| Age | Commit message (Collapse) | Author | Files | Lines |
|---|
|
Initialization happens once -- by way of credit_init_bits() -- and then
it never happens again. Therefore, it doesn't need to be in
crng_reseed(), which is a hot path that is called multiple times. It
also doesn't make sense to have there, as initialization activity is
better associated with initialization routines.
After the prior commit, crng_reseed() now won't be called by multiple
concurrent callers, which means that we can safely move the
"finialize_init" logic into crng_init_bits() unconditionally.
Reviewed-by: Dominik Brodowski <[email protected]>
Signed-off-by: Jason A. Donenfeld <[email protected]>
|
|
Since all changes of crng_init now go through credit_init_bits(), we can
fix a long standing race in which two concurrent callers of
credit_init_bits() have the new bit count >= some threshold, but are
doing so with crng_init as a lower threshold, checked outside of a lock,
resulting in crng_reseed() or similar being called twice.
In order to fix this, we can use the original cmpxchg value of the bit
count, and only change crng_init when the bit count transitions from
below a threshold to meeting the threshold.
Reviewed-by: Dominik Brodowski <[email protected]>
Signed-off-by: Jason A. Donenfeld <[email protected]>
|
|
crng_init represents a state machine, with three states, and various
rules for transitions. For the longest time, we've been managing these
with "0", "1", and "2", and expecting people to figure it out. To make
the code more obvious, replace these with proper enum values
representing the transition, and then redocument what each of these
states mean.
Reviewed-by: Dominik Brodowski <[email protected]>
Cc: Joe Perches <[email protected]>
Signed-off-by: Jason A. Donenfeld <[email protected]>
|
|
random32.c has two random number generators in it: one that is meant to
be used deterministically, with some predefined seed, and one that does
the same exact thing as random.c, except does it poorly. The first one
has some use cases. The second one no longer does and can be replaced
with calls to random.c's proper random number generator.
The relatively recent siphash-based bad random32.c code was added in
response to concerns that the prior random32.c was too deterministic.
Out of fears that random.c was (at the time) too slow, this code was
anonymously contributed. Then out of that emerged a kind of shadow
entropy gathering system, with its own tentacles throughout various net
code, added willy nilly.
Stop👏making👏bespoke👏random👏number👏generators👏.
Fortunately, recent advances in random.c mean that we can stop playing
with this sketchiness, and just use get_random_u32(), which is now fast
enough. In micro benchmarks using RDPMC, I'm seeing the same median
cycle count between the two functions, with the mean being _slightly_
higher due to batches refilling (which we can optimize further need be).
However, when doing *real* benchmarks of the net functions that actually
use these random numbers, the mean cycles actually *decreased* slightly
(with the median still staying the same), likely because the additional
prandom code means icache misses and complexity, whereas random.c is
generally already being used by something else nearby.
The biggest benefit of this is that there are many users of prandom who
probably should be using cryptographically secure random numbers. This
makes all of those accidental cases become secure by just flipping a
switch. Later on, we can do a tree-wide cleanup to remove the static
inline wrapper functions that this commit adds.
There are also some low-ish hanging fruits for making this even faster
in the future: a get_random_u16() function for use in the networking
stack will give a 2x performance boost there, using SIMD for ChaCha20
will let us compute 4 or 8 or 16 blocks of output in parallel, instead
of just one, giving us large buffers for cheap, and introducing a
get_random_*_bh() function that assumes irqs are already disabled will
shave off a few cycles for ordinary calls. These are things we can chip
away at down the road.
Acked-by: Jakub Kicinski <[email protected]>
Acked-by: Theodore Ts'o <[email protected]>
Signed-off-by: Jason A. Donenfeld <[email protected]>
|
|
The SipHash family of permutations is currently used in three places:
- siphash.c itself, used in the ordinary way it was intended.
- random32.c, in a construction from an anonymous contributor.
- random.c, as part of its fast_mix function.
Each one of these places reinvents the wheel with the same C code, same
rotation constants, and same symmetry-breaking constants.
This commit tidies things up a bit by placing macros for the
permutations and constants into siphash.h, where each of the three .c
users can access them. It also leaves a note dissuading more users of
them from emerging.
Signed-off-by: Jason A. Donenfeld <[email protected]>
|
|
Now that fast_mix() has more than one caller, gcc no longer inlines it.
That's fine. But it also doesn't handle the compound literal argument we
pass it very efficiently, nor does it handle the loop as well as it
could. So just expand the code to spell out this function so that it
generates the same code as it did before. Performance-wise, this now
behaves as it did before the last commit. The difference in actual code
size on x86 is 45 bytes, which is less than a cache line.
Signed-off-by: Jason A. Donenfeld <[email protected]>
|
|
Years ago, a separate fast pool was added for interrupts, so that the
cost associated with taking the input pool spinlocks and mixing into it
would be avoided in places where latency is critical. However, one
oversight was that add_input_randomness() and add_disk_randomness()
still sometimes are called directly from the interrupt handler, rather
than being deferred to a thread. This means that some unlucky interrupts
will be caught doing a blake2s_compress() call and potentially spinning
on input_pool.lock, which can also be taken by unprivileged users by
writing into /dev/urandom.
In order to fix this, add_timer_randomness() now checks whether it is
being called from a hard IRQ and if so, just mixes into the per-cpu IRQ
fast pool using fast_mix(), which is much faster and can be done
lock-free. A nice consequence of this, as well, is that it means hard
IRQ context FPU support is likely no longer useful.
