patch-2.4.0-test12 linux/arch/parisc/kernel/setup.c
Next file: linux/arch/parisc/kernel/signal.c
Previous file: linux/arch/parisc/kernel/semaphore.c
Back to the patch index
Back to the overall index
- Lines: 615
- Date:
Wed Dec 6 11:46:39 2000
- Orig file:
v2.4.0-test11/linux/arch/parisc/kernel/setup.c
- Orig date:
Wed Dec 31 16:00:00 1969
diff -u --recursive --new-file v2.4.0-test11/linux/arch/parisc/kernel/setup.c linux/arch/parisc/kernel/setup.c
@@ -0,0 +1,614 @@
+/* $Id: setup.c,v 1.8 2000/02/02 04:42:38 prumpf Exp $
+ *
+ * Initial setup-routines for HP 9000 based hardware.
+ *
+ * Copyright (C) 1991, 1992, 1995 Linus Torvalds
+ * Modifications for PA-RISC (C) 1999 Helge Deller <helge.deller@ruhr-uni-bochum.de>
+ * Modifications copyright 1999 SuSE GmbH (Philipp Rumpf)
+ * Modifications copyright 2000 Martin K. Petersen <mkp@mkp.net>
+ * Modifications copyright 2000 Philipp Rumpf <prumpf@tux.org>
+ *
+ * Initial PA-RISC Version: 04-23-1999 by Helge Deller
+ *
+ * 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, or (at your option)
+ * any later version.
+ *
+ * This program is distributed in the hope that it will be useful,
+ * but WITHOUT ANY WARRANTY; without even the implied warranty of
+ * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+ * GNU General Public License for more details.
+ *
+ * You should have received a copy of the GNU General Public License
+ * along with this program; if not, write to the Free Software
+ * Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.
+ *
+ */
+
+#include <linux/errno.h>
+#include <linux/kernel.h>
+#include <linux/malloc.h>
+#include <linux/mm.h>
+#include <linux/ptrace.h>
+#include <linux/sched.h>
+#include <linux/stddef.h>
+#include <linux/unistd.h>
+#include <linux/user.h>
+#include <linux/tty.h>
+#include <linux/config.h>
+#include <linux/fs.h>
+#include <linux/kdev_t.h>
+#include <linux/major.h>
+#include <linux/string.h>
+#include <linux/blk.h>
+#include <linux/init.h>
+#include <linux/interrupt.h>
+#include <linux/console.h>
+#include <linux/bootmem.h>
+#include <linux/delay.h>
+#include <linux/pci.h>
+#include <linux/threads.h>
+
+#include <asm/cache.h>
+#include <asm/hardware.h> /* for register_driver() stuff */
+#include <asm/processor.h>
+#include <asm/page.h>
+#include <asm/pdc.h>
+#include <asm/led.h>
+#include <asm/real.h>
+#include <asm/system.h>
+#include <asm/machdep.h> /* for pa7300lc_init() proto */
+
+#include <asm/irq.h> /* for struct irq_region */
+#include <asm/pdcpat.h> /* for PA_VIEW PDC_PAT_CPU_GET_NUMBER etc */
+
+#include <linux/proc_fs.h>
+
+#define COMMAND_LINE_SIZE 1024
+char saved_command_line[COMMAND_LINE_SIZE];
+
+/*
+** KLUGE ALERT!
+**
+** We *really* should be using a combination of request_resource()
+** and request_region()! But request_region() requires kmalloc since
+** returns a new struct resource. And kmalloc just isn't available
+** until after mem_init() is called from start_kernel().
+**
+** FIXME: assume contiguous memory initially.
+** Additional chunks of memory might be added to sysram_resource.sibling.
