On Thu, Jul 26, 2018 at 02:36:07PM +0100, James Morse wrote:

Hi Akashi,

On 24/07/18 07:57, AKASHI Takahiro wrote:

quoted

On arm64, purgatory would do almost nothing. So just invoke secondary
kernel directly by jumping into its entry code.

While, in this case, cpu_soft_restart() must be called with dtb address
in the fifth argument, the behavior still stays compatible with kexec_load
case as long as the argument is null.

This patch conflicts with commit 76f4e2da45b4 ("arm64: kexec: always reset to
EL2 if present") in the arm64 tree.

I haven't noticed Mark's patch.

I'm going to have to refresh my memory regarding why I introduced
el2_switch when I implemented kdump.
According to my current memory, however, I added
kvm_arch_hardware_enable/disable(), and associated functions, to gracefully
shutdown EL2 in case of kexec. Since we have no chance to call reset
function (via notifier) at kdump, I believed that el2_switch was necessary
for better chance of successful kdump.

Thanks,
-Takahiro AKASHI

Thanks,

James

quoted

diff --git a/arch/arm64/kernel/machine_kexec.c b/arch/arm64/kernel/machine_kexec.c
index f76ea92dff91..830a5063e09d 100644
--- a/arch/arm64/kernel/machine_kexec.c
+++ b/arch/arm64/kernel/machine_kexec.c

@@ -205,10 +205,18 @@ void machine_kexec(struct kimage *kimage)
 	 * uses physical addressing to relocate the new image to its final
 	 * position and transfers control to the image entry point when the
 	 * relocation is complete.
+	 * In kexec case, kimage->start points to purgatory assuming that
+	 * kernel entry and dtb address are embedded in purgatory by
+	 * userspace (kexec-tools).
+	 * In kexec_file case, the kernel starts directly without purgatory.
 	 */
-
 	cpu_soft_restart(kimage != kexec_crash_image,
-		reboot_code_buffer_phys, kimage->head, kimage->start, 0);
+		reboot_code_buffer_phys, kimage->head, kimage->start,
+#ifdef CONFIG_KEXEC_FILE
+						kimage->arch.dtb_mem);
+#else
+						0);
+#endif
 
 	BUG(); /* Should never get here. */
 }