_vdso_data is specific to x86 and __arch_get_k_vdso_data() is provided
so that all architectures can provide the requested pointer.
Do the same with _vdso_rng_data, provide __arch_get_k_vdso_rng_data()
and don't use x86 _vdso_rng_data directly.
Until now vdso/vsyscall.h was only included by time/vsyscall.c but now
it will also be included in char/random.c, leading to a duplicate
declaration of _vdso_data and _vdso_rng_data.
To fix this issue, move the declaration in a C file. vma.c looks like
the most appropriate candidate. We don't need to replace the definitions
in vsyscall.h by declarations as declarations are already in asm/vvar.h.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Performing SMP atomic operations on u64 fails on powerpc32:
CC drivers/char/random.o
In file included from <command-line>:
drivers/char/random.c: In function 'crng_reseed':
././include/linux/compiler_types.h:510:45: error: call to '__compiletime_assert_391' declared with attribute error: Need native word sized stores/loads for atomicity.
510 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^
././include/linux/compiler_types.h:491:25: note: in definition of macro '__compiletime_assert'
491 | prefix ## suffix(); \
| ^~~~~~
././include/linux/compiler_types.h:510:9: note: in expansion of macro '_compiletime_assert'
510 | _compiletime_assert(condition, msg, __compiletime_assert_, __COUNTER__)
| ^~~~~~~~~~~~~~~~~~~
././include/linux/compiler_types.h:513:9: note: in expansion of macro 'compiletime_assert'
513 | compiletime_assert(__native_word(t), \
| ^~~~~~~~~~~~~~~~~~
./arch/powerpc/include/asm/barrier.h:74:9: note: in expansion of macro 'compiletime_assert_atomic_type'
74 | compiletime_assert_atomic_type(*p); \
| ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
./include/asm-generic/barrier.h:172:55: note: in expansion of macro '__smp_store_release'
172 | #define smp_store_release(p, v) do { kcsan_release(); __smp_store_release(p, v); } while (0)
| ^~~~~~~~~~~~~~~~~~~
drivers/char/random.c:286:9: note: in expansion of macro 'smp_store_release'
286 | smp_store_release(&__arch_get_k_vdso_rng_data()->generation, next_gen + 1);
| ^~~~~~~~~~~~~~~~~
The kernel-side generation counter in the random driver is handled as an
unsigned long, not as a u64, in base_crng and struct crng.
But on the vDSO side, it needs to be an u64, not just an unsigned long,
in order to support a 32-bit vDSO atop a 64-bit kernel.
On kernel side, however, it is an unsigned long, hence a 32-bit value on
32-bit architectures, so just cast it to unsigned long for the
smp_store_release(). A side effect is that on big endian architectures
the store will be performed in the upper 32 bits. It is not an issue on
its own because the vDSO site doesn't mind the value, as it only checks
differences. Just make sure that the vDSO side checks the full 64 bits.
For that, the local current_generation has to be u64 as well.
Signed-off-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Suggested-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
const qualify the struct ctl_table argument in the proc_handler function
signatures. This is a prerequisite to moving the static ctl_table
structs into .rodata data which will ensure that proc_handler function
pointers cannot be modified.
This patch has been generated by the following coccinelle script:
```
virtual patch
@r1@
identifier ctl, write, buffer, lenp, ppos;
identifier func !~ "appldata_(timer|interval)_handler|sched_(rt|rr)_handler|rds_tcp_skbuf_handler|proc_sctp_do_(hmac_alg|rto_min|rto_max|udp_port|alpha_beta|auth|probe_interval)";
@@
int func(
- struct ctl_table *ctl
+ const struct ctl_table *ctl
,int write, void *buffer, size_t *lenp, loff_t *ppos);
@r2@
identifier func, ctl, write, buffer, lenp, ppos;
@@
int func(
- struct ctl_table *ctl
+ const struct ctl_table *ctl
,int write, void *buffer, size_t *lenp, loff_t *ppos)
{ ... }
@r3@
identifier func;
@@
int func(
- struct ctl_table *
+ const struct ctl_table *
,int , void *, size_t *, loff_t *);
@r4@
identifier func, ctl;
@@
int func(
- struct ctl_table *ctl
+ const struct ctl_table *ctl
,int , void *, size_t *, loff_t *);
@r5@
identifier func, write, buffer, lenp, ppos;
@@
int func(
- struct ctl_table *
+ const struct ctl_table *
,int write, void *buffer, size_t *lenp, loff_t *ppos);
```
* Code formatting was adjusted in xfs_sysctl.c to comply with code
conventions. The xfs_stats_clear_proc_handler,
xfs_panic_mask_proc_handler and xfs_deprecated_dointvec_minmax where
adjusted.
* The ctl_table argument in proc_watchdog_common was const qualified.
This is called from a proc_handler itself and is calling back into
another proc_handler, making it necessary to change it as part of the
proc_handler migration.
Co-developed-by: Thomas Weißschuh <linux@weissschuh.net>
Signed-off-by: Thomas Weißschuh <linux@weissschuh.net>
Co-developed-by: Joel Granados <j.granados@samsung.com>
Signed-off-by: Joel Granados <j.granados@samsung.com>
Provide a generic C vDSO getrandom() implementation, which operates on
an opaque state returned by vgetrandom_alloc() and produces random bytes
the same way as getrandom(). This has the following API signature:
ssize_t vgetrandom(void *buffer, size_t len, unsigned int flags,
void *opaque_state, size_t opaque_len);
The return value and the first three arguments are the same as ordinary
getrandom(), while the last two arguments are a pointer to the opaque
allocated state and its size. Were all five arguments passed to the
getrandom() syscall, nothing different would happen, and the functions
would have the exact same behavior.
