android_kernel_oneplus_msm8998/arch/mips/kernel/process.c

791 lines
19 KiB
C
Raw Normal View History

/*
* 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) 1994 - 1999, 2000 by Ralf Baechle and others.
* Copyright (C) 2005, 2006 by Ralf Baechle (ralf@linux-mips.org)
* Copyright (C) 1999, 2000 Silicon Graphics, Inc.
* Copyright (C) 2004 Thiemo Seufer
* Copyright (C) 2013 Imagination Technologies Ltd.
*/
#include <linux/errno.h>
#include <linux/sched.h>
#include <linux/tick.h>
#include <linux/kernel.h>
#include <linux/mm.h>
#include <linux/stddef.h>
#include <linux/unistd.h>
#include <linux/export.h>
#include <linux/ptrace.h>
#include <linux/mman.h>
#include <linux/personality.h>
#include <linux/sys.h>
#include <linux/init.h>
#include <linux/completion.h>
#include <linux/kallsyms.h>
#include <linux/random.h>
2015-01-08 12:17:37 +00:00
#include <linux/prctl.h>
#include <asm/asm.h>
#include <asm/bootinfo.h>
#include <asm/cpu.h>
#include <asm/dsp.h>
#include <asm/fpu.h>
#include <asm/irq.h>
#include <asm/msa.h>
#include <asm/pgtable.h>
#include <asm/mipsregs.h>
#include <asm/processor.h>
#include <asm/reg.h>
#include <asm/uaccess.h>
#include <asm/io.h>
#include <asm/elf.h>
#include <asm/isadep.h>
#include <asm/inst.h>
#include <asm/stacktrace.h>
#include <asm/irq_regs.h>
#ifdef CONFIG_HOTPLUG_CPU
void arch_cpu_idle_dead(void)
{
play_dead();
}
#endif
asmlinkage void ret_from_fork(void);
asmlinkage void ret_from_kernel_thread(void);
void start_thread(struct pt_regs * regs, unsigned long pc, unsigned long sp)
{
unsigned long status;
/* New thread loses kernel privileges. */
status = regs->cp0_status & ~(ST0_CU0|ST0_CU1|ST0_FR|KU_MASK);
status |= KU_USER;
regs->cp0_status = status;
clear_used_math();
clear_fpu_owner();
init_dsp();
clear_thread_flag(TIF_USEDMSA);
clear_thread_flag(TIF_MSA_CTX_LIVE);
disable_msa();
regs->cp0_epc = pc;
regs->regs[29] = sp;
}
void exit_thread(void)
{
}
void flush_thread(void)
{
}
MIPS: fork: Fix MSA/FPU/DSP context duplication race There is a race in the MIPS fork code which allows the child to get a stale copy of parent MSA/FPU/DSP state that is active in hardware registers when the fork() is called. This is because copy_thread() saves the live register state into the child context only if the hardware is currently in use, apparently on the assumption that the hardware state cannot have been saved and disabled since the initial duplication of the task_struct. However preemption is certainly possible during this window. An example sequence of events is as follows: 1) The parent userland process puts important data into saved floating point registers ($f20-$f31), which are then dirty compared to the process' stored context. 2) The parent process calls fork() which does a clone system call. 3) In the kernel, do_fork() -> copy_process() -> dup_task_struct() -> arch_dup_task_struct() (which uses the weakly defined default implementation). This duplicates the parent process' task context, which includes a stale version of its FP context from when it was last saved, probably some time before (1). 4) At some point before copy_process() calls copy_thread(), such as when duplicating the memory map, the process is desceduled. Perhaps it is preempted asynchronously, or perhaps it sleeps while blocked on a mutex. The dirty FP state in the FP registers is saved to the parent process' context and the FPU is disabled. 5) When the process is rescheduled again it continues copying state until it gets to copy_thread(), which checks whether the FPU is in use, so that it can copy that dirty state to the child process' task context. Because of the deschedule however the FPU is not in use, so the child process' context is left with stale FP context from the last time the parent saved it (some time before (1)). 6) When the new child process is scheduled it reads the important data from the saved floating point register, and ends up doing a NULL pointer dereference as a result of the stale data. This use of saved floating point registers across function calls can be triggered fairly easily by explicitly using inline asm with a current (MIPS R2) compiler, but is far more likely to happen unintentionally with a MIPS R6 compiler where the FP registers are more likely to get used as scratch registers for storing non-fp data. It is easily fixed, in the same way that other architectures do it, by overriding the implementation of arch_dup_task_struct() to sync the dirty hardware state to the parent process' task context *prior* to duplicating it, rather than copying straight to the child process' task context in copy_thread(). Note, the FPU hardware is not disabled so the parent process may continue executing with the live register context, but now the child process is guaranteed to have an identical copy of it at that point. Signed-off-by: James Hogan <james.hogan@imgtec.com> Reported-by: Matthew Fortune <matthew.fortune@imgtec.com> Tested-by: Markos Chandras <markos.chandras@imgtec.com> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Paul Burton <paul.burton@imgtec.com> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/9075/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org>
2015-01-19 10:30:54 +00:00
int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src)
{
/*
* Save any process state which is live in hardware registers to the
* parent context prior to duplication. This prevents the new child
* state becoming stale if the parent is preempted before copy_thread()
* gets a chance to save the parent's live hardware registers to the
* child context.
*/
preempt_disable();
if (is_msa_enabled())
save_msa(current);
else if (is_fpu_owner())
_save_fp(current);
save_dsp(current);
preempt_enable();
*dst = *src;
return 0;
}
/*
* Copy architecture-specific thread state
*/
int copy_thread(unsigned long clone_flags, unsigned long usp,
unsigned long kthread_arg, struct task_struct *p)
{
struct thread_info *ti = task_thread_info(p);
struct pt_regs *childregs, *regs = current_pt_regs();
unsigned long childksp;
childksp = (unsigned long)task_stack_page(p) + THREAD_SIZE - 32;
/* set up new TSS. */
childregs = (struct pt_regs *) childksp - 1;
/* Put the stack after the struct pt_regs. */
childksp = (unsigned long) childregs;
p->thread.cp0_status = read_c0_status() & ~(ST0_CU2|ST0_CU1);
if (unlikely(p->flags & PF_KTHREAD)) {
/* kernel thread */
unsigned long status = p->thread.cp0_status;
memset(childregs, 0, sizeof(struct pt_regs));
ti->addr_limit = KERNEL_DS;
p->thread.reg16 = usp; /* fn */
p->thread.reg17 = kthread_arg;
p->thread.reg29 = childksp;
p->thread.reg31 = (unsigned long) ret_from_kernel_thread;
#if defined(CONFIG_CPU_R3000) || defined(CONFIG_CPU_TX39XX)
status = (status & ~(ST0_KUP | ST0_IEP | ST0_IEC)) |
((status & (ST0_KUC | ST0_IEC)) << 2);
#else
status |= ST0_EXL;
#endif
childregs->cp0_status = status;
return 0;
}
/* user thread */
*childregs = *regs;
childregs->regs[7] = 0; /* Clear error flag */
childregs->regs[2] = 0; /* Child gets zero as return value */
if (usp)
childregs->regs[29] = usp;
ti->addr_limit = USER_DS;
p->thread.reg29 = (unsigned long) childregs;
p->thread.reg31 = (unsigned long) ret_from_fork;
/*
* New tasks lose permission to use the fpu. This accelerates context
* switching for most programs since they don't use the fpu.
*/
childregs->cp0_status &= ~(ST0_CU2|ST0_CU1);
clear_tsk_thread_flag(p, TIF_USEDFPU);
clear_tsk_thread_flag(p, TIF_USEDMSA);
clear_tsk_thread_flag(p, TIF_MSA_CTX_LIVE);
#ifdef CONFIG_MIPS_MT_FPAFF
clear_tsk_thread_flag(p, TIF_FPUBOUND);
#endif /* CONFIG_MIPS_MT_FPAFF */
if (clone_flags & CLONE_SETTLS)
ti->tp_value = regs->regs[7];
return 0;
}
#ifdef CONFIG_CC_STACKPROTECTOR
#include <linux/stackprotector.h>
unsigned long __stack_chk_guard __read_mostly;
EXPORT_SYMBOL(__stack_chk_guard);
#endif
struct mips_frame_info {
void *func;
unsigned long func_size;
int frame_size;
int pc_offset;
};
#define J_TARGET(pc,target) \
(((unsigned long)(pc) & 0xf0000000) | ((target) << 2))
MIPS: Calculate microMIPS ra properly when unwinding the stack commit bb9bc4689b9c635714fbcd5d335bad9934a7ebfc upstream. get_frame_info() calculates the offset of the return address within a stack frame simply by dividing a the bottom 16 bits of the instruction, treated as a signed integer, by the size of a long. Whilst this works for MIPS32 & MIPS64 ISAs where the sw or sd instructions are used, it's incorrect for microMIPS where encodings differ. The result is that we typically completely fail to unwind the stack on microMIPS. Fix this by adjusting is_ra_save_ins() to calculate the return address offset, and take into account the various different encodings there in the same place as we consider whether an instruction is storing the ra/$31 register. With this we are now able to unwind the stack for kernels targetting the microMIPS ISA, for example we can produce: Call Trace: [<80109e1f>] show_stack+0x63/0x7c [<8011ea17>] __warn+0x9b/0xac [<8011ea45>] warn_slowpath_fmt+0x1d/0x20 [<8013fe53>] register_console+0x43/0x314 [<8067c58d>] of_setup_earlycon+0x1dd/0x1ec [<8067f63f>] early_init_dt_scan_chosen_stdout+0xe7/0xf8 [<8066c115>] do_early_param+0x75/0xac [<801302f9>] parse_args+0x1dd/0x308 [<8066c459>] parse_early_options+0x25/0x28 [<8066c48b>] parse_early_param+0x2f/0x38 [<8066e8cf>] setup_arch+0x113/0x488 [<8066c4f3>] start_kernel+0x57/0x328 ---[ end trace 0000000000000000 ]--- Whereas previously we only produced: Call Trace: [<80109e1f>] show_stack+0x63/0x7c ---[ end trace 0000000000000000 ]--- Signed-off-by: Paul Burton <paul.burton@imgtec.com> Fixes: 34c2f668d0f6 ("MIPS: microMIPS: Add unaligned access support.") Cc: Leonid Yegoshin <leonid.yegoshin@imgtec.com> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/14532/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-07 15:07:06 +00:00
static inline int is_ra_save_ins(union mips_instruction *ip, int *poff)
{
#ifdef CONFIG_CPU_MICROMIPS
/*
* swsp ra,offset
* swm16 reglist,offset(sp)
* swm32 reglist,offset(sp)
* sw32 ra,offset(sp)
* jradiussp - NOT SUPPORTED
*
* microMIPS is way more fun...
