summaryrefslogtreecommitdiff
path: root/fs/efivarfs/super.c
AgeCommit message (Collapse)AuthorFilesLines
2025-05-26Merge tag 'vfs-6.16-rc1.super' of ↵Linus Torvalds1-144/+51
git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs Pull vfs freezing updates from Christian Brauner: "This contains various filesystem freezing related work for this cycle: - Allow the power subsystem to support filesystem freeze for suspend and hibernate. Now all the pieces are in place to actually allow the power subsystem to freeze/thaw filesystems during suspend/resume. Filesystems are only frozen and thawed if the power subsystem does actually own the freeze. If the filesystem is already frozen by the time we've frozen all userspace processes we don't care to freeze it again. That's userspace's job once the process resumes. We only actually freeze filesystems if we absolutely have to and we ignore other failures to freeze. We could bubble up errors and fail suspend/resume if the error isn't EBUSY (aka it's already frozen) but I don't think that this is worth it. Filesystem freezing during suspend/resume is best-effort. If the user has 500 ext4 filesystems mounted and 4 fail to freeze for whatever reason then we simply skip them. What we have now is already a big improvement and let's see how we fare with it before making our lives even harder (and uglier) than we have to. - Allow efivars to support freeze and thaw Allow efivarfs to partake to resync variable state during system hibernation and suspend. Add freeze/thaw support. This is a pretty straightforward implementation. We simply add regular freeze/thaw support for both userspace and the kernel. efivars is the first pseudofilesystem that adds support for filesystem freezing and thawing. The simplicity comes from the fact that we simply always resync variable state after efivarfs has been frozen. It doesn't matter whether that's because of suspend, userspace initiated freeze or hibernation. Efivars is simple enough that it doesn't matter that we walk all dentries. There are no directories and there aren't insane amounts of entries and both freeze/thaw are already heavy-handed operations. If userspace initiated a freeze/thaw cycle they would need CAP_SYS_ADMIN in the initial user namespace (as that's where efivarfs is mounted) so it can't be triggered by random userspace. IOW, we really really don't care" * tag 'vfs-6.16-rc1.super' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: f2fs: fix freezing filesystem during resize kernfs: add warning about implementing freeze/thaw efivarfs: support freeze/thaw power: freeze filesystems during suspend/resume libfs: export find_next_child() super: add filesystem freezing helpers for suspend and hibernate gfs2: pass through holder from the VFS for freeze/thaw super: use common iterator (Part 2) super: use a common iterator (Part 1) super: skip dying superblocks early super: simplify user_get_super() super: remove pointless s_root checks fs: allow all writers to be frozen locking/percpu-rwsem: add freezable alternative to down_read
2025-05-09efivarfs: support freeze/thawChristian Brauner1-144/+51
Allow efivarfs to partake to resync variable state during system hibernation and suspend. Add freeze/thaw support. This is a pretty straightforward implementation. We simply add regular freeze/thaw support for both userspace and the kernel. This works without any big issues and congrats afaict efivars is the first pseudofilesystem that adds support for filesystem freezing and thawing. The simplicity comes from the fact that we simply always resync variable state after efivarfs has been frozen. It doesn't matter whether that's because of suspend, userspace initiated freeze or hibernation. Efivars is simple enough that it doesn't matter that we walk all dentries. There are no directories and there aren't insane amounts of entries and both freeze/thaw are already heavy-handed operations. We really really don't need to care. Link: https://lore.kernel.org/r/20250331-work-freeze-v1-2-6dfbe8253b9f@kernel.org Signed-off-by: Christian Brauner <brauner@kernel.org>
2025-04-08Use try_lookup_noperm() instead of d_hash_and_lookup() outside of VFSNeilBrown1-10/+5
try_lookup_noperm() and d_hash_and_lookup() are nearly identical. The former does some validation of the name where the latter doesn't. Outside of the VFS that validation is likely valuable, and having only one exported function for this task is certainly a good idea. So make d_hash_and_lookup() local to VFS files and change all other callers to try_lookup_noperm(). Note that the arguments are swapped. Signed-off-by: NeilBrown <neilb@suse.de> Link: https://lore.kernel.org/r/20250319031545.2999807-6-neil@brown.name Signed-off-by: Christian Brauner <brauner@kernel.org>
2025-03-18efivarfs: fix NULL dereference on resumeJames Bottomley1-2/+48
LSMs often inspect the path.mnt of files in the security hooks, and this causes a NULL deref in efivarfs_pm_notify() because the path is constructed with a NULL path.mnt. Fix by obtaining from vfs_kern_mount() instead, and being very careful to ensure that deactivate_super() (potentially triggered by a racing userspace umount) is not called directly from the notifier, because it would deadlock when efivarfs_kill_sb() tried to unregister the notifier chain. [ Al notes: Umm... That's probably safe, but not as a long-term solution - it's too intimately dependent upon fs/super.c internals. The reasons why you can't run into ->s_umount deadlock here are non-trivial... ] Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Link: https://lore.kernel.org/r/e54e6a2f-1178-4980-b771-4d9bafc2aa47@tnxip.de Link: https://lore.kernel.org/r/3e998bf87638a442cbc6864cdcd3d8d9e08ce3e3.camel@HansenPartnership.com Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-03-17efivarfs: use I_MUTEX_CHILD nested lock to traverse variables on resumeArd Biesheuvel1-1/+1
