Module Name: src
Committed By: skrll
Date: Fri Apr 4 06:04:54 UTC 2014
Modified Files:
src/sys/arch/cats/cats: cats_machdep.c
src/sys/arch/cats/conf: files.cats
Log Message:
Revert previous - it wasn't complete.
To generate a diff of this commit:
cvs rdiff -u -r1.79 -r1.80 src/sys/arch/cats/cats/cats_machdep.c
cvs rdiff -u -r1.36 -r1.37 src/sys/arch/cats/conf/files.cats
Please note that diffs are not public domain; they are subject to the
copyright notices on the relevant files.
Modified files:
Index: src/sys/arch/cats/cats/cats_machdep.c
diff -u src/sys/arch/cats/cats/cats_machdep.c:1.79 src/sys/arch/cats/cats/cats_machdep.c:1.80
--- src/sys/arch/cats/cats/cats_machdep.c:1.79 Wed Apr 2 17:36:05 2014
+++ src/sys/arch/cats/cats/cats_machdep.c Fri Apr 4 06:04:54 2014
@@ -1,4 +1,4 @@
-/* $NetBSD: cats_machdep.c,v 1.79 2014/04/02 17:36:05 skrll Exp $ */
+/* $NetBSD: cats_machdep.c,v 1.80 2014/04/04 06:04:54 skrll Exp $ */
/*
* Copyright (c) 1997,1998 Mark Brinicombe.
@@ -40,7 +40,7 @@
*/
#include <sys/cdefs.h>
-__KERNEL_RCSID(0, "$NetBSD: cats_machdep.c,v 1.79 2014/04/02 17:36:05 skrll Exp $");
+__KERNEL_RCSID(0, "$NetBSD: cats_machdep.c,v 1.80 2014/04/04 06:04:54 skrll Exp $");
#include "opt_ddb.h"
#include "opt_modular.h"
@@ -74,8 +74,6 @@ __KERNEL_RCSID(0, "$NetBSD: cats_machdep
#include <arm/locore.h>
#include <arm/undefined.h>
#include <arm/arm32/machdep.h>
-
-#include <machine/pmap.h>
#include <machine/cyclone_boot.h>
#include <arm/footbridge/dc21285mem.h>
@@ -118,11 +116,30 @@ static char bootargs[MAX_BOOT_STRING + 1
char *boot_args = NULL;
char *boot_file = NULL;
+vm_offset_t physical_start;
+vm_offset_t physical_freestart;
+vm_offset_t physical_freeend;
+vm_offset_t physical_end;
+u_int free_pages;
+vm_offset_t pagetables_start;
+
+vm_offset_t msgbufphys;
#ifdef PMAP_DEBUG
extern int pmap_debug_level;
#endif
+#define KERNEL_PT_SYS 0 /* L2 table for mapping zero page */
+#define KERNEL_PT_KERNEL 1 /* L2 table for mapping kernel */
+#define KERNEL_PT_KERNEL_NUM 2
+
+/* now this could move into something more generic */
+ /* L2 tables for mapping kernel VM */
+#define KERNEL_PT_VMDATA (KERNEL_PT_KERNEL + KERNEL_PT_KERNEL_NUM)
+#define KERNEL_PT_VMDATA_NUM 4 /* 16MB kernel VM !*/
+#define NUM_KERNEL_PTS (KERNEL_PT_VMDATA + KERNEL_PT_VMDATA_NUM)
+
+pv_addr_t kernel_pt_table[NUM_KERNEL_PTS];
/* Prototypes */
@@ -173,7 +190,88 @@ extern void configure(void);
int comcnspeed = CONSPEED;
int comcnmode = CONMODE;
-static const struct pmap_devmap cats_devmap[] = {
+
+/*
+ * void cpu_reboot(int howto, char *bootstr)
+ *
+ * Reboots the system
+ *
+ * Deal with any syncing, unmounting, dumping and shutdown hooks,
+ * then reset the CPU.
