Merge tag 'powerpc-5.10-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux

Pull powerpc fixes from Michael Ellerman:

 - A fix for undetected data corruption on Power9 Nimbus <= DD2.1 in the
   emulation of VSX loads. The affected CPUs were not widely available.

 - Two fixes for machine check handling in guests under PowerVM.

 - A fix for our recent changes to SMP setup, when
   CONFIG_CPUMASK_OFFSTACK=y.

 - Three fixes for races in the handling of some of our powernv sysfs
   attributes.

 - One change to remove TM from the set of Power10 CPU features.

 - A couple of other minor fixes.

Thanks to: Aneesh Kumar K.V, Christophe Leroy, Ganesh Goudar, Jordan
Niethe, Mahesh Salgaonkar, Michael Neuling, Oliver O'Halloran, Qian Cai,
Srikar Dronamraju, Vasant Hegde.

* tag 'powerpc-5.10-2' of git://git.kernel.org/pub/scm/linux/kernel/git/powerpc/linux:
  powerpc/pseries: Avoid using addr_to_pfn in real mode
  powerpc/uaccess: Don't use "m<>" constraint with GCC 4.9
  powerpc/eeh: Fix eeh_dev_check_failure() for PE#0
  powerpc/64s: Remove TM from Power10 features
  selftests/powerpc: Make alignment handler test P9N DD2.1 vector CI load workaround
  powerpc: Fix undetected data corruption with P9N DD2.1 VSX CI load emulation
  powerpc/powernv/dump: Handle multiple writes to ack attribute
  powerpc/powernv/dump: Fix race while processing OPAL dump
  powerpc/smp: Use GFP_ATOMIC while allocating tmp mask
  powerpc/smp: Remove unnecessary variable
  powerpc/mce: Avoid nmi_enter/exit in real mode on pseries hash
  powerpc/opal_elog: Handle multiple writes to ack attribute

arch/powerpc/include/asm/asm-const.h

@@ -11,4 +11,17 @@
 #  define __ASM_CONST(x)	x##UL
 #  define ASM_CONST(x)		__ASM_CONST(x)
 #endif
+
+/*
+ * Inline assembly memory constraint
+ *
+ * GCC 4.9 doesn't properly handle pre update memory constraint "m<>"
+ *
+ */
+#if defined(GCC_VERSION) && GCC_VERSION < 50000
+#define UPD_CONSTR ""
+#else
+#define UPD_CONSTR "<>"
+#endif
+
 #endif /* _ASM_POWERPC_ASM_CONST_H */

arch/powerpc/include/asm/cputable.h

@@ -477,7 +477,7 @@ static inline void cpu_feature_keys_init(void) { }
 	    CPU_FTR_STCX_CHECKS_ADDRESS | CPU_FTR_POPCNTB | CPU_FTR_POPCNTD | \
 	    CPU_FTR_CFAR | CPU_FTR_HVMODE | CPU_FTR_VMX_COPY | \
 	    CPU_FTR_DBELL | CPU_FTR_HAS_PPR | CPU_FTR_ARCH_207S | \
-	    CPU_FTR_TM_COMP | CPU_FTR_ARCH_300 | CPU_FTR_ARCH_31 | \
+	    CPU_FTR_ARCH_300 | CPU_FTR_ARCH_31 | \
 	    CPU_FTR_DAWR | CPU_FTR_DAWR1)
 #define CPU_FTRS_CELL	(CPU_FTR_LWSYNC | \
 	    CPU_FTR_PPCAS_ARCH_V2 | CPU_FTR_CTRL | \

arch/powerpc/include/asm/uaccess.h

@@ -182,7 +182,7 @@ do {								\
 		"1:	" op "%U1%X1 %0,%1	# put_user\n"	\
 		EX_TABLE(1b, %l2)				\
 		:						\
-		: "r" (x), "m<>" (*addr)				\
+		: "r" (x), "m"UPD_CONSTR (*addr)		\
 		:						\
 		: label)
 
@@ -253,7 +253,7 @@ extern long __get_user_bad(void);
 		".previous\n"				\
 		EX_TABLE(1b, 3b)			\
 		: "=r" (err), "=r" (x)			\
-		: "m<>" (*addr), "i" (-EFAULT), "0" (err))
+		: "m"UPD_CONSTR (*addr), "i" (-EFAULT), "0" (err))
 