syzbot warns about a potential deadlock, but this is a false positive resulting from a missing lockdep annotation: iterate_dir() locks the parent whereas the inode_lock() it warns about locks the child, which is guaranteed to be a different lock. So use inode_lock_nested() instead with the appropriate lock class. Reported-by: syzbot+019072ad24ab1d948228@syzkaller.appspotmail.com Suggested-by: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-02-23efivarfs: Defer PM notifier registration until .fill_superArd Biesheuvel1-1/+2
syzbot reports an issue that turns out to be caused by the fact that the efivarfs PM notifier may be invoked before the efivarfs_fs_info::sb field is populated, resulting in a NULL deference. So defer the registration until efivarfs_fill_super() is invoked. Reported-by: syzbot+00d13e505ef530a45100@syzkaller.appspotmail.com Tested-by: syzbot+00d13e505ef530a45100@syzkaller.appspotmail.com Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-01-22efivarfs: add variable resync after hibernationJames Bottomley1-1/+150
Hibernation allows other OSs to boot and thus the variable state might be altered by the time the hibernation image is resumed. Resync the variable state by looping over all the dentries and update the size (in case of alteration) delete any which no-longer exist. Finally, loop over all efi variables creating any which don't have corresponding dentries. Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> [ardb: - apply error pointer fixup from Dan Carpenter - rebase onto latest version of James's efivarfs rework] Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-01-22efivarfs: abstract initial variable creation routineJames Bottomley1-12/+20
Reuse later for variable creation after hibernation Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-01-19efivarfs: fix error on write to new variable leaving remnantsJames Bottomley1-0/+1
Make variable cleanup go through the fops release mechanism and use zero inode size as the indicator to delete the file. Since all EFI variables must have an initial u32 attribute, zero size occurs either because the update deleted the variable or because an unsuccessful write after create caused the size never to be set in the first place. In the case of multiple racing opens and closes, the open is counted to ensure that the zero size check is done on the last close. Even though this fixes the bug that a create either not followed by a write or followed by a write that errored would leave a remnant file for the variable, the file will appear momentarily globally visible until the last close of the fd deletes it. This is safe because the normal filesystem operations will mediate any races; however, it is still possible for a directory listing at that instant between create and close contain a zero size variable that doesn't exist in the EFI table. Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-01-19efivarfs: remove unused efivarfs_listJames Bottomley1-10/+2
Remove all function helpers and mentions of the efivarfs_list now that all consumers of the list have been removed and entry management goes exclusively through the inode. Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-01-19efivarfs: move variable lifetime management into the inodesJames Bottomley1-22/+29
Make the inodes the default management vehicle for struct efivar_entry, so they are now all freed automatically if the file is removed and on unmount in kill_litter_super(). Remove the now superfluous iterator to free the entries after kill_litter_super(). Also fixes a bug where some entry freeing was missing causing efivarfs to leak memory. Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-01-09efivarfs: make variable_is_present use dcache lookupJames Bottomley1-0/+29
Instead of searching the variable entry list for a variable, use the dcache lookup functions to find it instead. Also add an efivarfs_ prefix to the function now it is no longer static. Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-01-09efivarfs: add helper to convert from UC16 name and GUID to utf8 nameJames Bottomley1-14/+3
These will be used by a later patch to check for uniqueness on initial EFI variable iteration. Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2025-01-09efivarfs: remove unused efi_variable.Attributes and efivar_entry.kobjJames Bottomley1-1/+1
These fields look to be remnants of older code: Attributes was likely meant to stash the variable attributes, but doesn't because we always read them from the variable store and kobj was likely left over from an older iteration of code where we manually created the objects instead of using a filesystem. [ ardb: these fields were used by the sysfs based 'efivars' precursor to efivarfs, which was removed in commit 0f5b2c69a4cb ("efi: vars: Remove deprecated 'efivars' sysfs interface") ] Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2024-12-09efivarfs: Fix error on non-existent fileJames Bottomley1-3/+0