+*/
+static struct resource sysrom_resource = {
+ name: "System ROM", start: 0x0f0000000UL, end: 0x0f00fffffUL,
+ flags: IORESOURCE_BUSY | IORESOURCE_MEM,
+ parent: &iomem_resource, sibling: NULL, child: NULL };
+
+static struct resource pdcdata_resource;
+
+static struct resource sysram_resource = {
+ name: "System RAM", start: 0UL, end: ~0UL /* bogus */,
+ flags: IORESOURCE_MEM,
+ parent: &iomem_resource, sibling: &sysrom_resource, child: &pdcdata_resource};
+
+extern char _text; /* start of kernel code, defined by linker */
+extern int data_start;
+extern char _edata; /* end of data, begin BSS, defined by linker */
+extern char _end; /* end of BSS, defined by linker */
+
+static struct resource data_resource = {
+ name: "kernel Data", start: virt_to_phys(&data_start), end: virt_to_phys(&_end)-1,
+ flags: IORESOURCE_BUSY | IORESOURCE_MEM,
+ parent: &sysram_resource, sibling: NULL, child: NULL};
+
+static struct resource code_resource = {
+ name: "Kernel Code", start: virt_to_phys(&_text), end: virt_to_phys(&data_start)-1,
+ flags: IORESOURCE_BUSY | IORESOURCE_MEM,
+ parent: &sysram_resource, sibling: &data_resource, child: NULL};
+
+static struct resource pdcdata_resource = {
+ name: "PDC data (Page Zero)", start: 0, end: 0x9ff,
+ flags: IORESOURCE_BUSY | IORESOURCE_MEM,
+ parent: &sysram_resource, sibling: &code_resource, child: NULL};
+
+
+
+
+struct system_cpuinfo_parisc boot_cpu_data;
+struct cpuinfo_parisc cpu_data[NR_CPUS];
+
+extern void do_inventory(void);
+extern void cache_init(void);
+extern struct hp_device * register_module(void *hpa);
+
+static int cpu_driver_callback(struct hp_device *, struct pa_iodc_driver *);
+
+static struct pa_iodc_driver cpu_drivers_for[] = {
+ {HPHW_NPROC, 0x0, 0x0, 0x0, 0, 0,
+ DRIVER_CHECK_HWTYPE,
+ "CPU", "PARISC", (void *) cpu_driver_callback},
+ {0,0,0,0,0,0,
+ 0,
+ (char *) NULL, (char *) NULL, (void *) NULL }
+};
+
+static long fallback_cpu_hpa[] = { 0xfffa0000L, 0xfffbe000L, 0x0 };
+
+
+/*
+** PARISC CPU driver - claim "device" and initialize CPU data structures.
+**
+** Consolidate per CPU initialization into (mostly) one module.
+** Monarch CPU will initialize boot_cpu_data which shouldn't
+** change once the system has booted.
+**
+** The callback *should* do per-instance initialization of
+** everything including the monarch. Some of the code that's
+** in setup.c:start_parisc() should migrate here and start_parisc()
+** should "register_driver(cpu_driver_for)" before calling
+** do_inventory().
+**
+** The goal of consolidating CPU initialization into one place is
+** to make sure all CPU's get initialized the same way.
+** It would be nice if the even the manarch through the exact same code path.
+** (up to rendevous at least).