The actual vDSO RNG algorithm implemented is the same one implemented by
drivers/char/random.c, using the same fast-erasure techniques as that.
Should the in-kernel implementation change, so too will the vDSO one.
It requires an implementation of ChaCha20 that does not use any stack,
in order to maintain forward secrecy if a multi-threaded program forks
(though this does not account for a similar issue with SA_SIGINFO
copying registers to the stack), so this is left as an
architecture-specific fill-in. Stack-less ChaCha20 is an easy algorithm
to implement on a variety of architectures, so this shouldn't be too
onerous.
Initially, the state is keyless, and so the first call makes a
getrandom() syscall to generate that key, and then uses it for
subsequent calls. By keeping track of a generation counter, it knows
when its key is invalidated and it should fetch a new one using the
syscall. Later, more than just a generation counter might be used.
Since MADV_WIPEONFORK is set on the opaque state, the key and related
state is wiped during a fork(), so secrets don't roll over into new
processes, and the same state doesn't accidentally generate the same
random stream. The generation counter, as well, is always >0, so that
the 0 counter is a useful indication of a fork() or otherwise
uninitialized state.
If the kernel RNG is not yet initialized, then the vDSO always calls the
syscall, because that behavior cannot be emulated in userspace, but
fortunately that state is short lived and only during early boot. If it
has been initialized, then there is no need to inspect the `flags`
argument, because the behavior does not change post-initialization
regardless of the `flags` value.
Since the opaque state passed to it is mutated, vDSO getrandom() is not
reentrant, when used with the same opaque state, which libc should be
mindful of.
The function works over an opaque per-thread state of a particular size,
which must be marked VM_WIPEONFORK, VM_DONTDUMP, VM_NORESERVE, and
VM_DROPPABLE for proper operation. Over time, the nuances of these
allocations may change or grow or even differ based on architectural
features.
The opaque state passed to vDSO getrandom() must be allocated using the
mmap_flags and mmap_prot parameters provided by the vgetrandom_opaque_params
struct, which also contains the size of each state. That struct can be
obtained with a call to vgetrandom(NULL, 0, 0, ¶ms, ~0UL). Then,
libc can call mmap(2) and slice up the returned array into a state per
each thread, while ensuring that no single state straddles a page
boundary. Libc is expected to allocate a chunk of these on first use,
and then dole them out to threads as they're created, allocating more
when needed.
vDSO getrandom() provides the ability for userspace to generate random
bytes quickly and safely, and is intended to be integrated into libc's
thread management. As an illustrative example, the introduced code in
the vdso_test_getrandom self test later in this series might be used to
do the same outside of libc. In a libc the various pthread-isms are
expected to be elided into libc internals.
Reviewed-by: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
The entropy accounting changes a static key when the RNG has
initialized, since it only ever initializes once. Static key changes,
however, cannot be made from atomic context, so depending on where the
last creditable entropy comes from, the static key change might need to
be deferred to a worker.
Previously the code used the execute_in_process_context() helper
function, which accounts for whether or not the caller is
in_interrupt(). However, that doesn't account for the case where the
caller is actually in process context but is holding a spinlock.
This turned out to be the case with input_handle_event() in
drivers/input/input.c contributing entropy:
[<ffffffd613025ba0>] die+0xa8/0x2fc
[<ffffffd613027428>] bug_handler+0x44/0xec
[<ffffffd613016964>] brk_handler+0x90/0x144
[<ffffffd613041e58>] do_debug_exception+0xa0/0x148
[<ffffffd61400c208>] el1_dbg+0x60/0x7c
[<ffffffd61400c000>] el1h_64_sync_handler+0x38/0x90
[<ffffffd613011294>] el1h_64_sync+0x64/0x6c
[<ffffffd613102d88>] __might_resched+0x1fc/0x2e8
[<ffffffd613102b54>] __might_sleep+0x44/0x7c
[<ffffffd6130b6eac>] cpus_read_lock+0x1c/0xec
[<ffffffd6132c2820>] static_key_enable+0x14/0x38
[<ffffffd61400ac08>] crng_set_ready+0x14/0x28
[<ffffffd6130df4dc>] execute_in_process_context+0xb8/0xf8
[<ffffffd61400ab30>] _credit_init_bits+0x118/0x1dc
[<ffffffd6138580c8>] add_timer_randomness+0x264/0x270
[<ffffffd613857e54>] add_input_randomness+0x38/0x48
[<ffffffd613a80f94>] input_handle_event+0x2b8/0x490
[<ffffffd613a81310>] input_event+0x6c/0x98
According to Guoyong, it's not really possible to refactor the various
drivers to never hold a spinlock there. And in_atomic() isn't reliable.
So, rather than trying to be too fancy, just punt the change in the
static key to a workqueue always. There's basically no drawback of doing
this, as the code already needed to account for the static key not
changing immediately, and given that it's just an optimization, there's
not exactly a hurry to change the static key right away, so deferal is
fine.
Reported-by: Guoyong Wang <guoyong.wang@mediatek.com>
Cc: stable@vger.kernel.org
Fixes: f5bda35fba ("random: use static branch for crng_ready()")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
With the removal of the 'iov' argument to import_single_range(), the two
functions are now fully identical. Convert the import_single_range()
callers to import_ubuf(), and remove the former fully.
