sbitmap suffers from code complexity, as demonstrated by recent fixes,
and eventual lost wake ups on nested I/O completion. The later happens,
from what I understand, due to the non-atomic nature of the updates to
wait_cnt, which needs to be subtracted and eventually reset when equal
to zero. This two step process can eventually miss an update when a
nested completion happens to interrupt the CPU in between the wait_cnt
updates. This is very hard to fix, as shown by the recent changes to
this code.
The code complexity arises mostly from the corner cases to avoid missed
wakes in this scenario. In addition, the handling of wake_batch
recalculation plus the synchronization with sbq_queue_wake_up is
non-trivial.
This patchset implements the idea originally proposed by Jan [1], which
removes the need for the two-step updates of wait_cnt. This is done by
tracking the number of completions and wakeups in always increasing,
per-bitmap counters. Instead of having to reset the wait_cnt when it
reaches zero, we simply keep counting, and attempt to wake up N threads
in a single wait queue whenever there is enough space for a batch.
Waking up less than batch_wake shouldn't be a problem, because we
haven't changed the conditions for wake up, and the existing batch
calculation guarantees at least enough remaining completions to wake up
a batch for each queue at any time.
Performance-wise, one should expect very similar performance to the
original algorithm for the case where there is no queueing. In both the
old algorithm and this implementation, the first thing is to check
ws_active, which bails out if there is no queueing to be managed. In the
new code, we took care to avoid accounting completions and wakeups when
there is no queueing, to not pay the cost of atomic operations
unnecessarily, since it doesn't skew the numbers.
For more interesting cases, where there is queueing, we need to take
into account the cross-communication of the atomic operations. I've
been benchmarking by running parallel fio jobs against a single hctx
nullb in different hardware queue depth scenarios, and verifying both
IOPS and queueing.
Each experiment was repeated 5 times on a 20-CPU box, with 20 parallel
jobs. fio was issuing fixed-size randwrites with qd=64 against nullb,
varying only the hardware queue length per test.
queue size 2 4 8 16 32 64
6.1-rc2 1681.1K (1.6K) 2633.0K (12.7K) 6940.8K (16.3K) 8172.3K (617.5K) 8391.7K (367.1K) 8606.1K (351.2K)
patched 1721.8K (15.1K) 3016.7K (3.8K) 7543.0K (89.4K) 8132.5K (303.4K) 8324.2K (230.6K) 8401.8K (284.7K)
The following is a similar experiment, ran against a nullb with a single
bitmap shared by 20 hctx spread across 2 NUMA nodes. This has 40
parallel fio jobs operating on the same device
queue size 2 4 8 16 32 64
6.1-rc2 1081.0K (2.3K) 957.2K (1.5K) 1699.1K (5.7K) 6178.2K (124.6K) 12227.9K (37.7K) 13286.6K (92.9K)
patched 1081.8K (2.8K) 1316.5K (5.4K) 2364.4K (1.8K) 6151.4K (20.0K) 11893.6K (17.5K) 12385.6K (18.4K)
It has also survived blktests and a 12h-stress run against nullb. I also
ran the code against nvme and a scsi SSD, and I didn't observe
performance regression in those. If there are other tests you think I
should run, please let me know and I will follow up with results.
[1] https://lore.kernel.org/all/aef9de29-e9f5-259a-f8be-12d1b734e72@google.com/
Cc: Hugh Dickins <hughd@google.com>
Cc: Keith Busch <kbusch@kernel.org>
Cc: Liu Song <liusong@linux.alibaba.com>
Suggested-by: Jan Kara <jack@suse.cz>
Signed-off-by: Gabriel Krisman Bertazi <krisman@suse.de>
Link: https://lore.kernel.org/r/20221105231055.25953-1-krisman@suse.de
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Implement a minimal library version of AES-GCM based on the existing
library implementations of AES and multiplication in GF(2^128). Using
these primitives, GCM can be implemented in a straight-forward manner.
