Merge tag 's390-5.12-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux

Pull s390 updates from Vasily Gorbik:

 - Convert to using the generic entry infrastructure.

 - Add vdso time namespace support.

 - Switch s390 and alpha to 64-bit ino_t. As discussed at

     https://lore.kernel.org/linux-mm/YCV7QiyoweJwvN+m@osiris/

 - Get rid of expensive stck (store clock) usages where possible.
   Utilize cpu alternatives to patch stckf when supported.

 - Make tod_clock usage less error prone by converting it to a union and
   rework code which is using it.

 - Machine check handler fixes and cleanups.

 - Drop couple of minor inline asm optimizations to fix clang build.

 - Default configs changes notably to make libvirt happy.

 - Various changes to rework and improve qdio code.

 - Other small various fixes and improvements all over the code.

* tag 's390-5.12-1' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (68 commits)
  s390/qdio: remove 'merge_pending' mechanism
  s390/qdio: improve handling of PENDING buffers for QEBSM devices
  s390/qdio: rework q->qdio_error indication
  s390/qdio: inline qdio_kick_handler()
  s390/time: remove get_tod_clock_ext()
  s390/crypto: use store_tod_clock_ext()
  s390/hypfs: use store_tod_clock_ext()
  s390/debug: use union tod_clock
  s390/kvm: use union tod_clock
  s390/vdso: use union tod_clock
  s390/time: convert tod_clock_base to union
  s390/time: introduce new store_tod_clock_ext()
  s390/time: rename store_tod_clock_ext() and use union tod_clock
  s390/time: introduce union tod_clock
  s390,alpha: switch to 64-bit ino_t
  s390: split cleanup_sie
  s390: use r13 in cleanup_sie as temp register
  s390: fix kernel asce loading when sie is interrupted
  s390: add stack for machine check handler
  s390: use WRITE_ONCE when re-allocating async stack
  ...

Documentation/scsi/scsi-parameters.rst

@@ -94,7 +94,7 @@ parameters may be changed at runtime by the command
 			(/proc/sys/dev/scsi/logging_level).
 			There is also a nice 'scsi_logging_level' script in the
 			S390-tools package, available for download at
-			https://github.com/ibm-s390-tools/s390-tools/blob/master/scripts/scsi_logging_level
+			https://github.com/ibm-s390-linux/s390-tools/blob/master/scripts/scsi_logging_level
 
 	scsi_mod.scan=	[SCSI] sync (default) scans SCSI busses as they are
 			discovered.  async scans them in kernel threads,

Documentation/virt/kvm/s390-pv-boot.rst

@@ -80,5 +80,5 @@ Keys
 ----
 Every CEC will have a unique public key to enable tooling to build
 encrypted images.
-See  `s390-tools <https://github.com/ibm-s390-tools/s390-tools/>`_
+See  `s390-tools <https://github.com/ibm-s390-linux/s390-tools/>`_
 for the tooling.

arch/Kconfig

@@ -295,6 +295,10 @@ config ARCH_32BIT_OFF_T
 	  still support 32-bit off_t. This option is enabled for all such
 	  architectures explicitly.
 
+# Selected by 64 bit architectures which have a 32 bit f_tinode in struct ustat
+config ARCH_32BIT_USTAT_F_TINODE
+	bool
+
 config HAVE_ASM_MODVERSIONS
 	bool
 	help

arch/alpha/Kconfig

@@ -2,6 +2,7 @@
 config ALPHA
 	bool
 	default y
+	select ARCH_32BIT_USTAT_F_TINODE
 	select ARCH_MIGHT_HAVE_PC_PARPORT
 	select ARCH_MIGHT_HAVE_PC_SERIO
 	select ARCH_NO_PREEMPT

arch/s390/Kconfig

@@ -58,6 +58,7 @@ config S390
 	# Note: keep this list sorted alphabetically
 	#
 	imply IMA_SECURE_AND_OR_TRUSTED_BOOT
+	select ARCH_32BIT_USTAT_F_TINODE
 	select ARCH_BINFMT_ELF_STATE
 	select ARCH_HAS_DEBUG_VM_PGTABLE
 	select ARCH_HAS_DEBUG_WX
@@ -123,11 +124,13 @@ config S390
 	select GENERIC_ALLOCATOR
 	select GENERIC_CPU_AUTOPROBE
 	select GENERIC_CPU_VULNERABILITIES
+	select GENERIC_ENTRY
 	select GENERIC_FIND_FIRST_BIT
 	select GENERIC_GETTIMEOFDAY
 	select GENERIC_PTDUMP
 	select GENERIC_SMP_IDLE_THREAD
 	select GENERIC_TIME_VSYSCALL
+	select GENERIC_VDSO_TIME_NS
 	select HAVE_ALIGNED_STRUCT_PAGE if SLUB
 	select HAVE_ARCH_AUDITSYSCALL
 	select HAVE_ARCH_JUMP_LABEL

arch/s390/Kconfig.debug

@@ -6,10 +6,12 @@ config TRACE_IRQFLAGS_SUPPORT
 config EARLY_PRINTK
 	def_bool y
 
-config DEBUG_USER_ASCE
+config DEBUG_ENTRY
-	bool "Debug User ASCE"
+	bool "Debug low-level entry code"
+	depends on DEBUG_KERNEL
 	help
-	  Check on exit to user space that address space control
+	  This option enables sanity checks in s390 low-level entry code.
-	  elements are setup correctly.
+	  Some of these sanity checks may slow down kernel entries and
+	  exits or otherwise impact performance.
 
 	  If unsure, say N.

arch/s390/configs/debug_defconfig

@@ -40,6 +40,7 @@ CONFIG_USERFAULTFD=y
 # CONFIG_COMPAT_BRK is not set
 CONFIG_PROFILING=y
 CONFIG_LIVEPATCH=y
+CONFIG_MARCH_ZEC12=y
 CONFIG_TUNE_ZEC12=y
 CONFIG_NR_CPUS=512
 CONFIG_NUMA=y
@@ -176,13 +177,17 @@ CONFIG_NF_CONNTRACK_TFTP=m
 CONFIG_NF_CT_NETLINK=m
 CONFIG_NF_CT_NETLINK_TIMEOUT=m
 CONFIG_NF_TABLES=m
+CONFIG_NF_TABLES_INET=y
 CONFIG_NFT_CT=m
 CONFIG_NFT_COUNTER=m
 CONFIG_NFT_LOG=m
 CONFIG_NFT_LIMIT=m
 CONFIG_NFT_NAT=m
+CONFIG_NFT_OBJREF=m
+CONFIG_NFT_REJECT=m
 CONFIG_NFT_COMPAT=m
 CONFIG_NFT_HASH=m
+CONFIG_NFT_FIB_INET=m
 CONFIG_NETFILTER_XT_SET=m
 CONFIG_NETFILTER_XT_TARGET_AUDIT=m
 CONFIG_NETFILTER_XT_TARGET_CHECKSUM=m
@@ -274,6 +279,7 @@ CONFIG_IP_VS_NQ=m
 CONFIG_IP_VS_FTP=m
 CONFIG_IP_VS_PE_SIP=m
 CONFIG_NF_TABLES_IPV4=y
+CONFIG_NFT_FIB_IPV4=m
 CONFIG_NF_TABLES_ARP=y
 CONFIG_IP_NF_IPTABLES=m
 CONFIG_IP_NF_MATCH_AH=m
@@ -294,6 +300,7 @@ CONFIG_IP_NF_ARPTABLES=m
 CONFIG_IP_NF_ARPFILTER=m
 CONFIG_IP_NF_ARP_MANGLE=m
 CONFIG_NF_TABLES_IPV6=y
+CONFIG_NFT_FIB_IPV6=m
 CONFIG_IP6_NF_IPTABLES=m
 CONFIG_IP6_NF_MATCH_AH=m
 CONFIG_IP6_NF_MATCH_EUI64=m
@@ -629,7 +636,6 @@ CONFIG_NTFS_RW=y
 CONFIG_PROC_KCORE=y
 CONFIG_TMPFS=y
 CONFIG_TMPFS_POSIX_ACL=y
-CONFIG_TMPFS_INODE64=y
 CONFIG_HUGETLBFS=y
 CONFIG_CONFIGFS_FS=m
 CONFIG_ECRYPT_FS=m
@@ -791,6 +797,8 @@ CONFIG_DEBUG_OBJECTS_RCU_HEAD=y
 CONFIG_DEBUG_OBJECTS_PERCPU_COUNTER=y
 CONFIG_SLUB_DEBUG_ON=y
 CONFIG_SLUB_STATS=y
+CONFIG_DEBUG_KMEMLEAK=y
+CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y
 CONFIG_DEBUG_STACK_USAGE=y
 CONFIG_DEBUG_VM=y
 CONFIG_DEBUG_VM_VMACACHE=y
@@ -831,7 +839,6 @@ CONFIG_BPF_KPROBE_OVERRIDE=y
 CONFIG_HIST_TRIGGERS=y
 CONFIG_FTRACE_STARTUP_TEST=y
 # CONFIG_EVENT_TRACE_STARTUP_TEST is not set
-CONFIG_DEBUG_USER_ASCE=y
 CONFIG_NOTIFIER_ERROR_INJECTION=m
 CONFIG_NETDEV_NOTIFIER_ERROR_INJECT=m
 CONFIG_FAULT_INJECTION=y
@@ -855,3 +862,4 @@ CONFIG_PERCPU_TEST=m
 CONFIG_ATOMIC64_SELFTEST=y
 CONFIG_TEST_BITOPS=m
 CONFIG_TEST_BPF=m
+CONFIG_DEBUG_ENTRY=y

arch/s390/configs/defconfig

@@ -38,6 +38,7 @@ CONFIG_USERFAULTFD=y
 # CONFIG_COMPAT_BRK is not set
 CONFIG_PROFILING=y
 CONFIG_LIVEPATCH=y
+CONFIG_MARCH_ZEC12=y
 CONFIG_TUNE_ZEC12=y
 CONFIG_NR_CPUS=512
 CONFIG_NUMA=y
@@ -167,13 +168,17 @@ CONFIG_NF_CONNTRACK_TFTP=m
 CONFIG_NF_CT_NETLINK=m
 CONFIG_NF_CT_NETLINK_TIMEOUT=m
 CONFIG_NF_TABLES=m
+CONFIG_NF_TABLES_INET=y
 CONFIG_NFT_CT=m
 CONFIG_NFT_COUNTER=m
 CONFIG_NFT_LOG=m
 CONFIG_NFT_LIMIT=m
 CONFIG_NFT_NAT=m
+CONFIG_NFT_OBJREF=m
+CONFIG_NFT_REJECT=m
 CONFIG_NFT_COMPAT=m
 CONFIG_NFT_HASH=m
+CONFIG_NFT_FIB_INET=m
 CONFIG_NETFILTER_XT_SET=m
 CONFIG_NETFILTER_XT_TARGET_AUDIT=m
 CONFIG_NETFILTER_XT_TARGET_CHECKSUM=m
@@ -265,6 +270,7 @@ CONFIG_IP_VS_NQ=m
 CONFIG_IP_VS_FTP=m
 CONFIG_IP_VS_PE_SIP=m
 CONFIG_NF_TABLES_IPV4=y
+CONFIG_NFT_FIB_IPV4=m
 CONFIG_NF_TABLES_ARP=y
 CONFIG_IP_NF_IPTABLES=m
 CONFIG_IP_NF_MATCH_AH=m
@@ -285,6 +291,7 @@ CONFIG_IP_NF_ARPTABLES=m
 CONFIG_IP_NF_ARPFILTER=m
 CONFIG_IP_NF_ARP_MANGLE=m
 CONFIG_NF_TABLES_IPV6=y
+CONFIG_NFT_FIB_IPV6=m
 CONFIG_IP6_NF_IPTABLES=m
 CONFIG_IP6_NF_MATCH_AH=m
 CONFIG_IP6_NF_MATCH_EUI64=m
@@ -617,7 +624,6 @@ CONFIG_NTFS_RW=y
 CONFIG_PROC_KCORE=y
 CONFIG_TMPFS=y
 CONFIG_TMPFS_POSIX_ACL=y
-CONFIG_TMPFS_INODE64=y
 CONFIG_HUGETLBFS=y
 CONFIG_CONFIGFS_FS=m
 CONFIG_ECRYPT_FS=m
@@ -779,7 +785,6 @@ CONFIG_FTRACE_SYSCALLS=y
 CONFIG_BLK_DEV_IO_TRACE=y
 CONFIG_BPF_KPROBE_OVERRIDE=y
 CONFIG_HIST_TRIGGERS=y
-CONFIG_DEBUG_USER_ASCE=y
 CONFIG_LKDTM=m
 CONFIG_PERCPU_TEST=m
 CONFIG_ATOMIC64_SELFTEST=y

arch/s390/configs/zfcpdump_defconfig

@@ -3,11 +3,13 @@ CONFIG_NO_HZ_IDLE=y
 CONFIG_HIGH_RES_TIMERS=y
 # CONFIG_CPU_ISOLATION is not set
 # CONFIG_UTS_NS is not set
+# CONFIG_TIME_NS is not set
 # CONFIG_PID_NS is not set
 # CONFIG_NET_NS is not set
 CONFIG_BLK_DEV_INITRD=y
 CONFIG_CC_OPTIMIZE_FOR_SIZE=y
 # CONFIG_COMPAT_BRK is not set
+CONFIG_MARCH_ZEC12=y
 CONFIG_TUNE_ZEC12=y
 # CONFIG_COMPAT is not set
 CONFIG_NR_CPUS=2

arch/s390/crypto/paes_s390.c

@@ -22,6 +22,7 @@
 #include <linux/init.h>
 #include <linux/mutex.h>
 #include <linux/spinlock.h>
+#include <linux/delay.h>
 #include <crypto/internal/skcipher.h>
 #include <crypto/xts.h>
 #include <asm/cpacf.h>
@@ -128,6 +129,9 @@ static inline int __paes_keyblob2pkey(struct key_blob *kb,
 
 	/* try three times in case of failure */
 	for (i = 0; i < 3; i++) {
+		if (i > 0 && ret == -EAGAIN && in_task())
+			if (msleep_interruptible(1000))
+				return -EINTR;
 		ret = pkey_keyblob2pkey(kb->key, kb->keylen, pk);
 		if (ret == 0)
 			break;
@@ -138,10 +142,12 @@ static inline int __paes_keyblob2pkey(struct key_blob *kb,
 
 static inline int __paes_convert_key(struct s390_paes_ctx *ctx)
 {
+	int ret;
 	struct pkey_protkey pkey;
 
-	if (__paes_keyblob2pkey(&ctx->kb, &pkey))
+	ret = __paes_keyblob2pkey(&ctx->kb, &pkey);
-		return -EINVAL;
+	if (ret)
+		return ret;
 
 	spin_lock_bh(&ctx->pk_lock);
 	memcpy(&ctx->pk, &pkey, sizeof(pkey));
@@ -169,10 +175,12 @@ static void ecb_paes_exit(struct crypto_skcipher *tfm)
 
 static inline int __ecb_paes_set_key(struct s390_paes_ctx *ctx)
 {
+	int rc;
 	unsigned long fc;
 
-	if (__paes_convert_key(ctx))
+	rc = __paes_convert_key(ctx);
-		return -EINVAL;
+	if (rc)
+		return rc;
 
 	/* Pick the correct function code based on the protected key type */
 	fc = (ctx->pk.type == PKEY_KEYTYPE_AES_128) ? CPACF_KM_PAES_128 :
@@ -282,10 +290,12 @@ static void cbc_paes_exit(struct crypto_skcipher *tfm)
 
 static inline int __cbc_paes_set_key(struct s390_paes_ctx *ctx)
 {
+	int rc;
 	unsigned long fc;
 
-	if (__paes_convert_key(ctx))
+	rc = __paes_convert_key(ctx);
-		return -EINVAL;
+	if (rc)
+		return rc;
 
 	/* Pick the correct function code based on the protected key type */
 	fc = (ctx->pk.type == PKEY_KEYTYPE_AES_128) ? CPACF_KMC_PAES_128 :
@@ -577,10 +587,12 @@ static void ctr_paes_exit(struct crypto_skcipher *tfm)
 
 static inline int __ctr_paes_set_key(struct s390_paes_ctx *ctx)
 {
+	int rc;
 	unsigned long fc;
 
-	if (__paes_convert_key(ctx))
+	rc = __paes_convert_key(ctx);
-		return -EINVAL;
+	if (rc)
+		return rc;
 
 	/* Pick the correct function code based on the protected key type */
 	fc = (ctx->pk.type == PKEY_KEYTYPE_AES_128) ? CPACF_KMCTR_PAES_128 :

arch/s390/crypto/prng.c

@@ -414,7 +414,7 @@ static int __init prng_sha512_instantiate(void)
 	}
 
 	/* append the seed by 16 bytes of unique nonce */
-	get_tod_clock_ext(seed + seedlen);
+	store_tod_clock_ext((union tod_clock *)(seed + seedlen));
 	seedlen += 16;
 
 	/* now initial seed of the prno drng */

arch/s390/hypfs/hypfs_diag0c.c

@@ -84,7 +84,7 @@ static int dbfs_diag0c_create(void **data, void **data_free_ptr, size_t *size)
 	if (IS_ERR(diag0c_data))
 		return PTR_ERR(diag0c_data);
 	memset(&diag0c_data->hdr, 0, sizeof(diag0c_data->hdr));
-	get_tod_clock_ext(diag0c_data->hdr.tod_ext);
+	store_tod_clock_ext((union tod_clock *)diag0c_data->hdr.tod_ext);
 	diag0c_data->hdr.len = count * sizeof(struct hypfs_diag0c_entry);
 	diag0c_data->hdr.version = DBFS_D0C_HDR_VERSION;
 	diag0c_data->hdr.count = count;

arch/s390/hypfs/hypfs_vm.c

@@ -234,7 +234,7 @@ failed:
 struct dbfs_d2fc_hdr {
 	u64	len;		/* Length of d2fc buffer without header */
 	u16	version;	/* Version of header */
-	char	tod_ext[STORE_CLOCK_EXT_SIZE]; /* TOD clock for d2fc */
+	union tod_clock tod_ext; /* TOD clock for d2fc */
 	u64	count;		/* Number of VM guests in d2fc buffer */
 	char	reserved[30];
 } __attribute__ ((packed));
@@ -252,7 +252,7 @@ static int dbfs_diag2fc_create(void **data, void **data_free_ptr, size_t *size)
 	d2fc = diag2fc_store(guest_query, &count, sizeof(d2fc->hdr));
 	if (IS_ERR(d2fc))
 		return PTR_ERR(d2fc);
-	get_tod_clock_ext(d2fc->hdr.tod_ext);
+	store_tod_clock_ext(&d2fc->hdr.tod_ext);
 	d2fc->hdr.len = count * sizeof(struct diag2fc_data);
 	d2fc->hdr.version = DBFS_D2FC_HDR_VERSION;
 	d2fc->hdr.count = count;

arch/s390/include/asm/alternative.h

@@ -145,6 +145,22 @@ void apply_alternatives(struct alt_instr *start, struct alt_instr *end);
 	asm_inline volatile(ALTERNATIVE_2(oldinstr, altinstr1, facility1,   \
 				   altinstr2, facility2) ::: "memory")
 
+/* Alternative inline assembly with input. */
+#define alternative_input(oldinstr, newinstr, feature, input...)	\
+	asm_inline volatile (ALTERNATIVE(oldinstr, newinstr, feature)	\
+		: : input)
+
+/* Like alternative_input, but with a single output argument */
+#define alternative_io(oldinstr, altinstr, facility, output, input...)	\
+	asm_inline volatile(ALTERNATIVE(oldinstr, altinstr, facility)	\
+		: output : input)
+
+/* Use this macro if more than one output parameter is needed. */
+#define ASM_OUTPUT2(a...) a
+
+/* Use this macro if clobbers are needed without inputs. */
+#define ASM_NO_INPUT_CLOBBER(clobber...) : clobber
+
 #endif /* __ASSEMBLY__ */
 
 #endif /* _ASM_S390_ALTERNATIVE_H */

arch/s390/include/asm/ap.h

@@ -368,7 +368,7 @@ static inline struct ap_queue_status ap_dqap(ap_qid_t qid,
 #if IS_ENABLED(CONFIG_ZCRYPT)
 void ap_bus_cfg_chg(void);
 #else
-static inline void ap_bus_cfg_chg(void){};
+static inline void ap_bus_cfg_chg(void){}
 #endif
 
 #endif /* _ASM_S390_AP_H_ */

arch/s390/include/asm/atomic.h

@@ -44,16 +44,6 @@ static inline int atomic_fetch_add(int i, atomic_t *v)
 
 static inline void atomic_add(int i, atomic_t *v)
 {
-#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
-	/*
-	 * Order of conditions is important to circumvent gcc 10 bug:
-	 * https://gcc.gnu.org/pipermail/gcc-patches/2020-July/549318.html
-	 */
-	if ((i > -129) && (i < 128) && __builtin_constant_p(i)) {
-		__atomic_add_const(i, &v->counter);
-		return;
-	}
-#endif
 	__atomic_add(i, &v->counter);
 }
 
@@ -115,16 +105,6 @@ static inline s64 atomic64_fetch_add(s64 i, atomic64_t *v)
 
 static inline void atomic64_add(s64 i, atomic64_t *v)
 {
-#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
-	/*
-	 * Order of conditions is important to circumvent gcc 10 bug:
-	 * https://gcc.gnu.org/pipermail/gcc-patches/2020-July/549318.html
-	 */
-	if ((i > -129) && (i < 128) && __builtin_constant_p(i)) {
-		__atomic64_add_const(i, (long *)&v->counter);
-		return;
-	}
-#endif
 	__atomic64_add(i, (long *)&v->counter);
 }
 

arch/s390/include/asm/bitops.h

@@ -61,18 +61,6 @@ static __always_inline void arch_set_bit(unsigned long nr, volatile unsigned lon
 	unsigned long *addr = __bitops_word(nr, ptr);
 	unsigned long mask;
 
-#ifdef CONFIG_HAVE_MARCH_ZEC12_FEATURES
-	if (__builtin_constant_p(nr)) {
-		unsigned char *caddr = __bitops_byte(nr, ptr);
-
-		asm volatile(
-			"oi	%0,%b1\n"
-			: "+Q" (*caddr)
-			: "i" (1 << (nr & 7))
-			: "cc", "memory");
-		return;
-	}
-#endif
 	mask = 1UL << (nr & (BITS_PER_LONG - 1));
 	__atomic64_or(mask, (long *)addr);
 }
@@ -82,18 +70,6 @@ static __always_inline void arch_clear_bit(unsigned long nr, volatile unsigned l
 	unsigned long *addr = __bitops_word(nr, ptr);
 	unsigned long mask;
 
