OBJCOPYFLAGS := -O binary -R .note -R .comment -S
LDFLAGS_vmlinux := -e stext
+check_gcc = $(shell if $(CC) $(1) -S -o /dev/null -xc /dev/null > /dev/null 2>&1 ; then echo "$(1)"; else echo "$(2)"; fi)
+
CFLAGS += -mno-red-zone
CFLAGS += -mcmodel=kernel
CFLAGS += -pipe
# actually it makes the kernel smaller too.
CFLAGS += -fno-reorder-blocks
# should lower this a lot and see how much .text is saves
-CFLAGS += -finline-limit=2000
+# CFLAGS += -finline-limit=2000
CFLAGS += -Wno-sign-compare
# don't enable this when you use kgdb:
ifneq ($(CONFIG_X86_REMOTE_DEBUG),y)
CFLAGS += -fno-asynchronous-unwind-tables
endif
-#CFLAGS += -funit-at-a-time
+CFLAGS += $(call check_gcc,-funit-at-a-time,)
head-y := arch/x86_64/kernel/head.o arch/x86_64/kernel/head64.o arch/x86_64/kernel/init_task.o
# Automatically generated make config: don't edit
#
CONFIG_X86_64=y
+CONFIG_64BIT=y
CONFIG_X86=y
CONFIG_MMU=y
CONFIG_RWSEM_GENERIC_SPINLOCK=y
# Code maturity level options
#
CONFIG_EXPERIMENTAL=y
-# CONFIG_BROKEN is not set
+CONFIG_CLEAN_COMPILE=y
+CONFIG_STANDALONE=y
#
# General setup
#
# Executable file formats / Emulations
#
-CONFIG_KCORE_ELF=y
CONFIG_BINFMT_ELF=y
# CONFIG_BINFMT_MISC is not set
CONFIG_IA32_EMULATION=y
# SCSI device support
#
CONFIG_SCSI=y
+# CONFIG_SCSI_PROC_FS is not set
#
# SCSI support type (disk, tape, CD-ROM)
# CONFIG_SCSI_AIC7XXX_OLD is not set
# CONFIG_SCSI_AIC79XX is not set
# CONFIG_SCSI_ADVANSYS is not set
-# CONFIG_SCSI_AM53C974 is not set
# CONFIG_SCSI_MEGARAID is not set
# CONFIG_SCSI_BUSLOGIC is not set
# CONFIG_SCSI_CPQFCTS is not set
# CONFIG_SCSI_FUTURE_DOMAIN is not set
# CONFIG_SCSI_GDTH is not set
CONFIG_SCSI_IPS=m
-# CONFIG_SCSI_INITIO is not set
# CONFIG_SCSI_INIA100 is not set
# CONFIG_SCSI_SYM53C8XX_2 is not set
-# CONFIG_SCSI_SYM53C8XX is not set
-# CONFIG_SCSI_PCI2000 is not set
-# CONFIG_SCSI_PCI2220I is not set
# CONFIG_SCSI_QLOGIC_ISP is not set
# CONFIG_SCSI_QLOGIC_FC is not set
# CONFIG_SCSI_QLOGIC_1280 is not set
# CONFIG_SCSI_DC395x is not set
-# CONFIG_SCSI_DC390T is not set
# CONFIG_SCSI_NSP32 is not set
# CONFIG_SCSI_DEBUG is not set
# CONFIG_BONDING is not set
# CONFIG_EQUALIZER is not set
# CONFIG_TUN is not set
-# CONFIG_ETHERTAP is not set
#
# Ethernet (10 or 100Mbit)
#
# Ftape, the floppy tape device driver
#
-# CONFIG_FTAPE is not set
CONFIG_AGP=y
CONFIG_AGP_AMD64=y
# CONFIG_DRM is not set
# CONFIG_SOUND_RME96XX is not set
# CONFIG_SOUND_SONICVIBES is not set
# CONFIG_SOUND_TRIDENT is not set
-# CONFIG_SOUND_MSNDCLAS is not set
-# CONFIG_SOUND_MSNDPIN is not set
# CONFIG_SOUND_VIA82CXXX is not set
# CONFIG_SOUND_OSS is not set
# CONFIG_SOUND_ALI5455 is not set
$(obj)/syscall32.o: $(src)/syscall32.c $(obj)/vsyscall.so
-# The DSO images are built using a special linker script.
