Rebase onto e4ad989506ec7d71f7302cc3067abd82730a4beb HEAD rootfs

Signed-off-by: Alyssa Ross <hi@alyssa.is>

2 files changed, 278 insertions, 0 deletions

diff --git a/pkgs/os-specific/linux/util-linux/bcachefs-patch-set.patch b/pkgs/os-specific/linux/util-linux/bcachefs-patch-set.patch
new file mode 100644
index 00000000000..068744d4f32
--- /dev/null
+++ b/pkgs/os-specific/linux/util-linux/bcachefs-patch-set.patch
@@ -0,0 +1,277 @@
+commit 68564ebb50f8afab5a9527c534417e247cca0b27
+Author: Filipe Manana <fdmanana@kernel.org>
+Date:   Thu Aug 17 10:20:13 2023 +0100
+
+    libmount: Fix regression when mounting with atime
+    
+    A regression was introduced in v2.39 that causes mounting with the atime
+    option to fail:
+    
+      $ mkfs.ext4 -F /dev/sdi
+      $ mount -o atime /dev/sdi /mnt/sdi
+      mount: /mnt/sdi: not mount point or bad option.
+             dmesg(1) may have more information after failed mount system call.
+    
+    The failure comes from the mount_setattr(2) call returning -EINVAL. This
+    is because we pass an invalid value for the attr_clr argument. From a
+    strace capture we have:
+    
+      mount_setattr(4, "", AT_EMPTY_PATH, {attr_set=0, attr_clr=MOUNT_ATTR_NOATIME, propagation=0 /* MS_??? */, userns_fd=0}, 32) = -1 EINVAL (Invalid argument)
+    
+    We can't pass MOUNT_ATTR_NOATIME to mount_setattr(2) through the attr_clr
+    argument because all atime options are exclusive, so in order to set atime
+    one has to pass MOUNT_ATTR__ATIME to attr_clr and leave attr_set as
+    MOUNT_ATTR_RELATIME (which is defined as a value of 0).
+    
+    This can be read from the man page for mount_setattr(2) and also from the
+    kernel source:
+    
+      $ cat fs/namespace.c
+      static int build_mount_kattr(const struct mount_attr *attr, size_t usize,
+                                   struct mount_kattr *kattr, unsigned int flags)
+      {
+          (...)
+          /*
+           * Since the MOUNT_ATTR_<atime> values are an enum, not a bitmap,
+           * users wanting to transition to a different atime setting cannot
+           * simply specify the atime setting in @attr_set, but must also
+           * specify MOUNT_ATTR__ATIME in the @attr_clr field.
+           * So ensure that MOUNT_ATTR__ATIME can't be partially set in
+           * @attr_clr and that @attr_set can't have any atime bits set if
+           * MOUNT_ATTR__ATIME isn't set in @attr_clr.
+           */
+          if (attr->attr_clr & MOUNT_ATTR__ATIME) {
+              if ((attr->attr_clr & MOUNT_ATTR__ATIME) != MOUNT_ATTR__ATIME)
+                  return -EINVAL;
+    
+                  /*
+                   * Clear all previous time settings as they are mutually
+                   * exclusive.
+                   */
+                  kattr->attr_clr |= MNT_RELATIME | MNT_NOATIME;
+                  switch (attr->attr_set & MOUNT_ATTR__ATIME) {
+                  case MOUNT_ATTR_RELATIME:
+                      kattr->attr_set |= MNT_RELATIME;
+                      break;
+                  case MOUNT_ATTR_NOATIME:
+                      kattr->attr_set |= MNT_NOATIME;
+                      break;
+                  case MOUNT_ATTR_STRICTATIME:
+                      break;
+                  default:
+                      return -EINVAL;
+                  }
+        (...)
+    
+    So fix this by setting attr_clr MOUNT_ATTR__ATIME if we want to clear any
+    atime related option.