GCM has a couple of sharp edges, i.e., the amount of input data
processed with the same initialization vector (IV) should be capped to
protect the counter from 32-bit rollover (or carry), and the size of the
authentication tag should be fixed for a given key. [0]
The former concern is addressed trivially, given that the function call
API uses 32-bit signed types for the input lengths. It is still up to
the caller to avoid IV reuse in general, but this is not something we
can police at the implementation level.
As for the latter concern, let's make the authentication tag size part
of the key schedule, and only permit it to be configured as part of the
key expansion routine.
Note that table based AES implementations are susceptible to known
plaintext timing attacks on the encryption key. The AES library already
attempts to mitigate this to some extent, but given that the counter
mode encryption used by GCM operates exclusively on known plaintext by
construction (the IV and therefore the initial counter value are known
to an attacker), let's take some extra care to mitigate this, by calling
the AES library with interrupts disabled.
[0] https://nvlpubs.nist.gov/nistpubs/legacy/sp/nistspecialpublication800-38d.pdf
Link: https://lore.kernel.org/all/c6fb9b25-a4b6-2e4a-2dd1-63adda055a49@amd.com/
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Tested-by: Nikunj A Dadhania <nikunj@amd.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The gf128mul library has different variants with different
memory/performance tradeoffs, where the faster ones use 4k or 64k lookup
tables precomputed at runtime, which are based on one of the
multiplication factors, which is commonly the key for keyed hash
algorithms such as GHASH.
The slowest variant is gf128_mul_lle() [and its bbe/ble counterparts],
which does not use precomputed lookup tables, but it still relies on a
single u16[256] lookup table which is input independent. The use of such
a table may cause the execution time of gf128_mul_lle() to correlate
with the value of the inputs, which is generally something that must be
avoided for cryptographic algorithms. On top of that, the function uses
a sequence of if () statements that conditionally invoke be128_xor()
based on which bits are set in the second argument of the function,
which is usually a pointer to the multiplication factor that represents
the key.
In order to remove the correlation between the execution time of
gf128_mul_lle() and the value of its inputs, let's address the
identified shortcomings:
- add a time invariant version of gf128mul_x8_lle() that replaces the
table lookup with the expression that is used at compile time to
populate the lookup table;
- make the invocations of be128_xor() unconditional, but pass a zero
vector as the third argument if the associated bit in the key is
cleared.
The resulting code is likely to be significantly slower. However, given
that this is the slowest version already, making it even slower in order
to make it more secure is assumed to be justified.
The bbe and ble counterparts could receive the same treatment, but the
former is never used anywhere in the kernel, and the latter is only
used in the driver for a asynchronous crypto h/w accelerator (Chelsio),
where timing variances are unlikely to matter.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
The gf128mul library does not depend on the crypto API at all, so it can
be moved into lib/crypto. This will allow us to use it in other library
code in a subsequent patch without having to depend on CONFIG_CRYPTO.
While at it, change the Kconfig symbol name to align with other crypto
library implementations. However, the source file name is retained, as
it is reflected in the module .ko filename, and changing this might
break things for users.
Signed-off-by: Ard Biesheuvel <ardb@kernel.org>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Consecutive zone device pages should not be merged into the same sgl
or bvec segment with other types of pages or if they belong to different
pgmaps. Otherwise getting the pgmap of a given segment is not possible
without scanning the entire segment. This helper returns true either if
both pages are not zone device pages or both pages are zone device
pages with the same pgmap.
Factor out the check for page mergability into a pages_are_mergable()
helper and add a check with zone_device_pages_are_mergeable().
Signed-off-by: Logan Gunthorpe <logang@deltatee.com>
Reviewed-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Chaitanya Kulkarni <kch@nvidia.com>
Link: https://lore.kernel.org/r/20221021174116.7200-6-logang@deltatee.com
Signed-off-by: Jens Axboe <axboe@kernel.dk>
Along the development cycle, the testing code support for module/in-kernel
compiles was removed. Restore this functionality by moving any internal
API tests to the userspace side, as well as threading tests. Fix the
lockdep issues and add a way to reduce memory usage so the tests can
complete with KASAN + memleak detection. Make the tests work on 32 bit
hosts where possible and detect 32 bit hosts in the radix test suite.