-$(obj)/vsyscall.so: $(src)/vsyscall.lds $(obj)/vsyscall.o
- $(CC) -m32 -nostdlib -shared -s -Wl,-soname=linux-gate.so.1 \
- -o $@ -Wl,-T,$^
+# Teach kbuild about targets
+targets := vsyscall.o vsyscall.so
-clean-files += vsyscall.so
+# The DSO images are built using a special linker script
+quiet_cmd_vsyscall = SYSCALL $@
+ cmd_vsyscall = $(CC) -m32 -nostdlib -shared -s \
+ -Wl,-soname=linux-gate.so.1 -o $@ \
+ -Wl,-T,$(filter-out FORCE,$^)
+$(obj)/vsyscall.so: $(src)/vsyscall.lds $(obj)/vsyscall.o FORCE
+ $(call if_changed,vsyscall)
AFLAGS_vsyscall.o = -m32
CFLAGS_ia32_ioctl.o += -Ifs/
#include <asm/thread_info.h>
#include <asm/hw_irq.h>
#include <asm/errno.h>
+#include <asm/dwarf2.h>
.code64
-#define PDAREF(field) %gs:field
-
#ifdef CONFIG_PREEMPT
#define preempt_stop cli
#else
/* %rsp:at FRAMEEND */
.macro FIXUP_TOP_OF_STACK tmp
- movq PDAREF(pda_oldrsp),\tmp
+ movq %gs:pda_oldrsp,\tmp
movq \tmp,RSP(%rsp)
movq $__USER_DS,SS(%rsp)
movq $__USER_CS,CS(%rsp)
.macro RESTORE_TOP_OF_STACK tmp,offset=0
movq RSP-\offset(%rsp),\tmp
- movq \tmp,PDAREF(pda_oldrsp)
+ movq \tmp,%gs:pda_oldrsp
movq EFLAGS-\offset(%rsp),\tmp
movq \tmp,R11-\offset(%rsp)
.endm
ENTRY(system_call)
swapgs
- movq %rsp,PDAREF(pda_oldrsp)
- movq PDAREF(pda_kernelstack),%rsp
+ movq %rsp,%gs:pda_oldrsp
+ movq %gs:pda_kernelstack,%rsp
sti
SAVE_ARGS 8,1
movq %rax,ORIG_RAX-ARGOFFSET(%rsp)
jnz sysret_careful
movq RIP-ARGOFFSET(%rsp),%rcx
RESTORE_ARGS 0,-ARG_SKIP,1
- movq PDAREF(pda_oldrsp),%rsp
+ movq %gs:pda_oldrsp,%rsp
swapgs
sysretq
testl $3,CS(%rdi)
je 1f
swapgs
-1: addl $1,PDAREF(pda_irqcount) # RED-PEN should check preempt count
- movq PDAREF(pda_irqstackptr),%rax
+1: addl $1,%gs:pda_irqcount # RED-PEN should check preempt count
+ movq %gs:pda_irqstackptr,%rax
cmoveq %rax,%rsp
pushq %rdi # save old stack
call \func
ret_from_intr:
popq %rdi
cli
- subl $1,PDAREF(pda_irqcount)
+ subl $1,%gs:pda_irqcount
leaq ARGOFFSET(%rdi),%rsp
exit_intr:
GET_THREAD_INFO(%rcx)
/* Set EXTENT bits starting at BASE in BITMAP to value TURN_ON. */
static void set_bitmap(unsigned long *bitmap, unsigned int base, unsigned int extent, int new_value)
{
- unsigned long mask;
- unsigned long *bitmap_base = bitmap + (base / BITS_PER_LONG);
- unsigned int low_index = base & (BITS_PER_LONG-1);
- int length = low_index + extent;
-
- if (low_index != 0) {
- mask = (~0UL << low_index);
- if (length < BITS_PER_LONG)
- mask &= ~(~0UL << length);
- if (new_value)
- *bitmap_base++ |= mask;
- else
- *bitmap_base++ &= ~mask;
- length -= BITS_PER_LONG;
- }
-
- mask = (new_value ? ~0UL : 0UL);
- while (length >= BITS_PER_LONG) {
- *bitmap_base++ = mask;
- length -= BITS_PER_LONG;
- }
-
- if (length > 0) {
- mask = ~(~0UL << length);
+ int i;
if (new_value)
- *bitmap_base++ |= mask;
+ for (i = base; i < base + extent; i++)
