kernel/linux.git/arch/xtensa/include/asm/bitops.h, branch v6.19.11Linux kernel stable tree (mirror)https://git.radix-linux.su/kernel/linux.git/atom?h=v6.19.112025-09-08T21:58:52+00:00xtensa: Add __attribute_const__ to ffs()-family implementations2025-09-08T21:58:52+00:00Kees Cookkees@kernel.org2025-08-04T16:44:11+00:00urn:sha1:945fc9dbd8377e48246de3a5f0104ebe82399eb0
While tracking down a problem where constant expressions used by
BUILD_BUG_ON() suddenly stopped working[1], we found that an added static
initializer was convincing the compiler that it couldn't track the state
of the prior statically initialized value. Tracing this down found that
ffs() was used in the initializer macro, but since it wasn't marked with
__attribute__const__, the compiler had to assume the function might
change variable states as a side-effect (which is not true for ffs(),
which provides deterministic math results).
Add missing __attribute_const__ annotations to Xtensa's implementations
of ffs(), __ffs(), fls(), __fls(), ffz() functions. These are pure
mathematical functions that always return the same result for the
same input with no side effects, making them eligible for compiler
optimization.
Build tested ARCH=xtensa defconfig with GCC xtensa-linux 15.1.0.
Link: https://github.com/KSPP/linux/issues/364 [1]
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lore.kernel.org/r/20250804164417.1612371-15-kees@kernel.org
Signed-off-by: Kees Cook <kees@kernel.org>
xtensa: enable KCSAN2022-05-02T02:51:22+00:00Max Filippovjcmvbkbc@gmail.com2019-10-05T06:33:31+00:00urn:sha1:725aea873261e8d986e527838fde2a721f0962d8
Prefix arch-specific barrier macros with '__' to make use of instrumented
generic macros.
Prefix arch-specific bitops with 'arch_' to make use of instrumented
generic functions.
Provide stubs for 64-bit atomics when building with KCSAN.
Disable KCSAN instrumentation in arch/xtensa/boot.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
Acked-by: Marco Elver <elver@google.com>
include: move find.h from asm_generic to linux2022-01-15T16:47:31+00:00Yury Norovyury.norov@gmail.com2021-08-14T21:16:59+00:00urn:sha1:47d8c15615c0a2046d2d90b04cb80b81ddf31fb1
find_bit API and bitmap API are closely related, but inclusion paths
are different - include/asm-generic and include/linux, correspondingly.
In the past it made a lot of troubles due to circular dependencies
and/or undefined symbols. Fix this by moving find.h under include/linux.
Signed-off-by: Yury Norov <yury.norov@gmail.com>
Tested-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Acked-by: Geert Uytterhoeven <geert@linux-m68k.org>
xtensa: use "m" constraint instead of "a" in bitops.h assembly2019-11-26T19:33:38+00:00Max Filippovjcmvbkbc@gmail.com2019-10-16T05:17:33+00:00urn:sha1:5bf67094a3a2d99d5f96db30be286f6c41988177
Use "m" constraint instead of "r" for the address, as "m" allows
compiler to access adjacent locations using base + offset, while "r"
requires updating the base register every time.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
xtensa: use named assembly arguments in bitops.h2019-11-26T19:33:38+00:00Max Filippovjcmvbkbc@gmail.com2019-10-16T05:14:15+00:00urn:sha1:e444917019258ef3bd564d0b4a432add2c26a2ae
Numeric assembly arguments are hard to understand and assembly code that
uses them is hard to modify. Use named arguments in BIT_OP and
TEST_AND_BIT_OP macros.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
xtensa: use macros to generate *_bit and test_and_*_bit functions2019-11-26T19:33:38+00:00Max Filippovjcmvbkbc@gmail.com2019-10-16T05:04:13+00:00urn:sha1:b387dc044efaa07cd8a47316c83fe2a5c08f9650
Parameterize macros with function name, opcode and inversion pattern.
This reduces code duplication removing 2/3 of definitions.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
xtensa: fix change_bit in exclusive access option2019-10-16T07:14:33+00:00Max Filippovjcmvbkbc@gmail.com2019-10-16T04:51:43+00:00urn:sha1:775fd6bfefc66a8c33e91dd9687ed530643b954d
change_bit implementation for XCHAL_HAVE_EXCLUSIVE case changes all bits
except the one required due to copy-paste error from clear_bit.
Cc: stable@vger.kernel.org # v5.2+
Fixes: f7c34874f04a ("xtensa: add exclusive atomics support")
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
xtensa: add exclusive atomics support2019-05-07T17:36:31+00:00Max Filippovjcmvbkbc@gmail.com2018-12-21T01:18:12+00:00urn:sha1:f7c34874f04a80d6c39a32f08da2529e59602d3c
Implement atomic primitives using exclusive access opcodes available in
the recent xtensa cores.
Since l32ex/s32ex don't have any memory ordering guarantees don't define
__smp_mb__before_atomic/__smp_mb__after_atomic to make them use memw.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
xtensa: drop ifdef __KERNEL__ from kernel-only headers2019-05-07T00:48:54+00:00Max Filippovjcmvbkbc@gmail.com2019-05-06T20:07:26+00:00urn:sha1:8e65986dcae8c49501c1920064dc192e704248bb
These headers are not exported to userspace, so they're never used
without __KERNEL__ defined. Drop these ifdef statements.
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>
xtensa: fix __ffs result type2016-07-24T03:33:29+00:00Max Filippovjcmvbkbc@gmail.com2016-04-25T01:06:37+00:00urn:sha1:12c8007dafbb7a9719d7beb04b79c9406c265f47
Make __ffs result type unsigned long to match generic asm
implementation. This fixes the following build warning:
mm/nobootmem.c: In function '__free_pages_memory':
include/linux/kernel.h:742:17: warning: comparison of distinct pointer
types lacks a cast
(void) (&_min1 == &_min2); \
^
mm/nobootmem.c:100:11: note: in expansion of macro 'min'
order = min(MAX_ORDER - 1UL, __ffs(start));
Signed-off-by: Max Filippov <jcmvbkbc@gmail.com>