+ */
+
+void
+cpu_reboot(int howto, char *bootstr)
+{
+#ifdef DIAGNOSTIC
+ /* info */
+ printf("boot: howto=%08x curlwp=%p\n", howto, curlwp);
+#endif
+
+ /*
+ * If we are still cold then hit the air brakes
+ * and crash to earth fast
+ */
+ if (cold) {
+ doshutdownhooks();
+ pmf_system_shutdown(boothowto);
+ printf("The operating system has halted.\n");
+ printf("Please press any key to reboot.\n\n");
+ cngetc();
+ printf("rebooting...\n");
+ cpu_reset();
+ /*NOTREACHED*/
+ }
+
+ /* Disable console buffering */
+/* cnpollc(1);*/
+
+ /*
+ * If RB_NOSYNC was not specified sync the discs.
+ * Note: Unless cold is set to 1 here, syslogd will die during the unmount.
+ * It looks like syslogd is getting woken up only to find that it cannot
+ * page part of the binary in as the filesystem has been unmounted.
+ */
+ if (!(howto & RB_NOSYNC))
+ bootsync();
+
+ /* Say NO to interrupts */
+ splhigh();
+
+ /* Do a dump if requested. */
+ if ((howto & (RB_DUMP | RB_HALT)) == RB_DUMP)
+ dumpsys();
+
+ /* Run any shutdown hooks */
+ doshutdownhooks();
+
+ pmf_system_shutdown(boothowto);
+
+ /* Make sure IRQ's are disabled */
+ IRQdisable;
+
+ if (howto & RB_HALT) {
+ printf("The operating system has halted.\n");
+ printf("Please press any key to reboot.\n\n");
+ cngetc();
+ }
+
+ printf("rebooting...\n");
+ cpu_reset();
+ /*NOTREACHED*/
+}
+
+/*
+ * Mapping table for core kernel memory. This memory is mapped at init
+ * time with section mappings.
+ */
+struct l1_sec_map {
+ vm_offset_t va;
+ vm_offset_t pa;
+ vm_size_t size;
+ vm_prot_t prot;
+ int cache;
+} l1_sec_table[] = {
/* Map 1MB for CSR space */
{ DC21285_ARMCSR_VBASE, DC21285_ARMCSR_BASE,
DC21285_ARMCSR_VSIZE, VM_PROT_READ|VM_PROT_WRITE,
@@ -212,14 +310,6 @@ static const struct pmap_devmap cats_dev
{ 0, 0, 0, 0, 0 }
};
-#define MAX_PHYSMEM 4
-static struct boot_physmem cats_physmem[MAX_PHYSMEM];
-int ncats_physmem = 0;
-
-extern struct bus_space footbridge_pci_io_bs_tag;
-extern struct bus_space footbridge_pci_mem_bs_tag;
-void footbridge_pci_bs_tag_init(void);
-
/*
* u_int initarm(struct ebsaboot *bootinfo)
*
@@ -238,6 +328,9 @@ u_int
initarm(void *arm_bootargs)
{
struct ebsaboot *bootinfo = arm_bootargs;
+ int loop;
+ int loop1;
+ u_int l1pagetable;
extern u_int cpu_get_control(void);
/*
@@ -265,10 +358,8 @@ initarm(void *arm_bootargs)
* Once all the memory map changes are complete we can call consinit()
* and not have to worry about things moving.
*/
- pmap_devmap_bootstrap((vaddr_t)ebsabootinfo.bt_l1, cats_devmap);
-
#ifdef FCOM_INIT_ARM
- fcomcnattach(DC21285_ARMCSR_VBASE, comcnspeed, comcnmode);
+ fcomcnattach(DC21285_ARMCSR_BASE, comcnspeed, comcnmode);
#endif
/* Talk to the user */
@@ -296,6 +387,37 @@ initarm(void *arm_bootargs)
printf("bt_vers = 0x%08x\n", ebsabootinfo.bt_vers);
printf("bt_features = 0x%08x\n", ebsabootinfo.bt_features);
#endif
+/* {
+ int loop;
+ for (loop = 0; loop < 8; ++loop) {
+ printf("%08x\n", *(((int *)bootinfo)+loop));
+ }
+ }*/
+
+ /*
+ * Ok we have the following memory map
+ *
+ * virtual address == physical address apart from the areas:
+ * 0x00000000 -> 0x000fffff which is mapped to
+ * top 1MB of physical memory
+ * 0x00100000 -> 0x0fffffff which is mapped to
+ * physical addresses 0x00100000 -> 0x0fffffff
+ * 0x10000000 -> 0x1fffffff which is mapped to
+ * physical addresses 0x00000000 -> 0x0fffffff
+ * 0x20000000 -> 0xefffffff which is mapped to
+ * physical addresses 0x20000000 -> 0xefffffff
+ * 0xf0000000 -> 0xf03fffff which is mapped to
+ * physical addresses 0x00000000 -> 0x003fffff
+ *
+ * This means that the kernel is mapped suitably for continuing
+ * execution, all I/O is mapped 1:1 virtual to physical and
+ * physical memory is accessible.