+ __set_bit(i, bitmap);
else
- *bitmap_base++ &= ~mask;
- }
+ for (i = base; i < base + extent; i++)
+ clear_bit(i, bitmap);
}
unsigned long mmu_cr4_features;
EXPORT_SYMBOL_GPL(mmu_cr4_features);
-int acpi_disabled __initdata = 0;
+int acpi_disabled = 0;
/* For PCI or other memory-mapped resources */
unsigned long pci_mem_start = 0x10000000;
cpuid(0x80000005, &dummy, &ebx, &ecx, &edx);
printk(KERN_INFO "CPU: L1 I Cache: %dK (%d bytes/line), D cache %dK (%d bytes/line)\n",
edx>>24, edx&0xFF, ecx>>24, ecx&0xFF);
- c->x86_cache_size = (ecx>>24)+(edx>>24);
+ c->x86_cache_size=(ecx>>24)+(edx>>24);
/* DTLB and ITLB together, but only 4K */
- c->x86_tlbsize = ((ebx >> 16) & 0xff) + (ebx & 0xff);
+ c->x86_tlbsize = ((ebx>>16)&0xff) + (ebx&0xff);
}
if (n >= 0x80000006) {
cpuid(0x80000006, &dummy, &ebx, &ecx, &edx);
ecx = cpuid_ecx(0x80000006);
c->x86_cache_size = ecx >> 16;
- c->x86_tlbsize += ((ebx >> 16) & 0xff) + (ebx & 0xff);
+ c->x86_tlbsize += ((ebx >> 16) & 0xfff) + (ebx & 0xfff);
printk(KERN_INFO "CPU: L2 Cache: %dK (%d bytes/line)\n",
c->x86_cache_size, ecx & 0xFF);
bust_spinlocks(1);
}
+void oops_end(void)
+{
+ die_owner = -1;
+ bust_spinlocks(0);
+ spin_unlock(&die_lock);
+ local_irq_enable(); /* make sure back scroll still works */
+}
+
void __die(const char * str, struct pt_regs * regs, long err)
{
static int die_counter;
- handle_BUG(regs);
- printk(KERN_EMERG "%s: %04lx [%u]\n", str, err & 0xffff, ++die_counter);
+ printk(KERN_EMERG "%s: %04lx [%u]\n", str, err & 0xffff,++die_counter);
notify_die(DIE_OOPS, (char *)str, regs, err, 255, SIGSEGV);
show_registers(regs);
- bust_spinlocks(0);
- die_owner = -1;
- spin_unlock_irq(&die_lock);
- do_exit(SIGSEGV);
+ /* Execute summary in case the oops scrolled away */
+ printk(KERN_EMERG "RIP ");
+ printk_address(regs->rip);
+ printk(" RSP <%016lx>\n", regs->rsp);
}
void die(const char * str, struct pt_regs * regs, long err)
{
oops_begin();
+ handle_BUG(regs);
__die(str, regs, err);
+ oops_end();
+ do_exit(SIGSEGV);
}
-
static inline void die_if_kernel(const char * str, struct pt_regs * regs, long err)
{
if (!(regs->eflags & VM_MASK) && (regs->cs == __KERNEL_CS))
copy_page:
prefetch (%rsi)
prefetch 1*64(%rsi)
- prefetch 2*64(%rsi)
- prefetch 3*64(%rsi)
- prefetch 4*64(%rsi)
prefetchw (%rdi)
prefetchw 1*64(%rdi)
- prefetchw 2*64(%rdi)
- prefetchw 3*64(%rdi)
- prefetchw 4*64(%rdi)
subq $3*8,%rsp
movq %rbx,(%rsp)
ignore
prefetch 1*64(%rdi)
ignore
- prefetch 2*64(%rdi)
- ignore
- prefetch 3*64(%rdi)
- ignore
- prefetch 4*64(%rdi)
- ignore
prefetchw (%rsi)
ignore
prefetchw 1*64(%rsi)
- ignore
- prefetchw 2*64(%rsi)
- ignore
- prefetchw 3*64(%rsi)
- ignore
- prefetchw 4*64(%rsi)
.Lignore:
subq $7*8,%rsp
}
}
+/* Sometimes the CPU reports invalid exceptions on prefetch.