 #ifdef __powerpc64__
 #define __get_user_asm2(x, addr, err)			\

arch/powerpc/kernel/cputable.c

@@ -121,9 +121,16 @@ extern void __restore_cpu_e6500(void);
 				 PPC_FEATURE2_DARN | \
 				 PPC_FEATURE2_SCV)
 #define COMMON_USER_POWER10	COMMON_USER_POWER9
-#define COMMON_USER2_POWER10	(COMMON_USER2_POWER9 | \
-				 PPC_FEATURE2_ARCH_3_1 | \
-				 PPC_FEATURE2_MMA)
+#define COMMON_USER2_POWER10	(PPC_FEATURE2_ARCH_3_1 | \
+				 PPC_FEATURE2_MMA | \
+				 PPC_FEATURE2_ARCH_3_00 | \
+				 PPC_FEATURE2_HAS_IEEE128 | \
+				 PPC_FEATURE2_DARN | \
+				 PPC_FEATURE2_SCV | \
+				 PPC_FEATURE2_ARCH_2_07 | \
+				 PPC_FEATURE2_DSCR | \
+				 PPC_FEATURE2_ISEL | PPC_FEATURE2_TAR | \
+				 PPC_FEATURE2_VEC_CRYPTO)
 
 #ifdef CONFIG_PPC_BOOK3E_64
 #define COMMON_USER_BOOKE	(COMMON_USER_PPC64 | PPC_FEATURE_BOOKE)

arch/powerpc/kernel/eeh.c

@@ -466,11 +466,6 @@ int eeh_dev_check_failure(struct eeh_dev *edev)
 		return 0;
 	}
 
-	if (!pe->addr) {
-		eeh_stats.no_cfg_addr++;
-		return 0;
-	}
-
 	/*
 	 * On PowerNV platform, we might already have fenced PHB
 	 * there and we need take care of that firstly.

arch/powerpc/kernel/mce.c

@@ -591,12 +591,11 @@ EXPORT_SYMBOL_GPL(machine_check_print_event_info);
 long notrace machine_check_early(struct pt_regs *regs)
 {
 	long handled = 0;
-	bool nested = in_nmi();
 	u8 ftrace_enabled = this_cpu_get_ftrace_enabled();
 
 	this_cpu_set_ftrace_enabled(0);
-
-	if (!nested)
+	/* Do not use nmi_enter/exit for pseries hpte guest */
+	if (radix_enabled() || !firmware_has_feature(FW_FEATURE_LPAR))
 		nmi_enter();
 
 	hv_nmi_check_nonrecoverable(regs);
@@ -607,7 +606,7 @@ long notrace machine_check_early(struct pt_regs *regs)
 	if (ppc_md.machine_check_early)
 		handled = ppc_md.machine_check_early(regs);
 
-	if (!nested)
+	if (radix_enabled() || !firmware_has_feature(FW_FEATURE_LPAR))
 		nmi_exit();
 
 	this_cpu_set_ftrace_enabled(ftrace_enabled);

arch/powerpc/kernel/smp.c

@@ -1240,43 +1240,33 @@ static struct device_node *cpu_to_l2cache(int cpu)
 	return cache;
 }
 
-static bool update_mask_by_l2(int cpu)
+static bool update_mask_by_l2(int cpu, cpumask_var_t *mask)
 {
 	struct cpumask *(*submask_fn)(int) = cpu_sibling_mask;
 	struct device_node *l2_cache, *np;
-	cpumask_var_t mask;
 	int i;
 
-	l2_cache = cpu_to_l2cache(cpu);
-	if (!l2_cache) {
-		struct cpumask *(*sibling_mask)(int) = cpu_sibling_mask;
-
-		/*
-		 * If no l2cache for this CPU, assume all siblings to share
-		 * cache with this CPU.
-		 */
-		if (has_big_cores)
-			sibling_mask = cpu_smallcore_mask;
+	if (has_big_cores)
+		submask_fn = cpu_smallcore_mask;
 