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Link: https://lore.kernel.org/r/20231204174827.1258875-3-axboe@kernel.dk
Signed-off-by: Christian Brauner <brauner@kernel.org>
This commit comes at the tail end of a greater effort to remove the
empty elements at the end of the ctl_table arrays (sentinels) which
will reduce the overall build time size of the kernel and run time
memory bloat by ~64 bytes per sentinel (further information Link :
https://lore.kernel.org/all/ZO5Yx5JFogGi%2FcBo@bombadil.infradead.org/)
Remove sentinel from impi_table and random_table
Signed-off-by: Joel Granados <j.granados@samsung.com>
Signed-off-by: Luis Chamberlain <mcgrof@kernel.org>
Use copy_splice_read() for tty, procfs, kernfs and random files rather
than going through generic_file_splice_read() as they just copy the file
into the output buffer and don't splice pages. This avoids the need for
them to have a ->read_folio() to satisfy filemap_splice_read().
Signed-off-by: David Howells <dhowells@redhat.com>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
cc: Christoph Hellwig <hch@lst.de>
cc: Jens Axboe <axboe@kernel.dk>
cc: Al Viro <viro@zeniv.linux.org.uk>
cc: John Hubbard <jhubbard@nvidia.com>
cc: David Hildenbrand <david@redhat.com>
cc: Matthew Wilcox <willy@infradead.org>
cc: Miklos Szeredi <miklos@szeredi.hu>
cc: Arnd Bergmann <arnd@arndb.de>
cc: linux-block@vger.kernel.org
cc: linux-fsdevel@vger.kernel.org
cc: linux-mm@kvack.org
Link: https://lore.kernel.org/r/20230522135018.2742245-13-dhowells@redhat.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
It turns out that commit 596ff4a09b ("cpumask: re-introduce
constant-sized cpumask optimizations") exposed a number of cases of
drivers not checking the result of "cpumask_next()" and friends
correctly.
The documented correct check for "no more cpus in the cpumask" is to
check for the result being equal or larger than the number of possible
CPU ids, exactly _because_ we've always done those constant-sized
cpumask scans using a widened type before. So the return value of a
cpumask scan should be checked with
if (cpu >= nr_cpu_ids)
...
because the cpumask scan did not necessarily stop exactly *at* that
maximum CPU id.
But a few cases ended up instead using checks like
if (cpu == nr_cpumask_bits)
...
which used that internal "widened" number of bits. And that used to
work pretty much by accident (ok, in this case "by accident" is simply
because it matched the historical internal implementation of the cpumask
scanning, so it was more of a "intentionally using implementation
details rather than an accident").
But the extended constant-sized optimizations then did that internal
implementation differently, and now that code that did things wrong but
matched the old implementation no longer worked at all.
Which then causes subsequent odd problems due to using what ends up
being an invalid CPU ID.
Most of these cases require either unusual hardware or special uses to
hit, but the random.c one triggers quite easily.
All you really need is to have a sufficiently small CONFIG_NR_CPUS value
for the bit scanning optimization to be triggered, but not enough CPUs
to then actually fill that widened cpumask. At that point, the cpumask
scanning will return the NR_CPUS constant, which is _not_ the same as
nr_cpumask_bits.
This just does the mindless fix with
sed -i 's/== nr_cpumask_bits/>= nr_cpu_ids/'
to fix the incorrect uses.
The ones in the SCSI lpfc driver in particular could probably be fixed
more cleanly by just removing that repeated pattern entirely, but I am
not emptionally invested enough in that driver to care.
Reported-and-tested-by: Guenter Roeck <linux@roeck-us.net>
Link: https://lore.kernel.org/lkml/481b19b5-83a0-4793-b4fd-194ad7b978c3@roeck-us.net/
Reported-and-tested-by: Geert Uytterhoeven <geert+renesas@glider.be>
Link: https://lore.kernel.org/lkml/CAMuHMdUKo_Sf7TjKzcNDa8Ve+6QrK+P8nSQrSQ=6LTRmcBKNww@mail.gmail.com/
Reported-by: Vernon Yang <vernon2gm@gmail.com>
Link: https://lore.kernel.org/lkml/20230306160651.2016767-1-vernon2gm@gmail.com/
Cc: Yury Norov <yury.norov@gmail.com>
Cc: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
The <asm/archrandom.h> header is a random.c private detail, not
something to be called by other code. As such, don't make it
automatically available by way of random.h.
Cc: Michael Ellerman <mpe@ellerman.id.au>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Reviewed-by: Christophe Leroy <christophe.leroy@csgroup.eu>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
direction misannotations and (hopefully) preventing
more of the same for the future.
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
-----BEGIN PGP SIGNATURE-----
iHQEABYIAB0WIQQqUNBr3gm4hGXdBJlZ7Krx/gZQ6wUCY5ZzQAAKCRBZ7Krx/gZQ
65RZAP4nTkvOn0NZLVFkuGOx8pgJelXAvrteyAuecVL8V6CR4AD40qCVY51PJp8N
MzwiRTeqnGDxTTF7mgd//IB6hoatAA==
=bcvF
-----END PGP SIGNATURE-----
Merge tag 'pull-iov_iter' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull iov_iter updates from Al Viro:
"iov_iter work; most of that is about getting rid of direction
misannotations and (hopefully) preventing more of the same for the
future"
* tag 'pull-iov_iter' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
use less confusing names for iov_iter direction initializers
iov_iter: saner checks for attempt to copy to/from iterator
[xen] fix "direction" argument of iov_iter_kvec()
[vhost] fix 'direction' argument of iov_iter_{init,bvec}()
[target] fix iov_iter_bvec() "direction" argument
[s390] memcpy_real(): WRITE is "data source", not destination...