+ Check that here and ignore.
+ Opcode checker based on code by Richard Brunner */
+static int is_prefetch(struct pt_regs *regs, unsigned long addr)
+{
+ unsigned char *instr = (unsigned char *)(regs->rip);
+ int scan_more = 1;
+ int prefetch = 0;
+ unsigned char *max_instr = instr + 15;
+
+ /* Avoid recursive faults for this common case */
+ if (regs->rip == addr)
+ return 0;
+
+ /* Code segments in LDT could have a non zero base. Don't check
+ when that's possible */
+ if (regs->cs & (1<<2))
+ return 0;
+
+ while (scan_more && instr < max_instr) {
+ unsigned char opcode;
+ unsigned char instr_hi;
+ unsigned char instr_lo;
+
+ if (__get_user(opcode, instr))
+ break;
+
+ instr_hi = opcode & 0xf0;
+ instr_lo = opcode & 0x0f;
+ instr++;
+
+ switch (instr_hi) {
+ case 0x20:
+ case 0x30:
+ /* Values 0x26,0x2E,0x36,0x3E are valid x86
+ prefixes. In long mode, the CPU will signal
+ invalid opcode if some of these prefixes are
+ present so we will never get here anyway */
+ scan_more = ((instr_lo & 7) == 0x6);
+ break;
+
+ case 0x40:
+ /* In AMD64 long mode, 0x40 to 0x4F are valid REX prefixes
+ Need to figure out under what instruction mode the
+ instruction was issued ... */
+ /* Could check the LDT for lm, but for now it's good
+ enough to assume that long mode only uses well known
+ segments or kernel. */
+ scan_more = ((regs->cs & 3) == 0) || (regs->cs == __USER_CS);
+ break;
+
+ case 0x60:
+ /* 0x64 thru 0x67 are valid prefixes in all modes. */
+ scan_more = (instr_lo & 0xC) == 0x4;
+ break;
+ case 0xF0:
+ /* 0xF0, 0xF2, and 0xF3 are valid prefixes in all modes. */
+ scan_more = !instr_lo || (instr_lo>>1) == 1;
+ break;
+ case 0x00:
+ /* Prefetch instruction is 0x0F0D or 0x0F18 */
+ scan_more = 0;
+ if (__get_user(opcode, instr))
+ break;
+ prefetch = (instr_lo == 0xF) &&
+ (opcode == 0x0D || opcode == 0x18);
+ break;
+ default:
+ scan_more = 0;
+ break;
+ }
+ }
+
+#if 1
+ if (prefetch)
+ printk("%s: prefetch caused page fault at %lx/%lx\n", current->comm,
+ regs->rip, addr);
+#endif
+ return prefetch;
+}
+
static int bad_address(void *p)
{
unsigned long dummy;
printk("BAD\n");
}
+static inline int unhandled_signal(struct task_struct *tsk, int sig)
+{
+ return (tsk->sighand->action[sig-1].sa.sa_handler == SIG_IGN) ||
+ (tsk->sighand->action[sig-1].sa.sa_handler == SIG_DFL);
+}
+
int page_fault_trace;
int exception_trace = 1;
* context, we must not take the fault..