The entropy estimation algorithm used by add_timer_randomness() is also
somewhat different than the one used for add_interrupt_randomness(). The
former looks at deltas of deltas of deltas, while the latter just waits
for 64 interrupts for one bit or for one second since the last bit. In
order to bridge these, and since add_interrupt_randomness() runs after
an add_timer_randomness() that's called from hard IRQ, we add to the
fast pool credit the related amount, and then subtract one to account
for add_interrupt_randomness()'s contribution.
A downside of this, however, is that the num argument is potentially
attacker controlled, which puts a bit more pressure on the fast_mix()
sponge to do more than it's really intended to do. As a mitigating
factor, the first 96 bits of input aren't attacker controlled (a cycle
counter followed by zeros), which means it's essentially two rounds of
siphash rather than one, which is somewhat better. It's also not that
much different from add_interrupt_randomness()'s use of the irq stack
instruction pointer register.
Cc: Thomas Gleixner <[email protected]>
Cc: Filipe Manana <[email protected]>
Cc: Peter Zijlstra <[email protected]>
Cc: Borislav Petkov <[email protected]>
Cc: Theodore Ts'o <[email protected]>
Signed-off-by: Jason A. Donenfeld <[email protected]>
|
|
If CONFIG_ARM64_SVE is not set:
arch/arm64/kernel/fpsimd.c:294:13: warning: ‘sve_free’ defined but not used [-Wunused-function]
Fix this by moving sve_free() and __sve_free() into the existing section
protected by "#ifdef CONFIG_ARM64_SVE", now the last user outside that
section has been removed.
Fixes: a1259dd80719 ("arm64/sve: Delay freeing memory in fpsimd_flush_thread()")
Signed-off-by: Geert Uytterhoeven <[email protected]>
Reviewed-by: Mark Brown <[email protected]>
Link: https://lore.kernel.org/r/cd633284683c24cb9469f8ff429915aedf67f868.1652798894.git.geert+renesas@glider.be
Signed-off-by: Catalin Marinas <[email protected]>
|
|
Add trailing comments to endmenu statements for better readability.
Signed-off-by: Juerg Haefliger <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Catalin Marinas <[email protected]>
|
|
The convention for indentation seems to be a single tab. Help text is
further indented by an additional two whitespaces. Fix the lines that
violate these rules.
While add it, add trailing comments to endif and endmenu statements for
better readability.
Signed-off-by: Juerg Haefliger <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Catalin Marinas <[email protected]>
|
|
The AST2600 when using the i210 NIC over NC-SI has been observed to
produce incorrect checksum results with specific MTU values. This was
first observed when sending data across a long distance set of networks.
On a local network, the following test was performed using a 1MB file of
random data.
On the receiver run this script:
#!/bin/bash
while [ 1 ]; do
# Zero the stats
nstat -r > /dev/null
nc -l 9899 > test-file
# Check for checksum errors
TcpInCsumErrors=$(nstat | grep TcpInCsumErrors)
if [ -z "$TcpInCsumErrors" ]; then
echo No TcpInCsumErrors
else
echo TcpInCsumErrors = $TcpInCsumErrors
fi
done
On an AST2600 system:
# nc <IP of receiver host> 9899 < test-file
The test was repeated with various MTU values:
# ip link set mtu 1410 dev eth0
The observed results:
1500 - good
1434 - bad
1400 - good
1410 - bad
1420 - good
The test was repeated after disabling tx checksumming:
# ethtool -K eth0 tx-checksumming off
And all MTU values tested resulted in transfers without error.
An issue with the driver cannot be ruled out, however there has been no
bug discovered so far.
David has done the work to take the original bug report of slow data
transfer between long distance connections and triaged it down to this
test case.
The vendor suspects this this is a hardware issue when using NC-SI. The
fixes line refers to the patch that introduced AST2600 support.
Reported-by: David Wilder <[email protected]>
Reviewed-by: Dylan Hung <[email protected]>
Signed-off-by: Joel Stanley <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
igb_read_phy_reg() will silently return, leaving phy_data untouched, if
hw->ops.read_reg isn't set. Depending on the uninitialized value of
phy_data, this led to the phy status check either succeeding immediately
or looping continuously for 2 seconds before emitting a noisy err-level
timeout. This message went out to the console even though there was no
actual problem.
Instead, first check if there is read_reg function pointer. If not,
proceed without trying to check the phy status register.
Fixes: b72f3f72005d ("igb: When GbE link up, wait for Remote receiver status condition")
Signed-off-by: Kevin Mitchell <[email protected]>
Tested-by: Gurucharan <[email protected]> (A Contingent worker at Intel)
Signed-off-by: Tony Nguyen <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
When removing the pn533 device (i2c or USB), there is a logic error. The
original code first cancels the worker (flush_delayed_work) and then
destroys the workqueue (destroy_workqueue), leaving the timer the last
one to be deleted (del_timer). This result in a possible race condition
in a multi-core preempt-able kernel. That is, if the cleanup
(pn53x_common_clean) is concurrently run with the timer handler
(pn533_listen_mode_timer), the timer can queue the poll_work to the
already destroyed workqueue, causing use-after-free.
This patch reorder the cleanup: it uses the del_timer_sync to make sure
the handler is finished before the routine will destroy the workqueue.