-#ifdef CONFIG_HAVE_MARCH_ZEC12_FEATURES
-	if (__builtin_constant_p(nr)) {
-		unsigned char *caddr = __bitops_byte(nr, ptr);
-
-		asm volatile(
-			"ni	%0,%b1\n"
-			: "+Q" (*caddr)
-			: "i" (~(1 << (nr & 7)))
-			: "cc", "memory");
-		return;
-	}
-#endif
 	mask = ~(1UL << (nr & (BITS_PER_LONG - 1)));
 	__atomic64_and(mask, (long *)addr);
 }
@@ -104,18 +80,6 @@ static __always_inline void arch_change_bit(unsigned long nr,
 	unsigned long *addr = __bitops_word(nr, ptr);
 	unsigned long mask;
 
-#ifdef CONFIG_HAVE_MARCH_ZEC12_FEATURES
-	if (__builtin_constant_p(nr)) {
-		unsigned char *caddr = __bitops_byte(nr, ptr);
-
-		asm volatile(
-			"xi	%0,%b1\n"
-			: "+Q" (*caddr)
-			: "i" (1 << (nr & 7))
-			: "cc", "memory");
-		return;
-	}
-#endif
 	mask = 1UL << (nr & (BITS_PER_LONG - 1));
 	__atomic64_xor(mask, (long *)addr);
 }

arch/s390/include/asm/cputime.h

@@ -35,4 +35,6 @@ u64 arch_cpu_idle_time(int cpu);
 
 #define arch_idle_time(cpu) arch_cpu_idle_time(cpu)
 
+void account_idle_time_irq(void);
+
 #endif /* _S390_CPUTIME_H */

arch/s390/include/asm/elf.h

@@ -233,8 +233,7 @@ extern char elf_platform[];
 do {								\
 	set_personality(PER_LINUX |				\
 		(current->personality & (~PER_MASK)));		\
-	current->thread.sys_call_table =			\
+	current->thread.sys_call_table = sys_call_table;	\
-		(unsigned long) &sys_call_table;		\
 } while (0)
 #else /* CONFIG_COMPAT */
 #define SET_PERSONALITY(ex)					\
@@ -245,11 +244,11 @@ do {								\
 	if ((ex).e_ident[EI_CLASS] == ELFCLASS32) {		\
 		set_thread_flag(TIF_31BIT);			\
 		current->thread.sys_call_table =		\
-			(unsigned long)	&sys_call_table_emu;	\
+			sys_call_table_emu;			\
 	} else {						\
 		clear_thread_flag(TIF_31BIT);			\
 		current->thread.sys_call_table =		\
-			(unsigned long) &sys_call_table;	\
+			sys_call_table;				\
 	}							\
 } while (0)
 #endif /* CONFIG_COMPAT */

arch/s390/include/asm/entry-common.h

@@ -0,0 +1,60 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef ARCH_S390_ENTRY_COMMON_H
+#define ARCH_S390_ENTRY_COMMON_H
+
+#include <linux/sched.h>
+#include <linux/audit.h>
+#include <linux/tracehook.h>
+#include <linux/processor.h>
+#include <linux/uaccess.h>
+#include <asm/fpu/api.h>
+
+#define ARCH_EXIT_TO_USER_MODE_WORK (_TIF_GUARDED_STORAGE | _TIF_PER_TRAP)
+
+void do_per_trap(struct pt_regs *regs);
+void do_syscall(struct pt_regs *regs);
+
+typedef void (*pgm_check_func)(struct pt_regs *regs);
+
+extern pgm_check_func pgm_check_table[128];
+
+#ifdef CONFIG_DEBUG_ENTRY
+static __always_inline void arch_check_user_regs(struct pt_regs *regs)
+{
+	debug_user_asce(0);
+}
+
+#define arch_check_user_regs arch_check_user_regs
+#endif /* CONFIG_DEBUG_ENTRY */
+
+static __always_inline void arch_exit_to_user_mode_work(struct pt_regs *regs,
+							unsigned long ti_work)
+{
+	if (ti_work & _TIF_PER_TRAP) {
+		clear_thread_flag(TIF_PER_TRAP);
+		do_per_trap(regs);
+	}
+
+	if (ti_work & _TIF_GUARDED_STORAGE)
+		gs_load_bc_cb(regs);
+}
+
+#define arch_exit_to_user_mode_work arch_exit_to_user_mode_work
+
+static __always_inline void arch_exit_to_user_mode(void)
+{
+	if (test_cpu_flag(CIF_FPU))
+		__load_fpu_regs();
+
+	if (IS_ENABLED(CONFIG_DEBUG_ENTRY))
+		debug_user_asce(1);
+}
+
+#define arch_exit_to_user_mode arch_exit_to_user_mode
+
+static inline bool on_thread_stack(void)
+{
+	return !(((unsigned long)(current->stack) ^ current_stack_pointer()) & ~(THREAD_SIZE - 1));
+}
+
+#endif

arch/s390/include/asm/fpu/api.h

@@ -47,6 +47,8 @@
 #include <linux/preempt.h>
 
 void save_fpu_regs(void);
+void load_fpu_regs(void);
+void __load_fpu_regs(void);
 
 static inline int test_fp_ctl(u32 fpc)
 {

arch/s390/include/asm/idle.h

@@ -20,11 +20,13 @@ struct s390_idle_data {
 	unsigned long long clock_idle_exit;
 	unsigned long long timer_idle_enter;
 	unsigned long long timer_idle_exit;
+	unsigned long mt_cycles_enter[8];
 };
 
 extern struct device_attribute dev_attr_idle_count;
 extern struct device_attribute dev_attr_idle_time_us;
 
-void psw_idle(struct s390_idle_data *, unsigned long);
+void psw_idle(struct s390_idle_data *data, unsigned long psw_mask);
+void psw_idle_exit(void);
 
 #endif /* _S390_IDLE_H */

arch/s390/include/asm/lowcore.h

@@ -81,8 +81,8 @@ struct lowcore {
 	psw_t	return_mcck_psw;		/* 0x02a0 */
 
 	/* CPU accounting and timing values. */
-	__u64	sync_enter_timer;		/* 0x02b0 */
+	__u64	sys_enter_timer;		/* 0x02b0 */
-	__u64	async_enter_timer;		/* 0x02b8 */
+	__u8	pad_0x02b8[0x02c0-0x02b8];	/* 0x02b8 */
 	__u64	mcck_enter_timer;		/* 0x02c0 */
 	__u64	exit_timer;			/* 0x02c8 */
 	__u64	user_timer;			/* 0x02d0 */
@@ -107,16 +107,15 @@ struct lowcore {
 	__u64	async_stack;			/* 0x0350 */
 	__u64	nodat_stack;			/* 0x0358 */
 	__u64	restart_stack;			/* 0x0360 */
-
+	__u64	mcck_stack;			/* 0x0368 */
 	/* Restart function and parameter. */
-	__u64	restart_fn;			/* 0x0368 */
+	__u64	restart_fn;			/* 0x0370 */
-	__u64	restart_data;			/* 0x0370 */
+	__u64	restart_data;			/* 0x0378 */
-	__u64	restart_source;			/* 0x0378 */
+	__u64	restart_source;			/* 0x0380 */
 
 	/* Address space pointer. */
-	__u64	kernel_asce;			/* 0x0380 */
+	__u64	kernel_asce;			/* 0x0388 */
-	__u64	user_asce;			/* 0x0388 */
+	__u64	user_asce;			/* 0x0390 */
-	__u8	pad_0x0390[0x0398-0x0390];	/* 0x0390 */
 
 	/*
 	 * The lpp and current_pid fields form a

arch/s390/include/asm/nmi.h

@@ -99,6 +99,7 @@ int nmi_alloc_per_cpu(struct lowcore *lc);
 void nmi_free_per_cpu(struct lowcore *lc);
 
 void s390_handle_mcck(void);
+void __s390_handle_mcck(void);
 int s390_do_machine_check(struct pt_regs *regs);
 
 #endif /* __ASSEMBLY__ */

arch/s390/include/asm/pci.h

@@ -201,7 +201,7 @@ extern unsigned int s390_pci_no_rid;
   Prototypes
 ----------------------------------------------------------------------------- */
 /* Base stuff */
-int zpci_create_device(struct zpci_dev *);
+int zpci_create_device(u32 fid, u32 fh, enum zpci_state state);
 void zpci_remove_device(struct zpci_dev *zdev);
 int zpci_enable_device(struct zpci_dev *);
 int zpci_disable_device(struct zpci_dev *);
@@ -212,7 +212,7 @@ void zpci_remove_reserved_devices(void);
 /* CLP */
 int clp_setup_writeback_mio(void);
 int clp_scan_pci_devices(void);
-int clp_add_pci_device(u32, u32, int);
+int clp_query_pci_fn(struct zpci_dev *zdev);
 int clp_enable_fh(struct zpci_dev *, u8);
 int clp_disable_fh(struct zpci_dev *);
 int clp_get_state(u32 fid, enum zpci_state *state);

arch/s390/include/asm/preempt.h

@@ -131,9 +131,9 @@ static inline bool should_resched(int preempt_offset)
 #endif /* CONFIG_HAVE_MARCH_Z196_FEATURES */
 
 #ifdef CONFIG_PREEMPTION
-extern asmlinkage void preempt_schedule(void);
+extern void preempt_schedule(void);
 #define __preempt_schedule() preempt_schedule()
-extern asmlinkage void preempt_schedule_notrace(void);
+extern void preempt_schedule_notrace(void);
 #define __preempt_schedule_notrace() preempt_schedule_notrace()
 #endif /* CONFIG_PREEMPTION */
 

arch/s390/include/asm/processor.h

@@ -38,6 +38,9 @@
 #include <asm/runtime_instr.h>
 #include <asm/fpu/types.h>
 #include <asm/fpu/internal.h>
+#include <asm/irqflags.h>
+
+typedef long (*sys_call_ptr_t)(struct pt_regs *regs);
 
 static inline void set_cpu_flag(int flag)
 {
@@ -101,31 +104,32 @@ extern void __bpon(void);
  */
 struct thread_struct {
 	unsigned int  acrs[NUM_ACRS];
-        unsigned long ksp;              /* kernel stack pointer             */
+	unsigned long ksp;			/* kernel stack pointer */
-	unsigned long user_timer;	/* task cputime in user space */
+	unsigned long user_timer;		/* task cputime in user space */
-	unsigned long guest_timer;	/* task cputime in kvm guest */
+	unsigned long guest_timer;		/* task cputime in kvm guest */
-	unsigned long system_timer;	/* task cputime in kernel space */
+	unsigned long system_timer;		/* task cputime in kernel space */
-	unsigned long hardirq_timer;	/* task cputime in hardirq context */
+	unsigned long hardirq_timer;		/* task cputime in hardirq context */
-	unsigned long softirq_timer;	/* task cputime in softirq context */
+	unsigned long softirq_timer;		/* task cputime in softirq context */
-	unsigned long sys_call_table;	/* system call table address */
+	const sys_call_ptr_t *sys_call_table;	/* system call table address */
-	unsigned long gmap_addr;	/* address of last gmap fault. */
+	unsigned long gmap_addr;		/* address of last gmap fault. */
-	unsigned int gmap_write_flag;	/* gmap fault write indication */
+	unsigned int gmap_write_flag;		/* gmap fault write indication */
-	unsigned int gmap_int_code;	/* int code of last gmap fault */
+	unsigned int gmap_int_code;		/* int code of last gmap fault */
-	unsigned int gmap_pfault;	/* signal of a pending guest pfault */
+	unsigned int gmap_pfault;		/* signal of a pending guest pfault */
+
 	/* Per-thread information related to debugging */
-	struct per_regs per_user;	/* User specified PER registers */
+	struct per_regs per_user;		/* User specified PER registers */
-	struct per_event per_event;	/* Cause of the last PER trap */
+	struct per_event per_event;		/* Cause of the last PER trap */
-	unsigned long per_flags;	/* Flags to control debug behavior */
+	unsigned long per_flags;		/* Flags to control debug behavior */
-	unsigned int system_call;	/* system call number in signal */
+	unsigned int system_call;		/* system call number in signal */
-	unsigned long last_break;	/* last breaking-event-address. */
+	unsigned long last_break;		/* last breaking-event-address. */
-        /* pfault_wait is used to block the process on a pfault event */
+	/* pfault_wait is used to block the process on a pfault event */
 	unsigned long pfault_wait;
 	struct list_head list;
 	/* cpu runtime instrumentation */
 	struct runtime_instr_cb *ri_cb;
-	struct gs_cb *gs_cb;		/* Current guarded storage cb */
+	struct gs_cb *gs_cb;			/* Current guarded storage cb */
-	struct gs_cb *gs_bc_cb;		/* Broadcast guarded storage cb */
+	struct gs_cb *gs_bc_cb;			/* Broadcast guarded storage cb */
-	unsigned char trap_tdb[256];	/* Transaction abort diagnose block */
+	unsigned char trap_tdb[256];		/* Transaction abort diagnose block */
 	/*
 	 * Warning: 'fpu' is dynamically-sized. It *MUST* be at
 	 * the end.
@@ -184,6 +188,7 @@ static inline void release_thread(struct task_struct *tsk) { }
 
 /* Free guarded storage control block */
 void guarded_storage_release(struct task_struct *tsk);
+void gs_load_bc_cb(struct pt_regs *regs);
 
 unsigned long get_wchan(struct task_struct *p);
 #define task_pt_regs(tsk) ((struct pt_regs *) \
@@ -324,6 +329,11 @@ extern void memcpy_absolute(void *, void *, size_t);
 extern int s390_isolate_bp(void);
 extern int s390_isolate_bp_guest(void);
 
+static __always_inline bool regs_irqs_disabled(struct pt_regs *regs)
+{
+	return arch_irqs_disabled_flags(regs->psw.mask);
+}
+
 #endif /* __ASSEMBLY__ */
 
 #endif /* __ASM_S390_PROCESSOR_H */

arch/s390/include/asm/ptrace.h

@@ -11,13 +11,13 @@
 #include <uapi/asm/ptrace.h>
 
 #define PIF_SYSCALL		0	/* inside a system call */
-#define PIF_PER_TRAP		1	/* deliver sigtrap on return to user */
+#define PIF_SYSCALL_RESTART	1	/* restart the current system call */
-#define PIF_SYSCALL_RESTART	2	/* restart the current system call */
+#define PIF_SYSCALL_RET_SET	2	/* return value was set via ptrace */
 #define PIF_GUEST_FAULT		3	/* indicates program check in sie64a */
 
 #define _PIF_SYSCALL		BIT(PIF_SYSCALL)
-#define _PIF_PER_TRAP		BIT(PIF_PER_TRAP)
 #define _PIF_SYSCALL_RESTART	BIT(PIF_SYSCALL_RESTART)
+#define _PIF_SYSCALL_RET_SET	BIT(PIF_SYSCALL_RET_SET)
 #define _PIF_GUEST_FAULT	BIT(PIF_GUEST_FAULT)
 
 #ifndef __ASSEMBLY__
@@ -68,6 +68,9 @@ enum {
 	&(*(struct psw_bits *)(&(__psw)));	\
 }))
 
+#define PGM_INT_CODE_MASK	0x7f
+#define PGM_INT_CODE_PER	0x80
+
 /*
  * The pt_regs struct defines the way the registers are stored on
  * the stack during a system call.

arch/s390/include/asm/qdio.h

@@ -250,17 +250,13 @@ struct slsb {
  * struct qdio_outbuf_state - SBAL related asynchronous operation information
  *   (for communication with upper layer programs)
  *   (only required for use with completion queues)
- * @flags: flags indicating state of buffer
  * @user: pointer to upper layer program's state information related to SBAL
  *        (stored in user1 data of QAOB)
  */
 struct qdio_outbuf_state {
-	u8 flags;
 	void *user;
 };
 
-#define QDIO_OUTBUF_STATE_FLAG_PENDING	0x01
-
 #define CHSC_AC1_INITIATE_INPUTQ	0x80
 
 
@@ -315,6 +311,7 @@ typedef void qdio_handler_t(struct ccw_device *, unsigned int, int,
 #define QDIO_ERROR_GET_BUF_STATE		0x0002
 #define QDIO_ERROR_SET_BUF_STATE		0x0004
 #define QDIO_ERROR_SLSB_STATE			0x0100
+#define QDIO_ERROR_SLSB_PENDING			0x0200
 
 #define QDIO_ERROR_FATAL			0x00ff
 #define QDIO_ERROR_TEMPORARY			0xff00
@@ -336,7 +333,7 @@ typedef void qdio_handler_t(struct ccw_device *, unsigned int, int,
  * @no_output_qs: number of output queues
  * @input_handler: handler to be called for input queues
  * @output_handler: handler to be called for output queues
- * @irq_poll: Data IRQ polling handler (NULL when not supported)
+ * @irq_poll: Data IRQ polling handler
  * @scan_threshold: # of in-use buffers that triggers scan on output queue
  * @int_parm: interruption parameter
  * @input_sbal_addr_array:  per-queue array, each element points to 128 SBALs

arch/s390/include/asm/scsw.h

@@ -525,8 +525,7 @@ static inline int scsw_cmd_is_valid_pno(union scsw *scsw)
 	return (scsw->cmd.fctl != 0) &&
 	       (scsw->cmd.stctl & SCSW_STCTL_STATUS_PEND) &&
 	       (!(scsw->cmd.stctl & SCSW_STCTL_INTER_STATUS) ||
-		 ((scsw->cmd.stctl & SCSW_STCTL_INTER_STATUS) &&
+		  (scsw->cmd.actl & SCSW_ACTL_SUSPENDED));
-		  (scsw->cmd.actl & SCSW_ACTL_SUSPENDED)));
 }
 
 /**

arch/s390/include/asm/syscall.h

@@ -14,8 +14,8 @@
 #include <linux/err.h>
 #include <asm/ptrace.h>
 
-extern const unsigned long sys_call_table[];
+extern const sys_call_ptr_t sys_call_table[];
-extern const unsigned long sys_call_table_emu[];
+extern const sys_call_ptr_t sys_call_table_emu[];
 
 static inline long syscall_get_nr(struct task_struct *task,
 				  struct pt_regs *regs)
@@ -56,6 +56,7 @@ static inline void syscall_set_return_value(struct task_struct *task,
 					    struct pt_regs *regs,
 					    int error, long val)
 {
+	set_pt_regs_flag(regs, PIF_SYSCALL_RET_SET);
 	regs->gprs[2] = error ? error : val;
 }
 
@@ -97,4 +98,10 @@ static inline int syscall_get_arch(struct task_struct *task)
 #endif
 	return AUDIT_ARCH_S390X;
 }
+
+static inline bool arch_syscall_is_vdso_sigreturn(struct pt_regs *regs)
+{
+	return false;
+}
+
 #endif	/* _ASM_SYSCALL_H */

arch/s390/include/asm/syscall_wrapper.h

@@ -7,6 +7,33 @@
 #ifndef _ASM_S390_SYSCALL_WRAPPER_H
 #define _ASM_S390_SYSCALL_WRAPPER_H
 
+#define __SC_TYPE(t, a) t
+
+#define SYSCALL_PT_ARG6(regs, m, t1, t2, t3, t4, t5, t6)\
+	SYSCALL_PT_ARG5(regs, m, t1, t2, t3, t4, t5),	\
+		m(t6, (regs->gprs[7]))
+
+#define SYSCALL_PT_ARG5(regs, m, t1, t2, t3, t4, t5)	\
+	SYSCALL_PT_ARG4(regs, m, t1, t2, t3, t4),	\
+		m(t5, (regs->gprs[6]))
+
+#define SYSCALL_PT_ARG4(regs, m, t1, t2, t3, t4)	\
+	SYSCALL_PT_ARG3(regs, m, t1, t2, t3),		\
+		m(t4, (regs->gprs[5]))
+
+#define SYSCALL_PT_ARG3(regs, m, t1, t2, t3)		\
+	SYSCALL_PT_ARG2(regs, m, t1, t2),		\
+		m(t3, (regs->gprs[4]))
+
+#define SYSCALL_PT_ARG2(regs, m, t1, t2)		\
+	SYSCALL_PT_ARG1(regs, m, t1),			\
+		m(t2, (regs->gprs[3]))
+
+#define SYSCALL_PT_ARG1(regs, m, t1)			\
+		m(t1, (regs->orig_gpr2))
+
+#define SYSCALL_PT_ARGS(x, ...) SYSCALL_PT_ARG##x(__VA_ARGS__)
+
 #ifdef CONFIG_COMPAT
 #define __SC_COMPAT_TYPE(t, a) \
 	__typeof(__builtin_choose_expr(sizeof(t) > 4, 0L, (t)0)) a
@@ -29,14 +56,15 @@
 	(t)__ReS;							\
 })
 
-#define __S390_SYS_STUBx(x, name, ...)					\
+#define __S390_SYS_STUBx(x, name, ...)						\
-	asmlinkage long __s390_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__));\
+	long __s390_sys##name(struct pt_regs *regs);				\
-	ALLOW_ERROR_INJECTION(__s390_sys##name, ERRNO);			\
+	ALLOW_ERROR_INJECTION(__s390_sys##name, ERRNO);				\
-	asmlinkage long __s390_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__))\
+	long __s390_sys##name(struct pt_regs *regs)				\
-	{								\
+	{									\
-		long ret = __s390x_sys##name(__MAP(x,__SC_COMPAT_CAST,__VA_ARGS__));\
+		long ret = __do_sys##name(SYSCALL_PT_ARGS(x, regs,		\
-		__MAP(x,__SC_TEST,__VA_ARGS__);				\
+			__SC_COMPAT_CAST, __MAP(x, __SC_TYPE, __VA_ARGS__)));	\
-		return ret;						\
+		__MAP(x,__SC_TEST,__VA_ARGS__);					\
+		return ret;							\
 	}
 
 /*
@@ -45,17 +73,17 @@
  */
 #define COMPAT_SYSCALL_DEFINE0(sname)					\
 	SYSCALL_METADATA(_##sname, 0);					\
-	asmlinkage long __s390_compat_sys_##sname(void);		\
+	long __s390_compat_sys_##sname(void);				\
 	ALLOW_ERROR_INJECTION(__s390_compat_sys_##sname, ERRNO);	\
-	asmlinkage long __s390_compat_sys_##sname(void)
+	long __s390_compat_sys_##sname(void)
 
 #define SYSCALL_DEFINE0(sname)						\
 	SYSCALL_METADATA(_##sname, 0);					\
-	asmlinkage long __s390x_sys_##sname(void);			\
+	long __s390x_sys_##sname(void);					\
 	ALLOW_ERROR_INJECTION(__s390x_sys_##sname, ERRNO);		\
-	asmlinkage long __s390_sys_##sname(void)			\
+	long __s390_sys_##sname(void)					\
 		__attribute__((alias(__stringify(__s390x_sys_##sname)))); \
-	asmlinkage long __s390x_sys_##sname(void)
+	long __s390x_sys_##sname(void)
 
 #define COND_SYSCALL(name)						\
 	cond_syscall(__s390x_sys_##name);				\
@@ -65,23 +93,24 @@
 	SYSCALL_ALIAS(__s390x_sys_##name, sys_ni_posix_timers);		\
 	SYSCALL_ALIAS(__s390_sys_##name, sys_ni_posix_timers)
 