*/
if (unlikely(in_atomic() || !mm))
- goto no_context;
+ goto bad_area_nosemaphore;
again:
down_read(&mm->mmap_sem);
/* User mode accesses just cause a SIGSEGV */
if (error_code & 4) {
+ if (is_prefetch(regs, address))
+ return;
+
if (exception_trace && !(tsk->ptrace & PT_PTRACED) &&
- (tsk->sighand->action[SIGSEGV-1].sa.sa_handler == SIG_IGN ||
- (tsk->sighand->action[SIGSEGV-1].sa.sa_handler == SIG_DFL)))
+ !unhandled_signal(tsk, SIGSEGV)) {
printk(KERN_INFO
"%s[%d]: segfault at %016lx rip %016lx rsp %016lx error %lx\n",
tsk->comm, tsk->pid, address, regs->rip,
regs->rsp, error_code);
+ }
tsk->thread.cr2 = address;
tsk->thread.error_code = error_code;
return;
}
+ if (is_prefetch(regs, address))
+ return;
+
/*
* Oops. The kernel tried to access some bad page. We'll have to
* terminate things with extreme prejudice.
printk(KERN_ALERT "Unable to handle kernel NULL pointer dereference");
else
printk(KERN_ALERT "Unable to handle kernel paging request");
- printk(" at virtual address %016lx\n",address);
- printk(" printing rip:\n");
- printk("%016lx\n", regs->rip);
+ printk(" at %016lx RIP: \n",address);
+ printk_address(regs->rip);
dump_pagetable(address);
__die("Oops", regs, error_code);
-
- /* never reached */
+ /* Execute summary in case the body of the oops scrolled away */
+ printk(KERN_EMERG "CR2: %016lx\n", address);
+ oops_end();
do_exit(SIGKILL);
/*
do_sigbus:
up_read(&mm->mmap_sem);
- /*
- * Send a sigbus, regardless of whether we were in kernel
- * or user mode.
- */
+ /* Kernel mode? Handle exceptions or die */
+ if (!(error_code & 4))
+ goto no_context;
+
tsk->thread.cr2 = address;
tsk->thread.error_code = error_code;
tsk->thread.trap_no = 14;
info.si_code = BUS_ADRERR;
info.si_addr = (void *)address;
force_sig_info(SIGBUS, &info, tsk);
-
- /* Kernel mode? Handle exceptions or die */
- if (!(error_code & 4))
- goto no_context;
return;
-
vmalloc_fault:
{
pgd_t *pgd;
#define LDT_entry_a(info) \
((((info)->base_addr & 0x0000ffff) << 16) | ((info)->limit & 0x0ffff))
+/* Don't allow setting of the lm bit. It is useless anyways because
+ 64bit system calls require __USER_CS. */
#define LDT_entry_b(info) \
(((info)->base_addr & 0xff000000) | \
(((info)->base_addr & 0x00ff0000) >> 16) | \
((info)->seg_32bit << 22) | \
((info)->limit_in_pages << 23) | \
((info)->useable << 20) | \
- ((info)->lm << 21) | \
+ /* ((info)->lm << 21) | */ \
0x7000)
#define LDT_empty(info) (\
--- /dev/null
+#ifndef _DWARF2_H
+#define _DWARF2_H 1
+
+#include <linux/config.h>
+
+#ifndef __ASSEMBLY__
+#warning "asm/dwarf2.h should be only included in pure assembly files"
+#endif
+
+/*
+ Macros for dwarf2 CFI unwind table entries.
+ See "as.info" for details on these pseudo ops. Unfortunately
+ they are only supported in very new binutils, so define them
+ away for older version.
