From d552a58d708020963f6973a8b3b690f19ac81c99 Mon Sep 17 00:00:00 2001
From: Thomas Richter <tmricht@linux.ibm.com>
Date: Wed, 28 Apr 2021 13:30:56 +0200
Subject: s390/cpumf: remove counter transaction call backs

The command 'perf stat -e cycles ...' triggers the following function
sequence in the CPU Measurement Facility counter device driver:

perf_pmu_event_init()
  __hw_perf_event_init()
    validate_ctr_auth()
    validate_ctr_version()

During event creation, the counter number is checked in functions
validate_ctr_auth() and validate_ctr_version() to verify it is a valid
counter and supported by the hardware. If this is not the case, both
functions return an error and the event is not created. System call
perf_event_open() returns an error in this case.

Later on the event is installed in the kernel event subsystem and the
driver functions cpumf_pmu_add() and cpumf_pmu_commit_txn() are called
to install the counter event by the hardware.

Since both events have been verified at event creation, there is no need
to re-evaluate the authorization state. This can not change since on
 * LPARs the authorization change requires a restart of the LPAR (and
   thus a reboot of the kernel)
 * DPMs can not take resources away, just add them.

Also the sequence of CPU Measurement facility counter device driver
calls is
  cpumf_pmu_start_txn
  cpumf_pmu_add
  cpumf_pmu_start
  cpumf_pmu_commit_txn
for every single event. Which means the condition in cpumf_pmu_add()
is never met and validate_ctr_auth() is never called.

This leaves the counter device driver transaction functions with
just one task:
start_txn: Verify a transaction is not in flight and call
	perf_pmu_disable()
cancel_txn, commit_txn: Verify a transaction is in flight and call
	perf_pmu_enable()

The same functionality is provided by the default transaction handling
functions in kernel/events/core.c. Use those by removing the
counter device driver private call back functions.

Suggested-by: Sumanth Korikkar <sumanthk@linux.ibm.com>
Signed-off-by: Thomas Richter <tmricht@linux.ibm.com>
Reviewed-by: Sumanth Korikkar <sumanthk@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/perf_cpum_cf.c        | 84 ----------------------------------
 arch/s390/kernel/perf_cpum_cf_common.c |  1 -
 2 files changed, 85 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/perf_cpum_cf.c b/arch/s390/kernel/perf_cpum_cf.c
index 31a605bcbc6e..91ee0f399aae 100644
--- a/arch/s390/kernel/perf_cpum_cf.c
+++ b/arch/s390/kernel/perf_cpum_cf.c
@@ -413,15 +413,6 @@ static int cpumf_pmu_add(struct perf_event *event, int flags)
 {
 	struct cpu_cf_events *cpuhw = this_cpu_ptr(&cpu_cf_events);
 
-	/* Check authorization for the counter set to which this
-	 * counter belongs.
-	 * For group events transaction, the authorization check is
-	 * done in cpumf_pmu_commit_txn().
-	 */
-	if (!(cpuhw->txn_flags & PERF_PMU_TXN_ADD))
-		if (validate_ctr_auth(&event->hw))
-			return -ENOENT;
-
 	ctr_set_enable(&cpuhw->state, event->hw.config_base);
 	event->hw.state = PERF_HES_UPTODATE | PERF_HES_STOPPED;
 
@@ -449,78 +440,6 @@ static void cpumf_pmu_del(struct perf_event *event, int flags)
 		ctr_set_disable(&cpuhw->state, event->hw.config_base);
 }
 
-/*
- * Start group events scheduling transaction.
- * Set flags to perform a single test at commit time.
- *
- * We only support PERF_PMU_TXN_ADD transactions. Save the
- * transaction flags but otherwise ignore non-PERF_PMU_TXN_ADD
- * transactions.
- */
-static void cpumf_pmu_start_txn(struct pmu *pmu, unsigned int txn_flags)
-{
-	struct cpu_cf_events *cpuhw = this_cpu_ptr(&cpu_cf_events);
-
-	WARN_ON_ONCE(cpuhw->txn_flags);		/* txn already in flight */
-
-	cpuhw->txn_flags = txn_flags;
-	if (txn_flags & ~PERF_PMU_TXN_ADD)
-		return;
-
-	perf_pmu_disable(pmu);
-	cpuhw->tx_state = cpuhw->state;
-}
-
-/*
- * Stop and cancel a group events scheduling tranctions.
- * Assumes cpumf_pmu_del() is called for each successful added
- * cpumf_pmu_add() during the transaction.
- */
-static void cpumf_pmu_cancel_txn(struct pmu *pmu)
-{
-	unsigned int txn_flags;
-	struct cpu_cf_events *cpuhw = this_cpu_ptr(&cpu_cf_events);
-
-	WARN_ON_ONCE(!cpuhw->txn_flags);	/* no txn in flight */
-
-	txn_flags = cpuhw->txn_flags;
-	cpuhw->txn_flags = 0;
-	if (txn_flags & ~PERF_PMU_TXN_ADD)
-		return;
-
-	WARN_ON(cpuhw->tx_state != cpuhw->state);
-
-	perf_pmu_enable(pmu);
-}
-
-/*
- * Commit the group events scheduling transaction.  On success, the
- * transaction is closed.   On error, the transaction is kept open
- * until cpumf_pmu_cancel_txn() is called.
- */
-static int cpumf_pmu_commit_txn(struct pmu *pmu)
-{
-	struct cpu_cf_events *cpuhw = this_cpu_ptr(&cpu_cf_events);
-	u64 state;
-
-	WARN_ON_ONCE(!cpuhw->txn_flags);	/* no txn in flight */
-
-	if (cpuhw->txn_flags & ~PERF_PMU_TXN_ADD) {
-		cpuhw->txn_flags = 0;
-		return 0;
-	}
-
-	/* check if the updated state can be scheduled */
-	state = cpuhw->state & ~((1 << CPUMF_LCCTL_ENABLE_SHIFT) - 1);
-	state >>= CPUMF_LCCTL_ENABLE_SHIFT;
-	if ((state & cpuhw->info.auth_ctl) != state)
-		return -ENOENT;
-
-	cpuhw->txn_flags = 0;
-	perf_pmu_enable(pmu);
-	return 0;
-}
-
 /* Performance monitoring unit for s390x */
 static struct pmu cpumf_pmu = {
 	.task_ctx_nr  = perf_sw_context,
@@ -533,9 +452,6 @@ static struct pmu cpumf_pmu = {
 	.start	      = cpumf_pmu_start,
 	.stop	      = cpumf_pmu_stop,
 	.read	      = cpumf_pmu_read,
-	.start_txn    = cpumf_pmu_start_txn,
-	.commit_txn   = cpumf_pmu_commit_txn,
-	.cancel_txn   = cpumf_pmu_cancel_txn,
 };
 
 static int __init cpumf_pmu_init(void)
diff --git a/arch/s390/kernel/perf_cpum_cf_common.c b/arch/s390/kernel/perf_cpum_cf_common.c
index 6d53215c8484..2300fbaac556 100644
--- a/arch/s390/kernel/perf_cpum_cf_common.c
+++ b/arch/s390/kernel/perf_cpum_cf_common.c
@@ -30,7 +30,6 @@ DEFINE_PER_CPU(struct cpu_cf_events, cpu_cf_events) = {
 	.alert = ATOMIC64_INIT(0),
 	.state = 0,
 	.flags = 0,
-	.txn_flags = 0,
 };
 /* Indicator whether the CPU-Measurement Counter Facility Support is ready */
 static bool cpum_cf_initalized;
-- 
cgit v1.2.3


