
-		}
 	}
 
 	bch2_trans_put(trans);
+
+	if (ret && !bch2_err_matches(ret, EROFS)) {
+		struct bkey_i *insert = bch2_keylist_front(&op->insert_keys);
+		bch2_write_op_error(op, bkey_start_offset(&insert->k),
+				    "btree update error: %s", bch2_err_str(ret));
+	}
+
+	if (ret)
+		op->error = ret;
 }
 
 static void __bch2_nocow_write_done(struct bch_write_op *op)
 {
 	if (unlikely(op->flags & BCH_WRITE_io_error)) {
-		op->error = -EIO;
+		op->error = -BCH_ERR_data_write_io;
 	} else if (unlikely(op->flags & BCH_WRITE_convert_unwritten))
 		bch2_nocow_write_convert_unwritten(op);
 }
@@ -1436,11 +1389,8 @@ err:
 	darray_exit(&buckets);
 
 	if (ret) {
-		struct printbuf buf = PRINTBUF;
-		bch2_write_op_error(&buf, op, op->pos.offset,
+		bch2_write_op_error(op, op->pos.offset,
 				    "%s(): btree lookup error: %s", __func__, bch2_err_str(ret));
-		bch_err_ratelimited(c, "%s", buf.buf);
-		printbuf_exit(&buf);
 		op->error = ret;
 		op->flags |= BCH_WRITE_submitted;
 	}
@@ -1480,7 +1430,7 @@ err_bucket_stale:
 				    "pointer to invalid bucket in nocow path on device %llu\n  %s",
 				    stale_at->b.inode,
 				    (bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
-		ret = -EIO;
+		ret = -BCH_ERR_data_write_invalid_ptr;
 	} else {
 		/* We can retry this: */
 		ret = -BCH_ERR_transaction_restart;
@@ -1558,13 +1508,9 @@ err:
 			op->flags |= BCH_WRITE_submitted;
 
 			if (unlikely(ret < 0)) {
-				if (!(op->flags & BCH_WRITE_alloc_nowait)) {
-					struct printbuf buf = PRINTBUF;
-					bch2_write_op_error(&buf, op, op->pos.offset,
+				if (!(op->flags & BCH_WRITE_alloc_nowait))
+					bch2_write_op_error(op, op->pos.offset,
 							    "%s(): %s", __func__, bch2_err_str(ret));
-					bch_err_ratelimited(c, "%s", buf.buf);
-					printbuf_exit(&buf);
-				}
 				op->error = ret;
 				break;
 			}
@@ -1691,11 +1637,8 @@ CLOSURE_CALLBACK(bch2_write)
 	wbio_init(bio)->put_bio = false;
 
 	if (unlikely(bio->bi_iter.bi_size & (c->opts.block_size - 1))) {
-		struct printbuf buf = PRINTBUF;
-		bch2_write_op_error(&buf, op, op->pos.offset,
-				    "misaligned write");
-		printbuf_exit(&buf);
-		op->error = -EIO;
+		bch2_write_op_error(op, op->pos.offset, "misaligned write");
+		op->error = -BCH_ERR_data_write_misaligned;
 		goto err;
 	}
 

libbcachefs/io_write.h

@@ -14,13 +14,8 @@ void bch2_bio_alloc_pages_pool(struct bch_fs *, struct bio *, size_t);
 void bch2_submit_wbio_replicas(struct bch_write_bio *, struct bch_fs *,
 			       enum bch_data_type, const struct bkey_i *, bool);
 
-__printf(5, 6)
-void bch2_write_op_error_trans(struct btree_trans *trans, struct printbuf *out,
-			       struct bch_write_op *op, u64, const char *, ...);
-
-__printf(4, 5)
-void bch2_write_op_error(struct printbuf *out, struct bch_write_op *op, u64,
-			 const char *, ...);
+__printf(3, 4)
+void bch2_write_op_error(struct bch_write_op *op, u64, const char *, ...);
 
 #define BCH_WRITE_FLAGS()		\
 	x(alloc_nowait)			\

libbcachefs/journal.c

@@ -62,8 +62,7 @@ static void bch2_journal_buf_to_text(struct printbuf *out, struct journal *j, u6
 		prt_newline(out);
 	}
 
-	prt_printf(out, "expires:\t");
-	prt_printf(out, "%li jiffies\n", buf->expires - jiffies);
+	prt_printf(out, "expires:\t%li jiffies\n", buf->expires - jiffies);
 
 	prt_printf(out, "flags:\t");
 	if (buf->noflush)
@@ -142,6 +141,8 @@ journal_error_check_stuck(struct journal *j, int error, unsigned flags)
 	bool stuck = false;
 	struct printbuf buf = PRINTBUF;
 
+	buf.atomic++;
+
 	if (!(error == -BCH_ERR_journal_full ||
 	      error == -BCH_ERR_journal_pin_full) ||
 	    nr_unwritten_journal_entries(j) ||
@@ -167,12 +168,12 @@ journal_error_check_stuck(struct journal *j, int error, unsigned flags)
 		return stuck;
 	}
 	j->err_seq = journal_cur_seq(j);
-	spin_unlock(&j->lock);
 
-	bch_err(c, "Journal stuck! Hava a pre-reservation but journal full (error %s)",
-		bch2_err_str(error));
-	bch2_journal_debug_to_text(&buf, j);
-	bch_err(c, "%s", buf.buf);
+	__bch2_journal_debug_to_text(&buf, j);
+	spin_unlock(&j->lock);
+	prt_printf(&buf, bch2_fmt(c, "Journal stuck! Hava a pre-reservation but journal full (error %s)"),
+				  bch2_err_str(error));
+	bch2_print_string_as_lines(KERN_ERR, buf.buf);
 
 	printbuf_reset(&buf);
 	bch2_journal_pins_to_text(&buf, j);
@@ -728,8 +729,8 @@ int bch2_journal_res_get_slowpath(struct journal *j, struct journal_res *res,
 
 	struct printbuf buf = PRINTBUF;
 	bch2_journal_debug_to_text(&buf, j);
-	bch_err(c, "Journal stuck? Waited for 10 seconds...\n%s",
-		buf.buf);
+	bch2_print_string_as_lines(KERN_ERR, buf.buf);
+	prt_printf(&buf, bch2_fmt(c, "Journal stuck? Waited for 10 seconds, err %s"), bch2_err_str(ret));
 	printbuf_exit(&buf);
 
 	closure_wait_event(&j->async_wait,
@@ -1510,7 +1511,7 @@ int bch2_dev_journal_init(struct bch_dev *ca, struct bch_sb *sb)
 	unsigned nr_bvecs = DIV_ROUND_UP(JOURNAL_ENTRY_SIZE_MAX, PAGE_SIZE);
 
 	for (unsigned i = 0; i < ARRAY_SIZE(ja->bio); i++) {
-		ja->bio[i] = kmalloc(struct_size(ja->bio[i], bio.bi_inline_vecs,
+		ja->bio[i] = kzalloc(struct_size(ja->bio[i], bio.bi_inline_vecs,
 				     nr_bvecs), GFP_KERNEL);
 		if (!ja->bio[i])
 			return -BCH_ERR_ENOMEM_dev_journal_init;

libbcachefs/journal_io.c

@@ -214,12 +214,12 @@ static int journal_entry_add(struct bch_fs *c, struct bch_dev *ca,
 
 		fsck_err_on(same_device,
 			    c, journal_entry_dup_same_device,
-			    "duplicate journal entry on same device\n  %s",
+			    "duplicate journal entry on same device\n%s",
 			    buf.buf);
 
 		fsck_err_on(not_identical,
 			    c, journal_entry_replicas_data_mismatch,
-			    "found duplicate but non identical journal entries\n  %s",
+			    "found duplicate but non identical journal entries\n%s",
 			    buf.buf);
 
 		if (entry_ptr.csum_good && !identical)
@@ -308,8 +308,8 @@ static void journal_entry_err_msg(struct printbuf *out,
 		break;							\
 	case WRITE:							\
 		bch2_sb_error_count(c, BCH_FSCK_ERR_##_err);		\
-		bch_err(c, "corrupt metadata before write: %s\n", _buf.buf);\
-		if (bch2_fs_inconsistent(c)) {				\
+		if (bch2_fs_inconsistent(c,				\
+				"corrupt metadata before write: %s\n", _buf.buf)) {\
 			ret = -BCH_ERR_fsck_errors_not_fixed;		\
 			goto fsck_err;					\
 		}							\
@@ -764,6 +764,23 @@ static void journal_entry_overwrite_to_text(struct printbuf *out, struct bch_fs
 	journal_entry_btree_keys_to_text(out, c, entry);
 }
 