When looking up a non-existent file, efivarfs returns -EINVAL if the file does not conform to the NAME-GUID format and -ENOENT if it does. This is caused by efivars_d_hash() returning -EINVAL if the name is not formatted correctly. This error is returned before simple_lookup() returns a negative dentry, and is the error value that the user sees. Fix by removing this check. If the file does not exist, simple_lookup() will return a negative dentry leading to -ENOENT and efivarfs_create() already has a validity check before it creates an entry (and will correctly return -EINVAL) Signed-off-by: James Bottomley <James.Bottomley@HansenPartnership.com> Cc: <stable@vger.kernel.org> [ardb: make efivarfs_valid_name() static] Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2024-07-02efivarfs: Convert to new uid/gid option parsing helpersEric Sandeen1-8/+4
Convert to new uid/gid option parsing helpers Signed-off-by: Eric Sandeen <sandeen@redhat.com> Link: https://lore.kernel.org/r/ff7c415c-43d4-4a1e-a987-246c781f8850@redhat.com Signed-off-by: Christian Brauner <brauner@kernel.org>
2024-02-25efivarfs: Drop 'duplicates' bool parameter on efivar_init()Ard Biesheuvel1-1/+1
The 'duplicates' bool argument is always true when efivar_init() is called from its only caller so let's just drop it instead. Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2024-02-25efivarfs: Drop redundant cleanup on fill_super() failureArd Biesheuvel1-6/+1
Al points out that kill_sb() will be called if efivarfs_fill_super() fails and so there is no point in cleaning up the efivar entry list. Reported-by: Alexander Viro <viro@zeniv.linux.org.uk> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-12-11efivarfs: automatically update super block flagMasahisa Kojima1-0/+27
efivar operation is updated when the tee_stmm_efi module is probed. tee_stmm_efi module supports SetVariable runtime service, but user needs to manually remount the efivarfs as RW to enable the write access if the previous efivar operation does not support SetVariable and efivarfs is mounted as read-only. This commit notifies the update of efivar operation to efivarfs subsystem, then drops SB_RDONLY flag if the efivar operation supports SetVariable. Signed-off-by: Masahisa Kojima <masahisa.kojima@linaro.org> [ardb: use per-superblock instance of the notifier block] Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-12-11efivarfs: Move efivarfs list into superblock s_fs_infoArd Biesheuvel1-9/+10
syzbot reports issues with concurrent fsopen()/fsconfig() invocations on efivarfs, which are the result of the fact that the efivarfs list (which caches the names and GUIDs of existing EFI variables) is a global structure. In normal use, these issues are unlikely to trigger, even in the presence of multiple mounts of efivarfs, but the execution pattern used by the syzkaller reproducer may result in multiple instances of the superblock that share the global efivarfs list, and this causes list corruption when the list is reinitialized by one user while another is traversing it. So let's move the list head into the superblock s_fs_info field, so that it will never be shared between distinct instances of the superblock. In the common case, there will still be a single instance of this list, but in the artificial syzkaller case, no list corruption can occur any longer. Reported-by: syzbot+1902c359bfcaf39c46f2@syzkaller.appspotmail.com Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-12-11efivarfs: Free s_fs_info on unmountArd Biesheuvel1-0/+3
Now that we allocate a s_fs_info struct on fs context creation, we should ensure that we free it again when the superblock goes away. Fixes: 5329aa5101f7 ("efivarfs: Add uid/gid mount options") Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-12-11efivarfs: Move efivar availability check into FS context initArd Biesheuvel1-6/+3
Instead of checking whether or not EFI variables are available when creating the superblock, check it one step earlier, when initializing the FS context for the mount. This way, no FS context will be created at all, and we can drop the second check at .kill_sb() time entirely. Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-12-11efivarfs: force RO when remounting if SetVariable is not supportedIlias Apalodimas1-0/+12
If SetVariable at runtime is not supported by the firmware we never assign a callback for that function. At the same time mount the efivarfs as RO so no one can call that. However, we never check the permission flags when someone remounts the filesystem as RW. As a result this leads to a crash looking like this: $ mount -o remount,rw /sys/firmware/efi/efivars $ efi-updatevar -f PK.auth PK [ 303.279166] Unable to handle kernel NULL pointer dereference at virtual address 0000000000000000 [ 303.280482] Mem abort info: [ 303.280854] ESR = 0x0000000086000004 [ 303.281338] EC = 0x21: IABT (current EL), IL = 32 bits [ 303.282016] SET = 0, FnV = 0 [ 303.282414] EA = 0, S1PTW = 0 [ 303.282821] FSC = 0x04: level 0 translation fault [ 303.283771] user pgtable: 4k pages, 48-bit VAs, pgdp=000000004258c000 [ 303.284913] [0000000000000000] pgd=0000000000000000, p4d=0000000000000000 [ 303.286076] Internal error: Oops: 0000000086000004 [#1] PREEMPT SMP [ 303.286936] Modules linked in: qrtr tpm_tis tpm_tis_core crct10dif_ce arm_smccc_trng rng_core drm fuse ip_tables x_tables ipv6 [ 303.288586] CPU: 1 PID: 755 Comm: efi-updatevar Not tainted 6.3.0-rc1-00108-gc7d0c4695c68 #1 [ 303.289748] Hardware name: Unknown Unknown Product/Unknown Product, BIOS 2023.04-00627-g88336918701d 04/01/2023 [ 303.291150] pstate: 60400005 (nZCv daif +PAN -UAO -TCO -DIT -SSBS BTYPE=--) [ 303.292123] pc : 0x0 [ 303.292443] lr : efivar_set_variable_locked+0x74/0xec [ 303.293156] sp : ffff800008673c10 [ 303.293619] x29: ffff800008673c10 x28: ffff0000037e8000 x27: 0000000000000000 [ 303.294592] x26: 0000000000000800 x25: ffff000002467400 x24: 0000000000000027 [ 303.295572] x23: ffffd49ea9832000 x22: ffff0000020c9800 x21: ffff000002467000 [ 303.296566] x20: 0000000000000001 x19: 00000000000007fc x18: 0000000000000000 [ 303.297531] x17: 0000000000000000 x16: 0000000000000000 x15: 0000aaaac807ab54 [ 303.298495] x14: ed37489f673633c0 x13: 71c45c606de13f80 x12: 47464259e219acf4 [ 303.299453] x11: ffff000002af7b01 x10: 0000000000000003 x9 : 0000000000000002 [ 303.300431] x8 : 0000000000000010 x7 : ffffd49ea8973230 x6 : 0000000000a85201 [ 303.301412] x5 : 0000000000000000 x4 : ffff0000020c9800 x3 : 00000000000007fc [ 303.302370] x2 : 0000000000000027 x1 : ffff000002467400 x0 : ffff000002467000 [ 303.303341] Call trace: [ 303.303679] 0x0 [ 303.303938] efivar_entry_set_get_size+0x98/0x16c [ 303.304585] efivarfs_file_write+0xd0/0x1a4 [ 303.305148] vfs_write+0xc4/0x2e4 [ 303.305601] ksys_write+0x70/0x104 [ 303.306073] __arm64_sys_write+0x1c/0x28 [ 303.306622] invoke_syscall+0x48/0x114 [ 303.307156] el0_svc_common.constprop.0+0x44/0xec [ 303.307803] do_el0_svc+0x38/0x98 [ 303.308268] el0_svc+0x2c/0x84 [ 303.308702] el0t_64_sync_handler+0xf4/0x120 [ 303.309293] el0t_64_sync+0x190/0x194 [ 303.309794] Code: ???????? ???????? ???????? ???????? (????????) [ 303.310612] ---[ end trace 0000000000000000 ]--- Fix this by adding a .reconfigure() function to the fs operations which we can use to check the requested flags and deny anything that's not RO if the firmware doesn't implement SetVariable at runtime. Fixes: f88814cc2578 ("efi/efivars: Expose RT service availability via efivars abstraction") Signed-off-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-11-07Merge tag 'vfs-6.7.fsid' of ↵Linus Torvalds1-0/+2
git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs Pull vfs fanotify fsid updates from Christian Brauner: "This work is part of the plan to enable fanotify to serve as a drop-in replacement for inotify. While inotify is availabe on all filesystems, fanotify currently isn't. In order to support fanotify on all filesystems two things are needed: (1) all filesystems need to support AT_HANDLE_FID (2) all filesystems need to report a non-zero f_fsid This contains (1) and allows filesystems to encode non-decodable file handlers for fanotify without implementing any exportfs operations by encoding a file id of type FILEID_INO64_GEN from i_ino and i_generation. Filesystems that want to opt out of encoding non-decodable file ids for fanotify that don't support NFS export can do so by providing an empty export_operations struct. This also partially addresses (2) by generating f_fsid for simple filesystems as well as freevxfs. Remaining filesystems will be dealt with by separate patches. Finally, this contains the patch from the current exportfs maintainers which moves exportfs under vfs with Chuck, Jeff, and Amir as maintainers and vfs.git as tree" * tag 'vfs-6.7.fsid' of git://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs: MAINTAINERS: create an entry for exportfs fs: fix build error with CONFIG_EXPORTFS=m or not defined freevxfs: derive f_fsid from bdev->bd_dev fs: report f_fsid from s_dev for "simple" filesystems exportfs: support encoding non-decodeable file handles by default exportfs: define FILEID_INO64_GEN* file handle types exportfs: make ->encode_fh() a mandatory method for NFS export exportfs: add helpers to check if filesystem can encode/decode file handles
2023-11-04Merge tag 'efi-next-for-v6.7' of ↵Linus Torvalds1-0/+68
git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi Pull EFI update from Ard Biesheuvel: "This is the only remaining EFI change, as everything else was taken via -tip this cycle: - implement uid/gid mount options for efivarfs" * tag 'efi-next-for-v6.7' of git://git.kernel.org/pub/scm/linux/kernel/git/efi/efi: efivarfs: Add uid/gid mount options
2023-10-28fs: report f_fsid from s_dev for "simple" filesystemsAmir Goldstein1-0/+2
There are many "simple" filesystems (*) that report null f_fsid in statfs(2). Those "simple" filesystems report sb->s_dev as the st_dev field of the stat syscalls for all inodes of the filesystem (**). In order to enable fanotify reporting of events with fsid on those "simple" filesystems, report the sb->s_dev number in f_fsid field of statfs(2). (*) For most of the "simple" filesystem refered to in this commit, the ->statfs() operation is simple_statfs(). Some of those fs assign the simple_statfs() method directly in their ->s_op struct and some assign it indirectly via a call to simple_fill_super() or to pseudo_fs_fill_super() with either custom or "simple" s_op. We also make the same change to efivarfs and hugetlbfs, although they do not use simple_statfs(), because they use the simple_* inode opreations (e.g. simple_lookup()). (**) For