linux_dsm_epyc7002/arch/sparc/kernel/setup_64.c

715 lines
17 KiB
C
Raw Normal View History

/*
* linux/arch/sparc64/kernel/setup.c
*
* Copyright (C) 1995,1996 David S. Miller (davem@caip.rutgers.edu)
* Copyright (C) 1997 Jakub Jelinek (jj@sunsite.mff.cuni.cz)
*/
#include <linux/errno.h>
#include <linux/sched.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/stddef.h>
#include <linux/unistd.h>
#include <linux/ptrace.h>
#include <asm/smp.h>
#include <linux/user.h>
#include <linux/screen_info.h>
#include <linux/delay.h>
#include <linux/fs.h>
#include <linux/seq_file.h>
#include <linux/syscalls.h>
#include <linux/kdev_t.h>
#include <linux/major.h>
#include <linux/string.h>
#include <linux/init.h>
#include <linux/inet.h>
#include <linux/console.h>
#include <linux/root_dev.h>
#include <linux/interrupt.h>
#include <linux/cpu.h>
#include <linux/initrd.h>
#include <linux/module.h>
sparc64: Fix register corruption in top-most kernel stack frame during boot. Meelis Roos reported that kernels built with gcc-4.9 do not boot, we eventually narrowed this down to only impacting machines using UltraSPARC-III and derivitive cpus. The crash happens right when the first user process is spawned: [ 54.451346] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 [ 54.451346] [ 54.571516] CPU: 1 PID: 1 Comm: init Not tainted 3.16.0-rc2-00211-gd7933ab #96 [ 54.666431] Call Trace: [ 54.698453] [0000000000762f8c] panic+0xb0/0x224 [ 54.759071] [000000000045cf68] do_exit+0x948/0x960 [ 54.823123] [000000000042cbc0] fault_in_user_windows+0xe0/0x100 [ 54.902036] [0000000000404ad0] __handle_user_windows+0x0/0x10 [ 54.978662] Press Stop-A (L1-A) to return to the boot prom [ 55.050713] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 Further investigation showed that compiling only per_cpu_patch() with an older compiler fixes the boot. Detailed analysis showed that the function is not being miscompiled by gcc-4.9, but it is using a different register allocation ordering. With the gcc-4.9 compiled function, something during the code patching causes some of the %i* input registers to get corrupted. Perhaps we have a TLB miss path into the firmware that is deep enough to cause a register window spill and subsequent restore when we get back from the TLB miss trap. Let's plug this up by doing two things: 1) Stop using the firmware stack for client interface calls into the firmware. Just use the kernel's stack. 2) As soon as we can, call into a new function "start_early_boot()" to put a one-register-window buffer between the firmware's deepest stack frame and the top-most initial kernel one. Reported-by: Meelis Roos <mroos@linux.ee> Tested-by: Meelis Roos <mroos@linux.ee> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-24 02:58:13 +07:00
#include <linux/start_kernel.h>
#include <linux/bootmem.h>
#include <asm/io.h>
#include <asm/processor.h>
#include <asm/oplib.h>
#include <asm/page.h>
#include <asm/pgtable.h>
#include <asm/idprom.h>
#include <asm/head.h>
#include <asm/starfire.h>
#include <asm/mmu_context.h>
#include <asm/timer.h>
#include <asm/sections.h>
#include <asm/setup.h>
#include <asm/mmu.h>
#include <asm/ns87303.h>
#include <asm/btext.h>
#include <asm/elf.h>
#include <asm/mdesc.h>
#include <asm/cacheflush.h>
#include <asm/dma.h>
#include <asm/irq.h>
#ifdef CONFIG_IP_PNP
#include <net/ipconfig.h>
#endif
#include "entry.h"
#include "kernel.h"
/* Used to synchronize accesses to NatSemi SUPER I/O chip configure
* operations in asm/ns87303.h
*/
DEFINE_SPINLOCK(ns87303_lock);
EXPORT_SYMBOL(ns87303_lock);
struct screen_info screen_info = {
0, 0, /* orig-x, orig-y */
0, /* unused */
0, /* orig-video-page */
0, /* orig-video-mode */
128, /* orig-video-cols */
0, 0, 0, /* unused, ega_bx, unused */
54, /* orig-video-lines */
0, /* orig-video-isVGA */
16 /* orig-video-points */
};
static void
prom_console_write(struct console *con, const char *s, unsigned int n)
{
prom_write(s, n);
}
/* Exported for mm/init.c:paging_init. */
unsigned long cmdline_memory_size = 0;
static struct console prom_early_console = {
.name = "earlyprom",
.write = prom_console_write,
.flags = CON_PRINTBUFFER | CON_BOOT | CON_ANYTIME,
.index = -1,
};
/*
* Process kernel command line switches that are specific to the
* SPARC or that require special low-level processing.