-#define COMPAT_SYSCALL_DEFINEx(x, name, ...)					\
+#define COMPAT_SYSCALL_DEFINEx(x, name, ...)						\
-	__diag_push();								\
+	__diag_push();									\
-	__diag_ignore(GCC, 8, "-Wattribute-alias",				\
+	__diag_ignore(GCC, 8, "-Wattribute-alias",					\
-		      "Type aliasing is used to sanitize syscall arguments");\
+		      "Type aliasing is used to sanitize syscall arguments");		\
-	asmlinkage long __s390_compat_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__));	\
+	long __s390_compat_sys##name(struct pt_regs *regs);				\
-	asmlinkage long __s390_compat_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__))	\
+	long __s390_compat_sys##name(struct pt_regs *regs)				\
-		__attribute__((alias(__stringify(__se_compat_sys##name))));	\
+		__attribute__((alias(__stringify(__se_compat_sys##name))));		\
-	ALLOW_ERROR_INJECTION(__s390_compat_sys##name, ERRNO);			\
+	ALLOW_ERROR_INJECTION(__s390_compat_sys##name, ERRNO);				\
-	static inline long __do_compat_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__));\
+	static inline long __do_compat_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__));	\
-	asmlinkage long __se_compat_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__));	\
+	long __se_compat_sys##name(struct pt_regs *regs);				\
-	asmlinkage long __se_compat_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__))	\
+	long __se_compat_sys##name(struct pt_regs *regs)				\
-	{									\
+	{										\
-		long ret = __do_compat_sys##name(__MAP(x,__SC_DELOUSE,__VA_ARGS__));\
+		long ret = __do_compat_sys##name(SYSCALL_PT_ARGS(x, regs, __SC_DELOUSE,	\
-		__MAP(x,__SC_TEST,__VA_ARGS__);					\
+						 __MAP(x, __SC_TYPE, __VA_ARGS__)));	\
-		return ret;							\
+		__MAP(x,__SC_TEST,__VA_ARGS__);						\
-	}									\
+		return ret;								\
-	__diag_pop();								\
+	}										\
+	__diag_pop();									\
 	static inline long __do_compat_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__))
 
 /*
@@ -101,9 +130,9 @@
 
 #define SYSCALL_DEFINE0(sname)						\
 	SYSCALL_METADATA(_##sname, 0);					\
-	asmlinkage long __s390x_sys_##sname(void);			\
+	long __s390x_sys_##sname(void);					\
 	ALLOW_ERROR_INJECTION(__s390x_sys_##sname, ERRNO);		\
-	asmlinkage long __s390x_sys_##sname(void)
+	long __s390x_sys_##sname(void)
 
 #define COND_SYSCALL(name)						\
 	cond_syscall(__s390x_sys_##name)
@@ -113,23 +142,24 @@
 
 #endif /* CONFIG_COMPAT */
 
-#define __SYSCALL_DEFINEx(x, name, ...)						\
+#define __SYSCALL_DEFINEx(x, name, ...)							\
-	__diag_push();								\
+	__diag_push();									\
-	__diag_ignore(GCC, 8, "-Wattribute-alias",				\
+	__diag_ignore(GCC, 8, "-Wattribute-alias",					\
-		      "Type aliasing is used to sanitize syscall arguments");\
+		      "Type aliasing is used to sanitize syscall arguments");		\
-	asmlinkage long __s390x_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__))	\
+	long __s390x_sys##name(struct pt_regs *regs)					\
-		__attribute__((alias(__stringify(__se_sys##name))));		\
+		__attribute__((alias(__stringify(__se_sys##name))));			\
-	ALLOW_ERROR_INJECTION(__s390x_sys##name, ERRNO);			\
+	ALLOW_ERROR_INJECTION(__s390x_sys##name, ERRNO);				\
-	long __se_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__));			\
+	static inline long __do_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__));		\
-	static inline long __do_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__));	\
+	long __se_sys##name(struct pt_regs *regs);					\
-	__S390_SYS_STUBx(x, name, __VA_ARGS__)					\
+	__S390_SYS_STUBx(x, name, __VA_ARGS__)						\
-	asmlinkage long __se_sys##name(__MAP(x,__SC_LONG,__VA_ARGS__))		\
+	long __se_sys##name(struct pt_regs *regs)					\
-	{									\
+	{										\
-		long ret = __do_sys##name(__MAP(x,__SC_CAST,__VA_ARGS__));	\
+		long ret = __do_sys##name(SYSCALL_PT_ARGS(x, regs,			\
-		__MAP(x,__SC_TEST,__VA_ARGS__);					\
+				    __SC_CAST, __MAP(x, __SC_TYPE, __VA_ARGS__)));	\
-		return ret;							\
+		__MAP(x,__SC_TEST,__VA_ARGS__);						\
-	}									\
+		return ret;								\
-	__diag_pop();								\
+	}										\
+	__diag_pop();									\
 	static inline long __do_sys##name(__MAP(x,__SC_DECL,__VA_ARGS__))
 
 #endif /* _ASM_X86_SYSCALL_WRAPPER_H */

arch/s390/include/asm/thread_info.h

@@ -36,6 +36,7 @@
  */
 struct thread_info {
 	unsigned long		flags;		/* low level flags */
+	unsigned long		syscall_work;	/* SYSCALL_WORK_ flags */
 };
 
 /*
@@ -46,6 +47,8 @@ struct thread_info {
 	.flags		= 0,			\
 }
 
+struct task_struct;
+
 void arch_release_task_struct(struct task_struct *tsk);
 int arch_dup_task_struct(struct task_struct *dst, struct task_struct *src);
 
@@ -68,6 +71,7 @@ void arch_setup_new_exec(void);
 #define TIF_NOTIFY_SIGNAL	7	/* signal notifications exist */
 #define TIF_ISOLATE_BP		8	/* Run process with isolated BP */
 #define TIF_ISOLATE_BP_GUEST	9	/* Run KVM guests with isolated BP */
+#define TIF_PER_TRAP		10	/* Need to handle PER trap on exit to usermode */
 
 #define TIF_31BIT		16	/* 32bit process */
 #define TIF_MEMDIE		17	/* is terminating due to OOM killer */
@@ -91,6 +95,7 @@ void arch_setup_new_exec(void);
 #define _TIF_PATCH_PENDING	BIT(TIF_PATCH_PENDING)
 #define _TIF_ISOLATE_BP		BIT(TIF_ISOLATE_BP)
 #define _TIF_ISOLATE_BP_GUEST	BIT(TIF_ISOLATE_BP_GUEST)
+#define _TIF_PER_TRAP		BIT(TIF_PER_TRAP)
 
 #define _TIF_31BIT		BIT(TIF_31BIT)
 #define _TIF_SINGLE_STEP	BIT(TIF_SINGLE_STEP)

arch/s390/include/asm/timex.h

@@ -19,6 +19,25 @@
 
 extern u64 clock_comparator_max;
 
+union tod_clock {
+	__uint128_t val;
+	struct {
+		__uint128_t ei	:  8; /* epoch index */
+		__uint128_t tod : 64; /* bits 0-63 of tod clock */
+		__uint128_t	: 40;
+		__uint128_t pf	: 16; /* programmable field */
+	};
+	struct {
+		__uint128_t eitod : 72; /* epoch index + bits 0-63 tod clock */
+		__uint128_t	  : 56;
+	};
+	struct {
+		__uint128_t us	: 60; /* micro-seconds */
+		__uint128_t sus	: 12; /* sub-microseconds */
+		__uint128_t	: 56;
+	};
+} __packed;
+
 /* Inline functions for clock register access. */
 static inline int set_tod_clock(__u64 time)
 {
@@ -32,18 +51,23 @@ static inline int set_tod_clock(__u64 time)
 	return cc;
 }
 
-static inline int store_tod_clock(__u64 *time)
+static inline int store_tod_clock_ext_cc(union tod_clock *clk)
 {
 	int cc;
 
 	asm volatile(
-		"   stck  %1\n"
+		"   stcke  %1\n"
 		"   ipm   %0\n"
 		"   srl   %0,28\n"
-		: "=d" (cc), "=Q" (*time) : : "cc");
+		: "=d" (cc), "=Q" (*clk) : : "cc");
 	return cc;
 }
 
+static inline void store_tod_clock_ext(union tod_clock *tod)
+{
+	asm volatile("stcke %0" : "=Q" (*tod) : : "cc");
+}
+
 static inline void set_clock_comparator(__u64 time)
 {
 	asm volatile("sckc %0" : : "Q" (time));
@@ -144,23 +168,15 @@ static inline void local_tick_enable(unsigned long long comp)
 }
 
 #define CLOCK_TICK_RATE		1193180 /* Underlying HZ */
-#define STORE_CLOCK_EXT_SIZE	16	/* stcke writes 16 bytes */
 
 typedef unsigned long long cycles_t;
 
-static inline void get_tod_clock_ext(char *clk)
-{
-	typedef struct { char _[STORE_CLOCK_EXT_SIZE]; } addrtype;
-
-	asm volatile("stcke %0" : "=Q" (*(addrtype *) clk) : : "cc");
-}
-
 static inline unsigned long long get_tod_clock(void)
 {
-	char clk[STORE_CLOCK_EXT_SIZE];
+	union tod_clock clk;
 
-	get_tod_clock_ext(clk);
+	store_tod_clock_ext(&clk);
-	return *((unsigned long long *)&clk[1]);
+	return clk.tod;
 }
 
 static inline unsigned long long get_tod_clock_fast(void)
@@ -183,7 +199,7 @@ static inline cycles_t get_cycles(void)
 int get_phys_clock(unsigned long *clock);
 void init_cpu_timer(void);
 
-extern unsigned char tod_clock_base[16] __aligned(8);
+extern union tod_clock tod_clock_base;
 
 /**
  * get_clock_monotonic - returns current time in clock rate units
@@ -197,7 +213,7 @@ static inline unsigned long long get_tod_clock_monotonic(void)
 	unsigned long long tod;
 
 	preempt_disable_notrace();
-	tod = get_tod_clock() - *(unsigned long long *) &tod_clock_base[1];
+	tod = get_tod_clock() - tod_clock_base.tod;
 	preempt_enable_notrace();
 	return tod;
 }

arch/s390/include/asm/tlb.h

@@ -66,7 +66,7 @@ static inline void pte_free_tlb(struct mmu_gather *tlb, pgtable_t pte,
 	__tlb_adjust_range(tlb, address, PAGE_SIZE);
 	tlb->mm->context.flush_mm = 1;
 	tlb->freed_tables = 1;
-	tlb->cleared_ptes = 1;
+	tlb->cleared_pmds = 1;
 	/*
 	 * page_table_free_rcu takes care of the allocation bit masks
 	 * of the 2K table fragments in the 4K page table page,
@@ -110,7 +110,6 @@ static inline void p4d_free_tlb(struct mmu_gather *tlb, p4d_t *p4d,
 	__tlb_adjust_range(tlb, address, PAGE_SIZE);
 	tlb->mm->context.flush_mm = 1;
 	tlb->freed_tables = 1;
-	tlb->cleared_p4ds = 1;
 	tlb_remove_table(tlb, p4d);
 }
 
@@ -128,7 +127,7 @@ static inline void pud_free_tlb(struct mmu_gather *tlb, pud_t *pud,
 		return;
 	tlb->mm->context.flush_mm = 1;
 	tlb->freed_tables = 1;
-	tlb->cleared_puds = 1;
+	tlb->cleared_p4ds = 1;
 	tlb_remove_table(tlb, pud);
 }
 

arch/s390/include/asm/uaccess.h

@@ -18,7 +18,7 @@
 #include <asm/extable.h>
 #include <asm/facility.h>
 
-void debug_user_asce(void);
+void debug_user_asce(int exit);
 
 static inline int __range_ok(unsigned long addr, unsigned long size)
 {

arch/s390/include/asm/vdso.h

@@ -4,17 +4,18 @@
 
 #include <vdso/datapage.h>
 
-/* Default link addresses for the vDSOs */
+/* Default link address for the vDSO */
-#define VDSO32_LBASE	0
 #define VDSO64_LBASE	0
 
+#define __VVAR_PAGES	2
+
 #define VDSO_VERSION_STRING	LINUX_2.6.29
 
 #ifndef __ASSEMBLY__
 
 extern struct vdso_data *vdso_data;
 
-void vdso_getcpu_init(void);
+int vdso_getcpu_init(void);
 
 #endif /* __ASSEMBLY__ */
 #endif /* __S390_VDSO_H__ */

arch/s390/include/asm/vdso/gettimeofday.h

@@ -24,13 +24,12 @@ static __always_inline const struct vdso_data *__arch_get_vdso_data(void)
 
 static inline u64 __arch_get_hw_counter(s32 clock_mode, const struct vdso_data *vd)
 {
-	const struct vdso_data *vdso = __arch_get_vdso_data();
 	u64 adj, now;
 
 	now = get_tod_clock();
-	adj = vdso->arch_data.tod_steering_end - now;
+	adj = vd->arch_data.tod_steering_end - now;
 	if (unlikely((s64) adj > 0))
-		now += (vdso->arch_data.tod_steering_delta < 0) ? (adj >> 15) : -(adj >> 15);
+		now += (vd->arch_data.tod_steering_delta < 0) ? (adj >> 15) : -(adj >> 15);
 	return now;
 }
 
@@ -68,4 +67,11 @@ long clock_getres_fallback(clockid_t clkid, struct __kernel_timespec *ts)
 	return r2;
 }
 
+#ifdef CONFIG_TIME_NS
+static __always_inline const struct vdso_data *__arch_get_timens_vdso_data(void)
+{
+	return _timens_data;
+}
+#endif
+
 #endif

arch/s390/include/asm/vtime.h

@@ -4,4 +4,18 @@
 
 #define __ARCH_HAS_VTIME_TASK_SWITCH
 
+static inline void update_timer_sys(void)
+{
+	S390_lowcore.system_timer += S390_lowcore.last_update_timer - S390_lowcore.exit_timer;
+	S390_lowcore.user_timer += S390_lowcore.exit_timer - S390_lowcore.sys_enter_timer;
+	S390_lowcore.last_update_timer = S390_lowcore.sys_enter_timer;
+}
+
+static inline void update_timer_mcck(void)
+{
+	S390_lowcore.system_timer += S390_lowcore.last_update_timer - S390_lowcore.exit_timer;
+	S390_lowcore.user_timer += S390_lowcore.exit_timer - S390_lowcore.mcck_enter_timer;
+	S390_lowcore.last_update_timer = S390_lowcore.mcck_enter_timer;
+}
+
 #endif /* _S390_VTIME_H */

arch/s390/include/uapi/asm/ptrace.h

@@ -179,8 +179,9 @@
 #define ACR_SIZE	4
 
 
-#define PTRACE_OLDSETOPTIONS	     21
+#define PTRACE_OLDSETOPTIONS		21
-
+#define PTRACE_SYSEMU			31
+#define PTRACE_SYSEMU_SINGLESTEP	32
 #ifndef __ASSEMBLY__
 #include <linux/stddef.h>
 #include <linux/types.h>

arch/s390/kernel/Makefile

@@ -34,7 +34,7 @@ CFLAGS_dumpstack.o	+= -fno-optimize-sibling-calls
 CFLAGS_unwind_bc.o	+= -fno-optimize-sibling-calls
 
 obj-y	:= traps.o time.o process.o base.o early.o setup.o idle.o vtime.o
-obj-y	+= processor.o sys_s390.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o
+obj-y	+= processor.o syscall.o ptrace.o signal.o cpcmd.o ebcdic.o nmi.o
 obj-y	+= debug.o irq.o ipl.o dis.o diag.o vdso.o
 obj-y	+= sysinfo.o lgr.o os_info.o machine_kexec.o pgm_check.o
 obj-y	+= runtime_instr.o cache.o fpu.o dumpstack.o guarded_storage.o sthyi.o

arch/s390/kernel/asm-offsets.c

@@ -26,26 +26,14 @@ int main(void)
 	BLANK();
 	/* thread struct offsets */
 	OFFSET(__THREAD_ksp, thread_struct, ksp);
-	OFFSET(__THREAD_sysc_table,  thread_struct, sys_call_table);
-	OFFSET(__THREAD_last_break, thread_struct, last_break);
-	OFFSET(__THREAD_FPU_fpc, thread_struct, fpu.fpc);
-	OFFSET(__THREAD_FPU_regs, thread_struct, fpu.regs);
-	OFFSET(__THREAD_per_cause, thread_struct, per_event.cause);
-	OFFSET(__THREAD_per_address, thread_struct, per_event.address);
-	OFFSET(__THREAD_per_paid, thread_struct, per_event.paid);
-	OFFSET(__THREAD_trap_tdb, thread_struct, trap_tdb);
 	BLANK();
 	/* thread info offsets */
 	OFFSET(__TI_flags, task_struct, thread_info.flags);
 	BLANK();
 	/* pt_regs offsets */
-	OFFSET(__PT_ARGS, pt_regs, args);
 	OFFSET(__PT_PSW, pt_regs, psw);
 	OFFSET(__PT_GPRS, pt_regs, gprs);
 	OFFSET(__PT_ORIG_GPR2, pt_regs, orig_gpr2);
-	OFFSET(__PT_INT_CODE, pt_regs, int_code);
-	OFFSET(__PT_INT_PARM, pt_regs, int_parm);
-	OFFSET(__PT_INT_PARM_LONG, pt_regs, int_parm_long);
 	OFFSET(__PT_FLAGS, pt_regs, flags);
 	OFFSET(__PT_CR1, pt_regs, cr1);
 	DEFINE(__PT_SIZE, sizeof(struct pt_regs));
@@ -64,6 +52,7 @@ int main(void)
 	OFFSET(__CLOCK_IDLE_EXIT, s390_idle_data, clock_idle_exit);
 	OFFSET(__TIMER_IDLE_ENTER, s390_idle_data, timer_idle_enter);
 	OFFSET(__TIMER_IDLE_EXIT, s390_idle_data, timer_idle_exit);
+	OFFSET(__MT_CYCLES_ENTER, s390_idle_data, mt_cycles_enter);
 	BLANK();
 	/* hardware defined lowcore locations 0x000 - 0x1ff */
 	OFFSET(__LC_EXT_PARAMS, lowcore, ext_params);
@@ -115,13 +104,9 @@ int main(void)
 	OFFSET(__LC_CPU_FLAGS, lowcore, cpu_flags);
 	OFFSET(__LC_RETURN_PSW, lowcore, return_psw);
 	OFFSET(__LC_RETURN_MCCK_PSW, lowcore, return_mcck_psw);
-	OFFSET(__LC_SYNC_ENTER_TIMER, lowcore, sync_enter_timer);
+	OFFSET(__LC_SYS_ENTER_TIMER, lowcore, sys_enter_timer);
-	OFFSET(__LC_ASYNC_ENTER_TIMER, lowcore, async_enter_timer);
 	OFFSET(__LC_MCCK_ENTER_TIMER, lowcore, mcck_enter_timer);
 	OFFSET(__LC_EXIT_TIMER, lowcore, exit_timer);
-	OFFSET(__LC_USER_TIMER, lowcore, user_timer);
-	OFFSET(__LC_SYSTEM_TIMER, lowcore, system_timer);
-	OFFSET(__LC_STEAL_TIMER, lowcore, steal_timer);
 	OFFSET(__LC_LAST_UPDATE_TIMER, lowcore, last_update_timer);
 	OFFSET(__LC_LAST_UPDATE_CLOCK, lowcore, last_update_clock);
 	OFFSET(__LC_INT_CLOCK, lowcore, int_clock);
@@ -133,6 +118,7 @@ int main(void)
 	OFFSET(__LC_ASYNC_STACK, lowcore, async_stack);
 	OFFSET(__LC_NODAT_STACK, lowcore, nodat_stack);
 	OFFSET(__LC_RESTART_STACK, lowcore, restart_stack);
+	OFFSET(__LC_MCCK_STACK, lowcore, mcck_stack);
 	OFFSET(__LC_RESTART_FN, lowcore, restart_fn);
 	OFFSET(__LC_RESTART_DATA, lowcore, restart_data);
 	OFFSET(__LC_RESTART_SOURCE, lowcore, restart_source);

arch/s390/kernel/compat_signal.c

@@ -118,6 +118,7 @@ static int restore_sigregs32(struct pt_regs *regs,_sigregs32 __user *sregs)
 	fpregs_load((_s390_fp_regs *) &user_sregs.fpregs, &current->thread.fpu);
 
 	clear_pt_regs_flag(regs, PIF_SYSCALL); /* No longer in a system call */
+	clear_pt_regs_flag(regs, PIF_SYSCALL_RESTART);
 	return 0;
 }
 

arch/s390/kernel/debug.c

@@ -829,11 +829,11 @@ static inline debug_entry_t *get_active_entry(debug_info_t *id)
 static inline void debug_finish_entry(debug_info_t *id, debug_entry_t *active,
 				      int level, int exception)
 {
-	unsigned char clk[STORE_CLOCK_EXT_SIZE];
 	unsigned long timestamp;
+	union tod_clock clk;
 
-	get_tod_clock_ext(clk);
+	store_tod_clock_ext(&clk);
-	timestamp = *(unsigned long *) &clk[0] >> 4;
+	timestamp = clk.us;
 	timestamp -= TOD_UNIX_EPOCH >> 12;
 	active->clock = timestamp;
 	active->cpu = smp_processor_id();

arch/s390/kernel/early.c

@@ -35,16 +35,16 @@
 
 static void __init reset_tod_clock(void)
 {
-	u64 time;
+	union tod_clock clk;
 
-	if (store_tod_clock(&time) == 0)
+	if (store_tod_clock_ext_cc(&clk) == 0)
 		return;
 	/* TOD clock not running. Set the clock to Unix Epoch. */
-	if (set_tod_clock(TOD_UNIX_EPOCH) != 0 || store_tod_clock(&time) != 0)
+	if (set_tod_clock(TOD_UNIX_EPOCH) || store_tod_clock_ext_cc(&clk))
 		disabled_wait();
 
-	memset(tod_clock_base, 0, 16);
+	memset(&tod_clock_base, 0, sizeof(tod_clock_base));
-	*(__u64 *) &tod_clock_base[1] = TOD_UNIX_EPOCH;
+	tod_clock_base.tod = TOD_UNIX_EPOCH;
 	S390_lowcore.last_update_clock = TOD_UNIX_EPOCH;
 }
 
@@ -230,7 +230,7 @@ static __init void detect_machine_facilities(void)
 	}
 	if (test_facility(133))
 		S390_lowcore.machine_flags |= MACHINE_FLAG_GS;
-	if (test_facility(139) && (tod_clock_base[1] & 0x80)) {
+	if (test_facility(139) && (tod_clock_base.tod >> 63)) {
 		/* Enabled signed clock comparator comparisons */
 		S390_lowcore.machine_flags |= MACHINE_FLAG_SCC;
 		clock_comparator_max = -1ULL >> 1;

arch/s390/kernel/entry.h

@@ -17,8 +17,9 @@ void io_int_handler(void);
 void mcck_int_handler(void);
 void restart_int_handler(void);
 