most of the "simple" filesystems, the ->getattr() method is not assigned, so stat() is implemented by generic_fillattr(). A few "simple" filesystem use the simple_getattr() method which also calls generic_fillattr() to fill most of the stat struct. The two exceptions are procfs and 9p. procfs implements several different ->getattr() methods, but they all end up calling generic_fillattr() to fill the st_dev field from sb->s_dev. 9p has more complicated ->getattr() methods, but they too, end up calling generic_fillattr() to fill the st_dev field from sb->s_dev. Note that 9p and kernfs also call simple_statfs() from custom ->statfs() methods which already fill the f_fsid field, but v9fs_statfs() calls simple_statfs() only in case f_fsid was not filled and kenrfs_statfs() overwrites f_fsid after calling simple_statfs(). Link: https://lore.kernel.org/r/20230919094820.g5bwharbmy2dq46w@quack3/ Signed-off-by: Amir Goldstein <amir73il@gmail.com> Link: https://lore.kernel.org/r/20231023143049.2944970-1-amir73il@gmail.com Reviewed-by: Jan Kara <jack@suse.cz> Signed-off-by: Christian Brauner <brauner@kernel.org>
2023-10-20efivarfs: Add uid/gid mount optionsJiao Zhou1-0/+68
Allow UEFI variables to be modified by non-root processes in order to run sandboxed code. This doesn't change the behavior of mounting efivarfs unless uid/gid are specified; by default both are set to root. Signed-off-by: Jiao Zhou <jiaozhou@google.com> Acked-by: Matthew Garrett <mgarrett@aurora.tech> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-09-11efivarfs: fix statfs() on efivarfsHeinrich Schuchardt1-4/+10
Some firmware (notably U-Boot) provides GetVariable() and GetNextVariableName() but not QueryVariableInfo(). With commit d86ff3333cb1 ("efivarfs: expose used and total size") the statfs syscall was broken for such firmware. If QueryVariableInfo() does not exist or returns EFI_UNSUPPORTED, just report the file system size as 0 as statfs_simple() previously did. Fixes: d86ff3333cb1 ("efivarfs: expose used and total size") Link: https://lore.kernel.org/all/20230910045445.41632-1-heinrich.schuchardt@canonical.com/ Signed-off-by: Heinrich Schuchardt <heinrich.schuchardt@canonical.com> [ardb: log warning on QueryVariableInfo() failure] Reviewed-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-05-17efivarfs: expose used and total sizeAnisse Astier1-1/+38
When writing EFI variables, one might get errors with no other message on why it fails. Being able to see how much is used by EFI variables helps analyzing such issues. Since this is not a conventional filesystem, block size is intentionally set to 1 instead of PAGE_SIZE. x86 quirks of reserved size are taken into account; so that available and free size can be different, further helping debugging space issues. With this patch, one can see the remaining space in EFI variable storage via efivarfs, like this: $ df -h /sys/firmware/efi/efivars/ Filesystem Size Used Avail Use% Mounted on efivarfs 176K 106K 66K 62% /sys/firmware/efi/efivars Signed-off-by: Anisse Astier <an.astier@criteo.com> [ardb: - rename efi_reserved_space() to efivar_reserved_space() - whitespace/coding style tweaks] Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-01-26efivarfs: always register filesystemJohan Hovold1-3/+6
The efivar ops are typically registered at subsys init time so that they are available when efivarfs is registered at module init time. Other efivars implementations, such as Google SMI, exist and can currently be built as modules which means that efivar may not be available when efivarfs is initialised. Move the efivar availability check from module init to when the filesystem is mounted to allow late registration of efivars. Signed-off-by: Johan Hovold <johan+linaro@kernel.org> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2023-01-17efi: efivars: drop kobject from efivars_register()Johan Hovold1-1/+1
Since commit 0f5b2c69a4cb ("efi: vars: Remove deprecated 'efivars' sysfs interface") and the removal of the sysfs interface there are no users of the efivars kobject. Drop the kobject argument from efivars_register() and add a new efivar_is_available() helper in favour of the old efivars_kobject(). Note that the new helper uses the prefix 'efivar' (i.e. without an 's') for consistency with efivar_supports_writes() and the rest of the interface (except the registration functions). For the benefit of drivers with optional EFI support, also provide a dummy implementation of efivar_is_available(). Signed-off-by: Johan Hovold <johan+linaro@kernel.org> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2022-12-01efi: vars: prohibit reading random seed variablesJason A. Donenfeld1-0/+3
In anticipation of putting random seeds in EFI variables, it's important that the random GUID namespace of variables remains hidden from userspace. We accomplish this by not populating efivarfs with entries from that GUID, as well as denying the creation of new ones in that GUID. Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2022-06-24efi: vars: Use locking version to iterate over efivars linked listsArd Biesheuvel1-6/+3