+static int journal_entry_log_bkey_validate(struct bch_fs *c,
+				struct jset *jset,
+				struct jset_entry *entry,
+				unsigned version, int big_endian,
+				struct bkey_validate_context from)
+{
+	from.flags = 0;
+	return journal_entry_btree_keys_validate(c, jset, entry,
+				version, big_endian, from);
+}
+
+static void journal_entry_log_bkey_to_text(struct printbuf *out, struct bch_fs *c,
+					   struct jset_entry *entry)
+{
+	journal_entry_btree_keys_to_text(out, c, entry);
+}
+
 static int journal_entry_write_buffer_keys_validate(struct bch_fs *c,
 				struct jset *jset,
 				struct jset_entry *entry,
@@ -1371,8 +1388,8 @@ int bch2_journal_read(struct bch_fs *c,
 			missing_end = seq - 1;
 			fsck_err(c, journal_entries_missing,
 				 "journal entries %llu-%llu missing! (replaying %llu-%llu)\n"
-				 "  prev at %s\n"
-				 "  next at %s, continue?",
+				 "prev at %s\n"
+				 "next at %s, continue?",
 				 missing_start, missing_end,
 				 *last_seq, *blacklist_seq - 1,
 				 buf1.buf, buf2.buf);
@@ -1426,7 +1443,7 @@ int bch2_journal_read(struct bch_fs *c,
 		    !bch2_replicas_marked(c, &replicas.e) &&
 		    (le64_to_cpu(i->j.seq) == *last_seq ||
 		     fsck_err(c, journal_entry_replicas_not_marked,
-			      "superblock not marked as containing replicas for journal entry %llu\n  %s",
+			      "superblock not marked as containing replicas for journal entry %llu\n%s",
 			      le64_to_cpu(i->j.seq), buf.buf))) {
 			ret = bch2_mark_replicas(c, &replicas.e);
 			if (ret)
@@ -1623,8 +1640,9 @@ static CLOSURE_CALLBACK(journal_write_done)
 			       : j->noflush_write_time, j->write_start_time);
 
 	if (!w->devs_written.nr) {
-		bch_err(c, "unable to write journal to sufficient devices");
-		err = -EIO;
+		if (!bch2_journal_error(j))
+			bch_err(c, "unable to write journal to sufficient devices");
+		err = -BCH_ERR_journal_write_err;
 	} else {
 		bch2_devlist_to_replicas(&replicas.e, BCH_DATA_journal,
 					 w->devs_written);
@@ -2081,12 +2099,12 @@ CLOSURE_CALLBACK(bch2_journal_write)
 		struct printbuf buf = PRINTBUF;
 		buf.atomic++;
 
+		__bch2_journal_debug_to_text(&buf, j);
+		spin_unlock(&j->lock);
 		prt_printf(&buf, bch2_fmt(c, "Unable to allocate journal write at seq %llu for %zu sectors: %s"),
 					  le64_to_cpu(w->data->seq),
 					  vstruct_sectors(w->data, c->block_bits),
 					  bch2_err_str(ret));
-		__bch2_journal_debug_to_text(&buf, j);
-		spin_unlock(&j->lock);
 		bch2_print_string_as_lines(KERN_ERR, buf.buf);
 		printbuf_exit(&buf);
 	}

libbcachefs/journal_reclaim.c

@@ -645,7 +645,6 @@ static u64 journal_seq_to_flush(struct journal *j)
  * @j:		journal object
  * @direct:	direct or background reclaim?
  * @kicked:	requested to run since we last ran?
- * Returns:	0 on success, or -EIO if the journal has been shutdown
  *
  * Background journal reclaim writes out btree nodes. It should be run
  * early enough so that we never completely run out of journal buckets.
@@ -685,10 +684,9 @@ static int __bch2_journal_reclaim(struct journal *j, bool direct, bool kicked)
 		if (kthread && kthread_should_stop())
 			break;
 
-		if (bch2_journal_error(j)) {
-			ret = -EIO;
+		ret = bch2_journal_error(j);
+		if (ret)
 			break;
-		}
 
 		bch2_journal_do_discards(j);
 

libbcachefs/lru.c

@@ -101,8 +101,7 @@ int bch2_lru_check_set(struct btree_trans *trans,
 			goto err;
 
 		if (fsck_err(trans, alloc_key_to_missing_lru_entry,
-			     "missing %s lru entry\n"
-			     "  %s",
+			     "missing %s lru entry\n%s",
 			     bch2_lru_types[lru_type(lru_k)],
 			     (bch2_bkey_val_to_text(&buf, c, referring_k), buf.buf))) {
 			ret = bch2_lru_set(trans, lru_id, dev_bucket, time);
@@ -190,8 +189,8 @@ static int bch2_check_lru_key(struct btree_trans *trans,
 
 		if (fsck_err(trans, lru_entry_bad,
 			     "incorrect lru entry: lru %s time %llu\n"
-			     "  %s\n"
-			     "  for %s",
+			     "%s\n"
+			     "for %s",
 			     bch2_lru_types[type],
 			     lru_pos_time(lru_k.k->p),
 			     (bch2_bkey_val_to_text(&buf1, c, lru_k), buf1.buf),

libbcachefs/move.c

@@ -528,6 +528,37 @@ int bch2_move_ratelimit(struct moving_context *ctxt)
 	return 0;
 }
 
+/*
+ * Move requires non extents iterators, and there's also no need for it to
+ * signal indirect_extent_missing_error:
+ */
+static struct bkey_s_c bch2_lookup_indirect_extent_for_move(struct btree_trans *trans,
+					    struct btree_iter *iter,
+					    struct bkey_s_c_reflink_p p)
+{
+	if (unlikely(REFLINK_P_ERROR(p.v)))
+		return bkey_s_c_null;
+
+	struct bpos reflink_pos = POS(0, REFLINK_P_IDX(p.v));
+
+	bch2_trans_iter_init(trans, iter,
+			     BTREE_ID_reflink, reflink_pos,
+			     BTREE_ITER_not_extents);
+
+	struct bkey_s_c k = bch2_btree_iter_peek(iter);
+	if (!k.k || bkey_err(k)) {
+		bch2_trans_iter_exit(trans, iter);
+		return k;
+	}
+
+	if (bkey_lt(reflink_pos, bkey_start_pos(k.k))) {
+		bch2_trans_iter_exit(trans, iter);
+		return bkey_s_c_null;
+	}
+
+	return k;
+}
+
 static int bch2_move_data_btree(struct moving_context *ctxt,
 				struct bpos start,
 				struct bpos end,
@@ -592,17 +623,16 @@ static int bch2_move_data_btree(struct moving_context *ctxt,
 		    k.k->type == KEY_TYPE_reflink_p &&
 		    REFLINK_P_MAY_UPDATE_OPTIONS(bkey_s_c_to_reflink_p(k).v)) {
 			struct bkey_s_c_reflink_p p = bkey_s_c_to_reflink_p(k);
-			s64 offset_into_extent	= 0;
 
 			bch2_trans_iter_exit(trans, &reflink_iter);
-			k = bch2_lookup_indirect_extent(trans, &reflink_iter, &offset_into_extent, p, true, 0);
+			k = bch2_lookup_indirect_extent_for_move(trans, &reflink_iter, p);
 			ret = bkey_err(k);
 			if (bch2_err_matches(ret, BCH_ERR_transaction_restart))
 				continue;
 			if (ret)
 				break;
 
-			if (bkey_deleted(k.k))
+			if (!k.k)
 				goto next_nondata;
 