-asmlinkage long do_syscall_trace_enter(struct pt_regs *regs);
+void __ret_from_fork(struct task_struct *prev, struct pt_regs *regs);
-asmlinkage void do_syscall_trace_exit(struct pt_regs *regs);
+void __do_pgm_check(struct pt_regs *regs);
+void __do_syscall(struct pt_regs *regs, int per_trap);
 
 void do_protection_exception(struct pt_regs *regs);
 void do_dat_exception(struct pt_regs *regs);
@@ -48,9 +49,7 @@ void translation_exception(struct pt_regs *regs);
 void vector_exception(struct pt_regs *regs);
 void monitor_event_exception(struct pt_regs *regs);
 
-void do_per_trap(struct pt_regs *regs);
 void do_report_trap(struct pt_regs *regs, int si_signo, int si_code, char *str);
-void syscall_trace(struct pt_regs *regs, int entryexit);
 void kernel_stack_overflow(struct pt_regs * regs);
 void do_signal(struct pt_regs *regs);
 void handle_signal32(struct ksignal *ksig, sigset_t *oldset,
@@ -58,7 +57,8 @@ void handle_signal32(struct ksignal *ksig, sigset_t *oldset,
 void do_notify_resume(struct pt_regs *regs);
 
 void __init init_IRQ(void);
-void do_IRQ(struct pt_regs *regs, int irq);
+void do_io_irq(struct pt_regs *regs);
+void do_ext_irq(struct pt_regs *regs);
 void do_restart(void);
 void __init startup_init(void);
 void die(struct pt_regs *regs, const char *str);
@@ -82,8 +82,6 @@ long sys_s390_sthyi(unsigned long function_code, void __user *buffer, u64 __user
 
 DECLARE_PER_CPU(u64, mt_cycles[8]);
 
-void gs_load_bc_cb(struct pt_regs *regs);
-
 unsigned long stack_alloc(void);
 void stack_free(unsigned long stack);
 

arch/s390/kernel/fpu.c

@@ -175,3 +175,91 @@ void __kernel_fpu_end(struct kernel_fpu *state, u32 flags)
 		: "1", "cc");
 }
 EXPORT_SYMBOL(__kernel_fpu_end);
+
+void __load_fpu_regs(void)
+{
+	struct fpu *state = &current->thread.fpu;
+	unsigned long *regs = current->thread.fpu.regs;
+
+	asm volatile("lfpc %0" : : "Q" (state->fpc));
+	if (likely(MACHINE_HAS_VX)) {
+		asm volatile("lgr	1,%0\n"
+			     "VLM	0,15,0,1\n"
+			     "VLM	16,31,256,1\n"
+			     :
+			     : "d" (regs)
+			     : "1", "cc", "memory");
+	} else {
+		asm volatile("ld 0,%0" : : "Q" (regs[0]));
+		asm volatile("ld 1,%0" : : "Q" (regs[1]));
+		asm volatile("ld 2,%0" : : "Q" (regs[2]));
+		asm volatile("ld 3,%0" : : "Q" (regs[3]));
+		asm volatile("ld 4,%0" : : "Q" (regs[4]));
+		asm volatile("ld 5,%0" : : "Q" (regs[5]));
+		asm volatile("ld 6,%0" : : "Q" (regs[6]));
+		asm volatile("ld 7,%0" : : "Q" (regs[7]));
+		asm volatile("ld 8,%0" : : "Q" (regs[8]));
+		asm volatile("ld 9,%0" : : "Q" (regs[9]));
+		asm volatile("ld 10,%0" : : "Q" (regs[10]));
+		asm volatile("ld 11,%0" : : "Q" (regs[11]));
+		asm volatile("ld 12,%0" : : "Q" (regs[12]));
+		asm volatile("ld 13,%0" : : "Q" (regs[13]));
+		asm volatile("ld 14,%0" : : "Q" (regs[14]));
+		asm volatile("ld 15,%0" : : "Q" (regs[15]));
+	}
+	clear_cpu_flag(CIF_FPU);
+}
+EXPORT_SYMBOL(__load_fpu_regs);
+
+void load_fpu_regs(void)
+{
+	raw_local_irq_disable();
+	__load_fpu_regs();
+	raw_local_irq_enable();
+}
+EXPORT_SYMBOL(load_fpu_regs);
+
+void save_fpu_regs(void)
+{
+	unsigned long flags, *regs;
+	struct fpu *state;
+
+	local_irq_save(flags);
+
+	if (test_cpu_flag(CIF_FPU))
+		goto out;
+
+	state = &current->thread.fpu;
+	regs = current->thread.fpu.regs;
+
+	asm volatile("stfpc %0" : "=Q" (state->fpc));
+	if (likely(MACHINE_HAS_VX)) {
+		asm volatile("lgr	1,%0\n"
+			     "VSTM	0,15,0,1\n"
+			     "VSTM	16,31,256,1\n"
+			     :
+			     : "d" (regs)
+			     : "1", "cc", "memory");
+	} else {
+		asm volatile("std 0,%0" : "=Q" (regs[0]));
+		asm volatile("std 1,%0" : "=Q" (regs[1]));
+		asm volatile("std 2,%0" : "=Q" (regs[2]));
+		asm volatile("std 3,%0" : "=Q" (regs[3]));
+		asm volatile("std 4,%0" : "=Q" (regs[4]));
+		asm volatile("std 5,%0" : "=Q" (regs[5]));
+		asm volatile("std 6,%0" : "=Q" (regs[6]));
+		asm volatile("std 7,%0" : "=Q" (regs[7]));
+		asm volatile("std 8,%0" : "=Q" (regs[8]));
+		asm volatile("std 9,%0" : "=Q" (regs[9]));
+		asm volatile("std 10,%0" : "=Q" (regs[10]));
+		asm volatile("std 11,%0" : "=Q" (regs[11]));
+		asm volatile("std 12,%0" : "=Q" (regs[12]));
+		asm volatile("std 13,%0" : "=Q" (regs[13]));
+		asm volatile("std 14,%0" : "=Q" (regs[14]));
+		asm volatile("std 15,%0" : "=Q" (regs[15]));
+	}
+	set_cpu_flag(CIF_FPU);
+out:
+	local_irq_restore(flags);
+}
+EXPORT_SYMBOL(save_fpu_regs);

arch/s390/kernel/idle.c

@@ -14,12 +14,36 @@
 #include <linux/cpu.h>
 #include <linux/sched/cputime.h>
 #include <trace/events/power.h>
+#include <asm/cpu_mf.h>
 #include <asm/nmi.h>
 #include <asm/smp.h>
 #include "entry.h"
 
 static DEFINE_PER_CPU(struct s390_idle_data, s390_idle);
 
+void account_idle_time_irq(void)
+{
+	struct s390_idle_data *idle = this_cpu_ptr(&s390_idle);
+	u64 cycles_new[8];
+	int i;
+
+	clear_cpu_flag(CIF_ENABLED_WAIT);
+	if (smp_cpu_mtid) {
+		stcctm(MT_DIAG, smp_cpu_mtid, cycles_new);
+		for (i = 0; i < smp_cpu_mtid; i++)
+			this_cpu_add(mt_cycles[i], cycles_new[i] - idle->mt_cycles_enter[i]);
+	}
+
+	idle->clock_idle_exit = S390_lowcore.int_clock;
+	idle->timer_idle_exit = S390_lowcore.sys_enter_timer;
+
+	S390_lowcore.steal_timer += idle->clock_idle_enter - S390_lowcore.last_update_clock;
+	S390_lowcore.last_update_clock = idle->clock_idle_exit;
+
+	S390_lowcore.system_timer += S390_lowcore.last_update_timer - idle->timer_idle_enter;
+	S390_lowcore.last_update_timer = idle->timer_idle_exit;
+}
+
 void arch_cpu_idle(void)
 {
 	struct s390_idle_data *idle = this_cpu_ptr(&s390_idle);

arch/s390/kernel/irq.c

@@ -21,6 +21,7 @@
 #include <linux/init.h>
 #include <linux/cpu.h>
 #include <linux/irq.h>
+#include <linux/entry-common.h>
 #include <asm/irq_regs.h>
 #include <asm/cputime.h>
 #include <asm/lowcore.h>
@@ -95,19 +96,97 @@ static const struct irq_class irqclass_sub_desc[] = {
 	{.irq = CPU_RST,    .name = "RST", .desc = "[CPU] CPU Restart"},
 };
 
-void do_IRQ(struct pt_regs *regs, int irq)
+static void do_IRQ(struct pt_regs *regs, int irq)
 {
-	struct pt_regs *old_regs;
-
-	old_regs = set_irq_regs(regs);
-	irq_enter();
 	if (tod_after_eq(S390_lowcore.int_clock,
 			 S390_lowcore.clock_comparator))
 		/* Serve timer interrupts first. */
 		clock_comparator_work();
 	generic_handle_irq(irq);
+}
+
+static int on_async_stack(void)
+{
+	unsigned long frame = current_frame_address();
+
+	return !!!((S390_lowcore.async_stack - frame) >> (PAGE_SHIFT + THREAD_SIZE_ORDER));
+}
+
+static void do_irq_async(struct pt_regs *regs, int irq)
+{
+	if (on_async_stack())
+		do_IRQ(regs, irq);
+	else
+		CALL_ON_STACK(do_IRQ, S390_lowcore.async_stack, 2, regs, irq);
+}
+
+static int irq_pending(struct pt_regs *regs)
+{
+	int cc;
+
+	asm volatile("tpi 0\n"
+		     "ipm %0" : "=d" (cc) : : "cc");
+	return cc >> 28;
+}
+
+void noinstr do_io_irq(struct pt_regs *regs)
+{
+	irqentry_state_t state = irqentry_enter(regs);
+	struct pt_regs *old_regs = set_irq_regs(regs);
+	int from_idle;
+
+	irq_enter();
+
+	if (user_mode(regs))
+		update_timer_sys();
+
+	from_idle = !user_mode(regs) && regs->psw.addr == (unsigned long)psw_idle_exit;
+	if (from_idle)
+		account_idle_time_irq();
+
+	do {
+		memcpy(&regs->int_code, &S390_lowcore.subchannel_id, 12);
+		if (S390_lowcore.io_int_word & BIT(31))
+			do_irq_async(regs, THIN_INTERRUPT);
+		else
+			do_irq_async(regs, IO_INTERRUPT);
+	} while (MACHINE_IS_LPAR && irq_pending(regs));
+
 	irq_exit();
 	set_irq_regs(old_regs);
+	irqentry_exit(regs, state);
+
+	if (from_idle)
+		regs->psw.mask &= ~(PSW_MASK_EXT | PSW_MASK_IO | PSW_MASK_WAIT);
+}
+
+void noinstr do_ext_irq(struct pt_regs *regs)
+{
+	irqentry_state_t state = irqentry_enter(regs);
+	struct pt_regs *old_regs = set_irq_regs(regs);
+	int from_idle;
+
+	irq_enter();
+
+	if (user_mode(regs))
+		update_timer_sys();
+
+	memcpy(&regs->int_code, &S390_lowcore.ext_cpu_addr, 4);
+	regs->int_parm = S390_lowcore.ext_params;
+	regs->int_parm_long = *(unsigned long *)S390_lowcore.ext_params2;
+
+	from_idle = !user_mode(regs) && regs->psw.addr == (unsigned long)psw_idle_exit;
+	if (from_idle)
+		account_idle_time_irq();
+
+	do_irq_async(regs, EXT_INTERRUPT);
+
+	irq_exit();
+	set_irq_regs(old_regs);
+	irqentry_exit(regs, state);
+
+	if (from_idle)
+		regs->psw.mask &= ~(PSW_MASK_EXT | PSW_MASK_IO | PSW_MASK_WAIT);
 }
 
 static void show_msi_interrupt(struct seq_file *p, int irq)

arch/s390/kernel/nmi.c

@@ -131,12 +131,11 @@ static notrace void s390_handle_damage(void)
 NOKPROBE_SYMBOL(s390_handle_damage);
 
 /*
- * Main machine check handler function. Will be called with interrupts enabled
+ * Main machine check handler function. Will be called with interrupts disabled
- * or disabled and machine checks enabled or disabled.
+ * and machine checks enabled.
  */
-void s390_handle_mcck(void)
+void __s390_handle_mcck(void)
 {
-	unsigned long flags;
 	struct mcck_struct mcck;
 
 	/*
@@ -144,12 +143,10 @@ void s390_handle_mcck(void)
 	 * machine checks. Afterwards delete the old state and enable machine
 	 * checks again.
 	 */
-	local_irq_save(flags);
 	local_mcck_disable();
 	mcck = *this_cpu_ptr(&cpu_mcck);
 	memset(this_cpu_ptr(&cpu_mcck), 0, sizeof(mcck));
 	local_mcck_enable();
-	local_irq_restore(flags);
 
 	if (mcck.channel_report)
 		crw_handle_channel_report();
@@ -181,8 +178,13 @@ void s390_handle_mcck(void)
 		do_exit(SIGSEGV);
 	}
 }
-EXPORT_SYMBOL_GPL(s390_handle_mcck);
 
+void noinstr s390_handle_mcck(void)
+{
+	trace_hardirqs_off();
+	__s390_handle_mcck();
+	trace_hardirqs_on();
+}
 /*
  * returns 0 if all required registers are available
  * returns 1 otherwise
@@ -344,6 +346,9 @@ int notrace s390_do_machine_check(struct pt_regs *regs)
 	int mcck_pending = 0;
 
 	nmi_enter();
+
+	if (user_mode(regs))
+		update_timer_mcck();
 	inc_irq_stat(NMI_NMI);
 	mci.val = S390_lowcore.mcck_interruption_code;
 	mcck = this_cpu_ptr(&cpu_mcck);

arch/s390/kernel/perf_cpum_cf_diag.c

@@ -118,8 +118,8 @@ static void cf_diag_trailer(struct cf_trailer_entry *te)
 	if (te->cpu_speed)
 		te->speed = 1;
 	te->clock_base = 1;			/* Save clock base */
-	memcpy(&te->tod_base, &tod_clock_base[1], 8);
+	te->tod_base = tod_clock_base.tod;
-	store_tod_clock((__u64 *)&te->timestamp);
+	te->timestamp = get_tod_clock_fast();
 }
 
 /*

arch/s390/kernel/perf_cpum_sf.c

@@ -1682,7 +1682,7 @@ static void aux_sdb_init(unsigned long sdb)
 
 	/* Save clock base */
 	te->clock_base = 1;
-	memcpy(&te->progusage2, &tod_clock_base[1], 8);
+	te->progusage2 = tod_clock_base.tod;
 }
 
 /*

arch/s390/kernel/process.c

@@ -29,6 +29,7 @@
 #include <linux/random.h>
 #include <linux/export.h>
 #include <linux/init_task.h>
+#include <linux/entry-common.h>
 #include <asm/cpu_mf.h>
 #include <asm/io.h>
 #include <asm/processor.h>
@@ -43,9 +44,22 @@
 #include <asm/unwind.h>
 #include "entry.h"
 
-asmlinkage void ret_from_fork(void) asm ("ret_from_fork");
+void ret_from_fork(void) asm("ret_from_fork");
 
-extern void kernel_thread_starter(void);
+void __ret_from_fork(struct task_struct *prev, struct pt_regs *regs)
+{
+	void (*func)(void *arg);
+
+	schedule_tail(prev);
+
+	if (!user_mode(regs)) {
+		/* Kernel thread */
+		func = (void *)regs->gprs[9];
+		func((void *)regs->gprs[10]);
+	}
+	clear_pt_regs_flag(regs, PIF_SYSCALL);
+	syscall_exit_to_user_mode(regs);
+}
 
 void flush_thread(void)
 {
@@ -108,10 +122,12 @@ int copy_thread(unsigned long clone_flags, unsigned long new_stackp,
 	p->thread.last_break = 1;
 
 	frame->sf.back_chain = 0;
+	frame->sf.gprs[5] = (unsigned long)frame + sizeof(struct stack_frame);
+	frame->sf.gprs[6] = (unsigned long)p;
 	/* new return point is ret_from_fork */
-	frame->sf.gprs[8] = (unsigned long) ret_from_fork;
+	frame->sf.gprs[8] = (unsigned long)ret_from_fork;
 	/* fake return stack for resume(), don't go back to schedule */
-	frame->sf.gprs[9] = (unsigned long) frame;
+	frame->sf.gprs[9] = (unsigned long)frame;
 
 	/* Store access registers to kernel stack of new process. */
 	if (unlikely(p->flags & PF_KTHREAD)) {
@@ -120,10 +136,10 @@ int copy_thread(unsigned long clone_flags, unsigned long new_stackp,
 		frame->childregs.psw.mask = PSW_KERNEL_BITS | PSW_MASK_DAT |
 				PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK;
 		frame->childregs.psw.addr =
-				(unsigned long) kernel_thread_starter;
+				(unsigned long)__ret_from_fork;
 		frame->childregs.gprs[9] = new_stackp; /* function */
 		frame->childregs.gprs[10] = arg;
-		frame->childregs.gprs[11] = (unsigned long) do_exit;
+		frame->childregs.gprs[11] = (unsigned long)do_exit;
 		frame->childregs.orig_gpr2 = -1;
 
 		return 0;
@@ -153,7 +169,7 @@ int copy_thread(unsigned long clone_flags, unsigned long new_stackp,
 	return 0;
 }
 
-asmlinkage void execve_tail(void)
+void execve_tail(void)
 {
 	current->thread.fpu.fpc = 0;
 	asm volatile("sfpc %0" : : "d" (0));

arch/s390/kernel/ptrace.c

@@ -7,6 +7,7 @@
  *               Martin Schwidefsky (schwidefsky@de.ibm.com)
  */
 
+#include "asm/ptrace.h"
 #include <linux/kernel.h>
 #include <linux/sched.h>
 #include <linux/sched/task_stack.h>
@@ -37,9 +38,6 @@
 #include "compat_ptrace.h"
 #endif
 
-#define CREATE_TRACE_POINTS
-#include <trace/events/syscalls.h>
-
 void update_cr_regs(struct task_struct *task)
 {
 	struct pt_regs *regs = task_pt_regs(task);
@@ -140,7 +138,7 @@ void ptrace_disable(struct task_struct *task)
 	memset(&task->thread.per_user, 0, sizeof(task->thread.per_user));
 	memset(&task->thread.per_event, 0, sizeof(task->thread.per_event));
 	clear_tsk_thread_flag(task, TIF_SINGLE_STEP);
-	clear_pt_regs_flag(task_pt_regs(task), PIF_PER_TRAP);
+	clear_tsk_thread_flag(task, TIF_PER_TRAP);
 	task->thread.per_flags = 0;
 }
 
@@ -322,25 +320,6 @@ static inline void __poke_user_per(struct task_struct *child,
 		child->thread.per_user.end = data;
 }
 
-static void fixup_int_code(struct task_struct *child, addr_t data)
-{
-	struct pt_regs *regs = task_pt_regs(child);
-	int ilc = regs->int_code >> 16;
-	u16 insn;
-
-	if (ilc > 6)
-		return;
-
-	if (ptrace_access_vm(child, regs->psw.addr - (regs->int_code >> 16),
-			&insn, sizeof(insn), FOLL_FORCE) != sizeof(insn))
-		return;
-
-	/* double check that tracee stopped on svc instruction */
-	if ((insn >> 8) != 0xa)
-		return;
-
-	regs->int_code = 0x20000 | (data & 0xffff);
-}
 /*
  * Write a word to the user area of a process at location addr. This
  * operation does have an additional problem compared to peek_user.
@@ -374,10 +353,12 @@ static int __poke_user(struct task_struct *child, addr_t addr, addr_t data)
 		}
 
 		if (test_pt_regs_flag(regs, PIF_SYSCALL) &&
-			addr == offsetof(struct user, regs.gprs[2]))
+			addr == offsetof(struct user, regs.gprs[2])) {
-			fixup_int_code(child, data);
+			struct pt_regs *regs = task_pt_regs(child);
-		*(addr_t *)((addr_t) &regs->psw + addr) = data;
 
+			regs->int_code = 0x20000 | (data & 0xffff);
+		}
+		*(addr_t *)((addr_t) &regs->psw + addr) = data;
 	} else if (addr < (addr_t) (&dummy->regs.orig_gpr2)) {
 		/*
 		 * access registers are stored in the thread structure
@@ -742,10 +723,12 @@ static int __poke_user_compat(struct task_struct *child,
 			regs->psw.mask = (regs->psw.mask & ~PSW_MASK_BA) |
 				(__u64)(tmp & PSW32_ADDR_AMODE);
 		} else {
-
 			if (test_pt_regs_flag(regs, PIF_SYSCALL) &&
-				addr == offsetof(struct compat_user, regs.gprs[2]))
+				addr == offsetof(struct compat_user, regs.gprs[2])) {
-				fixup_int_code(child, data);
+				struct pt_regs *regs = task_pt_regs(child);
+
+				regs->int_code = 0x20000 | (data & 0xffff);
+			}
 			/* gpr 0-15 */
 			*(__u32*)((addr_t) &regs->psw + addr*2 + 4) = tmp;
 		}
@@ -862,82 +845,6 @@ long compat_arch_ptrace(struct task_struct *child, compat_long_t request,
 }
 #endif
 
-asmlinkage long do_syscall_trace_enter(struct pt_regs *regs)
-{
-	unsigned long mask = -1UL;
-	long ret = -1;
-
-	if (is_compat_task())
-		mask = 0xffffffff;
-
-	/*
-	 * The sysc_tracesys code in entry.S stored the system
-	 * call number to gprs[2].
-	 */
-	if (test_thread_flag(TIF_SYSCALL_TRACE) &&
-	    tracehook_report_syscall_entry(regs)) {
-		/*
-		 * Tracing decided this syscall should not happen. Skip
-		 * the system call and the system call restart handling.
-		 */
-		goto skip;
-	}
-
-#ifdef CONFIG_SECCOMP
-	/* Do the secure computing check after ptrace. */
-	if (unlikely(test_thread_flag(TIF_SECCOMP))) {
-		struct seccomp_data sd;
-
-		if (is_compat_task()) {
-			sd.instruction_pointer = regs->psw.addr & 0x7fffffff;
-			sd.arch = AUDIT_ARCH_S390;
-		} else {
-			sd.instruction_pointer = regs->psw.addr;
-			sd.arch = AUDIT_ARCH_S390X;
-		}
-
-		sd.nr = regs->int_code & 0xffff;
-		sd.args[0] = regs->orig_gpr2 & mask;
-		sd.args[1] = regs->gprs[3] & mask;
-		sd.args[2] = regs->gprs[4] & mask;
-		sd.args[3] = regs->gprs[5] & mask;
-		sd.args[4] = regs->gprs[6] & mask;
-		sd.args[5] = regs->gprs[7] & mask;
-
-		if (__secure_computing(&sd) == -1)
-			goto skip;
-	}
-#endif /* CONFIG_SECCOMP */
-
-	if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT)))
-		trace_sys_enter(regs, regs->int_code & 0xffff);
-
-
-	audit_syscall_entry(regs->int_code & 0xffff, regs->orig_gpr2 & mask,
-			    regs->gprs[3] &mask, regs->gprs[4] &mask,
-			    regs->gprs[5] &mask);
-
-	if ((signed long)regs->gprs[2] >= NR_syscalls) {
-		regs->gprs[2] = -ENOSYS;
-		ret = -ENOSYS;
-	}
-	return regs->gprs[2];
-skip:
-	clear_pt_regs_flag(regs, PIF_SYSCALL);
-	return ret;
-}
-
-asmlinkage void do_syscall_trace_exit(struct pt_regs *regs)
-{
-	audit_syscall_exit(regs);
-
-	if (unlikely(test_thread_flag(TIF_SYSCALL_TRACEPOINT)))
-		trace_sys_exit(regs, regs->gprs[2]);
-
-	if (test_thread_flag(TIF_SYSCALL_TRACE))
-		tracehook_report_syscall_exit(regs, 0);
-}
-
 /*
  * user_regset definitions.
  */

arch/s390/kernel/setup.c

@@ -338,7 +338,7 @@ int __init arch_early_irq_init(void)
 	return 0;
 }
 
-static int __init async_stack_realloc(void)
+static int __init stack_realloc(void)
 {
 	unsigned long old, new;
 