*/
static void __init process_switch(char c)
{
switch (c) {
case 'd':
case 's':
break;
case 'h':
prom_printf("boot_flags_init: Halt!\n");
prom_halt();
break;
case 'p':
prom_early_console.flags &= ~CON_BOOT;
break;
case 'P':
/* Force UltraSPARC-III P-Cache on. */
if (tlb_type != cheetah) {
printk("BOOT: Ignoring P-Cache force option.\n");
break;
}
cheetah_pcache_forced_on = 1;
add_taint(TAINT_MACHINE_CHECK, LOCKDEP_NOW_UNRELIABLE);
cheetah_enable_pcache();
break;
default:
printk("Unknown boot switch (-%c)\n", c);
break;
}
}
static void __init boot_flags_init(char *commands)
{
while (*commands) {
/* Move to the start of the next "argument". */
while (*commands == ' ')
commands++;
/* Process any command switches, otherwise skip it. */
if (*commands == '\0')
break;
if (*commands == '-') {
commands++;
while (*commands && *commands != ' ')
process_switch(*commands++);
continue;
}
sparc64: mem boot option correction The "mem" boot option can result in many unexpected consequences. This patch attempts to prevent boot hangs which have been experienced on T4-4 and T5-8. Basically the boot loader allocates vmlinuz and initrd higher in available OBP physical memory. For example, on a 2Tb T5-8 it isn't possible to boot with mem=20G. The patch utilizes memblock to avoid reserved regions and trim memory which is only free. Other improvements are possible for a multi-node machine. This is a snippet of the boot log with mem=20G on T5-8 with the patch applied: MEMBLOCK configuration: <- before memory reduction memory size = 0x1ffad6ce000 reserved size = 0xa1adf44 memory.cnt = 0xb memory[0x0] [0x00000030400000-0x00003fdde47fff], 0x3fada48000 bytes memory[0x1] [0x00003fdde4e000-0x00003fdde4ffff], 0x2000 bytes memory[0x2] [0x00080000000000-0x00083fffffffff], 0x4000000000 bytes memory[0x3] [0x00100000000000-0x00103fffffffff], 0x4000000000 bytes memory[0x4] [0x00180000000000-0x00183fffffffff], 0x4000000000 bytes memory[0x5] [0x00200000000000-0x00203fffffffff], 0x4000000000 bytes memory[0x6] [0x00280000000000-0x00283fffffffff], 0x4000000000 bytes memory[0x7] [0x00300000000000-0x00303fffffffff], 0x4000000000 bytes memory[0x8] [0x00380000000000-0x00383fffc71fff], 0x3fffc72000 bytes memory[0x9] [0x00383fffc92000-0x00383fffca1fff], 0x10000 bytes memory[0xa] [0x00383fffcb4000-0x00383fffcb5fff], 0x2000 bytes reserved.cnt = 0x2 reserved[0x0] [0x00380000000000-0x0038000117e7f8], 0x117e7f9 bytes reserved[0x1] [0x00380004000000-0x0038000d02f74a], 0x902f74b bytes ... MEMBLOCK configuration: <- after reduction of memory memory size = 0x50a1adf44 reserved size = 0xa1adf44 memory.cnt = 0x4 memory[0x0] [0x00380000000000-0x0038000117e7f8], 0x117e7f9 bytes memory[0x1] [0x00380004000000-0x0038050d01d74a], 0x50901d74b bytes memory[0x2] [0x00383fffc92000-0x00383fffca1fff], 0x10000 bytes memory[0x3] [0x00383fffcb4000-0x00383fffcb5fff], 0x2000 bytes reserved.cnt = 0x2 reserved[0x0] [0x00380000000000-0x0038000117e7f8], 0x117e7f9 bytes reserved[0x1] [0x00380004000000-0x0038000d02f74a], 0x902f74b bytes ... Early memory node ranges node 7: [mem 0x380000000000-0x38000117dfff] node 7: [mem 0x380004000000-0x380f0d01bfff] node 7: [mem 0x383fffc92000-0x383fffca1fff] node 7: [mem 0x383fffcb4000-0x383fffcb5fff] Could not find start_pfn for node 0 Could not find start_pfn for node 1 Could not find start_pfn for node 2 Could not find start_pfn for node 3 Could not find start_pfn for node 4 Could not find start_pfn for node 5 Could not find start_pfn for node 6 . The patch was tested on T4-1, T5-8 and Jalap?no. Cc: sparclinux@vger.kernel.org Signed-off-by: Bob Picco <bob.picco@oracle.com> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-09-16 20:29:54 +07:00
if (!strncmp(commands, "mem=", 4))
cmdline_memory_size = memparse(commands + 4, &commands);
while (*commands && *commands != ' ')
commands++;
}
}
extern unsigned short root_flags;
extern unsigned short root_dev;
extern unsigned short ram_flags;
#define RAMDISK_IMAGE_START_MASK 0x07FF
#define RAMDISK_PROMPT_FLAG 0x8000
#define RAMDISK_LOAD_FLAG 0x4000
extern int root_mountflags;
char reboot_command[COMMAND_LINE_SIZE];
static struct pt_regs fake_swapper_regs = { { 0, }, 0, 0, 0, 0 };