[s390] zcore: WRITE is "data source", not destination...
[infiniband] READ is "data destination", not source...
[fsi] WRITE is "data source", not destination...
[s390] copy_oldmem_kernel() - WRITE is "data source", not destination
csum_and_copy_to_iter(): handle ITER_DISCARD
get rid of unlikely() on page_copy_sane() calls
The theory behind the jitter dance is that multiple things are poking at
the same cache line. This only works, however, if what's being poked at
is actually all in the same cache line. Ensure this is the case by
aligning the struct on the stack to the cache line size.
We can't use ____cacheline_aligned on a stack variable, because gcc
assumes 16 byte alignment when only 8 byte alignment is provided by the
kernel, which means gcc could technically do something pathological
like `(rsp & ~48) - 64`. It doesn't, but rather than risk it, just do
the stack alignment manually with PTR_ALIGN and an oversized buffer.
Fixes: 50ee7529ec ("random: try to actively add entropy rather than passively wait for it")
Cc: Eric Biggers <ebiggers@kernel.org>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Rather than just relying on interaction between cache lines of the timer
and the main loop, also explicitly take into account the fact that the
timer might fire at some time that's hard to predict, due to scheduling,
interrupts, or cross-CPU conditions. Mix in a cycle counter during the
firing of the timer, in addition to the existing one during the
scheduling of the timer. It can't hurt and can only help.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Rather than merely hoping that the callback gets called on another CPU,
arrange for that to actually happen, by round robining which CPU the
timer fires on. This way, on multiprocessor machines, we exacerbate
jitter by touching the same memory from multiple different cores.
There's a little bit of tricky bookkeeping involved here, because using
timer_setup_on_stack() + add_timer_on() + del_timer_sync() will result
in a use after free. See this sample code: <https://xn--4db.cc/xBdEiIKO/c>.
Instead, it's necessary to call [try_to_]del_timer_sync() before calling
add_timer_on(), so that the final call to del_timer_sync() at the end of
the function actually succeeds at making sure no handlers are running.
Cc: Sultan Alsawaf <sultan@kerneltoast.com>
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
READ/WRITE proved to be actively confusing - the meanings are
"data destination, as used with read(2)" and "data source, as
used with write(2)", but people keep interpreting those as
"we read data from it" and "we write data to it", i.e. exactly
the wrong way.
Call them ITER_DEST and ITER_SOURCE - at least that is harder
to misinterpret...
Signed-off-by: Al Viro <viro@zeniv.linux.org.uk>
This is required by vsprint, because it can't do things synchronously
from hardirq context, and it will be useful for an EFI notifier as well.
I didn't initially want to do this, but with two potential consumers
now, it seems worth it.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Currently, we reseed when random bytes are requested, if the current
seed is too old. Since random bytes can be requested from all contexts,
including hard IRQ, this means sometimes we wind up adding a bit of
latency to hard IRQ. This was so much of a problem on s390x that now
s390x just doesn't provide its architectural RNG from hard IRQ context,
so we miss out in that case.
Instead, let's just schedule a persistent delayed work, so that the
reseeding and potentially expensive operations will always happen from
process context, reducing unexpected latencies from hard IRQ.
This also has the nice effect of accumulating a transcript of random
inputs over time, since it means that we amass more input values. And it
should make future vDSO integration a bit easier.
Cc: Harald Freudenberger <freude@linux.ibm.com>
Cc: Juergen Christ <jchrist@linux.ibm.com>
Cc: Alexander Gordeev <agordeev@linux.ibm.com>
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Tejun Heo <tj@kernel.org>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Rather than calling add_device_randomness(), the add_early_randomness()
function should use add_hwgenerator_randomness(), so that the early
entropy can be potentially credited, which allows for the RNG to
initialize earlier without having to wait for the kthread to come up.
This requires some minor API refactoring, by adding a `sleep_after`
parameter to add_hwgenerator_randomness(), so that we don't hit a
blocking sleep from add_early_randomness().
Tested-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Tested-by: Marek Szyprowski <m.szyprowski@samsung.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Reviewed-by: Dominik Brodowski <linux@dominikbrodowski.net>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
The prior text was very old and made outdated references to TCP sequence
numbers, which should use one of the integer functions instead, since
batched entropy was introduced. The current way of describing the
quality of functions is just to say that it's as good as /dev/urandom,
which now all the functions are.
Fixes: f5b98461cb ("random: use chacha20 for get_random_int/long")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Since de492c83ca ("prandom: remove unused functions"),
get_random_int() no longer exists, so remove its reference from this
comment.
Fixes: de492c83ca ("prandom: remove unused functions")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
The arch_get_random*_early() abstraction is not completely useful and
adds complexity, because it's not a given that there will be no calls to
arch_get_random*() between random_init_early(), which uses
arch_get_random*_early(), and init_cpu_features(). During that gap,
crng_reseed() might be called, which uses arch_get_random*(), since it's
mostly not init code.
Instead we can test whether we're in the early phase in
arch_get_random*() itself, and in doing so avoid all ambiguity about
where we are. Fortunately, the only architecture that currently
implements arch_get_random*_early() also has an alternatives-based cpu
feature system, one flag of which determines whether the other flags
have been initialized. This makes it possible to do the early check with
zero cost once the system is initialized.