@@ -346,11 +346,18 @@ static int __init async_stack_realloc(void)
 	new = stack_alloc();
 	if (!new)
 		panic("Couldn't allocate async stack");
-	S390_lowcore.async_stack = new + STACK_INIT_OFFSET;
+	WRITE_ONCE(S390_lowcore.async_stack, new + STACK_INIT_OFFSET);
 	free_pages(old, THREAD_SIZE_ORDER);
+
+	old = S390_lowcore.mcck_stack - STACK_INIT_OFFSET;
+	new = stack_alloc();
+	if (!new)
+		panic("Couldn't allocate machine check stack");
+	WRITE_ONCE(S390_lowcore.mcck_stack, new + STACK_INIT_OFFSET);
+	memblock_free(old, THREAD_SIZE);
 	return 0;
 }
-early_initcall(async_stack_realloc);
+early_initcall(stack_realloc);
 
 void __init arch_call_rest_init(void)
 {
@@ -372,6 +379,7 @@ void __init arch_call_rest_init(void)
 static void __init setup_lowcore_dat_off(void)
 {
 	unsigned long int_psw_mask = PSW_KERNEL_BITS;
+	unsigned long mcck_stack;
 	struct lowcore *lc;
 
 	if (IS_ENABLED(CONFIG_KASAN))
@@ -411,8 +419,7 @@ static void __init setup_lowcore_dat_off(void)
 	memcpy(lc->alt_stfle_fac_list, S390_lowcore.alt_stfle_fac_list,
 	       sizeof(lc->alt_stfle_fac_list));
 	nmi_alloc_boot_cpu(lc);
-	lc->sync_enter_timer = S390_lowcore.sync_enter_timer;
+	lc->sys_enter_timer = S390_lowcore.sys_enter_timer;
-	lc->async_enter_timer = S390_lowcore.async_enter_timer;
 	lc->exit_timer = S390_lowcore.exit_timer;
 	lc->user_timer = S390_lowcore.user_timer;
 	lc->system_timer = S390_lowcore.system_timer;
@@ -440,6 +447,12 @@ static void __init setup_lowcore_dat_off(void)
 	lc->restart_data = 0;
 	lc->restart_source = -1UL;
 
+	mcck_stack = (unsigned long)memblock_alloc(THREAD_SIZE, THREAD_SIZE);
+	if (!mcck_stack)
+		panic("%s: Failed to allocate %lu bytes align=0x%lx\n",
+		      __func__, THREAD_SIZE, THREAD_SIZE);
+	lc->mcck_stack = mcck_stack + STACK_INIT_OFFSET;
+
 	/* Setup absolute zero lowcore */
 	mem_assign_absolute(S390_lowcore.restart_stack, lc->restart_stack);
 	mem_assign_absolute(S390_lowcore.restart_fn, lc->restart_fn);

arch/s390/kernel/signal.c

@@ -16,6 +16,7 @@
 #include <linux/smp.h>
 #include <linux/kernel.h>
 #include <linux/signal.h>
+#include <linux/entry-common.h>
 #include <linux/errno.h>
 #include <linux/wait.h>
 #include <linux/ptrace.h>
@@ -170,6 +171,7 @@ static int restore_sigregs(struct pt_regs *regs, _sigregs __user *sregs)
 	fpregs_load(&user_sregs.fpregs, &current->thread.fpu);
 
 	clear_pt_regs_flag(regs, PIF_SYSCALL); /* No longer in a system call */
+	clear_pt_regs_flag(regs, PIF_SYSCALL_RESTART);
 	return 0;
 }
 
@@ -459,7 +461,8 @@ static void handle_signal(struct ksignal *ksig, sigset_t *oldset,
  * the kernel can handle, and then we build all the user-level signal handling
  * stack-frames in one go after that.
  */
-void do_signal(struct pt_regs *regs)
+
+void arch_do_signal_or_restart(struct pt_regs *regs, bool has_signal)
 {
 	struct ksignal ksig;
 	sigset_t *oldset = sigmask_to_save();
@@ -472,7 +475,7 @@ void do_signal(struct pt_regs *regs)
 	current->thread.system_call =
 		test_pt_regs_flag(regs, PIF_SYSCALL) ? regs->int_code : 0;
 
-	if (test_thread_flag(TIF_SIGPENDING) && get_signal(&ksig)) {
+	if (has_signal && get_signal(&ksig)) {
 		/* Whee!  Actually deliver the signal.  */
 		if (current->thread.system_call) {
 			regs->int_code = current->thread.system_call;
@@ -498,6 +501,7 @@ void do_signal(struct pt_regs *regs)
 		}
 		/* No longer in a system call */
 		clear_pt_regs_flag(regs, PIF_SYSCALL);
+		clear_pt_regs_flag(regs, PIF_SYSCALL_RESTART);
 		rseq_signal_deliver(&ksig, regs);
 		if (is_compat_task())
 			handle_signal32(&ksig, oldset, regs);
@@ -508,6 +512,7 @@ void do_signal(struct pt_regs *regs)
 
 	/* No handlers present - check for system call restart */
 	clear_pt_regs_flag(regs, PIF_SYSCALL);
+	clear_pt_regs_flag(regs, PIF_SYSCALL_RESTART);
 	if (current->thread.system_call) {
 		regs->int_code = current->thread.system_call;
 		switch (regs->gprs[2]) {
@@ -520,9 +525,9 @@ void do_signal(struct pt_regs *regs)
 		case -ERESTARTNOINTR:
 			/* Restart system call with magic TIF bit. */
 			regs->gprs[2] = regs->orig_gpr2;
-			set_pt_regs_flag(regs, PIF_SYSCALL);
+			set_pt_regs_flag(regs, PIF_SYSCALL_RESTART);
 			if (test_thread_flag(TIF_SINGLE_STEP))
-				clear_pt_regs_flag(regs, PIF_PER_TRAP);
+				clear_thread_flag(TIF_PER_TRAP);
 			break;
 		}
 	}

arch/s390/kernel/smp.c

@@ -189,7 +189,7 @@ static void pcpu_ec_call(struct pcpu *pcpu, int ec_bit)
 
 static int pcpu_alloc_lowcore(struct pcpu *pcpu, int cpu)
 {
-	unsigned long async_stack, nodat_stack;
+	unsigned long async_stack, nodat_stack, mcck_stack;
 	struct lowcore *lc;
 
 	if (pcpu != &pcpu_devices[0]) {
@@ -202,13 +202,15 @@ static int pcpu_alloc_lowcore(struct pcpu *pcpu, int cpu)
 		nodat_stack = pcpu->lowcore->nodat_stack - STACK_INIT_OFFSET;
 	}
 	async_stack = stack_alloc();
-	if (!async_stack)
+	mcck_stack = stack_alloc();
-		goto out;
+	if (!async_stack || !mcck_stack)
+		goto out_stack;
 	lc = pcpu->lowcore;
 	memcpy(lc, &S390_lowcore, 512);
 	memset((char *) lc + 512, 0, sizeof(*lc) - 512);
 	lc->async_stack = async_stack + STACK_INIT_OFFSET;
 	lc->nodat_stack = nodat_stack + STACK_INIT_OFFSET;
+	lc->mcck_stack = mcck_stack + STACK_INIT_OFFSET;
 	lc->cpu_nr = cpu;
 	lc->spinlock_lockval = arch_spin_lockval(cpu);
 	lc->spinlock_index = 0;
@@ -216,12 +218,13 @@ static int pcpu_alloc_lowcore(struct pcpu *pcpu, int cpu)
 	lc->return_lpswe = gen_lpswe(__LC_RETURN_PSW);
 	lc->return_mcck_lpswe = gen_lpswe(__LC_RETURN_MCCK_PSW);
 	if (nmi_alloc_per_cpu(lc))
-		goto out_async;
+		goto out_stack;
 	lowcore_ptr[cpu] = lc;
 	pcpu_sigp_retry(pcpu, SIGP_SET_PREFIX, (u32)(unsigned long) lc);
 	return 0;
 
-out_async:
+out_stack:
+	stack_free(mcck_stack);
 	stack_free(async_stack);
 out:
 	if (pcpu != &pcpu_devices[0]) {
@@ -233,16 +236,18 @@ out:
 
 static void pcpu_free_lowcore(struct pcpu *pcpu)
 {
-	unsigned long async_stack, nodat_stack, lowcore;
+	unsigned long async_stack, nodat_stack, mcck_stack, lowcore;
 
 	nodat_stack = pcpu->lowcore->nodat_stack - STACK_INIT_OFFSET;
 	async_stack = pcpu->lowcore->async_stack - STACK_INIT_OFFSET;
+	mcck_stack = pcpu->lowcore->mcck_stack - STACK_INIT_OFFSET;
 	lowcore = (unsigned long) pcpu->lowcore;
 
 	pcpu_sigp_retry(pcpu, SIGP_SET_PREFIX, 0);
 	lowcore_ptr[pcpu - pcpu_devices] = NULL;
 	nmi_free_per_cpu(pcpu->lowcore);
 	stack_free(async_stack);
+	stack_free(mcck_stack);
 	if (pcpu == &pcpu_devices[0])
 		return;
 	free_pages(nodat_stack, THREAD_SIZE_ORDER);
@@ -499,7 +504,7 @@ static void smp_handle_ext_call(void)
 	if (test_bit(ec_call_function_single, &bits))
 		generic_smp_call_function_single_interrupt();
 	if (test_bit(ec_mcck_pending, &bits))
-		s390_handle_mcck();
+		__s390_handle_mcck();
 }
 
 static void do_ext_call_interrupt(struct ext_code ext_code,

arch/s390/kernel/syscall.c

@@ -29,6 +29,13 @@
 #include <linux/unistd.h>
 #include <linux/ipc.h>
 #include <linux/uaccess.h>
+#include <linux/string.h>
+#include <linux/thread_info.h>
+#include <linux/entry-common.h>
+
+#include <asm/ptrace.h>
+#include <asm/vtime.h>
+
 #include "entry.h"
 
 /*
@@ -100,3 +107,62 @@ SYSCALL_DEFINE0(ni_syscall)
 {
 	return -ENOSYS;
 }
+
+void do_syscall(struct pt_regs *regs)
+{
+	unsigned long nr;
+
+	nr = regs->int_code & 0xffff;
+	if (!nr) {
+		nr = regs->gprs[1] & 0xffff;
+		regs->int_code &= ~0xffffUL;
+		regs->int_code |= nr;
+	}
+
+	regs->gprs[2] = nr;
+
+	nr = syscall_enter_from_user_mode_work(regs, nr);
+
+	/*
+	 * In the s390 ptrace ABI, both the syscall number and the return value
+	 * use gpr2. However, userspace puts the syscall number either in the
+	 * svc instruction itself, or uses gpr1. To make at least skipping syscalls
+	 * work, the ptrace code sets PIF_SYSCALL_RET_SET, which is checked here
+	 * and if set, the syscall will be skipped.
+	 */
+	if (!test_pt_regs_flag(regs, PIF_SYSCALL_RET_SET)) {
+		regs->gprs[2] = -ENOSYS;
+		if (likely(nr < NR_syscalls))
+			regs->gprs[2] = current->thread.sys_call_table[nr](regs);
+	} else {
+		clear_pt_regs_flag(regs, PIF_SYSCALL_RET_SET);
+	}
+	syscall_exit_to_user_mode_work(regs);
+}
+
+void noinstr __do_syscall(struct pt_regs *regs, int per_trap)
+{
+	enter_from_user_mode(regs);
+
+	memcpy(&regs->gprs[8], S390_lowcore.save_area_sync, 8 * sizeof(unsigned long));
+	memcpy(&regs->int_code, &S390_lowcore.svc_ilc, sizeof(regs->int_code));
+	regs->psw = S390_lowcore.svc_old_psw;
+
+	update_timer_sys();
+
+	local_irq_enable();
+	regs->orig_gpr2 = regs->gprs[2];
+
+	if (per_trap)
+		set_thread_flag(TIF_PER_TRAP);
+
+	for (;;) {
+		regs->flags = 0;
+		set_pt_regs_flag(regs, PIF_SYSCALL);
+		do_syscall(regs);
+		if (!test_pt_regs_flag(regs, PIF_SYSCALL_RESTART))
+			break;
+		local_irq_enable();
+	}
+	exit_to_user_mode();
+}

arch/s390/kernel/time.c

@@ -55,11 +55,7 @@
 #include <asm/cio.h>
 #include "entry.h"
 
-unsigned char tod_clock_base[16] __aligned(8) = {
+union tod_clock tod_clock_base __section(".data");
-	/* Force to data section. */
-	0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff,
-	0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff, 0xff
-};
 EXPORT_SYMBOL_GPL(tod_clock_base);
 
 u64 clock_comparator_max = -1ULL;
@@ -86,7 +82,7 @@ void __init time_early_init(void)
 	struct ptff_qui qui;
 
 	/* Initialize TOD steering parameters */
-	tod_steering_end = *(unsigned long long *) &tod_clock_base[1];
+	tod_steering_end = tod_clock_base.tod;
 	vdso_data->arch_data.tod_steering_end = tod_steering_end;
 
 	if (!test_facility(28))
@@ -113,18 +109,13 @@ unsigned long long notrace sched_clock(void)
 }
 NOKPROBE_SYMBOL(sched_clock);
 
-static void ext_to_timespec64(unsigned char *clk, struct timespec64 *xt)
+static void ext_to_timespec64(union tod_clock *clk, struct timespec64 *xt)
 {
-	unsigned long long high, low, rem, sec, nsec;
+	unsigned long rem, sec, nsec;
 
-	/* Split extendnd TOD clock to micro-seconds and sub-micro-seconds */
+	sec = clk->us;
-	high = (*(unsigned long long *) clk) >> 4;
-	low = (*(unsigned long long *)&clk[7]) << 4;
-	/* Calculate seconds and nano-seconds */
-	sec = high;
 	rem = do_div(sec, 1000000);
-	nsec = (((low >> 32) + (rem << 32)) * 1000) >> 32;
+	nsec = ((clk->sus + (rem << 12)) * 125) >> 9;
-
 	xt->tv_sec = sec;
 	xt->tv_nsec = nsec;
 }
@@ -204,30 +195,26 @@ static void stp_reset(void);
 
 void read_persistent_clock64(struct timespec64 *ts)
 {
-	unsigned char clk[STORE_CLOCK_EXT_SIZE];
+	union tod_clock clk;
-	__u64 delta;
+	u64 delta;
 
 	delta = initial_leap_seconds + TOD_UNIX_EPOCH;
-	get_tod_clock_ext(clk);
+	store_tod_clock_ext(&clk);
-	*(__u64 *) &clk[1] -= delta;
+	clk.eitod -= delta;
-	if (*(__u64 *) &clk[1] > delta)
+	ext_to_timespec64(&clk, ts);
-		clk[0]--;
-	ext_to_timespec64(clk, ts);
 }
 
 void __init read_persistent_wall_and_boot_offset(struct timespec64 *wall_time,
 						 struct timespec64 *boot_offset)
 {
-	unsigned char clk[STORE_CLOCK_EXT_SIZE];
 	struct timespec64 boot_time;
-	__u64 delta;
+	union tod_clock clk;
+	u64 delta;
 
 	delta = initial_leap_seconds + TOD_UNIX_EPOCH;
-	memcpy(clk, tod_clock_base, STORE_CLOCK_EXT_SIZE);
+	clk = tod_clock_base;
-	*(__u64 *)&clk[1] -= delta;
+	clk.eitod -= delta;
-	if (*(__u64 *)&clk[1] > delta)
+	ext_to_timespec64(&clk, &boot_time);
-		clk[0]--;
-	ext_to_timespec64(clk, &boot_time);
 
 	read_persistent_clock64(wall_time);
 	*boot_offset = timespec64_sub(*wall_time, boot_time);
@@ -381,10 +368,7 @@ static void clock_sync_global(unsigned long long delta)
 	struct ptff_qto qto;
 
 	/* Fixup the monotonic sched clock. */
-	*(unsigned long long *) &tod_clock_base[1] += delta;
+	tod_clock_base.eitod += delta;
-	if (*(unsigned long long *) &tod_clock_base[1] < delta)
-		/* Epoch overflow */
-		tod_clock_base[0]++;
 	/* Adjust TOD steering parameters. */
 	now = get_tod_clock();
 	adj = tod_steering_end - now;

arch/s390/kernel/traps.c

@@ -13,6 +13,8 @@
  * 'Traps.c' handles hardware traps and faults after we have saved some
  * state in 'asm.s'.
  */
+#include "asm/irqflags.h"
+#include "asm/ptrace.h"
 #include <linux/kprobes.h>
 #include <linux/kdebug.h>
 #include <linux/extable.h>
@@ -23,7 +25,9 @@
 #include <linux/slab.h>
 #include <linux/uaccess.h>
 #include <linux/cpu.h>
+#include <linux/entry-common.h>
 #include <asm/fpu/api.h>
+#include <asm/vtime.h>
 #include "entry.h"
 
 static inline void __user *get_trap_ip(struct pt_regs *regs)
@@ -288,3 +292,64 @@ void __init trap_init(void)
 	local_mcck_enable();
 	test_monitor_call();
 }
+
+void noinstr __do_pgm_check(struct pt_regs *regs)
+{
+	unsigned long last_break = S390_lowcore.breaking_event_addr;
+	unsigned int trapnr, syscall_redirect = 0;
+	irqentry_state_t state;
+
+	regs->int_code = *(u32 *)&S390_lowcore.pgm_ilc;
+	regs->int_parm_long = S390_lowcore.trans_exc_code;
+
+	state = irqentry_enter(regs);
+
+	if (user_mode(regs)) {
+		update_timer_sys();
+		if (last_break < 4096)
+			last_break = 1;
+		current->thread.last_break = last_break;
+		regs->args[0] = last_break;
+	}
+
+	if (S390_lowcore.pgm_code & 0x0200) {
+		/* transaction abort */
+		memcpy(&current->thread.trap_tdb, &S390_lowcore.pgm_tdb, 256);
+	}
+
+	if (S390_lowcore.pgm_code & PGM_INT_CODE_PER) {
+		if (user_mode(regs)) {
+			struct per_event *ev = &current->thread.per_event;
+
+			set_thread_flag(TIF_PER_TRAP);
+			ev->address = S390_lowcore.per_address;
+			ev->cause = *(u16 *)&S390_lowcore.per_code;
+			ev->paid = S390_lowcore.per_access_id;
+		} else {
+			/* PER event in kernel is kprobes */
+			__arch_local_irq_ssm(regs->psw.mask & ~PSW_MASK_PER);
+			do_per_trap(regs);
+			goto out;
+		}
+	}
+
+	if (!irqs_disabled_flags(regs->psw.mask))
+		trace_hardirqs_on();
+	__arch_local_irq_ssm(regs->psw.mask & ~PSW_MASK_PER);
+
+	trapnr = regs->int_code & PGM_INT_CODE_MASK;
+	if (trapnr)
+		pgm_check_table[trapnr](regs);
+	syscall_redirect = user_mode(regs) && test_pt_regs_flag(regs, PIF_SYSCALL);
+out:
+	local_irq_disable();
+	irqentry_exit(regs, state);
+
+	if (syscall_redirect) {
+		enter_from_user_mode(regs);
+		local_irq_enable();
+		regs->orig_gpr2 = regs->gprs[2];
+		do_syscall(regs);
+		exit_to_user_mode();
+	}
+}

arch/s390/kernel/uprobes.c

@@ -32,7 +32,7 @@ int arch_uprobe_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
 		return -EINVAL;
 	if (!is_compat_task() && psw_bits(regs->psw).eaba == PSW_BITS_AMODE_31BIT)
 		return -EINVAL;
-	clear_pt_regs_flag(regs, PIF_PER_TRAP);
+	clear_thread_flag(TIF_PER_TRAP);
 	auprobe->saved_per = psw_bits(regs->psw).per;
 	auprobe->saved_int_code = regs->int_code;
 	regs->int_code = UPROBE_TRAP_NR;
@@ -103,7 +103,7 @@ int arch_uprobe_post_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
 		/* fix per address */
 		current->thread.per_event.address = utask->vaddr;
 		/* trigger per event */
-		set_pt_regs_flag(regs, PIF_PER_TRAP);
+		set_thread_flag(TIF_PER_TRAP);
 	}
 	return 0;
 }
@@ -259,7 +259,7 @@ static void sim_stor_event(struct pt_regs *regs, void *addr, int len)
 		return;
 	current->thread.per_event.address = regs->psw.addr;
 	current->thread.per_event.cause = PER_EVENT_STORE >> 16;
-	set_pt_regs_flag(regs, PIF_PER_TRAP);
+	set_thread_flag(TIF_PER_TRAP);
 }
 
 /*

arch/s390/kernel/vdso.c

@@ -6,73 +6,39 @@
  *  Author(s): Martin Schwidefsky (schwidefsky@de.ibm.com)
  */
 
-#include <linux/init.h>
+#include <linux/binfmts.h>
+#include <linux/compat.h>
+#include <linux/elf.h>
 #include <linux/errno.h>
-#include <linux/sched.h>
+#include <linux/init.h>
 #include <linux/kernel.h>
 #include <linux/mm.h>
-#include <linux/smp.h>
-#include <linux/stddef.h>
-#include <linux/unistd.h>
 #include <linux/slab.h>
-#include <linux/user.h>
+#include <linux/smp.h>
-#include <linux/elf.h>
+#include <linux/time_namespace.h>
-#include <linux/security.h>
-#include <linux/memblock.h>
-#include <linux/compat.h>
-#include <linux/binfmts.h>
 #include <vdso/datapage.h>
-#include <asm/asm-offsets.h>
-#include <asm/processor.h>
-#include <asm/mmu.h>
-#include <asm/mmu_context.h>
-#include <asm/sections.h>
 #include <asm/vdso.h>
-#include <asm/facility.h>
-#include <asm/timex.h>
 
-extern char vdso64_start, vdso64_end;
+extern char vdso64_start[], vdso64_end[];
-static void *vdso64_kbase = &vdso64_start;
+static unsigned int vdso_pages;
-static unsigned int vdso64_pages;
-static struct page **vdso64_pagelist;
 
