Import 2.3.23pre2

Documentation/Changes

@@ -43,7 +43,7 @@ Current Minimal Requirements
 encountered a bug!  If you're unsure what version you're currently
 running, the suggested command should tell you.
 
-- Kernel modutils	 2.3.5                   ; insmod -V
+- Kernel modutils	 2.3.6                   ; insmod -V
 - Gnu C 		 2.7.2.3                 ; gcc --version
 - Binutils		 2.8.1.0.23              ; ld -v
 - Linux libc5 C Library	 5.4.46                  ; ls -l /lib/libc*
@@ -567,8 +567,8 @@ ftp://metalab.unc.edu/pub/Linux/GCC/ld.so-1.9.9.tar.gz
 Modules utilities
 =================
 
-The 2.3.5 release:
-ftp://ftp.ocs.com.au/pub/modutils/v2.3/modutils-2.3.5.tar.gz
+The 2.3.6 release:
+ftp://ftp.ocs.com.au/pub/modutils/v2.3/modutils-2.3.6.tar.gz
 
 Procps utilities
 ================

Documentation/vm/locking

+Started Oct 1999 by Kanoj Sarcar <kanoj@sgi.com>
+
+The intent of this file is to have an uptodate, running commentary 
+from different people about how locking and synchronization is done 
+in the Linux vm code.
+
+vmlist_access_lock/vmlist_modify_lock
+--------------------------------------
+
+Page stealers pick processes out of the process pool and scan for 
+the best process to steal pages from. To guarantee the existance 
+of the victim mm, a mm_count inc and a mmdrop are done in swap_out().
+Page stealers hold kernel_lock to protect against a bunch of races.
+The vma list of the victim mm is also scanned by the stealer, 
+and the vmlist_lock is used to preserve list sanity against the
+process adding/deleting to the list. This also gurantees existance
+of the vma. Vma existance gurantee while invoking the driver
+swapout() method in try_to_swap_out() also relies on the fact
+that do_munmap() temporarily gets lock_kernel before decimating
+the vma, thus the swapout() method must snapshot all the vma 
+fields it needs before going to sleep (which will release the
+lock_kernel held by the page stealer). Currently, filemap_swapout
+is the only method that depends on this shaky interlocking.
+
+Any code that modifies the vmlist, or the vm_start/vm_end/
+vm_flags:VM_LOCKED/vm_next of any vma *in the list* must prevent 
+kswapd from looking at the chain. This does not include driver mmap() 
+methods, for example, since the vma is still not in the list.
+
+The rules are:
+1. To modify the vmlist (add/delete or change fields in an element), 
+you must hold mmap_sem to guard against clones doing mmap/munmap/faults, 
+(ie all vm system calls and faults), and from ptrace, swapin due to 
+swap deletion etc.
+2. To modify the vmlist (add/delete or change fields in an element), 
+you must also hold vmlist_modify_lock, to guard against page stealers 
+scanning the list.
+3. To scan the vmlist (find_vma()), you must either 
+        a. grab mmap_sem, which should be done by all cases except 
+	   page stealer.
+or
+        b. grab vmlist_access_lock, only done by page stealer.
+4. While holding the vmlist_modify_lock, you must be able to guarantee
+that no code path will lead to page stealing. A better guarantee is
+to claim non sleepability, which ensures that you are not sleeping
+for a lock, whose holder might in turn be doing page stealing.
+5. You must be able to guarantee that while holding vmlist_modify_lock
+or vmlist_access_lock of mm A, you will not try to get either lock
+for mm B.
+
+The caveats are:
+1. find_vma() makes use of, and updates, the mmap_cache pointer hint.
+The update of mmap_cache is racy (page stealer can race with other code
+that invokes find_vma with mmap_sem held), but that is okay, since it 
+is a hint. This can be fixed, if desired, by having find_vma grab the
+vmlist lock.
+
+
+Code that add/delete elements from the vmlist chain are
+1. callers of insert_vm_struct
+2. callers of merge_segments
+3. callers of avl_remove
+
+Code that changes vm_start/vm_end/vm_flags:VM_LOCKED of vma's on
+the list:
+1. expand_stack
+2. mprotect
+3. mlock
+4. mremap
+
+It is advisable that changes to vm_start/vm_end be protected, although 
+in some cases it is not really needed. Eg, vm_start is modified by 
+expand_stack(), it is hard to come up with a destructive scenario without 
+having the vmlist protection in this case.
+
+The vmlist lock nests with the inode i_shared_lock and the kmem cache
+c_spinlock spinlocks. This is okay, since code that holds i_shared_lock 
+never asks for memory, and the kmem code asks for pages after dropping
+c_spinlock.
+
+The vmlist lock can be a sleeping or spin lock. In either case, care
+must be taken that it is not held on entry to the driver methods, since
+those methods might sleep or ask for memory, causing deadlocks.
+
+The current implementation of the vmlist lock uses the page_table_lock,
+which is also the spinlock that page stealers use to protect changes to
+the victim process' ptes. Thus we have a reduction in the total number
+of locks. 

MAINTAINERS

@@ -831,6 +831,16 @@ STARMODE RADIO IP (STRIP) PROTOCOL DRIVER
 W:	http://mosquitonet.Stanford.EDU/strip.html
 S:	Unsupported ?
 
+SUPERH
+P:	Niibe Yutaka
+M:	gniibe@chroot.org
+P:	Kazumoto Kojima
+M:	kkojima@rr.iij4u.or.jp
+L:	linux-sh@m17n.org
+W:	http://www.m17n.org/linux-sh/
+W:	http://www.rr.iij4u.or.jp/~kkojima/linux-sh4.html
+S:	Maintained
+
 SVGA HANDLING
 P:	Martin Mares
 M:	mj@atrey.karlin.mff.cuni.cz

arch/arm/def-configs/footbridge

@@ -89,7 +89,7 @@ CONFIG_IDEDMA_AUTO=y
 # CONFIG_BLK_DEV_OPTI621 is not set
 # CONFIG_BLK_DEV_TRM290 is not set
 # CONFIG_BLK_DEV_NS87415 is not set
-# CONFIG_BLK_DEV_VIA82C586 is not set
+# CONFIG_BLK_DEV_VIA82CXXX is not set
 # CONFIG_BLK_DEV_CMD646 is not set
 CONFIG_BLK_DEV_SL82C105=y
 # CONFIG_IDE_CHIPSETS is not set

arch/arm/defconfig

@@ -89,7 +89,7 @@ CONFIG_IDEDMA_PCI_AUTO=y
 # CONFIG_BLK_DEV_OPTI621 is not set
 # CONFIG_BLK_DEV_TRM290 is not set
 # CONFIG_BLK_DEV_NS87415 is not set
-# CONFIG_BLK_DEV_VIA82C586 is not set
+# CONFIG_BLK_DEV_VIA82CXXX is not set
 # CONFIG_BLK_DEV_CMD646 is not set
 CONFIG_BLK_DEV_SL82C105=y
 # CONFIG_IDE_CHIPSETS is not set

arch/i386/Makefile

@@ -84,9 +84,6 @@ MAKEBOOT = $(MAKE) -C arch/$(ARCH)/boot
 
 vmlinux: arch/i386/vmlinux.lds
 
-arch/i386/vmlinux.lds: arch/i386/vmlinux.lds.S FORCE
-	$(CPP) -C -P -imacros $(HPATH)/asm-i386/page_offset.h -Ui386 arch/i386/vmlinux.lds.S >arch/i386/vmlinux.lds
-
 FORCE: ;
 
 .PHONY: zImage bzImage compressed zlilo bzlilo zdisk bzdisk install \
@@ -119,7 +116,6 @@ archclean:
 	@$(MAKEBOOT) clean
 
 archmrproper:
-	rm -f arch/i386/vmlinux.lds
 
 archdep:
 	@$(MAKEBOOT) dep

arch/i386/config.in

@@ -42,10 +42,6 @@ if [ "$CONFIG_MK7" = "y" ]; then
    define_bool CONFIG_X86_USE_3DNOW y
 fi
 
-choice 'Maximum Physical Memory' \
-	"1GB		CONFIG_1GB \
-	 2GB		CONFIG_2GB" 1GB
-
 bool 'Math emulation' CONFIG_MATH_EMULATION
 bool 'MTRR (Memory Type Range Register) support' CONFIG_MTRR
 bool 'Symmetric multi-processing support' CONFIG_SMP
@@ -63,7 +59,7 @@ endmenu
 mainmenu_option next_comment
 comment 'General setup'
 
-bool 'BIGMEM support' CONFIG_BIGMEM
+bool 'Support for over 1Gig of memory' CONFIG_BIGMEM
 bool 'Networking support' CONFIG_NET
 bool 'SGI Visual Workstation support' CONFIG_VISWS
 if [ "$CONFIG_VISWS" = "y" ]; then

arch/i386/kernel/pci-visws.c

@@ -14,6 +14,7 @@
 
 #include <asm/smp.h>
 #include <asm/lithium.h>
+#include <asm/io.h>
 
 #include "pci-i386.h"
 

arch/i386/kernel/visws_apic.c

@@ -37,7 +37,7 @@
 
 #include <asm/cobalt.h>
 
-#include "irq.h"
+#include <linux/irq.h>
 
 /*
  * This is the PIIX4-based 8259 that is wired up indirectly to Cobalt

arch/i386/vmlinux.lds.S → arch/i386/vmlinux.lds

@@ -6,7 +6,7 @@ OUTPUT_ARCH(i386)
 ENTRY(_start)
 SECTIONS
 {
-  . = PAGE_OFFSET_RAW + 0x100000;
+  . = 0xC0000000 + 0x100000;
   _text = .;			/* Text and read-only data */
   .text : {
 	*(.text)

arch/ppc/kernel/chrp_pci.c

@@ -16,7 +16,6 @@
 #include <asm/hydra.h>
 #include <asm/prom.h>
 #include <asm/gg2.h>
-#include <asm/ide.h>
 #include <asm/machdep.h>
 
 #include "pci.h"

arch/ppc/kernel/chrp_setup.c

@@ -28,6 +28,7 @@
 #include <linux/reboot.h>
 #include <linux/init.h>
 #include <linux/blk.h>
+#include <linux/ide.h>
 #include <linux/ioport.h>
 #include <linux/console.h>
 #include <linux/pci.h>
@@ -40,8 +41,6 @@
 #include <asm/processor.h>
 #include <asm/io.h>
 #include <asm/pgtable.h>
-#include <linux/ide.h>
-#include <asm/ide.h>
 #include <asm/prom.h>
 #include <asm/gg2.h>
 #include <asm/pci-bridge.h>

arch/ppc/kernel/mbx_setup.c

@@ -30,6 +30,7 @@
 #include <linux/reboot.h>
 #include <linux/init.h>
 #include <linux/blk.h>
+#include <linux/ide.h>
 #include <linux/ioport.h>
 
 #include <asm/mmu.h>
@@ -37,7 +38,6 @@
 #include <asm/residual.h>
 #include <asm/io.h>
 #include <asm/pgtable.h>
-#include <asm/ide.h>
 #include <asm/mbx.h>
 #include <asm/machdep.h>
 

arch/ppc/kernel/pmac_setup.c

@@ -39,6 +39,7 @@
 #include <linux/ioport.h>
 #include <linux/major.h>
 #include <linux/blk.h>
+#include <linux/ide.h>
 #include <linux/vt_kern.h>
 #include <linux/console.h>
 #include <linux/ide.h>
@@ -57,7 +58,6 @@
 #include <asm/ohare.h>
 #include <asm/mediabay.h>
 #include <asm/feature.h>
-#include <asm/ide.h>
 #include <asm/machdep.h>
 #include <asm/keyboard.h>
 #include <asm/dma.h>
@@ -421,7 +421,7 @@ note_scsi_host(struct device_node *node, void *host)
 #if defined(CONFIG_BLK_DEV_IDE) && defined(CONFIG_BLK_DEV_IDE_PMAC)
 extern int pmac_ide_count;
 extern struct device_node *pmac_ide_node[];
-static int ide_majors[] = { 3, 22, 33, 34, 56, 57, 88, 89 };
+static int ide_majors[] = { 3, 22, 33, 34, 56, 57, 88, 89, 90, 91 };
 
 kdev_t __init find_ide_boot(void)
 {

arch/ppc/kernel/prep_setup.c

@@ -41,7 +41,6 @@
 #include <asm/residual.h>
 #include <asm/io.h>
 #include <asm/pgtable.h>
-#include <asm/ide.h>
 #include <asm/cache.h>
 #include <asm/dma.h>
 #include <asm/machdep.h>

arch/ppc/kernel/setup.c

@@ -11,12 +11,11 @@
 #include <linux/reboot.h>
 #include <linux/delay.h>
 #include <linux/blk.h>
+#include <linux/ide.h>
 
 #include <asm/init.h>
 #include <asm/residual.h>
 #include <asm/io.h>
-#include <linux/ide.h>
-#include <asm/ide.h>
 #include <asm/prom.h>
 #include <asm/processor.h>
 #include <asm/pgtable.h>

arch/ppc/mbx_defconfig

@@ -74,7 +74,7 @@ CONFIG_IDEDMA_PCI_AUTO=y
 # CONFIG_BLK_DEV_OPTI621 is not set
 # CONFIG_BLK_DEV_TRM290 is not set
 # CONFIG_BLK_DEV_NS87415 is not set
-# CONFIG_BLK_DEV_VIA82C586 is not set
+# CONFIG_BLK_DEV_VIA82CXXX is not set
 # CONFIG_BLK_DEV_CMD646 is not set
 CONFIG_BLK_DEV_SL82C105=y
 # CONFIG_IDE_CHIPSETS is not set

arch/sh/Makefile

-# $Id$
+# $Id: Makefile,v 1.1 1999/09/18 16:55:51 gniibe Exp gniibe $
 #
 # This file is subject to the terms and conditions of the GNU General Public
 # License.  See the file "COPYING" in the main directory of this archive
 # for more details.
 #
+# Copyright (C) 1999  Kaz Kojima
+#
 # This file is included by the global makefile so that you can add your own
 # architecture-specific flags and dependencies. Remember to do have actions
 # for "archclean" and "archdep" for cleaning up and making dependencies for
@@ -13,25 +15,35 @@
 #
 # Select the object file format to substitute into the linker script.
 #
-tool-prefix	= sh-gniibe-
-oformat		= elf
+tool-prefix	= sh-elf
+
+ifdef CONFIG_LITTLE_ENDIAN
+CFLAGS		+= -ml
+AFLAGS		+= -ml
+# LINKFLAGS	+= -EL
+LDFLAGS		:= -EL
+
+LD	=$(CROSS_COMPILE)ld $(LDFLAGS)
+
+endif
 
 ifdef CONFIG_CROSSCOMPILE
 CROSS_COMPILE  = $(tool-prefix)
 endif
 
-LINKFLAGS	= # -EL # -static #-N
 MODFLAGS	+=
 
 #
 #
-CFLAGS		+= -m3 # -ml
-LINKFLAGS	+=
-LDFLAGS		+= # -EL
 
-#
-#
-HOSTCC        = cc
+ifdef CONFIG_CPU_SH3
+CFLAGS		+= -m3
+AFLAGS		+= -m3
+endif
+ifdef CONFIG_CPU_SH4
+CFLAGS		+= -m4
+AFLAGS		+= -m4
+endif
 
 #
 # Choosing incompatible machines durings configuration will result in
@@ -52,14 +64,16 @@ HEAD := arch/sh/kernel/head.o arch/sh/kernel/init_task.o
 
 SUBDIRS := $(SUBDIRS) $(addprefix arch/sh/, kernel mm lib)
 CORE_FILES := arch/sh/kernel/kernel.o arch/sh/mm/mm.o $(CORE_FILES)
-LIBS := $(TOPDIR)/arch/sh/lib/lib.a $(LIBS) $(TOPDIR)/arch/sh/lib/lib.a /home/niibe/lib/gcc-lib/sh-gniibe-elf/egcs-2.91.66/libgcc.a
+LIBGCC := $(shell $(CC) $(CFLAGS) -print-libgcc-file-name)
+LIBS := $(TOPDIR)/arch/sh/lib/lib.a $(LIBS) $(TOPDIR)/arch/sh/lib/lib.a \
+	$(LIBGCC)
 
 MAKEBOOT = $(MAKE) -C arch/$(ARCH)/boot
 
 vmlinux: arch/sh/vmlinux.lds
 
 arch/sh/vmlinux.lds: arch/sh/vmlinux.lds.S FORCE
-	gcc -E -C -P -I$(HPATH) -imacros $(HPATH)/linux/config.h -Ush arch/sh/vmlinux.lds.S >arch/sh/vmlinux.lds
+	gcc -E -C -P -I$(HPATH) -Ush arch/sh/vmlinux.lds.S >arch/sh/vmlinux.lds
 
 FORCE: ;
 
@@ -77,6 +91,7 @@ archclean:
 #	$(MAKE) -C arch/$(ARCH)/tools clean
 
 archmrproper:
+	rm -f arch/sh/vmlinux.lds
 
 archdep:
 	@$(MAKEBOOT) dep

arch/sh/config.in

@@ -12,28 +12,24 @@ endmenu
 mainmenu_option next_comment
 comment 'Processor type and features'
 choice 'Processor family' \
-	"SH3		CONFIG_CPU_SH3 \
-	 SH4		CONFIG_CPU_SH4" SH3
+   "SH-3	CONFIG_CPU_SH3 \
+    SH-4	CONFIG_CPU_SH4" SH-3
 bool 'Little Endian' CONFIG_LITTLE_ENDIAN
-hex 'Physical memory start address' CONFIG_MEMORY_START 0c000000
+hex 'Physical memory start address' CONFIG_MEMORY_START 08000000
+bool 'Use SH CPU internal real time clock' CONFIG_SH_CPU_RTC
 endmenu
 
 mainmenu_option next_comment
 comment 'Loadable module support'
 bool 'Enable loadable module support' CONFIG_MODULES
 if [ "$CONFIG_MODULES" = "y" ]; then
-   bool '  Set version information on all symbols for modules' CONFIG_MODVERSIONS
-   bool '  Kernel module loader' CONFIG_KMOD
+   bool 'Set version information on all symbols for modules' CONFIG_MODVERSIONS
+   bool 'Kernel module loader' CONFIG_KMOD
 fi
 endmenu
 
-define_bool CONFIG_SERIAL n
-define_bool CONFIG_SH3SCI_SERIAL y
-define_bool CONFIG_SERIAL_CONSOLE y
-
 mainmenu_option next_comment
-comment 'Floppy, IDE, and other block devices'
-
+comment 'General setup'
 bool 'Networking support' CONFIG_NET
 bool 'System V IPC' CONFIG_SYSVIPC
 bool 'BSD Process Accounting' CONFIG_BSD_PROCESS_ACCT
@@ -41,6 +37,18 @@ bool 'Sysctl support' CONFIG_SYSCTL
 
 tristate 'Kernel support for ELF binaries' CONFIG_BINFMT_ELF
 tristate 'Kernel support for MISC binaries' CONFIG_BINFMT_MISC
+endmenu
+
+mainmenu_option next_comment
+comment 'Character devices'
+define_bool CONFIG_SERIAL n
+define_bool CONFIG_SERIAL_CONSOLE y
+bool 'SuperH SCI support' CONFIG_SH_SCI_SERIAL
+bool 'SuperH SCIF support' CONFIG_SH_SCIF_SERIAL
+endmenu
+
+mainmenu_option next_comment
+comment 'Floppy, IDE, and other block devices'
 
 tristate 'RAM disk support' CONFIG_BLK_DEV_RAM
 if [ "$CONFIG_BLK_DEV_RAM" = "y" ]; then
@@ -75,4 +83,5 @@ mainmenu_option next_comment
 comment 'Kernel hacking'
 
 bool 'Magic SysRq key' CONFIG_MAGIC_SYSRQ
+bool 'GDB Stub kernel debug' CONFIG_DEBUG_KERNEL_WITH_GDB_STUB
 endmenu

arch/sh/defconfig

@@ -10,48 +10,41 @@
 #
 # Processor type and features
 #
-CONFIG_CPU_SH3=y
-# CONFIG_CPU_SH4 is not set
-# CONFIG_LITTLE_ENDIAN is not set
-CONFIG_MEMORY_START=0c000000
+# CONFIG_CPU_SH3 is not set
+CONFIG_CPU_SH4=y
+CONFIG_LITTLE_ENDIAN=y
+CONFIG_MEMORY_START=08000000
 
