2022-05-15 17:37:16

by Xianting Tian

Subject: [PATCH] RISC-V: Add fast call path of crash_kexec()

Currently, almost all archs (x86, arm64, mips...) support fast call
of crash_kexec() when "regs && kexec_should_crash()" is true. But
RISC-V not, it can only enter crash system via panic(). However panic()
doesn't pass the regs of the real accident scene to crash_kexec(),
it caused we can't get accurate backtrace via gdb,
$ riscv64-linux-gnu-gdb vmlinux vmcore
Reading symbols from vmlinux...
[New LWP 95]
#0 console_unlock () at kernel/printk/printk.c:2557
2557 if (do_cond_resched)
(gdb) bt
#0 console_unlock () at kernel/printk/printk.c:2557
#1 0x0000000000000000 in ?? ()

With the patch we can get the accurate backtrace,
$ riscv64-linux-gnu-gdb vmlinux vmcore
Reading symbols from vmlinux...
[New LWP 95]
#0 0xffffffe00063a4e0 in test_thread (data=<optimized out>) at drivers/virtio/virtio_mmio.c:806
806 *(int *)p = 0xdead;
(gdb)
(gdb) bt
#0 0xffffffe00063a4e0 in test_thread (data=<optimized out>) at drivers/virtio/virtio_mmio.c:806
#1 0x0000000000000000 in ?? ()

Test code to produce NULL address dereference,
+extern int panic_on_oops;
+static struct task_struct *k;
+static int test_thread(void *data) {
+
+ void *p = NULL;
+
+ while (!panic_on_oops)
+ msleep(2000);
+
+ *(int *)p = 0xdead;
+
+ return 0;
+}
+
static int __init virtio_mmio_init(void)
{
+ k = kthread_run(test_thread, NULL, "test_thread");
+ if (IS_ERR(k))
+ pr_err("Couldn't create test kthread\n");
+
return platform_driver_register(&virtio_mmio_driver);
}

Signed-off-by: Xianting Tian <[email protected]>
---
arch/riscv/kernel/traps.c | 4 ++++
1 file changed, 4 insertions(+)

diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
index fe92e119e6a3..e666ebfa2a64 100644
--- a/arch/riscv/kernel/traps.c
+++ b/arch/riscv/kernel/traps.c
@@ -16,6 +16,7 @@
#include <linux/mm.h>
#include <linux/module.h>
#include <linux/irq.h>
+#include <linux/kexec.h>

#include <asm/asm-prototypes.h>
#include <asm/bug.h>
@@ -44,6 +45,9 @@ void die(struct pt_regs *regs, const char *str)

ret = notify_die(DIE_OOPS, str, regs, 0, regs->cause, SIGSEGV);

+ if (regs && kexec_should_crash(current))
+ crash_kexec(regs);
+
bust_spinlocks(0);
add_taint(TAINT_DIE, LOCKDEP_NOW_UNRELIABLE);
spin_unlock_irq(&die_lock);
--
2.17.1

2022-05-16 20:17:05

by Guo Ren

Subject: Re: [PATCH] RISC-V: Add fast call path of crash_kexec()

Great Job, I think it's a fixup, we should add Fix: and Cc: [email protected]

Reviewed-by: Guo Ren <[email protected]>

On Sun, May 15, 2022 at 9:14 PM Xianting Tian
<[email protected]> wrote:
>
> Currently, almost all archs (x86, arm64, mips...) support fast call
> of crash_kexec() when "regs && kexec_should_crash()" is true. But
> RISC-V not, it can only enter crash system via panic(). However panic()
> doesn't pass the regs of the real accident scene to crash_kexec(),
> it caused we can't get accurate backtrace via gdb,
> $ riscv64-linux-gnu-gdb vmlinux vmcore
> Reading symbols from vmlinux...
> [New LWP 95]
> #0 console_unlock () at kernel/printk/printk.c:2557
> 2557 if (do_cond_resched)
> (gdb) bt
> #0 console_unlock () at kernel/printk/printk.c:2557
> #1 0x0000000000000000 in ?? ()
>
> With the patch we can get the accurate backtrace,
> $ riscv64-linux-gnu-gdb vmlinux vmcore
> Reading symbols from vmlinux...
> [New LWP 95]
> #0 0xffffffe00063a4e0 in test_thread (data=<optimized out>) at drivers/virtio/virtio_mmio.c:806
> 806 *(int *)p = 0xdead;
> (gdb)
> (gdb) bt
> #0 0xffffffe00063a4e0 in test_thread (data=<optimized out>) at drivers/virtio/virtio_mmio.c:806
> #1 0x0000000000000000 in ?? ()
>
> Test code to produce NULL address dereference,
> +extern int panic_on_oops;
> +static struct task_struct *k;
> +static int test_thread(void *data) {
> +
> + void *p = NULL;
> +
> + while (!panic_on_oops)
> + msleep(2000);
> +
> + *(int *)p = 0xdead;
> +
> + return 0;
> +}
> +
> static int __init virtio_mmio_init(void)
> {
> + k = kthread_run(test_thread, NULL, "test_thread");
> + if (IS_ERR(k))
> + pr_err("Couldn't create test kthread\n");
> +
> return platform_driver_register(&virtio_mmio_driver);
> }
>
> Signed-off-by: Xianting Tian <[email protected]>
> ---
> arch/riscv/kernel/traps.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
> index fe92e119e6a3..e666ebfa2a64 100644
> --- a/arch/riscv/kernel/traps.c
> +++ b/arch/riscv/kernel/traps.c
> @@ -16,6 +16,7 @@
> #include <linux/mm.h>
> #include <linux/module.h>
> #include <linux/irq.h>
> +#include <linux/kexec.h>
>
> #include <asm/asm-prototypes.h>
> #include <asm/bug.h>
> @@ -44,6 +45,9 @@ void die(struct pt_regs *regs, const char *str)
>
> ret = notify_die(DIE_OOPS, str, regs, 0, regs->cause, SIGSEGV);
>
> + if (regs && kexec_should_crash(current))
> + crash_kexec(regs);
> +
> bust_spinlocks(0);
> add_taint(TAINT_DIE, LOCKDEP_NOW_UNRELIABLE);
> spin_unlock_irq(&die_lock);
> --
> 2.17.1
>