-/*
+static struct vm_special_mapping vvar_mapping;
- * Should the kernel map a VDSO page into processes and pass its
- * address down to glibc upon exec()?
- */
-unsigned int __read_mostly vdso_enabled = 1;
 
-static vm_fault_t vdso_fault(const struct vm_special_mapping *sm,
+static union {
-		      struct vm_area_struct *vma, struct vm_fault *vmf)
+	struct vdso_data	data[CS_BASES];
-{
+	u8			page[PAGE_SIZE];
-	struct page **vdso_pagelist;
+} vdso_data_store __page_aligned_data;
-	unsigned long vdso_pages;
 
-	vdso_pagelist = vdso64_pagelist;
+struct vdso_data *vdso_data = vdso_data_store.data;
-	vdso_pages = vdso64_pages;
 
-	if (vmf->pgoff >= vdso_pages)
+enum vvar_pages {
-		return VM_FAULT_SIGBUS;
+	VVAR_DATA_PAGE_OFFSET,
-
+	VVAR_TIMENS_PAGE_OFFSET,
-	vmf->page = vdso_pagelist[vmf->pgoff];
+	VVAR_NR_PAGES,
-	get_page(vmf->page);
-	return 0;
-}
-
-static int vdso_mremap(const struct vm_special_mapping *sm,
-		       struct vm_area_struct *vma)
-{
-	current->mm->context.vdso_base = vma->vm_start;
-
-	return 0;
-}
-
-static const struct vm_special_mapping vdso_mapping = {
-	.name = "[vdso]",
-	.fault = vdso_fault,
-	.mremap = vdso_mremap,
 };
 
+unsigned int __read_mostly vdso_enabled = 1;
+
 static int __init vdso_setup(char *str)
 {
 	bool enabled;
@@ -83,109 +49,183 @@ static int __init vdso_setup(char *str)
 }
 __setup("vdso=", vdso_setup);
 
-/*
+#ifdef CONFIG_TIME_NS
- * The vdso data page
+struct vdso_data *arch_get_vdso_data(void *vvar_page)
- */
-static union {
-	struct vdso_data	data;
-	u8			page[PAGE_SIZE];
-} vdso_data_store __page_aligned_data;
-struct vdso_data *vdso_data = (struct vdso_data *)&vdso_data_store.data;
-
-void vdso_getcpu_init(void)
 {
-	set_tod_programmable_field(smp_processor_id());
+	return (struct vdso_data *)(vvar_page);
+}
+
+static struct page *find_timens_vvar_page(struct vm_area_struct *vma)
+{
+	if (likely(vma->vm_mm == current->mm))
+		return current->nsproxy->time_ns->vvar_page;
+	/*
+	 * VM_PFNMAP | VM_IO protect .fault() handler from being called
+	 * through interfaces like /proc/$pid/mem or
+	 * process_vm_{readv,writev}() as long as there's no .access()
+	 * in special_mapping_vmops().
+	 * For more details check_vma_flags() and __access_remote_vm()
+	 */
+	WARN(1, "vvar_page accessed remotely");
+	return NULL;
 }
 
 /*
- * This is called from binfmt_elf, we create the special vma for the
+ * The VVAR page layout depends on whether a task belongs to the root or
- * vDSO and insert it into the mm struct tree
+ * non-root time namespace. Whenever a task changes its namespace, the VVAR
+ * page tables are cleared and then they will be re-faulted with a
+ * corresponding layout.
+ * See also the comment near timens_setup_vdso_data() for details.
  */
+int vdso_join_timens(struct task_struct *task, struct time_namespace *ns)
+{
+	struct mm_struct *mm = task->mm;
+	struct vm_area_struct *vma;
+
+	mmap_read_lock(mm);
+	for (vma = mm->mmap; vma; vma = vma->vm_next) {
+		unsigned long size = vma->vm_end - vma->vm_start;
+
+		if (!vma_is_special_mapping(vma, &vvar_mapping))
+			continue;
+		zap_page_range(vma, vma->vm_start, size);
+		break;
+	}
+	mmap_read_unlock(mm);
+	return 0;
+}
+#else
+static inline struct page *find_timens_vvar_page(struct vm_area_struct *vma)
+{
+	return NULL;
+}
+#endif
+
+static vm_fault_t vvar_fault(const struct vm_special_mapping *sm,
+			     struct vm_area_struct *vma, struct vm_fault *vmf)
+{
+	struct page *timens_page = find_timens_vvar_page(vma);
+	unsigned long addr, pfn;
+	vm_fault_t err;
+
+	switch (vmf->pgoff) {
+	case VVAR_DATA_PAGE_OFFSET:
+		pfn = virt_to_pfn(vdso_data);
+		if (timens_page) {
+			/*
+			 * Fault in VVAR page too, since it will be accessed
+			 * to get clock data anyway.
+			 */
+			addr = vmf->address + VVAR_TIMENS_PAGE_OFFSET * PAGE_SIZE;
+			err = vmf_insert_pfn(vma, addr, pfn);
+			if (unlikely(err & VM_FAULT_ERROR))
+				return err;
+			pfn = page_to_pfn(timens_page);
+		}
+		break;
+#ifdef CONFIG_TIME_NS
+	case VVAR_TIMENS_PAGE_OFFSET:
+		/*
+		 * If a task belongs to a time namespace then a namespace
+		 * specific VVAR is mapped with the VVAR_DATA_PAGE_OFFSET and
+		 * the real VVAR page is mapped with the VVAR_TIMENS_PAGE_OFFSET
+		 * offset.
+		 * See also the comment near timens_setup_vdso_data().
+		 */
+		if (!timens_page)
+			return VM_FAULT_SIGBUS;
+		pfn = virt_to_pfn(vdso_data);
+		break;
+#endif /* CONFIG_TIME_NS */
+	default:
+		return VM_FAULT_SIGBUS;
+	}
+	return vmf_insert_pfn(vma, vmf->address, pfn);
+}
+
+static int vdso_mremap(const struct vm_special_mapping *sm,
+		       struct vm_area_struct *vma)
+{
+	current->mm->context.vdso_base = vma->vm_start;
+	return 0;
+}
+
+static struct vm_special_mapping vvar_mapping = {
+	.name = "[vvar]",
+	.fault = vvar_fault,
+};
+
+static struct vm_special_mapping vdso_mapping = {
+	.name = "[vdso]",
+	.mremap = vdso_mremap,
+};
+
+int vdso_getcpu_init(void)
+{
+	set_tod_programmable_field(smp_processor_id());
+	return 0;
+}
+early_initcall(vdso_getcpu_init); /* Must be called before SMP init */
+
 int arch_setup_additional_pages(struct linux_binprm *bprm, int uses_interp)
 {
+	unsigned long vdso_text_len, vdso_mapping_len;
+	unsigned long vvar_start, vdso_text_start;
 	struct mm_struct *mm = current->mm;
 	struct vm_area_struct *vma;
-	unsigned long vdso_pages;
-	unsigned long vdso_base;
 	int rc;
 
-	if (!vdso_enabled)
+	BUILD_BUG_ON(VVAR_NR_PAGES != __VVAR_PAGES);
+	if (!vdso_enabled || is_compat_task())
 		return 0;
-
-	if (is_compat_task())
-		return 0;
-
-	vdso_pages = vdso64_pages;
-	/*
-	 * vDSO has a problem and was disabled, just don't "enable" it for
-	 * the process
-	 */
-	if (vdso_pages == 0)
-		return 0;
-
-	/*
-	 * pick a base address for the vDSO in process space. We try to put
-	 * it at vdso_base which is the "natural" base for it, but we might
-	 * fail and end up putting it elsewhere.
-	 */
 	if (mmap_write_lock_killable(mm))
 		return -EINTR;
-	vdso_base = get_unmapped_area(NULL, 0, vdso_pages << PAGE_SHIFT, 0, 0);
+	vdso_text_len = vdso_pages << PAGE_SHIFT;
-	if (IS_ERR_VALUE(vdso_base)) {
+	vdso_mapping_len = vdso_text_len + VVAR_NR_PAGES * PAGE_SIZE;
-		rc = vdso_base;
+	vvar_start = get_unmapped_area(NULL, 0, vdso_mapping_len, 0, 0);
-		goto out_up;
+	rc = vvar_start;
-	}
+	if (IS_ERR_VALUE(vvar_start))
-
+		goto out;
-	/*
+	vma = _install_special_mapping(mm, vvar_start, VVAR_NR_PAGES*PAGE_SIZE,
-	 * our vma flags don't have VM_WRITE so by default, the process
+				       VM_READ|VM_MAYREAD|VM_IO|VM_DONTDUMP|
-	 * isn't allowed to write those pages.
+				       VM_PFNMAP,
-	 * gdb can break that with ptrace interface, and thus trigger COW
+				       &vvar_mapping);
-	 * on those pages but it's then your responsibility to never do that
+	rc = PTR_ERR(vma);
-	 * on the "data" page of the vDSO or you'll stop getting kernel
+	if (IS_ERR(vma))
-	 * updates and your nice userland gettimeofday will be totally dead.
+		goto out;
-	 * It's fine to use that for setting breakpoints in the vDSO code
+	vdso_text_start = vvar_start + VVAR_NR_PAGES * PAGE_SIZE;
-	 * pages though.
+	/* VM_MAYWRITE for COW so gdb can set breakpoints */
-	 */
+	vma = _install_special_mapping(mm, vdso_text_start, vdso_text_len,
-	vma = _install_special_mapping(mm, vdso_base, vdso_pages << PAGE_SHIFT,
 				       VM_READ|VM_EXEC|
 				       VM_MAYREAD|VM_MAYWRITE|VM_MAYEXEC,
 				       &vdso_mapping);
 	if (IS_ERR(vma)) {
+		do_munmap(mm, vvar_start, PAGE_SIZE, NULL);
 		rc = PTR_ERR(vma);
-		goto out_up;
+	} else {
+		current->mm->context.vdso_base = vdso_text_start;
+		rc = 0;
 	}
-
+out:
-	current->mm->context.vdso_base = vdso_base;
-	rc = 0;
-
-out_up:
 	mmap_write_unlock(mm);
 	return rc;
 }
 
 static int __init vdso_init(void)
 {
+	struct page **pages;
 	int i;
 
-	vdso_getcpu_init();
+	vdso_pages = (vdso64_end - vdso64_start) >> PAGE_SHIFT;
-	/* Calculate the size of the 64 bit vDSO */
+	pages = kcalloc(vdso_pages + 1, sizeof(struct page *), GFP_KERNEL);
-	vdso64_pages = ((&vdso64_end - &vdso64_start
+	if (!pages) {
-			 + PAGE_SIZE - 1) >> PAGE_SHIFT) + 1;
+		vdso_enabled = 0;
-
+		return -ENOMEM;
-	/* Make sure pages are in the correct state */
-	vdso64_pagelist = kcalloc(vdso64_pages + 1, sizeof(struct page *),
-				  GFP_KERNEL);
-	BUG_ON(vdso64_pagelist == NULL);
-	for (i = 0; i < vdso64_pages - 1; i++) {
-		struct page *pg = virt_to_page(vdso64_kbase + i*PAGE_SIZE);
-		get_page(pg);
-		vdso64_pagelist[i] = pg;
 	}
-	vdso64_pagelist[vdso64_pages - 1] = virt_to_page(vdso_data);
+	for (i = 0; i < vdso_pages; i++)
-	vdso64_pagelist[vdso64_pages] = NULL;
+		pages[i] = virt_to_page(vdso64_start + i * PAGE_SIZE);
-
+	pages[vdso_pages] = NULL;
-	get_page(virt_to_page(vdso_data));
+	vdso_mapping.pages = pages;
-
 	return 0;
 }
-early_initcall(vdso_init);
+arch_initcall(vdso_init);

arch/s390/kernel/vdso64/getcpu.c

@@ -8,12 +8,12 @@
 
 int __s390_vdso_getcpu(unsigned *cpu, unsigned *node, struct getcpu_cache *unused)
 {
-	__u16 todval[8];
+	union tod_clock clk;
 
 	/* CPU number is stored in the programmable field of the TOD clock */
-	get_tod_clock_ext((char *)todval);
+	store_tod_clock_ext(&clk);
 	if (cpu)
-		*cpu = todval[7];
+		*cpu = clk.pf;
 	/* NUMA node is always zero */
 	if (node)
 		*node = 0;

arch/s390/kernel/vdso64/vdso64.lds.S

@@ -13,6 +13,10 @@ ENTRY(_start)
 
 SECTIONS
 {
+	PROVIDE(_vdso_data = . - __VVAR_PAGES * PAGE_SIZE);
+#ifdef CONFIG_TIME_NS
+	PROVIDE(_timens_data = _vdso_data + PAGE_SIZE);
+#endif
 	. = VDSO64_LBASE + SIZEOF_HEADERS;
 
 	.hash		: { *(.hash) }			:text
@@ -94,9 +98,6 @@ SECTIONS
 	.debug_ranges	0 : { *(.debug_ranges) }
 	.gnu.attributes 0 : { KEEP (*(.gnu.attributes)) }
 
-	. = ALIGN(PAGE_SIZE);
-	PROVIDE(_vdso_data = .);
-
 	/DISCARD/	: {
 		*(.note.GNU-stack)
 		*(.branch_lt)

arch/s390/kernel/vtime.c

@@ -13,7 +13,7 @@
 #include <linux/timex.h>
 #include <linux/types.h>
 #include <linux/time.h>
-
+#include <asm/alternative.h>
 #include <asm/vtimer.h>
 #include <asm/vtime.h>
 #include <asm/cpu_mf.h>
@@ -128,15 +128,13 @@ static int do_account_vtime(struct task_struct *tsk)
 
 	timer = S390_lowcore.last_update_timer;
 	clock = S390_lowcore.last_update_clock;
-	asm volatile(
+	/* Use STORE CLOCK by default, STORE CLOCK FAST if available. */
-		"	stpt	%0\n"	/* Store current cpu timer value */
+	alternative_io("stpt %0\n .insn s,0xb2050000,%1\n",
-#ifdef CONFIG_HAVE_MARCH_Z9_109_FEATURES
+		       "stpt %0\n .insn s,0xb27c0000,%1\n",
-		"	stckf	%1"	/* Store current tod clock value */
+		       25,
-#else
+		       ASM_OUTPUT2("=Q" (S390_lowcore.last_update_timer),
-		"	stck	%1"	/* Store current tod clock value */
+				   "=Q" (S390_lowcore.last_update_clock)),
-#endif
+		       ASM_NO_INPUT_CLOBBER("cc"));
-		: "=Q" (S390_lowcore.last_update_timer),
-		  "=Q" (S390_lowcore.last_update_clock));
 	clock = S390_lowcore.last_update_clock - clock;
 	timer -= S390_lowcore.last_update_timer;
 

arch/s390/kvm/kvm-s390.c

@@ -45,6 +45,7 @@
 #include <asm/timex.h>
 #include <asm/ap.h>
 #include <asm/uv.h>
+#include <asm/fpu/api.h>
 #include "kvm-s390.h"
 #include "gaccess.h"
 
@@ -164,12 +165,6 @@ struct kvm_stats_debugfs_item debugfs_entries[] = {
 	{ NULL }
 };
 
-struct kvm_s390_tod_clock_ext {
-	__u8 epoch_idx;
-	__u64 tod;
-	__u8 reserved[7];
-} __packed;
-
 /* allow nested virtualization in KVM (if enabled by user space) */
 static int nested;
 module_param(nested, int, S_IRUGO);
@@ -1165,17 +1160,17 @@ static int kvm_s390_set_tod(struct kvm *kvm, struct kvm_device_attr *attr)
 static void kvm_s390_get_tod_clock(struct kvm *kvm,
 				   struct kvm_s390_vm_tod_clock *gtod)
 {
-	struct kvm_s390_tod_clock_ext htod;
+	union tod_clock clk;
 
 	preempt_disable();
 
-	get_tod_clock_ext((char *)&htod);
+	store_tod_clock_ext(&clk);
 
-	gtod->tod = htod.tod + kvm->arch.epoch;
+	gtod->tod = clk.tod + kvm->arch.epoch;
 	gtod->epoch_idx = 0;
 	if (test_kvm_facility(kvm, 139)) {
-		gtod->epoch_idx = htod.epoch_idx + kvm->arch.epdx;
+		gtod->epoch_idx = clk.ei + kvm->arch.epdx;
-		if (gtod->tod < htod.tod)
+		if (gtod->tod < clk.tod)
 			gtod->epoch_idx += 1;
 	}
 
@@ -3866,18 +3861,18 @@ void kvm_s390_set_tod_clock(struct kvm *kvm,
 			    const struct kvm_s390_vm_tod_clock *gtod)
 {
 	struct kvm_vcpu *vcpu;
-	struct kvm_s390_tod_clock_ext htod;
+	union tod_clock clk;
 	int i;
 
 	mutex_lock(&kvm->lock);
 	preempt_disable();
 
-	get_tod_clock_ext((char *)&htod);
+	store_tod_clock_ext(&clk);
 
-	kvm->arch.epoch = gtod->tod - htod.tod;
+	kvm->arch.epoch = gtod->tod - clk.tod;
 	kvm->arch.epdx = 0;
 	if (test_kvm_facility(kvm, 139)) {
-		kvm->arch.epdx = gtod->epoch_idx - htod.epoch_idx;
+		kvm->arch.epdx = gtod->epoch_idx - clk.ei;
 		if (kvm->arch.epoch > gtod->tod)
 			kvm->arch.epdx -= 1;
 	}
@@ -4147,6 +4142,8 @@ static int __vcpu_run(struct kvm_vcpu *vcpu)
 			       vcpu->run->s.regs.gprs,
 			       sizeof(sie_page->pv_grregs));
 		}
+		if (test_cpu_flag(CIF_FPU))
+			load_fpu_regs();
 		exit_reason = sie64a(vcpu->arch.sie_block,
 				     vcpu->run->s.regs.gprs);
 		if (kvm_s390_pv_cpu_is_protected(vcpu)) {

arch/s390/kvm/vsie.c

@@ -18,6 +18,7 @@
 #include <asm/sclp.h>
 #include <asm/nmi.h>
 #include <asm/dis.h>
+#include <asm/fpu/api.h>
 #include "kvm-s390.h"
 #include "gaccess.h"
 
@@ -1028,6 +1029,8 @@ static int do_vsie_run(struct kvm_vcpu *vcpu, struct vsie_page *vsie_page)
 	 */
 	vcpu->arch.sie_block->prog0c |= PROG_IN_SIE;
 	barrier();
+	if (test_cpu_flag(CIF_FPU))
+		load_fpu_regs();
 	if (!kvm_s390_vcpu_sie_inhibited(vcpu))
 		rc = sie64a(scb_s, vcpu->run->s.regs.gprs);
 	barrier();

arch/s390/lib/uaccess.c

@@ -16,8 +16,8 @@
 #include <asm/mmu_context.h>
 #include <asm/facility.h>
 
-#ifdef CONFIG_DEBUG_USER_ASCE
+#ifdef CONFIG_DEBUG_ENTRY
-void debug_user_asce(void)
+void debug_user_asce(int exit)
 {
 	unsigned long cr1, cr7;
 
@@ -25,12 +25,14 @@ void debug_user_asce(void)
 	__ctl_store(cr7, 7, 7);
 	if (cr1 == S390_lowcore.kernel_asce && cr7 == S390_lowcore.user_asce)
 		return;
-	panic("incorrect ASCE on kernel exit\n"
+	panic("incorrect ASCE on kernel %s\n"
 	      "cr1:    %016lx cr7:  %016lx\n"
 	      "kernel: %016llx user: %016llx\n",
-	      cr1, cr7, S390_lowcore.kernel_asce, S390_lowcore.user_asce);
+	      exit ? "exit" : "entry", cr1, cr7,
+	      S390_lowcore.kernel_asce, S390_lowcore.user_asce);
+
 }
-#endif /*CONFIG_DEBUG_USER_ASCE */
+#endif /*CONFIG_DEBUG_ENTRY */
 
 #ifndef CONFIG_HAVE_MARCH_Z10_FEATURES
 static DEFINE_STATIC_KEY_FALSE(have_mvcos);

arch/s390/mm/fault.c

@@ -385,7 +385,7 @@ static inline vm_fault_t do_exception(struct pt_regs *regs, int access)
 	 * The instruction that caused the program check has
 	 * been nullified. Don't signal single step via SIGTRAP.
 	 */
-	clear_pt_regs_flag(regs, PIF_PER_TRAP);
+	clear_thread_flag(TIF_PER_TRAP);
 
 	if (kprobe_page_fault(regs, 14))
 		return 0;

arch/s390/pci/pci.c

@@ -695,43 +695,68 @@ void zpci_remove_device(struct zpci_dev *zdev)
 	}
 }
 
-int zpci_create_device(struct zpci_dev *zdev)
+/**
+ * zpci_create_device() - Create a new zpci_dev and add it to the zbus
+ * @fid: Function ID of the device to be created
+ * @fh: Current Function Handle of the device to be created
+ * @state: Initial state after creation either Standby or Configured
+ *
+ * Creates a new zpci device and adds it to its, possibly newly created, zbus
+ * as well as zpci_list.
+ *
+ * Returns: 0 on success, an error value otherwise
+ */
+int zpci_create_device(u32 fid, u32 fh, enum zpci_state state)
 {
+	struct zpci_dev *zdev;
 	int rc;
 
+	zpci_dbg(3, "add fid:%x, fh:%x, c:%d\n", fid, fh, state);
+	zdev = kzalloc(sizeof(*zdev), GFP_KERNEL);
+	if (!zdev)
+		return -ENOMEM;
+
+	/* FID and Function Handle are the static/dynamic identifiers */
+	zdev->fid = fid;
+	zdev->fh = fh;
+
+	/* Query function properties and update zdev */
+	rc = clp_query_pci_fn(zdev);
+	if (rc)
+		goto error;
+	zdev->state =  state;
+
 	kref_init(&zdev->kref);
+	mutex_init(&zdev->lock);
+
+	rc = zpci_init_iommu(zdev);
+	if (rc)
+		goto error;
+
+	if (zdev->state == ZPCI_FN_STATE_CONFIGURED) {
+		rc = zpci_enable_device(zdev);
+		if (rc)
+			goto error_destroy_iommu;
+	}
+
+	rc = zpci_bus_device_register(zdev, &pci_root_ops);
+	if (rc)
+		goto error_disable;
 
 	spin_lock(&zpci_list_lock);
 	list_add_tail(&zdev->entry, &zpci_list);
 	spin_unlock(&zpci_list_lock);
 
-	rc = zpci_init_iommu(zdev);
-	if (rc)
-		goto out;
-
-	mutex_init(&zdev->lock);
-	if (zdev->state == ZPCI_FN_STATE_CONFIGURED) {
-		rc = zpci_enable_device(zdev);
-		if (rc)
-			goto out_destroy_iommu;
-	}
-
-	rc = zpci_bus_device_register(zdev, &pci_root_ops);
-	if (rc)
-		goto out_disable;
-
 	return 0;
 