sparc64: Fix register corruption in top-most kernel stack frame during boot. Meelis Roos reported that kernels built with gcc-4.9 do not boot, we eventually narrowed this down to only impacting machines using UltraSPARC-III and derivitive cpus. The crash happens right when the first user process is spawned: [ 54.451346] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 [ 54.451346] [ 54.571516] CPU: 1 PID: 1 Comm: init Not tainted 3.16.0-rc2-00211-gd7933ab #96 [ 54.666431] Call Trace: [ 54.698453] [0000000000762f8c] panic+0xb0/0x224 [ 54.759071] [000000000045cf68] do_exit+0x948/0x960 [ 54.823123] [000000000042cbc0] fault_in_user_windows+0xe0/0x100 [ 54.902036] [0000000000404ad0] __handle_user_windows+0x0/0x10 [ 54.978662] Press Stop-A (L1-A) to return to the boot prom [ 55.050713] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 Further investigation showed that compiling only per_cpu_patch() with an older compiler fixes the boot. Detailed analysis showed that the function is not being miscompiled by gcc-4.9, but it is using a different register allocation ordering. With the gcc-4.9 compiled function, something during the code patching causes some of the %i* input registers to get corrupted. Perhaps we have a TLB miss path into the firmware that is deep enough to cause a register window spill and subsequent restore when we get back from the TLB miss trap. Let's plug this up by doing two things: 1) Stop using the firmware stack for client interface calls into the firmware. Just use the kernel's stack. 2) As soon as we can, call into a new function "start_early_boot()" to put a one-register-window buffer between the firmware's deepest stack frame and the top-most initial kernel one. Reported-by: Meelis Roos <mroos@linux.ee> Tested-by: Meelis Roos <mroos@linux.ee> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-24 02:58:13 +07:00
static void __init per_cpu_patch(void)
{
struct cpuid_patch_entry *p;
unsigned long ver;
int is_jbus;
if (tlb_type == spitfire && !this_is_starfire)
return;
is_jbus = 0;
if (tlb_type != hypervisor) {
__asm__ ("rdpr %%ver, %0" : "=r" (ver));
is_jbus = ((ver >> 32UL) == __JALAPENO_ID ||
(ver >> 32UL) == __SERRANO_ID);
}
p = &__cpuid_patch;
while (p < &__cpuid_patch_end) {
unsigned long addr = p->addr;
unsigned int *insns;
switch (tlb_type) {
case spitfire:
insns = &p->starfire[0];
break;
case cheetah:
case cheetah_plus:
if (is_jbus)
insns = &p->cheetah_jbus[0];
else
insns = &p->cheetah_safari[0];
break;
case hypervisor:
insns = &p->sun4v[0];
break;
default:
prom_printf("Unknown cpu type, halting.\n");
prom_halt();
}
*(unsigned int *) (addr + 0) = insns[0];
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 0));
*(unsigned int *) (addr + 4) = insns[1];
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 4));
*(unsigned int *) (addr + 8) = insns[2];
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 8));
*(unsigned int *) (addr + 12) = insns[3];
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 12));
p++;
}
}
void sun4v_patch_1insn_range(struct sun4v_1insn_patch_entry *start,
struct sun4v_1insn_patch_entry *end)
{
while (start < end) {
unsigned long addr = start->addr;
*(unsigned int *) (addr + 0) = start->insn;
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 0));
start++;
}
}
void sun4v_patch_2insn_range(struct sun4v_2insn_patch_entry *start,
struct sun4v_2insn_patch_entry *end)
{
while (start < end) {
unsigned long addr = start->addr;
*(unsigned int *) (addr + 0) = start->insns[0];
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 0));
*(unsigned int *) (addr + 4) = start->insns[1];
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 4));
start++;
}
}
void sun_m7_patch_2insn_range(struct sun4v_2insn_patch_entry *start,
struct sun4v_2insn_patch_entry *end)
{
while (start < end) {
unsigned long addr = start->addr;
*(unsigned int *) (addr + 0) = start->insns[0];
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 0));
*(unsigned int *) (addr + 4) = start->insns[1];
wmb();
__asm__ __volatile__("flush %0" : : "r" (addr + 4));
start++;
}
}
sparc64: Fix register corruption in top-most kernel stack frame during boot. Meelis Roos reported that kernels built with gcc-4.9 do not boot, we eventually narrowed this down to only impacting machines using UltraSPARC-III and derivitive cpus. The crash happens right when the first user process is spawned: [ 54.451346] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 [ 54.451346] [ 54.571516] CPU: 1 PID: 1 Comm: init Not tainted 3.16.0-rc2-00211-gd7933ab #96 [ 54.666431] Call Trace: [ 54.698453] [0000000000762f8c] panic+0xb0/0x224 [ 54.759071] [000000000045cf68] do_exit+0x948/0x960 [ 54.823123] [000000000042cbc0] fault_in_user_windows+0xe0/0x100 [ 54.902036] [0000000000404ad0] __handle_user_windows+0x0/0x10 [ 54.978662] Press Stop-A (L1-A) to return to the boot prom [ 55.050713] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 