--
Best Regards
Guo Ren

ML: https://lore.kernel.org/linux-csky/

2022-05-24 04:30:55

by Xianting Tian

Subject: Re: [PATCH] RISC-V: Add fast call path of crash_kexec()

just a friendly ping

?? 2022/5/15 ????9:14, Xianting Tian д??:
> Currently, almost all archs (x86, arm64, mips...) support fast call
> of crash_kexec() when "regs && kexec_should_crash()" is true. But
> RISC-V not, it can only enter crash system via panic(). However panic()
> doesn't pass the regs of the real accident scene to crash_kexec(),
> it caused we can't get accurate backtrace via gdb,
> $ riscv64-linux-gnu-gdb vmlinux vmcore
> Reading symbols from vmlinux...
> [New LWP 95]
> #0 console_unlock () at kernel/printk/printk.c:2557
> 2557 if (do_cond_resched)
> (gdb) bt
> #0 console_unlock () at kernel/printk/printk.c:2557
> #1 0x0000000000000000 in ?? ()
>
> With the patch we can get the accurate backtrace,
> $ riscv64-linux-gnu-gdb vmlinux vmcore
> Reading symbols from vmlinux...
> [New LWP 95]
> #0 0xffffffe00063a4e0 in test_thread (data=<optimized out>) at drivers/virtio/virtio_mmio.c:806
> 806 *(int *)p = 0xdead;
> (gdb)
> (gdb) bt
> #0 0xffffffe00063a4e0 in test_thread (data=<optimized out>) at drivers/virtio/virtio_mmio.c:806
> #1 0x0000000000000000 in ?? ()
>
> Test code to produce NULL address dereference,
> +extern int panic_on_oops;
> +static struct task_struct *k;
> +static int test_thread(void *data) {
> +
> + void *p = NULL;
> +
> + while (!panic_on_oops)
> + msleep(2000);
> +
> + *(int *)p = 0xdead;
> +
> + return 0;
> +}
> +
> static int __init virtio_mmio_init(void)
> {
> + k = kthread_run(test_thread, NULL, "test_thread");
> + if (IS_ERR(k))
> + pr_err("Couldn't create test kthread\n");
> +
> return platform_driver_register(&virtio_mmio_driver);
> }
>
> Signed-off-by: Xianting Tian <[email protected]>
> ---
> arch/riscv/kernel/traps.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
> diff --git a/arch/riscv/kernel/traps.c b/arch/riscv/kernel/traps.c
> index fe92e119e6a3..e666ebfa2a64 100644
> --- a/arch/riscv/kernel/traps.c
> +++ b/arch/riscv/kernel/traps.c
> @@ -16,6 +16,7 @@
> #include <linux/mm.h>
> #include <linux/module.h>
> #include <linux/irq.h>
> +#include <linux/kexec.h>
>
> #include <asm/asm-prototypes.h>
> #include <asm/bug.h>
> @@ -44,6 +45,9 @@ void die(struct pt_regs *regs, const char *str)
>
> ret = notify_die(DIE_OOPS, str, regs, 0, regs->cause, SIGSEGV);
>
> + if (regs && kexec_should_crash(current))
> + crash_kexec(regs);
> +
> bust_spinlocks(0);
> add_taint(TAINT_DIE, LOCKDEP_NOW_UNRELIABLE);
> spin_unlock_irq(&die_lock);