Reviewed-by: Catalin Marinas <catalin.marinas@arm.com>
Cc: Will Deacon <will@kernel.org>
Cc: Ard Biesheuvel <ardb@kernel.org>
Cc: Jean-Philippe Brucker <jean-philippe@linaro.org>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
It's very unusual to have both a command line option and a compile time
option, and apparently that's confusing to people. Also, basically
everybody enables the compile time option now, which means people who
want to disable this wind up having to use the command line option to
ensure that anyway. So just reduce the number of moving pieces and nix
the compile time option in favor of the more versatile command line
option.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Now that we have get_random_u32_below(), it's nearly trivial to make
inline helpers to compute get_random_u32_above() and
get_random_u32_inclusive(), which will help clean up open coded loops
and manual computations throughout the tree.
One snag is that in order to make get_random_u32_inclusive() operate on
closed intervals, we have to do some (unlikely) special case handling if
get_random_u32_inclusive(0, U32_MAX) is called. The least expensive way
of doing this is actually to adjust the slowpath of
get_random_u32_below() to have its undefined 0 result just return the
output of get_random_u32(). We can make this basically free by calling
get_random_u32() before the branch, so that the branch latency gets
interleaved.
Cc: stable@vger.kernel.org # to ease future backports that use this api
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Until the very recent commits, many bounded random integers were
calculated using `get_random_u32() % max_plus_one`, which not only
incurs the price of a division -- indicating performance mostly was not
a real issue -- but also does not result in a uniformly distributed
output if max_plus_one is not a power of two. Recent commits moved to
using `prandom_u32_max(max_plus_one)`, which replaces the division with
a faster multiplication, but still does not solve the issue with
non-uniform output.
For some users, maybe this isn't a problem, and for others, maybe it is,
but for the majority of users, probably the question has never been
posed and analyzed, and nobody thought much about it, probably assuming
random is random is random. In other words, the unthinking expectation
of most users is likely that the resultant numbers are uniform.
So we implement here an efficient way of generating uniform bounded
random integers. Through use of compile-time evaluation, and avoiding
divisions as much as possible, this commit introduces no measurable
overhead. At least for hot-path uses tested, any potential difference
was lost in the noise. On both clang and gcc, code generation is pretty
small.
The new function, get_random_u32_below(), lives in random.h, rather than
prandom.h, and has a "get_random_xxx" function name, because it is
suitable for all uses, including cryptography.
In order to be efficient, we implement a kernel-specific variant of
Daniel Lemire's algorithm from "Fast Random Integer Generation in an
Interval", linked below. The kernel's variant takes advantage of
constant folding to avoid divisions entirely in the vast majority of
cases, works on both 32-bit and 64-bit architectures, and requests a
minimal amount of bytes from the RNG.
Link: https://arxiv.org/pdf/1805.10941.pdf
Cc: stable@vger.kernel.org # to ease future backports that use this api
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
While reworking the archrandom handling, commit d349ab99ee ("random:
handle archrandom with multiple longs") switched to the non-early
archrandom helpers in random_init(), which broke initialization of the
entropy pool from the arm64 random generator.
Indeed at that point the arm64 CPU features, which verify that all CPUs
have compatible capabilities, are not finalized so arch_get_random_seed_longs()
is unsuccessful. Instead random_init() should use the _early functions,
which check only the boot CPU on arm64. On other architectures the
_early functions directly call the normal ones.
Fixes: d349ab99ee ("random: handle archrandom with multiple longs")
Cc: stable@vger.kernel.org
Signed-off-by: Jean-Philippe Brucker <jean-philippe@linaro.org>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
With no callers left of prandom_u32() and prandom_bytes(), as well as
get_random_int(), remove these deprecated wrappers, in favor of
get_random_u32() and get_random_bytes().
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Yury Norov <yury.norov@gmail.com>
Acked-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
The commit that added the new get_random_{u8,u16}() functions neglected
to update the code that clears the batches when bringing up a new CPU.
It also forgot a few comments and helper defines, so add those in too.
Fixes: 585cd5fe9f ("random: add 8-bit and 16-bit batches")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Remove extra whitespace and add a missing word to a sentence describing
get_random_bytes().
Signed-off-by: William Zijl <postmaster@gusted.xyz>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Counterintuitively, mod_timer(..., jiffies + 1) will cause the timer to
fire not in the next jiffy, but in two jiffies. The way to cause
the timer to fire in the next jiffy is with mod_timer(..., jiffies).
Doing so then lets us bump the upper bound back up again.
Fixes: 50ee7529ec ("random: try to actively add entropy rather than passively wait for it")
Fixes: 829d680e82 ("random: cap jitter samples per bit to factor of HZ")
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Sultan Alsawaf <sultan@kerneltoast.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
There are numerous places in the kernel that would be sped up by having
smaller batches. Currently those callsites do `get_random_u32() & 0xff`
or similar. Since these are pretty spread out, and will require patches
to multiple different trees, let's get ahead of the curve and lay the
foundation for `get_random_u8()` and `get_random_u16()`, so that it's
then possible to start submitting conversion patches leisurely.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Rather than going through the current-> indirection for utsname, at this
point in boot, init_utsname()==utsname(), so just use it directly that
way. Additionally, init_utsname() appears to be available nearly always,
so move it into random_init_early().