+*/
+#undef ASSERT
+#define ASSERT(expr) \
+ if(!(expr)) { \
+ printk( "\n" __FILE__ ":%d: Assertion " #expr " failed!\n",__LINE__); \
+ panic(#expr); \
+ }
+
+static int
+cpu_driver_callback(struct hp_device *d, struct pa_iodc_driver *dri)
+{
+#ifdef __LP64__
+ extern int pdc_pat; /* arch/parisc/kernel/inventory.c */
+ static unsigned long pdc_result[32] __attribute__ ((aligned (8))) = {0,0,0,0};
+#endif
+ struct cpuinfo_parisc *p;
+
+#ifndef CONFIG_SMP
+ if (boot_cpu_data.cpu_count > 0) {
+ printk(KERN_INFO "CONFIG_SMP disabled - not claiming addional CPUs\n");
+ return(1);
+ }
+#endif
+
+ p = &cpu_data[boot_cpu_data.cpu_count];
+ boot_cpu_data.cpu_count++;
+
+/* TODO: Enable FP regs - done early in start_parisc() now */
+
+ /* initialize counters */
+ memset(p, 0, sizeof(struct cpuinfo_parisc));
+
+ p->hpa = (unsigned long) d->hpa; /* save CPU hpa */
+
+#ifdef __LP64__
+ if (pdc_pat) {
+ ulong status;
+ pdc_pat_cell_mod_maddr_block_t pa_pdc_cell;
+
+ status = pdc_pat_cell_module(& pdc_result, d->pcell_loc,
+ d->mod_index, PA_VIEW, & pa_pdc_cell);
+
+ ASSERT(PDC_RET_OK == status);
+
+ /* verify it's the same as what do_pat_inventory() found */
+ ASSERT(d->mod_info == pa_pdc_cell.mod_info);
+ ASSERT(d->pmod_loc == pa_pdc_cell.mod_location);
+ ASSERT(d->mod_path == pa_pdc_cell.mod_path);
+
+ p->txn_addr = pa_pdc_cell.mod[0]; /* id_eid for IO sapic */
+
+ /* get the cpu number */
+ status = mem_pdc_call( PDC_PAT_CPU, PDC_PAT_CPU_GET_NUMBER,
+ __pa(& pdc_result), d->hpa);
+
+ ASSERT(PDC_RET_OK == status);
+
+ p->cpuid = pdc_result[0];
+
+ } else
+#endif
+ {
+ p->txn_addr = (unsigned long) d->hpa; /* for normal parisc */
+
+ /* logical CPU ID and update global counter */
+ p->cpuid = boot_cpu_data.cpu_count - 1;
+ }
+
+ /*
+ ** itimer and ipi IRQ handlers are statically initialized in
+ ** arch/parisc/kernel/irq.c
+ */
+ p->region = irq_region[IRQ_FROM_REGION(CPU_IRQ_REGION)];
+
+ return(0);
+}
+
+
+void __xchg_called_with_bad_pointer(void)
+{
+ printk(KERN_EMERG "xchg() called with bad pointer !\n");
+}
+
+
+/* Some versions of IODC don't list the CPU, and since we don't walk
+ * the bus yet, we have to probe for processors at well known hpa
+ * addresses.
+ */
+
+void __init register_fallback_cpu (void)
+{
+ struct hp_device *d = NULL;
+ int i = 0;
+
+#ifdef CONFIG_SMP
+#error "Revisit CPU fallback addresses for SMP (Assuming bus walk hasn't been implemented)"
+#endif
+ printk ("No CPUs reported by firmware - probing...\n");
+
+ while (fallback_cpu_hpa[i]) {
+
+ d = register_module ((void *) fallback_cpu_hpa[i]);
+
+ if (d > 0) {
+ printk ("Found CPU at %lx\n", fallback_cpu_hpa[i]);
+ cpu_driver_callback (d, 0);
+ return;
+ }
+
+ i++;
+ }
+
+ panic ("No CPUs found. System halted.\n");
+ return;
+}
+
+
+/*
+ * Get CPU information and store it in the boot_cpu_data structure. */
+void __init collect_boot_cpu_data(void)
+{
+ memset(&boot_cpu_data,0,sizeof(boot_cpu_data));
+
+ boot_cpu_data.cpu_hz = 100 * PAGE0->mem_10msec; /* Hz of this PARISC */
+
+ /* get CPU-Model Information... */
+#define p ((unsigned long *)&boot_cpu_data.pdc.model)
+ if(pdc_model_info(&boot_cpu_data.pdc.model)==0)
+ printk("model %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx %08lx\n",