 			/*
@@ -611,7 +641,6 @@ static int bch2_move_data_btree(struct moving_context *ctxt,
 			 * pointer - need to fixup iter->k
 			 */
 			extent_iter = &reflink_iter;
-			offset_into_extent = 0;
 		}
 
 		if (!bkey_extent_is_direct_data(k.k))

libbcachefs/move_types.h

@@ -32,7 +32,7 @@ struct bch_move_stats {
 
 struct move_bucket_key {
 	struct bpos		bucket;
-	u8			gen;
+	unsigned		gen;
 };
 
 struct move_bucket {

libbcachefs/namei.c

@@ -4,8 +4,8 @@
 #include "acl.h"
 #include "btree_update.h"
 #include "dirent.h"
-#include "fs-common.h"
 #include "inode.h"
+#include "namei.h"
 #include "subvolume.h"
 #include "xattr.h"
 
@@ -564,6 +564,8 @@ err:
 	return ret;
 }
 
+/* inum_to_path */
+
 static inline void prt_bytes_reversed(struct printbuf *out, const void *b, unsigned n)
 {
 	bch2_printbuf_make_room(out, n);
@@ -654,3 +656,179 @@ disconnected:
 	prt_str_reversed(path, "(disconnected)");
 	goto out;
 }
+
+/* fsck */
+
+static int bch2_check_dirent_inode_dirent(struct btree_trans *trans,
+					  struct bkey_s_c_dirent d,
+					  struct bch_inode_unpacked *target,
+					  bool in_fsck)
+{
+	struct bch_fs *c = trans->c;
+	struct printbuf buf = PRINTBUF;
+	struct btree_iter bp_iter = { NULL };
+	int ret = 0;
+
+	if (inode_points_to_dirent(target, d))
+		return 0;
+
+	if (!target->bi_dir &&
+	    !target->bi_dir_offset) {
+		fsck_err_on(S_ISDIR(target->bi_mode),
+			    trans, inode_dir_missing_backpointer,
+			    "directory with missing backpointer\n%s",
+			    (printbuf_reset(&buf),
+			     bch2_bkey_val_to_text(&buf, c, d.s_c),
+			     prt_printf(&buf, "\n"),
+			     bch2_inode_unpacked_to_text(&buf, target),
+			     buf.buf));
+
+		fsck_err_on(target->bi_flags & BCH_INODE_unlinked,
+			    trans, inode_unlinked_but_has_dirent,
+			    "inode unlinked but has dirent\n%s",
+			    (printbuf_reset(&buf),
+			     bch2_bkey_val_to_text(&buf, c, d.s_c),
+			     prt_printf(&buf, "\n"),
+			     bch2_inode_unpacked_to_text(&buf, target),
+			     buf.buf));
+
+		target->bi_flags &= ~BCH_INODE_unlinked;
+		target->bi_dir		= d.k->p.inode;
+		target->bi_dir_offset	= d.k->p.offset;
+		return __bch2_fsck_write_inode(trans, target);
+	}
+
+	if (bch2_inode_should_have_single_bp(target) &&
+	    !fsck_err(trans, inode_wrong_backpointer,
+		      "dirent points to inode that does not point back:\n%s",
+		      (bch2_bkey_val_to_text(&buf, c, d.s_c),
+		       prt_newline(&buf),
+		       bch2_inode_unpacked_to_text(&buf, target),
+		       buf.buf)))
+		goto err;
+
+	struct bkey_s_c_dirent bp_dirent =
+		bch2_bkey_get_iter_typed(trans, &bp_iter, BTREE_ID_dirents,
+			      SPOS(target->bi_dir, target->bi_dir_offset, target->bi_snapshot),
+			      0, dirent);
+	ret = bkey_err(bp_dirent);
+	if (ret && !bch2_err_matches(ret, ENOENT))
+		goto err;
+
+	bool backpointer_exists = !ret;
+	ret = 0;
+
+	if (!backpointer_exists) {
+		if (fsck_err(trans, inode_wrong_backpointer,
+			     "inode %llu:%u has wrong backpointer:\n"
+			     "got       %llu:%llu\n"
+			     "should be %llu:%llu",
+			     target->bi_inum, target->bi_snapshot,
+			     target->bi_dir,
+			     target->bi_dir_offset,
+			     d.k->p.inode,
+			     d.k->p.offset)) {
+			target->bi_dir		= d.k->p.inode;
+			target->bi_dir_offset	= d.k->p.offset;
+			ret = __bch2_fsck_write_inode(trans, target);
+		}
+	} else {
+		bch2_bkey_val_to_text(&buf, c, d.s_c);
+		prt_newline(&buf);
+		bch2_bkey_val_to_text(&buf, c, bp_dirent.s_c);
+
+		if (S_ISDIR(target->bi_mode) || target->bi_subvol) {
+			/*
+			 * XXX: verify connectivity of the other dirent
+			 * up to the root before removing this one
+			 *
+			 * Additionally, bch2_lookup would need to cope with the
+			 * dirent it found being removed - or should we remove
+			 * the other one, even though the inode points to it?
+			 */
+			if (in_fsck) {
+				if (fsck_err(trans, inode_dir_multiple_links,
+					     "%s %llu:%u with multiple links\n%s",
+					     S_ISDIR(target->bi_mode) ? "directory" : "subvolume",
+					     target->bi_inum, target->bi_snapshot, buf.buf))
+					ret = bch2_fsck_remove_dirent(trans, d.k->p);
+			} else {
+				bch2_fs_inconsistent(c,
+						"%s %llu:%u with multiple links\n%s",
+						S_ISDIR(target->bi_mode) ? "directory" : "subvolume",
+						target->bi_inum, target->bi_snapshot, buf.buf);
+			}
+
+			goto out;
+		} else {
+			/*
+			 * hardlinked file with nlink 0:
+			 * We're just adjusting nlink here so check_nlinks() will pick
+			 * it up, it ignores inodes with nlink 0
+			 */
+			if (fsck_err_on(!target->bi_nlink,
+					trans, inode_multiple_links_but_nlink_0,
+					"inode %llu:%u type %s has multiple links but i_nlink 0\n%s",
+					target->bi_inum, target->bi_snapshot, bch2_d_types[d.v->d_type], buf.buf)) {
+				target->bi_nlink++;
+				target->bi_flags &= ~BCH_INODE_unlinked;
+				ret = __bch2_fsck_write_inode(trans, target);
+				if (ret)
+					goto err;
+			}
+		}
+	}
+out:
+err:
+fsck_err:
+	bch2_trans_iter_exit(trans, &bp_iter);
+	printbuf_exit(&buf);
+	bch_err_fn(c, ret);
+	return ret;
+}
+
+int __bch2_check_dirent_target(struct btree_trans *trans,
+			       struct btree_iter *dirent_iter,
+			       struct bkey_s_c_dirent d,
+			       struct bch_inode_unpacked *target,
+			       bool in_fsck)
+{
+	struct bch_fs *c = trans->c;
+	struct printbuf buf = PRINTBUF;
+	int ret = 0;
+
+	ret = bch2_check_dirent_inode_dirent(trans, d, target, in_fsck);
+	if (ret)
+		goto err;
+
+	if (fsck_err_on(d.v->d_type != inode_d_type(target),
+			trans, dirent_d_type_wrong,
+			"incorrect d_type: got %s, should be %s:\n%s",
+			bch2_d_type_str(d.v->d_type),
+			bch2_d_type_str(inode_d_type(target)),
+			(printbuf_reset(&buf),
+			 bch2_bkey_val_to_text(&buf, c, d.s_c), buf.buf))) {
+		struct bkey_i_dirent *n = bch2_trans_kmalloc(trans, bkey_bytes(d.k));
+		ret = PTR_ERR_OR_ZERO(n);
+		if (ret)
+			goto err;
+
+		bkey_reassemble(&n->k_i, d.s_c);
+		n->v.d_type = inode_d_type(target);
+		if (n->v.d_type == DT_SUBVOL) {
+			n->v.d_parent_subvol = cpu_to_le32(target->bi_parent_subvol);
+			n->v.d_child_subvol = cpu_to_le32(target->bi_subvol);
+		} else {
+			n->v.d_inum = cpu_to_le64(target->bi_inum);
+		}
+
+		ret = bch2_trans_update(trans, dirent_iter, &n->k_i, 0);
+		if (ret)
+			goto err;
+	}
+err:
+fsck_err:
+	printbuf_exit(&buf);
+	bch_err_fn(c, ret);
+	return ret;
+}

libbcachefs/namei.h

@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0 */
-#ifndef _BCACHEFS_FS_COMMON_H
-#define _BCACHEFS_FS_COMMON_H
+#ifndef _BCACHEFS_NAMEI_H
+#define _BCACHEFS_NAMEI_H
 