Suggested-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
The full RNG initialization relies on some timestamps, made possible
with initialization functions like time_init() and timekeeping_init().
However, these are only available rather late in initialization.
Meanwhile, other things, such as memory allocator functions, make use of
the RNG much earlier.
So split RNG initialization into two phases. We can provide arch
randomness very early on, and then later, after timekeeping and such are
available, initialize the rest.
This ensures that, for example, slabs are properly randomized if RDRAND
is available. Without this, CONFIG_SLAB_FREELIST_RANDOM=y loses a degree
of its security, because its random seed is potentially deterministic,
since it hasn't yet incorporated RDRAND. It also makes it possible to
use a better seed in kfence, which currently relies on only the cycle
counter.
Another positive consequence is that on systems with RDRAND, running
with CONFIG_WARN_ALL_UNSEEDED_RANDOM=y results in no warnings at all.
One subtle side effect of this change is that on systems with no RDRAND,
RDTSC is now only queried by random_init() once, committing the moment
of the function call, instead of multiple times as before. This is
intentional, as the multiple RDTSCs in a loop before weren't
accomplishing very much, with jitter being better provided by
try_to_generate_entropy(). Plus, filling blocks with RDTSC is still
being done in extract_entropy(), which is necessarily called before
random bytes are served anyway.
Cc: Andrew Morton <akpm@linux-foundation.org>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Previously, the fast pool was dumped into the main pool periodically in
the fast pool's hard IRQ handler. This worked fine and there weren't
problems with it, until RT came around. Since RT converts spinlocks into
sleeping locks, problems cropped up. Rather than switching to raw
spinlocks, the RT developers preferred we make the transformation from
originally doing:
do_some_stuff()
spin_lock()
do_some_other_stuff()
spin_unlock()
to doing:
do_some_stuff()
queue_work_on(some_other_stuff_worker)
This is an ordinary pattern done all over the kernel. However, Sherry
noticed a 10% performance regression in qperf TCP over a 40gbps
InfiniBand card. Quoting her message:
> MT27500 Family [ConnectX-3] cards:
> Infiniband device 'mlx4_0' port 1 status:
> default gid: fe80:0000:0000:0000:0010:e000:0178:9eb1
> base lid: 0x6
> sm lid: 0x1
> state: 4: ACTIVE
> phys state: 5: LinkUp
> rate: 40 Gb/sec (4X QDR)
> link_layer: InfiniBand
>
> Cards are configured with IP addresses on private subnet for IPoIB
> performance testing.
> Regression identified in this bug is in TCP latency in this stack as reported
> by qperf tcp_lat metric:
>
> We have one system listen as a qperf server:
> [root@yourQperfServer ~]# qperf
>
> Have the other system connect to qperf server as a client (in this
> case, it’s X7 server with Mellanox card):
> [root@yourQperfClient ~]# numactl -m0 -N0 qperf 20.20.20.101 -v -uu -ub --time 60 --wait_server 20 -oo msg_size:4K:1024K:*2 tcp_lat
Rather than incur the scheduling latency from queue_work_on, we can
instead switch to running on the next timer tick, on the same core. This
also batches things a bit more -- once per jiffy -- which is okay now
that mix_interrupt_randomness() can credit multiple bits at once.
Reported-by: Sherry Yang <sherry.yang@oracle.com>
Tested-by: Paul Webb <paul.x.webb@oracle.com>
Cc: Sherry Yang <sherry.yang@oracle.com>
Cc: Phillip Goerl <phillip.goerl@oracle.com>
Cc: Jack Vogel <jack.vogel@oracle.com>
Cc: Nicky Veitch <nicky.veitch@oracle.com>
Cc: Colm Harrington <colm.harrington@oracle.com>
Cc: Ramanan Govindarajan <ramanan.govindarajan@oracle.com>
Cc: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Cc: Tejun Heo <tj@kernel.org>
Cc: Sultan Alsawaf <sultan@kerneltoast.com>
Cc: stable@vger.kernel.org
Fixes: 58340f8e95 ("random: defer fast pool mixing to worker")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
In order to avoid reading and dirtying two cache lines on every IRQ,
move the work_struct to the bottom of the fast_pool struct. add_
interrupt_randomness() always touches .pool and .count, which are
currently split, because .mix pushes everything down. Instead, move .mix
to the bottom, so that .pool and .count are always in the first cache
line, since .mix is only accessed when the pool is full.
Fixes: 58340f8e95 ("random: defer fast pool mixing to worker")
Reviewed-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Since the most that's mixed into the pool is sizeof(long)*2, don't
credit more than that many bytes of entropy.
Fixes: e3e33fc2ea ("random: do not use input pool from hard IRQs")
Cc: stable@vger.kernel.org
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
If a hwrng source does not provide an entropy estimate, it currently
does not contribute at all to the CRNG. In order to help fix this, in
case add_hwgenerator_randomness() is called with the entropy parameter
set to zero, go to sleep until one reseed interval has passed.
While the hwrng thread currently only runs under conditions where this
is non-zero, this change is not harmful and prepares for future updates
to the hwrng core.
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Reviewed-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Mix in randomness from hw-rng sources more frequently during early
boot, approximately once for every rng reseed.
Signed-off-by: Dominik Brodowski <linux@dominikbrodowski.net>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Prior to 5.6, when /dev/random was opened with O_NONBLOCK, it would
return -EAGAIN if there was no entropy. When the pools were unified in
5.6, this was lost. The post 5.6 behavior of blocking until the pool is
initialized, and ignoring O_NONBLOCK in the process, went unnoticed,
with no reports about the regression received for two and a half years.