[akpm@linux-foundation.org: fix module export]
[akpm@linux-foundation.org: fix it some more]
[liam.howlett@oracle.com: fix compile warnings on 32bit build in check_find()]
Link: https://lkml.kernel.org/r/20221107203816.1260327-1-Liam.Howlett@oracle.com
Link: https://lkml.kernel.org/r/20221028180415.3074673-1-Liam.Howlett@oracle.com
Signed-off-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
clang-analyzer reported some Dead Stores in mas_anode_descend(). Upon
inspection, there were a few clean ups that would make the code cleaner:
The count variable was set from the mt_slots array and then updated but
never used again. Just use the array reference directly.
Also stop updating the type since it isn't used after the update.
Stop setting the gaps pointer to NULL at the start since it is always
set before the loop begins.
Link: https://lkml.kernel.org/r/20221026151413.4032730-1-Liam.Howlett@oracle.com
Signed-off-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Suggested-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Pull networking fixes from Paolo Abeni:
"Including fixes from bluetooth and netfilter.
Current release - regressions:
- net: several zerocopy flags fixes
- netfilter: fix possible memory leak in nf_nat_init()
- openvswitch: add missing .resv_start_op
Previous releases - regressions:
- neigh: fix null-ptr-deref in neigh_table_clear()
- sched: fix use after free in red_enqueue()
- dsa: fall back to default tagger if we can't load the one from DT
- bluetooth: fix use-after-free in l2cap_conn_del()
Previous releases - always broken:
- netfilter: netlink notifier might race to release objects
- nfc: fix potential memory leak of skb
- bluetooth: fix use-after-free caused by l2cap_reassemble_sdu
- bluetooth: use skb_put to set length
- eth: tun: fix bugs for oversize packet when napi frags enabled
- eth: lan966x: fixes for when MTU is changed
- eth: dwmac-loongson: fix invalid mdio_node"
* tag 'net-6.1-rc4' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (53 commits)
vsock: fix possible infinite sleep in vsock_connectible_wait_data()
vsock: remove the unused 'wait' in vsock_connectible_recvmsg()
ipv6: fix WARNING in ip6_route_net_exit_late()
bridge: Fix flushing of dynamic FDB entries
net, neigh: Fix null-ptr-deref in neigh_table_clear()
net/smc: Fix possible leaked pernet namespace in smc_init()
stmmac: dwmac-loongson: fix invalid mdio_node
ibmvnic: Free rwi on reset success
net: mdio: fix undefined behavior in bit shift for __mdiobus_register
Bluetooth: L2CAP: Fix attempting to access uninitialized memory
Bluetooth: L2CAP: Fix l2cap_global_chan_by_psm
Bluetooth: L2CAP: Fix accepting connection request for invalid SPSM
Bluetooth: hci_conn: Fix not restoring ISO buffer count on disconnect
Bluetooth: L2CAP: Fix memory leak in vhci_write
Bluetooth: L2CAP: fix use-after-free in l2cap_conn_del()
Bluetooth: virtio_bt: Use skb_put to set length
Bluetooth: hci_conn: Fix CIS connection dst_type handling
Bluetooth: L2CAP: Fix use-after-free caused by l2cap_reassemble_sdu
netfilter: ipset: enforce documented limit to prevent allocating huge memory
isdn: mISDN: netjet: fix wrong check of device registration
...
Implement a robust overflows_type() macro to test if a variable or
constant value would overflow another variable or type. This can be
used as a constant expression for static_assert() (which requires a
constant expression[1][2]) when used on constant values. This must be
constructed manually, since __builtin_add_overflow() does not produce
a constant expression[3].
Additionally adds castable_to_type(), similar to __same_type(), but for
checking if a constant value would overflow if cast to a given type.
Add unit tests for overflows_type(), __same_type(), and castable_to_type()
to the existing KUnit "overflow" test:
[16:03:33] ================== overflow (21 subtests) ==================
...