From 15e5b53ff4c92c8ea79094871f91d077bfc60526 Mon Sep 17 00:00:00 2001
From: Thomas Richter <tmricht@linux.ibm.com>
Date: Thu, 29 Apr 2021 15:04:30 +0200
Subject: s390/cpumf: remove WARN_ON_ONCE in counter start handler

Remove some WARN_ON_ONCE() warnings when a counter is started. Each
counter is installed function calls
event_sched_in() --> cpumf_pmu_add(..., PERF_EF_START).

This is done after the event has been created using
perf_pmu_event_init() which verifies the counter is valid.
Member hwc->config must be valid at this point.

Function cpumf_pmu_start(..., PERF_EF_RELOAD) is called from
function cpumf_pmu_add() for counter events. All other invocations of
cpumf_pmu_start(..., PERF_EF_RELOAD) are from the performance subsystem
for sampling events.

Signed-off-by: Thomas Richter <tmricht@linux.ibm.com>
Acked-by: Sumanth Korikkar <sumanthk@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/perf_cpum_cf.c | 8 +-------
 1 file changed, 1 insertion(+), 7 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/perf_cpum_cf.c b/arch/s390/kernel/perf_cpum_cf.c
index 91ee0f399aae..1b7a0525fbed 100644
--- a/arch/s390/kernel/perf_cpum_cf.c
+++ b/arch/s390/kernel/perf_cpum_cf.c
@@ -362,15 +362,9 @@ static void cpumf_pmu_start(struct perf_event *event, int flags)
 	struct cpu_cf_events *cpuhw = this_cpu_ptr(&cpu_cf_events);
 	struct hw_perf_event *hwc = &event->hw;
 
-	if (WARN_ON_ONCE(!(hwc->state & PERF_HES_STOPPED)))
+	if (!(hwc->state & PERF_HES_STOPPED))
 		return;
 
-	if (WARN_ON_ONCE(hwc->config == -1))
-		return;
-
-	if (flags & PERF_EF_RELOAD)
-		WARN_ON_ONCE(!(hwc->state & PERF_HES_UPTODATE));
-
 	hwc->state = 0;
 
 	/* (Re-)enable and activate the counter set */
-- 
cgit v1.2.3


From af9ad82290a58b3f1cc02d12459e3396eee04187 Mon Sep 17 00:00:00 2001
From: Sven Schnelle <svens@linux.ibm.com>
Date: Wed, 5 May 2021 22:01:09 +0200
Subject: s390/entry: use assignment to read intcode / asm to copy gprs

arch/s390/kernel/syscall.c: In function __do_syscall:
arch/s390/kernel/syscall.c:147:9: warning: memcpy reading 64 bytes from a region of size 0 [-Wstringop-overread]
  147 |         memcpy(&regs->gprs[8], S390_lowcore.save_area_sync, 8 * sizeof(unsigned long));
      |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
arch/s390/kernel/syscall.c:148:9: warning: memcpy reading 4 bytes from a region of size 0 [-Wstringop-overread]
  148 |         memcpy(&regs->int_code, &S390_lowcore.svc_ilc, sizeof(regs->int_code));
      |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fix this by moving the gprs restore from C to assembly, and use a assignment
for int_code instead of memcpy.

Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/include/asm/lowcore.h | 3 +--
 arch/s390/kernel/asm-offsets.c  | 2 --
 arch/s390/kernel/entry.S        | 1 +
 arch/s390/kernel/syscall.c      | 5 +----
 4 files changed, 3 insertions(+), 8 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h
index 22bceeeba4bc..484065b5fa7d 100644
--- a/arch/s390/include/asm/lowcore.h
+++ b/arch/s390/include/asm/lowcore.h
@@ -24,8 +24,7 @@ struct lowcore {
 	__u32	ext_params;			/* 0x0080 */
 	__u16	ext_cpu_addr;			/* 0x0084 */
 	__u16	ext_int_code;			/* 0x0086 */
-	__u16	svc_ilc;			/* 0x0088 */
-	__u16	svc_code;			/* 0x008a */
+	__u32	svc_int_code;			/* 0x0088 */
 	__u16	pgm_ilc;			/* 0x008c */
 	__u16	pgm_code;			/* 0x008e */
 	__u32	data_exc_code;			/* 0x0090 */
diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c
index 15e637728a4b..96270584ef69 100644
--- a/arch/s390/kernel/asm-offsets.c
+++ b/arch/s390/kernel/asm-offsets.c
@@ -58,8 +58,6 @@ int main(void)
 	OFFSET(__LC_EXT_PARAMS, lowcore, ext_params);
 	OFFSET(__LC_EXT_CPU_ADDR, lowcore, ext_cpu_addr);
 	OFFSET(__LC_EXT_INT_CODE, lowcore, ext_int_code);
-	OFFSET(__LC_SVC_ILC, lowcore, svc_ilc);
-	OFFSET(__LC_SVC_INT_CODE, lowcore, svc_code);
 	OFFSET(__LC_PGM_ILC, lowcore, pgm_ilc);
 	OFFSET(__LC_PGM_INT_CODE, lowcore, pgm_code);
 	OFFSET(__LC_DATA_EXC_CODE, lowcore, data_exc_code);
diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index 12de7a9c85b3..e7094ab8eb2c 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -276,6 +276,7 @@ ENTRY(system_call)
 	xgr	%r10,%r10
 	xgr	%r11,%r11
 	la	%r2,STACK_FRAME_OVERHEAD(%r15)	# pointer to pt_regs
+	mvc	__PT_R8(64,%r2),__LC_SAVE_AREA_SYNC
 	lgr	%r3,%r14
 	brasl	%r14,__do_syscall
 	lctlg	%c1,%c1,__LC_USER_ASCE
diff --git a/arch/s390/kernel/syscall.c b/arch/s390/kernel/syscall.c
index 4e5cc7d2364e..76f7916cc30f 100644
--- a/arch/s390/kernel/syscall.c
+++ b/arch/s390/kernel/syscall.c
@@ -144,11 +144,8 @@ void noinstr __do_syscall(struct pt_regs *regs, int per_trap)
 {
 	add_random_kstack_offset();
 	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;
-
+	regs->int_code = S390_lowcore.svc_int_code;
 	update_timer_sys();
 
 	local_irq_enable();
-- 
cgit v1.2.3


From 17e89e1340a377b2f14a14d7050f609328592793 Mon Sep 17 00:00:00 2001
From: Sven Schnelle <svens@linux.ibm.com>
Date: Wed, 5 May 2021 22:01:10 +0200
Subject: s390/facilities: move stfl information from lowcore to global data

With gcc-11, there are a lot of warnings because the facility functions
are accessing lowcore through a null pointer. Fix this by moving the
facility arrays away from lowcore.

Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/boot/als.c             |  6 +++---
 arch/s390/boot/ipl_parm.c        |  6 +++---
 arch/s390/boot/startup.c         |  3 +++
 arch/s390/include/asm/facility.h | 13 +++++++++----
 arch/s390/include/asm/lowcore.h  |  7 +------
 arch/s390/kernel/alternative.c   |  3 +--
 arch/s390/kernel/asm-offsets.c   |  2 --
 arch/s390/kernel/early.c         |  6 ++----
 arch/s390/kernel/nospec-branch.c | 14 +++++++-------
 arch/s390/kernel/nospec-sysfs.c  |  2 +-
 arch/s390/kernel/processor.c     |  4 +---
 arch/s390/kernel/setup.c         |  8 +++-----
 arch/s390/kernel/smp.c           |  4 ----
 arch/s390/kvm/kvm-s390.c         | 12 ++++++------
 14 files changed, 40 insertions(+), 50 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/boot/als.c b/arch/s390/boot/als.c
index ff6801d401c4..47c48fbfb563 100644
--- a/arch/s390/boot/als.c
+++ b/arch/s390/boot/als.c
@@ -68,7 +68,7 @@ void print_missing_facilities(void)
 
 	first = 1;
 	for (i = 0; i < ARRAY_SIZE(als); i++) {
-		val = ~S390_lowcore.stfle_fac_list[i] & als[i];
+		val = ~stfle_fac_list[i] & als[i];
 		for (j = 0; j < BITS_PER_LONG; j++) {
 			if (!(val & (1UL << (BITS_PER_LONG - 1 - j))))
 				continue;
@@ -106,9 +106,9 @@ void verify_facilities(void)
 {
 	int i;
 
-	__stfle(S390_lowcore.stfle_fac_list, ARRAY_SIZE(S390_lowcore.stfle_fac_list));
+	__stfle(stfle_fac_list, ARRAY_SIZE(stfle_fac_list));
 	for (i = 0; i < ARRAY_SIZE(als); i++) {
-		if ((S390_lowcore.stfle_fac_list[i] & als[i]) != als[i])
+		if ((stfle_fac_list[i] & als[i]) != als[i])
 			facility_mismatch();
 	}
 }
diff --git a/arch/s390/boot/ipl_parm.c b/arch/s390/boot/ipl_parm.c
index d372a45fe10e..865c66cb9b79 100644
--- a/arch/s390/boot/ipl_parm.c
+++ b/arch/s390/boot/ipl_parm.c
@@ -180,9 +180,9 @@ void setup_boot_command_line(void)
 static void modify_facility(unsigned long nr, bool clear)
 {
 	if (clear)
-		__clear_facility(nr, S390_lowcore.stfle_fac_list);
+		__clear_facility(nr, stfle_fac_list);
 	else
-		__set_facility(nr, S390_lowcore.stfle_fac_list);
+		__set_facility(nr, stfle_fac_list);
 }
 
 static void check_cleared_facilities(void)
@@ -191,7 +191,7 @@ static void check_cleared_facilities(void)
 	int i;
 
 	for (i = 0; i < ARRAY_SIZE(als); i++) {
-		if ((S390_lowcore.stfle_fac_list[i] & als[i]) != als[i]) {
+		if ((stfle_fac_list[i] & als[i]) != als[i]) {
 			sclp_early_printk("Warning: The Linux kernel requires facilities cleared via command line option\n");
 			print_missing_facilities();
 			break;
diff --git a/arch/s390/boot/startup.c b/arch/s390/boot/startup.c
index 05f8eefa3dcf..61a8ac4067e5 100644
--- a/arch/s390/boot/startup.c
+++ b/arch/s390/boot/startup.c
@@ -17,6 +17,9 @@ extern char __boot_data_preserved_start[], __boot_data_preserved_end[];
 unsigned long __bootdata_preserved(__kaslr_offset);
 unsigned long __bootdata(ident_map_size);
 
+u64 __bootdata_preserved(stfle_fac_list[16]);
+u64 __bootdata_preserved(alt_stfle_fac_list[16]);
+
 /*
  * Some code and data needs to stay below 2 GB, even when the kernel would be
  * relocated above 2 GB, because it has to use 31 bit addresses.
diff --git a/arch/s390/include/asm/facility.h b/arch/s390/include/asm/facility.h
index 91b5d714d28f..948e2616fe9c 100644
--- a/arch/s390/include/asm/facility.h
+++ b/arch/s390/include/asm/facility.h
@@ -13,7 +13,10 @@
 #include <linux/preempt.h>
 #include <asm/lowcore.h>
 
-#define MAX_FACILITY_BIT (sizeof(((struct lowcore *)0)->stfle_fac_list) * 8)
+#define MAX_FACILITY_BIT (sizeof(stfle_fac_list) * 8)
+
+extern u64 stfle_fac_list[16];
+extern u64 alt_stfle_fac_list[16];
 
 static inline void __set_facility(unsigned long nr, void *facilities)
 {
@@ -56,7 +59,7 @@ static inline int test_facility(unsigned long nr)
 		if (__test_facility(nr, &facilities_als))
 			return 1;
 	}
-	return __test_facility(nr, &S390_lowcore.stfle_fac_list);
+	return __test_facility(nr, &stfle_fac_list);
 }
 
 static inline unsigned long __stfle_asm(u64 *stfle_fac_list, int size)
@@ -79,13 +82,15 @@ static inline unsigned long __stfle_asm(u64 *stfle_fac_list, int size)
 static inline void __stfle(u64 *stfle_fac_list, int size)
 {
 	unsigned long nr;
+	u32 stfl_fac_list;
 