*/
if (mm_insn_16bit(ip->halfword[1])) {
MIPS: Calculate microMIPS ra properly when unwinding the stack commit bb9bc4689b9c635714fbcd5d335bad9934a7ebfc upstream. get_frame_info() calculates the offset of the return address within a stack frame simply by dividing a the bottom 16 bits of the instruction, treated as a signed integer, by the size of a long. Whilst this works for MIPS32 & MIPS64 ISAs where the sw or sd instructions are used, it's incorrect for microMIPS where encodings differ. The result is that we typically completely fail to unwind the stack on microMIPS. Fix this by adjusting is_ra_save_ins() to calculate the return address offset, and take into account the various different encodings there in the same place as we consider whether an instruction is storing the ra/$31 register. With this we are now able to unwind the stack for kernels targetting the microMIPS ISA, for example we can produce: Call Trace: [<80109e1f>] show_stack+0x63/0x7c [<8011ea17>] __warn+0x9b/0xac [<8011ea45>] warn_slowpath_fmt+0x1d/0x20 [<8013fe53>] register_console+0x43/0x314 [<8067c58d>] of_setup_earlycon+0x1dd/0x1ec [<8067f63f>] early_init_dt_scan_chosen_stdout+0xe7/0xf8 [<8066c115>] do_early_param+0x75/0xac [<801302f9>] parse_args+0x1dd/0x308 [<8066c459>] parse_early_options+0x25/0x28 [<8066c48b>] parse_early_param+0x2f/0x38 [<8066e8cf>] setup_arch+0x113/0x488 [<8066c4f3>] start_kernel+0x57/0x328 ---[ end trace 0000000000000000 ]--- Whereas previously we only produced: Call Trace: [<80109e1f>] show_stack+0x63/0x7c ---[ end trace 0000000000000000 ]--- Signed-off-by: Paul Burton <paul.burton@imgtec.com> Fixes: 34c2f668d0f6 ("MIPS: microMIPS: Add unaligned access support.") Cc: Leonid Yegoshin <leonid.yegoshin@imgtec.com> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/14532/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-07 15:07:06 +00:00
switch (ip->mm16_r5_format.opcode) {
case mm_swsp16_op:
if (ip->mm16_r5_format.rt != 31)
return 0;
*poff = ip->mm16_r5_format.imm;
MIPS: Calculate microMIPS ra properly when unwinding the stack commit bb9bc4689b9c635714fbcd5d335bad9934a7ebfc upstream. get_frame_info() calculates the offset of the return address within a stack frame simply by dividing a the bottom 16 bits of the instruction, treated as a signed integer, by the size of a long. Whilst this works for MIPS32 & MIPS64 ISAs where the sw or sd instructions are used, it's incorrect for microMIPS where encodings differ. The result is that we typically completely fail to unwind the stack on microMIPS. Fix this by adjusting is_ra_save_ins() to calculate the return address offset, and take into account the various different encodings there in the same place as we consider whether an instruction is storing the ra/$31 register. With this we are now able to unwind the stack for kernels targetting the microMIPS ISA, for example we can produce: Call Trace: [<80109e1f>] show_stack+0x63/0x7c [<8011ea17>] __warn+0x9b/0xac [<8011ea45>] warn_slowpath_fmt+0x1d/0x20 [<8013fe53>] register_console+0x43/0x314 [<8067c58d>] of_setup_earlycon+0x1dd/0x1ec [<8067f63f>] early_init_dt_scan_chosen_stdout+0xe7/0xf8 [<8066c115>] do_early_param+0x75/0xac [<801302f9>] parse_args+0x1dd/0x308 [<8066c459>] parse_early_options+0x25/0x28 [<8066c48b>] parse_early_param+0x2f/0x38 [<8066e8cf>] setup_arch+0x113/0x488 [<8066c4f3>] start_kernel+0x57/0x328 ---[ end trace 0000000000000000 ]--- Whereas previously we only produced: Call Trace: [<80109e1f>] show_stack+0x63/0x7c ---[ end trace 0000000000000000 ]--- Signed-off-by: Paul Burton <paul.burton@imgtec.com> Fixes: 34c2f668d0f6 ("MIPS: microMIPS: Add unaligned access support.") Cc: Leonid Yegoshin <leonid.yegoshin@imgtec.com> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/14532/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-07 15:07:06 +00:00
*poff = (*poff << 2) / sizeof(ulong);
return 1;
case mm_pool16c_op:
switch (ip->mm16_m_format.func) {
case mm_swm16_op:
*poff = ip->mm16_m_format.imm;
*poff += 1 + ip->mm16_m_format.rlist;
*poff = (*poff << 2) / sizeof(ulong);
return 1;
default:
return 0;
}
default:
return 0;
}
}
MIPS: Calculate microMIPS ra properly when unwinding the stack commit bb9bc4689b9c635714fbcd5d335bad9934a7ebfc upstream. get_frame_info() calculates the offset of the return address within a stack frame simply by dividing a the bottom 16 bits of the instruction, treated as a signed integer, by the size of a long. Whilst this works for MIPS32 & MIPS64 ISAs where the sw or sd instructions are used, it's incorrect for microMIPS where encodings differ. The result is that we typically completely fail to unwind the stack on microMIPS. Fix this by adjusting is_ra_save_ins() to calculate the return address offset, and take into account the various different encodings there in the same place as we consider whether an instruction is storing the ra/$31 register. With this we are now able to unwind the stack for kernels targetting the microMIPS ISA, for example we can produce: Call Trace: [<80109e1f>] show_stack+0x63/0x7c [<8011ea17>] __warn+0x9b/0xac [<8011ea45>] warn_slowpath_fmt+0x1d/0x20 [<8013fe53>] register_console+0x43/0x314 [<8067c58d>] of_setup_earlycon+0x1dd/0x1ec [<8067f63f>] early_init_dt_scan_chosen_stdout+0xe7/0xf8 [<8066c115>] do_early_param+0x75/0xac [<801302f9>] parse_args+0x1dd/0x308 [<8066c459>] parse_early_options+0x25/0x28 [<8066c48b>] parse_early_param+0x2f/0x38 [<8066e8cf>] setup_arch+0x113/0x488 [<8066c4f3>] start_kernel+0x57/0x328 ---[ end trace 0000000000000000 ]--- Whereas previously we only produced: Call Trace: [<80109e1f>] show_stack+0x63/0x7c ---[ end trace 0000000000000000 ]--- Signed-off-by: Paul Burton <paul.burton@imgtec.com> Fixes: 34c2f668d0f6 ("MIPS: microMIPS: Add unaligned access support.") Cc: Leonid Yegoshin <leonid.yegoshin@imgtec.com> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/14532/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-07 15:07:06 +00:00
switch (ip->i_format.opcode) {
case mm_sw32_op:
if (ip->i_format.rs != 29)
return 0;
if (ip->i_format.rt != 31)
return 0;
*poff = ip->i_format.simmediate / sizeof(ulong);
return 1;
case mm_pool32b_op:
switch (ip->mm_m_format.func) {
case mm_swm32_func:
if (ip->mm_m_format.rd < 0x10)
return 0;
if (ip->mm_m_format.base != 29)
return 0;
*poff = ip->mm_m_format.simmediate;
*poff += (ip->mm_m_format.rd & 0xf) * sizeof(u32);
*poff /= sizeof(ulong);
return 1;
default:
return 0;
}
default:
return 0;
}
#else
/* sw / sd $ra, offset($sp) */
MIPS: Calculate microMIPS ra properly when unwinding the stack commit bb9bc4689b9c635714fbcd5d335bad9934a7ebfc upstream. get_frame_info() calculates the offset of the return address within a stack frame simply by dividing a the bottom 16 bits of the instruction, treated as a signed integer, by the size of a long. Whilst this works for MIPS32 & MIPS64 ISAs where the sw or sd instructions are used, it's incorrect for microMIPS where encodings differ. The result is that we typically completely fail to unwind the stack on microMIPS. Fix this by adjusting is_ra_save_ins() to calculate the return address offset, and take into account the various different encodings there in the same place as we consider whether an instruction is storing the ra/$31 register. With this we are now able to unwind the stack for kernels targetting the microMIPS ISA, for example we can produce: Call Trace: [<80109e1f>] show_stack+0x63/0x7c [<8011ea17>] __warn+0x9b/0xac [<8011ea45>] warn_slowpath_fmt+0x1d/0x20 [<8013fe53>] register_console+0x43/0x314 [<8067c58d>] of_setup_earlycon+0x1dd/0x1ec [<8067f63f>] early_init_dt_scan_chosen_stdout+0xe7/0xf8 [<8066c115>] do_early_param+0x75/0xac [<801302f9>] parse_args+0x1dd/0x308 [<8066c459>] parse_early_options+0x25/0x28 [<8066c48b>] parse_early_param+0x2f/0x38 [<8066e8cf>] setup_arch+0x113/0x488 [<8066c4f3>] start_kernel+0x57/0x328 ---[ end trace 0000000000000000 ]--- Whereas previously we only produced: Call Trace: [<80109e1f>] show_stack+0x63/0x7c ---[ end trace 0000000000000000 ]--- Signed-off-by: Paul Burton <paul.burton@imgtec.com> Fixes: 34c2f668d0f6 ("MIPS: microMIPS: Add unaligned access support.") Cc: Leonid Yegoshin <leonid.yegoshin@imgtec.com> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/14532/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-07 15:07:06 +00:00
if ((ip->i_format.opcode == sw_op || ip->i_format.opcode == sd_op) &&
ip->i_format.rs == 29 && ip->i_format.rt == 31) {
*poff = ip->i_format.simmediate / sizeof(ulong);
return 1;
}
return 0;
#endif
}
static inline int is_jump_ins(union mips_instruction *ip)
{
#ifdef CONFIG_CPU_MICROMIPS
/*
* jr16,jrc,jalr16,jalr16
* jal
* jalr/jr,jalr.hb/jr.hb,jalrs,jalrs.hb
* jraddiusp - NOT SUPPORTED
*
* microMIPS is kind of more fun...
*/
if (mm_insn_16bit(ip->halfword[1])) {
if ((ip->mm16_r5_format.opcode == mm_pool16c_op &&
(ip->mm16_r5_format.rt & mm_jr16_op) == mm_jr16_op))
return 1;
return 0;
}
if (ip->j_format.opcode == mm_j32_op)
return 1;
if (ip->j_format.opcode == mm_jal32_op)
return 1;
if (ip->r_format.opcode != mm_pool32a_op ||
ip->r_format.func != mm_pool32axf_op)
return 0;
return ((ip->u_format.uimmediate >> 6) & mm_jalr_op) == mm_jalr_op;
#else
if (ip->j_format.opcode == j_op)
return 1;
if (ip->j_format.opcode == jal_op)
return 1;
if (ip->r_format.opcode != spec_op)
return 0;
return ip->r_format.func == jalr_op || ip->r_format.func == jr_op;
#endif
}
static inline int is_sp_move_ins(union mips_instruction *ip)
{
#ifdef CONFIG_CPU_MICROMIPS
/*
* addiusp -imm
* addius5 sp,-imm
* addiu32 sp,sp,-imm
* jradiussp - NOT SUPPORTED
*
* microMIPS is not more fun...