[16:03:33] [PASSED] overflows_type_test
[16:03:33] [PASSED] same_type_test
[16:03:33] [PASSED] castable_to_type_test
[16:03:33] ==================== [PASSED] overflow =====================
[16:03:33] ============================================================
[16:03:33] Testing complete. Ran 21 tests: passed: 21
[16:03:33] Elapsed time: 24.022s total, 0.002s configuring, 22.598s building, 0.767s running
[1] https://en.cppreference.com/w/c/language/_Static_assert
[2] C11 standard (ISO/IEC 9899:2011): 6.7.10 Static assertions
[3] https://gcc.gnu.org/onlinedocs/gcc/Integer-Overflow-Builtins.html
6.56 Built-in Functions to Perform Arithmetic with Overflow Checking
Built-in Function: bool __builtin_add_overflow (type1 a, type2 b,
Cc: Luc Van Oostenryck <luc.vanoostenryck@gmail.com>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Tom Rix <trix@redhat.com>
Cc: Daniel Latypov <dlatypov@google.com>
Cc: Vitor Massaru Iha <vitor@massaru.org>
Cc: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Cc: Jani Nikula <jani.nikula@intel.com>
Cc: Mauro Carvalho Chehab <mchehab@kernel.org>
Cc: linux-hardening@vger.kernel.org
Cc: llvm@lists.linux.dev
Co-developed-by: Gwan-gyeong Mun <gwan-gyeong.mun@intel.com>
Signed-off-by: Gwan-gyeong Mun <gwan-gyeong.mun@intel.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20221024201125.1416422-1-gwan-gyeong.mun@intel.com
Jakub reported that the addition of the "network_byte_order"
member in struct nla_policy increases size of 32bit platforms.
Instead of scraping the bit from elsewhere Johannes suggested
to add explicit NLA_BE types instead, so do this here.
NLA_POLICY_MAX_BE() macro is removed again, there is no need
for it: NLA_POLICY_MAX(NLA_BE.., ..) will do the right thing.
NLA_BE64 can be added later.
Fixes: 08724ef699 ("netlink: introduce NLA_POLICY_MAX_BE")
Reported-by: Jakub Kicinski <kuba@kernel.org>
Suggested-by: Johannes Berg <johannes@sipsolutions.net>
Signed-off-by: Florian Westphal <fw@strlen.de>
Link: https://lore.kernel.org/r/20221031123407.9158-1-fw@strlen.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
When building ARCH=i386 with CONFIG_LTO_CLANG_FULL=y, it's possible
(depending on additional configs which I have not been able to isolate)
to observe a failure during register allocation:
error: inline assembly requires more registers than available
when memmove is inlined into tcp_v4_fill_cb() or tcp_v6_fill_cb().
memmove is quite large and probably shouldn't be inlined due to size
alone. A noinline function attribute would be the simplest fix, but
there's a few things that stand out with the current definition:
In addition to having complex constraints that can't always be resolved,
the clobber list seems to be missing %bx. By using numbered operands
rather than symbolic operands, the constraints are quite obnoxious to
refactor.
Having a large function be 99% inline asm is a code smell that this
function should simply be written in stand-alone out-of-line assembler.
Moving this to out of line assembler guarantees that the
compiler cannot inline calls to memmove.
This has been done previously for 64b:
commit 9599ec0471 ("x86-64, mem: Convert memmove() to assembly file
and fix return value bug")
That gives the opportunity for other cleanups like fixing the
inconsistent use of tabs vs spaces and instruction suffixes, and the
label 3 appearing twice. Symbolic operands, local labels, and
additional comments would provide this code with a fresh coat of paint.
Finally, add a test that tickles the `rep movsl` implementation to test
it for correctness, since it has implicit operands.