Note that the timer cannot be activated by the worker again.
static void pn533_wq_poll(struct work_struct *work)
...
rc = pn533_send_poll_frame(dev);
if (rc)
return;
if (cur_mod->len == 0 && dev->poll_mod_count > 1)
mod_timer(&dev->listen_timer, ...);
That is, the mod_timer can be called only when pn533_send_poll_frame()
returns no error, which is impossible because the device is detaching
and the lower driver should return ENODEV code.
Signed-off-by: Lin Ma <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
for-5.19/drivers
Pull NVMe updates from Christoph:
"nvme updates for Linux 5.19
- tighten the PCI presence check (Stefan Roese):
- fix a potential NULL pointer dereference in an error path
(Kyle Miller Smith)
- fix interpretation of the DMRSL field (Tom Yan)
- relax the data transfer alignment (Keith Busch)
- verbose error logging improvements (Max Gurtovoy, Chaitanya Kulkarni)
- misc cleanups (Chaitanya Kulkarni, me)"
* tag 'nvme-5.19-2022-05-18' of git://git.infradead.org/nvme:
nvme: split the enum used for various register constants
nvme-fabrics: add a request timeout helper
nvme-pci: harden drive presence detect in nvme_dev_disable()
nvme-pci: fix a NULL pointer dereference in nvme_alloc_admin_tags
nvme: mark internal passthru request RQF_QUIET
nvme: remove unneeded include from constants file
nvme: add missing status values to verbose logging
nvme: set dma alignment to dword
nvme: fix interpretation of DMRSL
|
|
Accessing the file table needs a rcu_dereference_protected().
Signed-off-by: Christoph Hellwig <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>
|
|
POLL* are unannotated values for the userspace ABI, while everything
in-kernel should use EPOLL* and the __poll_t type.
Signed-off-by: Christoph Hellwig <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>
|
|
apoll_events is fed to vfs_poll and the poll tables, so it should be
a __poll_t.
Signed-off-by: Christoph Hellwig <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>
|
|
io_file_get_normal isn't marked inline, so don't claim it as such in the
forward declaration.
Signed-off-by: Christoph Hellwig <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>
|
|
ERR_PTR abuses the high bits of a pointer to transport error information.
This is only safe for kernel pointers and not user pointers. Fix
io_buffer_select and its helpers to just return NULL for failure and get
rid of this abuse.
Signed-off-by: Christoph Hellwig <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>
|
|
Use the proper type.
Signed-off-by: Christoph Hellwig <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>
|
|
Provided buffers allow an application to supply io_uring with buffers
that can then be grabbed for a read/receive request, when the data
source is ready to deliver data. The existing scheme relies on using
IORING_OP_PROVIDE_BUFFERS to do that, but it can be difficult to use
in real world applications. It's pretty efficient if the application
is able to supply back batches of provided buffers when they have been
consumed and the application is ready to recycle them, but if
fragmentation occurs in the buffer space, it can become difficult to
supply enough buffers at the time. This hurts efficiency.
Add a register op, IORING_REGISTER_PBUF_RING, which allows an application
to setup a shared queue for each buffer group of provided buffers. The
application can then supply buffers simply by adding them to this ring,
and the kernel can consume then just as easily. The ring shares the head
with the application, the tail remains private in the kernel.
Provided buffers setup with IORING_REGISTER_PBUF_RING cannot use
IORING_OP_{PROVIDE,REMOVE}_BUFFERS for adding or removing entries to the
ring, they must use the mapped ring. Mapped provided buffer rings can
co-exist with normal provided buffers, just not within the same group ID.
To gauge overhead of the existing scheme and evaluate the mapped ring
approach, a simple NOP benchmark was written. It uses a ring of 128
entries, and submits/completes 32 at the time. 'Replenish' is how
many buffers are provided back at the time after they have been
consumed:
Test Replenish NOPs/sec
================================================================
No provided buffers NA ~30M
Provided buffers 32 ~16M
Provided buffers 1 ~10M
Ring buffers 32 ~27M
Ring buffers 1 ~27M
The ring mapped buffers perform almost as well as not using provided
buffers at all, and they don't care if you provided 1 or more back at
the same time. This means application can just replenish as they go,
rather than need to batch and compact, further reducing overhead in the
application. The NOP benchmark above doesn't need to do any compaction,
so that overhead isn't even reflected in the above test.
Co-developed-by: Dylan Yudaken <[email protected]>
Signed-off-by: Jens Axboe <[email protected]>
|
|
Abstract this out from io_sqe_buffer_register() so we can use it
elsewhere too without duplicating this code.
No intended functional changes in this patch.
Signed-off-by: Jens Axboe <[email protected]>
|
|
Obviously not really useful since it's not transferring data, but it
is helpful in benchmarking overhead of provided buffers.
Signed-off-by: Jens Axboe <[email protected]>
|
|
io_provided_buffer_select() must drop the submit lock, if needed, even
in the error handling case. Failure to do so will leave us with the
ctx->uring_lock held, causing spew like:
====================================
WARNING: iou-wrk-366/368 still has locks held!
