Nyaaori/rockbox
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity
rockbox/firmware/asm/m68k/ffs.S

79 lines
3 KiB
ArmAsm
Raw Normal View History

Add a complete priority inheritance implementation to the scheduler (all mutex ownership and queue_send calls are inheritable). Priorities are differential so that dispatch depends on the runnable range of priorities. Codec priority can therefore be raised in small steps (pcmbuf updated to enable). Simplify the kernel functions to ease implementation and use the same kernel.c for both sim and target (I'm tired of maintaining two ;_). 1) Not sure if a minor audio break at first buffering issue will exist on large-sector disks (the main mutex speed issue was genuinely resolved earlier). At this point it's best dealt with at the buffering level. It seems a larger filechunk could be used again. 2) Perhaps 64-bit sims will have some minor issues (finicky) but a backroll of the code of concern there is a 5-minute job. All kernel objects become incompatible so a full rebuild and update is needed. git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16791 a1c6a512-1295-4272-9138-f99709370657
2008-03-25 02:34:12 +00:00
/***************************************************************************
* __________ __ ___.
* Open \______ \ ____ ____ | | _\_ |__ _______ ___
* Source | _// _ \_/ ___\| |/ /| __ \ / _ \ \/ /
* Jukebox | | ( <_> ) \___| < | \_\ ( <_> > < <
* Firmware |____|_ /\____/ \___ >__|_ \|___ /\____/__/\_ \
* \/ \/ \/ \/ \/
* $Id$
*
* Copyright (C) 2008 by Michael Sevakis
*
Updated our source code header to explicitly mention that we are GPL v2 or later. We still need to hunt down snippets used that are not. 1324 modified files... http://www.rockbox.org/mail/archive/rockbox-dev-archive-2008-06/0060.shtml git-svn-id: svn://svn.rockbox.org/rockbox/trunk@17847 a1c6a512-1295-4272-9138-f99709370657
2008-06-28 18:10:04 +00:00
* This program is free software; you can redistribute it and/or
* modify it under the terms of the GNU General Public License
* as published by the Free Software Foundation; either version 2
* of the License, or (at your option) any later version.
Add a complete priority inheritance implementation to the scheduler (all mutex ownership and queue_send calls are inheritable). Priorities are differential so that dispatch depends on the runnable range of priorities. Codec priority can therefore be raised in small steps (pcmbuf updated to enable). Simplify the kernel functions to ease implementation and use the same kernel.c for both sim and target (I'm tired of maintaining two ;_). 1) Not sure if a minor audio break at first buffering issue will exist on large-sector disks (the main mutex speed issue was genuinely resolved earlier). At this point it's best dealt with at the buffering level. It seems a larger filechunk could be used again. 2) Perhaps 64-bit sims will have some minor issues (finicky) but a backroll of the code of concern there is a 5-minute job. All kernel objects become incompatible so a full rebuild and update is needed. git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16791 a1c6a512-1295-4272-9138-f99709370657
2008-03-25 02:34:12 +00:00
*
* This software is distributed on an "AS IS" basis, WITHOUT WARRANTY OF ANY
* KIND, either express or implied.
*
****************************************************************************/
#include "config.h"
/****************************************************************************
* int find_first_set_bit(uint32_t val);
*
* Find the index of the least significant set bit in the 32-bit word.
*
* return values:
* 0 - bit 0 is set
* 1 - bit 1 is set
* ...
* 31 - bit 31 is set
* 32 - no bits set
****************************************************************************/
.text
.align 2
.global find_first_set_bit
.type find_first_set_bit,@function
find_first_set_bit:
| this is a coldfire version of the ffs algorithm devised by D.Seal
| and posted to comp.sys.arm on 16 Feb 1994.
|
| Output modified to suit rockbox purposes.