Suggested-by: Ingo Molnar <mingo@kernel.org>
Suggested-by: David Laight <David.Laight@aculab.com>
Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Dave Hansen <dave.hansen@linux.intel.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Tested-by: Kees Cook <keescook@chromium.org>
Tested-by: Nathan Chancellor <nathan@kernel.org>
Link: https://lore.kernel.org/all/20221018172155.287409-1-ndesaulniers%40google.com
Replacing compile-time safe calls of strcpy()-related functions with
strscpy() was always calling the full strscpy() logic when a builtin
would be better. For example:
char buf[16];
strcpy(buf, "yes");
would reduce to __builtin_memcpy(buf, "yes", 4), but not if it was:
strscpy(buf, yes, sizeof(buf));
Fix this by checking if all sizes are known at compile-time.
Cc: linux-hardening@vger.kernel.org
Tested-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Kees Cook <keescook@chromium.org>
Pull misc hotfixes from Andrew Morton:
"Eight fix pre-6.0 bugs and the remainder address issues which were
introduced in the 6.1-rc merge cycle, or address issues which aren't
considered sufficiently serious to warrant a -stable backport"
* tag 'mm-hotfixes-stable-2022-10-28' of git://git.kernel.org/pub/scm/linux/kernel/git/akpm/mm: (23 commits)
mm: multi-gen LRU: move lru_gen_add_mm() out of IRQ-off region
lib: maple_tree: remove unneeded initialization in mtree_range_walk()
mmap: fix remap_file_pages() regression
mm/shmem: ensure proper fallback if page faults
mm/userfaultfd: replace kmap/kmap_atomic() with kmap_local_page()
x86: fortify: kmsan: fix KMSAN fortify builds
x86: asm: make sure __put_user_size() evaluates pointer once
Kconfig.debug: disable CONFIG_FRAME_WARN for KMSAN by default
x86/purgatory: disable KMSAN instrumentation
mm: kmsan: export kmsan_copy_page_meta()
mm: migrate: fix return value if all subpages of THPs are migrated successfully
mm/uffd: fix vma check on userfault for wp
mm: prep_compound_tail() clear page->private
mm,madvise,hugetlb: fix unexpected data loss with MADV_DONTNEED on hugetlbfs
mm/page_isolation: fix clang deadcode warning
fs/ext4/super.c: remove unused `deprecated_msg'
ipc/msg.c: fix percpu_counter use after free
memory tier, sysfs: rename attribute "nodes" to "nodelist"
MAINTAINERS: git://github.com -> https://github.com for nilfs2
mm/kmemleak: prevent soft lockup in kmemleak_scan()'s object iteration loops
...
It's really difficult to debug when cgroup or css refs leak. Let's add a
debug option to force the refcnt function to not be inlined so that they can
be kprobed for debugging.
Signed-off-by: Tejun Heo <tj@kernel.org>
Before the do-while loop in mtree_range_walk(), the variables next, min,
max need to be initialized. The variables last, prev_min and prev_max are
set within the loop body before they are eventually used after exiting the
loop body.
As it is a do-while loop, the loop body is executed at least once, so the
variables last, prev_min and prev_max do not need to be initialized before
the loop body.
Remove unneeded initialization of last and prev_min.
The needless initialization was reported by clang-analyzer as Dead Stores.
As the compiler already identifies these assignments as unneeded, it
optimizes the assignments away. Hence:
No functional change. No change in object code.
Link: https://lkml.kernel.org/r/20221026120029.12555-2-lukas.bulwahn@gmail.com
Signed-off-by: Lukas Bulwahn <lukas.bulwahn@gmail.com>
Reviewed-by: Liam R. Howlett <Liam.Howlett@oracle.com>
Cc: Matthew Wilcox <willy@infradead.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Pull networking fixes from Jakub Kicinski:
"Including fixes from 802.15.4 (Zigbee et al).