 #
 # Loadable module support
 #
 # CONFIG_MODULES is not set
-# CONFIG_SERIAL is not set
-CONFIG_SH3SCI_SERIAL=y
-CONFIG_SERIAL_CONSOLE=y
 
 #
-# Floppy, IDE, and other block devices
+# General setup
 #
 # CONFIG_NET is not set
-CONFIG_SYSVIPC=y
+# CONFIG_SYSVIPC is not set
 # CONFIG_BSD_PROCESS_ACCT is not set
 # CONFIG_SYSCTL is not set
 CONFIG_BINFMT_ELF=y
 # CONFIG_BINFMT_MISC is not set
-CONFIG_BLK_DEV_RAM=y
-CONFIG_BLK_DEV_INITRD=y
-# CONFIG_BLK_DEV_LOOP is not set
-# CONFIG_BLK_DEV_NBD is not set
 
 #
-# Networking options
+# Character devices
 #
-# CONFIG_PACKET is not set
-# CONFIG_NETLINK is not set
-# CONFIG_FIREWALL is not set
-# CONFIG_FILTER is not set
-# CONFIG_UNIX is not set
-# CONFIG_INET is not set
+# CONFIG_SERIAL is not set
+CONFIG_SERIAL_CONSOLE=y
+# CONFIG_SH_SCI_SERIAL is not set
+CONFIG_SH_SCIF_SERIAL=y
 
 #
-#  
+# Floppy, IDE, and other block devices
 #
-# CONFIG_IPX is not set
-# CONFIG_ATALK is not set
+CONFIG_BLK_DEV_RAM=y
+CONFIG_BLK_DEV_INITRD=y
+# CONFIG_BLK_DEV_LOOP is not set
+# CONFIG_BLK_DEV_NBD is not set
 
 #
 # Unix 98 PTY support
@@ -66,8 +59,12 @@ CONFIG_BLK_DEV_INITRD=y
 # CONFIG_AFFS_FS is not set
 # CONFIG_HFS_FS is not set
 # CONFIG_FAT_FS is not set
+# CONFIG_MSDOS_FS is not set
+# CONFIG_UMSDOS_FS is not set
+# CONFIG_VFAT_FS is not set
 # CONFIG_ISO9660_FS is not set
 # CONFIG_JOLIET is not set
+# CONFIG_UDF_FS is not set
 # CONFIG_MINIX_FS is not set
 # CONFIG_NTFS_FS is not set
 # CONFIG_HPFS_FS is not set
@@ -77,17 +74,16 @@ CONFIG_EXT2_FS=y
 # CONFIG_SYSV_FS is not set
 # CONFIG_UFS_FS is not set
 
-#
-# Network File Systems
-#
-
 #
 # Partition Types
 #
 # CONFIG_PARTITION_ADVANCED is not set
 CONFIG_MSDOS_PARTITION=y
-# CONFIG_SMD_DISKLABEL is not set
-# CONFIG_SGI_DISKLABEL is not set
+# CONFIG_BSD_DISKLABEL is not set
+# CONFIG_SOLARIS_X86_PARTITION is not set
+# CONFIG_UNIXWARE_DISKLABEL is not set
+# CONFIG_SGI_PARTITION is not set
+# CONFIG_SUN_PARTITION is not set
 # CONFIG_NLS is not set
 
 #
@@ -99,3 +95,4 @@ CONFIG_MSDOS_PARTITION=y
 # Kernel hacking
 #
 # CONFIG_MAGIC_SYSRQ is not set
+CONFIG_DEBUG_KERNEL_WITH_GDB_STUB=y

arch/sh/kernel/Makefile

@@ -11,7 +11,7 @@
 
 O_TARGET := kernel.o
 O_OBJS   := process.o signal.o entry.o traps.o irq.o irq_onchip.o \
-            ptrace.o setup.o time.o sys_sh.o test-img.o semaphore.o
+            ptrace.o setup.o time.o sys_sh.o semaphore.o
 OX_OBJS  := sh_ksyms.o
 MX_OBJS  :=
 

arch/sh/kernel/head.S

-/* $Id$
+/* $Id: head.S,v 1.6 1999/10/05 12:34:16 gniibe Exp $
  *
  *  arch/sh/kernel/head.S
  *
- *  Copyright (C) 1999  Niibe Yutaka
+ *  Copyright (C) 1999  Niibe Yutaka & Kaz Kojima
  *
  * This file is subject to the terms and conditions of the GNU General Public
  * License.  See the file "COPYING" in the main directory of this archive
@@ -10,60 +10,74 @@
  *
  * Head.S contains the SH exception handlers and startup code.
  */
-#include <linux/config.h>
-#include <linux/threads.h>
 #include <linux/linkage.h>
-#include <asm/page.h>
-#include <asm/pgtable.h>
 
-#ifdef CONFIG_CPU_SH3
-/* Following values are assumed to be as small as immediate. */
-#define CCR		0xffffffec	/* Address of Cache Control Register */
-#define CACHE_INIT	0x00000009	/* 8k-byte cache, flush, enable */
-#elif CONFIG_CPU_SH4
-/* Should fill here. */
-#endif
+	.section	.empty_zero_page, "aw"
+ENTRY(empty_zero_page)
+	.long	1		/* MOUNT_ROOT_RDONLY */
+	.long	0		/* RAMDISK_FLAGS */
+	.long	0x0200		/* ORIG_ROOT_DEV */
+	.long	1		/* LOADER_TYPE */
+	.long	0x88400000	/* INITRD_START */
+	.long	0x00400000	/* INITRD_SIZE */
+	.long	0x89000000	/* MEMORY_END */
+	.long	0
 
+	.text	
+	.balign 4096,0,4096
+/*
+ * Condition at the entry of _stext:
+ *
+ *   BSC has already been initialized.
+ *   INTC may or may not be initialized.
+ *   VBR may or may not be initialized.
+ *   MMU may or may not be initialized.
+ *   Cache may or may not be initialized.
+ *   Hardware (including on-chip modules) may or may not be initialized. 
+ *
+ *   The register R4&R5 holds the address of the parameter block, which has
+ *   command-line data, etc.
+ *
+ */
 ENTRY(_stext)
-	!			Switch to register bank 0
-	stc	sr,r1		!
-	mov.l	1f,r0		! RB=0, BL=1
-	and	r1,r0
-	ldc	r0,sr
-	!			Enable cache
-#ifdef CONFIG_CPU_SH3
-	mov	#CCR,r1
-	mov.l	@r1,r0
-	cmp/eq	#1,r0		! If it's enabled already, don't flush it
-	bt/s	8f
-	 mov	#CACHE_INIT,r0
-	mov.l	r0,@r1
-#elif CONFIG_CPU_SH4
-	! Should fill here.
+#if defined(__SH4__)
+	!			Initialize FPSCR	
+	/* GCC (as of 2.95.1) assumes FPU with double precision mode. */
+	mov.l	7f,r0
+	lds	r0,fpscr
 #endif
-8:
+	!			Initialize Status Register
+	mov.l	1f,r0		! MD=1, RB=0, BL=1
+	ldc	r0,sr
 	!
 	mov.l	2f,r0
 	mov	r0,r15		! Set initial r15 (stack pointer)
 	ldc	r0,r4_bank	!   and stack base
+	!
+	!			Enable cache
+	mov.l	6f,r0
+	jsr	@r0
+	 nop
 	!			Clear BSS area
 	mov.l	3f,r1
+	add	#4,r1
 	mov.l	4f,r2
 	mov	#0,r0
-9:	mov.l	r0,@r1
-	cmp/hs	r2,r1
-	bf/s	9b
-	 add	#4,r1
+9:	cmp/hs	r2,r1
+	bf/s	9b		! while (r1 < r2)
+	 mov.l	r0,@-r2
 	!			Start kernel
 	mov.l	5f,r0
 	jmp	@r0
 	 nop
 
 	.balign 4
-1:	.long 0xdfffffff		! RB=0, BL=1
-2:	.long SYMBOL_NAME(stack)
-3:	.long SYMBOL_NAME(__bss_start)
-4:	.long SYMBOL_NAME(_end)
-5:	.long SYMBOL_NAME(start_kernel)
-
-.data
+1:	.long	0x50000000		! MD=1, RB=0, BL=1
+2:	.long	SYMBOL_NAME(stack)
+3:	.long	SYMBOL_NAME(__bss_start)
+4:	.long	SYMBOL_NAME(_end)
+5:	.long	SYMBOL_NAME(start_kernel)
+6:	.long	SYMBOL_NAME(cache_init)
+#if defined(__SH4__)
+7:	.long   0x00080000
+#endif

arch/sh/kernel/irq.c

-/*
+/* $Id: irq.c,v 1.4 1999/10/11 13:12:14 gniibe Exp $
+ *
  * linux/arch/sh/kernel/irq.c
  *
  *	Copyright (C) 1992, 1998 Linus Torvalds, Ingo Molnar
  *
  *
- *  SuperH version:  Copyright (C) 1999  Niibe Yutaka
+ * SuperH version:  Copyright (C) 1999  Niibe Yutaka
  */
 
 /*
@@ -48,7 +49,7 @@ spinlock_t irq_controller_lock = SPIN_LOCK_UNLOCKED;
 /*
  * Controller mappings for all interrupt sources:
  */
-irq_desc_t irq_desc[NR_IRQS] = { [0 ... NR_IRQS-1] = { 0, &no_irq_type, }};
+irq_desc_t irq_desc[NR_IRQS] __cacheline_aligned = { [0 ... NR_IRQS-1] = { 0, &no_irq_type, }};
 
 /*
  * Special irq handlers.
@@ -56,6 +57,37 @@ irq_desc_t irq_desc[NR_IRQS] = { [0 ... NR_IRQS-1] = { 0, &no_irq_type, }};
 
 void no_action(int cpl, void *dev_id, struct pt_regs *regs) { }
 
+/*
+ * Generic no controller code
+ */
+
+static void enable_none(unsigned int irq) { }
+static unsigned int startup_none(unsigned int irq) { return 0; }
+static void disable_none(unsigned int irq) { }
+static void ack_none(unsigned int irq)
+{
+/*
+ * 'what should we do if we get a hw irq event on an illegal vector'.
+ * each architecture has to answer this themselves, it doesnt deserve
+ * a generic callback i think.
+ */
+	printk("unexpected IRQ trap at vector %02x\n", irq);
+}
+
+/* startup is the same as "enable", shutdown is same as "disable" */
+#define shutdown_none	disable_none
+#define end_none	enable_none
+
+struct hw_interrupt_type no_irq_type = {
+	"none",
+	startup_none,
+	shutdown_none,
+	enable_none,
+	disable_none,
+	ack_none,
+	end_none
+};
+
 /*
  * Generic, controller-independent functions:
  */
@@ -203,6 +235,8 @@ asmlinkage int do_IRQ(unsigned long r4, unsigned long r5,
 	struct irqaction * action;
 	unsigned int status;
 
+	regs.syscall_nr = -1;	/* It's not system call */
+
 	/* Get IRQ number */
 	asm volatile("stc	r2_bank,%0\n\t"
 		     "shlr2	%0\n\t"
@@ -257,7 +291,7 @@ asmlinkage int do_IRQ(unsigned long r4, unsigned long r5,
 	for (;;) {
 		handle_IRQ_event(irq, &regs, action);
 		spin_lock(&irq_controller_lock);
-		
+
 		if (!(desc->status & IRQ_PENDING))
 			break;
 		desc->status &= ~IRQ_PENDING;
@@ -265,7 +299,7 @@ asmlinkage int do_IRQ(unsigned long r4, unsigned long r5,
 	}
 	desc->status &= ~IRQ_INPROGRESS;
 	if (!(desc->status & IRQ_DISABLED)){
-			irq_desc[irq].handler->end(irq);
+		irq_desc[irq].handler->end(irq);
 	}
 	spin_unlock(&irq_controller_lock);
 
@@ -334,16 +368,22 @@ void free_irq(unsigned int irq, void *dev_id)
 
 			/* Found it - now remove it from the list of entries */
 			*pp = action->next;
-			if (irq_desc[irq].action)
-				break;
-			irq_desc[irq].status |= IRQ_DISABLED;
-			irq_desc[irq].handler->shutdown(irq);
-			break;
+			if (!irq_desc[irq].action) {
+				irq_desc[irq].status |= IRQ_DISABLED;
+				irq_desc[irq].handler->shutdown(irq);
+			}
+			spin_unlock_irqrestore(&irq_controller_lock,flags);
+
+			/* Wait to make sure it's not being used on another CPU */
+			while (irq_desc[irq].status & IRQ_INPROGRESS)
+				barrier();
+			kfree(action);
+			return;
 		}
 		printk("Trying to free free IRQ%d\n",irq);
-		break;
+		spin_unlock_irqrestore(&irq_controller_lock,flags);
+		return;
 	}
-	spin_unlock_irqrestore(&irq_controller_lock,flags);
 }
 
 /*

arch/sh/kernel/irq_onchip.c

-/*
+/* $Id: irq_onchip.c,v 1.3 1999/10/11 13:12:19 gniibe Exp $
+ *
  * linux/arch/sh/kernel/irq_onchip.c
  *
  * Copyright (C) 1999  Niibe Yutaka
@@ -31,32 +32,6 @@
 
 #include <linux/irq.h>
 
-
-/*
- * SH (non-)specific no controller code
- */
-
-static void enable_none(unsigned int irq) { }
-static unsigned int startup_none(unsigned int irq) { return 0; }
-static void disable_none(unsigned int irq) { }
-static void ack_none(unsigned int irq)
-{
-}
-
-/* startup is the same as "enable", shutdown is same as "disable" */
-#define shutdown_none	disable_none
-#define end_none	enable_none
-
-struct hw_interrupt_type no_irq_type = {
-	"none",
-	startup_none,
-	shutdown_none,
-	enable_none,
-	disable_none,
-	ack_none,
-	end_none
-};
-
 struct ipr_data {
 	int offset;
 	int priority;
@@ -104,22 +79,25 @@ static struct hw_interrupt_type onChip_irq_type = {
  * IPRC  15-12  11-8  7-4  3-0
  *
  */
+#if defined(__sh3__)
 #define INTC_IPR	0xfffffee2UL	/* Word access */
+#define INTC_SIZE	0x2
+#elif defined(__SH4__)
+#define INTC_IPR	0xffd00004UL	/* Word access */
+#define INTC_SIZE	0x4
+#endif
 
 void disable_onChip_irq(unsigned int irq)
 {
 	/* Set priority in IPR to 0 */
 	int offset = ipr_data[irq-TIMER_IRQ].offset;
-	unsigned long intc_ipr_address = INTC_IPR + offset/16;
+	unsigned long intc_ipr_address = INTC_IPR + (offset/16*INTC_SIZE);
 	unsigned short mask = 0xffff ^ (0xf << (offset%16));
-	unsigned long __dummy;
-
-	asm volatile("mov.w	@%1,%0\n\t"
-		     "and	%2,%0\n\t"
-		     "mov.w	%0,@%1"
-		     : "=&z" (__dummy)
-		     : "r" (intc_ipr_address), "r" (mask)
-		     : "memory" );
+	unsigned long val;
+
+	val = ctrl_inw(intc_ipr_address);
+	val &= mask;
+	ctrl_outw(val, intc_ipr_address);
 }
 
 static void enable_onChip_irq(unsigned int irq)
@@ -127,16 +105,13 @@ static void enable_onChip_irq(unsigned int irq)
 	/* Set priority in IPR back to original value */
 	int offset = ipr_data[irq-TIMER_IRQ].offset;
 	int priority = ipr_data[irq-TIMER_IRQ].priority;
-	unsigned long intc_ipr_address = INTC_IPR + offset/16;
+	unsigned long intc_ipr_address = INTC_IPR + (offset/16*INTC_SIZE);
 	unsigned short value = (priority << (offset%16));
-	unsigned long __dummy;
-
-	asm volatile("mov.w	@%1,%0\n\t"
-		     "or	%2,%0\n\t"
-		     "mov.w	%0,@%1"
-		     : "=&z" (__dummy)
-		     : "r" (intc_ipr_address), "r" (value)
-		     : "memory" );
+	unsigned long val;
+
+	val = ctrl_inw(intc_ipr_address);
+	val |= value;
+	ctrl_outw(val, intc_ipr_address);
 }
 
 void make_onChip_irq(unsigned int irq)
@@ -149,13 +124,11 @@ void make_onChip_irq(unsigned int irq)
 static void mask_and_ack_onChip(unsigned int irq)
 {
 	disable_onChip_irq(irq);
-	sti();
 }
 
 static void end_onChip_irq(unsigned int irq)
 {
 	enable_onChip_irq(irq);
-	cli();
 }
 
 void __init init_IRQ(void)

arch/sh/kernel/process.c

-/*
+/* $Id: process.c,v 1.7 1999/09/23 00:05:41 gniibe Exp $
+ *
  *  linux/arch/sh/kernel/process.c
  *
  *  Copyright (C) 1995  Linus Torvalds
  *
- *  SuperH version:  Copyright (C) 1999  Niibe Yutaka
+ *  SuperH version:  Copyright (C) 1999  Niibe Yutaka & Kaz Kojima
  */
 
 /*
@@ -41,6 +42,10 @@
 
 #include <linux/irq.h>
 
+#if defined(__SH4__)
+struct task_struct *last_task_used_math = NULL;
+#endif
+
 static int hlt_counter=0;
 
 #define HARD_IDLE_TIMEOUT (HZ / 3)
@@ -92,22 +97,21 @@ void machine_power_off(void)
 void show_regs(struct pt_regs * regs)
 {
 	printk("\n");
-	printk("PC: [<%08lx>]", regs->pc);
-	printk(" SP: %08lx", regs->u_regs[UREG_SP]);
-	printk(" SR: %08lx\n", regs->sr);
-	printk("R0 : %08lx R1 : %08lx R2 : %08lx R3 : %08lx\n",
-	       regs->u_regs[0],regs->u_regs[1],
-	       regs->u_regs[2],regs->u_regs[3]);
-	printk("R4 : %08lx R5 : %08lx R6 : %08lx R7 : %08lx\n",
-	       regs->u_regs[4],regs->u_regs[5],
-	       regs->u_regs[6],regs->u_regs[7]);
-	printk("R8 : %08lx R9 : %08lx R10: %08lx R11: %08lx\n",
-	       regs->u_regs[8],regs->u_regs[9],
-	       regs->u_regs[10],regs->u_regs[11]);
-	printk("R12: %08lx R13: %08lx R14: %08lx\n",
-	       regs->u_regs[12],regs->u_regs[13],
-	       regs->u_regs[14]);
-	printk("MACH: %08lx MACL: %08lx GBR: %08lx PR: %08lx",
+	printk("PC  : %08lx SP  : %08lx SR  : %08lx TEA : %08lx\n",
+	       regs->pc, regs->sp, regs->sr, ctrl_inl(MMU_TEA));
+	printk("R0  : %08lx R1  : %08lx R2  : %08lx R3  : %08lx\n",
+	       regs->regs[0],regs->regs[1],
+	       regs->regs[2],regs->regs[3]);
+	printk("R4  : %08lx R5  : %08lx R6  : %08lx R7  : %08lx\n",
+	       regs->regs[4],regs->regs[5],
+	       regs->regs[6],regs->regs[7]);
+	printk("R8  : %08lx R9  : %08lx R10 : %08lx R11 : %08lx\n",
+	       regs->regs[8],regs->regs[9],
+	       regs->regs[10],regs->regs[11]);
+	printk("R12 : %08lx R13 : %08lx R14 : %08lx\n",
+	       regs->regs[12],regs->regs[13],
+	       regs->regs[14]);
+	printk("MACH: %08lx MACL: %08lx GBR : %08lx PR  : %08lx\n",
 	       regs->mach, regs->macl, regs->gbr, regs->pr);
 }
 
