aboutsummaryrefslogtreecommitdiffstats
path: root/arch
diff options
context:
space:
mode:
authorRussell King <rmk+kernel@armlinux.org.uk>2018-04-11 19:35:19 +0100
committerRussell King <rmk+kernel@armlinux.org.uk>2018-05-19 11:35:56 +0100
commit1c37963b1a1600b9686c4a99857ddcb6028be884 (patch)
tree64fc17a95c9c4f6b8345c9cb8e23c2bfd4a7602c /arch
parentARM: kexec: fix kdump register saving on panic() (diff)
downloadlinux-dev-1c37963b1a1600b9686c4a99857ddcb6028be884.tar.xz
linux-dev-1c37963b1a1600b9686c4a99857ddcb6028be884.zip
ARM: kexec: record parent context registers for non-crash CPUs
How we got to machine_crash_nonpanic_core() (iow, from an IPI, etc) is not interesting for debugging a crash. The more interesting context is the parent context prior to the IPI being received. Record the parent context register state rather than the register state in machine_crash_nonpanic_core(), which is more relevant to the failing condition. Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Diffstat (limited to 'arch')
-rw-r--r--arch/arm/kernel/machine_kexec.c2
1 files changed, 1 insertions, 1 deletions
diff --git a/arch/arm/kernel/machine_kexec.c b/arch/arm/kernel/machine_kexec.c
index c15318431986..dd2eb5f76b9f 100644
--- a/arch/arm/kernel/machine_kexec.c
+++ b/arch/arm/kernel/machine_kexec.c
@@ -83,7 +83,7 @@ void machine_crash_nonpanic_core(void *unused)
{
struct pt_regs regs;
- crash_setup_regs(&regs, NULL);
+ crash_setup_regs(&regs, get_irq_regs());
printk(KERN_DEBUG "CPU %u will stop doing anything useful since another CPU has crashed\n",
smp_processor_id());
crash_save_cpu(&regs, smp_processor_id());