Current release - regressions:
- ipa: fix bugs in the register conversion for IPA v3.1 and v3.5.1
Current release - new code bugs:
- mptcp: fix abba deadlock on fastopen
- eth: stmmac: rk3588: allow multiple gmac controllers in one system
Previous releases - regressions:
- ip: rework the fix for dflt addr selection for connected nexthop
- net: couple more fixes for misinterpreting bits in struct page
after the signature was added
Previous releases - always broken:
- ipv6: ensure sane device mtu in tunnels
- openvswitch: switch from WARN to pr_warn on a user-triggerable path
- ethtool: eeprom: fix null-deref on genl_info in dump
- ieee802154: more return code fixes for corner cases in
dgram_sendmsg
- mac802154: fix link-quality-indicator recording
- eth: mlx5: fixes for IPsec, PTP timestamps, OvS and conntrack
offload
- eth: fec: limit register access on i.MX6UL
- eth: bcm4908_enet: update TX stats after actual transmission
- can: rcar_canfd: improve IRQ handling for RZ/G2L
Misc:
- genetlink: piggy back on the newly added resv_op_start to enforce
more sanity checks on new commands"
* tag 'net-6.1-rc3-2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (57 commits)
net: enetc: survive memory pressure without crashing
kcm: do not sense pfmemalloc status in kcm_sendpage()
net: do not sense pfmemalloc status in skb_append_pagefrags()
net/mlx5e: Fix macsec sci endianness at rx sa update
net/mlx5e: Fix wrong bitwise comparison usage in macsec_fs_rx_add_rule function
net/mlx5e: Fix macsec rx security association (SA) update/delete
net/mlx5e: Fix macsec coverity issue at rx sa update
net/mlx5: Fix crash during sync firmware reset
net/mlx5: Update fw fatal reporter state on PCI handlers successful recover
net/mlx5e: TC, Fix cloned flow attr instance dests are not zeroed
net/mlx5e: TC, Reject forwarding from internal port to internal port
net/mlx5: Fix possible use-after-free in async command interface
net/mlx5: ASO, Create the ASO SQ with the correct timestamp format
net/mlx5e: Update restore chain id for slow path packets
net/mlx5e: Extend SKB room check to include PTP-SQ
net/mlx5: DR, Fix matcher disconnect error flow
net/mlx5: Wait for firmware to enable CRS before pci_restore_state
net/mlx5e: Do not increment ESN when updating IPsec ESN state
netdevsim: remove dir in nsim_dev_debugfs_init() when creating ports dir failed
netdevsim: fix memory leak in nsim_drv_probe() when nsim_dev_resources_register() failed
...
Augment the example_all_expect_macros_test with the KUNIT_EXPECT_MEMEQ
and KUNIT_EXPECT_MEMNEQ macros by creating a test with memory block
assertions.
Signed-off-by: Maíra Canal <mairacanal@riseup.net>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
Currently, in order to compare memory blocks in KUnit, the KUNIT_EXPECT_EQ
or KUNIT_EXPECT_FALSE macros are used in conjunction with the memcmp
function, such as:
KUNIT_EXPECT_EQ(test, memcmp(foo, bar, size), 0);
Although this usage produces correct results for the test cases, when
the expectation fails, the error message is not very helpful,
indicating only the return of the memcmp function.
Therefore, create a new set of macros KUNIT_EXPECT_MEMEQ and
KUNIT_EXPECT_MEMNEQ that compare memory blocks until a specified size.
In case of expectation failure, those macros print the hex dump of the
memory blocks, making it easier to debug test failures for memory blocks.
That said, the expectation
KUNIT_EXPECT_EQ(test, memcmp(foo, bar, size), 0);
would translate to the expectation
KUNIT_EXPECT_MEMEQ(test, foo, bar, size);
Signed-off-by: Maíra Canal <mairacanal@riseup.net>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: Muhammad Usama Anjum <usama.anjum@collabora.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
When KUNIT_EXPECT_EQ() or KUNIT_ASSERT_EQ() log a failure, they log the
two values being compared, with numerical values logged in decimal.