@@ -163,13 +167,35 @@ int kernel_thread(int (*fn)(void *), void * arg, unsigned long flags)
  */
 void exit_thread(void)
 {
+#if defined(__sh3__)
 	/* nothing to do ... */
+#elif defined(__SH4__)
+#if 0 /* for the time being... */
+	/* Forget lazy fpu state */
+	if (last_task_used_math == current) {
+		set_status_register (SR_FD, 0);
+		write_system_register (fpscr, FPSCR_PR);
+		last_task_used_math = NULL;
+	}
+#endif
+#endif
 }
 
 void flush_thread(void)
 {
+#if defined(__sh3__)
 	/* do nothing */
 	/* Possibly, set clear debug registers */
+#elif defined(__SH4__)
+#if 0 /* for the time being... */
+	/* Forget lazy fpu state */
+	if (last_task_used_math == current) {
+		set_status_register (SR_FD, 0);
+		write_system_register (fpscr, FPSCR_PR);
+		last_task_used_math = NULL;
+	}
+#endif
+#endif
 }
 
 void release_thread(struct task_struct *dead_task)
@@ -180,6 +206,15 @@ void release_thread(struct task_struct *dead_task)
 /* Fill in the fpu structure for a core dump.. */
 int dump_fpu(struct pt_regs *regs, elf_fpregset_t *r)
 {
+#if defined(__SH4__)
+#if 0 /* for the time being... */
+	    /* We store the FPU info in the task->thread area.  */
+	    if (! (regs->sr & SR_FD)) {
+		    memcpy (r, &current->thread.fpu, sizeof (*r));
+		    return 1;
+	    }
+#endif
+#endif
 	return 0; /* Task didn't use the fpu at all. */
 }
 
@@ -191,14 +226,26 @@ int copy_thread(int nr, unsigned long clone_flags, unsigned long usp,
 	struct pt_regs *childregs;
 
 	childregs = ((struct pt_regs *)(THREAD_SIZE + (unsigned long) p)) - 1;
-
 	*childregs = *regs;
+
+#if defined(__SH4__)
+#if 0 /* for the time being... */
+	if (last_task_used_math == current) {
+		set_status_register (SR_FD, 0);
+		sh4_save_fp (p);
+	}
+	/* New tasks loose permission to use the fpu. This accelerates context
+	   switching for most programs since they don't use the fpu.  */
+	p->thread.sr = (read_control_register (sr) &~ SR_MD) | SR_FD;
+	childregs->sr |= SR_FD;
+#endif
+#endif
 	if (user_mode(regs)) {
-		childregs->u_regs[UREG_SP] = usp;
+		childregs->sp = usp;
 	} else {
-		childregs->u_regs[UREG_SP] = (unsigned long)p+2*PAGE_SIZE;
+		childregs->sp = (unsigned long)p+2*PAGE_SIZE;
 	}
-	childregs->u_regs[0] = 0; /* Set return value for child */
+	childregs->regs[0] = 0; /* Set return value for child */
 
 	p->thread.sp = (unsigned long) childregs;
 	p->thread.pc = (unsigned long) ret_from_fork;
@@ -215,18 +262,22 @@ void dump_thread(struct pt_regs * regs, struct user * dump)
 {
 /* changed the size calculations - should hopefully work better. lbt */
 	dump->magic = CMAGIC;
-	dump->start_code = 0;
-	dump->start_stack = regs->u_regs[UREG_SP] & ~(PAGE_SIZE - 1);
-	dump->u_tsize = ((unsigned long) current->mm->end_code) >> PAGE_SHIFT;
-	dump->u_dsize = ((unsigned long) (current->mm->brk + (PAGE_SIZE-1))) >> PAGE_SHIFT;
-	dump->u_dsize -= dump->u_tsize;
-	dump->u_ssize = 0;
+	dump->start_code = current->mm->start_code;
+	dump->start_data  = current->mm->start_data;
+	dump->start_stack = regs->sp & ~(PAGE_SIZE - 1);
+	dump->u_tsize = (current->mm->end_code - dump->start_code) >> PAGE_SHIFT;
+	dump->u_dsize = (current->mm->brk + (PAGE_SIZE-1) - dump->start_data) >> PAGE_SHIFT;
+	dump->u_ssize = (current->mm->start_stack - dump->start_stack +
+			 PAGE_SIZE - 1) >> PAGE_SHIFT;
 	/* Debug registers will come here. */
 
-	if (dump->start_stack < TASK_SIZE)
-		dump->u_ssize = ((unsigned long) (TASK_SIZE - dump->start_stack)) >> PAGE_SHIFT;
-
 	dump->regs = *regs;
+
+#if 0 /* defined(__SH4__) */
+	/* FPU */
+	memcpy (&dump->regs[EF_SIZE/4], &current->thread.fpu,
+		sizeof (current->thread.fpu));
+#endif
 }
 
 /*
@@ -248,7 +299,7 @@ asmlinkage int sys_fork(unsigned long r4, unsigned long r5,
 			unsigned long r6, unsigned long r7,
 			struct pt_regs regs)
 {
-	return do_fork(SIGCHLD, regs.u_regs[UREG_SP], &regs);
+	return do_fork(SIGCHLD, regs.sp, &regs);
 }
 
 asmlinkage int sys_clone(unsigned long clone_flags, unsigned long newsp,
@@ -256,7 +307,7 @@ asmlinkage int sys_clone(unsigned long clone_flags, unsigned long newsp,
 			 struct pt_regs regs)
 {
 	if (!newsp)
-		newsp = regs.u_regs[UREG_SP];
+		newsp = regs.sp;
 	return do_fork(clone_flags, newsp, &regs);
 }
 
@@ -274,8 +325,7 @@ asmlinkage int sys_vfork(unsigned long r4, unsigned long r5,
 			 unsigned long r6, unsigned long r7,
 			 struct pt_regs regs)
 {
-	return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD,
-		       regs.u_regs[UREG_SP], &regs);
+	return do_fork(CLONE_VFORK | CLONE_VM | SIGCHLD, regs.sp, &regs);
 }
 
 /*

arch/sh/kernel/setup.c

-/*
+/* $Id: setup.c,v 1.4 1999/10/17 02:49:24 gniibe Exp $
+ *
  *  linux/arch/sh/kernel/setup.c
  *
  *  Copyright (C) 1999  Niibe Yutaka
@@ -29,6 +30,7 @@
 #endif
 #include <asm/processor.h>
 #include <linux/console.h>
+#include <asm/pgtable.h>
 #include <asm/uaccess.h>
 #include <asm/system.h>
 #include <asm/io.h>
@@ -38,8 +40,7 @@
  * Machine setup..
  */
 
-struct sh_cpuinfo boot_cpu_data = { 0, 0, 0, 0, };
-extern int _text, _etext, _edata, _end, _stext, __bss_start;
+struct sh_cpuinfo boot_cpu_data = { CPU_SH_NONE, 0, 0, 0, };
 
 #ifdef CONFIG_BLK_DEV_RAM
 extern int rd_doload;		/* 1 = load ramdisk, 0 = don't load */
@@ -48,13 +49,31 @@ extern int rd_image_start;	/* starting block # of image */
 #endif
 
 extern int root_mountflags;
+extern int _text, _etext, _edata, _end;
+
+/*
+ * This is set up by the setup-routine at boot-time
+ */
+#define PARAM	((unsigned char *)empty_zero_page)
+
+#define MOUNT_ROOT_RDONLY (*(unsigned long *) (PARAM+0x000))
+#define RAMDISK_FLAGS (*(unsigned long *) (PARAM+0x004))
+#define ORIG_ROOT_DEV (*(unsigned long *) (PARAM+0x008))
+#define LOADER_TYPE (*(unsigned long *) (PARAM+0x00c))
+#define INITRD_START (*(unsigned long *) (PARAM+0x010))
+#define INITRD_SIZE (*(unsigned long *) (PARAM+0x014))
+#define MEMORY_END (*(unsigned long *) (PARAM+0x018))
+/* ... */
+#define COMMAND_LINE ((char *) (PARAM+0x100))
+#define COMMAND_LINE_SIZE 256
+
+#define RAMDISK_IMAGE_START_MASK  	0x07FF
+#define RAMDISK_PROMPT_FLAG		0x8000
+#define RAMDISK_LOAD_FLAG		0x4000	
 
-#define COMMAND_LINE_SIZE 1024
 static char command_line[COMMAND_LINE_SIZE] = { 0, };
        char saved_command_line[COMMAND_LINE_SIZE];
 
-extern unsigned char *root_fs_image;
-
 struct resource standard_io_resources[] = {
 	{ "dma1", 0x00, 0x1f },
 	{ "pic1", 0x20, 0x3f },
@@ -68,7 +87,6 @@ struct resource standard_io_resources[] = {
 
 #define STANDARD_IO_RESOURCES (sizeof(standard_io_resources)/sizeof(struct resource))
 
-
 /* System RAM - interrupted by the 640kB-1M hole */
 #define code_resource (ram_resources[3])
 #define data_resource (ram_resources[4])
@@ -87,17 +105,26 @@ static struct resource rom_resources[MAXROMS] = {
 	{ "Video ROM", 0xc0000, 0xc7fff }
 };
 
-
 void __init setup_arch(char **cmdline_p,
 		       unsigned long * memory_start_p,
 		       unsigned long * memory_end_p)
 {
-	*cmdline_p = command_line;
-	*memory_start_p = (unsigned long) &_end;
-	*memory_end_p = 0x8c400000; /* For my board. */
-	ram_resources[1].end = *memory_end_p-1;
+	unsigned long memory_start, memory_end;
+
+	ROOT_DEV = to_kdev_t(ORIG_ROOT_DEV);
+
+#ifdef CONFIG_BLK_DEV_RAM
+	rd_image_start = RAMDISK_FLAGS & RAMDISK_IMAGE_START_MASK;
+	rd_prompt = ((RAMDISK_FLAGS & RAMDISK_PROMPT_FLAG) != 0);
+	rd_doload = ((RAMDISK_FLAGS & RAMDISK_LOAD_FLAG) != 0);
+#endif
+	if (!MOUNT_ROOT_RDONLY)
+		root_mountflags &= ~MS_RDONLY;
+
+	memory_start = (unsigned long) &_end;
+	memory_end = MEMORY_END;
 
-	init_mm.start_code = (unsigned long)&_stext;
+	init_mm.start_code = (unsigned long)&_text;
 	init_mm.end_code = (unsigned long) &_etext;
 	init_mm.end_data = (unsigned long) &_edata;
 	init_mm.brk = (unsigned long) &_end;
@@ -107,51 +134,25 @@ void __init setup_arch(char **cmdline_p,
 	data_resource.start = virt_to_bus(&_etext);
 	data_resource.end = virt_to_bus(&_edata)-1;
 
- 	ROOT_DEV = MKDEV(FLOPPY_MAJOR, 0);
+	/* Save unparsed command line copy for /proc/cmdline */
+	memcpy(saved_command_line, COMMAND_LINE, COMMAND_LINE_SIZE);
+	saved_command_line[COMMAND_LINE_SIZE-1] = '\0';
 
-	initrd_below_start_ok = 1;
-	initrd_start = (long)&root_fs_image;
-	initrd_end = (long)&__bss_start;
-	mount_initrd = 1;
+	memcpy(command_line, COMMAND_LINE, COMMAND_LINE_SIZE);
+	command_line[COMMAND_LINE_SIZE-1] = '\0';
 
+	/* Not support "mem=XXX[kKmM]" command line option. */
+	*cmdline_p = command_line;
 
-#if 0
-	/* Request the standard RAM and ROM resources - they eat up PCI memory space */
-	request_resource(&iomem_resource, ram_resources+0);
-	request_resource(&iomem_resource, ram_resources+1);
-	request_resource(&iomem_resource, ram_resources+2);
-	request_resource(ram_resources+1, &code_resource);
-	request_resource(ram_resources+1, &data_resource);
-#endif
-
-#if 0
-	for (i = 0; i < STANDARD_IO_RESOURCES; i++)
-		request_resource(&ioport_resource, standard_io_resources+i);
-#endif
-
-#if 0
-	rd_image_start = (long)root_fs_image;
-	rd_prompt = 0;
-	rd_doload = 1;
-#endif
+	memory_end &= PAGE_MASK;
+	ram_resources[1].end = memory_end-1;
 
-#if 0
- 	ROOT_DEV = to_kdev_t(ORIG_ROOT_DEV);
-
-#ifdef CONFIG_BLK_DEV_RAM
-	rd_image_start = RAMDISK_FLAGS & RAMDISK_IMAGE_START_MASK;
-	rd_prompt = ((RAMDISK_FLAGS & RAMDISK_PROMPT_FLAG) != 0);
-	rd_doload = ((RAMDISK_FLAGS & RAMDISK_LOAD_FLAG) != 0);
-#endif
-
-	if (!MOUNT_ROOT_RDONLY)
-		root_mountflags &= ~MS_RDONLY;
-#endif
+	*memory_start_p = memory_start;
+	*memory_end_p = memory_end;
 
 #ifdef CONFIG_BLK_DEV_INITRD
-#if 0
 	if (LOADER_TYPE) {
-		initrd_start = INITRD_START ? INITRD_START + PAGE_OFFSET : 0;
+		initrd_start = INITRD_START ? INITRD_START : 0;
 		initrd_end = initrd_start+INITRD_SIZE;
 		if (initrd_end > memory_end) {
 			printk("initrd extends beyond end of memory "
@@ -162,6 +163,29 @@ void __init setup_arch(char **cmdline_p,
 	}
 #endif
 
+#if 0
+	/*
+	 * Request the standard RAM and ROM resources -
+	 * they eat up PCI memory space
+	 */
+	request_resource(&iomem_resource, ram_resources+0);
+	request_resource(&iomem_resource, ram_resources+1);
+	request_resource(&iomem_resource, ram_resources+2);
+	request_resource(ram_resources+1, &code_resource);
+	request_resource(ram_resources+1, &data_resource);
+	probe_roms();
+
+	/* request I/O space for devices used on all i[345]86 PCs */
+	for (i = 0; i < STANDARD_IO_RESOURCES; i++)
+		request_resource(&ioport_resource, standard_io_resources+i);
+#endif
+
+#ifdef CONFIG_VT
+#if defined(CONFIG_VGA_CONSOLE)
+	conswitchp = &vga_con;
+#elif defined(CONFIG_DUMMY_CONSOLE)
+	conswitchp = &dummy_con;
+#endif
 #endif
 }
 
@@ -173,12 +197,12 @@ int get_cpuinfo(char *buffer)
 {
 	char *p = buffer;
 
-#ifdef CONFIG_CPU_SH3
-	p += sprintf(p,"cpu family\t: SH3\n"
+#if defined(__sh3__)
+	p += sprintf(p,"cpu family\t: SH-3\n"
 		       "cache size\t: 8K-byte\n");
-#elif CONFIG_CPU_SH4
-	p += sprintf(p,"cpu family\t: SH4\n"
-		       "cache size\t: ??K-byte\n");
+#elif defined(__SH4__)
+	p += sprintf(p,"cpu family\t: SH-4\n"
+		       "cache size\t: 8K-byte/16K-byte\n");
 #endif
 	p += sprintf(p, "bogomips\t: %lu.%02lu\n\n",
 		     (loops_per_sec+2500)/500000,

arch/sh/kernel/signal.c

-/*
+/* $Id: signal.c,v 1.10 1999/09/27 23:25:44 gniibe Exp $
+ *
  *  linux/arch/sh/kernel/signal.c
  *
  *  Copyright (C) 1991, 1992  Linus Torvalds
@@ -9,8 +10,6 @@
  *
  */
 
-#include <linux/config.h>
-
 #include <linux/sched.h>
 #include <linux/mm.h>
 #include <linux/smp.h>
@@ -24,6 +23,7 @@
 #include <linux/stddef.h>
 #include <asm/ucontext.h>
 #include <asm/uaccess.h>
+#include <asm/pgtable.h>
 
 #define DEBUG_SIG 0
 
@@ -50,7 +50,7 @@ sys_sigsuspend(old_sigset_t mask,
 	recalc_sigpending(current);
 	spin_unlock_irq(&current->sigmask_lock);
 
-	regs.u_regs[0] = -EINTR;
+	regs.regs[0] = -EINTR;
 	while (1) {
 		current->state = TASK_INTERRUPTIBLE;
 		schedule();
@@ -80,7 +80,7 @@ sys_rt_sigsuspend(sigset_t *unewset, size_t sigsetsize,
 	recalc_sigpending(current);
 	spin_unlock_irq(&current->sigmask_lock);
 
-	regs.u_regs[0] = -EINTR;
+	regs.regs[0] = -EINTR;
 	while (1) {
 		current->state = TASK_INTERRUPTIBLE;
 		schedule();
@@ -126,7 +126,7 @@ sys_sigaltstack(const stack_t *uss, stack_t *uoss,
 		unsigned long r6, unsigned long r7,
 		struct pt_regs regs)
 {
-	return do_sigaltstack(uss, uoss, regs.u_regs[UREG_SP]);
+	return do_sigaltstack(uss, uoss, regs.sp);
 }
 
 
@@ -137,7 +137,7 @@ sys_sigaltstack(const stack_t *uss, stack_t *uoss,
 struct sigframe
 {
 	struct sigcontext sc;
-	/* FPU should come here: SH-3 has no FPU */
+	/* FPU data should come here: SH-3 has no FPU */
 	unsigned long extramask[_NSIG_WORDS-1];
 	char retcode[4];
 };
@@ -158,22 +158,22 @@ restore_sigcontext(struct pt_regs *regs, struct sigcontext *sc, int *r0_p)
 {
 	unsigned int err = 0;
 
-#define COPY(x)		err |= __get_user(regs->x, &sc->x)
-				COPY(u_regs[1]);
-	COPY(u_regs[2]);	COPY(u_regs[3]);
-	COPY(u_regs[4]);	COPY(u_regs[5]);
-	COPY(u_regs[6]);	COPY(u_regs[7]);
-	COPY(u_regs[8]);	COPY(u_regs[9]);
-	COPY(u_regs[10]);	COPY(u_regs[11]);
-	COPY(u_regs[12]);	COPY(u_regs[13]);
-	COPY(u_regs[14]);	COPY(u_regs[15]);
-	COPY(gbr);		COPY(mach);
-	COPY(macl);		COPY(pr);
-	COPY(sr);		COPY(pc);
+#define COPY(x)		err |= __get_user(regs->x, &sc->sc_##x)
+			COPY(regs[1]);
+	COPY(regs[2]);	COPY(regs[3]);
+	COPY(regs[4]);	COPY(regs[5]);
+	COPY(regs[6]);	COPY(regs[7]);
+	COPY(regs[8]);	COPY(regs[9]);
+	COPY(regs[10]);	COPY(regs[11]);
+	COPY(regs[12]);	COPY(regs[13]);
+	COPY(regs[14]);	COPY(sp);
+	COPY(gbr);	COPY(mach);
+	COPY(macl);	COPY(pr);
+	COPY(sr);	COPY(pc);
 #undef COPY
 
 	regs->syscall_nr = -1;		/* disable syscall checks */
-	err |= __get_user(*r0_p, &sc->u_regs[0]);
+	err |= __get_user(*r0_p, &sc->sc_regs[0]);
 
 	return err;
 }
@@ -182,7 +182,7 @@ asmlinkage int sys_sigreturn(unsigned long r4, unsigned long r5,
 			     unsigned long r6, unsigned long r7,
 			     struct pt_regs regs)
 {
-	struct sigframe *frame = (struct sigframe *)regs.u_regs[UREG_SP];
+	struct sigframe *frame = (struct sigframe *)regs.sp;
 	sigset_t set;
 	int r0;
 
@@ -213,7 +213,7 @@ asmlinkage int sys_rt_sigreturn(unsigned long r4, unsigned long r5,
 				unsigned long r6, unsigned long r7,
 				struct pt_regs regs)
 {
-	struct rt_sigframe *frame = (struct rt_sigframe *)regs.u_regs[UREG_SP];
+	struct rt_sigframe *frame = (struct rt_sigframe *)regs.sp;
 	sigset_t set;
 	stack_t st;
 	int r0;
@@ -236,7 +236,7 @@ asmlinkage int sys_rt_sigreturn(unsigned long r4, unsigned long r5,
 		goto badframe;
 	/* It is more difficult to avoid calling this function than to
 	   call it and ignore errors.  */
-	do_sigaltstack(&st, NULL, regs.u_regs[UREG_SP]);
+	do_sigaltstack(&st, NULL, regs.sp);
 
 	return r0;
 
@@ -255,18 +255,18 @@ setup_sigcontext(struct sigcontext *sc, struct pt_regs *regs,
 {
 	int err = 0;
 