 #include "dirent.h"
 
@@ -44,4 +44,29 @@ bool bch2_reinherit_attrs(struct bch_inode_unpacked *,
 
 int bch2_inum_to_path(struct btree_trans *, subvol_inum, struct printbuf *);
 
-#endif /* _BCACHEFS_FS_COMMON_H */
+int __bch2_check_dirent_target(struct btree_trans *,
+			       struct btree_iter *,
+			       struct bkey_s_c_dirent,
+			       struct bch_inode_unpacked *, bool);
+
+static inline bool inode_points_to_dirent(struct bch_inode_unpacked *inode,
+					  struct bkey_s_c_dirent d)
+{
+	return  inode->bi_dir		== d.k->p.inode &&
+		inode->bi_dir_offset	== d.k->p.offset;
+}
+
+static inline int bch2_check_dirent_target(struct btree_trans *trans,
+					   struct btree_iter *dirent_iter,
+					   struct bkey_s_c_dirent d,
+					   struct bch_inode_unpacked *target,
+					   bool in_fsck)
+{
+	if (likely(inode_points_to_dirent(target, d) &&
+		   d.v->d_type == inode_d_type(target)))
+		return 0;
+
+	return __bch2_check_dirent_target(trans, dirent_iter, d, target, in_fsck);
+}
+
+#endif /* _BCACHEFS_NAMEI_H */

libbcachefs/opts.c

@@ -44,7 +44,7 @@ const char * const __bch2_btree_ids[] = {
 	NULL
 };
 
-static const char * const __bch2_csum_types[] = {
+const char * const __bch2_csum_types[] = {
 	BCH_CSUM_TYPES()
 	NULL
 };
@@ -219,10 +219,10 @@ typedef void (*sb_opt_set_fn)(struct bch_sb *, u64);
 typedef u64 (*member_opt_get_fn)(const struct bch_member *);
 typedef void (*member_opt_set_fn)(struct bch_member *, u64);
 
-static const sb_opt_get_fn	BCH2_NO_SB_OPT = NULL;
-static const sb_opt_set_fn	SET_BCH2_NO_SB_OPT = NULL;
-static const member_opt_get_fn	BCH2_NO_MEMBER_OPT = NULL;
-static const member_opt_set_fn	SET_BCH2_NO_MEMBER_OPT = NULL;
+__maybe_unused static const sb_opt_get_fn	BCH2_NO_SB_OPT = NULL;
+__maybe_unused static const sb_opt_set_fn	SET_BCH2_NO_SB_OPT = NULL;
+__maybe_unused static const member_opt_get_fn	BCH2_NO_MEMBER_OPT = NULL;
+__maybe_unused static const member_opt_set_fn	SET_BCH2_NO_MEMBER_OPT = NULL;
 
 #define type_compatible_or_null(_p, _type)				\
 	__builtin_choose_expr(						\
@@ -482,14 +482,12 @@ void bch2_opts_to_text(struct printbuf *out,
 
 int bch2_opt_check_may_set(struct bch_fs *c, struct bch_dev *ca, int id, u64 v)
 {
-	lockdep_assert_held(&c->state_lock);
-
 	int ret = 0;
 
 	switch (id) {
 	case Opt_state:
 		if (ca)
-			return __bch2_dev_set_state(c, ca, v, BCH_FORCE_IF_DEGRADED);
+			return bch2_dev_set_state(c, ca, v, BCH_FORCE_IF_DEGRADED);
 		break;
 
 	case Opt_compression:
@@ -551,14 +549,15 @@ int bch2_parse_one_mount_opt(struct bch_fs *c, struct bch_opts *opts,
 		goto bad_opt;
 
 	ret = bch2_opt_parse(c, &bch2_opt_table[id], val, &v, &err);
-	if (ret == -BCH_ERR_option_needs_open_fs && parse_later) {
-		prt_printf(parse_later, "%s=%s,", name, val);
-		if (parse_later->allocation_failure) {
-			ret = -ENOMEM;
-			goto out;
+	if (ret == -BCH_ERR_option_needs_open_fs) {
+		ret = 0;
+
+		if (parse_later) {
+			prt_printf(parse_later, "%s=%s,", name, val);
+			if (parse_later->allocation_failure)
+				ret = -ENOMEM;
 		}
 
-		ret = 0;
 		goto out;
 	}
 
@@ -569,28 +568,24 @@ int bch2_parse_one_mount_opt(struct bch_fs *c, struct bch_opts *opts,
 		bch2_opt_set_by_id(opts, id, v);
 
 	ret = 0;
-	goto out;
-
-bad_opt:
-	pr_err("Bad mount option %s", name);
-	ret = -BCH_ERR_option_name;
-	goto out;
-
-bad_val:
-	pr_err("Invalid mount option %s", err.buf);
-	ret = -BCH_ERR_option_value;
-
 out:
 	printbuf_exit(&err);
 	return ret;
+bad_opt:
+	ret = -BCH_ERR_option_name;
+	goto out;
+bad_val:
+	ret = -BCH_ERR_option_value;
+	goto out;
 }
 
 int bch2_parse_mount_opts(struct bch_fs *c, struct bch_opts *opts,
-			  struct printbuf *parse_later, char *options)
+			  struct printbuf *parse_later, char *options,
+			  bool ignore_unknown)
 {
 	char *copied_opts, *copied_opts_start;
 	char *opt, *name, *val;
-	int ret;
+	int ret = 0;
 
 	if (!options)
 		return 0;
@@ -615,14 +610,14 @@ int bch2_parse_mount_opts(struct bch_fs *c, struct bch_opts *opts,
 		val	= opt;
 
 		ret = bch2_parse_one_mount_opt(c, opts, parse_later, name, val);
-		if (ret < 0)
-			goto out;
+		if (ret == -BCH_ERR_option_name && ignore_unknown)
+			ret = 0;
+		if (ret) {
+			pr_err("Error parsing option %s: %s", name, bch2_err_str(ret));
+			break;
+		}
 	}
 
-	ret = 0;
-	goto out;
-
-out:
 	kfree(copied_opts_start);
 	return ret;
 }

libbcachefs/opts.h

@@ -16,6 +16,7 @@ extern const char * const bch2_version_upgrade_opts[];
 extern const char * const bch2_sb_features[];
 extern const char * const bch2_sb_compat[];
 extern const char * const __bch2_btree_ids[];
+extern const char * const __bch2_csum_types[];
 extern const char * const __bch2_csum_opts[];
 extern const char * const __bch2_compression_types[];
 extern const char * const bch2_compression_opts[];
@@ -499,11 +500,6 @@ enum fsck_err_opts {
 	  OPT_STR(bch2_member_states),					\
 	  BCH_MEMBER_STATE,		BCH_MEMBER_STATE_rw,		\
 	  "state",	"rw,ro,failed,spare")				\
-	x(fs_size,			u64,				\
-	  OPT_DEVICE|OPT_HIDDEN,					\
-	  OPT_UINT(0, S64_MAX),						\
-	  BCH2_NO_MEMBER_OPT,		0,				\
-	  "size",	"Size of filesystem on device")			\
 	x(bucket_size,			u32,				\
 	  OPT_DEVICE|OPT_HUMAN_READABLE|OPT_SB_FIELD_SECTORS,		\
 	  OPT_UINT(0, S64_MAX),						\
@@ -640,7 +636,7 @@ int bch2_opts_check_may_set(struct bch_fs *);
 int bch2_parse_one_mount_opt(struct bch_fs *, struct bch_opts *,
 			     struct printbuf *, const char *, const char *);
 int bch2_parse_mount_opts(struct bch_fs *, struct bch_opts *, struct printbuf *,
-			  char *);
+			  char *, bool);
 