+ *
+ * The initarm() has the responsibility for creating the kernel
+ * page tables.
+ * It must also set up various memory pointers that are used
+ * by pmap etc.
+ */
/*
* Examine the boot args string for options we need to know about
@@ -303,17 +425,186 @@ initarm(void *arm_bootargs)
*/
process_kernel_args(ebsabootinfo.bt_args);
- arm32_bootmem_init(ebsabootinfo.bt_memstart,
- ebsabootinfo.bt_memend - ebsabootinfo.bt_memstart,
- ebsabootinfo.bt_memstart);
+ printf("initarm: Configuring system ...\n");
- arm32_kernel_vm_init(KERNEL_VM_BASE, ARM_VECTORS_LOW, 0, cats_devmap,
- false);
+ /*
+ * Set up the variables that define the availablilty of
+ * physical memory
+ */
+ physical_start = ebsabootinfo.bt_memstart;
+ physical_freestart = physical_start;
+ physical_end = ebsabootinfo.bt_memend;
+ physical_freeend = physical_end;
+ free_pages = (physical_end - physical_start) / PAGE_SIZE;
+
+ physmem = (physical_end - physical_start) / PAGE_SIZE;
+
+ /* Tell the user about the memory */
+ printf("physmemory: %d pages at 0x%08lx -> 0x%08lx\n", physmem,
+ physical_start, physical_end - 1);
+
+ /*
+ * Ok the kernel occupies the bottom of physical memory.
+ * The first free page after the kernel can be found in
+ * ebsabootinfo->bt_memavail
+ * We now need to allocate some fixed page tables to get the kernel
+ * going.
+ * We allocate one page directory and a number page tables and store
+ * the physical addresses in the kernel_pt_table array.
+ *
+ * Ok the next bit of physical allocation may look complex but it is
+ * simple really. I have done it like this so that no memory gets
+ * wasted during the allocation of various pages and tables that are
+ * all different sizes.
+ * The start addresses will be page aligned.
+ * We allocate the kernel page directory on the first free 16KB boundry
+ * we find.
+ * We allocate the kernel page tables on the first 4KB boundry we find.
+ * Since we allocate at least 3 L2 pagetables we know that we must
+ * encounter at least one 16KB aligned address.