5.18.0-rc6-00294-gdf8dc7004331 #994 Not tainted
------------------------------------
1 lock held by iou-wrk-366/368:
#0: ffff0000c72598a8 (&ctx->uring_lock){+.+.}-{3:3}, at: io_ring_submit_lock+0x20/0x48
stack backtrace:
CPU: 4 PID: 368 Comm: iou-wrk-366 Not tainted 5.18.0-rc6-00294-gdf8dc7004331 #994
Hardware name: linux,dummy-virt (DT)
Call trace:
dump_backtrace.part.0+0xa4/0xd4
show_stack+0x14/0x5c
dump_stack_lvl+0x88/0xb0
dump_stack+0x14/0x2c
debug_check_no_locks_held+0x84/0x90
try_to_freeze.isra.0+0x18/0x44
get_signal+0x94/0x6ec
io_wqe_worker+0x1d8/0x2b4
ret_from_fork+0x10/0x20
and triggering later hangs off get_signal() because we attempt to
re-grab the lock.
Reported-by: [email protected]
Fixes: 149c69b04a90 ("io_uring: abstract out provided buffer list selection")
Signed-off-by: Jens Axboe <[email protected]>
|
|
Mat Martineau says:
====================
mptcp: Fix checksum byte order on little-endian
These patches address a bug in the byte ordering of MPTCP checksums on
little-endian architectures. The __sum16 type is always big endian, but
was being cast to u16 and then byte-swapped (on little-endian archs)
when reading/writing the checksum field in MPTCP option headers.
MPTCP checksums are off by default, but are enabled if one or both peers
request it in the SYN/SYNACK handshake.
The corrected code is verified to interoperate between big-endian and
little-endian machines.
Patch 1 fixes the checksum byte order, patch 2 partially mitigates
interoperation with peers sending bad checksums by falling back to TCP
instead of resetting the connection.
====================
Signed-off-by: David S. Miller <[email protected]>
|
|
RFC 8684 section 3.7 describes several opportunities for a MPTCP
connection to "fall back" to regular TCP early in the connection
process, before it has been confirmed that MPTCP options can be
successfully propagated on all SYN, SYN/ACK, and data packets. If a peer
acknowledges the first received data packet with a regular TCP header
(no MPTCP options), fallback is allowed.
If the recipient of that first data packet finds a MPTCP DSS checksum
error, this provides an opportunity to fail gracefully with a TCP
fallback rather than resetting the connection (as might happen if a
checksum failure were detected later).
This commit modifies the checksum failure code to attempt fallback on
the initial subflow of a MPTCP connection, only if it's a failure in the
first data mapping. In cases where the peer initiates the connection,
requests checksums, is the first to send data, and the peer is sending
incorrect checksums (see
https://github.com/multipath-tcp/mptcp_net-next/issues/275), this allows
the connection to proceed as TCP rather than reset.
Fixes: dd8bcd1768ff ("mptcp: validate the data checksum")
Acked-by: Paolo Abeni <[email protected]>
Signed-off-by: Mat Martineau <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
The MPTCP code typecasts the checksum value to u16 and
then converts it to big endian while storing the value into
the MPTCP option.
As a result, the wire encoding for little endian host is
wrong, and that causes interoperabilty interoperability
issues with other implementation or host with different endianness.
Address the issue writing in the packet the unmodified __sum16 value.
MPTCP checksum is disabled by default, interoperating with systems
with bad mptcp-level csum encoding should cause fallback to TCP.
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/275
Fixes: c5b39e26d003 ("mptcp: send out checksum for DSS")
Fixes: 390b95a5fb84 ("mptcp: receive checksum for DSS")
Signed-off-by: Paolo Abeni <[email protected]>
Signed-off-by: Mat Martineau <[email protected]>
Signed-off-by: David S. Miller <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2022-05-17
This series contains updates to ice driver only.
Arkadiusz prevents writing of timestamps when rings are being
configured to resolve null pointer dereference.
Paul changes a delayed call to baseline statistics to occur immediately
which was causing misreporting of statistics due to the delay.
Michal fixes incorrect restoration of interrupt moderation settings.
====================
Signed-off-by: David S. Miller <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec
Steffen Klassert says:
====================
pull request (net): ipsec 2022-05-18
1) Fix "disable_policy" flag use when arriving from different devices.
From Eyal Birger.
2) Fix error handling of pfkey_broadcast in function pfkey_process.
From Jiasheng Jiang.
3) Check the encryption module availability consistency in pfkey.
From Thomas Bartschies.
Please pull or let me know if there are problems.
====================
Signed-off-by: David S. Miller <[email protected]>
|
|
Get rid of old 31 bit leftovers within ipl code:
- convert everything to pc relative code
- use 64 bit addressing mode as early as possible
- use 64 bit arithmetics wherever possible
This way the code doesn't look as odd as before anymore.
Reviewed-by: Sven Schnelle <[email protected]>
Signed-off-by: Heiko Carstens <[email protected]>
|
|
|
|
In Thumb2, 'b . + 4' produces a branch instruction that uses a narrow
encoding, and so it does not jump to the following instruction as
expected. So use W(b) instead.
Fixes: 6c7cb60bff7a ("ARM: fix Thumb2 regression with Spectre BHB")
Signed-off-by: Ard Biesheuvel <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
|
|
The Spectre-BHB mitigations were inadvertently left disabled for
Cortex-A15, due to the fact that cpu_v7_bugs_init() is not called in
that case. So fix that.
Fixes: b9baf5c8c5c3 ("ARM: Spectre-BHB workaround")
Signed-off-by: Ard Biesheuvel <[email protected]>
Signed-off-by: Russell King (Oracle) <[email protected]>
|
|
git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5 fixes 2022-05-17
This series provides bug fixes to mlx5 driver.
Please pull and let me know if there is any problem.