+    
+    Signed-off-by: Filipe Manana <fdmanana@kernel.org>
+
+diff --git a/libmount/src/optlist.c b/libmount/src/optlist.c
+index 1e962ec6d..0702adae7 100644
+--- a/libmount/src/optlist.c
++++ b/libmount/src/optlist.c
+@@ -875,7 +875,18 @@ int mnt_optlist_get_attrs(struct libmnt_optlist *ls, uint64_t *set, uint64_t *cl
+ 
+ 		if (opt->ent->mask & MNT_INVERT) {
+ 			DBG(OPTLIST, ul_debugobj(ls, " clr: %s", opt->ent->name));
+-			*clr |= x;
++			/*
++			 * All atime settings are mutually exclusive so *clr must
++			 * have MOUNT_ATTR__ATIME set.
++			 *
++			 * See the function fs/namespace.c:build_mount_kattr()
++			 * in the linux kernel source.
++			 */
++			if (x == MOUNT_ATTR_RELATIME || x == MOUNT_ATTR_NOATIME ||
++			    x == MOUNT_ATTR_STRICTATIME)
++				*clr |= MOUNT_ATTR__ATIME;
++			else
++				*clr |= x;
+ 		} else {
+ 			DBG(OPTLIST, ul_debugobj(ls, " set: %s", opt->ent->name));
+ 			*set |= x;
+diff --git a/tests/expected/libmount/context-mount-flags b/tests/expected/libmount/context-mount-flags
+index 960641863..eb71323dd 100644
+--- a/tests/expected/libmount/context-mount-flags
++++ b/tests/expected/libmount/context-mount-flags
+@@ -3,3 +3,6 @@ ro,nosuid,noexec
+ successfully mounted
+ rw,nosuid,noexec
+ successfully umounted
++successfully mounted
++rw,relatime
++successfully umounted
+diff --git a/tests/ts/libmount/context b/tests/ts/libmount/context
+index f5b47185e..a5d2e81a3 100755
+--- a/tests/ts/libmount/context
++++ b/tests/ts/libmount/context
+@@ -116,8 +116,15 @@ $TS_CMD_FINDMNT --kernel --mountpoint $MOUNTPOINT -o VFS-OPTIONS -n >> $TS_OUTPU
+ 
+ ts_run $TESTPROG --umount $MOUNTPOINT >> $TS_OUTPUT 2>> $TS_ERRLOG
+ is_mounted $DEVICE && echo "$DEVICE still mounted" >> $TS_OUTPUT 2>> $TS_ERRLOG
+-ts_finalize_subtest
+ 
++# Test that the atime option works after the migration to use the new kernel mount APIs.
++ts_run $TESTPROG --mount -o atime $DEVICE $MOUNTPOINT >> $TS_OUTPUT 2>> $TS_ERRLOG
++$TS_CMD_FINDMNT --kernel --mountpoint $MOUNTPOINT -o VFS-OPTIONS -n >> $TS_OUTPUT 2>> $TS_ERRLOG
++is_mounted $DEVICE || echo "$DEVICE not mounted" >> $TS_OUTPUT 2>> $TS_ERRLOG
++ts_run $TESTPROG --umount $MOUNTPOINT >> $TS_OUTPUT 2>> $TS_ERRLOG
++is_mounted $DEVICE && echo "$DEVICE still mounted" >> $TS_OUTPUT 2>> $TS_ERRLOG
++
++ts_finalize_subtest
+ 
+ ts_init_subtest "mount-loopdev"
+ mkdir -p $MOUNTPOINT &> /dev/null
+
+commit 1ec71634aa4ef5ddca23d65c8a296f3614231e8a
+Author: Colin Gillespie <colin@cgillespie.xyz>
+Date:   Wed Aug 9 18:28:07 2023 +1000
+
+    libblkid: (bcachefs) fix not detecting large superblocks
+    
+    Probing does not detect bcachefs filesystems with a superblock larger
+    than 4KiB. Bcachefs superblocks grow in size and can become much larger
+    than this.
+    
+    Increase the superblock maximum size limit to 1MiB.
+    
+    Validate the superblock isn't larger than the maximum size defined in
+    the superblocks layout section.