-		for_each_cpu(i, sibling_mask(cpu))
+	l2_cache = cpu_to_l2cache(cpu);
+	if (!l2_cache || !*mask) {
+		/* Assume only core siblings share cache with this CPU */
+		for_each_cpu(i, submask_fn(cpu))
 			set_cpus_related(cpu, i, cpu_l2_cache_mask);
 
 		return false;
 	}
 
-	alloc_cpumask_var_node(&mask, GFP_KERNEL, cpu_to_node(cpu));
-	cpumask_and(mask, cpu_online_mask, cpu_cpu_mask(cpu));
-
-	if (has_big_cores)
-		submask_fn = cpu_smallcore_mask;
+	cpumask_and(*mask, cpu_online_mask, cpu_cpu_mask(cpu));
 
 	/* Update l2-cache mask with all the CPUs that are part of submask */
 	or_cpumasks_related(cpu, cpu, submask_fn, cpu_l2_cache_mask);
 
 	/* Skip all CPUs already part of current CPU l2-cache mask */
-	cpumask_andnot(mask, mask, cpu_l2_cache_mask(cpu));
+	cpumask_andnot(*mask, *mask, cpu_l2_cache_mask(cpu));
 
-	for_each_cpu(i, mask) {
+	for_each_cpu(i, *mask) {
 		/*
 		 * when updating the marks the current CPU has not been marked
 		 * online, but we need to update the cache masks
@@ -1286,15 +1276,14 @@ static bool update_mask_by_l2(int cpu)
 		/* Skip all CPUs already part of current CPU l2-cache */
 		if (np == l2_cache) {
 			or_cpumasks_related(cpu, i, submask_fn, cpu_l2_cache_mask);
-			cpumask_andnot(mask, mask, submask_fn(i));
+			cpumask_andnot(*mask, *mask, submask_fn(i));
 		} else {
-			cpumask_andnot(mask, mask, cpu_l2_cache_mask(i));
+			cpumask_andnot(*mask, *mask, cpu_l2_cache_mask(i));
 		}
 
 		of_node_put(np);
 	}
 	of_node_put(l2_cache);
-	free_cpumask_var(mask);
 
 	return true;
 }
@@ -1337,40 +1326,46 @@ static inline void add_cpu_to_smallcore_masks(int cpu)
 	}
 }
 
-static void update_coregroup_mask(int cpu)
+static void update_coregroup_mask(int cpu, cpumask_var_t *mask)
 {
 	struct cpumask *(*submask_fn)(int) = cpu_sibling_mask;
-	cpumask_var_t mask;
 	int coregroup_id = cpu_to_coregroup_id(cpu);
 	int i;
 
-	alloc_cpumask_var_node(&mask, GFP_KERNEL, cpu_to_node(cpu));
-	cpumask_and(mask, cpu_online_mask, cpu_cpu_mask(cpu));
-
 	if (shared_caches)
 		submask_fn = cpu_l2_cache_mask;
 
+	if (!*mask) {
+		/* Assume only siblings are part of this CPU's coregroup */
+		for_each_cpu(i, submask_fn(cpu))
+			set_cpus_related(cpu, i, cpu_coregroup_mask);
+
+		return;
+	}
+
+	cpumask_and(*mask, cpu_online_mask, cpu_cpu_mask(cpu));
+
 	/* Update coregroup mask with all the CPUs that are part of submask */
 	or_cpumasks_related(cpu, cpu, submask_fn, cpu_coregroup_mask);
 
 	/* Skip all CPUs already part of coregroup mask */
-	cpumask_andnot(mask, mask, cpu_coregroup_mask(cpu));
+	cpumask_andnot(*mask, *mask, cpu_coregroup_mask(cpu));
 
-	for_each_cpu(i, mask) {
+	for_each_cpu(i, *mask) {
 		/* Skip all CPUs not part of this coregroup */
 		if (coregroup_id == cpu_to_coregroup_id(i)) {
 			or_cpumasks_related(cpu, i, submask_fn, cpu_coregroup_mask);
-			cpumask_andnot(mask, mask, submask_fn(i));
+			cpumask_andnot(*mask, *mask, submask_fn(i));
 		} else {
-			cpumask_andnot(mask, mask, cpu_coregroup_mask(i));
+			cpumask_andnot(*mask, *mask, cpu_coregroup_mask(i));
 		}
 	}
-	free_cpumask_var(mask);
 }
 
 static void add_cpu_to_masks(int cpu)
 {
 	int first_thread = cpu_first_thread_sibling(cpu);
+	cpumask_var_t mask;
 	int i;
 