+ */
- struct exec *kernexec = (struct exec *)KERNEL_TEXT_BASE;
- bool isaout = (N_GETMAGIC(kernexec[0]) == ZMAGIC);
+#ifdef VERBOSE_INIT_ARM
+ printf("Allocating page tables");
+#endif
- printf("init subsystems: patch ");
+ /* Update the address of the first free page of physical memory */
+ physical_freestart = ebsabootinfo.bt_memavail;
+ free_pages -= (physical_freestart - physical_start) / PAGE_SIZE;
+
+#ifdef VERBOSE_INIT_ARM
+ printf(" above %p\n", (void *)physical_freestart);
+#endif
+ /* Define a macro to simplify memory allocation */
+#define valloc_pages(var, np) \
+ alloc_pages((var).pv_pa, (np)); \
+ (var).pv_va = KERNEL_BASE + (var).pv_pa - physical_start;
+
+#define alloc_pages(var, np) \
+ (var) = physical_freestart; \
+ physical_freestart += ((np) * PAGE_SIZE);\
+ free_pages -= (np); \
+ memset((char *)(var), 0, ((np) * PAGE_SIZE));
+
+ loop1 = 0;
+ for (loop = 0; loop <= NUM_KERNEL_PTS; ++loop) {
+ /* Are we 16KB aligned for an L1 ? */
+ if ((physical_freestart & (L1_TABLE_SIZE - 1)) == 0
+ && kernel_l1pt.pv_pa == 0) {
+ valloc_pages(kernel_l1pt, L1_TABLE_SIZE / PAGE_SIZE);
+ } else {
+ valloc_pages(kernel_pt_table[loop1],
+ L2_TABLE_SIZE / PAGE_SIZE);
+ ++loop1;
+ }
+ }
+
+#ifdef DIAGNOSTIC
+ /* This should never be able to happen but better confirm that. */
+ if (!kernel_l1pt.pv_pa || (kernel_l1pt.pv_pa & (L1_TABLE_SIZE-1)) != 0)
+ panic("initarm: Failed to align the kernel page directory");
+#endif
+
+ /*
+ * Allocate a page for the system page mapped to V0x00000000
+ * This page will just contain the system vectors and can be
+ * shared by all processes.
+ */
+ alloc_pages(systempage.pv_pa, 1);
+
+ /* Allocate stacks for all modes */
+ valloc_pages(irqstack, IRQ_STACK_SIZE);
+ valloc_pages(abtstack, ABT_STACK_SIZE);
+ valloc_pages(undstack, UND_STACK_SIZE);
+ valloc_pages(kernelstack, UPAGES);
+
+#ifdef VERBOSE_INIT_ARM
+ printf("IRQ stack: p0x%08lx v0x%08lx\n", irqstack.pv_pa, irqstack.pv_va);
+ printf("ABT stack: p0x%08lx v0x%08lx\n", abtstack.pv_pa, abtstack.pv_va);
+ printf("UND stack: p0x%08lx v0x%08lx\n", undstack.pv_pa, undstack.pv_va);
+ printf("SVC stack: p0x%08lx v0x%08lx\n", kernelstack.pv_pa, kernelstack.pv_va);
+#endif
+
+ alloc_pages(msgbufphys, round_page(MSGBUFSIZE) / PAGE_SIZE);
+
+ /*
+ * Ok we have allocated physical pages for the primary kernel
+ * page tables
+ */
+
+#ifdef VERBOSE_INIT_ARM
+ printf("Creating L1 page table\n");
+#endif
+
+ /*
+ * Now we start consturction of the L1 page table
+ * We start by mapping the L2 page tables into the L1.
+ * This means that we can replace L1 mappings later on if necessary
+ */
+ l1pagetable = kernel_l1pt.pv_pa;
+
+ /* Map the L2 pages tables in the L1 page table */
+ pmap_link_l2pt(l1pagetable, 0x00000000,
+ &kernel_pt_table[KERNEL_PT_SYS]);
+
+ for (loop = 0; loop < KERNEL_PT_KERNEL_NUM; loop++)
+ pmap_link_l2pt(l1pagetable, KERNEL_BASE + loop * 0x00400000,
+ &kernel_pt_table[KERNEL_PT_KERNEL + loop]);
+
+ for (loop = 0; loop < KERNEL_PT_VMDATA_NUM; ++loop)
+ pmap_link_l2pt(l1pagetable, KERNEL_VM_BASE + loop * 0x00400000,
+ &kernel_pt_table[KERNEL_PT_VMDATA + loop]);
+
+ /* update the top of the kernel VM */
+ pmap_curmaxkvaddr =
+ KERNEL_VM_BASE + (KERNEL_PT_VMDATA_NUM * 0x00400000);
+
+#ifdef VERBOSE_INIT_ARM
+ printf("Mapping kernel\n");
+#endif
+
+ /* Now we fill in the L2 pagetable for the kernel static code/data */
+ struct exec *kernexec = (struct exec *)KERNEL_TEXT_BASE;
+ if (N_GETMAGIC(kernexec[0]) != ZMAGIC) {
+ /*
+ * If it's not a.out, assume ELF.