 /* inode opts: */
 

libbcachefs/printbuf.c

@@ -276,6 +276,25 @@ void bch2_printbuf_indent_add(struct printbuf *buf, unsigned spaces)
 	buf->has_indent_or_tabstops = true;
 }
 
+/**
+ * bch2_printbuf_indent_add_nextline() - add to the current indent level for
+ * subsequent lines
+ *
+ * @buf: printbuf to control
+ * @spaces: number of spaces to add to the current indent level
+ *
+ * Subsequent lines - not the current line - will be indented by @spaces more
+ * spaces.
+ */
+void bch2_printbuf_indent_add_nextline(struct printbuf *buf, unsigned spaces)
+{
+	if (WARN_ON_ONCE(buf->indent + spaces < buf->indent))
+		spaces = 0;
+
+	buf->indent += spaces;
+	buf->has_indent_or_tabstops = true;
+}
+
 /**
  * bch2_printbuf_indent_sub() - subtract from the current indent level
  *

libbcachefs/printbuf.h

@@ -112,6 +112,7 @@ void bch2_printbuf_tabstop_pop(struct printbuf *);
 int bch2_printbuf_tabstop_push(struct printbuf *, unsigned);
 
 void bch2_printbuf_indent_add(struct printbuf *, unsigned);
+void bch2_printbuf_indent_add_nextline(struct printbuf *, unsigned);
 void bch2_printbuf_indent_sub(struct printbuf *, unsigned);
 
 void bch2_prt_newline(struct printbuf *);

libbcachefs/progress.c

@@ -16,10 +16,8 @@ void bch2_progress_init(struct progress_indicator_state *s,
 		if (!(btree_id_mask & BIT_ULL(i)))
 			continue;
 
-		struct disk_accounting_pos acc = {
-			.type		= BCH_DISK_ACCOUNTING_btree,
-			.btree.id	= i,
-		};
+		struct disk_accounting_pos acc;
+		disk_accounting_key_init(acc, btree, .id = i);
 
 		u64 v;
 		bch2_accounting_mem_read(c, disk_accounting_pos_to_bpos(&acc), &v, 1);

libbcachefs/rebalance.c

@@ -600,12 +600,13 @@ void bch2_rebalance_status_to_text(struct printbuf *out, struct bch_fs *c)
 	struct bch_fs_rebalance *r = &c->rebalance;
 
 	/* print pending work */
-	struct disk_accounting_pos acc = { .type = BCH_DISK_ACCOUNTING_rebalance_work, };
+	struct disk_accounting_pos acc;
+	disk_accounting_key_init(acc, rebalance_work);
 	u64 v;
 	bch2_accounting_mem_read(c, disk_accounting_pos_to_bpos(&acc), &v, 1);
 
 	prt_printf(out, "pending work:\t");
-	prt_human_readable_u64(out, v);
+	prt_human_readable_u64(out, v << 9);
 	prt_printf(out, "\n\n");
 
 	prt_str(out, bch2_rebalance_state_strs[r->state]);

libbcachefs/recovery.c

@@ -13,12 +13,12 @@
 #include "disk_accounting.h"
 #include "errcode.h"
 #include "error.h"
-#include "fs-common.h"
 #include "journal_io.h"
 #include "journal_reclaim.h"
 #include "journal_seq_blacklist.h"
 #include "logged_ops.h"
 #include "move.h"
+#include "namei.h"
 #include "quota.h"
 #include "rebalance.h"
 #include "recovery.h"

libbcachefs/recovery_passes.c

@@ -234,28 +234,22 @@ static int bch2_run_recovery_pass(struct bch_fs *c, enum bch_recovery_pass pass)
 
 int bch2_run_online_recovery_passes(struct bch_fs *c)
 {
-	int ret = 0;
-
-	down_read(&c->state_lock);
-
 	for (unsigned i = 0; i < ARRAY_SIZE(recovery_pass_fns); i++) {
 		struct recovery_pass_fn *p = recovery_pass_fns + i;
 
 		if (!(p->when & PASS_ONLINE))
 			continue;
 
-		ret = bch2_run_recovery_pass(c, i);
+		int ret = bch2_run_recovery_pass(c, i);
 		if (bch2_err_matches(ret, BCH_ERR_restart_recovery)) {
 			i = c->curr_recovery_pass;
 			continue;
 		}
 		if (ret)
-			break;
+			return ret;
 	}
 
-	up_read(&c->state_lock);
-
-	return ret;
+	return 0;
 }
 
 int bch2_run_recovery_passes(struct bch_fs *c)

libbcachefs/reflink.c

@@ -193,10 +193,10 @@ static int bch2_indirect_extent_missing_error(struct btree_trans *trans,
 	if (ret)
 		goto err;
 
-	prt_printf(&buf, "-%llu\n  ", (missing_pos.offset + (missing_end - missing_start)) << 9);
+	prt_printf(&buf, "-%llu\n", (missing_pos.offset + (missing_end - missing_start)) << 9);
 	bch2_bkey_val_to_text(&buf, c, p.s_c);
 
-	prt_printf(&buf, "\n  missing reflink btree range %llu-%llu",
+	prt_printf(&buf, "\nmissing reflink btree range %llu-%llu",
 		   missing_start, missing_end);
 
 	if (fsck_err(trans, reflink_p_to_missing_reflink_v, "%s", buf.buf)) {
@@ -323,10 +323,10 @@ static int trans_trigger_reflink_p_segment(struct btree_trans *trans,
 	__le64 *refcount = bkey_refcount(bkey_i_to_s(new));
 	if (!*refcount && (flags & BTREE_TRIGGER_overwrite)) {
 		bch2_bkey_val_to_text(&buf, c, p.s_c);
-		prt_printf(&buf, "\n  ");
+		prt_newline(&buf);
 		bch2_bkey_val_to_text(&buf, c, k);
 		log_fsck_err(trans, reflink_refcount_underflow,
-			     "indirect extent refcount underflow while marking\n  %s",
+			     "indirect extent refcount underflow while marking\n%s",
 			   buf.buf);
 		goto next;
 	}
@@ -795,8 +795,8 @@ static int bch2_gc_write_reflink_key(struct btree_trans *trans,
 	if (fsck_err_on(r->refcount != le64_to_cpu(*refcount),
 			trans, reflink_v_refcount_wrong,
 			"reflink key has wrong refcount:\n"
-			"  %s\n"
-			"  should be %u",
+			"%s\n"
+			"should be %u",
 			(bch2_bkey_val_to_text(&buf, c, k), buf.buf),
 			r->refcount)) {
 		struct bkey_i *new = bch2_bkey_make_mut_noupdate(trans, k);

libbcachefs/sb-errors_format.h

@@ -5,8 +5,7 @@
 enum bch_fsck_flags {
 	FSCK_CAN_FIX		= 1 << 0,
 	FSCK_CAN_IGNORE		= 1 << 1,
-	FSCK_NO_RATELIMIT	= 1 << 2,
-	FSCK_AUTOFIX		= 1 << 3,
+	FSCK_AUTOFIX		= 1 << 2,
 };
 