| Standard trick to isolate bottom bit in r0 or 0 if r0 = 0 on entry
move.l 4(%sp), %d1 | %d1 = %d1 & -%d1
lea.l L_ffs_table, %a0 | %a0 = table address
move.l %d1, %d0 |
neg.l %d1 |
and.l %d0, %d1 |
| now %d1 has at most one set bit, call this X
move.l #0x0450fbaf, %d0 | %d0 = multiplier
mulu.l %d0, %d1 | %d1 = X * 0x0450fbaf
| now lookup in table indexed on top 6 bits of %d0
moveq.l #26, %d0 | %d0 = final shift count
lsr.l %d0, %d1 |
Remove unneeded instruction from the coldfire ffs function and fix a wrong comment git-svn-id: svn://svn.rockbox.org/rockbox/trunk@24030 a1c6a512-1295-4272-9138-f99709370657
2009-12-16 16:39:24 +00:00
| the top 24 bits of d0 are 0 so just load the value over it
Add a complete priority inheritance implementation to the scheduler (all mutex ownership and queue_send calls are inheritable). Priorities are differential so that dispatch depends on the runnable range of priorities. Codec priority can therefore be raised in small steps (pcmbuf updated to enable). Simplify the kernel functions to ease implementation and use the same kernel.c for both sim and target (I'm tired of maintaining two ;_). 1) Not sure if a minor audio break at first buffering issue will exist on large-sector disks (the main mutex speed issue was genuinely resolved earlier). At this point it's best dealt with at the buffering level. It seems a larger filechunk could be used again. 2) Perhaps 64-bit sims will have some minor issues (finicky) but a backroll of the code of concern there is a 5-minute job. All kernel objects become incompatible so a full rebuild and update is needed. git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16791 a1c6a512-1295-4272-9138-f99709370657
2008-03-25 02:34:12 +00:00
move.b (%a0, %d1.l), %d0 |
rts |
find_first_set_bit tweaks Correct sizing in m68k asm. Make the gerneric version far more generic. Change-Id: I32945f7431771979b3fe0da5472bdb110fd054ae
2012-01-27 16:47:49 +00:00
.size find_first_set_bit, .-find_first_set_bit
Add a complete priority inheritance implementation to the scheduler (all mutex ownership and queue_send calls are inheritable). Priorities are differential so that dispatch depends on the runnable range of priorities. Codec priority can therefore be raised in small steps (pcmbuf updated to enable). Simplify the kernel functions to ease implementation and use the same kernel.c for both sim and target (I'm tired of maintaining two ;_). 1) Not sure if a minor audio break at first buffering issue will exist on large-sector disks (the main mutex speed issue was genuinely resolved earlier). At this point it's best dealt with at the buffering level. It seems a larger filechunk could be used again. 2) Perhaps 64-bit sims will have some minor issues (finicky) but a backroll of the code of concern there is a 5-minute job. All kernel objects become incompatible so a full rebuild and update is needed. git-svn-id: svn://svn.rockbox.org/rockbox/trunk@16791 a1c6a512-1295-4272-9138-f99709370657
2008-03-25 02:34:12 +00:00
Move optimized ffs to firmware/asm, using the new automatic-asm-picking infrastructure. Change-Id: I5e7bdc8be7de50ef604c16078857fff1b84650dc
2012-01-07 23:22:58 +00:00
.section .irodata
L_ffs_table:
| 0 1 2 3 4 5 6 7
|----------------------------------------------
.byte 32, 0, 1, 12, 2, 6, 0, 13 | 0- 7
.byte 3, 0, 7, 0, 0, 0, 0, 14 | 8-15
.byte 10, 4, 0, 0, 8, 0, 0, 25 | 16-23
.byte 0, 0, 0, 0, 0, 21, 27, 15 | 24-31
.byte 31, 11, 5, 0, 0, 0, 0, 0 | 32-39
.byte 9, 0, 0, 24, 0, 0, 20, 26 | 40-47
.byte 30, 0, 0, 0, 0, 23, 0, 19 | 48-55
.byte 29, 0, 22, 18, 28, 17, 16, 0 | 56-63
find_first_set_bit tweaks Correct sizing in m68k asm. Make the gerneric version far more generic. Change-Id: I32945f7431771979b3fe0da5472bdb110fd054ae
2012-01-27 16:47:49 +00:00
.size L_ffs_table, .-L_ffs_table
Reference in a new issue Copy permalink