+ */
+ extern char etext[], _end[];
+ size_t textsize = (uintptr_t) etext - KERNEL_BASE;
+ size_t totalsize = (uintptr_t) _end - KERNEL_BASE;
+ u_int logical;
+
+ textsize = round_page(textsize);
+ totalsize = round_page(totalsize);
+
+ logical = pmap_map_chunk(l1pagetable, KERNEL_BASE,
+ physical_start, textsize,
+ VM_PROT_READ|VM_PROT_WRITE, PTE_CACHE);
+
+ (void) pmap_map_chunk(l1pagetable, KERNEL_BASE + logical,
+ physical_start + logical, totalsize - textsize,
+ VM_PROT_READ|VM_PROT_WRITE, PTE_CACHE);
+ } else {
+ extern int end;
+ u_int logical;
+
+ logical = pmap_map_chunk(l1pagetable, KERNEL_TEXT_BASE,
+ physical_start, kernexec->a_text,
+ VM_PROT_READ, PTE_CACHE);
+ logical += pmap_map_chunk(l1pagetable,
+ KERNEL_TEXT_BASE + logical,
+ physical_start + logical, kernexec->a_data,
+ VM_PROT_READ|VM_PROT_WRITE, PTE_CACHE);
+ logical += pmap_map_chunk(l1pagetable,
+ KERNEL_TEXT_BASE + logical,
+ physical_start + logical, kernexec->a_bss,
+ VM_PROT_READ|VM_PROT_WRITE, PTE_CACHE);
+ logical += pmap_map_chunk(l1pagetable,
+ KERNEL_TEXT_BASE + logical,
+ physical_start + logical, kernexec->a_syms + sizeof(int)
+ + *(u_int *)((int)&end + kernexec->a_syms + sizeof(int)),
+ VM_PROT_READ|VM_PROT_WRITE, PTE_CACHE);
+ }
/*
* PATCH PATCH ...
@@ -520,11 +811,9 @@ initarm(void *arm_bootargs)
printf(" ISADMA 0x%lx -> 0x%lx\n", istart,
istart + isize - 1);
#endif
- bp = &cats_physmem[ncats_physmem++];
- KASSERT(ncats_physmem < MAX_PHYSMEM);
- bp->bp_start = atop(istart);
- bp->bp_pages = atop(isize);
- bp->bp_freelist = VM_FREELIST_ISADMA;
+ uvm_page_physload(atop(istart),
+ atop(istart + isize), atop(istart),
+ atop(istart + isize), VM_FREELIST_ISADMA);
/*
* Load the pieces that come before the
@@ -535,11 +824,9 @@ initarm(void *arm_bootargs)
printf(" BEFORE 0x%lx -> 0x%lx\n",
start, istart - 1);
#endif
- bp = &cats_physmem[ncats_physmem++];
- KASSERT(ncats_physmem < MAX_PHYSMEM);
- bp->bp_start = atop(start);
- bp->bp_pages = atop(istart - start);
- bp->bp_freelist = VM_FREELIST_DEFAULT;
+ uvm_page_physload(atop(start),
+ atop(istart), atop(start),
+ atop(istart), VM_FREELIST_DEFAULT);
}
/*
@@ -551,57 +838,50 @@ initarm(void *arm_bootargs)
printf(" AFTER 0x%lx -> 0x%lx\n",
(istart + isize), end - 1);
#endif
- bp = &cats_physmem[ncats_physmem++];
- KASSERT(ncats_physmem < MAX_PHYSMEM);
- bp->bp_start = atop(istart + isize);
- bp->bp_pages = atop(end - (istart + isize));
- bp->bp_freelist = VM_FREELIST_DEFAULT;
+ uvm_page_physload(atop(istart + isize),
+ atop(end), atop(istart + isize),
+ atop(end), VM_FREELIST_DEFAULT);
}
} else {
- bp = &cats_physmem[ncats_physmem++];
- KASSERT(ncats_physmem < MAX_PHYSMEM);
- bp->bp_start = atop(start);
- bp->bp_pages = atop(end - start);
- bp->bp_freelist = VM_FREELIST_DEFAULT;