 #define BCH_SB_ERRS()									\
@@ -311,13 +310,14 @@ enum bch_fsck_flags {
 	x(accounting_key_replicas_nr_required_bad,		279,	FSCK_AUTOFIX)	\
 	x(accounting_key_replicas_devs_unsorted,		280,	FSCK_AUTOFIX)	\
 	x(accounting_key_version_0,				282,	FSCK_AUTOFIX)	\
+	x(accounting_key_nr_counters_wrong,			307,	FSCK_AUTOFIX)	\
 	x(logged_op_but_clean,					283,	FSCK_AUTOFIX)	\
 	x(compression_opt_not_marked_in_sb,			295,	FSCK_AUTOFIX)	\
 	x(compression_type_not_marked_in_sb,			296,	FSCK_AUTOFIX)	\
 	x(directory_size_mismatch,				303,	FSCK_AUTOFIX)	\
 	x(dirent_cf_name_too_big,				304,	0)		\
 	x(dirent_stray_data_after_cf_name,			305,	0)		\
-	x(MAX,							307,	0)
+	x(MAX,							308,	0)
 
 enum bch_sb_error_id {
 #define x(t, n, ...) BCH_FSCK_ERR_##t = n,

libbcachefs/snapshot.c

@@ -485,7 +485,7 @@ static int check_snapshot_tree(struct btree_trans *trans,
 			root_id != bch2_snapshot_root(c, root_id) ||
 			st.k->p.offset != le32_to_cpu(s.tree),
 			trans, snapshot_tree_to_missing_snapshot,
-			"snapshot tree points to missing/incorrect snapshot:\n  %s",
+			"snapshot tree points to missing/incorrect snapshot:\n%s",
 			(bch2_bkey_val_to_text(&buf, c, st.s_c),
 			 prt_newline(&buf),
 			 ret
@@ -505,19 +505,19 @@ static int check_snapshot_tree(struct btree_trans *trans,
 
 	if (fsck_err_on(ret,
 			trans, snapshot_tree_to_missing_subvol,
-			"snapshot tree points to missing subvolume:\n  %s",
+			"snapshot tree points to missing subvolume:\n%s",
 			(printbuf_reset(&buf),
 			 bch2_bkey_val_to_text(&buf, c, st.s_c), buf.buf)) ||
 	    fsck_err_on(!bch2_snapshot_is_ancestor(c,
 						le32_to_cpu(subvol.snapshot),
 						root_id),
 			trans, snapshot_tree_to_wrong_subvol,
-			"snapshot tree points to subvolume that does not point to snapshot in this tree:\n  %s",
+			"snapshot tree points to subvolume that does not point to snapshot in this tree:\n%s",
 			(printbuf_reset(&buf),
 			 bch2_bkey_val_to_text(&buf, c, st.s_c), buf.buf)) ||
 	    fsck_err_on(BCH_SUBVOLUME_SNAP(&subvol),
 			trans, snapshot_tree_to_snapshot_subvol,
-			"snapshot tree points to snapshot subvolume:\n  %s",
+			"snapshot tree points to snapshot subvolume:\n%s",
 			(printbuf_reset(&buf),
 			 bch2_bkey_val_to_text(&buf, c, st.s_c), buf.buf))) {
 		struct bkey_i_snapshot_tree *u;
@@ -756,7 +756,7 @@ static int check_snapshot(struct btree_trans *trans,
 	} else {
 		if (fsck_err_on(s.subvol,
 				trans, snapshot_should_not_have_subvol,
-				"snapshot should not point to subvol:\n  %s",
+				"snapshot should not point to subvol:\n%s",
 				(bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
 			u = bch2_bkey_make_mut_typed(trans, iter, &k, 0, snapshot);
 			ret = PTR_ERR_OR_ZERO(u);
@@ -774,7 +774,7 @@ static int check_snapshot(struct btree_trans *trans,
 
 	if (fsck_err_on(!ret,
 			trans, snapshot_to_bad_snapshot_tree,
-			"snapshot points to missing/incorrect tree:\n  %s",
+			"snapshot points to missing/incorrect tree:\n%s",
 			(bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
 		ret = snapshot_tree_ptr_repair(trans, iter, k, &s);
 		if (ret)
@@ -786,7 +786,7 @@ static int check_snapshot(struct btree_trans *trans,
 
 	if (fsck_err_on(le32_to_cpu(s.depth) != real_depth,
 			trans, snapshot_bad_depth,
-			"snapshot with incorrect depth field, should be %u:\n  %s",
+			"snapshot with incorrect depth field, should be %u:\n%s",
 			real_depth, (bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
 		u = bch2_bkey_make_mut_typed(trans, iter, &k, 0, snapshot);
 		ret = PTR_ERR_OR_ZERO(u);
@@ -803,7 +803,7 @@ static int check_snapshot(struct btree_trans *trans,
 
 	if (fsck_err_on(!ret,
 			trans, snapshot_bad_skiplist,
-			"snapshot with bad skiplist field:\n  %s",
+			"snapshot with bad skiplist field:\n%s",
 			(bch2_bkey_val_to_text(&buf, c, k), buf.buf))) {
 		u = bch2_bkey_make_mut_typed(trans, iter, &k, 0, snapshot);
 		ret = PTR_ERR_OR_ZERO(u);

libbcachefs/str_hash.c

@@ -232,7 +232,7 @@ bad_hash:
 		goto out;
 
 	if (fsck_err(trans, hash_table_key_wrong_offset,
-		     "hash table key at wrong offset: btree %s inode %llu offset %llu, hashed to %llu\n  %s",
+		     "hash table key at wrong offset: btree %s inode %llu offset %llu, hashed to %llu\n%s",
 		     bch2_btree_id_str(desc->btree_id), hash_k.k->p.inode, hash_k.k->p.offset, hash,
 		     (printbuf_reset(&buf),
 		      bch2_bkey_val_to_text(&buf, c, hash_k), buf.buf))) {

libbcachefs/subvolume.c

@@ -561,6 +561,7 @@ int bch2_subvolume_unlink(struct btree_trans *trans, u32 subvolid)
 	}
 
 	SET_BCH_SUBVOLUME_UNLINKED(&n->v, true);
+	n->v.fs_path_parent = 0;
 	bch2_trans_iter_exit(trans, &iter);
 	return ret;
 }

libbcachefs/super.c

@@ -533,9 +533,11 @@ int bch2_fs_read_write(struct bch_fs *c)
 
 int bch2_fs_read_write_early(struct bch_fs *c)
 {
-	lockdep_assert_held(&c->state_lock);
+	down_write(&c->state_lock);
+	int ret = __bch2_fs_read_write(c, true);
+	up_write(&c->state_lock);
 
-	return __bch2_fs_read_write(c, true);
+	return ret;
 }
 
 /* Filesystem startup/shutdown: */
@@ -1019,38 +1021,39 @@ static void print_mount_opts(struct bch_fs *c)
 int bch2_fs_start(struct bch_fs *c)
 {
 	time64_t now = ktime_get_real_seconds();
-	int ret;
+	int ret = 0;
 
 	print_mount_opts(c);
 
 	down_write(&c->state_lock);
+	mutex_lock(&c->sb_lock);
 
 	BUG_ON(test_bit(BCH_FS_started, &c->flags));
 
-	mutex_lock(&c->sb_lock);
+	if (!bch2_sb_field_get_minsize(&c->disk_sb, ext,
+			sizeof(struct bch_sb_field_ext) / sizeof(u64))) {
+		mutex_unlock(&c->sb_lock);
+		up_write(&c->state_lock);
+		ret = -BCH_ERR_ENOSPC_sb;
+		goto err;
+	}
 
 	ret = bch2_sb_members_v2_init(c);
 	if (ret) {
 		mutex_unlock(&c->sb_lock);
+		up_write(&c->state_lock);
 		goto err;
 	}
 
 	for_each_online_member(c, ca)
 		bch2_members_v2_get_mut(c->disk_sb.sb, ca->dev_idx)->last_mount = cpu_to_le64(now);
 
-	struct bch_sb_field_ext *ext =
-		bch2_sb_field_get_minsize(&c->disk_sb, ext, sizeof(*ext) / sizeof(u64));
 	mutex_unlock(&c->sb_lock);
 
-	if (!ext) {
-		bch_err(c, "insufficient space in superblock for sb_field_ext");
-		ret = -BCH_ERR_ENOSPC_sb;
-		goto err;
-	}
-
 	for_each_rw_member(c, ca)
 		bch2_dev_allocator_add(c, ca);
 	bch2_recalc_capacity(c);
+	up_write(&c->state_lock);
 
 	c->recovery_task = current;
 	ret = BCH_SB_INITIALIZED(c->disk_sb.sb)
@@ -1066,31 +1069,28 @@ int bch2_fs_start(struct bch_fs *c)
 		goto err;
 
 	if (bch2_fs_init_fault("fs_start")) {
-		bch_err(c, "fs_start fault injected");
-		ret = -EINVAL;
+		ret = -BCH_ERR_injected_fs_start;
 		goto err;
 	}
 
 	set_bit(BCH_FS_started, &c->flags);
 	wake_up(&c->ro_ref_wait);
 
