RISC-V: Enable IPI CPU Backtrace

Add arch_trigger_cpumask_backtrace() which is a generic infrastructure
for sampling other CPUs' backtrace using IPI.

The feature is used when lockups are detected or in case of oops/panic
if parameters are set accordingly.

Below is the case of oops with the oops_all_cpu_backtrace enabled.

$ sysctl kernel.oops_all_cpu_backtrace=1

triggering oops shows:
[  212.214237] NMI backtrace for cpu 1
[  212.214390] CPU: 1 PID: 610 Comm: in:imklog Tainted: G           OE      6.10.0-rc6 #1
[  212.214570] Hardware name: riscv-virtio,qemu (DT)
[  212.214690] epc : fallback_scalar_usercopy+0x8/0xdc
[  212.214809]  ra : _copy_to_user+0x20/0x40
[  212.214913] epc : ffffffff80c3a930 ra : ffffffff8059ba7e sp : ff20000000eabb50
[  212.215061]  gp : ffffffff82066f90 tp : ff6000008e958000 t0 : 3463303866660000
[  212.215210]  t1 : 000000000000005b t2 : 3463303866666666 s0 : ff20000000eabb60
[  212.215358]  s1 : 0000000000000386 a0 : 00007ff6e81df926 a1 : ff600000824df800
[  212.215505]  a2 : 000000000000003f a3 : 7fffffffffffffc0 a4 : 0000000000000000
[  212.215651]  a5 : 000000000000003f a6 : 0000000000000000 a7 : 0000000000000000
[  212.215857]  s2 : ff600000824df800 s3 : ffffffff82066cc0 s4 : 0000000000001c1a
[  212.216074]  s5 : ffffffff8206a5a8 s6 : 00007ff6e81df926 s7 : ffffffff8206a5a0
[  212.216278]  s8 : ff600000824df800 s9 : ffffffff81e25de0 s10: 000000000000003f
[  212.216471]  s11: ffffffff8206a59d t3 : ff600000824df812 t4 : ff600000824df812
[  212.216651]  t5 : ff600000824df818 t6 : 0000000000040000
[  212.216796] status: 0000000000040120 badaddr: 0000000000000000 cause: 8000000000000001
[  212.217035] [<ffffffff80c3a930>] fallback_scalar_usercopy+0x8/0xdc
[  212.217207] [<ffffffff80095f56>] syslog_print+0x1f4/0x2b2
[  212.217362] [<ffffffff80096e5c>] do_syslog.part.0+0x94/0x2d8
[  212.217502] [<ffffffff800979e8>] do_syslog+0x66/0x88
[  212.217636] [<ffffffff803a5dda>] kmsg_read+0x44/0x5c
[  212.217764] [<ffffffff80392dbe>] proc_reg_read+0x7a/0xa8
[  212.217952] [<ffffffff802ff726>] vfs_read+0xb0/0x24e
[  212.218090] [<ffffffff803001ba>] ksys_read+0x64/0xe4
[  212.218264] [<ffffffff8030025a>] __riscv_sys_read+0x20/0x2c
[  212.218453] [<ffffffff80c4af9a>] do_trap_ecall_u+0x60/0x1d4
[  212.218664] [<ffffffff80c56998>] ret_from_exception+0x0/0x64

Signed-off-by: Ryo Takakura <takakura@valinux.co.jp>
Link: https://lore.kernel.org/r/20240718093659.158912-1-takakura@valinux.co.jp
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>

diff --git a/arch/riscv/include/asm/irq.h b/arch/riscv/include/asm/irq.h
index 8e10a94430a2c2bc6228ac7141f16ba832b474c0..8330d16b05b5e42046e4b4a165c7f067f526ff68 100644 (file)
--- a/arch/riscv/include/asm/irq.h
+++ b/arch/riscv/include/asm/irq.h
@@ -12,6 +12,11 @@
 
 #include <asm-generic/irq.h>
 
+#ifdef CONFIG_SMP
+void arch_trigger_cpumask_backtrace(const cpumask_t *mask, int exclude_cpu);
+#define arch_trigger_cpumask_backtrace arch_trigger_cpumask_backtrace
+#endif
+
 void riscv_set_intc_hwnode_fn(struct fwnode_handle *(*fn)(void));
 
 struct fwnode_handle *riscv_get_intc_hwnode(void);

diff --git a/arch/riscv/kernel/smp.c b/arch/riscv/kernel/smp.c
index 8e6eb64459afb0f1c89929be573b86aadabeccd7..9b047899791ce8765568d8635378961263676fcb 100644 (file)
--- a/arch/riscv/kernel/smp.c
+++ b/arch/riscv/kernel/smp.c
@@ -21,6 +21,7 @@
 #include <linux/delay.h>
 #include <linux/irq.h>
 #include <linux/irq_work.h>
+#include <linux/nmi.h>
 
 #include <asm/tlbflush.h>
 #include <asm/cacheflush.h>
@@ -33,6 +34,7 @@ enum ipi_message_type {
        IPI_CPU_CRASH_STOP,
        IPI_IRQ_WORK,
        IPI_TIMER,
+       IPI_CPU_BACKTRACE,
        IPI_MAX
 };
 
@@ -136,6 +138,9 @@ static irqreturn_t handle_IPI(int irq, void *data)
                tick_receive_broadcast();
                break;
 #endif
+       case IPI_CPU_BACKTRACE:
+               nmi_cpu_backtrace(get_irq_regs());
+               break;
        default:
                pr_warn("CPU%d: unhandled IPI%d\n", smp_processor_id(), ipi);
                break;
@@ -203,6 +208,7 @@ static const char * const ipi_names[] = {
        [IPI_CPU_CRASH_STOP]    = "CPU stop (for crash dump) interrupts",
        [IPI_IRQ_WORK]          = "IRQ work interrupts",
        [IPI_TIMER]             = "Timer broadcast interrupts",
+       [IPI_CPU_BACKTRACE]     = "CPU backtrace interrupts",
 };
 
 void show_ipi_stats(struct seq_file *p, int prec)
@@ -323,3 +329,13 @@ void arch_smp_send_reschedule(int cpu)
        send_ipi_single(cpu, IPI_RESCHEDULE);
 }
 EXPORT_SYMBOL_GPL(arch_smp_send_reschedule);
+
+static void riscv_backtrace_ipi(cpumask_t *mask)
+{
+       send_ipi_mask(mask, IPI_CPU_BACKTRACE);
+}
+
+void arch_trigger_cpumask_backtrace(const cpumask_t *mask, int exclude_cpu)
+{
+       nmi_trigger_cpumask_backtrace(mask, exclude_cpu, riscv_backtrace_ipi);
+}