Further investigation showed that compiling only per_cpu_patch() with an older compiler fixes the boot. Detailed analysis showed that the function is not being miscompiled by gcc-4.9, but it is using a different register allocation ordering. With the gcc-4.9 compiled function, something during the code patching causes some of the %i* input registers to get corrupted. Perhaps we have a TLB miss path into the firmware that is deep enough to cause a register window spill and subsequent restore when we get back from the TLB miss trap. Let's plug this up by doing two things: 1) Stop using the firmware stack for client interface calls into the firmware. Just use the kernel's stack. 2) As soon as we can, call into a new function "start_early_boot()" to put a one-register-window buffer between the firmware's deepest stack frame and the top-most initial kernel one. Reported-by: Meelis Roos <mroos@linux.ee> Tested-by: Meelis Roos <mroos@linux.ee> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-24 02:58:13 +07:00
static void __init sun4v_patch(void)
{
extern void sun4v_hvapi_init(void);
if (tlb_type != hypervisor)
return;
sun4v_patch_1insn_range(&__sun4v_1insn_patch,
&__sun4v_1insn_patch_end);
sun4v_patch_2insn_range(&__sun4v_2insn_patch,
&__sun4v_2insn_patch_end);
switch (sun4v_chip_type) {
case SUN4V_CHIP_SPARC_M7:
case SUN4V_CHIP_SPARC_M8:
case SUN4V_CHIP_SPARC_SN:
sun_m7_patch_2insn_range(&__sun_m7_2insn_patch,
&__sun_m7_2insn_patch_end);
break;
default:
break;
}
if (sun4v_chip_type != SUN4V_CHIP_NIAGARA1) {
sun4v_patch_1insn_range(&__fast_win_ctrl_1insn_patch,
&__fast_win_ctrl_1insn_patch_end);
}
sun4v_hvapi_init();
}
static void __init popc_patch(void)
{
struct popc_3insn_patch_entry *p3;
struct popc_6insn_patch_entry *p6;
p3 = &__popc_3insn_patch;
while (p3 < &__popc_3insn_patch_end) {
unsigned long i, addr = p3->addr;
for (i = 0; i < 3; i++) {
*(unsigned int *) (addr + (i * 4)) = p3->insns[i];
wmb();
__asm__ __volatile__("flush %0"
: : "r" (addr + (i * 4)));
}
p3++;
}
p6 = &__popc_6insn_patch;
while (p6 < &__popc_6insn_patch_end) {
unsigned long i, addr = p6->addr;
for (i = 0; i < 6; i++) {
*(unsigned int *) (addr + (i * 4)) = p6->insns[i];
wmb();
__asm__ __volatile__("flush %0"
: : "r" (addr + (i * 4)));
}
p6++;
}
}
static void __init pause_patch(void)
{
struct pause_patch_entry *p;
p = &__pause_3insn_patch;
while (p < &__pause_3insn_patch_end) {
unsigned long i, addr = p->addr;
for (i = 0; i < 3; i++) {
*(unsigned int *) (addr + (i * 4)) = p->insns[i];
wmb();
__asm__ __volatile__("flush %0"
: : "r" (addr + (i * 4)));
}
p++;
}
}
sparc64: Fix register corruption in top-most kernel stack frame during boot. Meelis Roos reported that kernels built with gcc-4.9 do not boot, we eventually narrowed this down to only impacting machines using UltraSPARC-III and derivitive cpus. The crash happens right when the first user process is spawned: [ 54.451346] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 [ 54.451346] [ 54.571516] CPU: 1 PID: 1 Comm: init Not tainted 3.16.0-rc2-00211-gd7933ab #96 [ 54.666431] Call Trace: [ 54.698453] [0000000000762f8c] panic+0xb0/0x224 [ 54.759071] [000000000045cf68] do_exit+0x948/0x960 [ 54.823123] [000000000042cbc0] fault_in_user_windows+0xe0/0x100 [ 54.902036] [0000000000404ad0] __handle_user_windows+0x0/0x10 [ 54.978662] Press Stop-A (L1-A) to return to the boot prom [ 55.050713] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 Further investigation showed that compiling only per_cpu_patch() with an older compiler fixes the boot. Detailed analysis showed that the function is not being miscompiled by gcc-4.9, but it is using a different register allocation ordering. With the gcc-4.9 compiled function, something during the code patching causes some of the %i* input registers to get corrupted. Perhaps we have a TLB miss path into the firmware that is deep enough to cause a register window spill and subsequent restore when we get back from the TLB miss trap. Let's plug this up by doing two things: 1) Stop using the firmware stack for client interface calls into the firmware. Just use the kernel's stack. 2) As soon as we can, call into a new function "start_early_boot()" to put a one-register-window buffer between the firmware's deepest stack frame and the top-most initial kernel one. Reported-by: Meelis Roos <mroos@linux.ee> Tested-by: Meelis Roos <mroos@linux.ee> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-24 02:58:13 +07:00
void __init start_early_boot(void)
{