+ p[0], p[1], p[2], p[3], p[4], p[5], p[6], p[7], p[8]);
+#undef p
+
+ if(pdc_model_versions(&boot_cpu_data.pdc.versions, 0)==0)
+ printk("vers %08lx\n", boot_cpu_data.pdc.versions.cpuid);
+
+ if(pdc_model_cpuid(&boot_cpu_data.pdc.cpuid)==0)
+ printk("cpuid %08lx\n", boot_cpu_data.pdc.cpuid.cpuid);
+
+ printk("CPUID vers %ld rev %ld\n",
+ (boot_cpu_data.pdc.cpuid.cpuid >> 5) & 127,
+ boot_cpu_data.pdc.cpuid.cpuid & 31);
+
+ if (pdc_model_sysmodel(boot_cpu_data.pdc.sys_model_name)==0)
+ printk("model %s\n",boot_cpu_data.pdc.sys_model_name);
+
+ boot_cpu_data.model_name = parisc_getHWdescription(HPHW_NPROC,
+ boot_cpu_data.pdc.model.hversion>>4,
+ boot_cpu_data.pdc.model.sversion>>8);
+
+ boot_cpu_data.hversion = boot_cpu_data.pdc.model.hversion;
+ boot_cpu_data.sversion = boot_cpu_data.pdc.model.sversion;
+
+ boot_cpu_data.cpu_type =
+ parisc_get_cpu_type(boot_cpu_data.pdc.model.hversion);
+
+ boot_cpu_data.cpu_name = cpu_name_version[boot_cpu_data.cpu_type][0];
+ boot_cpu_data.family_name = cpu_name_version[boot_cpu_data.cpu_type][1];
+}
+
+
+#ifdef __LP64__
+#define COMMAND_GLOBAL 0xfffffffffffe0030UL
+#else
+#define COMMAND_GLOBAL 0xfffe0030
+#endif
+
+#define CMD_RESET 5 /* reset any module */
+
+/*
+** The Wright Brothers and Gecko systems have a H/W problem
+** (Lasi...'nuf said) may cause a broadcast reset to lockup
+** the system. An HVERSION dependent PDC call was developed
+** to perform a "safe", platform specific broadcast reset instead
+** of kludging up all the code.
+**
+** Older machines which do not implement PDC_BROADCAST_RESET will
+** return (with an error) and the regular broadcast reset can be
+** issued. Obviously, if the PDC does implement PDC_BROADCAST_RESET
+** the PDC call will not return (the system will be reset).
+*/
+static int
+reset_parisc(struct notifier_block *self, unsigned long command, void *ptr)
+{
+ printk("%s: %s(cmd=%lu)\n", __FILE__, __FUNCTION__, command);
+
+ switch(command) {
+ case MACH_RESTART:
+#ifdef FASTBOOT_SELFTEST_SUPPORT
+ /*
+ ** If user has modified the Firmware Selftest Bitmap,
+ ** run the tests specified in the bitmap after the
+ ** system is rebooted w/PDC_DO_RESET.
+ **
+ ** ftc_bitmap = 0x1AUL "Skip destructive memory tests"
+ **
+ ** Using "directed resets" at each processor with the MEM_TOC
+ ** vector cleared will also avoid running destructive
+ ** memory self tests. (Not implemented yet)
+ */
+ if (ftc_bitmap) {
+ mem_pdc_call( PDC_BROADCAST_RESET,
+ PDC_DO_FIRM_TEST_RESET, PDC_FIRM_TEST_MAGIC,
+ ftc_bitmap);
+ }
+#endif
+
+ /* "Normal" system reset */
+ (void) mem_pdc_call(PDC_BROADCAST_RESET, PDC_DO_RESET,
+ 0L, 0L, 0L);
+
+ /* Nope...box should reset with just CMD_RESET now */
+ gsc_writel(CMD_RESET, COMMAND_GLOBAL);
+
+ /* Wait for RESET to lay us to rest. */
+ while (1) ;
+
+ break;
+ }
+ return NOTIFY_DONE;
+}
+
+static struct notifier_block parisc_block = { reset_parisc, NULL, 0 };
+
+
+/* start_parisc() will be called from head.S to setup our new memory_start
+ and actually start our kernel !
+ Memory-Layout is:
+ - Kernel-Image (code+data+BSS)
+ - Stack (stack-size see below!, stack-setup-code is in head.S)
+ - memory_start at end of stack..