+ uvm_page_physload(atop(start), atop(end),
+ atop(start), atop(end), VM_FREELIST_DEFAULT);
}
#else /* NISADMA > 0 */
- bp = &cats_physmem[ncats_physmem++];
- KASSERT(ncats_physmem < MAX_PHYSMEM);
- bp->bp_start = atop(start);
- bp->bp_pages = atop(end - start);
- bp->bp_freelist = VM_FREELIST_DEFAULT;
+ uvm_page_physload(atop(start), atop(end),
+ atop(start), atop(end), VM_FREELIST_DEFAULT);
#endif /* NISADMA > 0 */
}
- cpu_reset_address_paddr = DC21285_ROM_BASE;
+ /* Boot strap pmap telling it where the kernel page table is */
+ printf("pmap ");
+ pmap_bootstrap(KERNEL_VM_BASE, KERNEL_VM_BASE + KERNEL_VM_SIZE);
- /* initarm_common returns the new stack pointer address */
- u_int sp;
- sp = initarm_common(KERNEL_VM_BASE, KERNEL_VM_SIZE, cats_physmem,
- ncats_physmem);
+ cpu_reset_address_paddr = DC21285_ROM_BASE;
/* Setup the IRQ system */
- printf("init subsystems: irq ");
+ printf("irq ");
footbridge_intr_init();
printf("done.\n");
-#ifdef FCOM_INIT_ARM
- fcomcndetach();
-#endif
- if (isaout) {
+#if NKSYMS || defined(DDB) || defined(MODULAR)
+#ifndef __ELF__ /* XXX */
+ {
extern int end;
- extern int esym[];
+ extern int *esym;
ksyms_addsyms_elf(*(int *)&end, ((int *)&end) + 1, esym);
}
+#endif /* __ELF__ */
+#endif
- /*
- * XXX this should only be done in main() but it useful to
- * have output earlier ...
- */
- consinit();
+#ifdef DDB
+ db_machine_init();
+ if (boothowto & RB_KDB)
+ Debugger();
+#endif
- return sp;
+ /* We return the new stack pointer address */
+ return(kernelstack.pv_va + USPACE_SVC_STACK_TOP);
}
static void
@@ -634,6 +914,10 @@ process_kernel_args(const char *loader_a
parse_mi_bootargs(boot_args);
}
+extern struct bus_space footbridge_pci_io_bs_tag;
+extern struct bus_space footbridge_pci_mem_bs_tag;
+void footbridge_pci_bs_tag_init(void);
+
void
consinit(void)
{
@@ -645,6 +929,13 @@ consinit(void)
consinit_called = 1;
+#if NISA > 0
+ /* Initialise the ISA subsystem early ... */
+ isa_footbridge_init(DC21285_PCI_IO_VBASE, DC21285_PCI_ISA_MEM_VBASE);
+#endif
+
+ footbridge_pci_bs_tag_init();
+
get_bootconf_option(boot_args, "console", BOOTOPT_TYPE_STRING,
&console);
Index: src/sys/arch/cats/conf/files.cats
diff -u src/sys/arch/cats/conf/files.cats:1.36 src/sys/arch/cats/conf/files.cats:1.37
--- src/sys/arch/cats/conf/files.cats:1.36 Wed Apr 2 17:36:05 2014
+++ src/sys/arch/cats/conf/files.cats Fri Apr 4 06:04:54 2014
@@ -1,4 +1,4 @@
-# $NetBSD: files.cats,v 1.36 2014/04/02 17:36:05 skrll Exp $
+# $NetBSD: files.cats,v 1.37 2014/04/04 06:04:54 skrll Exp $
#
# CATS-specific configuration info
#
@@ -6,10 +6,6 @@
maxpartitions 8
maxusers 2 8 64
-file arch/arm/arm32/arm32_boot.c
-file arch/arm/arm32/arm32_kvminit.c
-file arch/arm/arm32/arm32_reboot.c
-
# Maintain Interrupt statistics
defflag IRQSTATS