====================
Signed-off-by: David S. Miller <[email protected]>
|
|
Since the recent introduction supporting the SM3 and SM4 hash algos for IPsec, the kernel
produces invalid pfkey acquire messages, when these encryption modules are disabled. This
happens because the availability of the algos wasn't checked in all necessary functions.
This patch adds these checks.
Signed-off-by: Thomas Bartschies <[email protected]>
Signed-off-by: Steffen Klassert <[email protected]>
|
|
If skb_clone() returns null pointer, pfkey_broadcast() will
return error.
Therefore, it should be better to check the return value of
pfkey_broadcast() and return error if fails.
Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Jiasheng Jiang <[email protected]>
Signed-off-by: Steffen Klassert <[email protected]>
|
|
That way percpu_ref_exit() is safe after failing percpu_ref_init().
At least one user (cgroup_create()) had a double-free that way;
there might be other similar bugs. Easier to fix in percpu_ref_init(),
rather than playing whack-a-mole in sloppy users...
Usual symptoms look like a messed refcounting in one of subsystems
that use percpu allocations (might be percpu-refcount, might be
something else). Having refcounts for two different objects share
memory is Not Nice(tm)...
Reported-by: [email protected]
Signed-off-by: Al Viro <[email protected]>
|
|
In case fw sync reset is called in parallel to device removal, device
might stuck in the following deadlock:
CPU 0 CPU 1
----- -----
remove_one
uninit_one (locks intf_state_mutex)
mlx5_sync_reset_now_event()
work in fw_reset->wq.
mlx5_enter_error_state()
mutex_lock (intf_state_mutex)
cleanup_once
fw_reset_cleanup()
destroy_workqueue(fw_reset->wq)
Drain the fw_reset WQ, and make sure no new work is being queued, before
entering uninit_one().
The Drain is done before devlink_unregister() since fw_reset, in some
flows, is using devlink API devlink_remote_reload_actions_performed().
Fixes: 38b9f903f22b ("net/mlx5: Handle sync reset request event")
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Moshe Shemesh <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
Cited patch sets flow_source to ANY overriding the provided spec
flow_source, avoiding the optimization done by commit c9c079b4deaa
("net/mlx5: CT: Set flow source hint from provided tuple device").
To fix the above, set the dr_rule flow_source from provided flow spec.
Fixes: 3ee61ebb0df1 ("net/mlx5: CT: Add software steering ct flow steering provider")
Signed-off-by: Paul Blakey <[email protected]>
Reviewed-by: Oz Shlomo <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
cited commit removed support for GRE tuples when software steering was enabled.
To bring back support for GRE tuples, add GRE ipv4/ipv6 matchers.
Fixes: 3ee61ebb0df1 ("net/mlx5: CT: Add software steering ct flow steering provider")
Signed-off-by: Paul Blakey <[email protected]>
Reviewed-by: Oz Shlomo <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
We got reports of certain HW-GRO flows causing kernel call traces, which
might be related to firmware. To be on the safe side, disable the
feature for now and re-enable it once a driver/firmware fix is found.
Fixes: 83439f3c37aa ("net/mlx5e: Add HW-GRO offload")
Signed-off-by: Gal Pressman <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
HW GRO is incompatible and mutually exclusive with XDP and XSK. However,
the needed checks are only made when enabling XDP. If HW GRO is enabled
when XDP is already active, the command will succeed, and XDP will be
skipped in the data path, although still enabled.
This commit fixes the bug by checking the XDP and XSK status in
mlx5e_fix_features and disabling HW GRO if XDP is enabled.
Fixes: 83439f3c37aa ("net/mlx5e: Add HW-GRO offload")
Signed-off-by: Maxim Mikityanskiy <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
LRO is incompatible and mutually exclusive with XDP. However, the needed
checks are only made when enabling XDP. If LRO is enabled when XDP is
already active, the command will succeed, and XDP will be skipped in the
data path, although still enabled.
This commit fixes the bug by checking the XDP status in
mlx5e_fix_features and disabling LRO if XDP is enabled.
Fixes: 86994156c736 ("net/mlx5e: XDP fast RX drop bpf programs support")
Signed-off-by: Maxim Mikityanskiy <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
When the driver is in switchdev mode and rx-gro-hw is set, the RQ needs
special CQE handling. Till then, block setting of rx-gro-hw feature in
switchdev mode, to avoid failure while setting the feature due to
failure while opening the RQ.
Fixes: f97d5c2a453e ("net/mlx5e: Add handle SHAMPO cqe support")
Signed-off-by: Aya Levin <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
The body of mlx5e_napi_poll is wrapped into rcu_read_lock to be able to
read the XDP program pointer using rcu_dereference. However, the trap RQ
NAPI doesn't use rcu_read_lock, because the trap RQ works only in the
non-linear mode, and mlx5e_skb_from_cqe_nonlinear, until recently,
didn't support XDP and didn't call rcu_dereference.
Starting from the cited commit, mlx5e_skb_from_cqe_nonlinear supports
XDP and calls rcu_dereference, but mlx5e_trap_napi_poll doesn't wrap it
into rcu_read_lock. It leads to RCU-lockdep warnings like this:
WARNING: suspicious RCU usage
This commit fixes the issue by adding an rcu_read_lock to
mlx5e_trap_napi_poll, similarly to mlx5e_napi_poll.
