diff options
author | Heiko Carstens <hca@linux.ibm.com> | 2024-02-03 13:45:03 +0300 |
---|---|---|
committer | Heiko Carstens <hca@linux.ibm.com> | 2024-02-16 16:30:14 +0300 |
commit | 045bad0800cec6098c30148e2873e20db742c750 (patch) | |
tree | 139c7b00c369d00545f685233a2bf27a18ff818e /arch/s390/include/asm/fpu-insn.h | |
parent | fd2527f20915d041e838b6e4a08122dbc73c7abc (diff) | |
download | linux-045bad0800cec6098c30148e2873e20db742c750.tar.xz |
s390/fpu: add documentation about fpu helper functions
Add documentation which describes what the fpu helper functions are
good for, and why they should be used.
Reviewed-by: Claudio Imbrenda <imbrenda@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Diffstat (limited to 'arch/s390/include/asm/fpu-insn.h')
-rw-r--r-- | arch/s390/include/asm/fpu-insn.h | 20 |
1 files changed, 20 insertions, 0 deletions
diff --git a/arch/s390/include/asm/fpu-insn.h b/arch/s390/include/asm/fpu-insn.h index 32d2894b60ca..ccdaacdeaa97 100644 --- a/arch/s390/include/asm/fpu-insn.h +++ b/arch/s390/include/asm/fpu-insn.h @@ -15,6 +15,26 @@ asm(".include \"asm/fpu-insn-asm.h\"\n"); +/* + * Various small helper functions, which can and should be used within + * kernel fpu code sections. Each function represents only one floating + * point or vector instruction (except for helper functions which require + * exception handling). + * + * This allows to use floating point and vector instructions like C + * functions, which has the advantage that all supporting code, like + * e.g. loops, can be written in easy to read C code. + * + * Each of the helper functions provides support for code instrumentation, + * like e.g. KASAN. Therefore instrumentation is also covered automatically + * when using these functions. + * + * In order to ensure that code generated with the helper functions stays + * within kernel fpu sections, which are guarded with kernel_fpu_begin() + * and kernel_fpu_end() calls, each function has a mandatory "memory" + * barrier. + */ + /** * sfpc_safe - Set floating point control register safely. * @fpc: new value for floating point control register |