*/
if (mm_insn_16bit(ip->halfword[1])) {
return (ip->mm16_r3_format.opcode == mm_pool16d_op &&
ip->mm16_r3_format.simmediate && mm_addiusp_func) ||
(ip->mm16_r5_format.opcode == mm_pool16d_op &&
ip->mm16_r5_format.rt == 29);
}
return ip->mm_i_format.opcode == mm_addiu32_op &&
ip->mm_i_format.rt == 29 && ip->mm_i_format.rs == 29;
#else
/* addiu/daddiu sp,sp,-imm */
if (ip->i_format.rs != 29 || ip->i_format.rt != 29)
return 0;
if (ip->i_format.opcode == addiu_op || ip->i_format.opcode == daddiu_op)
return 1;
#endif
return 0;
}
static int get_frame_info(struct mips_frame_info *info)
{
bool is_mmips = IS_ENABLED(CONFIG_CPU_MICROMIPS);
union mips_instruction insn, *ip;
const unsigned int max_insns = 128;
MIPS: Handle non word sized instructions when examining frame [ Upstream commit 11887ed172a6960673f130dad8f8fb42778f64d7 ] Commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") added fairly broken support for handling 16bit microMIPS instructions in get_frame_info(). It adjusts the instruction pointer by 16bits in the case of a 16bit sp move instruction, but not any other 16bit instruction. Commit b6c7a324df37 ("MIPS: Fix get_frame_info() handling of microMIPS function size") goes some way to fixing get_frame_info() to iterate over microMIPS instuctions, but the instruction pointer is still manipulated using a postincrement, and is of union mips_instruction type. Since the union is sized to the largest member (a word), but microMIPS instructions are a mix of halfword and word sizes, the function does not always iterate correctly, ending up misaligned with the instruction stream and interpreting it incorrectly. Since the instruction modifying the stack pointer is usually the first in the function, that one is usually handled correctly. But the instruction which saves the return address to the sp is some variable number of instructions into the frame and is frequently missed due to not being on a word boundary, leading to incomplete walking of the stack. Fix this by incrementing the instruction pointer based on the size of the previously decoded instruction (& remove the hack introduced by commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") which adjusts the instruction pointer in the case of a 16bit sp move instruction, but not any other). Fixes: 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Burton <paul.burton@imgtec.com> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/16953/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
2017-08-08 13:22:30 +01:00
unsigned int last_insn_size = 0;
unsigned int i;
info->pc_offset = -1;
info->frame_size = 0;
ip = (void *)msk_isa16_mode((ulong)info->func);
if (!ip)
goto err;
for (i = 0; i < max_insns; i++) {
MIPS: Handle non word sized instructions when examining frame [ Upstream commit 11887ed172a6960673f130dad8f8fb42778f64d7 ] Commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") added fairly broken support for handling 16bit microMIPS instructions in get_frame_info(). It adjusts the instruction pointer by 16bits in the case of a 16bit sp move instruction, but not any other 16bit instruction. Commit b6c7a324df37 ("MIPS: Fix get_frame_info() handling of microMIPS function size") goes some way to fixing get_frame_info() to iterate over microMIPS instuctions, but the instruction pointer is still manipulated using a postincrement, and is of union mips_instruction type. Since the union is sized to the largest member (a word), but microMIPS instructions are a mix of halfword and word sizes, the function does not always iterate correctly, ending up misaligned with the instruction stream and interpreting it incorrectly. Since the instruction modifying the stack pointer is usually the first in the function, that one is usually handled correctly. But the instruction which saves the return address to the sp is some variable number of instructions into the frame and is frequently missed due to not being on a word boundary, leading to incomplete walking of the stack. Fix this by incrementing the instruction pointer based on the size of the previously decoded instruction (& remove the hack introduced by commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") which adjusts the instruction pointer in the case of a 16bit sp move instruction, but not any other). Fixes: 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Burton <paul.burton@imgtec.com> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/16953/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
2017-08-08 13:22:30 +01:00
ip = (void *)ip + last_insn_size;
if (is_mmips && mm_insn_16bit(ip->halfword[0])) {
insn.halfword[0] = 0;
insn.halfword[1] = ip->halfword[0];
MIPS: Handle non word sized instructions when examining frame [ Upstream commit 11887ed172a6960673f130dad8f8fb42778f64d7 ] Commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") added fairly broken support for handling 16bit microMIPS instructions in get_frame_info(). It adjusts the instruction pointer by 16bits in the case of a 16bit sp move instruction, but not any other 16bit instruction. Commit b6c7a324df37 ("MIPS: Fix get_frame_info() handling of microMIPS function size") goes some way to fixing get_frame_info() to iterate over microMIPS instuctions, but the instruction pointer is still manipulated using a postincrement, and is of union mips_instruction type. Since the union is sized to the largest member (a word), but microMIPS instructions are a mix of halfword and word sizes, the function does not always iterate correctly, ending up misaligned with the instruction stream and interpreting it incorrectly. Since the instruction modifying the stack pointer is usually the first in the function, that one is usually handled correctly. But the instruction which saves the return address to the sp is some variable number of instructions into the frame and is frequently missed due to not being on a word boundary, leading to incomplete walking of the stack. Fix this by incrementing the instruction pointer based on the size of the previously decoded instruction (& remove the hack introduced by commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") which adjusts the instruction pointer in the case of a 16bit sp move instruction, but not any other). Fixes: 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Burton <paul.burton@imgtec.com> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/16953/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
2017-08-08 13:22:30 +01:00
last_insn_size = 2;
} else if (is_mmips) {
insn.halfword[0] = ip->halfword[1];
insn.halfword[1] = ip->halfword[0];
MIPS: Handle non word sized instructions when examining frame [ Upstream commit 11887ed172a6960673f130dad8f8fb42778f64d7 ] Commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") added fairly broken support for handling 16bit microMIPS instructions in get_frame_info(). It adjusts the instruction pointer by 16bits in the case of a 16bit sp move instruction, but not any other 16bit instruction. Commit b6c7a324df37 ("MIPS: Fix get_frame_info() handling of microMIPS function size") goes some way to fixing get_frame_info() to iterate over microMIPS instuctions, but the instruction pointer is still manipulated using a postincrement, and is of union mips_instruction type. Since the union is sized to the largest member (a word), but microMIPS instructions are a mix of halfword and word sizes, the function does not always iterate correctly, ending up misaligned with the instruction stream and interpreting it incorrectly. Since the instruction modifying the stack pointer is usually the first in the function, that one is usually handled correctly. But the instruction which saves the return address to the sp is some variable number of instructions into the frame and is frequently missed due to not being on a word boundary, leading to incomplete walking of the stack. Fix this by incrementing the instruction pointer based on the size of the previously decoded instruction (& remove the hack introduced by commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") which adjusts the instruction pointer in the case of a 16bit sp move instruction, but not any other). Fixes: 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Burton <paul.burton@imgtec.com> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/16953/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
2017-08-08 13:22:30 +01:00
last_insn_size = 4;
} else {
insn.word = ip->word;
MIPS: Handle non word sized instructions when examining frame [ Upstream commit 11887ed172a6960673f130dad8f8fb42778f64d7 ] Commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") added fairly broken support for handling 16bit microMIPS instructions in get_frame_info(). It adjusts the instruction pointer by 16bits in the case of a 16bit sp move instruction, but not any other 16bit instruction. Commit b6c7a324df37 ("MIPS: Fix get_frame_info() handling of microMIPS function size") goes some way to fixing get_frame_info() to iterate over microMIPS instuctions, but the instruction pointer is still manipulated using a postincrement, and is of union mips_instruction type. Since the union is sized to the largest member (a word), but microMIPS instructions are a mix of halfword and word sizes, the function does not always iterate correctly, ending up misaligned with the instruction stream and interpreting it incorrectly. Since the instruction modifying the stack pointer is usually the first in the function, that one is usually handled correctly. But the instruction which saves the return address to the sp is some variable number of instructions into the frame and is frequently missed due to not being on a word boundary, leading to incomplete walking of the stack. Fix this by incrementing the instruction pointer based on the size of the previously decoded instruction (& remove the hack introduced by commit 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") which adjusts the instruction pointer in the case of a 16bit sp move instruction, but not any other). Fixes: 34c2f668d0f6b ("MIPS: microMIPS: Add unaligned access support.") Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Paul Burton <paul.burton@imgtec.com> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/16953/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <sashal@kernel.org>
2017-08-08 13:22:30 +01:00
last_insn_size = 4;
}
if (is_jump_ins(&insn))
break;
if (!info->frame_size) {
if (is_sp_move_ins(&insn))
{
#ifdef CONFIG_CPU_MICROMIPS
if (mm_insn_16bit(ip->halfword[0]))
{
unsigned short tmp;
if (ip->halfword[0] & mm_addiusp_func)
{
tmp = (((ip->halfword[0] >> 1) & 0x1ff) << 2);
info->frame_size = -(signed short)(tmp | ((tmp & 0x100) ? 0xfe00 : 0));
} else {
tmp = (ip->halfword[0] >> 1);
info->frame_size = -(signed short)(tmp & 0xf);
}
} else
#endif
info->frame_size = - ip->i_format.simmediate;
}
continue;
}
MIPS: Calculate microMIPS ra properly when unwinding the stack commit bb9bc4689b9c635714fbcd5d335bad9934a7ebfc upstream. get_frame_info() calculates the offset of the return address within a stack frame simply by dividing a the bottom 16 bits of the instruction, treated as a signed integer, by the size of a long. Whilst this works for MIPS32 & MIPS64 ISAs where the sw or sd instructions are used, it's incorrect for microMIPS where encodings differ. The result is that we typically completely fail to unwind the stack on microMIPS. Fix this by adjusting is_ra_save_ins() to calculate the return address offset, and take into account the various different encodings there in the same place as we consider whether an instruction is storing the ra/$31 register. With this we are now able to unwind the stack for kernels targetting the microMIPS ISA, for example we can produce: Call Trace: [<80109e1f>] show_stack+0x63/0x7c [<8011ea17>] __warn+0x9b/0xac [<8011ea45>] warn_slowpath_fmt+0x1d/0x20 [<8013fe53>] register_console+0x43/0x314 [<8067c58d>] of_setup_earlycon+0x1dd/0x1ec [<8067f63f>] early_init_dt_scan_chosen_stdout+0xe7/0xf8 [<8066c115>] do_early_param+0x75/0xac [<801302f9>] parse_args+0x1dd/0x308 [<8066c459>] parse_early_options+0x25/0x28 [<8066c48b>] parse_early_param+0x2f/0x38 [<8066e8cf>] setup_arch+0x113/0x488 [<8066c4f3>] start_kernel+0x57/0x328 ---[ end trace 0000000000000000 ]--- Whereas previously we only produced: Call Trace: [<80109e1f>] show_stack+0x63/0x7c ---[ end trace 0000000000000000 ]--- Signed-off-by: Paul Burton <paul.burton@imgtec.com> Fixes: 34c2f668d0f6 ("MIPS: microMIPS: Add unaligned access support.") Cc: Leonid Yegoshin <leonid.yegoshin@imgtec.com> Cc: linux-mips@linux-mips.org Patchwork: https://patchwork.linux-mips.org/patch/14532/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2016-11-07 15:07:06 +00:00
if (info->pc_offset == -1 &&
is_ra_save_ins(&insn, &info->pc_offset))
break;
}
if (info->frame_size && info->pc_offset >= 0) /* nested */
return 0;
if (info->pc_offset < 0) /* leaf */
return 1;
/* prologue seems boggus... */
err:
return -1;
}
static struct mips_frame_info schedule_mfi __read_mostly;
#ifdef CONFIG_KALLSYMS
static unsigned long get___schedule_addr(void)
{
return kallsyms_lookup_name("__schedule");
}
#else
static unsigned long get___schedule_addr(void)
{
union mips_instruction *ip = (void *)schedule;
int max_insns = 8;
int i;
for (i = 0; i < max_insns; i++, ip++) {
if (ip->j_format.opcode == j_op)
return J_TARGET(ip, ip->j_format.target);
}
return 0;
}
#endif
static int __init frame_info_init(void)
{
unsigned long size = 0;
#ifdef CONFIG_KALLSYMS
unsigned long ofs;
#endif
unsigned long addr;
addr = get___schedule_addr();
if (!addr)
addr = (unsigned long)schedule;
#ifdef CONFIG_KALLSYMS
kallsyms_lookup_size_offset(addr, &size, &ofs);
#endif
schedule_mfi.func = (void *)addr;
schedule_mfi.func_size = size;
get_frame_info(&schedule_mfi);
/*
* Without schedule() frame info, result given by
* thread_saved_pc() and get_wchan() are not reliable.