+    
+    (cherry picked from commit 48d573797797650d96456979797c0155d58f61cb)
+
+diff --git a/libblkid/src/superblocks/bcache.c b/libblkid/src/superblocks/bcache.c
+index 40e702d75..236877042 100644
+--- a/libblkid/src/superblocks/bcache.c
++++ b/libblkid/src/superblocks/bcache.c
+@@ -102,6 +102,15 @@ union bcachefs_sb_csum {
+ 	uint8_t raw[16];
+ } __attribute__((packed));
+ 
++struct bcachefs_sb_layout {
++	uint8_t		magic[16];
++	uint8_t		layout_type;
++	uint8_t		sb_max_size_bits;
++	uint8_t		nr_superblocks;
++	uint8_t		pad[5];
++	uint64_t	sb_offset[61];
++} __attribute__((packed));
++
+ struct bcachefs_super_block {
+ 	union bcachefs_sb_csum	csum;
+ 	uint16_t	version;
+@@ -123,7 +132,7 @@ struct bcachefs_super_block {
+ 	uint64_t	flags[8];
+ 	uint64_t	features[2];
+ 	uint64_t	compat[2];
+-	uint8_t		layout[512];
++	struct bcachefs_sb_layout layout;
+ 	struct bcachefs_sb_field _start[];
+ }  __attribute__((packed));
+ 
+@@ -143,7 +152,7 @@ struct bcachefs_super_block {
+ /* granularity of offset and length fields within superblock */
+ #define BCACHEFS_SECTOR_SIZE   512
+ /* maximum superblock size */
+-#define BCACHEFS_SB_MAX_SIZE   4096
++#define BCACHEFS_SB_MAX_SIZE   0x100000
+ /* fields offset within super block */
+ #define BCACHEFS_SB_FIELDS_OFF offsetof(struct bcachefs_super_block, _start)
+ /* tag value for members field */
+@@ -302,6 +311,9 @@ static int probe_bcachefs(blkid_probe pr, const struct blkid_idmag *mag)
+ 		return BLKID_PROBE_NONE;
+ 
+ 	sb_size = BCACHEFS_SB_FIELDS_OFF + BYTES(bcs);
++	if (sb_size > BCACHEFS_SECTOR_SIZE << bcs->layout.sb_max_size_bits)
++		return BLKID_PROBE_NONE;
++
+ 	if (sb_size > BCACHEFS_SB_MAX_SIZE)
+ 		return BLKID_PROBE_NONE;
+ 
+
+commit acbf17ae8f8ee0f941fe98ed12f115f2b349bba8
+Author: Karel Zak <kzak@redhat.com>
+Date:   Wed Aug 23 11:53:45 2023 +0200
+
+    libblkid: (bcachefs) fix compiler warning [-Werror=sign-compare]
+    
+    Addresses: https://github.com/util-linux/util-linux/pull/2427
+    Signed-off-by: Karel Zak <kzak@redhat.com>
+    (cherry picked from commit 17873d38fc97913c0a31d4bd08cfbfe45c4de5be)
+
+diff --git a/libblkid/src/superblocks/bcache.c b/libblkid/src/superblocks/bcache.c
+index 236877042..6ab3fe9d4 100644
+--- a/libblkid/src/superblocks/bcache.c
++++ b/libblkid/src/superblocks/bcache.c
+@@ -311,7 +311,7 @@ static int probe_bcachefs(blkid_probe pr, const struct blkid_idmag *mag)
+ 		return BLKID_PROBE_NONE;
+ 
+ 	sb_size = BCACHEFS_SB_FIELDS_OFF + BYTES(bcs);
+-	if (sb_size > BCACHEFS_SECTOR_SIZE << bcs->layout.sb_max_size_bits)
++	if (sb_size > ((uint64_t) BCACHEFS_SECTOR_SIZE << bcs->layout.sb_max_size_bits))
+ 		return BLKID_PROBE_NONE;
+ 
+ 	if (sb_size > BCACHEFS_SB_MAX_SIZE)
+
+commit 6b9fda87c4e5d0c6f945d7565197f157b9fa3d5f
+Author: Thomas Weißschuh <thomas@t-8ch.de>
+Date:   Wed Aug 23 11:58:33 2023 +0200
+
+    libblkid: (bcachefs) fix size validation
+    
+    Avoid signed shift out-of-bounds.
