linux/kernel/kcsan
Marco Elver 75d75b7a4d kcsan: Support distinguishing volatile accesses
In the kernel, the "volatile" keyword is used in various concurrent
contexts, whether in low-level synchronization primitives or for
legacy reasons. If supported by the compiler, it will be assumed
that aligned volatile accesses up to sizeof(long long) (matching
compiletime_assert_rwonce_type()) are atomic.

Recent versions of Clang [1] (GCC tentative [2]) can instrument
volatile accesses differently. Add the option (required) to enable the
instrumentation, and provide the necessary runtime functions. None of
the updated compilers are widely available yet (Clang 11 will be the
first release to support the feature).

[1] 5a2c31116f
[2] https://gcc.gnu.org/pipermail/gcc-patches/2020-April/544452.html

This change allows removing of any explicit checks in primitives such as
READ_ONCE() and WRITE_ONCE().

 [ bp: Massage commit message a bit. ]

Signed-off-by: Marco Elver <elver@google.com>
Signed-off-by: Borislav Petkov <bp@suse.de>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Acked-by: Will Deacon <will@kernel.org>
Acked-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Link: https://lkml.kernel.org/r/20200521142047.169334-4-elver@google.com
2020-06-11 20:04:01 +02:00
..
Makefile kcsan, trace: Make KCSAN compatible with tracing 2020-03-21 09:44:41 +01:00
atomic.h kcsan: Add current->state to implicitly atomic accesses 2020-03-25 09:56:00 -07:00
core.c kcsan: Support distinguishing volatile accesses 2020-06-11 20:04:01 +02:00
debugfs.c kcsan: Use GFP_ATOMIC under spin lock 2020-04-27 11:10:10 -07:00
encoding.h kcsan: Prefer __always_inline for fast-path 2020-03-21 09:40:19 +01:00
kcsan.h kcsan: Avoid blocking producers in prepare_report() 2020-04-13 17:18:11 -07:00
report.c kcsan: Make reporting aware of KCSAN tests 2020-04-13 17:18:16 -07:00
test.c kcsan: Fix 0-sized checks 2020-03-21 09:42:42 +01:00