Lines Matching +full:four +full:- +full:byte

1 /* SPDX-License-Identifier: GPL-2.0 */
5 /* Because kmalloc only guarantees 8-byte alignment for kmalloc'd data,
6    and GCC only guarantees 8-byte alignment for stack locals, we can't
7    be assured of 16-byte alignment for atomic lock data even if we
9    we use a struct containing an array of four ints for the atomic lock
10    type and dynamically select the 16-byte aligned int from the array
16    16-byte alignment requirement for ldcw and ldcd is relaxed, and instead
17    they only require "natural" alignment (4-byte for ldcw, 8-byte for
22    require 16-byte alignment. If the address is unaligned, the operation
25    This hid the problem for years. So, restore the 16-byte alignment dropped
26    by Kyle McMartin in "Remove __ldcw_align for PA-RISC 2.0 processors". */
30 	unsigned long __ret = (unsigned long) &(a)->lock[0];	\
31 	__ret = (__ret + __PA_LDCW_ALIGNMENT - 1)		\
32 		& ~(__PA_LDCW_ALIGNMENT - 1);			\
42 /* LDCW, the only atomic read-write operation PA-RISC has. *sigh*.
45    reloaded when generating 64-bit PIC code.  Instead, we clobber