diff options
-rw-r--r-- | arch/arm64/include/asm/fpsimdmacros.h | 64 | ||||
-rw-r--r-- | arch/arm64/kernel/entry-fpsimd.S | 43 |
2 files changed, 67 insertions, 40 deletions
diff --git a/arch/arm64/include/asm/fpsimdmacros.h b/arch/arm64/include/asm/fpsimdmacros.h new file mode 100644 index 0000000..bbec599 --- /dev/null +++ b/arch/arm64/include/asm/fpsimdmacros.h @@ -0,0 +1,64 @@ +/* + * FP/SIMD state saving and restoring macros + * + * Copyright (C) 2012 ARM Ltd. + * Author: Catalin Marinas <catalin.marinas@arm.com> + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License version 2 as + * published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program. If not, see <http://www.gnu.org/licenses/>. + */ + +.macro fpsimd_save state, tmpnr + stp q0, q1, [\state, #16 * 0] + stp q2, q3, [\state, #16 * 2] + stp q4, q5, [\state, #16 * 4] + stp q6, q7, [\state, #16 * 6] + stp q8, q9, [\state, #16 * 8] + stp q10, q11, [\state, #16 * 10] + stp q12, q13, [\state, #16 * 12] + stp q14, q15, [\state, #16 * 14] + stp q16, q17, [\state, #16 * 16] + stp q18, q19, [\state, #16 * 18] + stp q20, q21, [\state, #16 * 20] + stp q22, q23, [\state, #16 * 22] + stp q24, q25, [\state, #16 * 24] + stp q26, q27, [\state, #16 * 26] + stp q28, q29, [\state, #16 * 28] + stp q30, q31, [\state, #16 * 30]! + mrs x\tmpnr, fpsr + str w\tmpnr, [\state, #16 * 2] + mrs x\tmpnr, fpcr + str w\tmpnr, [\state, #16 * 2 + 4] +.endm + +.macro fpsimd_restore state, tmpnr + ldp q0, q1, [\state, #16 * 0] + ldp q2, q3, [\state, #16 * 2] + ldp q4, q5, [\state, #16 * 4] + ldp q6, q7, [\state, #16 * 6] + ldp q8, q9, [\state, #16 * 8] + ldp q10, q11, [\state, #16 * 10] + ldp q12, q13, [\state, #16 * 12] + ldp q14, q15, [\state, #16 * 14] + ldp q16, q17, [\state, #16 * 16] + ldp q18, q19, [\state, #16 * 18] + ldp q20, q21, [\state, #16 * 20] + ldp q22, q23, [\state, #16 * 22] + ldp q24, q25, [\state, #16 * 24] + ldp q26, q27, [\state, #16 * 26] + ldp q28, q29, [\state, #16 * 28] + ldp q30, q31, [\state, #16 * 30]! + ldr w\tmpnr, [\state, #16 * 2] + msr fpsr, x\tmpnr + ldr w\tmpnr, [\state, #16 * 2 + 4] + msr fpcr, x\tmpnr +.endm diff --git a/arch/arm64/kernel/entry-fpsimd.S b/arch/arm64/kernel/entry-fpsimd.S index 17988a6..6a27cd6 100644 --- a/arch/arm64/kernel/entry-fpsimd.S +++ b/arch/arm64/kernel/entry-fpsimd.S @@ -20,6 +20,7 @@ #include <linux/linkage.h> #include <asm/assembler.h> +#include <asm/fpsimdmacros.h> /* * Save the FP registers. @@ -27,26 +28,7 @@ * x0 - pointer to struct fpsimd_state */ ENTRY(fpsimd_save_state) - stp q0, q1, [x0, #16 * 0] - stp q2, q3, [x0, #16 * 2] - stp q4, q5, [x0, #16 * 4] - stp q6, q7, [x0, #16 * 6] - stp q8, q9, [x0, #16 * 8] - stp q10, q11, [x0, #16 * 10] - stp q12, q13, [x0, #16 * 12] - stp q14, q15, [x0, #16 * 14] - stp q16, q17, [x0, #16 * 16] - stp q18, q19, [x0, #16 * 18] - stp q20, q21, [x0, #16 * 20] - stp q22, q23, [x0, #16 * 22] - stp q24, q25, [x0, #16 * 24] - stp q26, q27, [x0, #16 * 26] - stp q28, q29, [x0, #16 * 28] - stp q30, q31, [x0, #16 * 30]! - mrs x8, fpsr - str w8, [x0, #16 * 2] - mrs x8, fpcr - str w8, [x0, #16 * 2 + 4] + fpsimd_save x0, 8 ret ENDPROC(fpsimd_save_state) @@ -56,25 +38,6 @@ ENDPROC(fpsimd_save_state) * x0 - pointer to struct fpsimd_state */ ENTRY(fpsimd_load_state) - ldp q0, q1, [x0, #16 * 0] - ldp q2, q3, [x0, #16 * 2] - ldp q4, q5, [x0, #16 * 4] - ldp q6, q7, [x0, #16 * 6] - ldp q8, q9, [x0, #16 * 8] - ldp q10, q11, [x0, #16 * 10] - ldp q12, q13, [x0, #16 * 12] - ldp q14, q15, [x0, #16 * 14] - ldp q16, q17, [x0, #16 * 16] - ldp q18, q19, [x0, #16 * 18] - ldp q20, q21, [x0, #16 * 20] - ldp q22, q23, [x0, #16 * 22] - ldp q24, q25, [x0, #16 * 24] - ldp q26, q27, [x0, #16 * 26] - ldp q28, q29, [x0, #16 * 28] - ldp q30, q31, [x0, #16 * 30]! - ldr w8, [x0, #16 * 2] - ldr w9, [x0, #16 * 2 + 4] - msr fpsr, x8 - msr fpcr, x9 + fpsimd_restore x0, 8 ret ENDPROC(fpsimd_load_state) |