 	asm volatile(
 		"	stfl	0(0)\n"
 		: "=m" (S390_lowcore.stfl_fac_list));
+	stfl_fac_list = S390_lowcore.stfl_fac_list;
+	memcpy(stfle_fac_list, &stfl_fac_list, 4);
 	nr = 4; /* bytes stored by stfl */
-	memcpy(stfle_fac_list, &S390_lowcore.stfl_fac_list, 4);
-	if (S390_lowcore.stfl_fac_list & 0x01000000) {
+	if (stfl_fac_list & 0x01000000) {
 		/* More facility bits available with stfle */
 		nr = __stfle_asm(stfle_fac_list, size);
 		nr = min_t(unsigned long, (nr + 1) * 8, size * 8);
diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h
index 484065b5fa7d..a6f0b8feb12e 100644
--- a/arch/s390/include/asm/lowcore.h
+++ b/arch/s390/include/asm/lowcore.h
@@ -153,12 +153,7 @@ struct lowcore {
 	__u64	vmcore_info;			/* 0x0e0c */
 	__u8	pad_0x0e14[0x0e18-0x0e14];	/* 0x0e14 */
 	__u64	os_info;			/* 0x0e18 */
-	__u8	pad_0x0e20[0x0f00-0x0e20];	/* 0x0e20 */
-
-	/* Extended facility list */
-	__u64	stfle_fac_list[16];		/* 0x0f00 */
-	__u64	alt_stfle_fac_list[16];		/* 0x0f80 */
-	__u8	pad_0x1000[0x11b0-0x1000];	/* 0x1000 */
+	__u8	pad_0x0e20[0x11b0-0x0e20];	/* 0x0e20 */
 
 	/* Pointer to the machine check extended save area */
 	__u64	mcesad;				/* 0x11b0 */
diff --git a/arch/s390/kernel/alternative.c b/arch/s390/kernel/alternative.c
index 8e1f2aee85ef..c22ea1c3ef84 100644
--- a/arch/s390/kernel/alternative.c
+++ b/arch/s390/kernel/alternative.c
@@ -76,8 +76,7 @@ static void __init_or_module __apply_alternatives(struct alt_instr *start,
 		instr = (u8 *)&a->instr_offset + a->instr_offset;
 		replacement = (u8 *)&a->repl_offset + a->repl_offset;
 
-		if (!__test_facility(a->facility,
-				     S390_lowcore.alt_stfle_fac_list))
+		if (!__test_facility(a->facility, alt_stfle_fac_list))
 			continue;
 
 		if (unlikely(a->instrlen % 2 || a->replacementlen % 2)) {
diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c
index 96270584ef69..b4ca7cc011cb 100644
--- a/arch/s390/kernel/asm-offsets.c
+++ b/arch/s390/kernel/asm-offsets.c
@@ -75,8 +75,6 @@ int main(void)
 	OFFSET(__LC_SUBCHANNEL_NR, lowcore, subchannel_nr);
 	OFFSET(__LC_IO_INT_PARM, lowcore, io_int_parm);
 	OFFSET(__LC_IO_INT_WORD, lowcore, io_int_word);
-	OFFSET(__LC_STFL_FAC_LIST, lowcore, stfl_fac_list);
-	OFFSET(__LC_STFLE_FAC_LIST, lowcore, stfle_fac_list);
 	OFFSET(__LC_MCCK_CODE, lowcore, mcck_interruption_code);
 	OFFSET(__LC_EXT_DAMAGE_CODE, lowcore, external_damage_code);
 	OFFSET(__LC_MCCK_FAIL_STOR_ADDR, lowcore, failing_storage_address);
diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c
index a361d2e70025..c2cf79d353cf 100644
--- a/arch/s390/kernel/early.c
+++ b/arch/s390/kernel/early.c
@@ -180,11 +180,9 @@ static noinline __init void setup_lowcore_early(void)
 
 static noinline __init void setup_facility_list(void)
 {
-	memcpy(S390_lowcore.alt_stfle_fac_list,
-	       S390_lowcore.stfle_fac_list,
-	       sizeof(S390_lowcore.alt_stfle_fac_list));
+	memcpy(alt_stfle_fac_list, stfle_fac_list, sizeof(alt_stfle_fac_list));
 	if (!IS_ENABLED(CONFIG_KERNEL_NOBP))
-		__clear_facility(82, S390_lowcore.alt_stfle_fac_list);
+		__clear_facility(82, alt_stfle_fac_list);
 }
 
 static __init void detect_diag9c(void)
diff --git a/arch/s390/kernel/nospec-branch.c b/arch/s390/kernel/nospec-branch.c
index 29e511f5bf06..2c5c3756644b 100644
--- a/arch/s390/kernel/nospec-branch.c
+++ b/arch/s390/kernel/nospec-branch.c
@@ -17,11 +17,11 @@ static int __init nobp_setup_early(char *str)
 		 * The user explicitely requested nobp=1, enable it and
 		 * disable the expoline support.
 		 */
-		__set_facility(82, S390_lowcore.alt_stfle_fac_list);
+		__set_facility(82, alt_stfle_fac_list);
 		if (IS_ENABLED(CONFIG_EXPOLINE))
 			nospec_disable = 1;
 	} else {
-		__clear_facility(82, S390_lowcore.alt_stfle_fac_list);
+		__clear_facility(82, alt_stfle_fac_list);
 	}
 	return 0;
 }
@@ -29,7 +29,7 @@ early_param("nobp", nobp_setup_early);
 
 static int __init nospec_setup_early(char *str)
 {
-	__clear_facility(82, S390_lowcore.alt_stfle_fac_list);
+	__clear_facility(82, alt_stfle_fac_list);
 	return 0;
 }
 early_param("nospec", nospec_setup_early);
@@ -40,7 +40,7 @@ static int __init nospec_report(void)
 		pr_info("Spectre V2 mitigation: etokens\n");
 	if (__is_defined(CC_USING_EXPOLINE) && !nospec_disable)
 		pr_info("Spectre V2 mitigation: execute trampolines\n");
-	if (__test_facility(82, S390_lowcore.alt_stfle_fac_list))
+	if (__test_facility(82, alt_stfle_fac_list))
 		pr_info("Spectre V2 mitigation: limited branch prediction\n");
 	return 0;
 }
@@ -66,14 +66,14 @@ void __init nospec_auto_detect(void)
 		 */
 		if (__is_defined(CC_USING_EXPOLINE))
 			nospec_disable = 1;
-		__clear_facility(82, S390_lowcore.alt_stfle_fac_list);
+		__clear_facility(82, alt_stfle_fac_list);
 	} else if (__is_defined(CC_USING_EXPOLINE)) {
 		/*
 		 * The kernel has been compiled with expolines.
 		 * Keep expolines enabled and disable nobp.
 		 */
 		nospec_disable = 0;
-		__clear_facility(82, S390_lowcore.alt_stfle_fac_list);
+		__clear_facility(82, alt_stfle_fac_list);
 	}
 	/*
 	 * If the kernel has not been compiled with expolines the
@@ -86,7 +86,7 @@ static int __init spectre_v2_setup_early(char *str)
 {
 	if (str && !strncmp(str, "on", 2)) {
 		nospec_disable = 0;
-		__clear_facility(82, S390_lowcore.alt_stfle_fac_list);
+		__clear_facility(82, alt_stfle_fac_list);
 	}
 	if (str && !strncmp(str, "off", 3))
 		nospec_disable = 1;
diff --git a/arch/s390/kernel/nospec-sysfs.c b/arch/s390/kernel/nospec-sysfs.c
index 48f472bf9290..b4b5c8c21166 100644
--- a/arch/s390/kernel/nospec-sysfs.c
+++ b/arch/s390/kernel/nospec-sysfs.c
@@ -17,7 +17,7 @@ ssize_t cpu_show_spectre_v2(struct device *dev,
 		return sprintf(buf, "Mitigation: etokens\n");
 	if (__is_defined(CC_USING_EXPOLINE) && !nospec_disable)
 		return sprintf(buf, "Mitigation: execute trampolines\n");
-	if (__test_facility(82, S390_lowcore.alt_stfle_fac_list))
+	if (__test_facility(82, alt_stfle_fac_list))
 		return sprintf(buf, "Mitigation: limited branch prediction\n");
 	return sprintf(buf, "Vulnerable\n");
 }
diff --git a/arch/s390/kernel/processor.c b/arch/s390/kernel/processor.c
index c92d04f876cb..82df39b17bb5 100644
--- a/arch/s390/kernel/processor.c
+++ b/arch/s390/kernel/processor.c
@@ -103,11 +103,9 @@ EXPORT_SYMBOL(cpu_have_feature);
 static void show_facilities(struct seq_file *m)
 {
 	unsigned int bit;
-	long *facilities;
 