 	/*
@@ -1384,10 +1379,15 @@ static void add_cpu_to_masks(int cpu)
 			set_cpus_related(i, cpu, cpu_sibling_mask);
 
 	add_cpu_to_smallcore_masks(cpu);
-	update_mask_by_l2(cpu);
+
+	/* In CPU-hotplug path, hence use GFP_ATOMIC */
+	alloc_cpumask_var_node(&mask, GFP_ATOMIC, cpu_to_node(cpu));
+	update_mask_by_l2(cpu, &mask);
 
 	if (has_coregroup_support())
-		update_coregroup_mask(cpu);
+		update_coregroup_mask(cpu, &mask);
+
+	free_cpumask_var(mask);
 }
 
 /* Activate a secondary processor. */

arch/powerpc/kernel/traps.c

@@ -885,7 +885,7 @@ static void p9_hmi_special_emu(struct pt_regs *regs)
 {
 	unsigned int ra, rb, t, i, sel, instr, rc;
 	const void __user *addr;
-	u8 vbuf[16], *vdst;
+	u8 vbuf[16] __aligned(16), *vdst;
 	unsigned long ea, msr, msr_mask;
 	bool swap;
 

arch/powerpc/platforms/powernv/opal-dump.c

@@ -88,9 +88,14 @@ static ssize_t dump_ack_store(struct dump_obj *dump_obj,
 			      const char *buf,
 			      size_t count)
 {
-	dump_send_ack(dump_obj->id);
-	sysfs_remove_file_self(&dump_obj->kobj, &attr->attr);
-	kobject_put(&dump_obj->kobj);
+	/*
+	 * Try to self remove this attribute. If we are successful,
+	 * delete the kobject itself.
+	 */
+	if (sysfs_remove_file_self(&dump_obj->kobj, &attr->attr)) {
+		dump_send_ack(dump_obj->id);
+		kobject_put(&dump_obj->kobj);
+	}
 	return count;
 }
 
@@ -318,15 +323,14 @@ static ssize_t dump_attr_read(struct file *filep, struct kobject *kobj,
 	return count;
 }
 
-static struct dump_obj *create_dump_obj(uint32_t id, size_t size,
-					uint32_t type)
+static void create_dump_obj(uint32_t id, size_t size, uint32_t type)
 {
 	struct dump_obj *dump;
 	int rc;
 
 	dump = kzalloc(sizeof(*dump), GFP_KERNEL);
 	if (!dump)
-		return NULL;
+		return;
 
 	dump->kobj.kset = dump_kset;
 
@@ -346,21 +350,39 @@ static struct dump_obj *create_dump_obj(uint32_t id, size_t size,
 	rc = kobject_add(&dump->kobj, NULL, "0x%x-0x%x", type, id);
 	if (rc) {
 		kobject_put(&dump->kobj);
-		return NULL;
+		return;
 	}
 
+	/*
+	 * As soon as the sysfs file for this dump is created/activated there is
+	 * a chance the opal_errd daemon (or any userspace) might read and
+	 * acknowledge the dump before kobject_uevent() is called. If that
+	 * happens then there is a potential race between
+	 * dump_ack_store->kobject_put() and kobject_uevent() which leads to a
+	 * use-after-free of a kernfs object resulting in a kernel crash.
+	 *
+	 * To avoid that, we need to take a reference on behalf of the bin file,
+	 * so that our reference remains valid while we call kobject_uevent().
+	 * We then drop our reference before exiting the function, leaving the
+	 * bin file to drop the last reference (if it hasn't already).
+	 */
+
+	/* Take a reference for the bin file */
+	kobject_get(&dump->kobj);
 	rc = sysfs_create_bin_file(&dump->kobj, &dump->dump_attr);
-	if (rc) {
+	if (rc == 0) {
+		kobject_uevent(&dump->kobj, KOBJ_ADD);
+
+		pr_info("%s: New platform dump. ID = 0x%x Size %u\n",
+			__func__, dump->id, dump->size);
+	} else {
+		/* Drop reference count taken for bin file */
 		kobject_put(&dump->kobj);
-		return NULL;
 	}
 