sparc64: Fix register corruption in top-most kernel stack frame during boot. Meelis Roos reported that kernels built with gcc-4.9 do not boot, we eventually narrowed this down to only impacting machines using UltraSPARC-III and derivitive cpus. The crash happens right when the first user process is spawned: [ 54.451346] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 [ 54.451346] [ 54.571516] CPU: 1 PID: 1 Comm: init Not tainted 3.16.0-rc2-00211-gd7933ab #96 [ 54.666431] Call Trace: [ 54.698453] [0000000000762f8c] panic+0xb0/0x224 [ 54.759071] [000000000045cf68] do_exit+0x948/0x960 [ 54.823123] [000000000042cbc0] fault_in_user_windows+0xe0/0x100 [ 54.902036] [0000000000404ad0] __handle_user_windows+0x0/0x10 [ 54.978662] Press Stop-A (L1-A) to return to the boot prom [ 55.050713] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 Further investigation showed that compiling only per_cpu_patch() with an older compiler fixes the boot. Detailed analysis showed that the function is not being miscompiled by gcc-4.9, but it is using a different register allocation ordering. With the gcc-4.9 compiled function, something during the code patching causes some of the %i* input registers to get corrupted. Perhaps we have a TLB miss path into the firmware that is deep enough to cause a register window spill and subsequent restore when we get back from the TLB miss trap. Let's plug this up by doing two things: 1) Stop using the firmware stack for client interface calls into the firmware. Just use the kernel's stack. 2) As soon as we can, call into a new function "start_early_boot()" to put a one-register-window buffer between the firmware's deepest stack frame and the top-most initial kernel one. Reported-by: Meelis Roos <mroos@linux.ee> Tested-by: Meelis Roos <mroos@linux.ee> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-24 02:58:13 +07:00
int cpu;
check_if_starfire();
per_cpu_patch();
sun4v_patch();
smp_init_cpu_poke();
sparc64: Fix register corruption in top-most kernel stack frame during boot. Meelis Roos reported that kernels built with gcc-4.9 do not boot, we eventually narrowed this down to only impacting machines using UltraSPARC-III and derivitive cpus. The crash happens right when the first user process is spawned: [ 54.451346] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 [ 54.451346] [ 54.571516] CPU: 1 PID: 1 Comm: init Not tainted 3.16.0-rc2-00211-gd7933ab #96 [ 54.666431] Call Trace: [ 54.698453] [0000000000762f8c] panic+0xb0/0x224 [ 54.759071] [000000000045cf68] do_exit+0x948/0x960 [ 54.823123] [000000000042cbc0] fault_in_user_windows+0xe0/0x100 [ 54.902036] [0000000000404ad0] __handle_user_windows+0x0/0x10 [ 54.978662] Press Stop-A (L1-A) to return to the boot prom [ 55.050713] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 Further investigation showed that compiling only per_cpu_patch() with an older compiler fixes the boot. Detailed analysis showed that the function is not being miscompiled by gcc-4.9, but it is using a different register allocation ordering. With the gcc-4.9 compiled function, something during the code patching causes some of the %i* input registers to get corrupted. Perhaps we have a TLB miss path into the firmware that is deep enough to cause a register window spill and subsequent restore when we get back from the TLB miss trap. Let's plug this up by doing two things: 1) Stop using the firmware stack for client interface calls into the firmware. Just use the kernel's stack. 2) As soon as we can, call into a new function "start_early_boot()" to put a one-register-window buffer between the firmware's deepest stack frame and the top-most initial kernel one. Reported-by: Meelis Roos <mroos@linux.ee> Tested-by: Meelis Roos <mroos@linux.ee> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-24 02:58:13 +07:00
cpu = hard_smp_processor_id();
if (cpu >= NR_CPUS) {
prom_printf("Serious problem, boot cpu id (%d) >= NR_CPUS (%d)\n",
cpu, NR_CPUS);
prom_halt();
}
current_thread_info()->cpu = cpu;
time_init_early();