Both efivars and efivarfs uses __efivar_entry_iter() to go over the linked list that shadows the list of EFI variables held by the firmware, but fail to call the begin/end helpers that are documented as a prerequisite. So switch to the proper version, which is efivar_entry_iter(). Given that in both cases, efivar_entry_remove() is invoked with the lock held already, don't take the lock there anymore. Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2022-06-24efi: vars: Don't drop lock in the middle of efivar_init()Ard Biesheuvel1-4/+2
Even though the efivars_lock lock is documented as protecting the efivars->ops pointer (among other things), efivar_init() happily releases and reacquires the lock for every EFI variable that it enumerates. This used to be needed because the lock was originally a spinlock, which prevented the callback that is invoked for every variable from being able to sleep. However, releasing the lock could potentially invalidate the ops pointer, but more importantly, it might allow a SetVariable() runtime service call to take place concurrently, and the UEFI spec does not define how this affects an enumeration that is running in parallel using the GetNextVariable() runtime service, which is what efivar_init() uses. In the meantime, the lock has been converted into a semaphore, and the only reason we need to drop the lock is because the efivarfs pseudo filesystem driver will otherwise deadlock when it invokes the efivars API from the callback to create the efivar_entry items and insert them into the linked list. (EFI pstore is affected in a similar way) So let's switch to helpers that can be used while the lock is already taken. This way, we can hold on to the lock throughout the enumeration. Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2020-11-25efivarfs: revert "fix memory leak in efivarfs_create()"Ard Biesheuvel1-1/+0
The memory leak addressed by commit fe5186cf12e3 is a false positive: all allocations are recorded in a linked list, and freed when the filesystem is unmounted. This leads to double frees, and as reported by David, leads to crashes if SLUB is configured to self destruct when double frees occur. So drop the redundant kfree() again, and instead, mark the offending pointer variable so the allocation is ignored by kmemleak. Cc: Vamshi K Sthambamkadi <vamshi.k.sthambamkadi@gmail.com> Fixes: fe5186cf12e3 ("efivarfs: fix memory leak in efivarfs_create()") Reported-by: David Laight <David.Laight@aculab.com> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2020-10-26efivarfs: fix memory leak in efivarfs_create()Vamshi K Sthambamkadi1-0/+1
kmemleak report: unreferenced object 0xffff9b8915fcb000 (size 4096): comm "efivarfs.sh", pid 2360, jiffies 4294920096 (age 48.264s) hex dump (first 32 bytes): 2d 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 -............... 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 00 ................ backtrace: [<00000000cc4d897c>] kmem_cache_alloc_trace+0x155/0x4b0 [<000000007d1dfa72>] efivarfs_create+0x6e/0x1a0 [<00000000e6ee18fc>] path_openat+0xe4b/0x1120 [<000000000ad0414f>] do_filp_open+0x91/0x100 [<00000000ce93a198>] do_sys_openat2+0x20c/0x2d0 [<000000002a91be6d>] do_sys_open+0x46/0x80 [<000000000a854999>] __x64_sys_openat+0x20/0x30 [<00000000c50d89c9>] do_syscall_64+0x38/0x90 [<00000000cecd6b5f>] entry_SYSCALL_64_after_hwframe+0x44/0xa9 In efivarfs_create(), inode->i_private is setup with efivar_entry object which is never freed. Cc: <stable@vger.kernel.org> Signed-off-by: Vamshi K Sthambamkadi <vamshi.k.sthambamkadi@gmail.com> Link: https://lore.kernel.org/r/20201023115429.GA2479@cosmos Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2020-09-26efivarfs: Replace invalid slashes with exclamation marks in dentries.Michael Schaller1-0/+3
Without this patch efivarfs_alloc_dentry creates dentries with slashes in their name if the respective EFI variable has slashes in its name. This in turn causes EIO on getdents64, which prevents a complete directory listing of /sys/firmware/efi/efivars/. This patch replaces the invalid shlashes with exclamation marks like kobject_set_name_vargs does for /sys/firmware/efi/vars/ to have consistently named dentries under /sys/firmware/efi/vars/ and /sys/firmware/efi/efivars/. Signed-off-by: Michael Schaller <misch@google.com> Link: https://lore.kernel.org/r/20200925074502.150448-1-misch@google.com Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2020-07-09efi/efivars: Expose RT service availability via efivars abstractionArd Biesheuvel1-3/+3
Commit bf67fad19e493b ("efi: Use more granular check for availability for variable services") introduced a check into the efivarfs, efi-pstore and other drivers that aborts loading of the module if not all three variable runtime services (GetVariable, SetVariable and GetNextVariable) are supported. However, this results in efivarfs being unavailable entirely if only SetVariable support is missing, which is only needed if you want to make any modifications. Also, efi-pstore and the sysfs EFI variable interface could be backed by another implementation of the 'efivars' abstraction, in which case it is completely irrelevant which services are supported by the EFI firmware. So make the generic 'efivars' abstraction dependent on the availibility of the GetVariable and GetNextVariable EFI runtime services, and add a helper 'efivar_supports_writes()' to find out whether the currently active efivars abstraction supports writes (and wire it up to the availability of SetVariable for the generic one). Then, use the efivar_supports_writes() helper to decide whether to permit efivarfs to be mounted read-write, and whether to enable efi-pstore or the sysfs EFI variable interface altogether. Fixes: bf67fad19e493b ("efi: Use more granular check for availability for variable services") Reported-by: Heinrich Schuchardt <xypron.glpk@gmx.de> Acked-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Tested-by: Ilias Apalodimas <ilias.apalodimas@linaro.org> Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2020-02-23efi: Use more granular check for availability for variable servicesArd Biesheuvel1-1/+1