-	pr_info("%s: New platform dump. ID = 0x%x Size %u\n",
-		__func__, dump->id, dump->size);
-
-	kobject_uevent(&dump->kobj, KOBJ_ADD);
-
-	return dump;
+	/* Drop our reference */
+	kobject_put(&dump->kobj);
+	return;
 }
 
 static irqreturn_t process_dump(int irq, void *data)

arch/powerpc/platforms/powernv/opal-elog.c

@@ -72,9 +72,14 @@ static ssize_t elog_ack_store(struct elog_obj *elog_obj,
 			      const char *buf,
 			      size_t count)
 {
-	opal_send_ack_elog(elog_obj->id);
-	sysfs_remove_file_self(&elog_obj->kobj, &attr->attr);
-	kobject_put(&elog_obj->kobj);
+	/*
+	 * Try to self remove this attribute. If we are successful,
+	 * delete the kobject itself.
+	 */
+	if (sysfs_remove_file_self(&elog_obj->kobj, &attr->attr)) {
+		opal_send_ack_elog(elog_obj->id);
+		kobject_put(&elog_obj->kobj);
+	}
 	return count;
 }
 

arch/powerpc/platforms/pseries/ras.c

@@ -521,18 +521,55 @@ int pSeries_system_reset_exception(struct pt_regs *regs)
 	return 0; /* need to perform reset */
 }
 
+static int mce_handle_err_realmode(int disposition, u8 error_type)
+{
+#ifdef CONFIG_PPC_BOOK3S_64
+	if (disposition == RTAS_DISP_NOT_RECOVERED) {
+		switch (error_type) {
+		case	MC_ERROR_TYPE_SLB:
+		case	MC_ERROR_TYPE_ERAT:
+			/*
+			 * Store the old slb content in paca before flushing.
+			 * Print this when we go to virtual mode.
+			 * There are chances that we may hit MCE again if there
+			 * is a parity error on the SLB entry we trying to read
+			 * for saving. Hence limit the slb saving to single
+			 * level of recursion.
+			 */
+			if (local_paca->in_mce == 1)
+				slb_save_contents(local_paca->mce_faulty_slbs);
+			flush_and_reload_slb();
+			disposition = RTAS_DISP_FULLY_RECOVERED;
+			break;
+		default:
+			break;
+		}
+	} else if (disposition == RTAS_DISP_LIMITED_RECOVERY) {
+		/* Platform corrected itself but could be degraded */
+		pr_err("MCE: limited recovery, system may be degraded\n");
+		disposition = RTAS_DISP_FULLY_RECOVERED;
+	}
+#endif
+	return disposition;
+}
 
-static int mce_handle_error(struct pt_regs *regs, struct rtas_error_log *errp)
+static int mce_handle_err_virtmode(struct pt_regs *regs,
+				   struct rtas_error_log *errp,
+				   struct pseries_mc_errorlog *mce_log,
+				   int disposition)
 {
 	struct mce_error_info mce_err = { 0 };
-	unsigned long eaddr = 0, paddr = 0;
-	struct pseries_errorlog *pseries_log;
-	struct pseries_mc_errorlog *mce_log;
-	int disposition = rtas_error_disposition(errp);
 	int initiator = rtas_error_initiator(errp);
 	int severity = rtas_error_severity(errp);
+	unsigned long eaddr = 0, paddr = 0;
 	u8 error_type, err_sub_type;
 
+	if (!mce_log)
+		goto out;
+
+	error_type = mce_log->error_type;
+	err_sub_type = rtas_mc_error_sub_type(mce_log);
+
 	if (initiator == RTAS_INITIATOR_UNKNOWN)
 		mce_err.initiator = MCE_INITIATOR_UNKNOWN;
 	else if (initiator == RTAS_INITIATOR_CPU)
@@ -571,18 +608,7 @@ static int mce_handle_error(struct pt_regs *regs, struct rtas_error_log *errp)
 	mce_err.error_type = MCE_ERROR_TYPE_UNKNOWN;
 	mce_err.error_class = MCE_ECLASS_UNKNOWN;
 
