diff options
Diffstat (limited to 'kernel/power/zscal.S')
-rw-r--r-- | kernel/power/zscal.S | 385 |
1 files changed, 385 insertions, 0 deletions
diff --git a/kernel/power/zscal.S b/kernel/power/zscal.S new file mode 100644 index 000000000..7ffa80f19 --- /dev/null +++ b/kernel/power/zscal.S @@ -0,0 +1,385 @@ +/*********************************************************************/ +/* Copyright 2009, 2010 The University of Texas at Austin. */ +/* All rights reserved. */ +/* */ +/* 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 UNIVERSITY OF TEXAS AT */ +/* AUSTIN ``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 UNIVERSITY OF TEXAS AT */ +/* AUSTIN 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. */ +/* */ +/* The views and conclusions contained in the software and */ +/* documentation are those of the authors and should not be */ +/* interpreted as representing official policies, either expressed */ +/* or implied, of The University of Texas at Austin. */ +/*********************************************************************/ + +#define ASSEMBLER +#include "common.h" + +#define N r3 +#define XX r4 +#define PREA r5 + +#ifdef linux +#ifndef __64BIT__ +#define X r6 +#define INCX r7 +#else +#define X r8 +#define INCX r9 +#endif +#endif + +#if defined(_AIX) || defined(__APPLE__) +#if !defined(__64BIT__) && defined(DOUBLE) +#define X r10 +#define INCX r8 +#else +#define X r8 +#define INCX r9 +#endif +#endif + +#define FZERO f0 +#define ALPHA_R f1 +#define ALPHA_I f2 + + PROLOGUE + PROFCODE + + addi SP, SP, -8 + li r0, 0 + + stw r0, 0(SP) + lfs FZERO, 0(SP) + addi SP, SP, 8 + +#if (defined(_AIX) || defined(__APPLE__)) && !defined(__64BIT__) && defined(DOUBLE) + lwz INCX, 56(SP) +#endif + + slwi INCX, INCX, ZBASE_SHIFT + + li PREA, L1_PREFETCHSIZE + + cmpwi cr0, N, 0 + blelr- cr0 + + fcmpu cr0, FZERO, ALPHA_R + bne- cr0, LL(A1I1) + + fcmpu cr0, FZERO, ALPHA_I + bne- cr0, LL(A1I1) + + cmpwi cr0, INCX, 2 * SIZE + bne- cr0, LL(A0IN) + + srawi. r0, N, 3 + mtspr CTR, r0 + beq- cr0, LL(A0I1_Remain) + .align 4 + +LL(A0I1_kernel): + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + STFD FZERO, 2 * SIZE(X) + STFD FZERO, 3 * SIZE(X) + STFD FZERO, 4 * SIZE(X) + STFD FZERO, 5 * SIZE(X) + STFD FZERO, 6 * SIZE(X) + STFD FZERO, 7 * SIZE(X) + + STFD FZERO, 8 * SIZE(X) + STFD FZERO, 9 * SIZE(X) + STFD FZERO, 10 * SIZE(X) + STFD FZERO, 11 * SIZE(X) + STFD FZERO, 12 * SIZE(X) + STFD FZERO, 13 * SIZE(X) + STFD FZERO, 14 * SIZE(X) + STFD FZERO, 15 * SIZE(X) + + addi X, X, 16 * SIZE + bdnz LL(A0I1_kernel) + .align 4 + +LL(A0I1_Remain): + andi. r0, N, 7 + mtspr CTR, r0 + beqlr+ + .align 4 + +LL(A0I1_RemainKernel): + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + addi X, X, 2 * SIZE + bdnz LL(A0I1_RemainKernel) + blr + .align 4 + +LL(A0IN): + srawi. r0, N, 3 + mtspr CTR, r0 + beq- LL(A0IN_Remain) + .align 4 + +LL(A0IN_Kernel): + dcbtst X, PREA + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + bdnz LL(A0IN_Kernel) + .align 4 + +LL(A0IN_Remain): + andi. r0, N, 7 + mtspr CTR, r0 + beqlr+ + .align 4 + +LL(A0IN_RemainKernel): + STFD FZERO, 0 * SIZE(X) + STFD FZERO, 1 * SIZE(X) + add X, X, INCX + bdnz LL(A0IN_RemainKernel) + blr + .align 4 + +LL(A1I1): + cmpwi cr0, INCX, 2 * SIZE + bne- LL(A1IN) + + mr XX, X + srawi. r0, N, 3 + mtspr CTR, r0 + beq+ LL(A1I1_Remain) + .align 4 + +LL(A1I1_kernel): + LFD f3, 0 * SIZE(X) + LFD f4, 1 * SIZE(X) + LFD f5, 2 * SIZE(X) + LFD f6, 3 * SIZE(X) + LFD f7, 4 * SIZE(X) + LFD f8, 5 * SIZE(X) + LFD f9, 6 * SIZE(X) + LFD f10, 7 * SIZE(X) + + FMUL f0, ALPHA_I, f4 + FMUL f4, ALPHA_R, f4 + FMUL f11, ALPHA_I, f6 + FMUL f6, ALPHA_R, f6 + + FMUL f12, ALPHA_I, f8 + FMUL f8, ALPHA_R, f8 + FMUL f13, ALPHA_I, f10 + FMUL f10, ALPHA_R, f10 + + FMADD f4, ALPHA_I, f3, f4 + FMSUB f3, ALPHA_R, f3, f0 + FMADD f6, ALPHA_I, f5, f6 + FMSUB f5, ALPHA_R, f5, f11 + + FMADD f8, ALPHA_I, f7, f8 + FMSUB f7, ALPHA_R, f7, f12 + FMADD f10, ALPHA_I, f9, f10 + FMSUB f9, ALPHA_R, f9, f13 + + STFD f3, 0 * SIZE(X) + STFD f4, 1 * SIZE(X) + STFD f5, 2 * SIZE(X) + STFD f6, 3 * SIZE(X) + STFD f7, 4 * SIZE(X) + STFD f8, 5 * SIZE(X) + STFD f9, 6 * SIZE(X) + STFD f10, 7 * SIZE(X) + + LFD f3, 8 * SIZE(X) + LFD f4, 9 * SIZE(X) + LFD f5, 10 * SIZE(X) + LFD f6, 11 * SIZE(X) + LFD f7, 12 * SIZE(X) + LFD f8, 13 * SIZE(X) + LFD f9, 14 * SIZE(X) + LFD f10,15 * SIZE(X) + + FMUL f0, ALPHA_I, f4 + FMUL f4, ALPHA_R, f4 + FMUL f11, ALPHA_I, f6 + FMUL f6, ALPHA_R, f6 + + FMUL f12, ALPHA_I, f8 + FMUL f8, ALPHA_R, f8 + FMUL f13, ALPHA_I, f10 + FMUL f10, ALPHA_R, f10 + + FMADD f4, ALPHA_I, f3, f4 + FMSUB f3, ALPHA_R, f3, f0 + FMADD f6, ALPHA_I, f5, f6 + FMSUB f5, ALPHA_R, f5, f11 + + FMADD f8, ALPHA_I, f7, f8 + FMSUB f7, ALPHA_R, f7, f12 + FMADD f10, ALPHA_I, f9, f10 + FMSUB f9, ALPHA_R, f9, f13 + + STFD f3, 8 * SIZE(X) + STFD f4, 9 * SIZE(X) + STFD f5, 10 * SIZE(X) + STFD f6, 11 * SIZE(X) + STFD f7, 12 * SIZE(X) + STFD f8, 13 * SIZE(X) + STFD f9, 14 * SIZE(X) + STFD f10,15 * SIZE(X) + + addi X, X, 16 * SIZE + dcbtst X, PREA + bdnz LL(A1I1_kernel) + .align 4 + +LL(A1I1_Remain): + andi. r0, N, 7 + mtspr CTR, r0 + beqlr+ + .align 4 + +LL(A1I1_RemainKernel): + LFD f3, 0 * SIZE(X) + LFD f4, 1 * SIZE(X) + + FMUL f5, ALPHA_I, f4 + FMUL f4, ALPHA_R, f4 + FMADD f4, ALPHA_I, f3, f4 + FMSUB f3, ALPHA_R, f3, f5 + + STFD f3, 0 * SIZE(X) + STFD f4, 1 * SIZE(X) + addi X, X, 2 * SIZE + bdnz LL(A1I1_RemainKernel) + blr + .align 4 + +LL(A1IN): + mr XX, X + + srawi. r0, N, 2 + mtspr CTR, r0 + beq- LL(A1IN_Remain) + .align 4 + +LL(A1IN_Kernel): + LFD f3, 0 * SIZE(XX) + LFD f4, 1 * SIZE(XX) + add XX, XX, INCX + LFD f5, 0 * SIZE(XX) + LFD f6, 1 * SIZE(XX) + add XX, XX, INCX + LFD f7, 0 * SIZE(XX) + LFD f8, 1 * SIZE(XX) + add XX, XX, INCX + LFD f9, 0 * SIZE(XX) + LFD f10, 1 * SIZE(XX) + add XX, XX, INCX + + FMUL f0, ALPHA_I, f4 + FMUL f4, ALPHA_R, f4 + FMUL f11, ALPHA_I, f6 + FMUL f6, ALPHA_R, f6 + + FMUL f12, ALPHA_I, f8 + FMUL f8, ALPHA_R, f8 + FMUL f13, ALPHA_I, f10 + FMUL f10, ALPHA_R, f10 + + FMADD f4, ALPHA_I, f3, f4 + FMSUB f3, ALPHA_R, f3, f0 + FMADD f6, ALPHA_I, f5, f6 + FMSUB f5, ALPHA_R, f5, f11 + + FMADD f8, ALPHA_I, f7, f8 + FMSUB f7, ALPHA_R, f7, f12 + FMADD f10, ALPHA_I, f9, f10 + FMSUB f9, ALPHA_R, f9, f13 + + STFD f3, 0 * SIZE(X) + STFD f4, 1 * SIZE(X) + add X, X, INCX + STFD f5, 0 * SIZE(X) + STFD f6, 1 * SIZE(X) + add X, X, INCX + STFD f7, 0 * SIZE(X) + STFD f8, 1 * SIZE(X) + add X, X, INCX + STFD f9, 0 * SIZE(X) + STFD f10, 1 * SIZE(X) + add X, X, INCX + dcbtst X, PREA + bdnz LL(A1IN_Kernel) + .align 4 + +LL(A1IN_Remain): + andi. r0, N, 3 + mtspr CTR, r0 + beqlr+ + .align 4 + +LL(A1IN_RemainKernel): + LFD f3, 0 * SIZE(XX) + LFD f4, 1 * SIZE(XX) + add XX, XX, INCX + + FMUL f5, ALPHA_I, f4 + FMUL f4, ALPHA_R, f4 + FMADD f4, ALPHA_I, f3, f4 + FMSUB f3, ALPHA_R, f3, f5 + + STFD f3, 0 * SIZE(X) + STFD f4, 1 * SIZE(X) + add X, X, INCX + bdnz LL(A1IN_RemainKernel) + blr + + EPILOGUE |