-out_disable:
+error_disable:
 	if (zdev->state == ZPCI_FN_STATE_ONLINE)
 		zpci_disable_device(zdev);
-
+error_destroy_iommu:
-out_destroy_iommu:
 	zpci_destroy_iommu(zdev);
-out:
+error:
-	spin_lock(&zpci_list_lock);
+	zpci_dbg(0, "add fid:%x, rc:%d\n", fid, rc);
-	list_del(&zdev->entry);
+	kfree(zdev);
-	spin_unlock(&zpci_list_lock);
 	return rc;
 }
 

arch/s390/pci/pci_clp.c

@@ -181,7 +181,7 @@ static int clp_store_query_pci_fn(struct zpci_dev *zdev,
 	return 0;
 }
 
-static int clp_query_pci_fn(struct zpci_dev *zdev, u32 fh)
+int clp_query_pci_fn(struct zpci_dev *zdev)
 {
 	struct clp_req_rsp_query_pci *rrb;
 	int rc;
@@ -194,7 +194,7 @@ static int clp_query_pci_fn(struct zpci_dev *zdev, u32 fh)
 	rrb->request.hdr.len = sizeof(rrb->request);
 	rrb->request.hdr.cmd = CLP_QUERY_PCI_FN;
 	rrb->response.hdr.len = sizeof(rrb->response);
-	rrb->request.fh = fh;
+	rrb->request.fh = zdev->fh;
 
 	rc = clp_req(rrb, CLP_LPS_PCI);
 	if (!rc && rrb->response.hdr.rsp == CLP_RC_OK) {
@@ -212,40 +212,6 @@ out:
 	return rc;
 }
 
-int clp_add_pci_device(u32 fid, u32 fh, int configured)
-{
-	struct zpci_dev *zdev;
-	int rc = -ENOMEM;
-
-	zpci_dbg(3, "add fid:%x, fh:%x, c:%d\n", fid, fh, configured);
-	zdev = kzalloc(sizeof(*zdev), GFP_KERNEL);
-	if (!zdev)
-		goto error;
-
-	zdev->fh = fh;
-	zdev->fid = fid;
-
-	/* Query function properties and update zdev */
-	rc = clp_query_pci_fn(zdev, fh);
-	if (rc)
-		goto error;
-
-	if (configured)
-		zdev->state = ZPCI_FN_STATE_CONFIGURED;
-	else
-		zdev->state = ZPCI_FN_STATE_STANDBY;
-
-	rc = zpci_create_device(zdev);
-	if (rc)
-		goto error;
-	return 0;
-
-error:
-	zpci_dbg(0, "add fid:%x, rc:%d\n", fid, rc);
-	kfree(zdev);
-	return rc;
-}
-
 static int clp_refresh_fh(u32 fid);
 /*
  * Enable/Disable a given PCI function and update its function handle if
@@ -408,7 +374,7 @@ static void __clp_add(struct clp_fh_list_entry *entry, void *data)
 
 	zdev = get_zdev_by_fid(entry->fid);
 	if (!zdev)
-		clp_add_pci_device(entry->fid, entry->fh, entry->config_state);
+		zpci_create_device(entry->fid, entry->fh, entry->config_state);
 }
 
 int clp_scan_pci_devices(void)

arch/s390/pci/pci_event.c

@@ -80,7 +80,7 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 	enum zpci_state state;
 	int ret;
 
-	if (zdev && zdev->zbus && zdev->zbus->bus)
+	if (zdev && zdev->zbus->bus)
 		pdev = pci_get_slot(zdev->zbus->bus, zdev->devfn);
 
 	zpci_err("avail CCDF:\n");
@@ -89,7 +89,7 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 	switch (ccdf->pec) {
 	case 0x0301: /* Reserved|Standby -> Configured */
 		if (!zdev) {
-			ret = clp_add_pci_device(ccdf->fid, ccdf->fh, 1);
+			zpci_create_device(ccdf->fid, ccdf->fh, ZPCI_FN_STATE_CONFIGURED);
 			break;
 		}
 		/* the configuration request may be stale */
@@ -116,7 +116,7 @@ static void __zpci_event_availability(struct zpci_ccdf_avail *ccdf)
 		break;
 	case 0x0302: /* Reserved -> Standby */
 		if (!zdev) {
-			clp_add_pci_device(ccdf->fid, ccdf->fh, 0);
+			zpci_create_device(ccdf->fid, ccdf->fh, ZPCI_FN_STATE_STANDBY);
 			break;
 		}
 		zdev->fh = ccdf->fh;

drivers/s390/char/tape_3590.c

@@ -761,7 +761,7 @@ tape_3590_done(struct tape_device *device, struct tape_request *request)
  * This function is called, when error recovery was successful
  */
 static inline int
-tape_3590_erp_succeded(struct tape_device *device, struct tape_request *request)
+tape_3590_erp_succeeded(struct tape_device *device, struct tape_request *request)
 {
 	DBF_EVENT(3, "Error Recovery successful for %s\n",
 		  tape_op_verbose[request->op]);
@@ -831,7 +831,7 @@ tape_3590_erp_basic(struct tape_device *device, struct tape_request *request,
 	case SENSE_BRA_PER:
 		return tape_3590_erp_failed(device, request, irb, rc);
 	case SENSE_BRA_CONT:
-		return tape_3590_erp_succeded(device, request);
+		return tape_3590_erp_succeeded(device, request);
 	case SENSE_BRA_RE:
 		return tape_3590_erp_retry(device, request, irb);
 	case SENSE_BRA_DRE:

drivers/s390/cio/css.c

@@ -225,18 +225,23 @@ struct subchannel *css_alloc_subchannel(struct subchannel_id schid,
 
 	INIT_WORK(&sch->todo_work, css_sch_todo);
 	sch->dev.release = &css_subchannel_release;
+	sch->dev.dma_mask = &sch->dma_mask;
 	device_initialize(&sch->dev);
 	/*
-	 * The physical addresses of some the dma structures that can
+	 * The physical addresses for some of the dma structures that can
 	 * belong to a subchannel need to fit 31 bit width (e.g. ccw).
 	 */
-	sch->dev.coherent_dma_mask = DMA_BIT_MASK(31);
+	ret = dma_set_coherent_mask(&sch->dev, DMA_BIT_MASK(31));
+	if (ret)
+		goto err;
 	/*
 	 * But we don't have such restrictions imposed on the stuff that
 	 * is handled by the streaming API.
 	 */
-	sch->dma_mask = DMA_BIT_MASK(64);
+	ret = dma_set_mask(&sch->dev, DMA_BIT_MASK(64));
-	sch->dev.dma_mask = &sch->dma_mask;
+	if (ret)
+		goto err;
+
 	return sch;
 
 err:
@@ -970,8 +975,11 @@ static int __init setup_css(int nr)
 	 * css->device as the device argument with the DMA API)
 	 * and are fine with 64 bit addresses.
 	 */
-	css->device.coherent_dma_mask = DMA_BIT_MASK(64);
+	ret = dma_coerce_mask_and_coherent(&css->device, DMA_BIT_MASK(64));
-	css->device.dma_mask = &css->device.coherent_dma_mask;
+	if (ret) {
+		kfree(css);
+		goto out_err;
+	}
 
 	mutex_init(&css->mutex);
 	ret = chsc_get_cssid_iid(nr, &css->cssid, &css->iid);

drivers/s390/cio/device.c

@@ -621,14 +621,6 @@ static const struct attribute_group *ccwdev_attr_groups[] = {
 	NULL,
 };
 
-static int ccw_device_add(struct ccw_device *cdev)
-{
-	struct device *dev = &cdev->dev;
-
-	dev->bus = &ccw_bus_type;
-	return device_add(dev);
-}
-
 static int match_dev_id(struct device *dev, const void *data)
 {
 	struct ccw_device *cdev = to_ccwdev(dev);
@@ -687,33 +679,47 @@ static struct ccw_device * io_subchannel_allocate_dev(struct subchannel *sch)
 {
 	struct ccw_device *cdev;
 	struct gen_pool *dma_pool;
+	int ret;
 
 	cdev  = kzalloc(sizeof(*cdev), GFP_KERNEL);
-	if (!cdev)
+	if (!cdev) {
+		ret = -ENOMEM;
 		goto err_cdev;
+	}
 	cdev->private = kzalloc(sizeof(struct ccw_device_private),
 				GFP_KERNEL | GFP_DMA);
-	if (!cdev->private)
+	if (!cdev->private) {
+		ret = -ENOMEM;
 		goto err_priv;
-	cdev->dev.coherent_dma_mask = sch->dev.coherent_dma_mask;
+	}
+
 	cdev->dev.dma_mask = sch->dev.dma_mask;
+	ret = dma_set_coherent_mask(&cdev->dev, sch->dev.coherent_dma_mask);
+	if (ret)
+		goto err_coherent_mask;
+
 	dma_pool = cio_gp_dma_create(&cdev->dev, 1);
-	if (!dma_pool)
+	if (!dma_pool) {
+		ret = -ENOMEM;
 		goto err_dma_pool;
+	}
 	cdev->private->dma_pool = dma_pool;
 	cdev->private->dma_area = cio_gp_dma_zalloc(dma_pool, &cdev->dev,
 					sizeof(*cdev->private->dma_area));
-	if (!cdev->private->dma_area)
+	if (!cdev->private->dma_area) {
+		ret = -ENOMEM;
 		goto err_dma_area;
+	}
 	return cdev;
 err_dma_area:
 	cio_gp_dma_destroy(dma_pool, &cdev->dev);
 err_dma_pool:
+err_coherent_mask:
 	kfree(cdev->private);
 err_priv:
 	kfree(cdev);
 err_cdev:
-	return ERR_PTR(-ENOMEM);
+	return ERR_PTR(ret);
 }
 
 static void ccw_device_todo(struct work_struct *work);
@@ -739,6 +745,7 @@ static int io_subchannel_initialize_dev(struct subchannel *sch,
 	cdev->ccwlock = sch->lock;
 	cdev->dev.parent = &sch->dev;
 	cdev->dev.release = ccw_device_release;
+	cdev->dev.bus = &ccw_bus_type;
 	cdev->dev.groups = ccwdev_attr_groups;
 	/* Do first half of device_register. */
 	device_initialize(&cdev->dev);
@@ -840,7 +847,7 @@ static void io_subchannel_register(struct ccw_device *cdev)
 		kobject_uevent(&sch->dev.kobj, KOBJ_ADD);
 	}
 	/* make it known to the system */
-	ret = ccw_device_add(cdev);
+	ret = device_add(&cdev->dev);
 	if (ret) {
 		CIO_MSG_EVENT(0, "Could not register ccw dev 0.%x.%04x: %d\n",
 			      cdev->private->dev_id.ssid,
@@ -1052,7 +1059,7 @@ static int io_subchannel_probe(struct subchannel *sch)
 			kobject_uevent(&sch->dev.kobj, KOBJ_ADD);
 		}
 		cdev = sch_get_cdev(sch);
-		rc = ccw_device_add(cdev);
+		rc = device_add(&cdev->dev);
 		if (rc) {
 			/* Release online reference. */
 			put_device(&cdev->dev);

drivers/s390/cio/qdio.h

@@ -139,9 +139,6 @@ struct qdio_dev_perf_stat {
 	unsigned int qdio_int;
 	unsigned int pci_request_int;
 
-	unsigned int tasklet_inbound;
-	unsigned int tasklet_inbound_resched;
-	unsigned int tasklet_inbound_resched2;
 	unsigned int tasklet_outbound;
 
 	unsigned int siga_read;
@@ -149,7 +146,6 @@ struct qdio_dev_perf_stat {
 	unsigned int siga_sync;
 
 	unsigned int inbound_call;
-	unsigned int inbound_handler;
 	unsigned int stop_polling;
 	unsigned int inbound_queue_full;
 	unsigned int outbound_call;
@@ -193,6 +189,8 @@ struct qdio_output_q {
 	struct qdio_outbuf_state *sbal_state;
 	/* timer to check for more outbound work */
 	struct timer_list timer;
+	/* tasklet to check for completions */
+	struct tasklet_struct tasklet;
 };
 
 /*
@@ -216,13 +214,9 @@ struct qdio_q {
 	/* number of buffers in use by the adapter */
 	atomic_t nr_buf_used;
 
-	/* error condition during a data transfer */
-	unsigned int qdio_error;
-
 	/* last scan of the queue */
 	u64 timestamp;
 
-	struct tasklet_struct tasklet;
 	struct qdio_queue_perf_stat q_stats;
 
 	struct qdio_buffer *sbal[QDIO_MAX_BUFFERS_PER_Q] ____cacheline_aligned;
@@ -254,6 +248,7 @@ struct qdio_irq {
 	struct ccw_device *cdev;
 	struct list_head entry;		/* list of thinint devices */
 	struct dentry *debugfs_dev;
+	u64 last_data_irq_time;
 
 	unsigned long int_parm;
 	struct subchannel_id schid;
@@ -324,6 +319,14 @@ static inline int multicast_outbound(struct qdio_q *q)
 	       (q->nr == q->irq_ptr->nr_output_qs - 1);
 }
 
+static inline void qdio_deliver_irq(struct qdio_irq *irq)
+{
+	if (!test_and_set_bit(QDIO_IRQ_DISABLED, &irq->poll_state))
+		irq->irq_poll(irq->cdev, irq->int_parm);
+	else
+		QDIO_PERF_STAT_INC(irq, int_discarded);
+}
+
 #define pci_out_supported(irq) ((irq)->qib.ac & QIB_AC_OUTBOUND_PCI_SUPPORTED)
 #define is_qebsm(q)			(q->irq_ptr->sch_token != 0)
 
@@ -357,16 +360,12 @@ extern u64 last_ai_time;
 /* prototypes for thin interrupt */
 int qdio_establish_thinint(struct qdio_irq *irq_ptr);
 void qdio_shutdown_thinint(struct qdio_irq *irq_ptr);
-void tiqdio_add_device(struct qdio_irq *irq_ptr);
-void tiqdio_remove_device(struct qdio_irq *irq_ptr);
-void tiqdio_inbound_processing(unsigned long q);
 int qdio_thinint_init(void);
 void qdio_thinint_exit(void);
 int test_nonshared_ind(struct qdio_irq *);
 
 /* prototypes for setup */
-void qdio_inbound_processing(unsigned long data);
+void qdio_outbound_tasklet(struct tasklet_struct *t);
-void qdio_outbound_processing(unsigned long data);
 void qdio_outbound_timer(struct timer_list *t);
 void qdio_int_handler(struct ccw_device *cdev, unsigned long intparm,
 		      struct irb *irb);

drivers/s390/cio/qdio_debug.c

@@ -105,8 +105,9 @@ static int qstat_show(struct seq_file *m, void *v)
 	if (!q)
 		return 0;
 
-	seq_printf(m, "Timestamp: %Lx  Last AI: %Lx\n",
+	seq_printf(m, "Timestamp: %llx\n", q->timestamp);
-		   q->timestamp, last_ai_time);
+	seq_printf(m, "Last Data IRQ: %llx  Last AI: %llx\n",
+		   q->irq_ptr->last_data_irq_time, last_ai_time);
 	seq_printf(m, "nr_used: %d  ftc: %d\n",
 		   atomic_read(&q->nr_buf_used), q->first_to_check);
 	if (q->is_input_q) {
@@ -197,15 +198,11 @@ static char *qperf_names[] = {
 	"Assumed adapter interrupts",
 	"QDIO interrupts",
 	"Requested PCIs",
-	"Inbound tasklet runs",
-	"Inbound tasklet resched",
-	"Inbound tasklet resched2",
 	"Outbound tasklet runs",
 	"SIGA read",
 	"SIGA write",
 	"SIGA sync",
 	"Inbound calls",
-	"Inbound handler",
 	"Inbound stop_polling",
 	"Inbound queue full",
 	"Outbound calls",

drivers/s390/cio/qdio_main.c

@@ -202,7 +202,7 @@ again:
  */
 static inline int get_buf_states(struct qdio_q *q, unsigned int bufnr,
 				 unsigned char *state, unsigned int count,
-				 int auto_ack, int merge_pending)
+				 int auto_ack)
 {
 	unsigned char __state = 0;
 	int i = 1;
@@ -217,18 +217,9 @@ static inline int get_buf_states(struct qdio_q *q, unsigned int bufnr,
 	if (__state & SLSB_OWNER_CU)
 		goto out;
 
-	if (merge_pending && __state == SLSB_P_OUTPUT_PENDING)
-		__state = SLSB_P_OUTPUT_EMPTY;
-
 	for (; i < count; i++) {
 		bufnr = next_buf(bufnr);
 
-		/* merge PENDING into EMPTY: */
-		if (merge_pending &&
-		    q->slsb.val[bufnr] == SLSB_P_OUTPUT_PENDING &&
-		    __state == SLSB_P_OUTPUT_EMPTY)
-			continue;
-
 		/* stop if next state differs from initial state: */
 		if (q->slsb.val[bufnr] != __state)
 			break;
@@ -242,7 +233,7 @@ out:
 static inline int get_buf_state(struct qdio_q *q, unsigned int bufnr,
 				unsigned char *state, int auto_ack)
 {
-	return get_buf_states(q, bufnr, state, 1, auto_ack, 0);
+	return get_buf_states(q, bufnr, state, 1, auto_ack);
 }
 
 /* wrap-around safe setting of slsb states, returns number of changed buffers */
@@ -420,8 +411,6 @@ static inline void account_sbals(struct qdio_q *q, unsigned int count)
 static void process_buffer_error(struct qdio_q *q, unsigned int start,
 				 int count)
 {
-	q->qdio_error = QDIO_ERROR_SLSB_STATE;
-
 	/* special handling for no target buffer empty */
 	if (queue_type(q) == QDIO_IQDIO_QFMT && !q->is_input_q &&
 	    q->sbal[start]->element[15].sflags == 0x10) {
@@ -450,7 +439,8 @@ static inline void inbound_handle_work(struct qdio_q *q, unsigned int start,
 	q->u.in.batch_count += count;
 }
 
-static int get_inbound_buffer_frontier(struct qdio_q *q, unsigned int start)
+static int get_inbound_buffer_frontier(struct qdio_q *q, unsigned int start,
+				       unsigned int *error)
 {
 	unsigned char state = 0;
 	int count;
@@ -465,7 +455,7 @@ static int get_inbound_buffer_frontier(struct qdio_q *q, unsigned int start)
 	 * No siga sync here, as a PCI or we after a thin interrupt
 	 * already sync'ed the queues.
 	 */
-	count = get_buf_states(q, start, &state, count, 1, 0);
+	count = get_buf_states(q, start, &state, count, 1);
 	if (!count)
 		return 0;
 
@@ -484,6 +474,7 @@ static int get_inbound_buffer_frontier(struct qdio_q *q, unsigned int start)
 		DBF_DEV_EVENT(DBF_INFO, q->irq_ptr, "in err:%1d %02x", q->nr,
 			      count);
 
+		*error = QDIO_ERROR_SLSB_STATE;
 		process_buffer_error(q, start, count);
 		inbound_handle_work(q, start, count, false);
 		if (atomic_sub_return(count, &q->nr_buf_used) == 0)
@@ -508,11 +499,6 @@ static int get_inbound_buffer_frontier(struct qdio_q *q, unsigned int start)
 	}
 }
 
-static int qdio_inbound_q_moved(struct qdio_q *q, unsigned int start)
-{
-	return get_inbound_buffer_frontier(q, start);
-}
-
 static inline int qdio_inbound_q_done(struct qdio_q *q, unsigned int start)
 {
 	unsigned char state = 0;
@@ -546,96 +532,23 @@ static inline unsigned long qdio_aob_for_buffer(struct qdio_output_q *q,
 		WARN_ON_ONCE(phys_aob & 0xFF);
 	}
 
-	q->sbal_state[bufnr].flags = 0;
 	return phys_aob;
 }
 
-static void qdio_kick_handler(struct qdio_q *q, unsigned int start,
-			      unsigned int count)
-{
-	if (unlikely(q->irq_ptr->state != QDIO_IRQ_STATE_ACTIVE))
-		return;
-
-	if (q->is_input_q) {
-		qperf_inc(q, inbound_handler);
-		DBF_DEV_EVENT(DBF_INFO, q->irq_ptr, "kih s:%02x c:%02x", start, count);
-	} else {
-		qperf_inc(q, outbound_handler);
-		DBF_DEV_EVENT(DBF_INFO, q->irq_ptr, "koh: s:%02x c:%02x",
-			      start, count);
-	}
-
-	q->handler(q->irq_ptr->cdev, q->qdio_error, q->nr, start, count,
-		   q->irq_ptr->int_parm);
-
-	/* for the next time */
-	q->qdio_error = 0;
-}
-
 static inline int qdio_tasklet_schedule(struct qdio_q *q)
 {
 	if (likely(q->irq_ptr->state == QDIO_IRQ_STATE_ACTIVE)) {
-		tasklet_schedule(&q->tasklet);
+		tasklet_schedule(&q->u.out.tasklet);
 		return 0;
 	}
 	return -EPERM;
 }
 
-static void __qdio_inbound_processing(struct qdio_q *q)
+static int get_outbound_buffer_frontier(struct qdio_q *q, unsigned int start,
-{
+					unsigned int *error)
-	unsigned int start = q->first_to_check;
-	int count;
-
-	qperf_inc(q, tasklet_inbound);
-
-	count = qdio_inbound_q_moved(q, start);
-	if (count == 0)
-		return;
-
-	qdio_kick_handler(q, start, count);
-	start = add_buf(start, count);
-	q->first_to_check = start;
-
-	if (!qdio_inbound_q_done(q, start)) {
-		/* means poll time is not yet over */
-		qperf_inc(q, tasklet_inbound_resched);
-		if (!qdio_tasklet_schedule(q))
-			return;
-	}
-
-	qdio_stop_polling(q);
-	/*
-	 * We need to check again to not lose initiative after
-	 * resetting the ACK state.
-	 */
-	if (!qdio_inbound_q_done(q, start)) {
-		qperf_inc(q, tasklet_inbound_resched2);
-		qdio_tasklet_schedule(q);
-	}
-}
-
-void qdio_inbound_processing(unsigned long data)
-{
-	struct qdio_q *q = (struct qdio_q *)data;
-	__qdio_inbound_processing(q);
-}
-
-static void qdio_check_pending(struct qdio_q *q, unsigned int index)
-{
-	unsigned char state;
-
-	if (get_buf_state(q, index, &state, 0) > 0 &&
-	    state == SLSB_P_OUTPUT_PENDING &&
-	    q->u.out.aobs[index]) {
-		q->u.out.sbal_state[index].flags |=
-			QDIO_OUTBUF_STATE_FLAG_PENDING;
-		q->u.out.aobs[index] = NULL;
-	}
-}
-
-static int get_outbound_buffer_frontier(struct qdio_q *q, unsigned int start)
 {
 	unsigned char state = 0;
+	unsigned int i;
 	int count;
 
 	q->timestamp = get_tod_clock_fast();
@@ -651,13 +564,19 @@ static int get_outbound_buffer_frontier(struct qdio_q *q, unsigned int start)
 	if (!count)
 		return 0;
 