-#define COPY(x)		err |= __put_user(regs->x, &sc->x)
-	COPY(u_regs[0]);	COPY(u_regs[1]);
-	COPY(u_regs[2]);	COPY(u_regs[3]);
-	COPY(u_regs[4]);	COPY(u_regs[5]);
-	COPY(u_regs[6]);	COPY(u_regs[7]);
-	COPY(u_regs[8]);	COPY(u_regs[9]);
-	COPY(u_regs[10]);	COPY(u_regs[11]);
-	COPY(u_regs[12]);	COPY(u_regs[13]);
-	COPY(u_regs[14]);	COPY(u_regs[15]);
-	COPY(gbr);		COPY(mach);
-	COPY(macl);		COPY(pr);
-	COPY(sr);		COPY(pc);
+#define COPY(x)		err |= __put_user(regs->x, &sc->sc_##x)
+	COPY(regs[0]);	COPY(regs[1]);
+	COPY(regs[2]);	COPY(regs[3]);
+	COPY(regs[4]);	COPY(regs[5]);
+	COPY(regs[6]);	COPY(regs[7]);
+	COPY(regs[8]);	COPY(regs[9]);
+	COPY(regs[10]);	COPY(regs[11]);
+	COPY(regs[12]);	COPY(regs[13]);
+	COPY(regs[14]);	COPY(sp);
+	COPY(gbr);	COPY(mach);
+	COPY(macl);	COPY(pr);
+	COPY(sr);	COPY(pc);
 #undef COPY
 
 	/* non-iBCS2 extensions.. */
@@ -294,7 +294,7 @@ static void setup_frame(int sig, struct k_sigaction *ka,
 	int err = 0;
 	int signal;
 
-	frame = get_sigframe(ka, regs->u_regs[UREG_SP], sizeof(*frame));
+	frame = get_sigframe(ka, regs->sp, sizeof(*frame));
 
 	if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
 		goto give_sigsegv;
@@ -318,8 +318,8 @@ static void setup_frame(int sig, struct k_sigaction *ka,
 		regs->pr = (unsigned long) ka->sa.sa_restorer;
 	} else {
 		/* This is ; mov  #__NR_sigreturn,r0 ; trapa #0 */
-#ifdef CONFIG_LITTLE_ENDIAN
-		unsigned long code = 0x00c300e0 | (__NR_sigreturn << 8);
+#ifdef __LITTLE_ENDIAN__
+		unsigned long code = 0xc300e000 | (__NR_sigreturn);
 #else
 		unsigned long code = 0xe000c300 | (__NR_sigreturn << 16);
 #endif
@@ -332,8 +332,8 @@ static void setup_frame(int sig, struct k_sigaction *ka,
 		goto give_sigsegv;
 
 	/* Set up registers for signal handler */
-	regs->u_regs[UREG_SP] = (unsigned long) frame;
-	regs->u_regs[4] = signal; /* Arg for signal handler */
+	regs->sp = (unsigned long) frame;
+	regs->regs[4] = signal; /* Arg for signal handler */
 	regs->pc = (unsigned long) ka->sa.sa_handler;
 
 	set_fs(USER_DS);
@@ -343,6 +343,7 @@ static void setup_frame(int sig, struct k_sigaction *ka,
 		current->comm, current->pid, frame, regs->pc, regs->pr);
 #endif
 
+	flush_icache_range(regs->pr, regs->pr+4);
 	return;
 
 give_sigsegv:
@@ -358,7 +359,7 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 	int err = 0;
 	int signal;
 
-	frame = get_sigframe(ka, regs->u_regs[UREG_SP], sizeof(*frame));
+	frame = get_sigframe(ka, regs->sp, sizeof(*frame));
 
 	if (!access_ok(VERIFY_WRITE, frame, sizeof(*frame)))
 		goto give_sigsegv;
@@ -376,9 +377,9 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 	/* Create the ucontext.  */
 	err |= __put_user(0, &frame->uc.uc_flags);
 	err |= __put_user(0, &frame->uc.uc_link);
-	err |= __put_user(current->sas_ss_sp, &frame->uc.uc_stack.ss_sp);
-	err |= __put_user(sas_ss_flags(regs->u_regs[UREG_SP]),
-			  &frame->uc.uc_stack.ss_flags);
+	err |= __put_user((void *)current->sas_ss_sp,
+			  &frame->uc.uc_stack.ss_sp);
+	err |= __put_user(sas_ss_flags(regs->sp), &frame->uc.uc_stack.ss_flags);
 	err |= __put_user(current->sas_ss_size, &frame->uc.uc_stack.ss_size);
 	err |= setup_sigcontext(&frame->uc.uc_mcontext,
 			        regs, set->sig[0]);
@@ -390,8 +391,8 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 		regs->pr = (unsigned long) ka->sa.sa_restorer;
 	} else {
 		/* This is ; mov  #__NR_sigreturn,r0 ; trapa #0 */
-#ifdef CONFIG_LITTLE_ENDIAN
-		unsigned long code = 0x00c300e0 | (__NR_sigreturn << 8);
+#ifdef __LITTLE_ENDIAN__
+		unsigned long code = 0xc300e000 | (__NR_sigreturn);
 #else
 		unsigned long code = 0xe000c300 | (__NR_sigreturn << 16);
 #endif
@@ -404,8 +405,8 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 		goto give_sigsegv;
 
 	/* Set up registers for signal handler */
-	regs->u_regs[UREG_SP] = (unsigned long) frame;
-	regs->u_regs[4] = signal; /* Arg for signal handler */
+	regs->sp = (unsigned long) frame;
+	regs->regs[4] = signal; /* Arg for signal handler */
 	regs->pc = (unsigned long) ka->sa.sa_handler;
 
 	set_fs(USER_DS);
@@ -415,6 +416,7 @@ static void setup_rt_frame(int sig, struct k_sigaction *ka, siginfo_t *info,
 		current->comm, current->pid, frame, regs->pc, regs->pr);
 #endif
 
+	flush_icache_range(regs->pr, regs->pr+4);
 	return;
 
 give_sigsegv:
@@ -434,19 +436,19 @@ handle_signal(unsigned long sig, struct k_sigaction *ka,
 	/* Are we from a system call? */
 	if (regs->syscall_nr >= 0) {
 		/* If so, check system call restarting.. */
-		switch (regs->u_regs[0]) {
+		switch (regs->regs[0]) {
 			case -ERESTARTNOHAND:
-				regs->u_regs[0] = -EINTR;
+				regs->regs[0] = -EINTR;
 				break;
 
 			case -ERESTARTSYS:
 				if (!(ka->sa.sa_flags & SA_RESTART)) {
-					regs->u_regs[0] = -EINTR;
+					regs->regs[0] = -EINTR;
 					break;
 				}
 			/* fallthrough */
 			case -ERESTARTNOINTR:
-				regs->u_regs[0] = regs->syscall_nr;
+				regs->regs[0] = regs->syscall_nr;
 				regs->pc -= 2;
 		}
 	}
@@ -577,7 +579,6 @@ int do_signal(struct pt_regs *regs, sigset_t *oldset)
 				/* NOTREACHED */
 			}
 		}
-
 		/* Whee!  Actually deliver the signal.  */
 		handle_signal(signr, ka, &info, oldset, regs);
 		return 1;
@@ -586,10 +587,10 @@ int do_signal(struct pt_regs *regs, sigset_t *oldset)
 	/* Did we come from a system call? */
 	if (regs->syscall_nr >= 0) {
 		/* Restart the system call - no handlers present */
-		if (regs->u_regs[0] == -ERESTARTNOHAND ||
-		    regs->u_regs[0] == -ERESTARTSYS ||
-		    regs->u_regs[0] == -ERESTARTNOINTR) {
-			regs->u_regs[0] = regs->syscall_nr;
+		if (regs->regs[0] == -ERESTARTNOHAND ||
+		    regs->regs[0] == -ERESTARTSYS ||
+		    regs->regs[0] == -ERESTARTNOINTR) {
+			regs->regs[0] = regs->syscall_nr;
 			regs->pc -= 2;
 		}
 	}

arch/sh/kernel/sys_sh.c

 /*
- * linux/arch/i386/kernel/sys_i386.c
+ * linux/arch/sh/kernel/sys_sh.c
  *
  * This file contains various random system calls that
- * have a non-standard calling sequence on the Linux/i386
+ * have a non-standard calling sequence on the Linux/SuperH
  * platform.
+ *
+ * Taken from i386 version.
  */
 
 #include <linux/errno.h>
@@ -41,66 +43,32 @@ asmlinkage int sys_pipe(unsigned long * fildes)
 	return error;
 }
 
-/*
- * Perform the select(nd, in, out, ex, tv) and mmap() system
- * calls. Linux/i386 didn't use to be able to handle more than
- * 4 system call parameters, so these system calls used a memory
- * block for parameter passing..
- */
-
-struct mmap_arg_struct {
-	unsigned long addr;
-	unsigned long len;
-	unsigned long prot;
-	unsigned long flags;
-	unsigned long fd;
-	unsigned long offset;
-};
-
-asmlinkage int old_mmap(struct mmap_arg_struct *arg)
+asmlinkage unsigned long
+sys_mmap(int fd, unsigned long addr,
+	 unsigned long len, unsigned long prot,
+	 unsigned long flags, unsigned long off)
 {
 	int error = -EFAULT;
-	struct file * file = NULL;
-	struct mmap_arg_struct a;
-
-	if (copy_from_user(&a, arg, sizeof(a)))
-		return -EFAULT;
+	struct file *file = NULL;
 
 	down(&current->mm->mmap_sem);
 	lock_kernel();
-	if (!(a.flags & MAP_ANONYMOUS)) {
+	if (!(flags & MAP_ANONYMOUS)) {
 		error = -EBADF;
-		file = fget(a.fd);
+		file = fget(fd);
 		if (!file)
 			goto out;
 	}
-	a.flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE);
+	flags &= ~(MAP_EXECUTABLE | MAP_DENYWRITE);
 
-	error = do_mmap(file, a.addr, a.len, a.prot, a.flags, a.offset);
+	error = do_mmap(file, addr, len, prot, flags, off);
 	if (file)
 		fput(file);
 out:
 	unlock_kernel();
 	up(&current->mm->mmap_sem);
-	return error;
-}
-
-extern asmlinkage int sys_select(int, fd_set *, fd_set *, fd_set *, struct timeval *);
-
-struct sel_arg_struct {
-	unsigned long n;
-	fd_set *inp, *outp, *exp;
-	struct timeval *tvp;
-};
-
-asmlinkage int old_select(struct sel_arg_struct *arg)
-{
-	struct sel_arg_struct a;
 
-	if (copy_from_user(&a, arg, sizeof(a)))
-		return -EFAULT;
-	/* sys_select() does the appropriate kernel locking */
-	return sys_select(a.n, a.inp, a.outp, a.exp, a.tvp);
+	return error;
 }
 
 /*
@@ -198,9 +166,6 @@ asmlinkage int sys_ipc (uint call, int first, int second,
 	return -EINVAL;
 }
 
-/*
- * Old cruft
- */
 asmlinkage int sys_uname(struct old_utsname * name)
 {
 	int err;
@@ -212,35 +177,6 @@ asmlinkage int sys_uname(struct old_utsname * name)
 	return err?-EFAULT:0;
 }
 
-asmlinkage int sys_olduname(struct oldold_utsname * name)
-{
-	int error;
-
-	if (!name)
-		return -EFAULT;
-	if (!access_ok(VERIFY_WRITE,name,sizeof(struct oldold_utsname)))
-		return -EFAULT;
-  
-  	down(&uts_sem);
-	
-	error = __copy_to_user(&name->sysname,&system_utsname.sysname,__OLD_UTS_LEN);
-	error |= __put_user(0,name->sysname+__OLD_UTS_LEN);
-	error |= __copy_to_user(&name->nodename,&system_utsname.nodename,__OLD_UTS_LEN);
-	error |= __put_user(0,name->nodename+__OLD_UTS_LEN);
-	error |= __copy_to_user(&name->release,&system_utsname.release,__OLD_UTS_LEN);
-	error |= __put_user(0,name->release+__OLD_UTS_LEN);
-	error |= __copy_to_user(&name->version,&system_utsname.version,__OLD_UTS_LEN);
-	error |= __put_user(0,name->version+__OLD_UTS_LEN);
-	error |= __copy_to_user(&name->machine,&system_utsname.machine,__OLD_UTS_LEN);
-	error |= __put_user(0,name->machine+__OLD_UTS_LEN);
-	
-	up(&uts_sem);
-	
-	error = error ? -EFAULT : 0;
-
-	return error;
-}
-
 asmlinkage int sys_pause(void)
 {
 	current->state = TASK_INTERRUPTIBLE;

arch/sh/kernel/test-img.c

-unsigned char root_fs_image[]
-__attribute__((__section__(".data.disk_image"))) 
-= {
-0x1f,0x8b,0x08,0x08,0x5d,0xd5,0xc7,0x37,0x00,0x03,0x72,0x2e,0x62,0x69,0x6e,0x00,
-0xed,0xdc,0x3f,0x6c,0x1b,0x55,0x1c,0xc0,0xf1,0xdf,0xf9,0xdc,0x04,0x27,0x69,0xb1,
-0x93,0x14,0x10,0x48,0x91,0xd3,0x02,0x4d,0x8a,0xb8,0xd4,0x21,0x8a,0x09,0x02,0x02,
-0xb5,0x4a,0xab,0x52,0x65,0x69,0x11,0x03,0x42,0xc2,0xb1,0x8f,0xc4,0x92,0xe3,0x03,
-0x9f,0x8d,0xca,0x14,0xd8,0x88,0x2a,0xa6,0x0e,0x88,0xa9,0x20,0xb1,0x87,0x8d,0xa5,
-0x5b,0x86,0xcc,0x90,0x78,0x77,0xd4,0x60,0x75,0xa9,0x40,0xe2,0xdf,0xd0,0x42,0x78,
-0x77,0xef,0x9c,0x38,0x24,0x72,0x49,0x20,0xc9,0x70,0xdf,0x8f,0xf2,0xf3,0xd9,0x77,
-0xbf,0xf3,0xbb,0x67,0xbf,0xdf,0xf9,0x4f,0xf4,0x2c,0x02,0x20,0xac,0xe2,0x2a,0x5e,
-0x53,0x61,0xaa,0x18,0x0e,0xd6,0x19,0xad,0x09,0x49,0x1d,0x5e,0x5e,0x7d,0x75,0x39,
-0xfd,0x6c,0x6d,0x39,0x6d,0x48,0xbf,0x5c,0xfd,0xc9,0xf0,0xf3,0x56,0xd5,0x3a,0x99,
-0xba,0xf7,0xd0,0x76,0x8a,0x53,0x5f,0xc4,0xdf,0xcd,0x24,0x56,0x6e,0x9e,0x59,0xb9,
-0x30,0x3e,0x73,0x3b,0xf7,0x3f,0x76,0x01,0xc0,0x3e,0x79,0x75,0x1f,0x55,0x71,0x4c,
-0x74,0xfd,0x47,0x8f,0xf6,0x70,0x00,0x1c,0xa2,0x8d,0x8d,0x49,0x6f,0xf1,0xc9,0x06,
-0x00,0x00,0x08,0x8d,0xe6,0xfb,0x00,0xef,0x73,0x7c,0x33,0x0e,0xf3,0xfd,0xc7,0xbd,
-0xd7,0xc5,0xff,0xd0,0x31,0x5a,0x5b,0x4e,0xf7,0x05,0xa1,0xb7,0x1c,0x93,0x48,0x4b,
-0x5e,0xe7,0x61,0x1e,0x14,0x80,0x50,0xf0,0xcf,0x3f,0xe7,0x76,0x3b,0xff,0x45,0xe4,
-0x89,0x96,0xbc,0x47,0x54,0xc4,0x54,0x74,0xa9,0xe8,0x56,0xd1,0xa3,0xe2,0xb8,0x8a,
-0x13,0x2a,0x1e,0x15,0xfd,0xfd,0x68,0x42,0x45,0xaf,0x8a,0xbe,0xbd,0xb6,0xaf,0xce,
-0x7f,0x7f,0xaa,0x76,0xef,0x07,0xd1,0x6c,0xbf,0xf5,0xfc,0xd7,0xbf,0xf7,0xae,0x6d,
-0x32,0xda,0x6c,0x6b,0xb6,0x7f,0x56,0x9d,0x77,0x4f,0x05,0xb1,0x5b,0xfb,0x27,0x0f,
-0xa8,0xfd,0x6f,0x06,0xf5,0xf2,0xfe,0x8e,0xfe,0xff,0x63,0xaf,0xff,0xf0,0xc5,0x54,
-0xdb,0xfe,0x7f,0x7a,0xeb,0xf2,0x15,0x53,0xe4,0xe6,0xaa,0x7e,0xed,0x19,0x0b,0xda,
-0xbf,0x75,0xd9,0xd8,0xd6,0xff,0xc7,0xf6,0xdf,0x7c,0xdb,0xf6,0x37,0xbe,0xd6,0x63,
-0x6a,0xe7,0xe3,0xbf,0x7d,0x2f,0xcb,0x1a,0x29,0x16,0x4a,0xd5,0xeb,0xe5,0x7d,0x7c,
-0x73,0xde,0xae,0x7d,0xaf,0x8f,0x3d,0x2a,0xc3,0xda,0xbc,0x1e,0x51,0x6d,0xe9,0x31,
-0xde,0xaf,0x8e,0xac,0xe8,0xb8,0x95,0xe7,0xde,0x77,0xaa,0xa5,0xbc,0x1e,0xf3,0x3d,
-0x62,0x4a,0xde,0xfe,0xc8,0x1f,0xfb,0x3d,0xea,0x49,0x71,0xa7,0x0b,0x25,0x6f,0xfc,
-0xdf,0x36,0x3b,0x65,0xdf,0x07,0x08,0xe0,0x48,0xe8,0xd7,0xb2,0xad,0xfa,0xff,0xd5,
-0xd4,0xf5,0x0f,0x20,0x24,0xf8,0xa7,0x1f,0x10,0x5e,0xd4,0x3f,0x10,0x5e,0xd4,0x3f,
-0x10,0x5e,0xd4,0x3f,0x10,0x5e,0xd4,0x3f,0x10,0x5e,0xd4,0x3f,0x10,0x5e,0xd4,0x3f,
-0x10,0x5e,0xd4,0x3f,0x10,0x5e,0xd4,0x3f,0x10,0x5e,0xd4,0x3f,0x10,0x5e,0xd4,0x3f,
-0x10,0x4a,0x7a,0x4e,0xcf,0xce,0xf9,0x3f,0xde,0xbc,0xb6,0xbb,0x66,0xa7,0xe4,0x9c,
-0x92,0xeb,0x14,0xed,0xa3,0x3d,0x48,0x00,0x07,0x42,0xcf,0xe3,0xdb,0x59,0xff,0xde,
-0x7c,0xd6,0xbb,0x66,0x54,0x0a,0xa5,0x42,0xe5,0x68,0x8f,0x10,0xc0,0x41,0x99,0xbf,
-0x70,0xe5,0x0d,0x23,0xd2,0x32,0x43,0x38,0x22,0x67,0xc5,0x9f,0x32,0x1c,0xff,0x4a,
-0x2d,0xc7,0xd4,0xd5,0x75,0x7f,0xfd,0x98,0x24,0xd5,0xb6,0x21,0x89,0xf9,0x53,0xe1,
-0x83,0x1d,0xe2,0x41,0x18,0xd3,0x3a,0xfc,0x9f,0x11,0x34,0x74,0x78,0xb7,0x07,0x83,
-0xd8,0xd4,0xe1,0x27,0x67,0xd6,0x8d,0x46,0x7c,0xa4,0x51,0x8f,0xd6,0x3a,0x4a,0xbf,
-0x2c,0xc9,0x7b,0xa7,0x3f,0x33,0x16,0xcc,0x5a,0xb4,0x61,0xd6,0xa3,0x4b,0xe2,0xdc,
-0x91,0xee,0xd2,0xef,0x22,0x89,0xa7,0x55,0xbc,0x38,0xd2,0x98,0xff,0xb9,0x1e,0xf1,
-0xb2,0xa6,0xcd,0xf3,0x89,0x85,0xce,0x75,0xa3,0xf6,0x78,0xe3,0xa4,0x97,0x27,0xb1,
-0xc5,0xbf,0x24,0x76,0x6a,0x68,0xa1,0x7b,0xa5,0x6f,0x4d,0x3e,0x34,0x52,0xe9,0x1b,
-0x0f,0xf2,0xa7,0x7f,0x34,0xea,0xcf,0x2c,0xc9,0xe2,0x1f,0x6b,0x6a,0xfb,0xf7,0x27,
-0xd6,0x0d,0xab,0xd7,0xbe,0xb3,0x26,0x03,0x89,0x86,0x0c,0xf4,0xd6,0x33,0x03,0x7d,
-0x4b,0xf2,0x43,0xd7,0xba,0x21,0xb1,0x5a,0xac,0x71,0xdc,0xbb,0x17,0x2f,0x4f,0xed,
-0x7b,0xe6,0xc6,0x83,0xc5,0xdf,0xbc,0xf5,0xaa,0xcd,0x97,0xe5,0x9d,0xcf,0xe7,0x55,
-0xbf,0x2a,0xf2,0xdd,0x93,0x1b,0xea,0xf6,0xb5,0x6b,0xb3,0x05,0x37,0xa9,0xfe,0xae,
-0x56,0x3f,0xb0,0xcb,0x97,0x06,0xbd,0xe9,0xda,0x32,0x39,0xd9,0x25,0xae,0x33,0x67,
-0x57,0x66,0x0b,0xa5,0x99,0x64,0xb5,0x54,0x75,0xab,0xd9,0xa2,0x65,0x59,0xde,0xc6,
-0x4b,0x76,0xb1,0xe8,0x24,0xdf,0x76,0xca,0xc5,0xbc,0x97,0x7c,0x31,0x93,0x79,0x29,
-0x39,0x74,0x71,0xea,0xad,0xe1,0xa4,0x3d,0x93,0x73,0x9f,0x1f,0xb5,0x26,0x52,0xd6,
-0xf8,0x78,0x32,0x35,0x31,0x31,0x71,0xee,0x85,0xd4,0x58,0x72,0xc8,0x5f,0x9d,0xb2,
-0x52,0xd6,0x68,0xb2,0x6c,0x17,0xed,0xac,0x6b,0x0f,0x8b,0x58,0xee,0xc7,0x73,0x95,
-0xec,0xb4,0x5a,0x56,0xca,0x7a,0x39,0xdb,0xbc,0x56,0xb1,0xaf,0x57,0xc4,0x2a,0x3b,
-0xf9,0x6c,0x25,0x2b,0x96,0xbe,0xcc,0x55,0x9c,0xb2,0xab,0x6e,0xe8,0xc5,0xb4,0xab,
-0x2e,0x72,0xce,0xdc,0x9c,0x5d,0xda,0xd3,0xe9,0xfb,0xa9,0xe0,0xf9,0xeb,0xf0,0xfb,
-0x2f,0xe2,0xc5,0xb7,0x2d,0xdb,0x9b,0x9f,0x14,0x07,0x83,0xbc,0x88,0x7e,0x9e,0x0c,
-0x15,0xf2,0xea,0x2e,0x79,0xc3,0x41,0x9e,0xa9,0xc7,0x81,0xd1,0x3a,0x16,0x64,0x6b,
-0x1c,0xc9,0xc8,0xd6,0xb8,0x69,0x9b,0x37,0xfe,0x2f,0xf3,0x5e,0x11,0xfd,0x93,0x0d,
-0x0f,0x6b,0xf7,0xbc,0x6c,0x9b,0x1e,0xef,0xe7,0xa5,0x77,0xc9,0x4b,0xe8,0xfb,0xda,
-0x5c,0xfd,0xa5,0xba,0x78,0x73,0x97,0x3c,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,0x00,
-0x00,0x00,0x00,0x00,0x00,0x87,0xe8,0x6f,0x20,0x01,0xec,0xc5,0x00,0x00,0x01,0x00,
-};

