From: Nathan Lynch nathanl@linux.ibm.com
[ Upstream commit e610a466d16a086e321f0bd421e2fc75cff28605 ]
It's common for the platform to replace the cache device nodes after a migration. Since the cacheinfo code is never informed about this, it never drops its references to the source system's cache nodes, causing it to wind up in an inconsistent state resulting in warnings and oopses as soon as CPU online/offline occurs after the migration, e.g.
cache for /cpus/l3-cache@3113(Unified) refers to cache for /cpus/l2-cache@200d(Unified) WARNING: CPU: 15 PID: 86 at arch/powerpc/kernel/cacheinfo.c:176 release_cache+0x1bc/0x1d0 [...] NIP release_cache+0x1bc/0x1d0 LR release_cache+0x1b8/0x1d0 Call Trace: release_cache+0x1b8/0x1d0 (unreliable) cacheinfo_cpu_offline+0x1c4/0x2c0 unregister_cpu_online+0x1b8/0x260 cpuhp_invoke_callback+0x114/0xf40 cpuhp_thread_fun+0x270/0x310 smpboot_thread_fn+0x2c8/0x390 kthread+0x1b8/0x1c0 ret_from_kernel_thread+0x5c/0x68
Using device tree notifiers won't work since we want to rebuild the hierarchy only after all the removals and additions have occurred and the device tree is in a consistent state. Call cacheinfo_teardown() before processing device tree updates, and rebuild the hierarchy afterward.
Fixes: 410bccf97881 ("powerpc/pseries: Partition migration in the kernel") Signed-off-by: Nathan Lynch nathanl@linux.ibm.com Reviewed-by: Gautham R. Shenoy ego@linux.vnet.ibm.com Signed-off-by: Michael Ellerman mpe@ellerman.id.au Signed-off-by: Sasha Levin sashal@kernel.org --- arch/powerpc/platforms/pseries/mobility.c | 10 ++++++++++ 1 file changed, 10 insertions(+)
diff --git a/arch/powerpc/platforms/pseries/mobility.c b/arch/powerpc/platforms/pseries/mobility.c index e4ea71383383..70744b4fbd9e 100644 --- a/arch/powerpc/platforms/pseries/mobility.c +++ b/arch/powerpc/platforms/pseries/mobility.c @@ -24,6 +24,7 @@ #include <asm/machdep.h> #include <asm/rtas.h> #include "pseries.h" +#include "../../kernel/cacheinfo.h"
static struct kobject *mobility_kobj;
@@ -360,11 +361,20 @@ void post_mobility_fixup(void) */ cpus_read_lock();
+ /* + * It's common for the destination firmware to replace cache + * nodes. Release all of the cacheinfo hierarchy's references + * before updating the device tree. + */ + cacheinfo_teardown(); + rc = pseries_devicetree_update(MIGRATION_SCOPE); if (rc) printk(KERN_ERR "Post-mobility device tree update " "failed: %d\n", rc);
+ cacheinfo_rebuild(); + cpus_read_unlock();
/* Possibly switch to a new RFI flush type */