-	facilities = (long *)&S390_lowcore.stfle_fac_list;
 	seq_puts(m, "facilities      :");
-	for_each_set_bit_inv(bit, facilities, MAX_FACILITY_BIT)
+	for_each_set_bit_inv(bit, (long *)&stfle_fac_list, MAX_FACILITY_BIT)
 		seq_printf(m, " %d", bit);
 	seq_putc(m, '\n');
 }
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index 5aab59ad5688..59c21df9ea79 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -108,6 +108,9 @@ unsigned long __bootdata_preserved(__edma);
 unsigned long __bootdata_preserved(__kaslr_offset);
 unsigned int __bootdata_preserved(zlib_dfltcc_support);
 EXPORT_SYMBOL(zlib_dfltcc_support);
+u64 __bootdata_preserved(stfle_fac_list[16]);
+EXPORT_SYMBOL(stfle_fac_list);
+u64 __bootdata_preserved(alt_stfle_fac_list[16]);
 
 unsigned long VMALLOC_START;
 EXPORT_SYMBOL(VMALLOC_START);
@@ -413,11 +416,6 @@ static void __init setup_lowcore_dat_off(void)
 	lc->lpp = LPP_MAGIC;
 	lc->machine_flags = S390_lowcore.machine_flags;
 	lc->preempt_count = S390_lowcore.preempt_count;
-	lc->stfl_fac_list = S390_lowcore.stfl_fac_list;
-	memcpy(lc->stfle_fac_list, S390_lowcore.stfle_fac_list,
-	       sizeof(lc->stfle_fac_list));
-	memcpy(lc->alt_stfle_fac_list, S390_lowcore.alt_stfle_fac_list,
-	       sizeof(lc->alt_stfle_fac_list));
 	nmi_alloc_boot_cpu(lc);
 	lc->sys_enter_timer = S390_lowcore.sys_enter_timer;
 	lc->exit_timer = S390_lowcore.exit_timer;
diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 2fec2b80d35d..4b439967ca5d 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -275,10 +275,6 @@ static void pcpu_prepare_secondary(struct pcpu *pcpu, int cpu)
 	lc->cregs_save_area[1] = lc->kernel_asce;
 	lc->cregs_save_area[7] = lc->user_asce;
 	save_access_regs((unsigned int *) lc->access_regs_save_area);
-	memcpy(lc->stfle_fac_list, S390_lowcore.stfle_fac_list,
-	       sizeof(lc->stfle_fac_list));
-	memcpy(lc->alt_stfle_fac_list, S390_lowcore.alt_stfle_fac_list,
-	       sizeof(lc->alt_stfle_fac_list));
 	arch_spin_lock_setup(cpu);
 }
 
diff --git a/arch/s390/kvm/kvm-s390.c b/arch/s390/kvm/kvm-s390.c
index 1296fc10f80c..582e947675e6 100644
--- a/arch/s390/kvm/kvm-s390.c
+++ b/arch/s390/kvm/kvm-s390.c
@@ -214,7 +214,7 @@ static unsigned long kvm_s390_fac_size(void)
 	BUILD_BUG_ON(SIZE_INTERNAL > S390_ARCH_FAC_MASK_SIZE_U64);
 	BUILD_BUG_ON(SIZE_INTERNAL > S390_ARCH_FAC_LIST_SIZE_U64);
 	BUILD_BUG_ON(SIZE_INTERNAL * sizeof(unsigned long) >
-		sizeof(S390_lowcore.stfle_fac_list));
+		sizeof(stfle_fac_list));
 
 	return SIZE_INTERNAL;
 }
@@ -1458,8 +1458,8 @@ static int kvm_s390_get_machine(struct kvm *kvm, struct kvm_device_attr *attr)
 	mach->ibc = sclp.ibc;
 	memcpy(&mach->fac_mask, kvm->arch.model.fac_mask,
 	       S390_ARCH_FAC_LIST_SIZE_BYTE);
-	memcpy((unsigned long *)&mach->fac_list, S390_lowcore.stfle_fac_list,
-	       sizeof(S390_lowcore.stfle_fac_list));
+	memcpy((unsigned long *)&mach->fac_list, stfle_fac_list,
+	       sizeof(stfle_fac_list));
 	VM_EVENT(kvm, 3, "GET: host ibc:  0x%4.4x, host cpuid:  0x%16.16llx",
 		 kvm->arch.model.ibc,
 		 kvm->arch.model.cpuid);
@@ -2683,10 +2683,10 @@ int kvm_arch_init_vm(struct kvm *kvm, unsigned long type)
 	kvm->arch.model.fac_list = kvm->arch.sie_page2->fac_list;
 
 	for (i = 0; i < kvm_s390_fac_size(); i++) {
-		kvm->arch.model.fac_mask[i] = S390_lowcore.stfle_fac_list[i] &
+		kvm->arch.model.fac_mask[i] = stfle_fac_list[i] &
 					      (kvm_s390_fac_base[i] |
 					       kvm_s390_fac_ext[i]);
-		kvm->arch.model.fac_list[i] = S390_lowcore.stfle_fac_list[i] &
+		kvm->arch.model.fac_list[i] = stfle_fac_list[i] &
 					      kvm_s390_fac_base[i];
 	}
 	kvm->arch.model.subfuncs = kvm_s390_available_subfunc;
@@ -5055,7 +5055,7 @@ static int __init kvm_s390_init(void)
 
 	for (i = 0; i < 16; i++)
 		kvm_s390_fac_base[i] |=
-			S390_lowcore.stfle_fac_list[i] & nonhyp_mask(i);
+			stfle_fac_list[i] & nonhyp_mask(i);
 
 	return kvm_init(NULL, sizeof(struct kvm_vcpu), 0, THIS_MODULE);
 }
-- 
cgit v1.2.3


From 6c6a07fc7c98f31fffb0a00d2d7a5344df8b4ba8 Mon Sep 17 00:00:00 2001
From: Sven Schnelle <svens@linux.ibm.com>
Date: Wed, 5 May 2021 22:01:11 +0200
Subject: s390/irq: add union/struct to access external interrupt parameters

gcc-11 warns:

arch/s390/kernel/irq.c: In function do_ext_irq:
arch/s390/kernel/irq.c:175:9: warning: memcpy reading 4 bytes from a region of size 0 [-Wstringop-overread]
  175 |         memcpy(&regs->int_code, &S390_lowcore.ext_cpu_addr, 4);
      |         ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fix this by adding a struct for int_code to struct lowcore.

Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/include/asm/lowcore.h | 9 +++++++--
 arch/s390/kernel/irq.c          | 2 +-
 2 files changed, 8 insertions(+), 3 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h
index a6f0b8feb12e..d08813c7f462 100644
--- a/arch/s390/include/asm/lowcore.h
+++ b/arch/s390/include/asm/lowcore.h
@@ -22,8 +22,13 @@ struct lowcore {
 	__u32	ipl_parmblock_ptr;		/* 0x0014 */
 	__u8	pad_0x0018[0x0080-0x0018];	/* 0x0018 */
 	__u32	ext_params;			/* 0x0080 */
-	__u16	ext_cpu_addr;			/* 0x0084 */
-	__u16	ext_int_code;			/* 0x0086 */
+	union {
+		struct {
+			__u16 ext_cpu_addr;	/* 0x0084 */
+			__u16 ext_int_code;	/* 0x0086 */
+		} __packed;
+		__u32 ext_int_code_addr;
+	} __packed;
 	__u32	svc_int_code;			/* 0x0088 */
 	__u16	pgm_ilc;			/* 0x008c */
 	__u16	pgm_code;			/* 0x008e */
diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c
index 714269e10eec..169258012530 100644
--- a/arch/s390/kernel/irq.c
+++ b/arch/s390/kernel/irq.c
@@ -172,7 +172,7 @@ void noinstr do_ext_irq(struct pt_regs *regs)
 	if (user_mode(regs))
 		update_timer_sys();
 
-	memcpy(&regs->int_code, &S390_lowcore.ext_cpu_addr, 4);
+	regs->int_code = S390_lowcore.ext_int_code_addr;
 	regs->int_parm = S390_lowcore.ext_params;
 	regs->int_parm_long = S390_lowcore.ext_params2;
 
-- 
cgit v1.2.3


From 755112b35cdd181dffd704b42ca6a788cc0e3326 Mon Sep 17 00:00:00 2001
From: Sven Schnelle <svens@linux.ibm.com>
Date: Wed, 5 May 2021 22:01:16 +0200
Subject: s390/traps: add struct to access transactional diagnostic block

gcc-11 warns:

arch/s390/kernel/traps.c: In function __do_pgm_check:
arch/s390/kernel/traps.c:319:17: warning: memcpy reading 256 bytes from a region of size 0 [-Wstringop-overread]
  319 |                 memcpy(&current->thread.trap_tdb, &S390_lowcore.pgm_tdb, 256);
      |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Fix this by adding a struct pgm_tdb to struct lowcore and copy that.

Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/include/asm/lowcore.h   | 6 +++++-
 arch/s390/include/asm/processor.h | 2 +-
 arch/s390/kernel/ptrace.c         | 4 +++-
 arch/s390/kernel/traps.c          | 4 ++--
 4 files changed, 11 insertions(+), 5 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h
index d08813c7f462..3c85c60f56ac 100644
--- a/arch/s390/include/asm/lowcore.h
+++ b/arch/s390/include/asm/lowcore.h
@@ -17,6 +17,10 @@
 #define LC_ORDER 1
 #define LC_PAGES 2
 
+struct pgm_tdb {
+	u64 data[32];
+};
+
 struct lowcore {
 	__u8	pad_0x0000[0x0014-0x0000];	/* 0x0000 */
 	__u32	ipl_parmblock_ptr;		/* 0x0014 */
@@ -184,7 +188,7 @@ struct lowcore {
 	__u8	pad_0x1400[0x1800-0x1400];	/* 0x1400 */
 
 	/* Transaction abort diagnostic block */
-	__u8	pgm_tdb[256];			/* 0x1800 */
+	struct pgm_tdb pgm_tdb;			/* 0x1800 */
 	__u8	pad_0x1900[0x2000-0x1900];	/* 0x1900 */
 } __packed __aligned(8192);
 
diff --git a/arch/s390/include/asm/processor.h b/arch/s390/include/asm/processor.h
index 023a15dc25a3..6d3055f7329a 100644
--- a/arch/s390/include/asm/processor.h
+++ b/arch/s390/include/asm/processor.h
@@ -129,7 +129,7 @@ struct thread_struct {
 	struct runtime_instr_cb *ri_cb;
 	struct gs_cb *gs_cb;			/* Current guarded storage cb */
 	struct gs_cb *gs_bc_cb;			/* Broadcast guarded storage cb */
-	unsigned char trap_tdb[256];		/* Transaction abort diagnose block */
+	struct pgm_tdb trap_tdb;		/* Transaction abort diagnose block */
 	/*
 	 * Warning: 'fpu' is dynamically-sized. It *MUST* be at
 	 * the end.
diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c
index 18b3416fd663..0ea3d02b378d 100644
--- a/arch/s390/kernel/ptrace.c
+++ b/arch/s390/kernel/ptrace.c
@@ -975,10 +975,12 @@ static int s390_tdb_get(struct task_struct *target,
 			struct membuf to)
 {
 	struct pt_regs *regs = task_pt_regs(target);
+	size_t size;
 
 	if (!(regs->int_code & 0x200))
 		return -ENODATA;
-	return membuf_write(&to, target->thread.trap_tdb, 256);
+	size = sizeof(target->thread.trap_tdb.data);
+	return membuf_write(&to, target->thread.trap_tdb.data, size);
 }
 
 static int s390_tdb_set(struct task_struct *target,
diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c
index 8dd23c703718..019c5748b607 100644
--- a/arch/s390/kernel/traps.c
+++ b/arch/s390/kernel/traps.c
@@ -36,7 +36,7 @@ static inline void __user *get_trap_ip(struct pt_regs *regs)
 	unsigned long address;
 
 	if (regs->int_code & 0x200)
-		address = *(unsigned long *)(current->thread.trap_tdb + 24);
+		address = current->thread.trap_tdb.data[3];
 	else
 		address = regs->psw.addr;
 	return (void __user *) (address - (regs->int_code >> 16));
@@ -318,7 +318,7 @@ void noinstr __do_pgm_check(struct pt_regs *regs)
 
 	if (S390_lowcore.pgm_code & 0x0200) {
 		/* transaction abort */
-		memcpy(&current->thread.trap_tdb, &S390_lowcore.pgm_tdb, 256);
+		current->thread.trap_tdb = S390_lowcore.pgm_tdb;
 	}
 
 	if (S390_lowcore.pgm_code & PGM_INT_CODE_PER) {
-- 
cgit v1.2.3


From b7d91d230a119fdcc334d10c9889ce9c5e15118b Mon Sep 17 00:00:00 2001
From: Valentin Vidic <vvidic@valentin-vidic.from.hr>
Date: Tue, 27 Apr 2021 21:40:10 +0200
Subject: s390/sclp_vt220: fix console name to match device