*/
if (schedule_mfi.pc_offset < 0)
printk("Can't analyze schedule() prologue at %p\n", schedule);
return 0;
}
arch_initcall(frame_info_init);
/*
* Return saved PC of a blocked thread.
*/
unsigned long thread_saved_pc(struct task_struct *tsk)
{
struct thread_struct *t = &tsk->thread;
/* New born processes are a special case */
if (t->reg31 == (unsigned long) ret_from_fork)
return t->reg31;
if (schedule_mfi.pc_offset < 0)
return 0;
return ((unsigned long *)t->reg29)[schedule_mfi.pc_offset];
}
#ifdef CONFIG_KALLSYMS
/* generic stack unwinding function */
unsigned long notrace unwind_stack_by_address(unsigned long stack_page,
unsigned long *sp,
unsigned long pc,
unsigned long *ra)
{
MIPS: IRQ Stack: Unwind IRQ stack onto task stack [ Upstream commit db8466c581cca1a08b505f1319c3ecd246f16fa8 ] When the separate IRQ stack was introduced, stack unwinding only proceeded as far as the top of the IRQ stack, leading to kernel backtraces being less useful, lacking the trace of what was interrupted. Fix this by providing a means for the kernel to unwind the IRQ stack onto the interrupted task stack. The processor state is saved to the kernel task stack on interrupt. The IRQ_STACK_START macro reserves an unsigned long at the top of the IRQ stack where the interrupted task stack pointer can be saved. After the active stack is switched to the IRQ stack, save the interrupted tasks stack pointer to the reserved location. Fix the stack unwinding code to look for the frame being the top of the IRQ stack and if so get the next frame from the saved location. The existing test does not work with the separate stack since the ra is no longer pointed at ret_from_{irq,exception}. The test to stop unwinding the stack 32 bytes from the top of a stack must be modified to allow unwinding to continue up to the location of the saved task stack pointer when on the IRQ stack. The low / high marks of the stack are set depending on whether the sp is on an irq stack or not. Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: Masanari Iida <standby24x7@gmail.com> Cc: Chris Metcalf <cmetcalf@mellanox.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Paul Burton <paul.burton@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jason A. Donenfeld <jason@zx2c4.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/15788/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <alexander.levin@verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-03-21 14:52:25 +00:00
unsigned long low, high, irq_stack_high;
struct mips_frame_info info;
unsigned long size, ofs;
MIPS: IRQ Stack: Unwind IRQ stack onto task stack [ Upstream commit db8466c581cca1a08b505f1319c3ecd246f16fa8 ] When the separate IRQ stack was introduced, stack unwinding only proceeded as far as the top of the IRQ stack, leading to kernel backtraces being less useful, lacking the trace of what was interrupted. Fix this by providing a means for the kernel to unwind the IRQ stack onto the interrupted task stack. The processor state is saved to the kernel task stack on interrupt. The IRQ_STACK_START macro reserves an unsigned long at the top of the IRQ stack where the interrupted task stack pointer can be saved. After the active stack is switched to the IRQ stack, save the interrupted tasks stack pointer to the reserved location. Fix the stack unwinding code to look for the frame being the top of the IRQ stack and if so get the next frame from the saved location. The existing test does not work with the separate stack since the ra is no longer pointed at ret_from_{irq,exception}. The test to stop unwinding the stack 32 bytes from the top of a stack must be modified to allow unwinding to continue up to the location of the saved task stack pointer when on the IRQ stack. The low / high marks of the stack are set depending on whether the sp is on an irq stack or not. Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: Masanari Iida <standby24x7@gmail.com> Cc: Chris Metcalf <cmetcalf@mellanox.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Paul Burton <paul.burton@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jason A. Donenfeld <jason@zx2c4.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/15788/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <alexander.levin@verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-03-21 14:52:25 +00:00
struct pt_regs *regs;
int leaf;
if (!stack_page)
return 0;
/*
MIPS: IRQ Stack: Unwind IRQ stack onto task stack [ Upstream commit db8466c581cca1a08b505f1319c3ecd246f16fa8 ] When the separate IRQ stack was introduced, stack unwinding only proceeded as far as the top of the IRQ stack, leading to kernel backtraces being less useful, lacking the trace of what was interrupted. Fix this by providing a means for the kernel to unwind the IRQ stack onto the interrupted task stack. The processor state is saved to the kernel task stack on interrupt. The IRQ_STACK_START macro reserves an unsigned long at the top of the IRQ stack where the interrupted task stack pointer can be saved. After the active stack is switched to the IRQ stack, save the interrupted tasks stack pointer to the reserved location. Fix the stack unwinding code to look for the frame being the top of the IRQ stack and if so get the next frame from the saved location. The existing test does not work with the separate stack since the ra is no longer pointed at ret_from_{irq,exception}. The test to stop unwinding the stack 32 bytes from the top of a stack must be modified to allow unwinding to continue up to the location of the saved task stack pointer when on the IRQ stack. The low / high marks of the stack are set depending on whether the sp is on an irq stack or not. Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: Masanari Iida <standby24x7@gmail.com> Cc: Chris Metcalf <cmetcalf@mellanox.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Paul Burton <paul.burton@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jason A. Donenfeld <jason@zx2c4.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/15788/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <alexander.levin@verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-03-21 14:52:25 +00:00
* IRQ stacks start at IRQ_STACK_START
* task stacks at THREAD_SIZE - 32
*/
MIPS: IRQ Stack: Unwind IRQ stack onto task stack [ Upstream commit db8466c581cca1a08b505f1319c3ecd246f16fa8 ] When the separate IRQ stack was introduced, stack unwinding only proceeded as far as the top of the IRQ stack, leading to kernel backtraces being less useful, lacking the trace of what was interrupted. Fix this by providing a means for the kernel to unwind the IRQ stack onto the interrupted task stack. The processor state is saved to the kernel task stack on interrupt. The IRQ_STACK_START macro reserves an unsigned long at the top of the IRQ stack where the interrupted task stack pointer can be saved. After the active stack is switched to the IRQ stack, save the interrupted tasks stack pointer to the reserved location. Fix the stack unwinding code to look for the frame being the top of the IRQ stack and if so get the next frame from the saved location. The existing test does not work with the separate stack since the ra is no longer pointed at ret_from_{irq,exception}. The test to stop unwinding the stack 32 bytes from the top of a stack must be modified to allow unwinding to continue up to the location of the saved task stack pointer when on the IRQ stack. The low / high marks of the stack are set depending on whether the sp is on an irq stack or not. Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: Masanari Iida <standby24x7@gmail.com> Cc: Chris Metcalf <cmetcalf@mellanox.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Paul Burton <paul.burton@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jason A. Donenfeld <jason@zx2c4.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/15788/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <alexander.levin@verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-03-21 14:52:25 +00:00
low = stack_page;
if (!preemptible() && on_irq_stack(raw_smp_processor_id(), *sp)) {
high = stack_page + IRQ_STACK_START;
irq_stack_high = high;
} else {
high = stack_page + THREAD_SIZE - 32;
irq_stack_high = 0;
}
/*
* If we reached the top of the interrupt stack, start unwinding
* the interrupted task stack.
*/
if (unlikely(*sp == irq_stack_high)) {
unsigned long task_sp = *(unsigned long *)*sp;
/*
* Check that the pointer saved in the IRQ stack head points to
* something within the stack of the current task
*/
if (!object_is_on_stack((void *)task_sp))
return 0;
/*
* Follow pointer to tasks kernel stack frame where interrupted
* state was saved.
*/
regs = (struct pt_regs *)task_sp;
pc = regs->cp0_epc;
if (!user_mode(regs) && __kernel_text_address(pc)) {
*sp = regs->regs[29];
*ra = regs->regs[31];
return pc;
}
return 0;
}
if (!kallsyms_lookup_size_offset(pc, &size, &ofs))
return 0;
/*
* Return ra if an exception occurred at the first instruction
*/
if (unlikely(ofs == 0)) {
pc = *ra;
*ra = 0;
return pc;
}
info.func = (void *)(pc - ofs);
info.func_size = ofs; /* analyze from start to ofs */
leaf = get_frame_info(&info);
if (leaf < 0)
return 0;
MIPS: IRQ Stack: Unwind IRQ stack onto task stack [ Upstream commit db8466c581cca1a08b505f1319c3ecd246f16fa8 ] When the separate IRQ stack was introduced, stack unwinding only proceeded as far as the top of the IRQ stack, leading to kernel backtraces being less useful, lacking the trace of what was interrupted. Fix this by providing a means for the kernel to unwind the IRQ stack onto the interrupted task stack. The processor state is saved to the kernel task stack on interrupt. The IRQ_STACK_START macro reserves an unsigned long at the top of the IRQ stack where the interrupted task stack pointer can be saved. After the active stack is switched to the IRQ stack, save the interrupted tasks stack pointer to the reserved location. Fix the stack unwinding code to look for the frame being the top of the IRQ stack and if so get the next frame from the saved location. The existing test does not work with the separate stack since the ra is no longer pointed at ret_from_{irq,exception}. The test to stop unwinding the stack 32 bytes from the top of a stack must be modified to allow unwinding to continue up to the location of the saved task stack pointer when on the IRQ stack. The low / high marks of the stack are set depending on whether the sp is on an irq stack or not. Signed-off-by: Matt Redfearn <matt.redfearn@imgtec.com> Cc: Paolo Bonzini <pbonzini@redhat.com> Cc: Marcin Nowakowski <marcin.nowakowski@imgtec.com> Cc: Masanari Iida <standby24x7@gmail.com> Cc: Chris Metcalf <cmetcalf@mellanox.com> Cc: James Hogan <james.hogan@imgtec.com> Cc: Paul Burton <paul.burton@imgtec.com> Cc: Ingo Molnar <mingo@kernel.org> Cc: Jason A. Donenfeld <jason@zx2c4.com> Cc: Andrew Morton <akpm@linux-foundation.org> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/15788/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Sasha Levin <alexander.levin@verizon.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-03-21 14:52:25 +00:00
if (*sp < low || *sp + info.frame_size > high)
return 0;
if (leaf)
/*
* For some extreme cases, get_frame_info() can
* consider wrongly a nested function as a leaf
* one. In that cases avoid to return always the
* same value.