-	if (!rtas_error_extended(errp))
-		goto out;
-
-	pseries_log = get_pseries_errorlog(errp, PSERIES_ELOG_SECT_ID_MCE);
-	if (pseries_log == NULL)
-		goto out;
-
-	mce_log = (struct pseries_mc_errorlog *)pseries_log->data;
-	error_type = mce_log->error_type;
-	err_sub_type = rtas_mc_error_sub_type(mce_log);
-
-	switch (mce_log->error_type) {
+	switch (error_type) {
 	case MC_ERROR_TYPE_UE:
 		mce_err.error_type = MCE_ERROR_TYPE_UE;
 		mce_common_process_ue(regs, &mce_err);
@@ -682,37 +708,31 @@ static int mce_handle_error(struct pt_regs *regs, struct rtas_error_log *errp)
 		mce_err.error_type = MCE_ERROR_TYPE_UNKNOWN;
 		break;
 	}
+out:
+	save_mce_event(regs, disposition == RTAS_DISP_FULLY_RECOVERED,
+		       &mce_err, regs->nip, eaddr, paddr);
+	return disposition;
+}
 
-#ifdef CONFIG_PPC_BOOK3S_64
-	if (disposition == RTAS_DISP_NOT_RECOVERED) {
-		switch (error_type) {
-		case	MC_ERROR_TYPE_SLB:
-		case	MC_ERROR_TYPE_ERAT:
-			/*
-			 * Store the old slb content in paca before flushing.
-			 * Print this when we go to virtual mode.
-			 * There are chances that we may hit MCE again if there
-			 * is a parity error on the SLB entry we trying to read
-			 * for saving. Hence limit the slb saving to single
-			 * level of recursion.
-			 */
-			if (local_paca->in_mce == 1)
-				slb_save_contents(local_paca->mce_faulty_slbs);
-			flush_and_reload_slb();
-			disposition = RTAS_DISP_FULLY_RECOVERED;
-			break;
-		default:
-			break;
-		}
-	} else if (disposition == RTAS_DISP_LIMITED_RECOVERY) {
-		/* Platform corrected itself but could be degraded */
-		printk(KERN_ERR "MCE: limited recovery, system may "
-		       "be degraded\n");
-		disposition = RTAS_DISP_FULLY_RECOVERED;
-	}
-#endif
+static int mce_handle_error(struct pt_regs *regs, struct rtas_error_log *errp)
+{
+	struct pseries_errorlog *pseries_log;
+	struct pseries_mc_errorlog *mce_log = NULL;
+	int disposition = rtas_error_disposition(errp);
+	u8 error_type;
+
+	if (!rtas_error_extended(errp))
+		goto out;
+
+	pseries_log = get_pseries_errorlog(errp, PSERIES_ELOG_SECT_ID_MCE);
+	if (!pseries_log)
+		goto out;
+
+	mce_log = (struct pseries_mc_errorlog *)pseries_log->data;
+	error_type = mce_log->error_type;
+
+	disposition = mce_handle_err_realmode(disposition, error_type);
 
-out:
 	/*
 	 * Enable translation as we will be accessing per-cpu variables
 	 * in save_mce_event() which may fall outside RMO region, also
@@ -723,10 +743,10 @@ static int mce_handle_error(struct pt_regs *regs, struct rtas_error_log *errp)
 	 * Note: All the realmode handling like flushing SLB entries for
 	 *       SLB multihit is done by now.
 	 */
+out:
 	mtmsr(mfmsr() | MSR_IR | MSR_DR);
-	save_mce_event(regs, disposition == RTAS_DISP_FULLY_RECOVERED,
-			&mce_err, regs->nip, eaddr, paddr);
-
+	disposition = mce_handle_err_virtmode(regs, errp, mce_log,
+					      disposition);
 	return disposition;
 }
 

tools/testing/selftests/powerpc/alignment/alignment_handler.c

@@ -266,8 +266,12 @@ int do_test(char *test_name, void (*test_func)(char *, char *))
 	}
 
 	rc = 0;
-	/* offset = 0 no alignment fault, so skip */
-	for (offset = 1; offset < 16; offset++) {
+	/*
+	 * offset = 0 is aligned but tests the workaround for the P9N
+	 * DD2.1 vector CI load issue (see 5080332c2c89 "powerpc/64s:
+	 * Add workaround for P9 vector CI load issue")
+	 */
+	for (offset = 0; offset < 16; offset++) {
 		width = 16; /* vsx == 16 bytes */
 		r = 0;