However, eventually this indeed did break somebody's userspace.
So we restore the old behavior, by returning -EAGAIN if the pool is not
initialized. Unlike the old /dev/random, this can only occur during
early boot, after which it never blocks again.
In order to make this O_NONBLOCK behavior consistent with other
expectations, also respect users reading with preadv2(RWF_NOWAIT) and
similar.
Fixes: 30c08efec8 ("random: make /dev/random be almost like /dev/urandom")
Reported-by: Guozihua <guozihua@huawei.com>
Reported-by: Zhongguohua <zhongguohua1@huawei.com>
Cc: Al Viro <viro@zeniv.linux.org.uk>
Cc: Theodore Ts'o <tytso@mit.edu>
Cc: Andrew Lutomirski <luto@kernel.org>
Cc: stable@vger.kernel.org
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Here is the large set of char and misc and other driver subsystem
changes for 6.0-rc1.
Highlights include:
- large set of IIO driver updates, additions, and cleanups
- new habanalabs device support added (loads of register maps
much like GPUs have)
- soundwire driver updates
- phy driver updates
- slimbus driver updates
- tiny virt driver fixes and updates
- misc driver fixes and updates
- interconnect driver updates
- hwtracing driver updates
- fpga driver updates
- extcon driver updates
- firmware driver updates
- counter driver update
- mhi driver fixes and updates
- binder driver fixes and updates
- speakup driver fixes
Full details are in the long shortlog contents.
All of these have been in linux-next for a while without any reported
problems.
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
-----BEGIN PGP SIGNATURE-----
iG0EABECAC0WIQT0tgzFv3jCIUoxPcsxR9QN2y37KQUCYup9QQ8cZ3JlZ0Brcm9h
aC5jb20ACgkQMUfUDdst+ylBKQCfaSuzl9ZP9dTvAw2FPp14oRqXnpoAnicvWAoq
1vU9Vtq2c73uBVLdZm4m
=AwP3
-----END PGP SIGNATURE-----
Merge tag 'char-misc-6.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc
Pull char / misc driver updates from Greg KH:
"Here is the large set of char and misc and other driver subsystem
changes for 6.0-rc1.
Highlights include:
- large set of IIO driver updates, additions, and cleanups
- new habanalabs device support added (loads of register maps much
like GPUs have)
- soundwire driver updates
- phy driver updates
- slimbus driver updates
- tiny virt driver fixes and updates
- misc driver fixes and updates
- interconnect driver updates
- hwtracing driver updates
- fpga driver updates
- extcon driver updates
- firmware driver updates
- counter driver update
- mhi driver fixes and updates
- binder driver fixes and updates
- speakup driver fixes
All of these have been in linux-next for a while without any reported
problems"
* tag 'char-misc-6.0-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc: (634 commits)
drivers: lkdtm: fix clang -Wformat warning
char: remove VR41XX related char driver
misc: Mark MICROCODE_MINOR unused
spmi: trace: fix stack-out-of-bound access in SPMI tracing functions
dt-bindings: iio: adc: Add compatible for MT8188
iio: light: isl29028: Fix the warning in isl29028_remove()
iio: accel: sca3300: Extend the trigger buffer from 16 to 32 bytes
iio: fix iio_format_avail_range() printing for none IIO_VAL_INT
iio: adc: max1027: unlock on error path in max1027_read_single_value()
iio: proximity: sx9324: add empty line in front of bullet list
iio: magnetometer: hmc5843: Remove duplicate 'the'
iio: magn: yas530: Use DEFINE_RUNTIME_DEV_PM_OPS() and pm_ptr() macros
iio: magnetometer: ak8974: Use DEFINE_RUNTIME_DEV_PM_OPS() and pm_ptr() macros
iio: light: veml6030: Use DEFINE_RUNTIME_DEV_PM_OPS() and pm_ptr() macros
iio: light: vcnl4035: Use DEFINE_RUNTIME_DEV_PM_OPS() and pm_ptr() macros
iio: light: vcnl4000: Use DEFINE_RUNTIME_DEV_PM_OPS() and pm_ptr() macros
iio: light: tsl2591: Use DEFINE_RUNTIME_DEV_PM_OPS() and pm_ptr()
iio: light: tsl2583: Use DEFINE_RUNTIME_DEV_PM_OPS and pm_ptr()
iio: light: isl29028: Use DEFINE_RUNTIME_DEV_PM_OPS() and pm_ptr()
iio: light: gp2ap002: Switch to DEFINE_RUNTIME_DEV_PM_OPS and pm_ptr()
...
It was missing an 'r'.
Fixes: 186873c549 ("random: use simpler fast key erasure flow on per-cpu keys")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
The archrandom interface was originally designed for x86, which supplies
RDRAND/RDSEED for receiving random words into registers, resulting in
one function to generate an int and another to generate a long. However,
other architectures don't follow this.
On arm64, the SMCCC TRNG interface can return between one and three
longs. On s390, the CPACF TRNG interface can return arbitrary amounts,
with four longs having the same cost as one. On UML, the os_getrandom()
interface can return arbitrary amounts.
So change the api signature to take a "max_longs" parameter designating
the maximum number of longs requested, and then return the number of
longs generated.