*/
pc = pc != *ra ? *ra : 0;
else
pc = ((unsigned long *)(*sp))[info.pc_offset];
*sp += info.frame_size;
*ra = 0;
return __kernel_text_address(pc) ? pc : 0;
}
EXPORT_SYMBOL(unwind_stack_by_address);
/* used by show_backtrace() */
unsigned long unwind_stack(struct task_struct *task, unsigned long *sp,
unsigned long pc, unsigned long *ra)
{
unsigned long stack_page = 0;
int cpu;
for_each_possible_cpu(cpu) {
if (on_irq_stack(cpu, *sp)) {
stack_page = (unsigned long)irq_stack[cpu];
break;
}
}
if (!stack_page)
stack_page = (unsigned long)task_stack_page(task);
return unwind_stack_by_address(stack_page, sp, pc, ra);
}
#endif
/*
* get_wchan - a maintenance nightmare^W^Wpain in the ass ...
*/
unsigned long get_wchan(struct task_struct *task)
{
unsigned long pc = 0;
#ifdef CONFIG_KALLSYMS
unsigned long sp;
unsigned long ra = 0;
#endif
if (!task || task == current || task->state == TASK_RUNNING)
goto out;
if (!task_stack_page(task))
goto out;
pc = thread_saved_pc(task);
#ifdef CONFIG_KALLSYMS
sp = task->thread.reg29 + schedule_mfi.frame_size;
while (in_sched_functions(pc))
pc = unwind_stack(task, &sp, pc, &ra);
#endif
out:
return pc;
}
/*
* Don't forget that the stack pointer must be aligned on a 8 bytes
* boundary for 32-bits ABI and 16 bytes for 64-bits ABI.
*/
unsigned long arch_align_stack(unsigned long sp)
{
if (!(current->personality & ADDR_NO_RANDOMIZE) && randomize_va_space)
sp -= get_random_int() & ~PAGE_MASK;
return sp & ALMASK;
}
MIPS: Use async IPIs for arch_trigger_cpumask_backtrace() commit b63e132b6433a41cf311e8bc382d33fd2b73b505 upstream. The current MIPS implementation of arch_trigger_cpumask_backtrace() is broken because it attempts to use synchronous IPIs despite the fact that it may be run with interrupts disabled. This means that when arch_trigger_cpumask_backtrace() is invoked, for example by the RCU CPU stall watchdog, we may: - Deadlock due to use of synchronous IPIs with interrupts disabled, causing the CPU that's attempting to generate the backtrace output to hang itself. - Not succeed in generating the desired output from remote CPUs. - Produce warnings about this from smp_call_function_many(), for example: [42760.526910] INFO: rcu_sched detected stalls on CPUs/tasks: [42760.535755] 0-...!: (1 GPs behind) idle=ade/140000000000000/0 softirq=526944/526945 fqs=0 [42760.547874] 1-...!: (0 ticks this GP) idle=e4a/140000000000000/0 softirq=547885/547885 fqs=0 [42760.559869] (detected by 2, t=2162 jiffies, g=266689, c=266688, q=33) [42760.568927] ------------[ cut here ]------------ [42760.576146] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:416 smp_call_function_many+0x88/0x20c [42760.587839] Modules linked in: [42760.593152] CPU: 2 PID: 1216 Comm: sh Not tainted 4.15.4-00373-gee058bb4d0c2 #2 [42760.603767] Stack : 8e09bd20 8e09bd20 8e09bd20 fffffff0 00000007 00000006 00000000 8e09bca8 [42760.616937] 95b2b379 95b2b379 807a0080 00000007 81944518 0000018a 00000032 00000000 [42760.630095] 00000000 00000030 80000000 00000000 806eca74 00000009 8017e2b8 000001a0 [42760.643169] 00000000 00000002 00000000 8e09baa4 00000008 808b8008 86d69080 8e09bca0 [42760.656282] 8e09ad50 805e20aa 00000000 00000000 00000000 8017e2b8 00000009 801070ca [42760.669424] ... [42760.673919] Call Trace: [42760.678672] [<27fde568>] show_stack+0x70/0xf0 [42760.685417] [<84751641>] dump_stack+0xaa/0xd0 [42760.692188] [<699d671c>] __warn+0x80/0x92 [42760.698549] [<68915d41>] warn_slowpath_null+0x28/0x36 [42760.705912] [<f7c76c1c>] smp_call_function_many+0x88/0x20c [42760.713696] [<6bbdfc2a>] arch_trigger_cpumask_backtrace+0x30/0x4a [42760.722216] [<f845bd33>] rcu_dump_cpu_stacks+0x6a/0x98 [42760.729580] [<796e7629>] rcu_check_callbacks+0x672/0x6ac [42760.737476] [<059b3b43>] update_process_times+0x18/0x34 [42760.744981] [<6eb94941>] tick_sched_handle.isra.5+0x26/0x38 [42760.752793] [<478d3d70>] tick_sched_timer+0x1c/0x50 [42760.759882] [<e56ea39f>] __hrtimer_run_queues+0xc6/0x226 [42760.767418] [<e88bbcae>] hrtimer_interrupt+0x88/0x19a [42760.775031] [<6765a19e>] gic_compare_interrupt+0x2e/0x3a [42760.782761] [<0558bf5f>] handle_percpu_devid_irq+0x78/0x168 [42760.790795] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.798117] [<1b6d462c>] gic_handle_local_int+0x38/0x86 [42760.805545] [<b2ada1c7>] gic_irq_dispatch+0xa/0x14 [42760.812534] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.820086] [<c7521934>] do_IRQ+0x16/0x20 [42760.826274] [<9aef3ce6>] plat_irq_dispatch+0x62/0x94 [42760.833458] [<6a94b53c>] except_vec_vi_end+0x70/0x78 [42760.840655] [<22284043>] smp_call_function_many+0x1ba/0x20c [42760.848501] [<54022b58>] smp_call_function+0x1e/0x2c [42760.855693] [<ab9fc705>] flush_tlb_mm+0x2a/0x98 [42760.862730] [<0844cdd0>] tlb_flush_mmu+0x1c/0x44 [42760.869628] [<cb259b74>] arch_tlb_finish_mmu+0x26/0x3e [42760.877021] [<1aeaaf74>] tlb_finish_mmu+0x18/0x66 [42760.883907] [<b3fce717>] exit_mmap+0x76/0xea [42760.890428] [<c4c8a2f6>] mmput+0x80/0x11a [42760.896632] [<a41a08f4>] do_exit+0x1f4/0x80c [42760.903158] [<ee01cef6>] do_group_exit+0x20/0x7e [42760.909990] [<13fa8d54>] __wake_up_parent+0x0/0x1e [42760.917045] [<46cf89d0>] smp_call_function_many+0x1a2/0x20c [42760.924893] [<8c21a93b>] syscall_common+0x14/0x1c [42760.931765] ---[ end trace 02aa09da9dc52a60 ]--- [42760.938342] ------------[ cut here ]------------ [42760.945311] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:291 smp_call_function_single+0xee/0xf8 ... This patch switches MIPS' arch_trigger_cpumask_backtrace() to use async IPIs & smp_call_function_single_async() in order to resolve this problem. We ensure use of the pre-allocated call_single_data_t structures is serialized by maintaining a cpumask indicating that they're busy, and refusing to attempt to send an IPI when a CPU's bit is set in this mask. This should only happen if a CPU hasn't responded to a previous backtrace IPI - ie. if it's hung - and we print a warning to the console in this case. I've marked this for stable branches as far back as v4.9, to which it applies cleanly. Strictly speaking the faulty MIPS implementation can be traced further back to commit 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") in v3.19, but kernel versions v3.19 through v4.8 will require further work to backport due to the rework performed in commit 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods"). Signed-off-by: Paul Burton <paul.burton@mips.com> Patchwork: https://patchwork.linux-mips.org/patch/19597/ Cc: James Hogan <jhogan@kernel.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Huacai Chen <chenhc@lemote.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # v4.9+ Fixes: 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") Fixes: 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods") [ Huacai: backported to 4.4: Restruction since generic NMI solution is unavailable ] Signed-off-by: Huacai Chen <chenhc@lemote.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-06-22 10:55:46 -07:00
static DEFINE_PER_CPU(struct call_single_data, backtrace_csd);
static struct cpumask backtrace_csd_busy;
static void arch_dump_stack(void *info)
{
struct pt_regs *regs;
MIPS: Use async IPIs for arch_trigger_cpumask_backtrace() commit b63e132b6433a41cf311e8bc382d33fd2b73b505 upstream. The current MIPS implementation of arch_trigger_cpumask_backtrace() is broken because it attempts to use synchronous IPIs despite the fact that it may be run with interrupts disabled. This means that when arch_trigger_cpumask_backtrace() is invoked, for example by the RCU CPU stall watchdog, we may: - Deadlock due to use of synchronous IPIs with interrupts disabled, causing the CPU that's attempting to generate the backtrace output to hang itself. - Not succeed in generating the desired output from remote CPUs. - Produce warnings about this from smp_call_function_many(), for example: [42760.526910] INFO: rcu_sched detected stalls on CPUs/tasks: [42760.535755] 0-...!: (1 GPs behind) idle=ade/140000000000000/0 softirq=526944/526945 fqs=0 [42760.547874] 1-...!: (0 ticks this GP) idle=e4a/140000000000000/0 softirq=547885/547885 fqs=0 [42760.559869] (detected by 2, t=2162 jiffies, g=266689, c=266688, q=33) [42760.568927] ------------[ cut here ]------------ [42760.576146] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:416 smp_call_function_many+0x88/0x20c [42760.587839] Modules linked in: [42760.593152] CPU: 2 PID: 1216 Comm: sh Not tainted 4.15.4-00373-gee058bb4d0c2 #2 [42760.603767] Stack : 8e09bd20 8e09bd20 8e09bd20 fffffff0 00000007 00000006 00000000 8e09bca8 [42760.616937] 95b2b379 95b2b379 807a0080 00000007 81944518 0000018a 00000032 00000000 [42760.630095] 00000000 00000030 80000000 00000000 806eca74 00000009 8017e2b8 000001a0 [42760.643169] 00000000 00000002 00000000 8e09baa4 00000008 808b8008 86d69080 8e09bca0 [42760.656282] 8e09ad50 805e20aa 00000000 00000000 00000000 8017e2b8 00000009 801070ca [42760.669424] ... [42760.673919] Call Trace: [42760.678672] [<27fde568>] show_stack+0x70/0xf0 [42760.685417] [<84751641>] dump_stack+0xaa/0xd0 [42760.692188] [<699d671c>] __warn+0x80/0x92 [42760.698549] [<68915d41>] warn_slowpath_null+0x28/0x36 [42760.705912] [<f7c76c1c>] smp_call_function_many+0x88/0x20c [42760.713696] [<6bbdfc2a>] arch_trigger_cpumask_backtrace+0x30/0x4a [42760.722216] [<f845bd33>] rcu_dump_cpu_stacks+0x6a/0x98 [42760.729580] [<796e7629>] rcu_check_callbacks+0x672/0x6ac [42760.737476] [<059b3b43>] update_process_times+0x18/0x34 [42760.744981] [<6eb94941>] tick_sched_handle.isra.5+0x26/0x38 [42760.752793] [<478d3d70>] tick_sched_timer+0x1c/0x50 [42760.759882] [<e56ea39f>] __hrtimer_run_queues+0xc6/0x226 [42760.767418] [<e88bbcae>] hrtimer_interrupt+0x88/0x19a [42760.775031] [<6765a19e>] gic_compare_interrupt+0x2e/0x3a [42760.782761] [<0558bf5f>] handle_percpu_devid_irq+0x78/0x168 [42760.790795] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.798117] [<1b6d462c>] gic_handle_local_int+0x38/0x86 [42760.805545] [<b2ada1c7>] gic_irq_dispatch+0xa/0x14 [42760.812534] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.820086] [<c7521934>] do_IRQ+0x16/0x20 [42760.826274] [<9aef3ce6>] plat_irq_dispatch+0x62/0x94 [42760.833458] [<6a94b53c>] except_vec_vi_end+0x70/0x78 [42760.840655] [<22284043>] smp_call_function_many+0x1ba/0x20c [42760.848501] [<54022b58>] smp_call_function+0x1e/0x2c [42760.855693] [<ab9fc705>] flush_tlb_mm+0x2a/0x98 [42760.862730] [<0844cdd0>] tlb_flush_mmu+0x1c/0x44 [42760.869628] [<cb259b74>] arch_tlb_finish_mmu+0x26/0x3e [42760.877021] [<1aeaaf74>] tlb_finish_mmu+0x18/0x66 [42760.883907] [<b3fce717>] exit_mmap+0x76/0xea [42760.890428] [<c4c8a2f6>] mmput+0x80/0x11a [42760.896632] [<a41a08f4>] do_exit+0x1f4/0x80c [42760.903158] [<ee01cef6>] do_group_exit+0x20/0x7e [42760.909990] [<13fa8d54>] __wake_up_parent+0x0/0x1e [42760.917045] [<46cf89d0>] smp_call_function_many+0x1a2/0x20c [42760.924893] [<8c21a93b>] syscall_common+0x14/0x1c [42760.931765] ---[ end trace 02aa09da9dc52a60 ]--- [42760.938342] ------------[ cut here ]------------ [42760.945311] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:291 smp_call_function_single+0xee/0xf8 ... This patch switches MIPS' arch_trigger_cpumask_backtrace() to use async IPIs & smp_call_function_single_async() in order to resolve this problem. We ensure use of the pre-allocated call_single_data_t structures is serialized by maintaining a cpumask indicating that they're busy, and refusing to attempt to send an IPI when a CPU's bit is set in this mask. This should only happen if a CPU hasn't responded to a previous backtrace IPI - ie. if it's hung - and we print a warning to the console in this case. I've marked this for stable branches as far back as v4.9, to which it applies cleanly. Strictly speaking the faulty MIPS implementation can be traced further back to commit 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") in v3.19, but kernel versions v3.19 through v4.8 will require further work to backport due to the rework performed in commit 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods"). Signed-off-by: Paul Burton <paul.burton@mips.com> Patchwork: https://patchwork.linux-mips.org/patch/19597/ Cc: James Hogan <jhogan@kernel.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Huacai Chen <chenhc@lemote.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # v4.9+ Fixes: 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") Fixes: 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods") [ Huacai: backported to 4.4: Restruction since generic NMI solution is unavailable ] Signed-off-by: Huacai Chen <chenhc@lemote.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-06-22 10:55:46 -07:00
static arch_spinlock_t lock = __ARCH_SPIN_LOCK_UNLOCKED;
MIPS: Use async IPIs for arch_trigger_cpumask_backtrace() commit b63e132b6433a41cf311e8bc382d33fd2b73b505 upstream. The current MIPS implementation of arch_trigger_cpumask_backtrace() is broken because it attempts to use synchronous IPIs despite the fact that it may be run with interrupts disabled. This means that when arch_trigger_cpumask_backtrace() is invoked, for example by the RCU CPU stall watchdog, we may: - Deadlock due to use of synchronous IPIs with interrupts disabled, causing the CPU that's attempting to generate the backtrace output to hang itself. - Not succeed in generating the desired output from remote CPUs. - Produce warnings about this from smp_call_function_many(), for example: [42760.526910] INFO: rcu_sched detected stalls on CPUs/tasks: [42760.535755] 0-...!: (1 GPs behind) idle=ade/140000000000000/0 softirq=526944/526945 fqs=0 [42760.547874] 1-...!: (0 ticks this GP) idle=e4a/140000000000000/0 softirq=547885/547885 fqs=0 [42760.559869] (detected by 2, t=2162 jiffies, g=266689, c=266688, q=33) [42760.568927] ------------[ cut here ]------------ [42760.576146] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:416 smp_call_function_many+0x88/0x20c [42760.587839] Modules linked in: [42760.593152] CPU: 2 PID: 1216 Comm: sh Not tainted 4.15.4-00373-gee058bb4d0c2 #2 [42760.603767] Stack : 8e09bd20 8e09bd20 8e09bd20 fffffff0 00000007 00000006 00000000 8e09bca8 [42760.616937] 95b2b379 95b2b379 807a0080 00000007 81944518 0000018a 00000032 00000000 [42760.630095] 00000000 00000030 80000000 00000000 806eca74 00000009 8017e2b8 000001a0 [42760.643169] 00000000 00000002 00000000 8e09baa4 00000008 808b8008 86d69080 8e09bca0 [42760.656282] 8e09ad50 805e20aa 00000000 00000000 00000000 8017e2b8 00000009 801070ca [42760.669424] ... [42760.673919] Call Trace: [42760.678672] [<27fde568>] show_stack+0x70/0xf0 [42760.685417] [<84751641>] dump_stack+0xaa/0xd0 [42760.692188] [<699d671c>] __warn+0x80/0x92 [42760.698549] [<68915d41>] warn_slowpath_null+0x28/0x36 [42760.705912] [<f7c76c1c>] smp_call_function_many+0x88/0x20c [42760.713696] [<6bbdfc2a>] arch_trigger_cpumask_backtrace+0x30/0x4a [42760.722216] [<f845bd33>] rcu_dump_cpu_stacks+0x6a/0x98 [42760.729580] [<796e7629>] rcu_check_callbacks+0x672/0x6ac [42760.737476] [<059b3b43>] update_process_times+0x18/0x34 [42760.744981] [<6eb94941>] tick_sched_handle.isra.5+0x26/0x38 [42760.752793] [<478d3d70>] tick_sched_timer+0x1c/0x50 [42760.759882] [<e56ea39f>] __hrtimer_run_queues+0xc6/0x226 [42760.767418] [<e88bbcae>] hrtimer_interrupt+0x88/0x19a [42760.775031] [<6765a19e>] gic_compare_interrupt+0x2e/0x3a [42760.782761] [<0558bf5f>] handle_percpu_devid_irq+0x78/0x168 [42760.790795] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.798117] [<1b6d462c>] gic_handle_local_int+0x38/0x86 [42760.805545] [<b2ada1c7>] gic_irq_dispatch+0xa/0x14 [42760.812534] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.820086] [<c7521934>] do_IRQ+0x16/0x20 [42760.826274] [<9aef3ce6>] plat_irq_dispatch+0x62/0x94 [42760.833458] [<6a94b53c>] except_vec_vi_end+0x70/0x78 [42760.840655] [<22284043>] smp_call_function_many+0x1ba/0x20c [42760.848501] [<54022b58>] smp_call_function+0x1e/0x2c [42760.855693] [<ab9fc705>] flush_tlb_mm+0x2a/0x98 [42760.862730] [<0844cdd0>] tlb_flush_mmu+0x1c/0x44 [42760.869628] [<cb259b74>] arch_tlb_finish_mmu+0x26/0x3e [42760.877021] [<1aeaaf74>] tlb_finish_mmu+0x18/0x66 [42760.883907] [<b3fce717>] exit_mmap+0x76/0xea [42760.890428] [<c4c8a2f6>] mmput+0x80/0x11a [42760.896632] [<a41a08f4>] do_exit+0x1f4/0x80c [42760.903158] [<ee01cef6>] do_group_exit+0x20/0x7e [42760.909990] [<13fa8d54>] __wake_up_parent+0x0/0x1e [42760.917045] [<46cf89d0>] smp_call_function_many+0x1a2/0x20c [42760.924893] [<8c21a93b>] syscall_common+0x14/0x1c [42760.931765] ---[ end trace 02aa09da9dc52a60 ]--- [42760.938342] ------------[ cut here ]------------ [42760.945311] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:291 smp_call_function_single+0xee/0xf8 ... This patch switches MIPS' arch_trigger_cpumask_backtrace() to use async IPIs & smp_call_function_single_async() in order to resolve this problem. We ensure use of the pre-allocated call_single_data_t structures is serialized by maintaining a cpumask indicating that they're busy, and refusing to attempt to send an IPI when a CPU's bit is set in this mask. This should only happen if a CPU hasn't responded to a previous backtrace IPI - ie. if it's hung - and we print a warning to the console in this case. I've marked this for stable branches as far back as v4.9, to which it applies cleanly. Strictly speaking the faulty MIPS implementation can be traced further back to commit 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") in v3.19, but kernel versions v3.19 through v4.8 will require further work to backport due to the rework performed in commit 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods"). Signed-off-by: Paul Burton <paul.burton@mips.com> Patchwork: https://patchwork.linux-mips.org/patch/19597/ Cc: James Hogan <jhogan@kernel.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Huacai Chen <chenhc@lemote.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # v4.9+ Fixes: 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") Fixes: 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods") [ Huacai: backported to 4.4: Restruction since generic NMI solution is unavailable ] Signed-off-by: Huacai Chen <chenhc@lemote.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-06-22 10:55:46 -07:00
arch_spin_lock(&lock);
regs = get_irq_regs();
if (regs)
show_regs(regs);
MIPS: Call dump_stack() from show_regs() commit 5a267832c2ec47b2dad0fdb291a96bb5b8869315 upstream. The generic nmi_cpu_backtrace() function calls show_regs() when a struct pt_regs is available, and dump_stack() otherwise. If we were to make use of the generic nmi_cpu_backtrace() with MIPS' current implementation of show_regs() this would mean that we see only register data with no accompanying stack information, in contrast with our current implementation which calls dump_stack() regardless of whether register state is available. In preparation for making use of the generic nmi_cpu_backtrace() to implement arch_trigger_cpumask_backtrace(), have our implementation of show_regs() call dump_stack() and drop the explicit dump_stack() call in arch_dump_stack() which is invoked by arch_trigger_cpumask_backtrace(). This will allow the output we produce to remain the same after a later patch switches to using nmi_cpu_backtrace(). It may mean that we produce extra stack output in other uses of show_regs(), but this: 1) Seems harmless. 2) Is good for consistency between arch_trigger_cpumask_backtrace() and other users of show_regs(). 3) Matches the behaviour of the ARM & PowerPC architectures. Marked for stable back to v4.9 as a prerequisite of the following patch "MIPS: Call dump_stack() from show_regs()". Signed-off-by: Paul Burton <paul.burton@mips.com> Patchwork: https://patchwork.linux-mips.org/patch/19596/ Cc: James Hogan <jhogan@kernel.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Huacai Chen <chenhc@lemote.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # v4.9+ Signed-off-by: Huacai Chen <chenhc@lemote.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-06-22 10:55:45 -07:00
else
dump_stack();