Fixes: ea5d49bdae8b ("net/mlx5e: Add XDP multi buffer support to the non-linear legacy RQ")
Signed-off-by: Maxim Mikityanskiy <[email protected]>
Reviewed-by: Tariq Toukan <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
When modifying TTL, packet's csum has to be recalculated.
Due to HW issue in ConnectX-5, csum recalculation for modify
TTL on RX is supported through a work-around that is specifically
enabled by configuration.
If the work-around isn't enabled, rather than adding an unsupported
action the modify TTL action on RX should be ignored.
Ignoring modify TTL action might result in zero actions, so in such
cases we will not convert the match STE to modify STE, as it is done
by FW in DMFS.
This patch fixes an issue where modify TTL action was ignored both
on RX and TX instead of only on RX.
Fixes: 4ff725e1d4ad ("net/mlx5: DR, Ignore modify TTL if device doesn't support it")
Signed-off-by: Yevgeny Kliteynik <[email protected]>
Reviewed-by: Alex Vesker <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
Currently, software objects of flow steering are created and destroyed
during reload flow. In case a device is unloaded, the following error
is printed during grace period:
mlx5_core 0000:00:0b.0: mlx5_fw_fatal_reporter_err_work:690:(pid 95):
Driver is in error state. Unloading
As a solution to fix use-after-free bugs, where we try to access
these objects, when reading the value of flow_steering_mode devlink
param[1], let's split flow steering creation and destruction into two
routines:
* init and cleanup: memory, cache, and pools allocation/free.
* create and destroy: namespaces initialization and cleanup.
While at it, re-order the cleanup function to mirror the init function.
[1]
Kasan trace:
[ 385.119849 ] BUG: KASAN: use-after-free in mlx5_devlink_fs_mode_get+0x3b/0xa0
[ 385.119849 ] Read of size 4 at addr ffff888104b79308 by task bash/291
[ 385.119849 ]
[ 385.119849 ] CPU: 1 PID: 291 Comm: bash Not tainted 5.17.0-rc1+ #2
[ 385.119849 ] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.13.0-2.fc32 04/01/2014
[ 385.119849 ] Call Trace:
[ 385.119849 ] <TASK>
[ 385.119849 ] dump_stack_lvl+0x6e/0x91
[ 385.119849 ] print_address_description.constprop.0+0x1f/0x160
[ 385.119849 ] ? mlx5_devlink_fs_mode_get+0x3b/0xa0
[ 385.119849 ] ? mlx5_devlink_fs_mode_get+0x3b/0xa0
[ 385.119849 ] kasan_report.cold+0x83/0xdf
[ 385.119849 ] ? devlink_param_notify+0x20/0x190
[ 385.119849 ] ? mlx5_devlink_fs_mode_get+0x3b/0xa0
[ 385.119849 ] mlx5_devlink_fs_mode_get+0x3b/0xa0
[ 385.119849 ] devlink_nl_param_fill+0x18a/0xa50
[ 385.119849 ] ? _raw_spin_lock_irqsave+0x8d/0xe0
[ 385.119849 ] ? devlink_flash_update_timeout_notify+0xf0/0xf0
[ 385.119849 ] ? __wake_up_common+0x4b/0x1e0
[ 385.119849 ] ? preempt_count_sub+0x14/0xc0
[ 385.119849 ] ? _raw_spin_unlock_irqrestore+0x28/0x40
[ 385.119849 ] ? __wake_up_common_lock+0xe3/0x140
[ 385.119849 ] ? __wake_up_common+0x1e0/0x1e0
[ 385.119849 ] ? __sanitizer_cov_trace_const_cmp8+0x27/0x80
[ 385.119849 ] ? __rcu_read_unlock+0x48/0x70
[ 385.119849 ] ? kasan_unpoison+0x23/0x50
[ 385.119849 ] ? __kasan_slab_alloc+0x2c/0x80
[ 385.119849 ] ? memset+0x20/0x40
[ 385.119849 ] ? __sanitizer_cov_trace_const_cmp4+0x25/0x80
[ 385.119849 ] devlink_param_notify+0xce/0x190
[ 385.119849 ] devlink_unregister+0x92/0x2b0
[ 385.119849 ] remove_one+0x41/0x140
[ 385.119849 ] pci_device_remove+0x68/0x140
[ 385.119849 ] ? pcibios_free_irq+0x10/0x10
[ 385.119849 ] __device_release_driver+0x294/0x3f0
[ 385.119849 ] device_driver_detach+0x82/0x130
[ 385.119849 ] unbind_store+0x193/0x1b0
[ 385.119849 ] ? subsys_interface_unregister+0x270/0x270
[ 385.119849 ] drv_attr_store+0x4e/0x70
[ 385.119849 ] ? drv_attr_show+0x60/0x60
[ 385.119849 ] sysfs_kf_write+0xa7/0xc0
[ 385.119849 ] kernfs_fop_write_iter+0x23a/0x2f0
[ 385.119849 ] ? sysfs_kf_bin_read+0x160/0x160
[ 385.119849 ] new_sync_write+0x311/0x430
[ 385.119849 ] ? new_sync_read+0x480/0x480
[ 385.119849 ] ? _raw_spin_lock+0x87/0xe0
[ 385.119849 ] ? __sanitizer_cov_trace_cmp4+0x25/0x80
[ 385.119849 ] ? security_file_permission+0x94/0xa0