In some cases, decimal output is painful to consume, and hexadecimal
output would be more helpful. For example, this is the case for tests
I'm currently developing for the arm64 insn encoding/decoding code,
where comparing two 32-bit instruction opcodes results in output such
as:
| # test_insn_add_shifted_reg: EXPECTATION FAILED at arch/arm64/lib/test_insn.c:2791
| Expected obj_insn == gen_insn, but
| obj_insn == 2332164128
| gen_insn == 1258422304
To make this easier to consume, this patch logs the values in both
decimal and hexadecimal:
| # test_insn_add_shifted_reg: EXPECTATION FAILED at arch/arm64/lib/test_insn.c:2791
| Expected obj_insn == gen_insn, but
| obj_insn == 2332164128 (0x8b020020)
| gen_insn == 1258422304 (0x4b020020)
As can be seen from the example, having hexadecimal makes it
significantly easier for a human to spot which specific bits are
incorrect.
Signed-off-by: Mark Rutland <mark.rutland@arm.com>
Cc: Brendan Higgins <brendan.higgins@linux.dev>
Cc: David Gow <davidgow@google.com>
Cc: linux-kselftest@vger.kernel.org
Cc: kunit-dev@googlegroups.com
Acked-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
The "random rhlist add/delete operations" actually wasn't very random, as all
cases tested the same bit. Since the later parts of this loop depend on the
first case execute this unconditionally, and then test on different bits for the
remaining tests. While at it only request as much random bits as are actually
used.
Signed-off-by: Rolf Eike Beer <eike-kernel@sf-tec.de>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>
Convert test exclusion into test skipping. This brings the logic for
why a test is being skipped into the test itself, instead of having to
spread ifdefs around the code. This will make cleanup easier as minimum
tests get raised. Drop __maybe_unused so missed tests will be noticed
again and clean up whitespace.
For example, clang-11 on i386:
[15:52:32] ================== overflow (18 subtests) ==================
[15:52:32] [PASSED] u8_u8__u8_overflow_test
[15:52:32] [PASSED] s8_s8__s8_overflow_test
[15:52:32] [PASSED] u16_u16__u16_overflow_test
[15:52:32] [PASSED] s16_s16__s16_overflow_test
[15:52:32] [PASSED] u32_u32__u32_overflow_test
[15:52:32] [PASSED] s32_s32__s32_overflow_test
[15:52:32] [SKIPPED] u64_u64__u64_overflow_test
[15:52:32] [SKIPPED] s64_s64__s64_overflow_test
[15:52:32] [SKIPPED] u32_u32__int_overflow_test
[15:52:32] [PASSED] u32_u32__u8_overflow_test
[15:52:32] [PASSED] u8_u8__int_overflow_test
[15:52:32] [PASSED] int_int__u8_overflow_test
[15:52:32] [PASSED] shift_sane_test
[15:52:32] [PASSED] shift_overflow_test
[15:52:32] [PASSED] shift_truncate_test
[15:52:32] [PASSED] shift_nonsense_test
[15:52:32] [PASSED] overflow_allocation_test
[15:52:32] [PASSED] overflow_size_helpers_test
[15:52:32] ==================== [PASSED] overflow =====================
[15:52:32] ============================================================
[15:52:32] Testing complete. Ran 18 tests: passed: 15, skipped: 3
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Tom Rix <trix@redhat.com>
Cc: Daniel Latypov <dlatypov@google.com>
Cc: "Gustavo A. R. Silva" <gustavoars@kernel.org>
Cc: Gwan-gyeong Mun <gwan-gyeong.mun@intel.com>
Cc: llvm@lists.linux.dev
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Nick Desaulniers <ndesaulniers@google.com>
Tested-by: Nick Desaulniers <ndesaulniers@google.com>
Link: https://lore.kernel.org/r/20221006230017.1833458-1-keescook@chromium.org
Building the overflow kunit tests with clang-11 fails with:
$ ./tools/testing/kunit/kunit.py run --arch=arm --make_options LLVM=1 \
overflow
...
ld.lld: error: undefined symbol: __mulodi4
...
Clang 11 and earlier generate unwanted libcalls for signed output,
unsigned input.
Disable these tests for now, but should these become used in the kernel
we might consider that as justification for dropping clang-11 support.
Keep the clang-11 build alive a little bit longer.