+ */
+
+#ifdef CONFIG_CFI_BINUTILS
+
+#define CFI_STARTPROC .cfi_startproc
+#define CFI_ENDPROC .cfi_endproc
+#define CFI_DEF_CFA .cfi_def_cfa
+#define CFI_DEF_CFA_REGISTER .cfi_def_cfa_register
+#define CFI_DEF_CFA_OFFSET .cfi_def_cfa_offset
+#define CFI_ADJUST_CFA_OFFSET .cfi_adjust_cfa_offset
+#define CFI_OFFSET .cfi_offset
+#define CFI_REL_OFFSET .cfi_rel_offset
+
+#else
+
+#ifdef __ASSEMBLY__
+ .macro nothing
+ .endm
+ .macro nothing1 a
+ .endm
+ .macro nothing2 a,b
+ .endm
+#endif
+
+#define CFI_STARTPROC nothing
+#define CFI_ENDPROC nothing
+#define CFI_DEF_CFA nothing2
+#define CFI_DEF_CFA_REGISTER nothing1
+#define CFI_DEF_CFA_OFFSET nothing1
+#define CFI_ADJUST_CFA_OFFSET nothing1
+#define CFI_OFFSET nothing2
+#define CFI_REL_OFFSET nothing2
+
+#endif
+
+#endif
return notifier_call_chain(&die_chain, val, &args);
}
-int printk_address(unsigned long address);
+extern int printk_address(unsigned long address);
+extern void die(const char *,struct pt_regs *,long);
+extern void __die(const char *,struct pt_regs *,long);
+extern void show_registers(struct pt_regs *regs);
+extern void dump_pagetable(unsigned long);
+extern void oops_begin(void);
+extern void oops_end(void);
#endif
#define __ARCH_SI_BAND_T long
+#define SIGEV_PAD_SIZE ((SIGEV_MAX_SIZE/sizeof(int)) - 4)
+
#include <asm-generic/siginfo.h>
#endif
({ \
int __pu_err = -EFAULT; \
__typeof__(*(ptr)) *__pu_addr = (ptr); \
- if (access_ok(VERIFY_WRITE,__pu_addr,size)) \
+ if (likely(access_ok(VERIFY_WRITE,__pu_addr,size))) \
__put_user_size((x),__pu_addr,(size),__pu_err); \
__pu_err; \
})
return ret;
case 10:
__get_user_asm(*(u64*)dst,(u64*)src,ret,"q","","=r",16);
- if (ret) return ret;
+ if (unlikely(ret)) return ret;
__get_user_asm(*(u16*)(8+(char*)dst),(u16*)(8+(char*)src),ret,"w","w","=r",2);
return ret;
case 16:
__get_user_asm(*(u64*)dst,(u64*)src,ret,"q","","=r",16);
- if (ret) return ret;
+ if (unlikely(ret)) return ret;
__get_user_asm(*(u64*)(8+(char*)dst),(u64*)(8+(char*)src),ret,"q","","=r",8);
return ret;
default:
return ret;
case 10:
__put_user_asm(*(u64*)src,(u64*)dst,ret,"q","","ir",10);
- if (ret) return ret;
+ if (unlikely(ret)) return ret;
asm("":::"memory");
__put_user_asm(4[(u16*)src],4+(u16*)dst,ret,"w","w","ir",2);
return ret;
case 16:
__put_user_asm(*(u64*)src,(u64*)dst,ret,"q","","ir",16);
- if (ret) return ret;
+ if (unlikely(ret)) return ret;
asm("":::"memory");
__put_user_asm(1[(u64*)src],1+(u64*)dst,ret,"q","","ir",8);
return ret;
case 1: {
u8 tmp;
__get_user_asm(tmp,(u8 *)src,ret,"b","b","=q",1);
- if (!ret)
+ if (likely(!ret))
__put_user_asm(tmp,(u8 *)dst,ret,"b","b","iq",1);
return ret;
}
case 2: {
u16 tmp;
__get_user_asm(tmp,(u16 *)src,ret,"w","w","=r",2);
- if (!ret)
+ if (likely(!ret))
__put_user_asm(tmp,(u16 *)dst,ret,"w","w","ir",2);
return ret;
}
case 4: {
u32 tmp;
__get_user_asm(tmp,(u32 *)src,ret,"l","k","=r",4);
- if (!ret)
+ if (likely(!ret))
__put_user_asm(tmp,(u32 *)dst,ret,"l","k","ir",4);
return ret;
}
case 8: {
u64 tmp;
__get_user_asm(tmp,(u64 *)src,ret,"q","","=r",8);
- if (!ret)
+ if (likely(!ret))
__put_user_asm(tmp,(u64 *)dst,ret,"q","","ir",8);
return ret;
}
projects / linux / commitdiff