+	down_write(&c->state_lock);
 	if (c->opts.read_only) {
 		bch2_fs_read_only(c);
 	} else {
 		ret = !test_bit(BCH_FS_rw, &c->flags)
 			? bch2_fs_read_write(c)
 			: bch2_fs_read_write_late(c);
-		if (ret)
-			goto err;
 	}
+	up_write(&c->state_lock);
 
-	ret = 0;
 err:
 	if (ret)
 		bch_err_msg(c, ret, "starting filesystem");
 	else
 		bch_verbose(c, "done starting filesystem");
-	up_write(&c->state_lock);
 	return ret;
 }
 
@@ -1990,15 +1990,12 @@ int bch2_dev_resize(struct bch_fs *c, struct bch_dev *ca, u64 nbuckets)
 	mutex_unlock(&c->sb_lock);
 
 	if (ca->mi.freespace_initialized) {
-		struct disk_accounting_pos acc = {
-			.type = BCH_DISK_ACCOUNTING_dev_data_type,
-			.dev_data_type.dev = ca->dev_idx,
-			.dev_data_type.data_type = BCH_DATA_free,
-		};
 		u64 v[3] = { nbuckets - old_nbuckets, 0, 0 };
 
 		ret   = bch2_trans_commit_do(ca->fs, NULL, NULL, 0,
-				bch2_disk_accounting_mod(trans, &acc, v, ARRAY_SIZE(v), false)) ?:
+				bch2_disk_accounting_mod2(trans, false, v, dev_data_type,
+							  .dev = ca->dev_idx,
+							  .data_type = BCH_DATA_free)) ?:
 			bch2_dev_freespace_init(c, ca, old_nbuckets, nbuckets);
 		if (ret)
 			goto err;
@@ -2262,7 +2259,7 @@ BCH_DEBUG_PARAMS()
 
 __maybe_unused
 static unsigned bch2_metadata_version = bcachefs_metadata_version_current;
-module_param_named(version, bch2_metadata_version, uint, 0400);
+module_param_named(version, bch2_metadata_version, uint, 0444);
 
 module_exit(bcachefs_exit);
 module_init(bcachefs_init);

libbcachefs/sysfs.c

@@ -257,10 +257,8 @@ static int bch2_compression_stats_to_text(struct printbuf *out, struct bch_fs *c
 	prt_printf(out, "type\tcompressed\runcompressed\raverage extent size\r\n");
 
 	for (unsigned i = 1; i < BCH_COMPRESSION_TYPE_NR; i++) {
-		struct disk_accounting_pos a = {
-			.type			= BCH_DISK_ACCOUNTING_compression,
-			.compression.type	= i,
-		};
+		struct disk_accounting_pos a;
+		disk_accounting_key_init(a, compression, .type = i);
 		struct bpos p = disk_accounting_pos_to_bpos(&a);
 		u64 v[3];
 		bch2_accounting_mem_read(c, p, v, ARRAY_SIZE(v));
@@ -631,8 +629,6 @@ static ssize_t sysfs_opt_store(struct bch_fs *c,
 	if (unlikely(!bch2_write_ref_tryget(c, BCH_WRITE_REF_sysfs)))
 		return -EROFS;
 
-	down_write(&c->state_lock);
-
 	char *tmp = kstrdup(buf, GFP_KERNEL);
 	if (!tmp) {
 		ret = -ENOMEM;
@@ -675,7 +671,6 @@ static ssize_t sysfs_opt_store(struct bch_fs *c,
 
 	ret = size;
 err:
-	up_write(&c->state_lock);
 	bch2_write_ref_put(c, BCH_WRITE_REF_sysfs);
 	return ret;
 }

libbcachefs/time_stats.c

@@ -10,6 +10,9 @@
 #include "eytzinger.h"
 #include "time_stats.h"
 
+/* disable automatic switching to percpu mode */
+#define TIME_STATS_NONPCPU	((unsigned long) 1)
+
 static const struct time_unit time_units[] = {
 	{ "ns",		1		 },
 	{ "us",		NSEC_PER_USEC	 },
@@ -123,11 +126,12 @@ void __bch2_time_stats_update(struct bch2_time_stats *stats, u64 start, u64 end)
 {
 	unsigned long flags;
 
-	if (!stats->buffer) {
+	if ((unsigned long) stats->buffer <= TIME_STATS_NONPCPU) {
 		spin_lock_irqsave(&stats->lock, flags);
 		time_stats_update_one(stats, start, end);
 
-		if (mean_and_variance_weighted_get_mean(stats->freq_stats_weighted, TIME_STATS_MV_WEIGHT) < 32 &&
+		if (!stats->buffer &&
+		    mean_and_variance_weighted_get_mean(stats->freq_stats_weighted, TIME_STATS_MV_WEIGHT) < 32 &&
 		    stats->duration_stats.n > 1024)
 			stats->buffer =
 				alloc_percpu_gfp(struct time_stat_buffer,
@@ -157,7 +161,7 @@ void bch2_time_stats_reset(struct bch2_time_stats *stats)
 	unsigned offset = offsetof(struct bch2_time_stats, min_duration);
 	memset((void *) stats + offset, 0, sizeof(*stats) - offset);
 
-	if (stats->buffer) {
+	if ((unsigned long) stats->buffer > TIME_STATS_NONPCPU) {
 		int cpu;
 		for_each_possible_cpu(cpu)
 			per_cpu_ptr(stats->buffer, cpu)->nr = 0;
@@ -167,7 +171,9 @@ void bch2_time_stats_reset(struct bch2_time_stats *stats)
 
 void bch2_time_stats_exit(struct bch2_time_stats *stats)
 {
-	free_percpu(stats->buffer);
+	if ((unsigned long) stats->buffer > TIME_STATS_NONPCPU)
+		free_percpu(stats->buffer);
+	stats->buffer = NULL;
 }
 
 void bch2_time_stats_init(struct bch2_time_stats *stats)
@@ -177,3 +183,9 @@ void bch2_time_stats_init(struct bch2_time_stats *stats)
 	stats->min_freq = U64_MAX;
 	spin_lock_init(&stats->lock);
 }
+
+void bch2_time_stats_init_no_pcpu(struct bch2_time_stats *stats)
+{
+	bch2_time_stats_init(stats);
+	stats->buffer = (struct time_stat_buffer __percpu *) TIME_STATS_NONPCPU;
+}

libbcachefs/time_stats.h

@@ -145,6 +145,7 @@ static inline bool track_event_change(struct bch2_time_stats *stats, bool v)
 void bch2_time_stats_reset(struct bch2_time_stats *);
 void bch2_time_stats_exit(struct bch2_time_stats *);
 void bch2_time_stats_init(struct bch2_time_stats *);
+void bch2_time_stats_init_no_pcpu(struct bch2_time_stats *);
 
 static inline void bch2_time_stats_quantiles_exit(struct bch2_time_stats_quantiles *statq)
 {

libbcachefs/util.c

@@ -270,7 +270,7 @@ static void __bch2_print_string_as_lines(const char *prefix, const char *lines,
 		locked = console_trylock();
 	}
 
-	while (1) {
+	while (*lines) {
 		p = strchrnul(lines, '\n');
 		printk("%s%.*s\n", prefix, (int) (p - lines), lines);
 		if (!*p)

libbcachefs/util.h

@@ -94,6 +94,7 @@ do {									\
 #define printbuf_tabstop_push(_buf, _n)	bch2_printbuf_tabstop_push(_buf, _n)
 
 #define printbuf_indent_add(_out, _n)	bch2_printbuf_indent_add(_out, _n)
+#define printbuf_indent_add_nextline(_out, _n)	bch2_printbuf_indent_add_nextline(_out, _n)
 #define printbuf_indent_sub(_out, _n)	bch2_printbuf_indent_sub(_out, _n)
 
 #define prt_newline(_out)		bch2_prt_newline(_out)
@@ -431,7 +432,7 @@ static inline void memcpy_u64s_small(void *dst, const void *src,
 static inline void __memcpy_u64s(void *dst, const void *src,
 				 unsigned u64s)
 {
-#ifdef CONFIG_X86_64
+#if defined(CONFIG_X86_64) && !defined(CONFIG_KMSAN)
 	long d0, d1, d2;
 
 	asm volatile("rep ; movsq"
@@ -508,7 +509,7 @@ static inline void __memmove_u64s_up(void *_dst, const void *_src,
 	u64 *dst = (u64 *) _dst + u64s - 1;
 	u64 *src = (u64 *) _src + u64s - 1;
 