Avoid -Wunused-function warnings via __maybe_unused. To test W=1:
$ make LLVM=1 -j128 defconfig
$ ./scripts/config -e KUNIT -e KUNIT_ALL
$ make LLVM=1 -j128 olddefconfig lib/overflow_kunit.o W=1
Link: https://github.com/ClangBuiltLinux/linux/issues/1711
Link: 3203143f13
Reported-by: Nathan Chancellor <nathan@kernel.org>
Signed-off-by: Nick Desaulniers <ndesaulniers@google.com>
Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20221006171751.3444575-1-ndesaulniers@google.com
Inject fault while loading module, kset_register() may fail.
If it fails, the kset.kobj.name allocated by kobject_set_name()
which must be called before a call to kset_register() may be
leaked, since refcount of kobj was set in kset_init().
To mitigate this, we free the name in kset_register() when an
error is encountered, i.e. when kset_register() returns an error.
A kset may be embedded in a larger structure which may be dynamically
allocated in callers, it needs to be freed in ktype.release() or error
path in callers, in this case, we can not call kset_put() in kset_register(),
or it will cause double free, so just call kfree_const() to free the
name and set it to NULL to avoid accessing bad pointer in callers.
With this fix, the callers don't need care about freeing the name
and may call kset_put() if kset_register() fails.
Suggested-by: Luben Tuikov <luben.tuikov@amd.com>
Signed-off-by: Yang Yingliang <yangyingliang@huawei.com>
Reviewed-by: <luben.tuikov@amd.com>
Link: https://lore.kernel.org/r/20221025071549.1280528-1-yangyingliang@huawei.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Updates the kernel's zstd library to v1.5.2, the latest zstd release.
The upstream tag it is updated to is `v1.5.2-kernel`, which contains
several cherry-picked commits on top of the v1.5.2 release which are
required for the kernel update. I will create this tag once the PR is
ready to merge, until then reference the temporary upstream branch
`v1.5.2-kernel-cherrypicks`.
I plan to submit this patch as part of the v6.2 merge window.
I've done basic build testing & testing on x86-64, i386, and aarch64.
I'm merging these patches into my `zstd-next` branch, which is pulled
into `linux-next` for further testing.
I've benchmarked BtrFS with zstd compression on a x86-64 machine, and
saw these results. Decompression speed is a small win across the board.
The lower compression levels 1-4 see both compression speed and
compression ratio wins. The higher compression levels see a small
compression speed loss and about neutral ratio. I expect the lower
compression levels to be used much more heavily than the high
compression levels, so this should be a net win.
Level CTime DTime Ratio
1 -2.95% -1.1% -0.7%
3 -3.5% -1.2% -0.5%
5 +3.7% -1.0% +0.0%
7 +3.2% -0.9% +0.0%
9 -4.3% -0.8% +0.1%
Signed-off-by: Nick Terrell <terrelln@fb.com>
The zstd codebase is imported from the upstream zstd repo, and is over-written on
every update. Upstream keeps the kernel specific code separate from the main
library. So the module definition is moved into the zstd_common_module.c file.
This matches the pattern followed by the zstd-compress and zstd-decompress files.
I've done build and boot testing on x86-64, i386, and aarch64. I've
verified that zstd built both as modules and built-in build and boot.
Signed-off-by: Nick Terrell <terrelln@fb.com>
Fix trivial typo in Kconfig help comments in KCSAN_SKIP_WATCH and
KCSAN_SKIP_WATCH_RANDOMIZE
Signed-off-by: Ryosuke Yasuoka <ryasuoka@redhat.com>
Reviewed-by: Marco Elver <elver@google.com>
Signed-off-by: Paul E. McKenney <paulmck@kernel.org>
The alloc_string_stream() functions were changed from returning NULL on
error to returning error pointers so these caller needs to be updated
as well.
Fixes: 78b1c6584f ("kunit: string-stream: Simplify resource use")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Reviewed-by: Daniel Latypov <dlatypov@google.com>
Reviewed-by: David Gow <davidgow@google.com>
Signed-off-by: Shuah Khan <skhan@linuxfoundation.org>
