/* $NetBSD: pmap_subr.s,v 1.1 2021/07/10 20:22:37 thorpej Exp $ */ /*- * Copyright (c) 2021 The NetBSD Foundation, Inc. * All rights reserved. * * This code is derived from software contributed to The NetBSD Foundation * by Jason R. Thorpe. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE NETBSD FOUNDATION, INC. AND CONTRIBUTORS * ``AS IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED * TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR * PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE FOUNDATION OR CONTRIBUTORS * BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR * CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF * SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS * INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN * CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) * ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE * POSSIBILITY OF SUCH DAMAGE. */ __KERNEL_RCSID(7, "$NetBSD: pmap_subr.s,v 1.1 2021/07/10 20:22:37 thorpej Exp $") /* * Optimized pmap subroutines. */ .text inc7: .stabs __FILE__,132,0,0,inc7; .loc 1 __LINE__ /* * pmap_zero_page: [ INTERFACE ] * * Zero the specified (machine independent) page by mapping the page * into virtual memory and clear its contents, one machine dependent * page at a time. * * Note: no locking is necessary in this function. */ .p2align 4 LEAF(pmap_zero_page, 1) /* No global references - skip LDGP() */ /* * Code here is arranged to keep branch targets on 16-byte * boundaries, minimize result latencies in the loop, unroll * the loop to at least 20 insns, and to dual-issue when * feasible. * * In the setup, we use nop and unop to minimize pipline stalls * on dependent instruction pairs. */ /* ---- */ lda t0, -1 nop sll t0, 42, t0 /* t0 = ALPHA_K0SEG_BASE */ /* * Loop counter: * PAGE_SIZE / 8 bytes per store / 16 stores per iteration */ lda v0, ((ALPHA_PGBYTES / 8) / 16) /* ---- */ or a0, t0, a0 /* a0 = ALPHA_PHYS_TO_K0SEG(a0) */ nop addq a0, (8*8), a2 /* a2 = a0 + 8-quads */ unop /* ---- */ 1: stq zero, (0*8)(a0) /* 0 */ stq zero, (1*8)(a0) /* 1 */ stq zero, (2*8)(a0) /* 2 */ stq zero, (3*8)(a0) /* 3 */ /* ---- */ stq zero, (4*8)(a0) /* 4 */ stq zero, (5*8)(a0) /* 5 */ stq zero, (6*8)(a0) /* 6 */ stq zero, (7*8)(a0) /* 7 */ /* ---- */ addq a2, (8*8), a0 /* a0 = a2 + 8-quads */ stq zero, (0*8)(a2) /* 8 */ stq zero, (1*8)(a2) /* 9 */ stq zero, (2*8)(a2) /* 10 */ /* --- */ subq v0, 1, v0 /* count-- */ stq zero, (3*8)(a2) /* 11 */ stq zero, (4*8)(a2) /* 12 */ stq zero, (5*8)(a2) /* 13 */ /* ---- */ stq zero, (6*8)(a2) /* 14 */ stq zero, (7*8)(a2) /* 15 */ addq a0, (8*8), a2 /* a2 = a0 + 8-quads */ bne v0, 1b /* loop around if count != 0 */ /* ---- */ RET END(pmap_zero_page) /* * pmap_copy_page: [ INTERFACE ] * * Copy the specified (machine independent) page by mapping the page * into virtual memory and copying the page, one machine dependent * page at a time. * * Note: no locking is necessary in this function. */ .p2align 4 LEAF(pmap_copy_page, 2) /* No global references - skip LDGP() */ /* See above. */ /* ---- */ lda t0, -1 nop sll t0, 42, t0 /* t0 = ALPHA_K0SEG_BASE */ /* * Loop counter: * PAGE_SIZE / 8 bytes per store / 8 stores per iteration */ lda v0, ((ALPHA_PGBYTES / 8) / 8) /* ---- */ or a0, t0, a0 /* a0 = ALPHA_PHYS_TO_K0SEG(a0) */ unop or a1, t0, a1 /* a1 = ALPHA_PHYS_TO_K0SEG(a1) */ unop /* ---- */ 1: ldq t0, (0*8)(a0) /* load 0 */ ldq t1, (1*8)(a0) /* load 1 */ ldq t2, (2*8)(a0) /* load 2 */ ldq t3, (3*8)(a0) /* load 3 */ /* ---- */ ldq t4, (4*8)(a0) /* load 4 */ ldq t5, (5*8)(a0) /* load 5 */ ldq t6, (6*8)(a0) /* load 6 */ ldq t7, (7*8)(a0) /* load 7 */ /* ---- */ addq a0, (8*8), a0 /* a0 = a0 + 8-quads */ stq t0, (0*8)(a1) /* store 0 */ stq t1, (1*8)(a1) /* store 1 */ stq t2, (2*8)(a1) /* store 2 */ /* ---- */ subq v0, 1, v0 /* count-- */ stq t3, (3*8)(a1) /* store 3 */ stq t4, (4*8)(a1) /* store 4 */ stq t5, (5*8)(a1) /* store 5 */ /* ---- */ stq t6, (6*8)(a1) /* store 6 */ stq t7, (7*8)(a1) /* store 7 */ addq a1, (8*8), a1 /* a1 = a1 + 8-quads */ bne v0, 1b /* loop around if count != 0 */ /* ---- */ RET END(pmap_copy_page)
