Lines Matching +full:sub +full:- +full:processor
8 * of the Xtensa link-time HAL, and those files may differ per
9 * processor configuration. Patching the kernel for another
10 * processor configuration includes replacing the hal files, and we
11 * could lose the special functionality for accessing user-space
37 * This code tries to use fall-through braches for the common
97 addi a4, a4, -1
100 .Ldst2mod4: # dst 16-bit aligned
109 addi a4, a4, -2
138 * Destination and source are word-aligned.
140 # copy 16 bytes per iteration for word-aligned dst and word-aligned src
205 # copy 16 bytes per iteration for word-aligned dst and unaligned src
207 sub a3, a3, a10 # align a3 (to avoid sim warnings only; not needed for hardware)
292 * bytes_copied = a5 - a2
293 * retval = bytes_not_copied = original len - bytes_copied
294 * retval = a11 - (a5 - a2)
299 sub a2, a5, a2 /* a2 <-- bytes copied */
300 sub a2, a11, a2 /* a2 <-- bytes not copied */