-	count = get_buf_states(q, start, &state, count, 0, q->u.out.use_cq);
+	count = get_buf_states(q, start, &state, count, 0);
 	if (!count)
 		return 0;
 
 	switch (state) {
-	case SLSB_P_OUTPUT_EMPTY:
 	case SLSB_P_OUTPUT_PENDING:
+		/* detach the utilized QAOBs: */
+		for (i = 0; i < count; i++)
+			q->u.out.aobs[QDIO_BUFNR(start + i)] = NULL;
+
+		*error = QDIO_ERROR_SLSB_PENDING;
+		fallthrough;
+	case SLSB_P_OUTPUT_EMPTY:
 		/* the adapter got it */
 		DBF_DEV_EVENT(DBF_INFO, q->irq_ptr,
 			"out empty:%1d %02x", q->nr, count);
@@ -667,6 +586,10 @@ static int get_outbound_buffer_frontier(struct qdio_q *q, unsigned int start)
 			account_sbals(q, count);
 		return count;
 	case SLSB_P_OUTPUT_ERROR:
+		DBF_DEV_EVENT(DBF_INFO, q->irq_ptr, "out error:%1d %02x",
+			      q->nr, count);
+
+		*error = QDIO_ERROR_SLSB_STATE;
 		process_buffer_error(q, start, count);
 		atomic_sub(count, &q->nr_buf_used);
 		if (q->irq_ptr->perf_stat_enabled)
@@ -697,26 +620,6 @@ static inline int qdio_outbound_q_done(struct qdio_q *q)
 	return atomic_read(&q->nr_buf_used) == 0;
 }
 
-static inline int qdio_outbound_q_moved(struct qdio_q *q, unsigned int start)
-{
-	int count;
-
-	count = get_outbound_buffer_frontier(q, start);
-
-	if (count) {
-		DBF_DEV_EVENT(DBF_INFO, q->irq_ptr, "out moved:%1d", q->nr);
-
-		if (q->u.out.use_cq) {
-			unsigned int i;
-
-			for (i = 0; i < count; i++)
-				qdio_check_pending(q, QDIO_BUFNR(start + i));
-		}
-	}
-
-	return count;
-}
-
 static int qdio_kick_outbound_q(struct qdio_q *q, unsigned int count,
 				unsigned long aob)
 {
@@ -760,18 +663,29 @@ retry:
 	return cc;
 }
 
-static void __qdio_outbound_processing(struct qdio_q *q)
+void qdio_outbound_tasklet(struct tasklet_struct *t)
 {
+	struct qdio_output_q *out_q = from_tasklet(out_q, t, tasklet);
+	struct qdio_q *q = container_of(out_q, struct qdio_q, u.out);
 	unsigned int start = q->first_to_check;
+	unsigned int error = 0;
 	int count;
 
 	qperf_inc(q, tasklet_outbound);
 	WARN_ON_ONCE(atomic_read(&q->nr_buf_used) < 0);
 
-	count = qdio_outbound_q_moved(q, start);
+	count = get_outbound_buffer_frontier(q, start, &error);
 	if (count) {
 		q->first_to_check = add_buf(start, count);
-		qdio_kick_handler(q, start, count);
+
+		if (q->irq_ptr->state == QDIO_IRQ_STATE_ACTIVE) {
+			qperf_inc(q, outbound_handler);
+			DBF_DEV_EVENT(DBF_INFO, q->irq_ptr, "koh: s:%02x c:%02x",
+				      start, count);
+
+			q->handler(q->irq_ptr->cdev, error, q->nr, start,
+				   count, q->irq_ptr->int_parm);
+		}
 	}
 
 	if (queue_type(q) == QDIO_ZFCP_QFMT && !pci_out_supported(q->irq_ptr) &&
@@ -798,13 +712,6 @@ sched:
 	qdio_tasklet_schedule(q);
 }
 
-/* outbound tasklet */
-void qdio_outbound_processing(unsigned long data)
-{
-	struct qdio_q *q = (struct qdio_q *)data;
-	__qdio_outbound_processing(q);
-}
-
 void qdio_outbound_timer(struct timer_list *t)
 {
 	struct qdio_q *q = from_timer(q, t, u.out.timer);
@@ -825,19 +732,6 @@ static inline void qdio_check_outbound_pci_queues(struct qdio_irq *irq)
 			qdio_tasklet_schedule(out);
 }
 
-void tiqdio_inbound_processing(unsigned long data)
-{
-	struct qdio_q *q = (struct qdio_q *)data;
-
-	if (need_siga_sync(q) && need_siga_sync_after_ai(q))
-		qdio_sync_queues(q);
-
-	/* The interrupt could be caused by a PCI request: */
-	qdio_check_outbound_pci_queues(q->irq_ptr);
-
-	__qdio_inbound_processing(q);
-}
-
 static inline void qdio_set_state(struct qdio_irq *irq_ptr,
 				  enum qdio_irq_states state)
 {
@@ -865,15 +759,8 @@ static void qdio_int_handler_pci(struct qdio_irq *irq_ptr)
 	if (unlikely(irq_ptr->state != QDIO_IRQ_STATE_ACTIVE))
 		return;
 
-	if (irq_ptr->irq_poll) {
+	qdio_deliver_irq(irq_ptr);
-		if (!test_and_set_bit(QDIO_IRQ_DISABLED, &irq_ptr->poll_state))
+	irq_ptr->last_data_irq_time = S390_lowcore.int_clock;
-			irq_ptr->irq_poll(irq_ptr->cdev, irq_ptr->int_parm);
-		else
-			QDIO_PERF_STAT_INC(irq_ptr, int_discarded);
-	} else {
-		for_each_input_queue(irq_ptr, q, i)
-			tasklet_schedule(&q->tasklet);
-	}
 
 	if (!pci_out_supported(irq_ptr) || !irq_ptr->scan_threshold)
 		return;
@@ -1016,12 +903,9 @@ static void qdio_shutdown_queues(struct qdio_irq *irq_ptr)
 	struct qdio_q *q;
 	int i;
 
-	for_each_input_queue(irq_ptr, q, i)
-		tasklet_kill(&q->tasklet);
-
 	for_each_output_queue(irq_ptr, q, i) {
 		del_timer_sync(&q->u.out.timer);
-		tasklet_kill(&q->tasklet);
+		tasklet_kill(&q->u.out.tasklet);
 	}
 }
 
@@ -1059,7 +943,6 @@ int qdio_shutdown(struct ccw_device *cdev, int how)
 	 */
 	qdio_set_state(irq_ptr, QDIO_IRQ_STATE_STOPPED);
 
-	tiqdio_remove_device(irq_ptr);
 	qdio_shutdown_queues(irq_ptr);
 	qdio_shutdown_debug_entries(irq_ptr);
 
@@ -1177,7 +1060,6 @@ int qdio_allocate(struct ccw_device *cdev, unsigned int no_input_qs,
 	if (rc)
 		goto err_queues;
 
-	INIT_LIST_HEAD(&irq_ptr->entry);
 	cdev->private->qdio_data = irq_ptr;
 	qdio_set_state(irq_ptr, QDIO_IRQ_STATE_INACTIVE);
 	return 0;
@@ -1263,6 +1145,9 @@ int qdio_establish(struct ccw_device *cdev,
 	    !init_data->output_sbal_addr_array)
 		return -EINVAL;
 
+	if (!init_data->irq_poll)
+		return -EINVAL;
+
 	mutex_lock(&irq_ptr->setup_mutex);
 	qdio_trace_init_data(irq_ptr, init_data);
 	qdio_setup_irq(irq_ptr, init_data);
@@ -1357,9 +1242,6 @@ int qdio_activate(struct ccw_device *cdev)
 		goto out;
 	}
 
-	if (is_thinint_irq(irq_ptr))
-		tiqdio_add_device(irq_ptr);
-
 	/* wait for subchannel to become active */
 	msleep(5);
 
@@ -1557,17 +1439,16 @@ static int __qdio_inspect_queue(struct qdio_q *q, unsigned int *bufnr,
 	unsigned int start = q->first_to_check;
 	int count;
 
-	count = q->is_input_q ? qdio_inbound_q_moved(q, start) :
+	*error = 0;
-				qdio_outbound_q_moved(q, start);
+	count = q->is_input_q ? get_inbound_buffer_frontier(q, start, error) :
+				get_outbound_buffer_frontier(q, start, error);
 	if (count == 0)
 		return 0;
 
 	*bufnr = start;
-	*error = q->qdio_error;
 
 	/* for the next time */
 	q->first_to_check = add_buf(start, count);
-	q->qdio_error = 0;
 
 	return count;
 }

drivers/s390/cio/qdio_setup.c

@@ -259,14 +259,6 @@ static void setup_queues(struct qdio_irq *irq_ptr,
 
 		setup_storage_lists(q, irq_ptr,
 				    qdio_init->input_sbal_addr_array[i], i);
-
-		if (is_thinint_irq(irq_ptr)) {
-			tasklet_init(&q->tasklet, tiqdio_inbound_processing,
-				     (unsigned long) q);
-		} else {
-			tasklet_init(&q->tasklet, qdio_inbound_processing,
-				     (unsigned long) q);
-		}
 	}
 
 	for_each_output_queue(irq_ptr, q, i) {
@@ -280,8 +272,7 @@ static void setup_queues(struct qdio_irq *irq_ptr,
 		setup_storage_lists(q, irq_ptr,
 				    qdio_init->output_sbal_addr_array[i], i);
 
-		tasklet_init(&q->tasklet, qdio_outbound_processing,
+		tasklet_setup(&q->u.out.tasklet, qdio_outbound_tasklet);
-			     (unsigned long) q);
 		timer_setup(&q->u.out.timer, qdio_outbound_timer, 0);
 	}
 }
@@ -483,12 +474,8 @@ int qdio_setup_irq(struct qdio_irq *irq_ptr, struct qdio_initialize *init_data)
 	ccw_device_get_schid(cdev, &irq_ptr->schid);
 	setup_queues(irq_ptr, init_data);
 
-	if (init_data->irq_poll) {
+	irq_ptr->irq_poll = init_data->irq_poll;
-		irq_ptr->irq_poll = init_data->irq_poll;
+	set_bit(QDIO_IRQ_DISABLED, &irq_ptr->poll_state);
-		set_bit(QDIO_IRQ_DISABLED, &irq_ptr->poll_state);
-	} else {
-		irq_ptr->irq_poll = NULL;
-	}
 
 	setup_qib(irq_ptr, init_data);
 	set_impl_params(irq_ptr, init_data->qib_param_field_format,

drivers/s390/cio/qdio_thinint.c

@@ -66,22 +66,6 @@ static void put_indicator(u32 *addr)
 	atomic_dec(&ind->count);
 }
 
-void tiqdio_add_device(struct qdio_irq *irq_ptr)
-{
-	mutex_lock(&tiq_list_lock);
-	list_add_rcu(&irq_ptr->entry, &tiq_list);
-	mutex_unlock(&tiq_list_lock);
-}
-
-void tiqdio_remove_device(struct qdio_irq *irq_ptr)
-{
-	mutex_lock(&tiq_list_lock);
-	list_del_rcu(&irq_ptr->entry);
-	mutex_unlock(&tiq_list_lock);
-	synchronize_rcu();
-	INIT_LIST_HEAD(&irq_ptr->entry);
-}
-
 static inline int references_shared_dsci(struct qdio_irq *irq_ptr)
 {
 	return irq_ptr->dsci == &q_indicators[TIQDIO_SHARED_IND].ind;
@@ -106,32 +90,6 @@ static inline u32 clear_shared_ind(void)
 	return xchg(&q_indicators[TIQDIO_SHARED_IND].ind, 0);
 }
 
-static inline void tiqdio_call_inq_handlers(struct qdio_irq *irq)
-{
-	struct qdio_q *q;
-	int i;
-
-	if (!references_shared_dsci(irq))
-		xchg(irq->dsci, 0);
-
-	if (irq->irq_poll) {
-		if (!test_and_set_bit(QDIO_IRQ_DISABLED, &irq->poll_state))
-			irq->irq_poll(irq->cdev, irq->int_parm);
-		else
-			QDIO_PERF_STAT_INC(irq, int_discarded);
-
-		return;
-	}
-
-	for_each_input_queue(irq, q, i) {
-		/*
-		 * Call inbound processing but not directly
-		 * since that could starve other thinint queues.
-		 */
-		tasklet_schedule(&q->tasklet);
-	}
-}
-
 /**
  * tiqdio_thinint_handler - thin interrupt handler for qdio
  * @airq: pointer to adapter interrupt descriptor
@@ -139,10 +97,11 @@ static inline void tiqdio_call_inq_handlers(struct qdio_irq *irq)
  */
 static void tiqdio_thinint_handler(struct airq_struct *airq, bool floating)
 {
+	u64 irq_time = S390_lowcore.int_clock;
 	u32 si_used = clear_shared_ind();
 	struct qdio_irq *irq;
 
-	last_ai_time = S390_lowcore.int_clock;
+	last_ai_time = irq_time;
 	inc_irq_stat(IRQIO_QAI);
 
 	/* protect tiq_list entries, only changed in activate or shutdown */
@@ -153,10 +112,15 @@ static void tiqdio_thinint_handler(struct airq_struct *airq, bool floating)
 		if (unlikely(references_shared_dsci(irq))) {
 			if (!si_used)
 				continue;
-		} else if (!*irq->dsci)
+		} else {
-			continue;
+			if (!*irq->dsci)
+				continue;
 
-		tiqdio_call_inq_handlers(irq);
+			xchg(irq->dsci, 0);
+		}
+
+		qdio_deliver_irq(irq);
+		irq->last_data_irq_time = irq_time;
 
 		QDIO_PERF_STAT_INC(irq, adapter_int);
 	}
@@ -208,10 +172,15 @@ int qdio_establish_thinint(struct qdio_irq *irq_ptr)
 	DBF_HEX(&irq_ptr->dsci, sizeof(void *));
 
 	rc = set_subchannel_ind(irq_ptr, 0);
-	if (rc)
+	if (rc) {
 		put_indicator(irq_ptr->dsci);
+		return rc;
+	}
 
-	return rc;
+	mutex_lock(&tiq_list_lock);
+	list_add_rcu(&irq_ptr->entry, &tiq_list);
+	mutex_unlock(&tiq_list_lock);
+	return 0;
 }
 
 void qdio_shutdown_thinint(struct qdio_irq *irq_ptr)
@@ -219,6 +188,11 @@ void qdio_shutdown_thinint(struct qdio_irq *irq_ptr)
 	if (!is_thinint_irq(irq_ptr))
 		return;
 
+	mutex_lock(&tiq_list_lock);
+	list_del_rcu(&irq_ptr->entry);
+	mutex_unlock(&tiq_list_lock);
+	synchronize_rcu();
+
 	/* reset adapter interrupt indicators */
 	set_subchannel_ind(irq_ptr, 1);
 	put_indicator(irq_ptr->dsci);

drivers/s390/crypto/zcrypt_api.c

@@ -1438,6 +1438,8 @@ static int icarsamodexpo_ioctl(struct ap_perms *perms, unsigned long arg)
 			if (rc == -EAGAIN)
 				tr.again_counter++;
 		} while (rc == -EAGAIN && tr.again_counter < TRACK_AGAIN_MAX);
+	if (rc == -EAGAIN && tr.again_counter >= TRACK_AGAIN_MAX)
+		rc = -EIO;
 	if (rc) {
 		ZCRYPT_DBF(DBF_DEBUG, "ioctl ICARSAMODEXPO rc=%d\n", rc);
 		return rc;
@@ -1481,6 +1483,8 @@ static int icarsacrt_ioctl(struct ap_perms *perms, unsigned long arg)
 			if (rc == -EAGAIN)
 				tr.again_counter++;
 		} while (rc == -EAGAIN && tr.again_counter < TRACK_AGAIN_MAX);
+	if (rc == -EAGAIN && tr.again_counter >= TRACK_AGAIN_MAX)
+		rc = -EIO;
 	if (rc) {
 		ZCRYPT_DBF(DBF_DEBUG, "ioctl ICARSACRT rc=%d\n", rc);
 		return rc;
@@ -1524,6 +1528,8 @@ static int zsecsendcprb_ioctl(struct ap_perms *perms, unsigned long arg)
 			if (rc == -EAGAIN)
 				tr.again_counter++;
 		} while (rc == -EAGAIN && tr.again_counter < TRACK_AGAIN_MAX);
+	if (rc == -EAGAIN && tr.again_counter >= TRACK_AGAIN_MAX)
+		rc = -EIO;
 	if (rc)
 		ZCRYPT_DBF(DBF_DEBUG, "ioctl ZSENDCPRB rc=%d status=0x%x\n",
 			   rc, xcRB.status);
@@ -1568,6 +1574,8 @@ static int zsendep11cprb_ioctl(struct ap_perms *perms, unsigned long arg)
 			if (rc == -EAGAIN)
 				tr.again_counter++;
 		} while (rc == -EAGAIN && tr.again_counter < TRACK_AGAIN_MAX);
+	if (rc == -EAGAIN && tr.again_counter >= TRACK_AGAIN_MAX)
+		rc = -EIO;
 	if (rc)
 		ZCRYPT_DBF(DBF_DEBUG, "ioctl ZSENDEP11CPRB rc=%d\n", rc);
 	if (copy_to_user(uxcrb, &xcrb, sizeof(xcrb)))
@@ -1744,6 +1752,8 @@ static long trans_modexpo32(struct ap_perms *perms, struct file *filp,
 			if (rc == -EAGAIN)
 				tr.again_counter++;
 		} while (rc == -EAGAIN && tr.again_counter < TRACK_AGAIN_MAX);
+	if (rc == -EAGAIN && tr.again_counter >= TRACK_AGAIN_MAX)
+		rc = -EIO;
 	if (rc)
 		return rc;
 	return put_user(mex64.outputdatalength,
@@ -1795,6 +1805,8 @@ static long trans_modexpo_crt32(struct ap_perms *perms, struct file *filp,
 			if (rc == -EAGAIN)
 				tr.again_counter++;
 		} while (rc == -EAGAIN && tr.again_counter < TRACK_AGAIN_MAX);
+	if (rc == -EAGAIN && tr.again_counter >= TRACK_AGAIN_MAX)
+		rc = -EIO;
 	if (rc)
 		return rc;
 	return put_user(crt64.outputdatalength,
@@ -1865,6 +1877,8 @@ static long trans_xcRB32(struct ap_perms *perms, struct file *filp,
 			if (rc == -EAGAIN)
 				tr.again_counter++;
 		} while (rc == -EAGAIN && tr.again_counter < TRACK_AGAIN_MAX);
+	if (rc == -EAGAIN && tr.again_counter >= TRACK_AGAIN_MAX)
+		rc = -EIO;
 	xcRB32.reply_control_blk_length = xcRB64.reply_control_blk_length;
 	xcRB32.reply_data_length = xcRB64.reply_data_length;
 	xcRB32.status = xcRB64.status;

drivers/s390/crypto/zcrypt_ccamisc.c

@@ -662,7 +662,10 @@ int cca_sec2protkey(u16 cardnr, u16 domain,
 			  __func__,
 			  (int) prepcblk->ccp_rtcode,
 			  (int) prepcblk->ccp_rscode);
-		rc = -EIO;
+		if (prepcblk->ccp_rtcode == 8 && prepcblk->ccp_rscode == 2290)
+			rc = -EAGAIN;
+		else
+			rc = -EIO;
 		goto out;
 	}
 	if (prepcblk->ccp_rscode != 0) {
@@ -1275,7 +1278,10 @@ int cca_cipher2protkey(u16 cardnr, u16 domain, const u8 *ckey,
 			__func__,
 			(int) prepcblk->ccp_rtcode,
 			(int) prepcblk->ccp_rscode);
-		rc = -EIO;
+		if (prepcblk->ccp_rtcode == 8 && prepcblk->ccp_rscode == 2290)
+			rc = -EAGAIN;
+		else
+			rc = -EIO;
 		goto out;
 	}
 	if (prepcblk->ccp_rscode != 0) {
@@ -1441,7 +1447,10 @@ int cca_ecc2protkey(u16 cardnr, u16 domain, const u8 *key,
 			__func__,
 			(int) prepcblk->ccp_rtcode,
 			(int) prepcblk->ccp_rscode);
-		rc = -EIO;
+		if (prepcblk->ccp_rtcode == 8 && prepcblk->ccp_rscode == 2290)
+			rc = -EAGAIN;
+		else
+			rc = -EIO;
 		goto out;
 	}
 	if (prepcblk->ccp_rscode != 0) {

drivers/s390/net/qeth_core_main.c

@@ -6074,14 +6074,15 @@ int qeth_poll(struct napi_struct *napi, int budget)
 EXPORT_SYMBOL_GPL(qeth_poll);
 
 static void qeth_iqd_tx_complete(struct qeth_qdio_out_q *queue,
-				 unsigned int bidx, bool error, int budget)
+				 unsigned int bidx, unsigned int qdio_error,
+				 int budget)
 {
 	struct qeth_qdio_out_buffer *buffer = queue->bufs[bidx];
 	u8 sflags = buffer->buffer->element[15].sflags;
 	struct qeth_card *card = queue->card;
+	bool error = !!qdio_error;
 
-	if (queue->bufstates && (queue->bufstates[bidx].flags &
+	if (qdio_error == QDIO_ERROR_SLSB_PENDING) {
-				 QDIO_OUTBUF_STATE_FLAG_PENDING)) {
 		WARN_ON_ONCE(card->options.cq != QETH_CQ_ENABLED);
 
 		QETH_CARD_TEXT_(card, 5, "pel%u", bidx);

fs/statfs.c

@@ -255,7 +255,10 @@ SYSCALL_DEFINE2(ustat, unsigned, dev, struct ustat __user *, ubuf)
 
 	memset(&tmp,0,sizeof(struct ustat));
 	tmp.f_tfree = sbuf.f_bfree;
-	tmp.f_tinode = sbuf.f_ffree;
+	if (IS_ENABLED(CONFIG_ARCH_32BIT_USTAT_F_TINODE))
+		tmp.f_tinode = min_t(u64, sbuf.f_ffree, UINT_MAX);
+	else
+		tmp.f_tinode = sbuf.f_ffree;
 
 	return copy_to_user(ubuf, &tmp, sizeof(struct ustat)) ? -EFAULT : 0;
 }

include/linux/types.h

@@ -14,7 +14,7 @@ typedef u32 __kernel_dev_t;
 
 typedef __kernel_fd_set		fd_set;
 typedef __kernel_dev_t		dev_t;
-typedef __kernel_ino_t		ino_t;
+typedef __kernel_ulong_t	ino_t;
 typedef __kernel_mode_t		mode_t;
 typedef unsigned short		umode_t;
 typedef u32			nlink_t;
@@ -189,7 +189,11 @@ struct hlist_node {
 
 struct ustat {
 	__kernel_daddr_t	f_tfree;
-	__kernel_ino_t		f_tinode;
+#ifdef CONFIG_ARCH_32BIT_USTAT_F_TINODE
+	unsigned int		f_tinode;
+#else
+	unsigned long		f_tinode;
+#endif
 	char			f_fname[6];
 	char			f_fpack[6];
 };