Since callers need to check this return value and loop anyway, each arch
implementation does not bother implementing its own loop to try again to
fill the maximum number of longs. Additionally, all existing callers
pass in a constant max_longs parameter. Taken together, these two things
mean that the codegen doesn't really change much for one-word-at-a-time
platforms, while performance is greatly improved on platforms such as
s390.
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Acked-by: Catalin Marinas <catalin.marinas@arm.com>
Acked-by: Mark Rutland <mark.rutland@arm.com>
Acked-by: Michael Ellerman <mpe@ellerman.id.au>
Acked-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Use `!try_cmpxchg(ptr, &orig, new)` instead of `cmpxchg(ptr, orig, new)
!= orig` in _credit_init_bits. This has two benefits:
- The x86 cmpxchg instruction returns success in the ZF flag, so this
change saves a compare after cmpxchg, as well as a related move
instruction in front of cmpxchg.
- try_cmpxchg implicitly assigns the *ptr value to &orig when cmpxchg
fails, enabling further code simplifications.
This patch has no functional change.
Signed-off-by: Uros Bizjak <ubizjak@gmail.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Currently the jitter mechanism will require two timer ticks per
iteration, and it requires N iterations per bit. This N is determined
with a small measurement, and if it's too big, it won't waste time with
jitter entropy because it'd take too long or not have sufficient entropy
anyway.
With the current max N of 32, there are large timeouts on systems with a
small CONFIG_HZ. Rather than set that maximum to 32, instead choose a
factor of CONFIG_HZ. In this case, 1/30 seems to yield sane values for
different configurations of CONFIG_HZ.
Reported-by: Vladimir Murzin <vladimir.murzin@arm.com>
Fixes: 78c768e619 ("random: vary jitter iterations based on cycle counter speed")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Tested-by: Vladimir Murzin <vladimir.murzin@arm.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
Systems that initiate frequent suspend/resume from userspace
can make the kernel aware by enabling PM_USERSPACE_AUTOSLEEP
config.
This allows for certain sleep-sensitive code (wireguard/rng) to
decide on what preparatory work should be performed (or not) in
their pm_notification callbacks.
This patch was prompted by the discussion at [1] which attempts
to remove CONFIG_ANDROID that currently guards these code paths.
[1] https://lore.kernel.org/r/20220629150102.1582425-1-hch@lst.de/
Suggested-by: Jason A. Donenfeld <Jason@zx2c4.com>
Acked-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Kalesh Singh <kaleshsingh@google.com>
Link: https://lore.kernel.org/r/20220630191230.235306-1-kaleshsingh@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
This comment wasn't updated when we moved from read() to read_iter(), so
this patch makes the trivial fix.
Fixes: 1b388e7765 ("random: convert to using fops->read_iter()")
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
random.c ratelimits how much it warns about uninitialized urandom reads
using __ratelimit(). When the RNG is finally initialized, it prints the
number of missed messages due to ratelimiting.
It has been this way since that functionality was introduced back in
2018. Recently, cc1e127bfa ("random: remove ratelimiting for in-kernel
unseeded randomness") put a bit more stress on the urandom ratelimiting,
which teased out a bug in the implementation.
Specifically, when under pressure, __ratelimit() will print its own
message and reset the count back to 0, making the final message at the
end less useful. Secondly, it does so as a pr_warn(), which apparently
is undesirable for people's CI.
Fortunately, __ratelimit() has the RATELIMIT_MSG_ON_RELEASE flag exactly
for this purpose, so we set the flag.
Fixes: 4e00b339e2 ("random: rate limit unseeded randomness warnings")
Cc: stable@vger.kernel.org
Reported-by: Jon Hunter <jonathanh@nvidia.com>
Reported-by: Ron Economos <re@w6rz.net>
Tested-by: Ron Economos <re@w6rz.net>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
It used to be that mix_interrupt_randomness() would credit 1 bit each
time it ran, and so add_interrupt_randomness() would schedule mix() to
run every 64 interrupts, a fairly arbitrary number, but nonetheless
considered to be a decent enough conservative estimate.
Since e3e33fc2ea ("random: do not use input pool from hard IRQs"),
mix() is now able to credit multiple bits, depending on the number of
calls to add(). This was done for reasons separate from this commit, but
it has the nice side effect of enabling this patch to schedule mix()
less often.
Currently the rules are:
a) Credit 1 bit for every 64 calls to add().
b) Schedule mix() once a second that add() is called.
c) Schedule mix() once every 64 calls to add().
Rules (a) and (c) no longer need to be coupled. It's still important to
have _some_ value in (c), so that we don't "over-saturate" the fast
pool, but the once per second we get from rule (b) is a plenty enough
baseline. So, by increasing the 64 in rule (c) to something larger, we
avoid calling queue_work_on() as frequently during irq storms.
This commit changes that 64 in rule (c) to be 1024, which means we
schedule mix() 16 times less often. And it does *not* need to change the
64 in rule (a).
Fixes: 58340f8e95 ("random: defer fast pool mixing to worker")
Cc: stable@vger.kernel.org
Cc: Dominik Brodowski <linux@dominikbrodowski.net>
Acked-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
With arch randomness being used by every distro and enabled in
defconfigs, the distinction between rng_has_arch_random() and
rng_is_initialized() is now rather small. In fact, the places where they
differ are now places where paranoid users and system builders really
don't want arch randomness to be used, in which case we should respect
that choice, or places where arch randomness is known to be broken, in
which case that choice is all the more important. So this commit just
removes the function and its one user.
Reviewed-by: Petr Mladek <pmladek@suse.com> # for vsprintf.c
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