sparc64: Fix register corruption in top-most kernel stack frame during boot. Meelis Roos reported that kernels built with gcc-4.9 do not boot, we eventually narrowed this down to only impacting machines using UltraSPARC-III and derivitive cpus. The crash happens right when the first user process is spawned: [ 54.451346] Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 [ 54.451346] [ 54.571516] CPU: 1 PID: 1 Comm: init Not tainted 3.16.0-rc2-00211-gd7933ab #96 [ 54.666431] Call Trace: [ 54.698453] [0000000000762f8c] panic+0xb0/0x224 [ 54.759071] [000000000045cf68] do_exit+0x948/0x960 [ 54.823123] [000000000042cbc0] fault_in_user_windows+0xe0/0x100 [ 54.902036] [0000000000404ad0] __handle_user_windows+0x0/0x10 [ 54.978662] Press Stop-A (L1-A) to return to the boot prom [ 55.050713] ---[ end Kernel panic - not syncing: Attempted to kill init! exitcode=0x00000004 Further investigation showed that compiling only per_cpu_patch() with an older compiler fixes the boot. Detailed analysis showed that the function is not being miscompiled by gcc-4.9, but it is using a different register allocation ordering. With the gcc-4.9 compiled function, something during the code patching causes some of the %i* input registers to get corrupted. Perhaps we have a TLB miss path into the firmware that is deep enough to cause a register window spill and subsequent restore when we get back from the TLB miss trap. Let's plug this up by doing two things: 1) Stop using the firmware stack for client interface calls into the firmware. Just use the kernel's stack. 2) As soon as we can, call into a new function "start_early_boot()" to put a one-register-window buffer between the firmware's deepest stack frame and the top-most initial kernel one. Reported-by: Meelis Roos <mroos@linux.ee> Tested-by: Meelis Roos <mroos@linux.ee> Signed-off-by: David S. Miller <davem@davemloft.net>
2014-10-24 02:58:13 +07:00
prom_init_report();
start_kernel();
}
/* On Ultra, we support all of the v8 capabilities. */
unsigned long sparc64_elf_hwcap = (HWCAP_SPARC_FLUSH | HWCAP_SPARC_STBAR |
HWCAP_SPARC_SWAP | HWCAP_SPARC_MULDIV |
HWCAP_SPARC_V9);
EXPORT_SYMBOL(sparc64_elf_hwcap);
static const char *hwcaps[] = {
"flush", "stbar", "swap", "muldiv", "v9",
"ultra3", "blkinit", "n2",
/* These strings are as they appear in the machine description
* 'hwcap-list' property for cpu nodes.
*/
"mul32", "div32", "fsmuld", "v8plus", "popc", "vis", "vis2",
"ASIBlkInit", "fmaf", "vis3", "hpc", "random", "trans", "fjfmau",
"ima", "cspare", "pause", "cbcond", NULL /*reserved for crypto */,
"adp",
};
static const char *crypto_hwcaps[] = {
"aes", "des", "kasumi", "camellia", "md5", "sha1", "sha256",
"sha512", "mpmul", "montmul", "montsqr", "crc32c",
};
void cpucap_info(struct seq_file *m)
{
unsigned long caps = sparc64_elf_hwcap;
int i, printed = 0;
seq_puts(m, "cpucaps\t\t: ");
for (i = 0; i < ARRAY_SIZE(hwcaps); i++) {
unsigned long bit = 1UL << i;
if (hwcaps[i] && (caps & bit)) {
seq_printf(m, "%s%s",
printed ? "," : "", hwcaps[i]);
printed++;
}
}
if (caps & HWCAP_SPARC_CRYPTO) {
unsigned long cfr;
__asm__ __volatile__("rd %%asr26, %0" : "=r" (cfr));
for (i = 0; i < ARRAY_SIZE(crypto_hwcaps); i++) {
unsigned long bit = 1UL << i;
if (cfr & bit) {
seq_printf(m, "%s%s",
printed ? "," : "", crypto_hwcaps[i]);
printed++;
}
}
}
seq_putc(m, '\n');
}
static void __init report_one_hwcap(int *printed, const char *name)
{
if ((*printed) == 0)
printk(KERN_INFO "CPU CAPS: [");
printk(KERN_CONT "%s%s",
(*printed) ? "," : "", name);
if (++(*printed) == 8) {
printk(KERN_CONT "]\n");
*printed = 0;
}
}
static void __init report_crypto_hwcaps(int *printed)
{
unsigned long cfr;
int i;
__asm__ __volatile__("rd %%asr26, %0" : "=r" (cfr));
for (i = 0; i < ARRAY_SIZE(crypto_hwcaps); i++) {
unsigned long bit = 1UL << i;
if (cfr & bit)
report_one_hwcap(printed, crypto_hwcaps[i]);
}
}
static void __init report_hwcaps(unsigned long caps)
{
int i, printed = 0;
for (i = 0; i < ARRAY_SIZE(hwcaps); i++) {
unsigned long bit = 1UL << i;
if (hwcaps[i] && (caps & bit))
report_one_hwcap(&printed, hwcaps[i]);
}
if (caps & HWCAP_SPARC_CRYPTO)
report_crypto_hwcaps(&printed);
if (printed != 0)
printk(KERN_CONT "]\n");
}
static unsigned long __init mdesc_cpu_hwcap_list(void)
{
struct mdesc_handle *hp;
unsigned long caps = 0;
const char *prop;
int len;
u64 pn;
hp = mdesc_grab();
if (!hp)
return 0;
pn = mdesc_node_by_name(hp, MDESC_NODE_NULL, "cpu");
if (pn == MDESC_NODE_NULL)
goto out;
prop = mdesc_get_property(hp, pn, "hwcap-list", &len);
if (!prop)
goto out;
while (len) {
int i, plen;
for (i = 0; i < ARRAY_SIZE(hwcaps); i++) {
unsigned long bit = 1UL << i;
if (hwcaps[i] && !strcmp(prop, hwcaps[i])) {
caps |= bit;
break;
}
}
for (i = 0; i < ARRAY_SIZE(crypto_hwcaps); i++) {
if (!strcmp(prop, crypto_hwcaps[i]))
caps |= HWCAP_SPARC_CRYPTO;
}
plen = strlen(prop) + 1;
prop += plen;
len -= plen;
}
out:
mdesc_release(hp);
return caps;
}
/* This yields a mask that user programs can use to figure out what
* instruction set this cpu supports.