The UEFI spec rev 2.8 permits firmware implementations to support only a subset of EFI runtime services at OS runtime (i.e., after the call to ExitBootServices()), so let's take this into account in the drivers that rely specifically on the availability of the EFI variable services. Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
2019-07-19Merge branch 'work.mount0' of ↵Linus Torvalds1-10/+15
git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs Pull vfs mount updates from Al Viro: "The first part of mount updates. Convert filesystems to use the new mount API" * 'work.mount0' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: (63 commits) mnt_init(): call shmem_init() unconditionally constify ksys_mount() string arguments don't bother with registering rootfs init_rootfs(): don't bother with init_ramfs_fs() vfs: Convert smackfs to use the new mount API vfs: Convert selinuxfs to use the new mount API vfs: Convert securityfs to use the new mount API vfs: Convert apparmorfs to use the new mount API vfs: Convert openpromfs to use the new mount API vfs: Convert xenfs to use the new mount API vfs: Convert gadgetfs to use the new mount API vfs: Convert oprofilefs to use the new mount API vfs: Convert ibmasmfs to use the new mount API vfs: Convert qib_fs/ipathfs to use the new mount API vfs: Convert efivarfs to use the new mount API vfs: Convert configfs to use the new mount API vfs: Convert binfmt_misc to use the new mount API convenience helper: get_tree_single() convenience helper get_tree_nodev() vfs: Kill sget_userns() ...
2019-07-05vfs: Convert efivarfs to use the new mount APIDavid Howells1-10/+15
Convert the efivarfs filesystem to the new internal mount API as the old one will be obsoleted and removed. This allows greater flexibility in communication of mount parameters between userspace, the VFS and the filesystem. [AV: get rid of efivarfs_sb nonsense - it has never been used] See Documentation/filesystems/mount_api.txt for more information. Signed-off-by: David Howells <dhowells@redhat.com> cc: Matthew Garrett <matthew.garrett@nebula.com> cc: Jeremy Kerr <jk@ozlabs.org> cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> cc: linux-efi@vger.kernel.org Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2019-06-19treewide: Replace GPLv2 boilerplate/reference with SPDX - rule 500Thomas Gleixner1-4/+1
Based on 2 normalized pattern(s): this program is free software you can redistribute it and or modify it under the terms of the gnu general public license version 2 as published by the free software foundation this program is free software you can redistribute it and or modify it under the terms of the gnu general public license version 2 as published by the free software foundation # extracted by the scancode license scanner the SPDX license identifier GPL-2.0-only has been chosen to replace the boilerplate/reference in 4122 file(s). Signed-off-by: Thomas Gleixner <tglx@linutronix.de> Reviewed-by: Enrico Weigelt <info@metux.net> Reviewed-by: Kate Stewart <kstewart@linuxfoundation.org> Reviewed-by: Allison Randal <allison@lohutok.net> Cc: linux-spdx@vger.kernel.org Link: https://lkml.kernel.org/r/20190604081206.933168790@linutronix.de Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-07-11VFS: Kill off s_options and helpersDavid Howells1-1/+0
Kill off s_options, save/replace_mount_options() and generic_show_options() as all filesystems now implement ->show_options() for themselves. This should make it easier to implement a context-based mount where the mount options can be passed individually over a file descriptor. Signed-off-by: David Howells <dhowells@redhat.com> Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-09-09fs/efivarfs: Fix double kfree() in error pathMatt Fleming1-3/+3
Julia reported that we may double free 'name' in efivarfs_callback(), and that this bug was introduced by commit 0d22f33bc37c ("efi: Don't use spinlocks for efi vars"). Move one of the kfree()s until after the point at which we know we are definitely on the success path. Reported-by: Julia Lawall <julia.lawall@lip6.fr> Acked-by: Julia Lawall <julia.lawall@lip6.fr> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Sylvain Chouleur <sylvain.chouleur@gmail.com> Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
2016-09-09efi: Don't use spinlocks for efi varsSylvain Chouleur1-2/+7
All efivars operations are protected by a spinlock which prevents interruptions and preemption. This is too restricted, we just need a lock preventing concurrency. The idea is to use a semaphore of count 1 and to have two ways of locking, depending on the context: - In interrupt context, we call down_trylock(), if it fails we return an error - In normal context, we call down_interruptible() We don't use a mutex here because the mutex_trylock() function must not be called from interrupt context, whereas the down_trylock() can. Signed-off-by: Sylvain Chouleur <sylvain.chouleur@intel.com> Signed-off-by: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Leif Lindholm <leif.lindholm@linaro.org> Cc: Mark Rutland <mark.rutland@arm.com> Cc: Sylvain Chouleur <sylvain.chouleur@gmail.com> Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