[ 385.119849 ] vfs_write+0x4c7/0x590
[ 385.119849 ] ksys_write+0xf6/0x1e0
[ 385.119849 ] ? __x64_sys_read+0x50/0x50
[ 385.119849 ] ? fpregs_assert_state_consistent+0x99/0xa0
[ 385.119849 ] do_syscall_64+0x3d/0x90
[ 385.119849 ] entry_SYSCALL_64_after_hwframe+0x44/0xae
[ 385.119849 ] RIP: 0033:0x7fc36ef38504
[ 385.119849 ] Code: 00 f7 d8 64 89 02 48 c7 c0 ff ff ff ff eb b3 0f 1f
80 00 00 00 00 48 8d 05 f9 61 0d 00 8b 00 85 c0 75 13 b8 01 00 00 00 0f
05 <48> 3d 00 f0 ff ff 77 54 c3 0f 1f 00 41 54 49 89 d4 55 48 89 f5 53
[ 385.119849 ] RSP: 002b:00007ffde0ff3d08 EFLAGS: 00000246 ORIG_RAX: 0000000000000001
[ 385.119849 ] RAX: ffffffffffffffda RBX: 000000000000000c RCX: 00007fc36ef38504
[ 385.119849 ] RDX: 000000000000000c RSI: 00007fc370521040 RDI: 0000000000000001
[ 385.119849 ] RBP: 00007fc370521040 R08: 00007fc36f00b8c0 R09: 00007fc36ee4b740
[ 385.119849 ] R10: 0000000000000000 R11: 0000000000000246 R12: 00007fc36f00a760
[ 385.119849 ] R13: 000000000000000c R14: 00007fc36f005760 R15: 000000000000000c
[ 385.119849 ] </TASK>
[ 385.119849 ]
[ 385.119849 ] Allocated by task 65:
[ 385.119849 ] kasan_save_stack+0x1e/0x40
[ 385.119849 ] __kasan_kmalloc+0x81/0xa0
[ 385.119849 ] mlx5_init_fs+0x11b/0x1160
[ 385.119849 ] mlx5_load+0x13c/0x220
[ 385.119849 ] mlx5_load_one+0xda/0x160
[ 385.119849 ] mlx5_recover_device+0xb8/0x100
[ 385.119849 ] mlx5_health_try_recover+0x2f9/0x3a1
[ 385.119849 ] devlink_health_reporter_recover+0x75/0x100
[ 385.119849 ] devlink_health_report+0x26c/0x4b0
[ 385.275909 ] mlx5_fw_fatal_reporter_err_work+0x11e/0x1b0
[ 385.275909 ] process_one_work+0x520/0x970
[ 385.275909 ] worker_thread+0x378/0x950
[ 385.275909 ] kthread+0x1bb/0x200
[ 385.275909 ] ret_from_fork+0x1f/0x30
[ 385.275909 ]
[ 385.275909 ] Freed by task 65:
[ 385.275909 ] kasan_save_stack+0x1e/0x40
[ 385.275909 ] kasan_set_track+0x21/0x30
[ 385.275909 ] kasan_set_free_info+0x20/0x30
[ 385.275909 ] __kasan_slab_free+0xfc/0x140
[ 385.275909 ] kfree+0xa5/0x3b0
[ 385.275909 ] mlx5_unload+0x2e/0xb0
[ 385.275909 ] mlx5_unload_one+0x86/0xb0
[ 385.275909 ] mlx5_fw_fatal_reporter_err_work.cold+0xca/0xcf
[ 385.275909 ] process_one_work+0x520/0x970
[ 385.275909 ] worker_thread+0x378/0x950
[ 385.275909 ] kthread+0x1bb/0x200
[ 385.275909 ] ret_from_fork+0x1f/0x30
[ 385.275909 ]
[ 385.275909 ] The buggy address belongs to the object at ffff888104b79300
[ 385.275909 ] which belongs to the cache kmalloc-128 of size 128
[ 385.275909 ] The buggy address is located 8 bytes inside of
[ 385.275909 ] 128-byte region [ffff888104b79300, ffff888104b79380)
[ 385.275909 ] The buggy address belongs to the page:
[ 385.275909 ] page:00000000de44dd39 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x104b78
[ 385.275909 ] head:00000000de44dd39 order:1 compound_mapcount:0
[ 385.275909 ] flags: 0x8000000000010200(slab|head|zone=2)
[ 385.275909 ] raw: 8000000000010200 0000000000000000 dead000000000122 ffff8881000428c0
[ 385.275909 ] raw: 0000000000000000 0000000080200020 00000001ffffffff 0000000000000000
[ 385.275909 ] page dumped because: kasan: bad access detected
[ 385.275909 ]
[ 385.275909 ] Memory state around the buggy address:
[ 385.275909 ] ffff888104b79200: 00 00 00 00 00 00 00 00 00 00 00 00 00 00 fc fc
[ 385.275909 ] ffff888104b79280: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[ 385.275909 ] >ffff888104b79300: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 385.275909 ] ^
[ 385.275909 ] ffff888104b79380: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc
[ 385.275909 ] ffff888104b79400: fa fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
[ 385.275909 ]]
Fixes: e890acd5ff18 ("net/mlx5: Add devlink flow_steering_mode parameter")
Signed-off-by: Shay Drory <[email protected]>
Reviewed-by: Mark Bloch <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
In order to support multiple destination FTEs with SW steering
FW table is created with single FTE with multiple actions and
SW steering rule forward to it. When creating this table, flow
source isn't set according to the original FTE.