arch/sh/kernel/time.c

-/*
+/* $Id: time.c,v 1.2 1999/10/11 13:12:02 gniibe Exp $
+ *
  *  linux/arch/sh/kernel/time.c
  *
- *  Copyright (C) 1999  Niibe Yutaka
+ *  Copyright (C) 1999  Tetsuya Okada & Niibe Yutaka
  *
  *  Some code taken from i386 version.
  *    Copyright (C) 1991, 1992, 1995  Linus Torvalds
  */
 
+#include <linux/config.h>
+
 #include <linux/errno.h>
 #include <linux/sched.h>
 #include <linux/kernel.h>
@@ -28,6 +31,11 @@
 #include <linux/timex.h>
 #include <linux/irq.h>
 
+#define TMU_TOCR_INIT	0x00
+#define TMU0_TCR_INIT	0x0020
+#define TMU_TSTR_INIT	1
+
+#if defined(__sh3__)
 #define TMU_TOCR	0xfffffe90	/* Byte access */
 #define TMU_TSTR	0xfffffe92	/* Byte access */
 
@@ -35,13 +43,63 @@
 #define TMU0_TCNT	0xfffffe98	/* Long access */
 #define TMU0_TCR	0xfffffe9c	/* Word access */
 
-#define TMU_TOCR_INIT	0x00
-#define TMU0_TCR_INIT	0x0020
-#define TMU_TSTR_INIT	1
-
-#define CLOCK_MHZ	(60/4)
 #define INTERVAL	37500 /* (1000000*CLOCK_MHZ/HZ/2) ??? */
 
+/* SH-3 RTC */
+#define R64CNT  	0xfffffec0
+#define RSECCNT 	0xfffffec2
+#define RMINCNT 	0xfffffec4
+#define RHRCNT  	0xfffffec6
+#define RWKCNT  	0xfffffec8
+#define RDAYCNT 	0xfffffeca
+#define RMONCNT 	0xfffffecc
+#define RYRCNT  	0xfffffece
+#define RSECAR  	0xfffffed0
+#define RMINAR  	0xfffffed2
+#define RHRAR   	0xfffffed4
+#define RWKAR   	0xfffffed6
+#define RDAYAR  	0xfffffed8
+#define RMONAR  	0xfffffeda
+#define RCR1    	0xfffffedc
+#define RCR2    	0xfffffede
+
+#elif defined(__SH4__)
+#define TMU_TOCR	0xffd80000	/* Byte access */
+#define TMU_TSTR	0xffd80004	/* Byte access */
+
+#define TMU0_TCOR	0xffd80008	/* Long access */
+#define TMU0_TCNT	0xffd8000c	/* Long access */
+#define TMU0_TCR	0xffd80010	/* Word access */
+
+#define INTERVAL	83333
+
+/* SH-4 RTC */
+#define R64CNT  	0xffc80000
+#define RSECCNT 	0xffc80004
+#define RMINCNT 	0xffc80008
+#define RHRCNT  	0xffc8000c
+#define RWKCNT  	0xffc80010
+#define RDAYCNT 	0xffc80014
+#define RMONCNT 	0xffc80018
+#define RYRCNT  	0xffc8001c  /* 16bit */
+#define RSECAR  	0xffc80020
+#define RMINAR  	0xffc80024
+#define RHRAR   	0xffc80028
+#define RWKAR   	0xffc8002c
+#define RDAYAR  	0xffc80030
+#define RMONAR  	0xffc80034
+#define RCR1    	0xffc80038
+#define RCR2    	0xffc8003c
+#endif
+
+#ifndef BCD_TO_BIN
+#define BCD_TO_BIN(val) ((val)=((val)&15) + ((val)>>4)*10)
+#endif
+
+#ifndef BIN_TO_BCD
+#define BIN_TO_BCD(val) ((val)=(((val)/10)<<4) + (val)%10)
+#endif
+
 extern rwlock_t xtime_lock;
 #define TICK_SIZE tick
 
@@ -82,14 +140,48 @@ void do_settimeofday(struct timeval *tv)
 	write_unlock_irq(&xtime_lock);
 }
 
-/*
- */
 static int set_rtc_time(unsigned long nowtime)
 {
-/* XXX  should be implemented XXXXXXXXXX */
-	int retval = -1;
+#ifdef CONFIG_SH_CPU_RTC
+	int retval = 0;
+	int real_seconds, real_minutes, cmos_minutes;
+
+	ctrl_outb(2, RCR2);  /* reset pre-scaler & stop RTC */
+
+	cmos_minutes = ctrl_inb(RMINCNT);
+	BCD_TO_BIN(cmos_minutes);
+
+	/*
+	 * since we're only adjusting minutes and seconds,
+	 * don't interfere with hour overflow. This avoids
+	 * messing with unknown time zones but requires your
+	 * RTC not to be off by more than 15 minutes
+	 */
+	real_seconds = nowtime % 60;
+	real_minutes = nowtime / 60;
+	if (((abs(real_minutes - cmos_minutes) + 15)/30) & 1)
+		real_minutes += 30;	/* correct for half hour time zone */
+	real_minutes %= 60;
+
+	if (abs(real_minutes - cmos_minutes) < 30) {
+		BIN_TO_BCD(real_seconds);
+		BIN_TO_BCD(real_minutes);
+		ctrl_outb(real_seconds, RSECCNT);
+		ctrl_outb(real_minutes, RMINCNT);
+	} else {
+		printk(KERN_WARNING
+		       "set_rtc_time: can't update from %d to %d\n",
+		       cmos_minutes, real_minutes);
+		retval = -1;
+	}
+
+	ctrl_outb(2, RCR2);  /* start RTC */
 
 	return retval;
+#else
+	/* XXX should support other clock devices? */
+	return -1;
+#endif
 }
 
 /* last time the RTC clock got updated */
@@ -131,14 +223,12 @@ static inline void do_timer_interrupt(int irq, void *dev_id, struct pt_regs *reg
  */
 static void timer_interrupt(int irq, void *dev_id, struct pt_regs *regs)
 {
-	unsigned long __dummy;
+	unsigned long timer_status;
 
 	/* Clear UNF bit */
-	asm volatile("mov.w	%1,%0\n\t"
-		     "and	%2,%0\n\t"
-		     "mov.w	%0,%1"
-		     : "=&z" (__dummy)
-		     : "m" (__m(TMU0_TCR)), "r" (~0x100));
+	timer_status = ctrl_inw(TMU0_TCR);
+	timer_status &= ~0x100;
+	ctrl_outw(timer_status, TMU0_TCR);
 
 	/*
 	 * Here we are in the timer irq handler. We just have irqs locally
@@ -187,16 +277,67 @@ static inline unsigned long mktime(unsigned int year, unsigned int mon,
 
 static unsigned long get_rtc_time(void)
 {
-/* XXX not implemented yet */
+#ifdef CONFIG_SH_CPU_RTC
+	unsigned int sec, min, hr, wk, day, mon, yr, yr100;
+
+ again:
+	ctrl_outb(1, RCR1);  /* clear CF bit */
+	do {
+		sec = ctrl_inb(RSECCNT);
+		min = ctrl_inb(RMINCNT);
+		hr  = ctrl_inb(RHRCNT);
+		wk  = ctrl_inb(RWKCNT);
+		day = ctrl_inb(RDAYCNT);
+		mon = ctrl_inb(RMONCNT);
+#if defined(__SH4__)
+		yr  = ctrl_inw(RYRCNT);
+		yr100 = (yr >> 8);
+		yr &= 0xff;
+#else
+		yr  = ctrl_inb(RYRCNT);
+		yr100 = (yr == 0x99) ? 0x19 : 0x20;
+#endif
+	} while ((ctrl_inb(RCR1) & 0x80) != 0);
+
+	BCD_TO_BIN(yr100);
+	BCD_TO_BIN(yr);
+	BCD_TO_BIN(mon);
+	BCD_TO_BIN(day);
+	BCD_TO_BIN(hr);
+	BCD_TO_BIN(min);
+	BCD_TO_BIN(sec);
+
+	if (yr > 99 || mon < 1 || mon > 12 || day > 31 || day < 1 ||
+	    hr > 23 || min > 59 || sec > 59) {
+		printk(KERN_ERR
+		       "SH RTC: invalid value, resetting to 1 Jan 2000\n");
+		ctrl_outb(2, RCR2);  /* reset, stop */
+		ctrl_outb(0, RSECCNT);
+		ctrl_outb(0, RMINCNT);
+		ctrl_outb(0, RHRCNT);
+		ctrl_outb(6, RWKCNT);
+		ctrl_outb(1, RDAYCNT);
+		ctrl_outb(1, RMONCNT);
+#if defined(__SH4__)
+		ctrl_outw(0x2000, RYRCNT);
+#else
+		ctrl_outb(0, RYRCNT);
+#endif
+		ctrl_outb(1, RCR2);  /* start */
+		goto again;
+	}
+
+	return mktime(yr100 * 100 + yr, mon, day, hr, min, sec);
+#else
+	/* XXX should support other clock devices? */
 	return 0;
+#endif
 }
 
 static struct irqaction irq0  = { timer_interrupt, SA_INTERRUPT, 0, "timer", NULL, NULL};
 
 void __init time_init(void)
 {
-	unsigned long __dummy;
-
 	xtime.tv_sec = get_rtc_time();
 	xtime.tv_usec = 0;
 
@@ -204,19 +345,12 @@ void __init time_init(void)
 	setup_irq(TIMER_IRQ, &irq0);
 
 	/* Start TMU0 */
-	asm volatile("mov	%1,%0\n\t"
-		     "mov.b	%0,%2		! external clock input\n\t"
-		     "mov	%3,%0\n\t"
-		     "mov.w	%0,%4		! enable timer0 interrupt\n\t"
-		     "mov.l	%5,%6\n\t"
-		     "mov.l	%5,%7\n\t"
-		     "mov	%8,%0\n\t"
-		     "mov.b	%0,%9"
-		     : "=&z" (__dummy)
-		     : "i" (TMU_TOCR_INIT), "m" (__m(TMU_TOCR)),
-		       "i" (TMU0_TCR_INIT), "m" (__m(TMU0_TCR)),
-		       "r" (INTERVAL), "m" (__m(TMU0_TCOR)), "m" (__m(TMU0_TCNT)),
-		       "i" (TMU_TSTR_INIT), "m" (__m(TMU_TSTR)));
+	ctrl_outb(TMU_TOCR_INIT,TMU_TOCR);
+	ctrl_outw(TMU0_TCR_INIT,TMU0_TCR);
+	ctrl_outl(INTERVAL,TMU0_TCOR);
+	ctrl_outl(INTERVAL,TMU0_TCNT);
+	ctrl_outb(TMU_TSTR_INIT,TMU_TSTR);
+
 #if 0
 	/* Start RTC */
 	asm volatile("");

arch/sh/kernel/traps.c

-/*
+/* $Id: traps.c,v 1.3 1999/09/21 14:37:19 gniibe Exp $
+ *
  *  linux/arch/sh/traps.c
  *
  *  SuperH version: Copyright (C) 1999  Niibe Yutaka
@@ -56,10 +57,6 @@ asmlinkage void do_##name(unsigned long r4, unsigned long r5, \
 #define VMALLOC_OFFSET (8*1024*1024)
 #define MODULE_RANGE (8*1024*1024)
 
-static void show_registers(struct pt_regs *regs)
-{/* Not implemented yet. */
-}	
-
 spinlock_t die_lock;
 
 void die(const char * str, struct pt_regs * regs, long err)
@@ -67,7 +64,7 @@ void die(const char * str, struct pt_regs * regs, long err)
 	console_verbose();
 	spin_lock_irq(&die_lock);
 	printk("%s: %04lx\n", str, err & 0xffff);
-	show_registers(regs);
+	show_regs(regs);
 	spin_unlock_irq(&die_lock);
 	do_exit(SIGSEGV);
 }

arch/sh/lib/Makefile

@@ -6,9 +6,7 @@
 	$(CC) -D__ASSEMBLY__ $(AFLAGS) -traditional -c $< -o $*.o
 
 L_TARGET = lib.a
-# L_OBJS  = checksum.o old-checksum.o semaphore.o delay.o \
-# 	usercopy.o getuser.o putuser.o
-L_OBJS  = delay.o memcpy.o memset.o memmove.o csum_partial_copy.o \
-	  wordcopy.o checksum.o # usercopy.o getuser.o putuser.o
+L_OBJS  = delay.o memcpy.o memset.o memmove.o old-checksum.o \
+	  checksum.o
 
