diff mbox series

[10/16] core/init: enable machine check on secondaries

Message ID 20190107140428.16388-11-npiggin@gmail.com
State Accepted
Headers show
Series assorted MCE and SRESET handling and reporting | expand

Checks

Context Check Description
snowpatch_ozlabs/apply_patch success master/apply_patch Successfully applied
snowpatch_ozlabs/snowpatch_job_snowpatch-skiboot success Test snowpatch/job/snowpatch-skiboot on branch master

Commit Message

Nicholas Piggin Jan. 7, 2019, 2:04 p.m. UTC 
Secondary CPUs currently run with MSR[ME]=0 during boot, whih means
if they take a machine check, the system will checkstop.

Enable ME where possible and allow them to print registers.

Signed-off-by: Nicholas Piggin <npiggin@gmail.com>
---
 core/fast-reboot.c |  2 ++
 core/init.c        | 38 ++++++++++++++++++++++++++++++++++++++
 2 files changed, 40 insertions(+)
diff mbox series

Patch

diff --git a/core/fast-reboot.c b/core/fast-reboot.c
index eb1b348ae..d3cc8cf5d 100644
--- a/core/fast-reboot.c
+++ b/core/fast-reboot.c
@@ -362,6 +362,8 @@  void __noreturn fast_reboot_entry(void)
 		}
 		sync();
 		cleanup_cpu_state();
+		enable_machine_check();
+
 		__secondary_cpu_entry();
 	}
 
diff --git a/core/init.c b/core/init.c
index 0ad2bfb8a..3f7dd14ee 100644
--- a/core/init.c
+++ b/core/init.c
@@ -475,6 +475,39 @@  static void load_initramfs(void)
 	}
 }
 
+static void cpu_disable_ME_one(void *param __unused)
+{
+	disable_machine_check();
+}
+
+static int64_t cpu_disable_ME_all(void)
+{
+	struct cpu_thread *cpu;
+	struct cpu_job **jobs;
+
+	jobs = zalloc(sizeof(struct cpu_job *) * (cpu_max_pir + 1));
+	assert(jobs);
+
+	for_each_available_cpu(cpu) {
+		if (cpu == this_cpu())
+			continue;
+		jobs[cpu->pir] = cpu_queue_job(cpu, "cpu_disable_ME",
+						cpu_disable_ME_one, NULL);
+	}
+
+	/* this cpu */
+	cpu_disable_ME_one(NULL);
+
+	for_each_available_cpu(cpu) {
+		if (jobs[cpu->pir])
+			cpu_wait_job(jobs[cpu->pir], true);
+	}
+
+	free(jobs);
+
+	return OPAL_SUCCESS;
+}
+
 void *fdt;
 
 void __noreturn load_and_boot_kernel(bool is_reboot)
@@ -586,6 +619,9 @@  void __noreturn load_and_boot_kernel(bool is_reboot)
 	printf("INIT: Starting kernel at 0x%llx, fdt at %p %u bytes\n",
 	       kernel_entry, fdt, fdt_totalsize(fdt));
 
+	/* Disable machine checks on all */
+	cpu_disable_ME_all();
+
 	debug_descriptor.state_flags |= OPAL_BOOT_COMPLETE;
 
 	cpu_give_self_os();
@@ -1234,6 +1270,8 @@  void __noreturn __secondary_cpu_entry(void)
 	/* Secondary CPU called in */
 	cpu_callin(cpu);
 
+	enable_machine_check();
+
 	/* Some XIVE setup */
 	xive_cpu_callin(cpu);