Now that the necessary functions from log.h have been moved to macro.h,
we can stop including log.h in macro.h. This requires modifying source
files all over the tree to include log.h instead.
To resolve conflict with sys/mount.h and linux/mount.h or linux/fs.h.
The conflict between sys/mount.h and linux/mount.h is resolved in
glibc-2.37 (774058d72942249f71d74e7f2b639f77184160a6), but our baseline
is still glibc-2.31. Also, even with the version or newer, still
sys/mount.h conflicts with linux/fs.h, which is included by
linux/btrfs.h.
This introduces our own implementation of sys/mount.h, that can be
simultaneously included with linux/mount.h and linux/fs.h. This also
imports linux/fs.h, linux/mount.h, and several other dependent headers.
The introduced sys/mount.h header itself may not be enough simple, but
by using the header, we can drop most of workarounds in other source files.
This has been depracted since v254 (2023). Let's kill it for
good now, it has been long enough with 2y. Noone has shown up who wants
to keep it. And given it doesn't work in SB world anyway, and is not
measured is quite problematic security wise.
Unfortunately kernel reports EOF if there's an inconsistency between efivarfs var list
and what's actually stored in firmware, c.f. #34304. A zero size env var is not allowed in
efi and hence the variable doesn't really exist in the backing store as long as it is zero
sized, and the kernel calls this "uncommitted". Hence we translate EOF back to ENOENT here,
as with kernel behavior before
https://github.com/torvalds/linux/commit/3fab70c165795431f00ddf9be8b84ddd07bd1f8f
If the kernel changes behaviour (to flush dentries on resume), we can drop
this at some point in the future. But note that the commit is 11
years old at this point so we'll need to deal with the current behaviour for
a long time.
Fix#34304.
The names of these conflict with macros from efi.h that we'll move
to efi-fundamental.h in a later commit. Let's avoid the conflict by
getting rid of these helpers. Arguably this also improves readability
by clearly indicating we're passing arbitrary strings and not constants
to the macros when we invoke them.
Let's make utf8_to_utf16() and utf16_to_utf8() a bit nicer to use by
adding shortcuts for common cases.
This is particularly relevant for utf16_to_utf8() since the
multiplication with 2 is easy to forget.
"static inline" makes sense in .h files. But in .c files it's useless
decoration, the compiler should just make its own decisions there, and
it can do that.
hence, replace all remaining uses of "static line" by a simple" static"
in all .c files (but keep them in .h files, where they make sense)
We use usec_t for storing time value, which is 64bit.
However, usleep() takes useconds_t that is (typically?) 32bit.
Also, usleep() may only support the range [0, 1000000].
This introduce usleep_safe() which takes usec_t.
-1 was used everywhere, but -EBADF or -EBADFD started being used in various
places. Let's make things consistent in the new style.
Note that there are two candidates:
EBADF 9 Bad file descriptor
EBADFD 77 File descriptor in bad state
Since we're initializating the fd, we're just assigning a value that means
"no fd yet", so it's just a bad file descriptor, and the first errno fits
better. If instead we had a valid file descriptor that became invalid because
of some operation or state change, the other errno would fit better.
In some places, initialization is dropped if unnecessary.
On systems lacking EFI or the SecureBoot efi var the caching of this
info didn#t work, since we'd see ENOENT when reading the var, and cache
that, which we then use as reason to retry next time.
Let's fix that and convert ENOENT to "secure boot", because that's what
it really means. All other errors are left as is (and reason to retry).
But let's add some debug logging for that case.
In order to minimize EFI variable NVRAM wear, do not rewrite variables
if they are already in the wanted state (i.e. same data and attributes).
This allows e.g. performing repeat calls of "bootctl install" (which
always rewrites the EFI boot entry) without consuming EFI NVRAM write
cycles.
The logic is that if the options are updated after boot, we *don't* use
the new value. But we still want to print out the changed contents in
bootctl as to not confuse people.
Fixes#19597.
Also https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=988450.
$ build/bootctl systemd-efi-options
quiet
Note: SystemdOptions EFI variable has been modified since boot. New value: debug
The hint is printed to stderr, so scripts should not be confused.
Creating those string dynamically at runtime is slow and unnecessary.
Let's use static strings with a bit of macro magic and the let the compiler
coalesce as much as possible.
$ size build/src/shared/libsystemd-shared-248.so{.old,}
text data bss dec hex filename
2813453 94572 4584 2912609 2c7161 build/src/shared/libsystemd-shared-248.so.old
2812309 94564 4584 2911457 2c6ce1 build/src/shared/libsystemd-shared-248.so
A nice side-effect is that the same form is used everywhere, so it's easier to
figure out all variables that are used, and where each specific variable is
used.
C.f. 2b0445262a.
Note: 'const char *foo = alloca(…);' seems OK. Our coding style document and
alloca(3) only warn against using alloca() in function invocations. Declaring
both stack variable and alloca at the same time should be fine: no matter in
which order they happen, i.e. if the pointer variable is above the contents,
or the contents are above the pointer, or even if the pointer is elided by the
compiler, everything should be fine.
With some versions of the compiler, the _cleanup_ attr makes it think
the variable might be freed/closed when uninitialized, even though it
cannot happen. The added cost is small enough to be worth the benefit,
and optimized builds will help reduce it even further.
https://tools.ietf.org/html/draft-knodel-terminology-02https://lwn.net/Articles/823224/
This gets rid of most but not occasions of these loaded terms:
1. scsi_id and friends are something that is supposed to be removed from
our tree (see #7594)
2. The test suite defines an API used by the ubuntu CI. We can remove
this too later, but this needs to be done in sync with the ubuntu CI.