+*/
+
+unsigned long mem_start, mem_max;
+unsigned long start_pfn, max_pfn;
+extern asmlinkage void __init start_kernel(void);
+
+#define PFN_UP(x) (((x) + PAGE_SIZE-1) >> PAGE_SHIFT)
+#define PFN_DOWN(x) ((x) >> PAGE_SHIFT)
+#define PFN_PHYS(x) ((x) << PAGE_SHIFT)
+
+void __init start_parisc(unsigned arg0, unsigned arg1,
+ unsigned arg2, unsigned arg3)
+{
+ register unsigned long ccr;
+ unsigned long memory_start;
+
+ /* Clear BSS */
+
+ {
+ char *p = &_edata, *q = &_end;
+
+ while (p < q) {
+ *p++ = 0;
+ }
+ }
+
+
+ pdc_console_init();
+
+#ifdef __LP64__
+ printk("The 64-bit Kernel has started...\n");
+#else
+ printk("The 32-bit Kernel has started...\n");
+#endif
+
+ /*
+ ** Enable FP coprocessor
+ **
+ ** REVISIT: ccr should be set by PDC_COPROC results to support PA1.0.
+ ** Hardcoding works for PA1.1 processors.
+ **
+ ** REVISIT: this could be done in the "code 22" trap handler.
+ ** (frowands idea - that way we know which processes need FP
+ ** registers saved on the interrupt stack.)
+ **
+ ** NEWS FLASH: wide kernels need FP coprocessor enabled to handle
+ ** formatted printing of %lx for example (double divides I think)
+ */
+ ccr = 0xc0;
+ mtctl(ccr, 10);
+ printk("Enabled FP coprocessor\n");
+
+#ifdef __LP64__
+ printk( "If this is the LAST MESSAGE YOU SEE, you're probably using\n"
+ "32-bit millicode by mistake.\n");
+#endif
+
+ memory_start = (unsigned long) &_end;
+ memory_start = (memory_start + PAGE_SIZE) & PAGE_MASK;
+ printk("Free memory starts at: 0x%lx\n", memory_start);
+
+ /* Collect stuff passed in from the boot loader */
+ printk(KERN_WARNING "%s(0x%x,0x%x,0x%x,0x%x)\n",
+ __FUNCTION__, arg0, arg1, arg2, arg3);
+
+ /* arg0 is free-mem start, arg1 is ptr to command line */
+ if (arg0 < 64) {
+ /* called from hpux boot loader */
+ saved_command_line[0] = '\0';
+ } else {
+ strcpy(saved_command_line, (char *)__va(arg1));
+ printk("PALO command line: '%s'\nPALO initrd %x-%x\n",
+ saved_command_line, arg2, arg3);
+
+#ifdef CONFIG_BLK_DEV_INITRD
+ if (arg2 != 0) /* did palo pass us a ramdisk? */
+ {
+ initrd_start = (unsigned long)__va(arg2);
+ initrd_end = (unsigned long)__va(arg3);
+ }
+#endif
+ }
+
+ mem_start = __pa(memory_start);
+#define MAX_MEM (512*1024*1024)
+ mem_max = (PAGE0->imm_max_mem > MAX_MEM ? MAX_MEM : PAGE0->imm_max_mem);
+
+ collect_boot_cpu_data();
+
+ /* initialize the LCD/LED after boot_cpu_data is available ! */
+ led_init(); /* LCD/LED initialization */
+
+ do_inventory(); /* probe for hardware */
+ register_driver(cpu_drivers_for); /* claim all the CPUs */
+
+ if (boot_cpu_data.cpu_count == 0)
+ register_fallback_cpu();
+
+ printk("CPU(s): %d x %s at %d.%06d MHz\n",
+ boot_cpu_data.cpu_count,
+ boot_cpu_data.cpu_name,
+ boot_cpu_data.cpu_hz / 1000000,
+ boot_cpu_data.cpu_hz % 1000000 );
+
+ switch (boot_cpu_data.cpu_type) {
+ case pcx:
+ case pcxs:
+ case pcxt:
+ hppa_dma_ops = &pcx_dma_ops;
+ break;
+ case pcxl2:
+ pa7300lc_init();