-#ifdef CONFIG_X86_64
+#if defined(CONFIG_X86_64) && !defined(CONFIG_KMSAN)
 	long d0, d1, d2;
 
 	asm volatile("std ;\n"

src/commands/mod.rs

@@ -25,28 +25,3 @@ enum Subcommands {
     #[command(visible_aliases = ["subvol"])]
     Subvolume(subvolume::Cli),
 }
-
-// FIXME: Can be removed after bumping MSRV >= 1.77 in favor of `c""` literals
-#[macro_export]
-macro_rules! c_str {
-    ($lit:expr) => {
-        ::std::ffi::CStr::from_bytes_with_nul(concat!($lit, "\0").as_bytes())
-            .unwrap()
-            .as_ptr()
-    };
-}
-
-#[cfg(test)]
-mod tests {
-    use std::ffi::CStr;
-
-    #[test]
-    fn check_cstr_macro() {
-        let literal = c_str!("hello");
-
-        assert_eq!(
-            literal,
-            CStr::from_bytes_with_nul(b"hello\0").unwrap().as_ptr()
-        );
-    }
-}

src/commands/mount.rs

@@ -9,6 +9,7 @@ use std::{
 
 use anyhow::{ensure, Result};
 use bch_bindgen::{bcachefs, bcachefs::bch_sb_handle, opt_set, path_to_cstr};
+use bcachefs::bch_opts;
 use clap::Parser;
 use log::{debug, error, info};
 use uuid::Uuid;
@@ -78,7 +79,7 @@ fn mount_inner(
 
 /// Parse a comma-separated mount options and split out mountflags and filesystem
 /// specific options.
-fn parse_mount_options(options: impl AsRef<str>) -> (Option<String>, libc::c_ulong) {
+fn parse_mountflag_options(options: impl AsRef<str>) -> (Option<String>, libc::c_ulong) {
     use either::Either::{Left, Right};
 
     debug!("parsing mount options: {}", options.as_ref());
@@ -120,8 +121,7 @@ fn parse_mount_options(options: impl AsRef<str>) -> (Option<String>, libc::c_ulo
     )
 }
 
-fn read_super_silent(path: impl AsRef<Path>) -> anyhow::Result<bch_sb_handle> {
-    let mut opts = bcachefs::bch_opts::default();
+fn read_super_silent(path: impl AsRef<Path>, mut opts: bch_opts) -> anyhow::Result<bch_sb_handle> {
     opt_set!(opts, noexcl, 1);
 
     bch_bindgen::sb_io::read_super_silent(path.as_ref(), opts)
@@ -170,11 +170,11 @@ fn udev_bcachefs_info() -> anyhow::Result<HashMap<String, Vec<String>>> {
     Ok(info)
 }
 
-fn get_super_blocks(uuid: Uuid, devices: &[String]) -> Vec<(PathBuf, bch_sb_handle)> {
+fn get_super_blocks(uuid: Uuid, devices: &[String], opts: &bch_opts) -> Vec<(PathBuf, bch_sb_handle)> {
     devices
         .iter()
         .filter_map(|dev| {
-            read_super_silent(PathBuf::from(dev))
+            read_super_silent(PathBuf::from(dev), *opts)
                 .ok()
                 .map(|sb| (PathBuf::from(dev), sb))
         })
@@ -202,6 +202,7 @@ fn get_all_block_devnodes() -> anyhow::Result<Vec<String>> {
 fn get_devices_by_uuid(
     udev_bcachefs: &HashMap<String, Vec<String>>,
     uuid: Uuid,
+    opts: &bch_opts
 ) -> anyhow::Result<Vec<(PathBuf, bch_sb_handle)>> {
     let devices = {
         if !udev_bcachefs.is_empty() {
@@ -216,16 +217,17 @@ fn get_devices_by_uuid(
         }
     };
 
-    Ok(get_super_blocks(uuid, &devices))
+    Ok(get_super_blocks(uuid, &devices, opts))
 }
 
 fn devs_str_sbs_from_uuid(
     udev_info: &HashMap<String, Vec<String>>,
     uuid: &str,
+    opts: &bch_opts
 ) -> anyhow::Result<(String, Vec<bch_sb_handle>)> {
     debug!("enumerating devices with UUID {}", uuid);
 
-    let devs_sbs = Uuid::parse_str(uuid).map(|uuid| get_devices_by_uuid(udev_info, uuid))??;
+    let devs_sbs = Uuid::parse_str(uuid).map(|uuid| get_devices_by_uuid(udev_info, uuid, opts))??;
 
     let devs_str = devs_sbs
         .iter()
@@ -241,15 +243,16 @@ fn devs_str_sbs_from_uuid(
 fn devs_str_sbs_from_device(
     udev_info: &HashMap<String, Vec<String>>,
     device: &Path,
+    opts: &bch_opts
 ) -> anyhow::Result<(String, Vec<bch_sb_handle>)> {
-    let dev_sb = read_super_silent(device)?;
+    let dev_sb = read_super_silent(device, *opts)?;
 
     if dev_sb.sb().number_of_devices() == 1 {
         Ok((device.as_os_str().to_str().unwrap().to_string(), vec![dev_sb]))
     } else {
         let uuid = dev_sb.sb().uuid();
 
-        devs_str_sbs_from_uuid(udev_info, &uuid.to_string())
+        devs_str_sbs_from_uuid(udev_info, &uuid.to_string(), opts)
     }
 }
 
@@ -271,12 +274,16 @@ fn handle_unlock(cli: &Cli, sb: &bch_sb_handle) -> Result<KeyHandle> {
 }
 
 fn cmd_mount_inner(cli: &Cli) -> Result<()> {
+    let (optstr, mountflags) = parse_mountflag_options(&cli.options);
+    let opts = bch_bindgen::opts::parse_mount_opts(None, optstr.as_deref(), true)
+        .unwrap_or_default();
+
     // Grab the udev information once
     let udev_info = udev_bcachefs_info()?;
 
     let (devices, mut sbs) =
         if let Some(("UUID" | "OLD_BLKID_UUID", uuid)) = cli.dev.split_once('=') {
-            devs_str_sbs_from_uuid(&udev_info, uuid)?
+            devs_str_sbs_from_uuid(&udev_info, uuid, &opts)?
         } else if cli.dev.contains(':') {
             // If the device string contains ":" we will assume the user knows the
             // entire list. If they supply a single device it could be either the FS
@@ -287,12 +294,12 @@ fn cmd_mount_inner(cli: &Cli) -> Result<()> {
             let sbs = cli
                 .dev
                 .split(':')
-                .map(read_super_silent)
+                .map(|path| read_super_silent(path, opts))
                 .collect::<Result<Vec<_>>>()?;
 
             (cli.dev.clone(), sbs)
         } else {
-            devs_str_sbs_from_device(&udev_info, Path::new(&cli.dev))?
+            devs_str_sbs_from_device(&udev_info, Path::new(&cli.dev), &opts)?
         };
 
     ensure!(!sbs.is_empty(), "No device(s) to mount specified");
@@ -317,8 +324,7 @@ fn cmd_mount_inner(cli: &Cli) -> Result<()> {
             &cli.options
         );
 
-        let (data, mountflags) = parse_mount_options(&cli.options);
-        mount_inner(devices, mountpoint, "bcachefs", mountflags, data)
+        mount_inner(devices, mountpoint, "bcachefs", mountflags, optstr)
     } else {
         info!(
             "would mount with params: device: {}, options: {}",

src/commands/subvolume.rs

@@ -18,12 +18,14 @@ enum Subcommands {
     #[command(visible_aliases = ["new"])]
     Create {
         /// Paths
+        #[arg(required = true)]
         targets: Vec<PathBuf>,
     },
 
     #[command(visible_aliases = ["del"])]
     Delete {
         /// Path
+        #[arg(required = true)]
         targets: Vec<PathBuf>,
     },