 include $(TOPDIR)/Rules.make

arch/sh/lib/checksum.S

+/* $Id: checksum.S,v 1.1 1999/09/18 16:56:53 gniibe Exp $
+ *
+ * INET		An implementation of the TCP/IP protocol suite for the LINUX
+ *		operating system.  INET is implemented using the  BSD Socket
+ *		interface as the means of communication with the user level.
+ *
+ *		IP/TCP/UDP checksumming routines
+ *
+ * Authors:	Jorge Cwik, <jorge@laser.satlink.net>
+ *		Arnt Gulbrandsen, <agulbra@nvg.unit.no>
+ *		Tom May, <ftom@netcom.com>
+ *              Pentium Pro/II routines:
+ *              Alexander Kjeldaas <astor@guardian.no>
+ *              Finn Arne Gangstad <finnag@guardian.no>
+ *		Lots of code moved from tcp.c and ip.c; see those files
+ *		for more names.
+ *
+ * Changes:     Ingo Molnar, converted csum_partial_copy() to 2.1 exception
+ *			     handling.
+ *		Andi Kleen,  add zeroing on error
+ *                   converted to pure assembler
+ *
+ * SuperH version:  Copyright (C) 1999  Niibe Yutaka
+ *
+ *		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.
+ */
+
+#include <asm/errno.h>
+#include <linux/linkage.h>
+
+/*
+ * computes a partial checksum, e.g. for TCP/UDP fragments
+ */
+
+/*	
+ * unsigned int csum_partial(const unsigned char *buf, int len,
+ *                           unsigned int sum);
+ */
+
+.text
+ENTRY(csum_partial)
+	  /*
+	   * Experiments with Ethernet and SLIP connections show that buff
+	   * is aligned on either a 2-byte or 4-byte boundary.  We get at
+	   * least a twofold speedup on 486 and Pentium if it is 4-byte aligned.
+	   * Fortunately, it is easy to convert 2-byte alignment to 4-byte
+	   * alignment for the unrolled loop.
+	   */
+	mov	r5,r1
+	mov	r4,r0
+	tst	#2,r0		! Check alignment.
+	bt	2f		! Jump if alignment is ok.
+	!
+	add	#-2,r5		! Alignment uses up two bytes.
+	cmp/pz	r5		!
+	bt/s	1f		! Jump if we had at least two bytes.
+	 clrt
+	bra	6f
+	 add	#2,r5		! r5 was < 2.  Deal with it.
+1:
+	mov.w	@r4+,r0
+	extu.w	r0,r0
+	addc	r0,r6
+	bf	2f
+	add	#1,r6
+2:
+	mov	#-5,r0
+	shld	r0,r5
+	tst	r5,r5
+	bt/s	4f		! if it's =0, go to 4f
+	 clrt
+3:
+	mov.l	@r4+,r0
+	addc	r0,r6
+	mov.l	@r4+,r0
+	addc	r0,r6
+	mov.l	@r4+,r0
+	addc	r0,r6
+	mov.l	@r4+,r0
+	addc	r0,r6
+	mov.l	@r4+,r0
+	addc	r0,r6
+	mov.l	@r4+,r0
+	addc	r0,r6
+	mov.l	@r4+,r0
+	addc	r0,r6
+	mov.l	@r4+,r0
+	addc	r0,r6
+	movt	r0
+	dt	r5
+	bf/s	3b
+	 cmp/eq	#1,r0
+	mov	#0,r0
+	addc	r0,r6
+4:
+	mov	r1,r5
+	mov	#0x1c,r0
+	and	r0,r5
+	tst	r5,r5
+	bt/s	6f
+	 clrt
+	shlr2	r5
+5:
+	mov.l	@r4+,r0
+	addc	r0,r6
+	movt	r0
+	dt	r5
+	bf/s	5b
+	 cmp/eq	#1,r0
+	mov	#0,r0
+	addc	r0,r6
+6:
+	mov	r1,r5
+	mov	#3,r0
+	and	r0,r5
+	tst	r5,r5
+	bt	9f		! if it's =0 go to 9f
+	mov	#2,r1
+	cmp/hs  r1,r5
+	bf	7f
+	mov.w	@r4+,r0
+	extu.w	r0,r0
+	cmp/eq	r1,r5
+	bt/s	8f
+	 clrt
+	shll16	r0
+	addc	r0,r6
+7:
+	mov.b	@r4+,r0
+	extu.b	r0,r0
+8:
+	addc	r0,r6
+	mov	#0,r0
+	addc	r0,r6 
+9:
+	rts
+	 mov	r6,r0
+
+/*
+unsigned int csum_partial_copy_generic (const char *src, char *dst, int len, 
+					int sum, int *src_err_ptr, int *dst_err_ptr)
+ */ 
+
+/*
+ * Copy from ds while checksumming, otherwise like csum_partial
+ *
+ * The macros SRC and DST specify the type of access for the instruction.
+ * thus we can call a custom exception handler for all access types.
+ *
+ * FIXME: could someone double-check whether I haven't mixed up some SRC and
+ *	  DST definitions? It's damn hard to trigger all cases.  I hope I got
+ *	  them all but there's no guarantee.
+ */
+
+#define SRC(y...)			\
+	9999: y;			\
+	.section __ex_table, "a";	\
+	.long 9999b, 6001f	;	\
+	.previous
+
+#define DST(y...)			\
+	9999: y;			\
+	.section __ex_table, "a";	\
+	.long 9999b, 6002f	;	\
+	.previous
+
+ENTRY(csum_partial_copy_generic)
+	mov.l	r5,@-r15
+	mov.l	r6,@-r15
+
+	mov	#2,r0
+	tst	r0,r5			! Check alignment. 
+	bt	2f			! Jump if alignment is ok.
+	add	#-2,r6			! Alignment uses up two bytes.
+	cmp/pz	r6			! Jump if we had at least two bytes.
+	bt/s	1f
+	 clrt
+	bra	4f
+	 add	#2,r6			! ecx was < 2.  Deal with it.
+SRC(1:	mov.w	@r4+,r0	)
+DST(	mov.w	r0,@r5	)
+	add	#2,r5
+	extu.w	r0,r0
+	addc	r0,r7
+	mov	#0,r0
+	addc	r0,r7
+2:
+	mov	r6,r2
+	mov	#-5,r0
+	shld	r0,r6
+	tst	r6,r6
+	bf/s	2f
+	 clrt
+SRC(1:	mov.l	@r4+,r0	)
+SRC(	mov.l	@r4+,r1	)
+	addc	r0,r7
+DST(	mov.l	r0,@r5	)
+	add	#4,r5
+	addc	r1,r7
+DST(	mov.l	r1,@r5	)
+	add	#4,r5
+
+SRC(	mov.l	@r4+,r0	)
+SRC(	mov.l	@r4+,r1	)
+	addc	r0,r7
+DST(	mov.l	r0,@r5	)
+	add	#4,r5
+	addc	r1,r7
+DST(	mov.l	r1,@r5	)
+	add	#4,r5
+
+SRC(	mov.l	@r4+,r0 	)
+SRC(	mov.l	@r4+,r1	)
+	addc	r0,r7
+DST(	mov.l	r0,@r5	)
+	add	#4,r5
+	addc	r1,r7
+DST(	mov.l	r1,@r5	)
+	add	#4,r5
+
+SRC(	mov.l	@r4+,r0	)
+SRC(	mov.l	@r4+,r1	)
+	addc	r0,r7
+DST(	mov.l	r0,@r5	)
+	add	#4,r5
+	addc	r1,r7
+DST(	mov.l	r1,@r5	)
+	add	#4,r5
+	movt	r0
+	dt	r6
+	bf/s	1b
+	 cmp/eq	#1,r0
+	mov	#0,r0
+	addc	r0,r7
+
+2:	mov	r2,r6
+	mov	#0x1c,r0
+	and	r0,r6
+	cmp/pl	r6
+	bf/s	4f
+	 clrt
+	shlr2	r6
+SRC(3:	mov.l	@r4+,r0	)
+	addc	r0,r7
+DST(	mov.l	r0,@r5	)
+	add	#4,r5
+	movt	r0
+	dt	r6
+	bf/s	3b
+	 cmp/eq	#1,r0
+	mov	#0,r0
+	addc	r0,r7
+4:	mov	r2,r6
+	mov	#3,r0
+	and	r0,r6
+	cmp/pl	r6
+	bf	7f
+	mov	#2,r1
+	cmp/hs	r1,r6
+	bf	5f
+SRC(	mov.w	@r4+,r0	)
+DST(	mov.w	r0,@r5	)
+	extu.w	r0,r0
+	add	#2,r5
+	cmp/eq	r1,r6
+	bt/s	6f
+	 clrt
+	shll16	r0
+	addc	r0,r7
+SRC(5:	mov.b	@r4+,r0	)
+DST(	mov.b	r0,@r5	)
+	extu.b	r0,r0
+6:	addc	r0,r7
+	mov	#0,r0
+	addc	r0,r7
+7:
+5000:
+
+# Exception handler:
+.section .fixup, "ax"							
+
+6001:
+	mov.l	@(8,r15),r0			! src_err_ptr
+	mov	#-EFAULT,r1
+	mov.l	r1,@r0
+
+	! zero the complete destination - computing the rest
+	! is too much work 
+	mov.l	@(4,r15),r5		! dst
+	mov.l	@r15,r6			! len
+	mov	#0,r7
+1:	mov.b	r7,@r5
+	dt	r6
+	bf/s	1b
+	 add	#1,r5
+	mov.l	8000f,r0
+	jmp	@r0
+	 nop
+	.balign 4
+8000:	.long	5000b
+
+6002:
+	mov.l	@(12,r15),r0			! dst_err_ptr
+	mov	#-EFAULT,r1
+	mov.l	r1,@r0
+	mov.l	8001f,r0
+	jmp	@r0
+	 nop
+	.balign 4
+8001:	.long	5000b
+
+.previous
+	add	#8,r15
+	rts
+	 mov	r7,r0

arch/sh/lib/checksum.c

-/*
- * Taken from:
- * arch/alpha/lib/checksum.c
- *
- * This file contains network checksum routines that are better done
- * in an architecture-specific manner due to speed..
- */
- 
-#include <linux/string.h>
-
-#include <asm/byteorder.h>
-
-static inline unsigned short from64to16(unsigned long long x)
-{
-	/* add up 32-bit words for 33 bits */
-	x = (x & 0xffffffff) + (x >> 32);
-	/* add up 16-bit and 17-bit words for 17+c bits */
-	x = (x & 0xffff) + (x >> 16);
-	/* add up 16-bit and 2-bit for 16+c bit */
-	x = (x & 0xffff) + (x >> 16);
-	/* add up carry.. */
-	x = (x & 0xffff) + (x >> 16);
-	return x;
-}
-
-/*
- * computes the checksum of the TCP/UDP pseudo-header
- * returns a 16-bit checksum, already complemented.
- */
-unsigned short int csum_tcpudp_magic(unsigned long saddr,
-				   unsigned long daddr,
-				   unsigned short len,
-				   unsigned short proto,
-				   unsigned int sum)
-{
-	return ~from64to16(saddr + daddr + sum +
-		((unsigned long) ntohs(len) << 16) +
-		((unsigned long) proto << 8));
-}
-
-unsigned int csum_tcpudp_nofold(unsigned long saddr,
-				unsigned long daddr,
-				unsigned short len,
-				unsigned short proto,
-				unsigned int sum)
-{
-	unsigned long long result;
-
-	result = (saddr + daddr + sum +
-		  ((unsigned long) ntohs(len) << 16) +
-		  ((unsigned long) proto << 8));
-
-	/* Fold down to 32-bits so we don't loose in the typedef-less 
-	   network stack.  */
-	/* 64 to 33 */
-	result = (result & 0xffffffff) + (result >> 32);
-	/* 33 to 32 */
-	result = (result & 0xffffffff) + (result >> 32);
-	return result;
-}
-
-/*
- * Do a 64-bit checksum on an arbitrary memory area..
- *
- * This isn't a great routine, but it's not _horrible_ either. The
- * inner loop could be unrolled a bit further, and there are better
- * ways to do the carry, but this is reasonable.
- */
-static inline unsigned long do_csum(const unsigned char * buff, int len)
-{
-	int odd, count;
-	unsigned long long result = 0;
-
-	if (len <= 0)
-		goto out;
-	odd = 1 & (unsigned long) buff;
-	if (odd) {
-		result = *buff << 8;
-		len--;
-		buff++;
-	}
-	count = len >> 1;		/* nr of 16-bit words.. */
-	if (count) {
-		if (2 & (unsigned long) buff) {
-			result += *(unsigned short *) buff;
-			count--;
-			len -= 2;
-			buff += 2;
-		}
-		count >>= 1;		/* nr of 32-bit words.. */
-		if (count) {
-			if (4 & (unsigned long) buff) {
-				result += *(unsigned int *) buff;
-				count--;
-				len -= 4;
-				buff += 4;
-			}
-			count >>= 1;	/* nr of 64-bit words.. */
-			if (count) {
-				unsigned long carry = 0;
-				do {
-					unsigned long w = *(unsigned long *) buff;
-					count--;
-					buff += 8;
-					result += carry;
-					result += w;
-					carry = (w > result);
-				} while (count);
-				result += carry;
-				result = (result & 0xffffffff) + (result >> 32);
-			}
-			if (len & 4) {
-				result += *(unsigned int *) buff;
-				buff += 4;
-			}
-		}
-		if (len & 2) {
-			result += *(unsigned short *) buff;
-			buff += 2;
-		}
-	}
-	if (len & 1)
-		result += *buff;
-	result = from64to16(result);
-	if (odd)
-		result = ((result >> 8) & 0xff) | ((result & 0xff) << 8);
-out:
-	return result;
-}
-
-/*
- *	This is a version of ip_compute_csum() optimized for IP headers,
- *	which always checksum on 4 octet boundaries.
- */
-unsigned short ip_fast_csum(unsigned char * iph, unsigned int ihl)
-{
-	return ~do_csum(iph,ihl*4);
-}
-
-/*
- * computes the checksum of a memory block at buff, length len,
- * and adds in "sum" (32-bit)
- *
- * returns a 32-bit number suitable for feeding into itself
- * or csum_tcpudp_magic
- *
- * this function must be called with even lengths, except
- * for the last fragment, which may be odd
- *
- * it's best to have buff aligned on a 32-bit boundary
- */
-unsigned int csum_partial(const unsigned char * buff, int len, unsigned int sum)
-{
-	unsigned long long result = do_csum(buff, len);
-
-	/* add in old sum, and carry.. */
-	result += sum;
-	/* 32+c bits -> 32 bits */
-	result = (result & 0xffffffff) + (result >> 32);
-	return result;
-}
-
-/*
- * this routine is used for miscellaneous IP-like checksums, mainly
- * in icmp.c
- */
-unsigned short ip_compute_csum(unsigned char * buff, int len)
-{
-	return ~from64to16(do_csum(buff,len));
-}

arch/sh/lib/csum_partial_copy.c

-/*
- * INET		An implementation of the TCP/IP protocol suite for the LINUX
- *		operating system.  INET is implemented using the  BSD Socket
- *		interface as the means of communication with the user level.
- *
- *		MIPS specific IP/TCP/UDP checksumming routines
- *
- * Authors:	Ralf Baechle, <ralf@waldorf-gmbh.de>
- *		Lots of code moved from tcp.c and ip.c; see those files
- *		for more names.
- *
- *		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.
- *
- * $Id: csum_partial_copy.c,v 1.2 1998/09/16 13:29:32 ralf Exp $
- */
-#include <net/checksum.h>
-#include <linux/types.h>
-#include <asm/byteorder.h>
-#include <asm/string.h>
-#include <asm/uaccess.h>
-
-/*
- * copy while checksumming, otherwise like csum_partial
- */
-unsigned int csum_partial_copy(const char *src, char *dst, 
-                               int len, unsigned int sum)
-{
-	/*
-	 * It's 2:30 am and I don't feel like doing it real ...
-	 * This is lots slower than the real thing (tm)
-	 */
-	sum = csum_partial(src, len, sum);
-	memcpy(dst, src, len);
-
-	return sum;
-}
-
-/*
- * Copy from userspace and compute checksum.  If we catch an exception
- * then zero the rest of the buffer.
- */
-unsigned int csum_partial_copy_from_user (const char *src, char *dst,
-                                          int len, unsigned int sum,
-                                          int *err_ptr)
-{
-	int missing;
-
-	missing = copy_from_user(dst, src, len);
-	if (missing) {
-		memset(dst + len - missing, 0, missing);
-		*err_ptr = -EFAULT;
-	}
-		
-	return csum_partial(dst, len, sum);
-}
-
-/*
- * Copy to userspace and compute checksum.
- */
-unsigned int csum_partial_copy_to_user (const char *src, char *dst,
-                                        int len, unsigned int sum,
-                                        int *err_ptr)
-{
-	sum = csum_partial(src, len, sum);
-
-	if (copy_to_user(dst, src, len)) {
-		*err_ptr = -EFAULT;
-		return sum;
-	}
-
-	return sum;
-}

arch/sh/lib/memcpy.S

-! Taken from newlib-1.8.0
+/* $Id: memcpy.S,v 1.3 1999/09/28 11:32:48 gniibe Exp $
+ *
+ * "memcpy" implementation of SuperH
+ *
+ * Copyright (C) 1999  Niibe Yutaka
+ *
+ */
 
-!
-! Fast SH memcpy
-!
-! by Toshiyasu Morita (tm@netcom.com)
-! hacked by J"orn Rernnecke (amylaar@cygnus.co.uk) ("o for o-umlaut)
-!
-! Entry: r4: destination pointer
-!        r5: source pointer
-!        r6: byte count
-!
-! Exit:  r0: destination pointer
-!        r1-r7: trashed
-!
-! Notes: Usually one wants to do small reads and write a longword, but
-!        unfortunately it is difficult in some cases to concatanate bytes
-!        into a longword on the SH, so this does a longword read and small
-!        writes.
-!
-! This implementation makes two assumptions about how it is called:
-!
-! 1.: If the byte count is nonzero, the address of the last byte to be
-!     copied is unsigned greater than the address of the first byte to
-!     be copied.  This could be easily swapped for a signed comparison,
-!     but the algorithm used needs some comparison.
-!
-! 2.: When there are two or three bytes in the last word of an 11-or-bore
-!     bytes memory chunk to b copied, the rest of the word can be read
-!     without size effects.
-!     This could be easily changed by increasing the minumum size of
-!     a fast memcpy and the amount subtracted from r7 before L_2l_loop be 2,
-!     however, this would cost a few extra cyles on average.
-!
+/*
+ * void *memcpy(void *dst, const void *src, size_t n);
+ * No overlap between the memory of DST and of SRC are assumed.
+ */
 
 #include <linux/linkage.h>
 ENTRY(memcpy)
-	! Big endian version copies with decreasing addresses.
-	mov	r4,r0
-	add	r6,r0
-	sub	r4,r5
-	mov	#11,r1
-	cmp/hs	r1,r6
-	bf/s	L_small
+	tst	r6,r6
+	bt/s	9f		! if n=0, do nothing
+	 mov	r4,r0
+	sub	r4,r5		! From here, r5 has the distance to r0
+	add	r6,r0		! From here, r0 points the end of copying point
+	mov	#12,r1
+	cmp/gt	r6,r1
+	bt/s	7f		! if it's too small, copy a byte at once
 	 add	#-1,r5
-	mov	r5,r3
-	add	r0,r3
-	shlr	r3
-	bt/s	L_even
-	 mov	r4,r7
-	mov.b	@(r0,r5),r2
-	add	#-1,r3
-	mov.b	r2,@-r0
-L_even:
-	tst	#1,r0
-	add	#-1,r5
-	bf/s	L_odddst
-	 add	#8,r7
-	tst	#2,r0
-	bt	L_al4dst
-	add	#-1,r3
-	mov.w	@(r0,r5),r1
-	mov.w	r1,@-r0
-L_al4dst:
-	shlr	r3
-	bt	L_al4both
-	mov.w	@(r0,r5),r1
-	swap.w	r1,r1
-	add	#4,r7
-	add	#-4,r5
-	.align	2
-L_2l_loop:
-	mov.l	@(r0,r5),r2
-	xtrct	r2,r1
-	mov.l	r1,@-r0
-	cmp/hs	r7,r0
-	mov.l	@(r0,r5),r1
-	xtrct	r1,r2
-	mov.l	r2,@-r0
-	bt	L_2l_loop
-	bra	L_cleanup
-	 add	#5,r5
+	add	#1,r5
+	!			From here, r6 is free
+	!
+	!      r4   -->  [ ...  ] DST             [ ...  ] SRC
+	!	         [ ...  ]                 [ ...  ]
+	!	           :                        :
+	!      r0   -->  [ ...  ]       r0+r5 --> [ ...  ]
+	!
+	!
+	mov	r5,r1
+	mov	#3,r2
+	and	r2,r1
+	shll2	r1
+	mov	r0,r3		! Save the value on R0 to R3
+	mova	jmptable,r0
+	add	r1,r0
+	mov.l	@r0,r1
+	jmp	@r1
+	 mov	r3,r0		! and back to R0
+	.balign	4
+jmptable:
+	.long	case0
+	.long	case1
+	.long	case2
+	.long	case3
 
-	nop ! avoid nop in executed code.
-L_al4both:
-	add	#-2,r5
-	.align	2
-L_al4both_loop:
-	mov.l	@(r0,r5),r1
-	cmp/hs	r7,r0
-	bt/s	L_al4both_loop
+	! copy a byte at once
+7:	mov	r4,r2
+	add	#1,r2
+8:
+	cmp/hi	r2,r0
+	mov.b	@(r0,r5),r1
+	bt/s	8b			! while (r0>r2)
+	 mov.b	r1,@-r0
+9:
+	rts
+	 nop
+
+case0:
+	!
+	!	GHIJ KLMN OPQR -->  GHIJ KLMN OPQR
+	!
+	! First, align to long word boundary
+	mov	r0,r3
+	and	r2,r3
+	tst	r3,r3
+	bt/s	2f
+	 add	#-4,r5
+	add	#3,r5
+1:	dt	r3
+	mov.b	@(r0,r5),r1
+	bf/s	1b
+	 mov.b	r1,@-r0
+	!
+	add	#-3,r5
+2:	! Second, copy a long word at once
+	mov	r4,r2
+	add	#7,r2
+3:	mov.l	@(r0,r5),r1
+	cmp/hi	r2,r0
+	bt/s	3b
 	 mov.l	r1,@-r0
-	bra	L_cleanup
+	!
+	! Third, copy a byte at once, if necessary
+	cmp/eq	r4,r0
+	bt/s	9b
 	 add	#3,r5
+	bra	8b
+	 add	#-6,r2
 