*/
static void __init init_sparc64_elf_hwcap(void)
{
unsigned long cap = sparc64_elf_hwcap;
unsigned long mdesc_caps;
if (tlb_type == cheetah || tlb_type == cheetah_plus)
cap |= HWCAP_SPARC_ULTRA3;
else if (tlb_type == hypervisor) {
if (sun4v_chip_type == SUN4V_CHIP_NIAGARA1 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA2 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA3 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA4 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA5 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M6 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M7 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M8 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_SN ||
sun4v_chip_type == SUN4V_CHIP_SPARC64X)
cap |= HWCAP_SPARC_BLKINIT;
if (sun4v_chip_type == SUN4V_CHIP_NIAGARA2 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA3 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA4 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA5 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M6 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M7 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M8 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_SN ||
sun4v_chip_type == SUN4V_CHIP_SPARC64X)
cap |= HWCAP_SPARC_N2;
}
cap |= (AV_SPARC_MUL32 | AV_SPARC_DIV32 | AV_SPARC_V8PLUS);
mdesc_caps = mdesc_cpu_hwcap_list();
if (!mdesc_caps) {
if (tlb_type == spitfire)
cap |= AV_SPARC_VIS;
if (tlb_type == cheetah || tlb_type == cheetah_plus)
cap |= AV_SPARC_VIS | AV_SPARC_VIS2;
if (tlb_type == cheetah_plus) {
unsigned long impl, ver;
__asm__ __volatile__("rdpr %%ver, %0" : "=r" (ver));
impl = ((ver >> 32) & 0xffff);
if (impl == PANTHER_IMPL)
cap |= AV_SPARC_POPC;
}
if (tlb_type == hypervisor) {
if (sun4v_chip_type == SUN4V_CHIP_NIAGARA1)
cap |= AV_SPARC_ASI_BLK_INIT;
if (sun4v_chip_type == SUN4V_CHIP_NIAGARA2 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA3 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA4 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA5 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M6 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M7 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M8 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_SN ||
sun4v_chip_type == SUN4V_CHIP_SPARC64X)
cap |= (AV_SPARC_VIS | AV_SPARC_VIS2 |
AV_SPARC_ASI_BLK_INIT |
AV_SPARC_POPC);
if (sun4v_chip_type == SUN4V_CHIP_NIAGARA3 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA4 ||
sun4v_chip_type == SUN4V_CHIP_NIAGARA5 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M6 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M7 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_M8 ||
sun4v_chip_type == SUN4V_CHIP_SPARC_SN ||
sun4v_chip_type == SUN4V_CHIP_SPARC64X)
cap |= (AV_SPARC_VIS3 | AV_SPARC_HPC |
AV_SPARC_FMAF);
}
}
sparc64_elf_hwcap = cap | mdesc_caps;
report_hwcaps(sparc64_elf_hwcap);
if (sparc64_elf_hwcap & AV_SPARC_POPC)
popc_patch();
if (sparc64_elf_hwcap & AV_SPARC_PAUSE)
pause_patch();
}
void __init alloc_irqstack_bootmem(void)
{
unsigned int i, node;
for_each_possible_cpu(i) {
node = cpu_to_node(i);
softirq_stack[i] = __alloc_bootmem_node(NODE_DATA(node),
THREAD_SIZE,
THREAD_SIZE, 0);
hardirq_stack[i] = __alloc_bootmem_node(NODE_DATA(node),
THREAD_SIZE,
THREAD_SIZE, 0);
}
}
void __init setup_arch(char **cmdline_p)
{
/* Initialize PROM console and command line. */
*cmdline_p = prom_getbootargs();
strlcpy(boot_command_line, *cmdline_p, COMMAND_LINE_SIZE);
parse_early_param();
boot_flags_init(*cmdline_p);
#ifdef CONFIG_EARLYFB
if (btext_find_display())
#endif
register_console(&prom_early_console);
if (tlb_type == hypervisor)
printk("ARCH: SUN4V\n");
else
printk("ARCH: SUN4U\n");
#ifdef CONFIG_DUMMY_CONSOLE
conswitchp = &dummy_con;
#endif
idprom_init();
if (!root_flags)
root_mountflags &= ~MS_RDONLY;
ROOT_DEV = old_decode_dev(root_dev);
#ifdef CONFIG_BLK_DEV_RAM
rd_image_start = ram_flags & RAMDISK_IMAGE_START_MASK;
rd_prompt = ((ram_flags & RAMDISK_PROMPT_FLAG) != 0);
rd_doload = ((ram_flags & RAMDISK_LOAD_FLAG) != 0);
#endif
task_thread_info(&init_task)->kregs = &fake_swapper_regs;
#ifdef CONFIG_IP_PNP
if (!ic_set_manually) {
phandle chosen = prom_finddevice("/chosen");
u32 cl, sv, gw;
cl = prom_getintdefault (chosen, "client-ip", 0);
sv = prom_getintdefault (chosen, "server-ip", 0);
gw = prom_getintdefault (chosen, "gateway-ip", 0);
if (cl && sv) {
ic_myaddr = cl;
ic_servaddr = sv;
if (gw)
ic_gateway = gw;
#if defined(CONFIG_IP_PNP_BOOTP) || defined(CONFIG_IP_PNP_RARP)
ic_proto_enabled = 0;
#endif
}