Fix this by passing the original FTE flow source to the created
FW table.
Fixes: 34583beea4b7 ("net/mlx5: DR, Create multi-destination table for SW-steering use")
Signed-off-by: Maor Dickman <[email protected]>
Reviewed-by: Yevgeny Kliteynik <[email protected]>
Signed-off-by: Saeed Mahameed <[email protected]>
|
|
In commit d72d827f2f26, I used 'cpumask_t' incorrectly:
void iscsit_thread_get_cpumask(struct iscsi_conn *conn)
{
int ord, cpu;
cpumask_t conn_allowed_cpumask;
......
}
static ssize_t lio_target_wwn_cpus_allowed_list_store(
struct config_item *item, const char *page, size_t count)
{
int ret;
char *orig;
cpumask_t new_allowed_cpumask;
......
}
The correct pattern should be as follows:
cpumask_var_t mask;
if (!zalloc_cpumask_var(&mask, GFP_KERNEL))
return -ENOMEM;
... use 'mask' here ...
free_cpumask_var(mask);
Link: https://lore.kernel.org/r/[email protected]
Fixes: d72d827f2f26 ("scsi: target: Add iscsi/cpus_allowed_list in configfs")
Reported-by: Test Bot <[email protected]>
Reviewed-by: Mike Christie <[email protected]>
Signed-off-by: Mingzhe Zou <[email protected]>
Signed-off-by: Martin K. Petersen <[email protected]>
|
|
1.In current process, all bio will set the BIO_THROTTLED flag
after __blk_throtl_bio().
2.If bio needs to be throttled, it will start the timer and
stop submit bio directly. Bio will submit in
blk_throtl_dispatch_work_fn() when the timer expires.But in
the current process, if bio is throttled. The BIO_THROTTLED
will be set to bio after timer start. If the bio has been
completed, it may cause use-after-free blow.
BUG: KASAN: use-after-free in blk_throtl_bio+0x12f0/0x2c70
Read of size 2 at addr ffff88801b8902d4 by task fio/26380
dump_stack+0x9b/0xce
print_address_description.constprop.6+0x3e/0x60
kasan_report.cold.9+0x22/0x3a
blk_throtl_bio+0x12f0/0x2c70
submit_bio_checks+0x701/0x1550
submit_bio_noacct+0x83/0xc80
submit_bio+0xa7/0x330
mpage_readahead+0x380/0x500
read_pages+0x1c1/0xbf0
page_cache_ra_unbounded+0x471/0x6f0
do_page_cache_ra+0xda/0x110
ondemand_readahead+0x442/0xae0
page_cache_async_ra+0x210/0x300
generic_file_buffered_read+0x4d9/0x2130
generic_file_read_iter+0x315/0x490
blkdev_read_iter+0x113/0x1b0
aio_read+0x2ad/0x450
io_submit_one+0xc8e/0x1d60
__se_sys_io_submit+0x125/0x350
do_syscall_64+0x2d/0x40
entry_SYSCALL_64_after_hwframe+0x44/0xa9
Allocated by task 26380:
kasan_save_stack+0x19/0x40
__kasan_kmalloc.constprop.2+0xc1/0xd0
kmem_cache_alloc+0x146/0x440
mempool_alloc+0x125/0x2f0
bio_alloc_bioset+0x353/0x590
mpage_alloc+0x3b/0x240
do_mpage_readpage+0xddf/0x1ef0
mpage_readahead+0x264/0x500
read_pages+0x1c1/0xbf0
page_cache_ra_unbounded+0x471/0x6f0
do_page_cache_ra+0xda/0x110
ondemand_readahead+0x442/0xae0
page_cache_async_ra+0x210/0x300
generic_file_buffered_read+0x4d9/0x2130
generic_file_read_iter+0x315/0x490
blkdev_read_iter+0x113/0x1b0
aio_read+0x2ad/0x450
io_submit_one+0xc8e/0x1d60
__se_sys_io_submit+0x125/0x350
do_syscall_64+0x2d/0x40
entry_SYSCALL_64_after_hwframe+0x44/0xa9
Freed by task 0:
kasan_save_stack+0x19/0x40
kasan_set_track+0x1c/0x30
kasan_set_free_info+0x1b/0x30
__kasan_slab_free+0x111/0x160
kmem_cache_free+0x94/0x460
mempool_free+0xd6/0x320
bio_free+0xe0/0x130
bio_put+0xab/0xe0
bio_endio+0x3a6/0x5d0
blk_update_request+0x590/0x1370
scsi_end_request+0x7d/0x400
scsi_io_completion+0x1aa/0xe50
scsi_softirq_done+0x11b/0x240
blk_mq_complete_request+0xd4/0x120
scsi_mq_done+0xf0/0x200
virtscsi_vq_done+0xbc/0x150
vring_interrupt+0x179/0x390
__handle_irq_event_percpu+0xf7/0x490
handle_irq_event_percpu+0x7b/0x160
handle_irq_event+0xcc/0x170
handle_edge_irq+0x215/0xb20
common_interrupt+0x60/0x120
asm_common_interrupt+0x1e/0x40
Fix this by move BIO_THROTTLED set into the queue_lock.
Signed-off-by: Laibin Qiu <[email protected]>
Reviewed-by: Ming Lei <[email protected]>
Link: https://lore.kernel.org/r/[email protected]
Signed-off-by: Jens Axboe <[email protected]>
|