-	nop ! avoid nop in executed code.
-L_odddst:
-	shlr	r3
-	bt	L_al4src
-	mov.w	@(r0,r5),r1
-	mov.b	r1,@-r0
-	shlr8	r1
-	mov.b	r1,@-r0
-L_al4src:
-	add	#-2,r5
-	.align	2
-L_odd_loop:
-	mov.l	@(r0,r5),r2
-	cmp/hs	r7,r0
-	mov.b	r2,@-r0
-	shlr8	r2
-	mov.w	r2,@-r0
-	shlr16	r2
-	mov.b	r2,@-r0
-	bt	L_odd_loop
-
-	add	#3,r5
-L_cleanup:
-L_small:
+case1:
+	!
+	!	GHIJ KLMN OPQR -->  ...G HIJK LMNO PQR.
+	!
+	! First, align to long word boundary
+	mov	r0,r3
+	and	r2,r3
+	tst	r3,r3
+	bt/s	2f
+	 add	#-1,r5
+1:	dt	r3
+	mov.b	@(r0,r5),r1
+	bf/s	1b
+	 mov.b	r1,@-r0
+	!
+2:	! Second, read a long word and write a long word at once
+	mov.l	@(r0,r5),r1
+	add	#-4,r5
+	mov	r4,r2
+	add	#7,r2
+	!
+#ifdef __LITTLE_ENDIAN__
+3:	mov	r1,r3		! RQPO
+	shll16	r3
+	shll8	r3		! Oxxx
+	mov.l	@(r0,r5),r1	! NMLK
+	mov	r1,r6
+	shlr8	r6		! xNML
+	or	r6,r3		! ONML
+	cmp/hi	r2,r0
+	bt/s	3b
+	 mov.l	r3,@-r0
+#else
+3:	mov	r1,r3		! OPQR
+	shlr16	r3
+	shlr8	r3		! xxxO
+	mov.l	@(r0,r5),r1	! KLMN
+	mov	r1,r6
+	shll8	r6		! LMNx
+	or	r6,r3		! LMNO
+	cmp/hi	r2,r0
+	bt/s	3b
+	 mov.l	r3,@-r0
+#endif
+	!
+	! Third, copy a byte at once, if necessary
 	cmp/eq	r4,r0
-	bt	L_ready
-	add	#1,r4
-	.align	2
-L_cleanup_loop:
-	mov.b	@(r0,r5),r2
+	bt/s	9b
+	 add	#4,r5
+	bra	8b
+	 add	#-6,r2
+
+case2:
+	!
+	!	GHIJ KLMN OPQR -->  ..GH IJKL MNOP QR..
+	!
+	! First, align to word boundary
+	tst	#1,r0
+	bt/s	2f
+	 add	#-1,r5
+	mov.b	@(r0,r5),r1
+	mov.b	r1,@-r0
+	!
+2:	! Second, read a word and write a word at once
+	add	#-1,r5
+	mov	r4,r2
+	add	#3,r2
+	!
+3:	mov.w	@(r0,r5),r1
+	cmp/hi	r2,r0
+	bt/s	3b
+	 mov.w	r1,@-r0
+	!
+	! Third, copy a byte at once, if necessary
 	cmp/eq	r4,r0
-	mov.b	r2,@-r0
-	bf	L_cleanup_loop
-L_ready:
+	bt/s	9b
+	 add	#1,r5
+	mov.b	@(r0,r5),r1
 	rts
-	 nop
+	 mov.b	r1,@-r0
+
+case3:
+	!
+	!	GHIJ KLMN OPQR -->  .GHI JKLM NOPQ R...
+	!
+	! First, align to long word boundary
+	mov	r0,r3
+	and	r2,r3
+	tst	r3,r3
+	bt/s	2f
+	 add	#-1,r5
+1:	dt	r3
+	mov.b	@(r0,r5),r1
+	bf/s	1b
+	 mov.b	r1,@-r0
+	!
+2:	! Second, read a long word and write a long word at once
+	add	#-2,r5
+	mov.l	@(r0,r5),r1
+	add	#-4,r5
+	mov	r4,r2
+	add	#7,r2
+	!
+#ifdef __LITTLE_ENDIAN__
+3:	mov	r1,r3		! RQPO
+	shll8	r3		! QPOx
+	mov.l	@(r0,r5),r1	! NMLK
+	mov	r1,r6
+	shlr16	r6
+	shlr8	r6		! xxxN
+	or	r6,r3		! QPON
+	cmp/hi	r2,r0
+	bt/s	3b
+	 mov.l	r3,@-r0
+#else
+3:	mov	r1,r3		! OPQR
+	shlr8	r3		! xOPQ
+	mov.l	@(r0,r5),r1	! KLMN
+	mov	r1,r6
+	shll16	r6
+	shll8	r6		! Nxxx
+	or	r6,r3		! NOPQ
+	cmp/hi	r2,r0
+	bt/s	3b
+	 mov.l	r3,@-r0
+#endif
+	!
+	! Third, copy a byte at once, if necessary
+	cmp/eq	r4,r0
+	bt/s	9b
+	 add	#6,r5
+	bra	8b
+	 add	#-6,r2

arch/sh/lib/memmove.S

+/* $Id: memmove.S,v 1.2 1999/09/21 12:55:49 gniibe Exp $
+ *
+ * "memmove" implementation of SuperH
+ *
+ * Copyright (C) 1999  Niibe Yutaka
+ *
+ */
+
+/*
+ * void *memmove(void *dst, const void *src, size_t n);
+ * The memory areas may overlap.
+ */
+
 #include <linux/linkage.h>
 ENTRY(memmove)
-	mov.l	r8,@-r15
-	mov.l	r9,@-r15
-	mov.l	r14,@-r15
-	sts.l	pr,@-r15
-	add	#-28,r15
-	mov	r15,r14
-	mov.l	r4,@r14
-	mov.l	r5,@(4,r14)
-	mov.l	r6,@(8,r14)
-	mov.l	@r14,r1
-	mov.l	r1,@(12,r14)
-	mov.l	@(4,r14),r1
-	mov.l	r1,@(16,r14)
-	mov.l	@(12,r14),r1
-	mov.l	@(16,r14),r2
-	sub	r2,r1
-	mov.l	@(8,r14),r2
-	cmp/hs	r2,r1
-	bt	.L54
-	bra	.L2
-	nop
-.L54:
-	mov.l	@(8,r14),r1
-	mov	#15,r2
-	cmp/gt	r2,r1
-	bt	.LF100
-	bra	.L52
-	nop
-.LF100:
-	mov.l	@(12,r14),r2
-	neg	r2,r1
-	mov	#3,r2
-	and	r1,r2
-	mov.l	@(8,r14),r1
-	mov	r1,r9
-	sub	r2,r9
-	mov	r9,r2
-	mov.l	r2,@(8,r14)
-.L4:
-	mov.l	@(12,r14),r2
-	neg	r2,r1
-	mov	#3,r2
-	and	r1,r2
-	mov.l	r2,@(20,r14)
-.L7:
-	mov.l	@(20,r14),r1
-	cmp/pl	r1
-	bt	.L9
-	bra	.L6
-	nop
-	.align 2
-.L9:
-	mov	r14,r2
-	mov	r14,r1
-	add	#24,r1
-	mov.l	@(16,r14),r2
-	mov.b	@r2,r3
-	mov.b	r3,@r1
-	mov.l	@(16,r14),r1
-	mov	r1,r2
-	add	#1,r2
-	mov.l	r2,@(16,r14)
-	mov.l	@(20,r14),r1
-	mov	r1,r2
-	add	#-1,r2
-	mov.l	r2,@(20,r14)
-	mov.l	@(12,r14),r1
-	mov	r14,r2
-	mov	r14,r3
-	add	#24,r3
-	mov.b	@r3,r2
-	mov.b	r2,@r1
-	mov.l	@(12,r14),r1
-	mov	r1,r2
-	add	#1,r2
-	mov.l	r2,@(12,r14)
-	bra	.L7
-	nop
-	.align 2
-.L8:
-.L6:
-	bra	.L5
-	nop
-	.align 2
-.L10:
-	bra	.L4
-	nop
-	.align 2
-.L5:
-	nop
-.L11:
-	mov.l	@(16,r14),r1
+	! if dest > src, call memcpy (it copies in decreasing order)
+	cmp/hi	r5,r4
+	bf	1f
+	mov.l	2f,r0
+	jmp	@r0
+	 nop
+	.balign 4
+2:	.long	SYMBOL_NAME(memcpy)
+1:
+	sub	r5,r4		! From here, r4 has the distance to r0
+	tst	r6,r6
+	bt/s	9f		! if n=0, do nothing
+	 mov	r5,r0
+	add	r6,r5
+	mov	#12,r1
+	cmp/gt	r6,r1
+	bt/s	8f		! if it's too small, copy a byte at once
+	 add	#-1,r4
+	add	#1,r4
+	!
+	!                [ ...  ] DST             [ ...  ] SRC
+	!	         [ ...  ]                 [ ...  ]
+	!	           :                        :
+	!      r0+r4-->  [ ...  ]       r0    --> [ ...  ]
+	!	           :                        :
+	!	         [ ...  ]                 [ ...  ]
+	!			        r5    -->
+	!
+	mov	r4,r1
 	mov	#3,r2
-	and	r1,r2
-	tst	r2,r2
-	bf	.L14
-	mov	r15,r2
-	mov.l	@(12,r14),r1
-	mov.l	@(16,r14),r2
-	mov.l	@(8,r14),r7
-	mov	r7,r3
-	shlr2	r3
-	mov	r1,r4
-	mov	r2,r5
-	mov	r3,r6
-	mov.l	.L46,r8
-	jsr	@r8
-	nop
-	bra	.L15
-	nop
-	.align 2
-.L14:
-	mov	r15,r2
-	mov.l	@(12,r14),r1
-	mov.l	@(16,r14),r2
-	mov.l	@(8,r14),r7
-	mov	r7,r3
-	shlr2	r3
-	mov	r1,r4
-	mov	r2,r5
-	mov	r3,r6
-	mov.l	.L47,r8
-	jsr	@r8
-	nop
-.L15:
-	mov.l	@(8,r14),r1
-	mov	#-4,r2
-	and	r2,r1
-	mov.l	@(16,r14),r2
-	add	r2,r1
-	mov.l	r1,@(16,r14)
-	mov.l	@(8,r14),r1
-	mov	#-4,r2
 	and	r2,r1
-	mov.l	@(12,r14),r2
-	add	r2,r1
-	mov.l	r1,@(12,r14)
-	mov.l	@(8,r14),r1
-	mov	#3,r2
-	and	r1,r2
-	mov.l	r2,@(8,r14)
-.L13:
-.L52:
-	bra	.L3
-	nop
-	.align 2
-.L16:
-	bra	.L11
-	nop
-	.align 2
-.L12:
-.L3:
-	nop
-.L17:
-	mov.l	@(8,r14),r1
-	mov.l	r1,@(20,r14)
-.L20:
-	mov.l	@(20,r14),r1
-	cmp/pl	r1
-	bt	.L22
-	bra	.L19
-	nop
-	.align 2
-.L22:
-	mov	r14,r2
-	mov	r14,r1
-	add	#24,r1
-	mov.l	@(16,r14),r2
-	mov.b	@r2,r3
-	mov.b	r3,@r1
-	mov.l	@(16,r14),r1
-	mov	r1,r2
-	add	#1,r2
-	mov.l	r2,@(16,r14)
-	mov.l	@(20,r14),r1
-	mov	r1,r2
-	add	#-1,r2
-	mov.l	r2,@(20,r14)
-	mov.l	@(12,r14),r1
-	mov	r14,r2
-	mov	r14,r3
-	add	#24,r3
-	mov.b	@r3,r2
-	mov.b	r2,@r1
-	mov.l	@(12,r14),r1
+	shll2	r1
+	mov	r0,r3		! Save the value on R0 to R3
+	mova	jmptable,r0
+	add	r1,r0
+	mov.l	@r0,r1
+	jmp	@r1
+	 mov	r3,r0		! and back to R0
+	.balign	4
+jmptable:
+	.long	case0
+	.long	case1
+	.long	case2
+	.long	case3
+
+	! copy a byte at once
+8:	mov.b	@r0+,r1
+	cmp/hs	r5,r0
+	bf/s	8b			! while (r0<r5)
+	 mov.b	r1,@(r0,r4)
+	add	#1,r4
+9:
+	add	r4,r0
+	rts
+	 sub	r6,r0
+
+case_none:
+	bra	8b
+	 add	#-1,r4
+
+case0:
+	!
+	!	GHIJ KLMN OPQR -->  GHIJ KLMN OPQR
+	!
+	! First, align to long word boundary
+	mov	r0,r3
+	and	r2,r3
+	tst	r3,r3
+	bt/s	2f
+	 add	#-1,r4
+	mov	#4,r2
+	sub	r3,r2
+1:	dt	r2
+	mov.b	@r0+,r1
+	bf/s	1b
+	 mov.b	r1,@(r0,r4)
+	!
+2:	! Second, copy a long word at once
+	add	#-3,r4
+	add	#-3,r5
+3:	mov.l	@r0+,r1
+	cmp/hs	r5,r0
+	bf/s	3b
+	 mov.l	r1,@(r0,r4)
+	add	#3,r5
+	!
+	! Third, copy a byte at once, if necessary
+	cmp/eq	r5,r0
+	bt/s	9b
+	 add	#4,r4
+	bra	8b
+	 add	#-1,r4
+
+case3:
+	!
+	!	GHIJ KLMN OPQR -->  ...G HIJK LMNO PQR.
+	!
+	! First, align to long word boundary
+	mov	r0,r3
+	and	r2,r3
+	tst	r3,r3
+	bt/s	2f
+	 add	#-1,r4
+	mov	#4,r2
+	sub	r3,r2
+1:	dt	r2
+	mov.b	@r0+,r1
+	bf/s	1b
+	 mov.b	r1,@(r0,r4)
+	!
+2:	! Second, read a long word and write a long word at once
+	add	#-2,r4
+	mov.l	@(r0,r4),r1
+	add	#-7,r5
+	add	#-4,r4
+	!
+#ifdef __LITTLE_ENDIAN__
+	shll8	r1
+3:	mov	r1,r3		! JIHG
+	shlr8	r3		! xJIH
+	mov.l	@r0+,r1		! NMLK
 	mov	r1,r2
-	add	#1,r2
-	mov.l	r2,@(12,r14)
-	bra	.L20
-	nop
-	.align 2
-.L21:
-.L19:
-	bra	.L18
-	nop
-	.align 2
-.L23:
-	bra	.L17
-	nop
-	.align 2
-.L18:
-	bra	.L24
-	nop
-	.align 2
-.L2:
-	mov.l	@(16,r14),r1
-	mov.l	@(8,r14),r2
-	add	r2,r1
-	mov.l	r1,@(16,r14)
-	mov.l	@(12,r14),r1
-	mov.l	@(8,r14),r2
-	add	r2,r1
-	mov.l	r1,@(12,r14)
-	mov.l	@(8,r14),r1
-	mov	#15,r2
-	cmp/gt	r2,r1
-	bt	.LF101
-	bra	.L53
-	nop
-.LF101:
-	mov.l	@(12,r14),r1
-	mov	#3,r2
-	and	r1,r2
-	mov.l	@(8,r14),r1
-	mov	r1,r9
-	sub	r2,r9
-	mov	r9,r2
-	mov.l	r2,@(8,r14)
-.L26:
-	mov.l	@(12,r14),r1
-	mov	#3,r2
-	and	r1,r2
-	mov.l	r2,@(20,r14)
-.L29:
-	mov.l	@(20,r14),r1
-	cmp/pl	r1
-	bt	.L31
-	bra	.L28
-	nop
-	.align 2
-.L31:
-	mov.l	@(16,r14),r1
-	mov	r1,r2
-	add	#-1,r2
-	mov.l	r2,@(16,r14)
-	mov	r14,r2
-	mov	r14,r1
-	add	#24,r1
-	mov.l	@(16,r14),r2
-	mov.b	@r2,r3
-	mov.b	r3,@r1
-	mov.l	@(12,r14),r1
-	mov	r1,r2
-	add	#-1,r2
-	mov.l	r2,@(12,r14)
-	mov.l	@(20,r14),r1
-	mov	r1,r2
-	add	#-1,r2
-	mov.l	r2,@(20,r14)
-	mov.l	@(12,r14),r1
-	mov	r14,r2
-	mov	r14,r3
-	add	#24,r3
-	mov.b	@r3,r2
-	mov.b	r2,@r1
-	bra	.L29
-	nop
-	.align 2
-.L30:
-.L28:
-	bra	.L27
-	nop
-	.align 2
-.L32:
-	bra	.L26
-	nop
-	.align 2
-.L27:
-	nop
-.L33:
-	mov.l	@(16,r14),r1
-	mov	#3,r2
-	and	r1,r2
-	tst	r2,r2
-	bf	.L36
-	mov	r15,r2
-	mov.l	@(12,r14),r1
-	mov.l	@(16,r14),r2
-	mov.l	@(8,r14),r7
-	mov	r7,r3
-	shlr2	r3
-	mov	r1,r4
-	mov	r2,r5
-	mov	r3,r6
-	mov.l	.L48,r8
-	jsr	@r8
-	nop
-	bra	.L37
-	nop
-	.align 2
-.L36:
-	mov	r15,r2
-	mov.l	@(12,r14),r1
-	mov.l	@(16,r14),r2
-	mov.l	@(8,r14),r7
-	mov	r7,r3
-	shlr2	r3
-	mov	r1,r4
-	mov	r2,r5
-	mov	r3,r6
-	mov.l	.L49,r8
-	jsr	@r8
-	nop
-.L37:
-	mov.l	@(8,r14),r1
-	mov	#-4,r2
-	and	r2,r1
-	mov.l	@(16,r14),r2
-	mov	r2,r9
-	sub	r1,r9
-	mov	r9,r1
-	mov.l	r1,@(16,r14)
-	mov.l	@(8,r14),r1
-	mov	#-4,r2
-	and	r2,r1
-	mov.l	@(12,r14),r2
-	mov	r2,r9
-	sub	r1,r9
-	mov	r9,r1
-	mov.l	r1,@(12,r14)
-	mov.l	@(8,r14),r1
-	mov	#3,r2
-	and	r1,r2
-	mov.l	r2,@(8,r14)
-.L35:
-.L53:
-	bra	.L25
-	nop
-	.align 2
-.L38:
-	bra	.L33
-	nop
-	.align 2
-.L34:
-.L25:
-	nop
-.L39:
-	mov.l	@(8,r14),r1
-	mov.l	r1,@(20,r14)
-.L42:
-	mov.l	@(20,r14),r1
-	cmp/pl	r1
-	bt	.L44
-	bra	.L41
-	nop
-	.align 2
-.L44:
-	mov.l	@(16,r14),r1
+	shll16	r2
+	shll8	r2		! Kxxx
+	or	r2,r3		! KJIH
+	cmp/hs	r5,r0
+	bf/s	3b
+	 mov.l	r3,@(r0,r4)
+#else
+	shlr8	r1
+3:	mov	r1,r3		! GHIJ
+	shll8	r3		! HIJx
+	mov.l	@r0+,r1		! KLMN
 	mov	r1,r2
-	add	#-1,r2
-	mov.l	r2,@(16,r14)
-	mov	r14,r2
-	mov	r14,r1
-	add	#24,r1
-	mov.l	@(16,r14),r2
-	mov.b	@r2,r3
-	mov.b	r3,@r1
-	mov.l	@(12,r14),r1
+	shlr16	r2
+	shlr8	r2		! xxxK
+	or	r2,r3		! HIJK
+	cmp/hs	r5,r0
+	bf/s	3b
+	 mov.l	r3,@(r0,r4)
+#endif
+	add	#7,r5
+	!
+	! Third, copy a byte at once, if necessary
+	cmp/eq	r5,r0
+	bt/s	9b
+	 add	#7,r4
+	add	#-3,r0
+	bra	8b
+	 add	#-1,r4
+
+case2:
+	!
+	!	GHIJ KLMN OPQR -->  ..GH IJKL MNOP QR..
+	!
+	! First, align to word boundary
+	tst	#1,r0
+	bt/s	2f
+	 add	#-1,r4
+	mov.b	@r0+,r1
+	mov.b	r1,@(r0,r4)
+	!
+2:	! Second, read a word and write a word at once
+	add	#-1,r4
+	add	#-1,r5
+	!
+3:	mov.w	@r0+,r1
+	cmp/hs	r5,r0
+	bf/s	3b
+	 mov.w	r1,@(r0,r4)
+	add	#1,r5
+	!
+	! Third, copy a byte at once, if necessary
+	cmp/eq	r5,r0
+	bt/s	9b
+	 add	#2,r4
+	mov.b	@r0,r1
+	mov.b	r1,@(r0,r4)
+	bra	9b
+	 add	#1,r0
+
+case1:
+	!
+	!	GHIJ KLMN OPQR -->  .GHI JKLM NOPQ R...
+	!
+	! First, align to long word boundary
+	mov	r0,r3
+	and	r2,r3
+	tst	r3,r3
+	bt/s	2f
+	 add	#-1,r4
+	mov	#4,r2
+	sub	r3,r2
+1:	dt	r2
+	mov.b	@r0+,r1
+	bf/s	1b
+	 mov.b	r1,@(r0,r4)
+	!
+2:	! Second, read a long word and write a long word at once
+	mov.l	@(r0,r4),r1
+	add	#-7,r5
+	add	#-4,r4
+	!
+#ifdef __LITTLE_ENDIAN__
+	shll16	r1
+	shll8	r1
+3:	mov	r1,r3		! JIHG
+	shlr16	r3
+	shlr8	r3		! xxxJ
+	mov.l	@r0+,r1		! NMLK
 	mov	r1,r2
-	add	#-1,r2
-	mov.l	r2,@(12,r14)
-	mov.l	@(20,r14),r1
+	shll8	r2		! MLKx
+	or	r2,r3		! MLKJ
+	cmp/hs	r5,r0
+	bf/s	3b
+	 mov.l	r3,@(r0,r4)
+#else
+	shlr16	r1
+	shlr8	r1
+3:	mov	r1,r3		! GHIJ
+	shll16	r3
+	shll8	r3		! Jxxx
+	mov.l	@r0+,r1		! KLMN
 	mov	r1,r2
-	add	#-1,r2
-	mov.l	r2,@(20,r14)
-	mov.l	@(12,r14),r1
-	mov	r14,r2
-	mov	r14,r3
-	add	#24,r3
-	mov.b	@r3,r2
-	mov.b	r2,@r1
-	bra	.L42
-	nop
-	.align 2
-.L43:
-.L41:
-	bra	.L24
-	nop
-	.align 2
-.L45:
-	bra	.L39
-	nop
-	.align 2
-.L40:
-.L24:
-	mov.l	@r14,r1
-	mov	r1,r0
-	bra	.L1
-	nop
-	.align 2
-.L1:
-	add	#28,r14
-	mov	r14,r15
-	lds.l	@r15+,pr
-	mov.l	@r15+,r14
-	mov.l	@r15+,r9
-	mov.l	@r15+,r8
-	rts	
-	 nop
-.L50:
-	.align 2
-.L46:
-	.long	__wordcopy_fwd_aligned
-.L47:
-	.long	__wordcopy_fwd_dest_aligned
-.L48:
-	.long	__wordcopy_bwd_aligned
-.L49:
-	.long	__wordcopy_bwd_dest_aligned
-.Lfe1:
+	shlr8	r2		! xKLM
+	or	r2,r3		! JKLM
+	cmp/hs	r5,r0
+	bf/s	3b		! while(r0<r5)
+	 mov.l	r3,@(r0,r4)
+#endif
+	add	#7,r5
+	!
+	! Third, copy a byte at once, if necessary
+	cmp/eq	r5,r0
+	bt/s	9b
+	 add	#5,r4
+	add	#-3,r0
+	bra	8b
+	 add	#-1,r4

arch/sh/lib/memset.S

-! Taken from newlib-1.8.0
+/* $Id: memset.S,v 1.1 1999/09/18 16:57:09 gniibe Exp $
+ *
+ * "memset" implementation of SuperH
+ *
+ * Copyright (C) 1999  Niibe Yutaka
+ *
+ */
+
+/*
+ *            void *memset(void *s, int c, size_t n);
+ */
 