MIPS: Use async IPIs for arch_trigger_cpumask_backtrace() commit b63e132b6433a41cf311e8bc382d33fd2b73b505 upstream. The current MIPS implementation of arch_trigger_cpumask_backtrace() is broken because it attempts to use synchronous IPIs despite the fact that it may be run with interrupts disabled. This means that when arch_trigger_cpumask_backtrace() is invoked, for example by the RCU CPU stall watchdog, we may: - Deadlock due to use of synchronous IPIs with interrupts disabled, causing the CPU that's attempting to generate the backtrace output to hang itself. - Not succeed in generating the desired output from remote CPUs. - Produce warnings about this from smp_call_function_many(), for example: [42760.526910] INFO: rcu_sched detected stalls on CPUs/tasks: [42760.535755] 0-...!: (1 GPs behind) idle=ade/140000000000000/0 softirq=526944/526945 fqs=0 [42760.547874] 1-...!: (0 ticks this GP) idle=e4a/140000000000000/0 softirq=547885/547885 fqs=0 [42760.559869] (detected by 2, t=2162 jiffies, g=266689, c=266688, q=33) [42760.568927] ------------[ cut here ]------------ [42760.576146] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:416 smp_call_function_many+0x88/0x20c [42760.587839] Modules linked in: [42760.593152] CPU: 2 PID: 1216 Comm: sh Not tainted 4.15.4-00373-gee058bb4d0c2 #2 [42760.603767] Stack : 8e09bd20 8e09bd20 8e09bd20 fffffff0 00000007 00000006 00000000 8e09bca8 [42760.616937] 95b2b379 95b2b379 807a0080 00000007 81944518 0000018a 00000032 00000000 [42760.630095] 00000000 00000030 80000000 00000000 806eca74 00000009 8017e2b8 000001a0 [42760.643169] 00000000 00000002 00000000 8e09baa4 00000008 808b8008 86d69080 8e09bca0 [42760.656282] 8e09ad50 805e20aa 00000000 00000000 00000000 8017e2b8 00000009 801070ca [42760.669424] ... [42760.673919] Call Trace: [42760.678672] [<27fde568>] show_stack+0x70/0xf0 [42760.685417] [<84751641>] dump_stack+0xaa/0xd0 [42760.692188] [<699d671c>] __warn+0x80/0x92 [42760.698549] [<68915d41>] warn_slowpath_null+0x28/0x36 [42760.705912] [<f7c76c1c>] smp_call_function_many+0x88/0x20c [42760.713696] [<6bbdfc2a>] arch_trigger_cpumask_backtrace+0x30/0x4a [42760.722216] [<f845bd33>] rcu_dump_cpu_stacks+0x6a/0x98 [42760.729580] [<796e7629>] rcu_check_callbacks+0x672/0x6ac [42760.737476] [<059b3b43>] update_process_times+0x18/0x34 [42760.744981] [<6eb94941>] tick_sched_handle.isra.5+0x26/0x38 [42760.752793] [<478d3d70>] tick_sched_timer+0x1c/0x50 [42760.759882] [<e56ea39f>] __hrtimer_run_queues+0xc6/0x226 [42760.767418] [<e88bbcae>] hrtimer_interrupt+0x88/0x19a [42760.775031] [<6765a19e>] gic_compare_interrupt+0x2e/0x3a [42760.782761] [<0558bf5f>] handle_percpu_devid_irq+0x78/0x168 [42760.790795] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.798117] [<1b6d462c>] gic_handle_local_int+0x38/0x86 [42760.805545] [<b2ada1c7>] gic_irq_dispatch+0xa/0x14 [42760.812534] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.820086] [<c7521934>] do_IRQ+0x16/0x20 [42760.826274] [<9aef3ce6>] plat_irq_dispatch+0x62/0x94 [42760.833458] [<6a94b53c>] except_vec_vi_end+0x70/0x78 [42760.840655] [<22284043>] smp_call_function_many+0x1ba/0x20c [42760.848501] [<54022b58>] smp_call_function+0x1e/0x2c [42760.855693] [<ab9fc705>] flush_tlb_mm+0x2a/0x98 [42760.862730] [<0844cdd0>] tlb_flush_mmu+0x1c/0x44 [42760.869628] [<cb259b74>] arch_tlb_finish_mmu+0x26/0x3e [42760.877021] [<1aeaaf74>] tlb_finish_mmu+0x18/0x66 [42760.883907] [<b3fce717>] exit_mmap+0x76/0xea [42760.890428] [<c4c8a2f6>] mmput+0x80/0x11a [42760.896632] [<a41a08f4>] do_exit+0x1f4/0x80c [42760.903158] [<ee01cef6>] do_group_exit+0x20/0x7e [42760.909990] [<13fa8d54>] __wake_up_parent+0x0/0x1e [42760.917045] [<46cf89d0>] smp_call_function_many+0x1a2/0x20c [42760.924893] [<8c21a93b>] syscall_common+0x14/0x1c [42760.931765] ---[ end trace 02aa09da9dc52a60 ]--- [42760.938342] ------------[ cut here ]------------ [42760.945311] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:291 smp_call_function_single+0xee/0xf8 ... This patch switches MIPS' arch_trigger_cpumask_backtrace() to use async IPIs & smp_call_function_single_async() in order to resolve this problem. We ensure use of the pre-allocated call_single_data_t structures is serialized by maintaining a cpumask indicating that they're busy, and refusing to attempt to send an IPI when a CPU's bit is set in this mask. This should only happen if a CPU hasn't responded to a previous backtrace IPI - ie. if it's hung - and we print a warning to the console in this case. I've marked this for stable branches as far back as v4.9, to which it applies cleanly. Strictly speaking the faulty MIPS implementation can be traced further back to commit 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") in v3.19, but kernel versions v3.19 through v4.8 will require further work to backport due to the rework performed in commit 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods"). Signed-off-by: Paul Burton <paul.burton@mips.com> Patchwork: https://patchwork.linux-mips.org/patch/19597/ Cc: James Hogan <jhogan@kernel.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Huacai Chen <chenhc@lemote.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # v4.9+ Fixes: 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") Fixes: 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods") [ Huacai: backported to 4.4: Restruction since generic NMI solution is unavailable ] Signed-off-by: Huacai Chen <chenhc@lemote.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-06-22 10:55:46 -07:00
arch_spin_unlock(&lock);
cpumask_clear_cpu(smp_processor_id(), &backtrace_csd_busy);
}
void arch_trigger_all_cpu_backtrace(bool include_self)
{
MIPS: Use async IPIs for arch_trigger_cpumask_backtrace() commit b63e132b6433a41cf311e8bc382d33fd2b73b505 upstream. The current MIPS implementation of arch_trigger_cpumask_backtrace() is broken because it attempts to use synchronous IPIs despite the fact that it may be run with interrupts disabled. This means that when arch_trigger_cpumask_backtrace() is invoked, for example by the RCU CPU stall watchdog, we may: - Deadlock due to use of synchronous IPIs with interrupts disabled, causing the CPU that's attempting to generate the backtrace output to hang itself. - Not succeed in generating the desired output from remote CPUs. - Produce warnings about this from smp_call_function_many(), for example: [42760.526910] INFO: rcu_sched detected stalls on CPUs/tasks: [42760.535755] 0-...!: (1 GPs behind) idle=ade/140000000000000/0 softirq=526944/526945 fqs=0 [42760.547874] 1-...!: (0 ticks this GP) idle=e4a/140000000000000/0 softirq=547885/547885 fqs=0 [42760.559869] (detected by 2, t=2162 jiffies, g=266689, c=266688, q=33) [42760.568927] ------------[ cut here ]------------ [42760.576146] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:416 smp_call_function_many+0x88/0x20c [42760.587839] Modules linked in: [42760.593152] CPU: 2 PID: 1216 Comm: sh Not tainted 4.15.4-00373-gee058bb4d0c2 #2 [42760.603767] Stack : 8e09bd20 8e09bd20 8e09bd20 fffffff0 00000007 00000006 00000000 8e09bca8 [42760.616937] 95b2b379 95b2b379 807a0080 00000007 81944518 0000018a 00000032 00000000 [42760.630095] 00000000 00000030 80000000 00000000 806eca74 00000009 8017e2b8 000001a0 [42760.643169] 00000000 00000002 00000000 8e09baa4 00000008 808b8008 86d69080 8e09bca0 [42760.656282] 8e09ad50 805e20aa 00000000 00000000 00000000 8017e2b8 00000009 801070ca [42760.669424] ... [42760.673919] Call Trace: [42760.678672] [<27fde568>] show_stack+0x70/0xf0 [42760.685417] [<84751641>] dump_stack+0xaa/0xd0 [42760.692188] [<699d671c>] __warn+0x80/0x92 [42760.698549] [<68915d41>] warn_slowpath_null+0x28/0x36 [42760.705912] [<f7c76c1c>] smp_call_function_many+0x88/0x20c [42760.713696] [<6bbdfc2a>] arch_trigger_cpumask_backtrace+0x30/0x4a [42760.722216] [<f845bd33>] rcu_dump_cpu_stacks+0x6a/0x98 [42760.729580] [<796e7629>] rcu_check_callbacks+0x672/0x6ac [42760.737476] [<059b3b43>] update_process_times+0x18/0x34 [42760.744981] [<6eb94941>] tick_sched_handle.isra.5+0x26/0x38 [42760.752793] [<478d3d70>] tick_sched_timer+0x1c/0x50 [42760.759882] [<e56ea39f>] __hrtimer_run_queues+0xc6/0x226 [42760.767418] [<e88bbcae>] hrtimer_interrupt+0x88/0x19a [42760.775031] [<6765a19e>] gic_compare_interrupt+0x2e/0x3a [42760.782761] [<0558bf5f>] handle_percpu_devid_irq+0x78/0x168 [42760.790795] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.798117] [<1b6d462c>] gic_handle_local_int+0x38/0x86 [42760.805545] [<b2ada1c7>] gic_irq_dispatch+0xa/0x14 [42760.812534] [<90c11ba2>] generic_handle_irq+0x1e/0x2c [42760.820086] [<c7521934>] do_IRQ+0x16/0x20 [42760.826274] [<9aef3ce6>] plat_irq_dispatch+0x62/0x94 [42760.833458] [<6a94b53c>] except_vec_vi_end+0x70/0x78 [42760.840655] [<22284043>] smp_call_function_many+0x1ba/0x20c [42760.848501] [<54022b58>] smp_call_function+0x1e/0x2c [42760.855693] [<ab9fc705>] flush_tlb_mm+0x2a/0x98 [42760.862730] [<0844cdd0>] tlb_flush_mmu+0x1c/0x44 [42760.869628] [<cb259b74>] arch_tlb_finish_mmu+0x26/0x3e [42760.877021] [<1aeaaf74>] tlb_finish_mmu+0x18/0x66 [42760.883907] [<b3fce717>] exit_mmap+0x76/0xea [42760.890428] [<c4c8a2f6>] mmput+0x80/0x11a [42760.896632] [<a41a08f4>] do_exit+0x1f4/0x80c [42760.903158] [<ee01cef6>] do_group_exit+0x20/0x7e [42760.909990] [<13fa8d54>] __wake_up_parent+0x0/0x1e [42760.917045] [<46cf89d0>] smp_call_function_many+0x1a2/0x20c [42760.924893] [<8c21a93b>] syscall_common+0x14/0x1c [42760.931765] ---[ end trace 02aa09da9dc52a60 ]--- [42760.938342] ------------[ cut here ]------------ [42760.945311] WARNING: CPU: 2 PID: 1216 at kernel/smp.c:291 smp_call_function_single+0xee/0xf8 ... This patch switches MIPS' arch_trigger_cpumask_backtrace() to use async IPIs & smp_call_function_single_async() in order to resolve this problem. We ensure use of the pre-allocated call_single_data_t structures is serialized by maintaining a cpumask indicating that they're busy, and refusing to attempt to send an IPI when a CPU's bit is set in this mask. This should only happen if a CPU hasn't responded to a previous backtrace IPI - ie. if it's hung - and we print a warning to the console in this case. I've marked this for stable branches as far back as v4.9, to which it applies cleanly. Strictly speaking the faulty MIPS implementation can be traced further back to commit 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") in v3.19, but kernel versions v3.19 through v4.8 will require further work to backport due to the rework performed in commit 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods"). Signed-off-by: Paul Burton <paul.burton@mips.com> Patchwork: https://patchwork.linux-mips.org/patch/19597/ Cc: James Hogan <jhogan@kernel.org> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: Huacai Chen <chenhc@lemote.com> Cc: linux-mips@linux-mips.org Cc: stable@vger.kernel.org # v4.9+ Fixes: 856839b76836 ("MIPS: Add arch_trigger_all_cpu_backtrace() function") Fixes: 9a01c3ed5cdb ("nmi_backtrace: add more trigger_*_cpu_backtrace() methods") [ Huacai: backported to 4.4: Restruction since generic NMI solution is unavailable ] Signed-off-by: Huacai Chen <chenhc@lemote.com> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-06-22 10:55:46 -07:00
struct call_single_data *csd;
int cpu;
for_each_cpu(cpu, cpu_online_mask) {
/*
* If we previously sent an IPI to the target CPU & it hasn't
* cleared its bit in the busy cpumask then it didn't handle
* our previous IPI & it's not safe for us to reuse the
* call_single_data_t.