2016-07-31get rid of 'parent' argument of ->d_compare()Al Viro1-2/+1
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
2016-06-11vfs: make the string hashes salt the hashLinus Torvalds1-2/+2
We always mixed in the parent pointer into the dentry name hash, but we did it late at lookup time. It turns out that we can simplify that lookup-time action by salting the hash with the parent pointer early instead of late. A few other users of our string hashes also wanted to mix in their own pointers into the hash, and those are updated to use the same mechanism. Hash users that don't have any particular initial salt can just use the NULL pointer as a no-salt. Cc: Vegard Nossum <vegard.nossum@oracle.com> Cc: George Spelvin <linux@sciencehorizons.net> Cc: Al Viro <viro@zeniv.linux.org.uk> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-05-07efi: Merge boolean flag argumentsJulia Lawall1-2/+1
The parameters atomic and duplicates of efivar_init always have opposite values. Drop the parameter atomic, replace the uses of !atomic with duplicates, and update the call sites accordingly. The code using duplicates is slightly reorganized with an 'else', to avoid duplicating the lock code. Signed-off-by: Julia Lawall <Julia.Lawall@lip6.fr> Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk> Cc: Andy Lutomirski <luto@amacapital.net> Cc: Ard Biesheuvel <ard.biesheuvel@linaro.org> Cc: Borislav Petkov <bp@alien8.de> Cc: Brian Gerst <brgerst@gmail.com> Cc: Denys Vlasenko <dvlasenk@redhat.com> Cc: H. Peter Anvin <hpa@zytor.com> Cc: Jeremy Kerr <jk@ozlabs.org> Cc: Linus Torvalds <torvalds@linux-foundation.org> Cc: Matthew Garrett <mjg59@srcf.ucam.org> Cc: Peter Zijlstra <peterz@infradead.org> Cc: Saurabh Sengar <saurabh.truth@gmail.com> Cc: Thomas Gleixner <tglx@linutronix.de> Cc: Vaishali Thakkar <vaishali.thakkar@oracle.com> Cc: linux-efi@vger.kernel.org Link: http://lkml.kernel.org/r/1462570771-13324-5-git-send-email-matt@codeblueprint.co.uk Signed-off-by: Ingo Molnar <mingo@kernel.org>
2016-04-04mm, fs: get rid of PAGE_CACHE_* and page_cache_{get,release} macrosKirill A. Shutemov1-2/+2
PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} macros were introduced *long* time ago with promise that one day it will be possible to implement page cache with bigger chunks than PAGE_SIZE. This promise never materialized. And unlikely will. We have many places where PAGE_CACHE_SIZE assumed to be equal to PAGE_SIZE. And it's constant source of confusion on whether PAGE_CACHE_* or PAGE_* constant should be used in a particular case, especially on the border between fs and mm. Global switching to PAGE_CACHE_SIZE != PAGE_SIZE would cause to much breakage to be doable. Let's stop pretending that pages in page cache are special. They are not. The changes are pretty straight-forward: - <foo> << (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - <foo> >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) -> <foo>; - PAGE_CACHE_{SIZE,SHIFT,MASK,ALIGN} -> PAGE_{SIZE,SHIFT,MASK,ALIGN}; - page_cache_get() -> get_page(); - page_cache_release() -> put_page(); This patch contains automated changes generated with coccinelle using script below. For some reason, coccinelle doesn't patch header files. I've called spatch for them manually. The only adjustment after coccinelle is revert of changes to PAGE_CAHCE_ALIGN definition: we are going to drop it later. There are few places in the code where coccinelle didn't reach. I'll fix them manually in a separate patch. Comments and documentation also will be addressed with the separate patch. virtual patch @@ expression E; @@ - E << (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ expression E; @@ - E >> (PAGE_CACHE_SHIFT - PAGE_SHIFT) + E @@ @@ - PAGE_CACHE_SHIFT + PAGE_SHIFT @@ @@ - PAGE_CACHE_SIZE + PAGE_SIZE @@ @@ - PAGE_CACHE_MASK + PAGE_MASK @@ expression E; @@ - PAGE_CACHE_ALIGN(E) + PAGE_ALIGN(E) @@ expression E; @@ - page_cache_get(E) + get_page(E) @@ expression E; @@ - page_cache_release(E) + put_page(E) Signed-off-by: Kirill A. Shutemov <kirill.shutemov@linux.intel.com> Acked-by: Michal Hocko <mhocko@suse.com> Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2016-02-10efi: Make efivarfs entries immutable by defaultPeter Jones1-2/+7
"rm -rf" is bricking some peoples' laptops because of variables being used to store non-reinitializable firmware driver data that's required to POST the hardware. These are 100% bugs, and they need to be fixed, but in the mean time it shouldn't be easy to *accidentally* brick machines. We have to have delete working, and picking which variables do and don't work for deletion is quite intractable, so instead make everything immutable by default (except for a whitelist), and make tools that aren't quite so broad-spectrum unset the immutable flag. Signed-off-by: Peter Jones <pjones@redhat.com> Tested-by: Lee, Chun-Yi <jlee@suse.com> Acked-by: Matthew Garrett <mjg59@coreos.com> Signed-off-by: Matt Fleming <matt@codeblueprint.co.uk>
generated by cgit v1.2.3 (git 2.39.0) at 2025-09-25 16:14:36 +0300