-!
-! Fast SH memset
-!
-! by Toshiyasu Morita (tm@netcom.com)
-!
-! Entry: r4: destination pointer
-!        r5: fill value
-!        r6: byte count
-!
-! Exit:  r0-r3: trashed
-!
 #include <linux/linkage.h>
 
 ENTRY(memset)
-	mov	r4,r3	! Save return value
-	
-	mov	r6,r0	! Check explicitly for zero
-	cmp/eq	#0,r0
-	bt	L_exit
-
-	mov	#12,r0	! Check for small number of bytes
+	tst	r6,r6
+	bt/s	5f		! if n=0, do nothing
+	 add	r6,r4
+	mov	#12,r0
 	cmp/gt	r6,r0
-	bt	L_store_byte_loop
-
-	neg	r4,r0	! Align destination
-	add	#4,r0
+	bt/s	4f		! if it's too small, set a byte at once
+	 mov	r4,r0
 	and	#3,r0
-	tst	r0,r0
-	bt	L_dup_bytes
-	.balignw 4,0x0009
-L_align_loop:
-	mov.b	r5,@r4
-	add	#-1,r6
-	add	#1,r4
+	cmp/eq	#0,r0
+	bt/s	2f		! It's aligned
+	 sub	r0,r6
+1:
 	dt	r0
-	bf	L_align_loop
-
-L_dup_bytes:	
-	extu.b	r5,r5	! Duplicate bytes across longword
-	swap.b	r5,r0
-	or	r0,r5
-	swap.w	r5,r0
-	or	r0,r5
-
-	mov	r6,r2	! Calculate number of double longwords
-	shlr2	r2
-	shlr	r2
-
-	.balignw 4,0x0009
-L_store_long_loop:
-	mov.l	r5,@r4	! Store double longs to memory
-	dt	r2
-	mov.l	r5,@(4,r4)
-	add	#8,r4
-	bf	L_store_long_loop
-
+	bf/s	1b
+	 mov.b	r5,@-r4
+2:				! make VVVV
+	swap.b	r5,r0		!   V0
+	or	r0,r5		!   VV
+	swap.w	r5,r0		! VV00
+	or	r0,r5		! VVVV
+	!
+	mov	r6,r0
+	shlr2	r0
+	shlr	r0		! r0 = r6 >> 3
+3:
+	dt	r0
+	mov.l	r5,@-r4		! set 8-byte at once
+	bf/s	3b
+	 mov.l	r5,@-r4
+	!
 	mov	#7,r0
 	and	r0,r6
 	tst	r6,r6
-	bt	L_exit
-	.balignw 4,0x0009
-L_store_byte_loop:
-	mov.b	r5,@r4	! Store bytes to memory
-	add	#1,r4
+	bt	5f
+	! fill bytes
+4:
 	dt	r6
-	bf	L_store_byte_loop
-
-L_exit:
+	bf/s	4b
+	 mov.b	r5,@-r4
+5:
 	rts
-	 mov	r3,r0
+	 mov	r4,r0

arch/sh/lib/old-checksum.c

@@ -15,5 +15,3 @@ unsigned int csum_partial_copy( const char *src, char *dst, int len, int sum)
 
 	return sum;
 }
-
-

arch/sh/mm/Makefile

@@ -8,6 +8,6 @@
 # Note 2! The CFLAGS definition is now in the main makefile...
 
 O_TARGET := mm.o
-O_OBJS	 := init.o fault.o ioremap.o extable.o
+O_OBJS	 := init.o fault.o ioremap.o extable.o cache.o
 
 include $(TOPDIR)/Rules.make

arch/sh/mm/cache.c

+/* $Id: cache.c,v 1.7 1999/09/23 11:43:07 gniibe Exp $
+ *
+ *  linux/arch/sh/mm/cache.c
+ *
+ * Copyright (C) 1999  Niibe Yutaka
+ *
+ */
+
+#include <linux/init.h>
+#include <linux/mman.h>
+#include <linux/mm.h>
+#include <linux/threads.h>
+#include <asm/page.h>
+#include <asm/pgtable.h>
+#include <asm/processor.h>
+#include <asm/cache.h>
+#include <asm/io.h>
+#include <asm/uaccess.h>
+
+#if defined(__sh3__)
+#define CCR		0xffffffec	/* Address of Cache Control Register */
+#define CCR_CACHE_VAL	0x00000005	/* 8k-byte cache, P1-wb, enable */
+#define CCR_CACHE_INIT	0x0000000d	/* 8k-byte cache, CF, P1-wb, enable */
+#define CCR_CACHE_ENABLE	 1
+
+#define CACHE_IC_ADDRESS_ARRAY 0xf0000000 /* SH-3 has unified cache system */
+#define CACHE_OC_ADDRESS_ARRAY 0xf0000000
+#define CACHE_VALID	  1
+#define CACHE_UPDATED	  2
+
+/* 7709A/7729 has 16K cache (256-entry), while 7702 has only 2K(direct)
+   7702 is not supported (yet) */
+struct _cache_system_info {
+	int way_shift;
+	int entry_mask;
+	int num_entries;
+};
+
+static struct _cache_system_info cache_system_info;
+
+#define CACHE_OC_WAY_SHIFT	(cache_system_info.way_shift)
+#define CACHE_IC_WAY_SHIFT	(cache_system_info.way_shift)
+#define CACHE_OC_ENTRY_SHIFT    4
+#define CACHE_OC_ENTRY_MASK	(cache_system_info.entry_mask)
+#define CACHE_IC_ENTRY_MASK	(cache_system_info.entry_mask)
+#define CACHE_OC_NUM_ENTRIES	(cache_system_info.num_entries)
+#define CACHE_OC_NUM_WAYS	4
+#define CACHE_IC_NUM_WAYS	4
+#elif defined(__SH4__)
+#define CCR		 0xff00001c	/* Address of Cache Control Register */
+#define CCR_CACHE_VAL	 0x00000105	/* 8k+16k-byte cache,P1-wb,enable */
+#define CCR_CACHE_INIT	 0x0000090d	/* 8k+16k-byte cache,CF,P1-wb,enable */
+#define CCR_CACHE_ENABLE 0x00000101
+
+#define CACHE_IC_ADDRESS_ARRAY 0xf0000000
+#define CACHE_OC_ADDRESS_ARRAY 0xf4000000
+#define CACHE_VALID	  1
+#define CACHE_UPDATED	  2
+
+#define CACHE_OC_WAY_SHIFT       13
+#define CACHE_IC_WAY_SHIFT       13
+#define CACHE_OC_ENTRY_SHIFT      5
+#define CACHE_OC_ENTRY_MASK  0x3fe0
+#define CACHE_IC_ENTRY_MASK  0x1fe0
+#define CACHE_OC_NUM_ENTRIES	512
+#define CACHE_OC_NUM_WAYS	  1
+#define CACHE_IC_NUM_WAYS	  1
+#endif
+
+#define jump_to_p2(__dummy) \
+	asm volatile("mova	1f,%0\n\t" \
+		     "add	%1,%0\n\t" \
+		     "jmp	@r0		! Jump to P2 area\n\t" \
+		     " nop\n\t" \
+		     ".balign 4\n" \
+		     "1:" \
+		     : "=&z" (__dummy) \
+		     : "r" (0x20000000))
+
+#define back_to_p1(__dummy) \
+	asm volatile("nop;nop;nop;nop;nop;nop\n\t" \
+		     "mova	9f,%0\n\t" \
+		     "sub	%1,%0\n\t" \
+		     "jmp	@r0		! Back to P1 area\n\t" \
+		     " nop\n\t" \
+		     ".balign 4\n" \
+		     "9:" \
+		     : "=&z" (__dummy) \
+		     : "r" (0x20000000), "0" (__dummy))
+
+/* Write back caches to memory (if needed) and invalidates the caches */
+void cache_flush_area(unsigned long start, unsigned long end)
+{
+	unsigned long flags, __dummy;
+	unsigned long addr, data, v, p;
+
+	start &= ~(L1_CACHE_BYTES-1);
+	save_and_cli(flags);
+	jump_to_p2(__dummy);
+
+	for (v = start; v < end; v+=L1_CACHE_BYTES) {
+		p = __pa(v);
+		addr = CACHE_IC_ADDRESS_ARRAY |
+			(v&CACHE_IC_ENTRY_MASK) | 0x8 /* A-bit */;
+		data = (v&0xfffffc00); /* U=0, V=0 */
+		ctrl_outl(data,addr);
+#if CACHE_IC_ADDRESS_ARRAY != CACHE_OC_ADDRESS_ARRAY
+		asm volatile("ocbp	%0"
+			     : /* no output */
+			     : "m" (__m(v)));
+#endif
+	}
+	back_to_p1(__dummy);
+	restore_flags(flags);
+}
+
+/* Purge (just invalidate, no write back) the caches */
+/* This is expected to work well.. but..
+
+   On SH7708S, the write-back cache is written back on "purge".
+   (it's not expected, though).
+
+   It seems that we have no way to just purge (with no write back action)
+   the cache line. */
+void cache_purge_area(unsigned long start, unsigned long end)
+{
+	unsigned long flags, __dummy;
+	unsigned long addr, data, v, p, j;
+
+	start &= ~(L1_CACHE_BYTES-1);
+	save_and_cli(flags);
+	jump_to_p2(__dummy);
+
+	for (v = start; v < end; v+=L1_CACHE_BYTES) {
+		p = __pa(v);
+		for (j=0; j<CACHE_IC_NUM_WAYS; j++) {
+			addr = CACHE_IC_ADDRESS_ARRAY|(j<<CACHE_IC_WAY_SHIFT)|
+				(v&CACHE_IC_ENTRY_MASK);
+			data = ctrl_inl(addr);
+			if ((data & 0xfffffc00) == (p&0xfffffc00)
+			    && (data & CACHE_VALID)) {
+				data &= ~CACHE_VALID;
+				ctrl_outl(data,addr);
+				break;
+			}
+		}
+#if CACHE_IC_ADDRESS_ARRAY != CACHE_OC_ADDRESS_ARRAY
+		asm volatile("ocbi	%0"
+			     : /* no output */
+			     : "m" (__m(v)));
+#endif
+	}
+	back_to_p1(__dummy);
+	restore_flags(flags);
+}
+
+/* write back the dirty cache, but not invalidate the cache */
+void cache_wback_area(unsigned long start, unsigned long end)
+{
+	unsigned long flags, __dummy;
+	unsigned long v;
+
+	start &= ~(L1_CACHE_BYTES-1);
+	save_and_cli(flags);
+	jump_to_p2(__dummy);
+
+	for (v = start; v < end; v+=L1_CACHE_BYTES) {
+#if CACHE_IC_ADDRESS_ARRAY == CACHE_OC_ADDRESS_ARRAY
+		unsigned long addr, data, j;
+		unsigned long p = __pa(v);
+
+		for (j=0; j<CACHE_OC_NUM_WAYS; j++) {
+			addr = CACHE_OC_ADDRESS_ARRAY|(j<<CACHE_OC_WAY_SHIFT)|
+				(v&CACHE_OC_ENTRY_MASK);
+			data = ctrl_inl(addr);
+			if ((data & 0xfffffc00) == (p&0xfffffc00)
+			    && (data & CACHE_VALID)
+			    && (data & CACHE_UPDATED)) {
+				data &= ~CACHE_UPDATED;
+				ctrl_outl(data,addr);
+				break;
+			}
+		}
+#else
+		asm volatile("ocbwb	%0"
+			     : /* no output */
+			     : "m" (__m(v)));
+#endif
+	}
+	back_to_p1(__dummy);
+	restore_flags(flags);
+}
+
+/*
+ * Write back the cache.
+ *
+ * For SH-4, flush (write back) Operand Cache, as Instruction Cache
+ * doesn't have "updated" data.
+ */
+static void cache_wback_all(void)
+{
+	unsigned long flags, __dummy;
+	unsigned long addr, data, i, j;
+
+	save_and_cli(flags);
+	jump_to_p2(__dummy);
+
+	for (i=0; i<CACHE_OC_NUM_ENTRIES; i++) {
+		for (j=0; j<CACHE_OC_NUM_WAYS; j++) {
+			addr = CACHE_OC_ADDRESS_ARRAY|(j<<CACHE_OC_WAY_SHIFT)|
+				(i<<CACHE_OC_ENTRY_SHIFT);
+			data = ctrl_inl(addr);
+			if (data & CACHE_VALID) {
+				data &= ~(CACHE_VALID|CACHE_UPDATED);
+				ctrl_outl(data,addr);
+			}
+		}
+	}
+
+	back_to_p1(__dummy);
+	restore_flags(flags);
+}
+
+static void
+detect_cpu_and_cache_system(void)
+{
+#if defined(__sh3__)
+	unsigned long __dummy, addr0, addr1, data0, data1, data2, data3;
+
+	jump_to_p2(__dummy);
+	/* Check if the entry shadows or not.
+	 * When shadowed, it's 128-entry system.
+	 * Otherwise, it's 256-entry system.
+	 */
+	addr0 = CACHE_OC_ADDRESS_ARRAY + (3 << 12);
+	addr1 = CACHE_OC_ADDRESS_ARRAY + (1 << 12);
+	data0  = ctrl_inl(addr0);
+	data0  ^= 0x00000001;
+	ctrl_outl(data0,addr0);
+	data1 = ctrl_inl(addr1);
+	data2 = data1 ^ 0x00000001;
+	ctrl_outl(data2,addr1);
+	data3 = ctrl_inl(addr0);
+
+	/* Invaliate them, in case the cache has been enabled already. */
+	ctrl_outl(data0&~0x00000001,addr0);
+	ctrl_outl(data2&~0x00000001,addr1);
+	back_to_p1(__dummy);
+
+	if (data0 == data1 && data2 == data3) {	/* Shadow */
+		cache_system_info.way_shift = 11;
+		cache_system_info.entry_mask = 0x7f0;
+		cache_system_info.num_entries = 128;
+		cpu_data->type = CPU_SH7708;
+	} else {				/* 7709A or 7729  */
+		cache_system_info.way_shift = 12;
+		cache_system_info.entry_mask = 0xff0;
+		cache_system_info.num_entries = 256;
+		cpu_data->type = CPU_SH7729;
+	}
+#elif defined(__SH4__)
+	cpu_data->type = CPU_SH7750;
+#endif
+}
+
+void __init cache_init(void)
+{
+	unsigned long __dummy, ccr;
+
+	detect_cpu_and_cache_system();
+
+	ccr = ctrl_inl(CCR);
+	if (ccr == CCR_CACHE_VAL)
+		return;
+	if (ccr & CCR_CACHE_ENABLE)
+		/* Should check RA here. If RA was 1,
+		   we only need to flush the half of the caches. */
+		cache_wback_all();
+
+	jump_to_p2(__dummy);
+	ctrl_outl(CCR_CACHE_INIT, CCR);
+	back_to_p1(__dummy);
+}
+
+#if defined(__SH4__)
+/* Write back data caches, and invalidates instructiin caches */
+void flush_icache_range(unsigned long start, unsigned long end)
+{
+	unsigned long flags, __dummy;
+	unsigned long addr, data, v;
+
+	start &= ~(L1_CACHE_BYTES-1);
+	save_and_cli(flags);
+	jump_to_p2(__dummy);
+
+	for (v = start; v < end; v+=L1_CACHE_BYTES) {
+		/* Write back O Cache */
+		asm volatile("ocbwb	%0"
+			     : /* no output */
+			     : "m" (__m(v)));
+		/* Invalidate I Cache */
+		addr = CACHE_IC_ADDRESS_ARRAY |
+			(v&CACHE_IC_ENTRY_MASK) | 0x8 /* A-bit */;
+		data = (v&0xfffffc00); /* Valid=0 */
+		ctrl_outl(data,addr);
+	}
+	back_to_p1(__dummy);
+	restore_flags(flags);
+}
+
+void flush_cache_all(void)
+{
+	unsigned long flags,__dummy;
+
+	/* Write back Operand Cache */
+	cache_wback_all ();
+
+	/* Then, invalidate Instruction Cache and Operand Cache */
+	save_and_cli(flags);
+	jump_to_p2(__dummy);
+	ctrl_outl(CCR_CACHE_INIT, CCR);
+	back_to_p1(__dummy);
+	restore_flags(flags);
+}
+
+void flush_cache_mm(struct mm_struct *mm)
+{
+	/* Is there any good way? */
+	/* XXX: possibly call flush_cache_range for each vm area */
+	flush_cache_all();
+}
+
+void flush_cache_range(struct mm_struct *mm, unsigned long start,
+		       unsigned long end)
+{
+	unsigned long flags, __dummy;
+	unsigned long addr, data, v;
+
+	start &= ~(L1_CACHE_BYTES-1);
+	save_and_cli(flags);
+	jump_to_p2(__dummy);
+
+	for (v = start; v < end; v+=L1_CACHE_BYTES) {
+		addr = CACHE_IC_ADDRESS_ARRAY |
+			(v&CACHE_IC_ENTRY_MASK) | 0x8 /* A-bit */;
+		data = (v&0xfffffc00); /* Update=0, Valid=0 */
+		ctrl_outl(data,addr);
+		addr = CACHE_OC_ADDRESS_ARRAY |
+			(v&CACHE_OC_ENTRY_MASK) | 0x8 /* A-bit */;
+		ctrl_outl(data,addr);
+	}
+	back_to_p1(__dummy);
+	restore_flags(flags);
+}
+
+void flush_cache_page(struct vm_area_struct *vma, unsigned long addr)
+{
+	flush_cache_range(vma->vm_mm, addr, addr+PAGE_SIZE);
+}
+
+void flush_page_to_ram(unsigned long page)
+{				/* Page is in physical address */
+	/* XXX: for the time being... */
+	flush_cache_all();
+}
+#endif