*/
if (cpumask_test_and_set_cpu(cpu, &backtrace_csd_busy)) {
pr_warn("Unable to send backtrace IPI to CPU%u - perhaps it hung?\n",
cpu);
continue;
}
csd = &per_cpu(backtrace_csd, cpu);
csd->func = arch_dump_stack;
smp_call_function_single_async(cpu, csd);
}
}
2015-01-08 12:17:37 +00:00
int mips_get_process_fp_mode(struct task_struct *task)
{
int value = 0;
if (!test_tsk_thread_flag(task, TIF_32BIT_FPREGS))
value |= PR_FP_MODE_FR;
if (test_tsk_thread_flag(task, TIF_HYBRID_FPREGS))
value |= PR_FP_MODE_FRE;
return value;
}
int mips_set_process_fp_mode(struct task_struct *task, unsigned int value)
{
const unsigned int known_bits = PR_FP_MODE_FR | PR_FP_MODE_FRE;
unsigned long switch_count;
struct task_struct *t;
MIPS: Validate PR_SET_FP_MODE prctl(2) requests against the ABI of the task commit b67336eee3fcb8ecedc6c13e2bf88aacfa3151e2 upstream. Fix an API loophole introduced with commit 9791554b45a2 ("MIPS,prctl: add PR_[GS]ET_FP_MODE prctl options for MIPS"), where the caller of prctl(2) is incorrectly allowed to make a change to CP0.Status.FR or CP0.Config5.FRE register bits even if CONFIG_MIPS_O32_FP64_SUPPORT has not been enabled, despite that an executable requesting the mode requested via ELF file annotation would not be allowed to run in the first place, or for n64 and n64 ABI tasks which do not have non-default modes defined at all. Add suitable checks to `mips_set_process_fp_mode' and bail out if an invalid mode change has been requested for the ABI in effect, even if the FPU hardware or emulation would otherwise allow it. Always succeed however without taking any further action if the mode requested is the same as one already in effect, regardless of whether any mode change, should it be requested, would actually be allowed for the task concerned. Signed-off-by: Maciej W. Rozycki <macro@mips.com> Fixes: 9791554b45a2 ("MIPS,prctl: add PR_[GS]ET_FP_MODE prctl options for MIPS") Reviewed-by: Paul Burton <paul.burton@mips.com> Cc: James Hogan <james.hogan@mips.com> Cc: linux-mips@linux-mips.org Cc: linux-kernel@vger.kernel.org Patchwork: https://patchwork.linux-mips.org/patch/17800/ Signed-off-by: Ralf Baechle <ralf@linux-mips.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2017-11-27 09:33:03 +00:00
/* If nothing to change, return right away, successfully. */
if (value == mips_get_process_fp_mode(task))
return 0;
/* Only accept a mode change if 64-bit FP enabled for o32. */
if (!IS_ENABLED(CONFIG_MIPS_O32_FP64_SUPPORT))
return -EOPNOTSUPP;
/* And only for o32 tasks. */
if (IS_ENABLED(CONFIG_64BIT) && !test_thread_flag(TIF_32BIT_REGS))
return -EOPNOTSUPP;
2015-01-08 12:17:37 +00:00
/* Check the value is valid */
if (value & ~known_bits)
return -EOPNOTSUPP;
MIPS: prctl: Disallow FRE without FR with PR_SET_FP_MODE requests commit 28e4213dd331e944e7fca1954a946829162ed9d4 upstream. Having PR_FP_MODE_FRE (i.e. Config5.FRE) set without PR_FP_MODE_FR (i.e. Status.FR) is not supported as the lone purpose of Config5.FRE is to emulate Status.FR=0 handling on FPU hardware that has Status.FR=1 hardwired[1][2]. Also we do not handle this case elsewhere, and assume throughout our code that TIF_HYBRID_FPREGS and TIF_32BIT_FPREGS cannot be set both at once for a task, leading to inconsistent behaviour if this does happen. Return unsuccessfully then from prctl(2) PR_SET_FP_MODE calls requesting PR_FP_MODE_FRE to be set with PR_FP_MODE_FR clear. This corresponds to modes allowed by `mips_set_personality_fp'. References: [1] "MIPS Architecture For Programmers, Vol. III: MIPS32 / microMIPS32 Privileged Resource Architecture", Imagination Technologies, Document Number: MD00090, Revision 6.02, July 10, 2015, Table 9.69 "Config5 Register Field Descriptions", p. 262 [2] "MIPS Architecture For Programmers, Volume III: MIPS64 / microMIPS64 Privileged Resource Architecture", Imagination Technologies, Document Number: MD00091, Revision 6.03, December 22, 2015, Table 9.72 "Config5 Register Field Descriptions", p. 288 Fixes: 9791554b45a2 ("MIPS,prctl: add PR_[GS]ET_FP_MODE prctl options for MIPS") Signed-off-by: Maciej W. Rozycki <macro@mips.com> Cc: Ralf Baechle <ralf@linux-mips.org> Cc: linux-mips@linux-mips.org Cc: <stable@vger.kernel.org> # 4.0+ Patchwork: https://patchwork.linux-mips.org/patch/19327/ Signed-off-by: James Hogan <jhogan@kernel.org> Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-05-15 23:04:44 +01:00
/* Setting FRE without FR is not supported. */
if ((value & (PR_FP_MODE_FR | PR_FP_MODE_FRE)) == PR_FP_MODE_FRE)
return -EOPNOTSUPP;
2015-01-08 12:17:37 +00:00
/* Avoid inadvertently triggering emulation */
if ((value & PR_FP_MODE_FR) && raw_cpu_has_fpu &&
!(raw_current_cpu_data.fpu_id & MIPS_FPIR_F64))
2015-01-08 12:17:37 +00:00
return -EOPNOTSUPP;
if ((value & PR_FP_MODE_FRE) && raw_cpu_has_fpu && !cpu_has_fre)
2015-01-08 12:17:37 +00:00
return -EOPNOTSUPP;
/* FR = 0 not supported in MIPS R6 */
if (!(value & PR_FP_MODE_FR) && raw_cpu_has_fpu && cpu_has_mips_r6)
return -EOPNOTSUPP;
/* Proceed with the mode switch */
preempt_disable();
2015-01-08 12:17:37 +00:00
/* Save FP & vector context, then disable FPU & MSA */
if (task->signal == current->signal)
lose_fpu(1);
/* Prevent any threads from obtaining live FP context */
atomic_set(&task->mm->context.fp_mode_switching, 1);
smp_mb__after_atomic();
/*
* If there are multiple online CPUs then wait until all threads whose
* FP mode is about to change have been context switched. This approach
* allows us to only worry about whether an FP mode switch is in
* progress when FP is first used in a tasks time slice. Pretty much all
* of the mode switch overhead can thus be confined to cases where mode
* switches are actually occuring. That is, to here. However for the
* thread performing the mode switch it may take a while...
*/
if (num_online_cpus() > 1) {
spin_lock_irq(&task->sighand->siglock);
for_each_thread(task, t) {
if (t == current)
continue;
switch_count = t->nvcsw + t->nivcsw;
do {
spin_unlock_irq(&task->sighand->siglock);
cond_resched();
spin_lock_irq(&task->sighand->siglock);
} while ((t->nvcsw + t->nivcsw) == switch_count);
}
spin_unlock_irq(&task->sighand->siglock);
}
/*
* There are now no threads of the process with live FP context, so it
* is safe to proceed with the FP mode switch.
*/
for_each_thread(task, t) {
/* Update desired FP register width */
if (value & PR_FP_MODE_FR) {
clear_tsk_thread_flag(t, TIF_32BIT_FPREGS);
} else {
set_tsk_thread_flag(t, TIF_32BIT_FPREGS);
clear_tsk_thread_flag(t, TIF_MSA_CTX_LIVE);
}
/* Update desired FP single layout */
if (value & PR_FP_MODE_FRE)
set_tsk_thread_flag(t, TIF_HYBRID_FPREGS);
else
clear_tsk_thread_flag(t, TIF_HYBRID_FPREGS);
}
/* Allow threads to use FP again */
atomic_set(&task->mm->context.fp_mode_switching, 0);
preempt_enable();
2015-01-08 12:17:37 +00:00
return 0;
}