+    
+    Also mark the constants explitly as unsigned instead of casting.
+    
+    Signed-off-by: Thomas Weißschuh <thomas@t-8ch.de>
+    (cherry picked from commit befe455f59de8c7bc66b85ed52aae8cbc95325fa)
+
+diff --git a/libblkid/src/superblocks/bcache.c b/libblkid/src/superblocks/bcache.c
+index 6ab3fe9d4..28ac4b52b 100644
+--- a/libblkid/src/superblocks/bcache.c
++++ b/libblkid/src/superblocks/bcache.c
+@@ -142,17 +142,19 @@ struct bcachefs_super_block {
+ /* magic string len */
+ #define BCACHE_SB_MAGIC_LEN (sizeof(BCACHE_SB_MAGIC) - 1)
+ /* super block offset */
+-#define BCACHE_SB_OFF       0x1000
++#define BCACHE_SB_OFF       0x1000U
+ /* supper block offset in kB */
+ #define BCACHE_SB_KBOFF     (BCACHE_SB_OFF >> 10)
+ /* magic string offset within super block */
+ #define BCACHE_SB_MAGIC_OFF offsetof(struct bcache_super_block, magic)
+ /* start of checksummed data within superblock */
+-#define BCACHE_SB_CSUMMED_START 8
++#define BCACHE_SB_CSUMMED_START 8U
+ /* granularity of offset and length fields within superblock */
+-#define BCACHEFS_SECTOR_SIZE   512
++#define BCACHEFS_SECTOR_SIZE   512U
++/* maximum superblock size shift */
++#define BCACHEFS_SB_MAX_SIZE_SHIFT   0x10U
+ /* maximum superblock size */
+-#define BCACHEFS_SB_MAX_SIZE   0x100000
++#define BCACHEFS_SB_MAX_SIZE   (1U << BCACHEFS_SB_MAX_SIZE_SHIFT)
+ /* fields offset within super block */
+ #define BCACHEFS_SB_FIELDS_OFF offsetof(struct bcachefs_super_block, _start)
+ /* tag value for members field */
+@@ -311,12 +313,16 @@ static int probe_bcachefs(blkid_probe pr, const struct blkid_idmag *mag)
+ 		return BLKID_PROBE_NONE;
+ 
+ 	sb_size = BCACHEFS_SB_FIELDS_OFF + BYTES(bcs);
+-	if (sb_size > ((uint64_t) BCACHEFS_SECTOR_SIZE << bcs->layout.sb_max_size_bits))
+-		return BLKID_PROBE_NONE;
+ 
+ 	if (sb_size > BCACHEFS_SB_MAX_SIZE)
+ 		return BLKID_PROBE_NONE;
+ 
++	if (bcs->layout.sb_max_size_bits > BCACHEFS_SB_MAX_SIZE_SHIFT)
++		return BLKID_PROBE_NONE;
++
++	if (sb_size > (BCACHEFS_SECTOR_SIZE << bcs->layout.sb_max_size_bits))
++		return BLKID_PROBE_NONE;
++
+ 	sb = blkid_probe_get_sb_buffer(pr, mag, sb_size);
+ 	if (!sb)
+ 		return BLKID_PROBE_NONE;
diff --git a/pkgs/os-specific/linux/util-linux/default.nix b/pkgs/os-specific/linux/util-linux/default.nix
index 45c0d57cbff..d710fabb7ac 100644
--- a/pkgs/os-specific/linux/util-linux/default.nix
+++ b/pkgs/os-specific/linux/util-linux/default.nix
@@ -29,6 +29,7 @@ stdenv.mkDerivation rec {
 
   patches = [
     ./rtcwake-search-PATH-for-shutdown.patch
+    ./bcachefs-patch-set.patch
   ];
 
   # We separate some of the utilities into their own outputs. This