3. In some cases the terms are part of APIs we call or where we expose
concepts the kernel names the way it names them. (In particular all
remaining uses of the word "slave" in our codebase are like this,
it's used by the POSIX PTY layer, by the network subsystem, the mount
API and the block device subsystem). Getting rid of the term in these
contexts would mean doing some major fixes of the kernel ABI first.
Regarding the replacements: when whitelist/blacklist is used as noun we
replace with with allow list/deny list, and when used as verb with
allow-list/deny-list.
Cache it early in startup of the system manager, right after `/run/systemd` is
created, so that further access to it can be done without accessing the EFI
filesystem at all.
Quoting https://github.com/systemd/systemd/issues/14828#issuecomment-635212615:
> [kernel uses] msleep_interruptible() and that means when the process receives
> any kind of signal masked or not this will abort with EINTR. systemd-logind
> gets signals from the TTY layer all the time though.
> Here's what might be happening: while logind reads the EFI stuff it gets a
> series of signals from the TTY layer, which causes the read() to be aborted
> with EINTR, which means logind will wait 50ms and retry. Which will be
> aborted again, and so on, until quite some time passed. If we'd not wait for
> the 50ms otoh we wouldn't wait so long, as then on each signal we'd
> immediately retry again.
This is a safey net anyway, let's make it fully safe: if the data ends
on an uneven byte, then we need to complete the UTF-16 codepoint first,
before adding the final NUL byte pair. Hence let's suffix with three
NULs, instead of just two.
EFI variable access is nowadays subject to rate limiting by the kernel.
Thus, let's cache the results of checking them, in order to minimize how
often we access them.
Fixes: #14828
On my laptop (Lenovo X1carbo 4th) I very occasionally see test-boot-timestamps
fail with this tb:
262/494 test-boot-timestamps FAIL 0.7348453998565674 s (killed by signal 6 SIGABRT)
08:12:48 SYSTEMD_LANGUAGE_FALLBACK_MAP='/home/zbyszek/src/systemd/src/locale/language-fallback-map' SYSTEMD_KBD_MODEL_MAP='/home/zbyszek/src/systemd/src/locale/kbd-model-map' PATH='/home/zbyszek/src/systemd/build:/home/zbyszek/.local/bin:/usr/lib64/qt-3.3/bin:/usr/share/Modules/bin:/usr/condabin:/usr/lib64/ccache:/usr/local/bin:/usr/local/sbin:/usr/bin:/usr/sbin:/home/zbyszek/bin:/var/lib/snapd/snap/bin' /home/zbyszek/src/systemd/build/test-boot-timestamps
--- stderr ---
Failed to read $container of PID 1, ignoring: Permission denied
Found container virtualization none.
Failed to get SystemdOptions EFI variable, ignoring: Interrupted system call
Failed to read ACPI FPDT: Permission denied
Failed to read LoaderTimeInitUSec: Interrupted system call
Failed to read EFI loader data: Interrupted system call
Assertion 'q >= 0' failed at src/test/test-boot-timestamps.c:84, function main(). Aborting.
Normally it takes ~0.02s, but here there's a slowdown to 0.73 and things fail with EINTR.
This happens only occasionally, and I haven't been able to capture a strace.
It would be to ignore that case in test-boot-timestamps or always translate
EINTR to -ENODATA. Nevertheless, I think it's better to retry, since this gives
as more resilient behaviour and avoids a transient failure.
See
https://github.com/torvalds/linux/blob/master/fs/efivarfs/file.c#L75
and
https://github.com/torvalds/linux/commit/bef3efbeb897b56867e271cdbc5f8adaacaeb9cd.
In SecureBoot mode this is probably not what you want. As your cmdline
is cryptographically signed like when using Type #2 EFI Unified Kernel
Images (https://systemd.io/BOOT_LOADER_SPECIFICATION/) The user's
intention is then that the cmdline should not be modified. You want to
make sure that the system starts up as exactly specified in the signed
artifact.
On my machine stat returns size 22, but only 20 bytes are read:
openat(AT_FDCWD, "/sys/firmware/efi/efivars/LoaderTimeInitUSec-4a67b082-0a4c-41cf-b6c7-440b29bb8c4f", O_RDONLY|O_NOCTTY|O_CLOEXEC) = 3
fstat(3, {st_mode=S_IFREG|0644, st_size=22, ...}) = 0
read(3, "\6\0\0\0", 4) = 4
read(3, "7\0001\0001\0003\0005\0002\0007\0\0\0", 18) = 16
Failed to read LoaderTimeInitUSec: Input/output error
Let's just accept that the kernel is returning inconsistent results.
It seems to happen two only two variables on my machine:
/sys/firmware/efi/efivars/LoaderTimeInitUSec-4a67b082-0a4c-41cf-b6c7-440b29bb8c4f
/sys/firmware/efi/efivars/LoaderTimeMenuUSec-4a67b082-0a4c-41cf-b6c7-440b29bb8c4f
so it might be related to the way we write them.
This makes the naming more consistent: we now have
bootctl systemd-efi-options,
$SYSTEMD_EFI_OPTIONS
and the SystemdOptions EFI variable.
(SystemdEFIOptions would be redundant, because it is only used in the context
of efivars, and users don't interact with that name directly.)
bootctl is adjusted to use 2sp indentation, similarly to systemctl and other
programs.
Remove the prefix with the old name from 'bootctl systemd-efi-options' output,
since it's redundant and we don't want the old name anyway.
Daan De Meyer
Yu Watanabe
Lennart Poettering
wrvsrx
Mike Yuan
Zbigniew Jędrzejewski-Szmek
Anssi Hannula
Jan Janssen
Frantisek Sumsal
Luca Boccassi
Filipe Brandenburger
Arian van Putten