}
#endif
[SPARC64]: Elminate all usage of hard-coded trap globals. UltraSPARC has special sets of global registers which are switched to for certain trap types. There is one set for MMU related traps, one set of Interrupt Vector processing, and another set (called the Alternate globals) for all other trap types. For what seems like forever we've hard coded the values in some of these trap registers. Some examples include: 1) Interrupt Vector global %g6 holds current processors interrupt work struct where received interrupts are managed for IRQ handler dispatch. 2) MMU global %g7 holds the base of the page tables of the currently active address space. 3) Alternate global %g6 held the current_thread_info() value. Such hardcoding has resulted in some serious issues in many areas. There are some code sequences where having another register available would help clean up the implementation. Taking traps such as cross-calls from the OBP firmware requires some trick code sequences wherein we have to save away and restore all of the special sets of global registers when we enter/exit OBP. We were also using the IMMU TSB register on SMP to hold the per-cpu area base address, which doesn't work any longer now that we actually use the TSB facility of the cpu. The implementation is pretty straight forward. One tricky bit is getting the current processor ID as that is different on different cpu variants. We use a stub with a fancy calling convention which we patch at boot time. The calling convention is that the stub is branched to and the (PC - 4) to return to is in register %g1. The cpu number is left in %g6. This stub can be invoked by using the __GET_CPUID macro. We use an array of per-cpu trap state to store the current thread and physical address of the current address space's page tables. The TRAP_LOAD_THREAD_REG loads %g6 with the current thread from this table, it uses __GET_CPUID and also clobbers %g1. TRAP_LOAD_IRQ_WORK is used by the interrupt vector processing to load the current processor's IRQ software state into %g6. It also uses __GET_CPUID and clobbers %g1. Finally, TRAP_LOAD_PGD_PHYS loads the physical address base of the current address space's page tables into %g7, it clobbers %g1 and uses __GET_CPUID. Many refinements are possible, as well as some tuning, with this stuff in place. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-02-27 14:24:22 +07:00
/* Get boot processor trap_block[] setup. */
[SPARC64]: Get SUN4V SMP working. The sibling cpu bringup is extremely fragile. We can only perform the most basic calls until we take over the trap table from the firmware/hypervisor on the new cpu. This means no accesses to %g4, %g5, %g6 since those can't be TLB translated without our trap handlers. In order to achieve this: 1) Change sun4v_init_mondo_queues() so that it can operate in several modes. It can allocate the queues, or install them in the current processor, or both. The boot cpu does both in it's call early on. Later, the boot cpu allocates the sibling cpu queue, starts the sibling cpu, then the sibling cpu loads them in. 2) init_cur_cpu_trap() is changed to take the current_thread_info() as an argument instead of reading %g6 directly on the current cpu. 3) Create a trampoline stack for the sibling cpus. We do our basic kernel calls using this stack, which is locked into the kernel image, then go to our proper thread stack after taking over the trap table. 4) While we are in this delicate startup state, we put 0xdeadbeef into %g4/%g5/%g6 in order to catch accidental accesses. 5) On the final prom_set_trap_table*() call, we put &init_thread_union into %g6. This is a hack to make prom_world(0) work. All that wants to do is restore the %asi register using get_thread_current_ds(). Longer term we should just do the OBP calls to set the trap table by hand just like we do for everything else. This would avoid that silly prom_world(0) issue, then we can remove the init_thread_union hack. Signed-off-by: David S. Miller <davem@davemloft.net>
2006-02-17 16:29:17 +07:00
init_cur_cpu_trap(current_thread_info());
paging_init();
init_sparc64_elf_hwcap();
smp_fill_in_cpu_possible_map();
/*
* Once the OF device tree and MDESC have been setup and nr_cpus has
* been parsed, we know the list of possible cpus. Therefore we can
* allocate the IRQ stacks.
*/
alloc_irqstack_bootmem();
}
extern int stop_a_enabled;
void sun_do_break(void)
{
if (!stop_a_enabled)
return;
prom_printf("\n");
flush_user_windows();
prom_cmdline();
}
EXPORT_SYMBOL(sun_do_break);
int stop_a_enabled = 1;
EXPORT_SYMBOL(stop_a_enabled);