+ case pcxl: /* falls through */
+ hppa_dma_ops = &pcxl_dma_ops;
+ break;
+ default:
+ break;
+ }
+
+#if 1
+ /* KLUGE! this really belongs in kernel/resource.c! */
+ iomem_resource.end = ~0UL;
+#endif
+ sysram_resource.end = mem_max - 1;
+ notifier_chain_register(&mach_notifier, &parisc_block);
+ start_kernel(); /* now back to arch-generic code... */
+}
+
+void __init setup_arch(char **cmdline_p)
+{
+ unsigned long bootmap_size;
+ unsigned long start_pfn;
+ unsigned long mem_free;
+
+ *cmdline_p = saved_command_line;
+
+ /* initialize bootmem */
+
+ start_pfn = PFN_UP(mem_start);
+ max_pfn = PFN_DOWN(mem_max);
+
+ bootmap_size = init_bootmem(start_pfn, max_pfn);
+
+ mem_start += bootmap_size;
+ mem_free = mem_max - mem_start;
+
+ /* free_bootmem handles rounding nicely */
+ printk("free_bootmem(0x%lx, 0x%lx)\n", (unsigned long)mem_start,
+ (unsigned long)mem_free);
+ free_bootmem(mem_start, mem_free);
+
+#ifdef CONFIG_BLK_DEV_INITRD
+ printk("initrd: %08x-%08x\n", (int) initrd_start, (int) initrd_end);
+
+ if (initrd_end != 0) {
+ initrd_below_start_ok = 1;
+ reserve_bootmem(__pa(initrd_start), initrd_end - initrd_start);
+ }
+#endif
+
+ cache_init();
+
+ paging_init();
+
+ if((unsigned long)&init_task_union&(INIT_TASK_SIZE - 1)) {
+ printk("init_task_union not aligned. Please recompile the kernel after changing the first line in arch/parisc/kernel/init_task.c from \n\"#define PAD 0\" to\n\"#define PAD 1\" or vice versa\n");
+ for(;;);
+ }
+
+
+#ifdef CONFIG_SERIAL_CONSOLE
+ /* nothing */
+#elif CONFIG_VT
+#if defined(CONFIG_STI_CONSOLE)
+ conswitchp = &dummy_con; /* we use take_over_console() later ! */
+#elif defined(CONFIG_IODC_CONSOLE)
+ conswitchp = &prom_con; /* it's currently really "prom_con" */
+#elif defined(CONFIG_DUMMY_CONSOLE)
+ conswitchp = &dummy_con;
+#endif
+#endif
+
+}
+
+#ifdef CONFIG_PROC_FS
+/*
+ * Get CPU information for use by procfs.
+ */
+
+int get_cpuinfo(char *buffer)
+{
+ char *p = buffer;
+ int n;
+
+ for(n=0; n<boot_cpu_data.cpu_count; n++) {
+#ifdef CONFIG_SMP
+ if (!(cpu_online_map & (1<<n)))
+ continue;
+#endif
+ p += sprintf(p, "processor\t: %d\n"
+ "cpu family\t: PA-RISC %s\n",
+ n, boot_cpu_data.family_name);
+
+ p += sprintf(p, "cpu\t\t: %s\n", boot_cpu_data.cpu_name );
+
+ /* cpu MHz */
+ p += sprintf(p, "cpu MHz\t\t: %d.%06d\n",
+ boot_cpu_data.cpu_hz / 1000000,
+ boot_cpu_data.cpu_hz % 1000000 );
+
+ p += sprintf(p, "model\t\t: %s\n"
+ "model name\t: %s\n",
+ boot_cpu_data.pdc.sys_model_name,
+ boot_cpu_data.model_name);
+
+ p += sprintf(p, "hversion\t: 0x%08x\n"
+ "sversion\t: 0x%08x\n",
+ boot_cpu_data.hversion,
+ boot_cpu_data.sversion );
+
+ p += get_cache_info(p);
+ /* print cachesize info ? */
+ p += sprintf(p, "bogomips\t: %lu.%02lu\n",
+ (loops_per_sec+2500)/500000,
+ ((loops_per_sec+2500)/5000) % 100);
+ }
+ return p - buffer;
+}
+#endif
+
FUNET's LINUX-ADM group, linux-adm@nic.funet.fi
TCL-scripts by Sam Shen (who was at: slshen@lbl.gov)