Console name reported in /proc/consoles:

  ttyS1                -W- (EC p  )    4:65

does not match the char device name:

  crw--w----    1 root     root        4,  65 May 17 12:18 /dev/ttysclp0

so debian-installer inside a QEMU s390x instance gets confused and fails
to start with the following error:

  steal-ctty: No such file or directory

Signed-off-by: Valentin Vidic <vvidic@valentin-vidic.from.hr>
Link: https://lore.kernel.org/r/20210427194010.9330-1-vvidic@valentin-vidic.from.hr
Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/setup.c       | 2 +-
 drivers/s390/char/sclp_vt220.c | 4 ++--
 2 files changed, 3 insertions(+), 3 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index 59c21df9ea79..f069083a7517 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -168,7 +168,7 @@ static void __init set_preferred_console(void)
 	else if (CONSOLE_IS_3270)
 		add_preferred_console("tty3270", 0, NULL);
 	else if (CONSOLE_IS_VT220)
-		add_preferred_console("ttyS", 1, NULL);
+		add_preferred_console("ttysclp", 0, NULL);
 	else if (CONSOLE_IS_HVC)
 		add_preferred_console("hvc", 0, NULL);
 }
diff --git a/drivers/s390/char/sclp_vt220.c b/drivers/s390/char/sclp_vt220.c
index 7f4445b0f819..2f96c31e9b7b 100644
--- a/drivers/s390/char/sclp_vt220.c
+++ b/drivers/s390/char/sclp_vt220.c
@@ -35,8 +35,8 @@
 #define SCLP_VT220_MINOR		65
 #define SCLP_VT220_DRIVER_NAME		"sclp_vt220"
 #define SCLP_VT220_DEVICE_NAME		"ttysclp"
-#define SCLP_VT220_CONSOLE_NAME		"ttyS"
-#define SCLP_VT220_CONSOLE_INDEX	1	/* console=ttyS1 */
+#define SCLP_VT220_CONSOLE_NAME		"ttysclp"
+#define SCLP_VT220_CONSOLE_INDEX	0	/* console=ttysclp0 */
 
 /* Representation of a single write request */
 struct sclp_vt220_request {
-- 
cgit v1.2.3


From f73c632d387a5f1528cca6032c646489610bec13 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <hca@linux.ibm.com>
Date: Fri, 7 May 2021 19:12:39 +0200
Subject: s390/ipl: make parameter area accessible via struct parmarea

Since commit 9a965ea95135 ("s390/kexec_file: Simplify parmarea
access") we have struct parmarea which describes the layout of the
kernel parameter area.

Make the kernel parameter area available as global variable parmarea
of type struct parmarea, which allows to easily access its members.

Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/boot/head.S          | 3 +++
 arch/s390/include/asm/setup.h  | 2 ++
 arch/s390/kernel/asm-offsets.c | 3 +++
 3 files changed, 8 insertions(+)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/boot/head.S b/arch/s390/boot/head.S
index dacb7813f982..51693cfb65c2 100644
--- a/arch/s390/boot/head.S
+++ b/arch/s390/boot/head.S
@@ -401,6 +401,7 @@ SYM_CODE_END(startup_pgm_check_handler)
 # Must be keept in sync with struct parmarea in setup.h
 #
 	.org	PARMAREA
+SYM_DATA_START(parmarea)
 	.quad	0			# IPL_DEVICE
 	.quad	0			# INITRD_START
 	.quad	0			# INITRD_SIZE
@@ -411,6 +412,8 @@ SYM_CODE_END(startup_pgm_check_handler)
 	.org	COMMAND_LINE
 	.byte	"root=/dev/ram0 ro"
 	.byte	0
+	.org	PARMAREA+__PARMAREA_SIZE
+SYM_DATA_END(parmarea)
 
 	.org	EARLY_SCCB_OFFSET
 	.fill	4096
diff --git a/arch/s390/include/asm/setup.h b/arch/s390/include/asm/setup.h
index 3e388fa208d4..29baab03f091 100644
--- a/arch/s390/include/asm/setup.h
+++ b/arch/s390/include/asm/setup.h
@@ -78,6 +78,8 @@ struct parmarea {
 	char command_line[ARCH_COMMAND_LINE_SIZE];	/* 0x10480 */
 };
 
+extern struct parmarea parmarea;
+
 extern unsigned int zlib_dfltcc_support;
 #define ZLIB_DFLTCC_DISABLED		0
 #define ZLIB_DFLTCC_FULL		1
diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c
index b4ca7cc011cb..f53605a3dfcd 100644
--- a/arch/s390/kernel/asm-offsets.c
+++ b/arch/s390/kernel/asm-offsets.c
@@ -15,6 +15,7 @@
 #include <asm/idle.h>
 #include <asm/gmap.h>
 #include <asm/nmi.h>
+#include <asm/setup.h>
 #include <asm/stacktrace.h>
 
 int main(void)
@@ -155,5 +156,7 @@ int main(void)
 	OFFSET(__KEXEC_SHA_REGION_START, kexec_sha_region, start);
 	OFFSET(__KEXEC_SHA_REGION_LEN, kexec_sha_region, len);
 	DEFINE(__KEXEC_SHA_REGION_SIZE, sizeof(struct kexec_sha_region));
+	/* sizeof kernel parameter area */
+	DEFINE(__PARMAREA_SIZE, sizeof(struct parmarea));
 	return 0;
 }
-- 
cgit v1.2.3


From 5789284710aa121416524acc54ac0d8c27c3c2ef Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@linux.ibm.com>
Date: Thu, 6 May 2021 16:26:52 +0200
Subject: s390/smp: reallocate IPL CPU lowcore

The lowcore for IPL CPU is special. It is allocated early
in the boot process using memblock and never freed since.
The reason is pcpu_alloc_lowcore() and pcpu_free_lowcore()
routines use page allocator which is not available when
the IPL CPU is getting initialized.

Similar problem is already addressed for stacks - once the
virtual memory is available the early boot stacks get re-
allocated. Doing the same for lowcore will allow freeing
the IPL CPU lowcore and make no difference between the
boot and secondary CPUs.

Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/setup.c | 21 -------------------
 arch/s390/kernel/smp.c   | 52 ++++++++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 52 insertions(+), 21 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index f069083a7517..146d01700a55 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -341,27 +341,6 @@ int __init arch_early_irq_init(void)
 	return 0;
 }
 
-static int __init stack_realloc(void)
-{
-	unsigned long old, new;
-
-	old = S390_lowcore.async_stack - STACK_INIT_OFFSET;
-	new = stack_alloc();
-	if (!new)
-		panic("Couldn't allocate async stack");
-	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_late(old, THREAD_SIZE);
-	return 0;
-}
-early_initcall(stack_realloc);
-
 void __init arch_call_rest_init(void)
 {
 	unsigned long stack;
diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 4b439967ca5d..1e1d61a4d294 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -1234,3 +1234,55 @@ out:
 	return rc;
 }
 subsys_initcall(s390_smp_init);
+
+static __always_inline void set_new_lowcore(struct lowcore *lc)
+{
+	struct lowcore *old_lc = &S390_lowcore;
+	struct lowcore *new_lc = lc;
+	u32 pfx;
+	register struct lowcore *reg2 asm ("2") = new_lc;
+	register unsigned long	 reg3 asm ("3") = sizeof(*reg2);
+	register struct lowcore *reg4 asm ("4") = old_lc;
+	register unsigned long	 reg5 asm ("5") = sizeof(*reg4);
+
+	asm volatile(
+		"	st	2,%[pfx]\n"
+		"	mvcl	2,4\n"
+		"	spx	%[pfx]\n"
+		: "+&d" (reg2), "+&d" (reg3),
+		  "+&d" (reg4), "+&d" (reg5), [pfx] "=Q" (pfx)
+		: : "memory", "cc");
+}
+
+static int __init smp_reinit_ipl_cpu(void)
+{
+	unsigned long async_stack, nodat_stack, mcck_stack;
+	struct lowcore *lc, *lc_ipl;
+	unsigned long flags;
+
+	lc_ipl = lowcore_ptr[0];
+	lc = (struct lowcore *)	__get_free_pages(GFP_KERNEL | GFP_DMA, LC_ORDER);
+	nodat_stack = __get_free_pages(GFP_KERNEL, THREAD_SIZE_ORDER);
+	async_stack = stack_alloc();
+	mcck_stack = stack_alloc();
+	if (!lc || !nodat_stack || !async_stack || !mcck_stack)
+		panic("Couldn't allocate memory");
+
+	local_irq_save(flags);
+	local_mcck_disable();
+	set_new_lowcore(lc);
+	S390_lowcore.nodat_stack = nodat_stack + STACK_INIT_OFFSET;
+	S390_lowcore.async_stack = async_stack + STACK_INIT_OFFSET;
+	S390_lowcore.mcck_stack = mcck_stack + STACK_INIT_OFFSET;
+	lowcore_ptr[0] = lc;
+	pcpu_devices[0].lowcore = lc;
+	local_mcck_enable();
+	local_irq_restore(flags);
+
+	free_pages(lc_ipl->async_stack - STACK_INIT_OFFSET, THREAD_SIZE_ORDER);
+	memblock_free_late(lc_ipl->mcck_stack - STACK_INIT_OFFSET, THREAD_SIZE);
+	memblock_free_late((unsigned long) lc_ipl, sizeof(*lc_ipl));
+
+	return 0;
+}
+early_initcall(smp_reinit_ipl_cpu);
-- 
cgit v1.2.3


From 587704efb3dea5685a2f571b75bd3dc47f73fec1 Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@linux.ibm.com>
Date: Thu, 6 May 2021 16:26:53 +0200
Subject: s390/smp: do not preserve boot CPU lowcore on hotplug

Once the kernel is running the boot CPU lowcore becomes
freeable and does not differ from the secondary CPU ones
in any way. Make use of it and do not preserve the boot
CPU lowcore on unplugging. That allows returning unused
memory when the boot CPU is offline and makes the code
more clear.

Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/smp.c | 47 ++++++++++++++++++-----------------------------
 1 file changed, 18 insertions(+), 29 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 1e1d61a4d294..eaca08b1688f 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -194,20 +194,12 @@ static int pcpu_alloc_lowcore(struct pcpu *pcpu, int cpu)
 	unsigned long async_stack, nodat_stack, mcck_stack;
 	struct lowcore *lc;
 
-	if (pcpu != &pcpu_devices[0]) {
-		pcpu->lowcore =	(struct lowcore *)
-			__get_free_pages(GFP_KERNEL | GFP_DMA, LC_ORDER);
-		nodat_stack = __get_free_pages(GFP_KERNEL, THREAD_SIZE_ORDER);
-		if (!pcpu->lowcore || !nodat_stack)
-			goto out;
-	} else {
-		nodat_stack = pcpu->lowcore->nodat_stack - STACK_INIT_OFFSET;
-	}
+	lc = (struct lowcore *) __get_free_pages(GFP_KERNEL | GFP_DMA, LC_ORDER);
+	nodat_stack = __get_free_pages(GFP_KERNEL, THREAD_SIZE_ORDER);
 	async_stack = stack_alloc();
 	mcck_stack = stack_alloc();
-	if (!async_stack || !mcck_stack)
-		goto out_stack;
-	lc = pcpu->lowcore;
+	if (!lc || !nodat_stack || !async_stack || !mcck_stack)
+		goto out;
 	memcpy(lc, &S390_lowcore, 512);
 	memset((char *) lc + 512, 0, sizeof(*lc) - 512);
 	lc->async_stack = async_stack + STACK_INIT_OFFSET;
@@ -220,40 +212,37 @@ 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_stack;
+		goto out;
 	lowcore_ptr[cpu] = lc;
+	pcpu->lowcore = lc;
 	pcpu_sigp_retry(pcpu, SIGP_SET_PREFIX, (u32)(unsigned long) lc);
 	return 0;
 
-out_stack:
+out:
 	stack_free(mcck_stack);
 	stack_free(async_stack);
-out:
-	if (pcpu != &pcpu_devices[0]) {
-		free_pages(nodat_stack, THREAD_SIZE_ORDER);
-		free_pages((unsigned long) pcpu->lowcore, LC_ORDER);
-	}
+	free_pages(nodat_stack, THREAD_SIZE_ORDER);
+	free_pages((unsigned long) lc, LC_ORDER);
 	return -ENOMEM;
 }
 
 static void pcpu_free_lowcore(struct pcpu *pcpu)
 {
-	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;
+	unsigned long async_stack, nodat_stack, mcck_stack;
+	struct lowcore *lc;
 
+	lc = pcpu->lowcore;
+	nodat_stack = lc->nodat_stack - STACK_INIT_OFFSET;
+	async_stack = lc->async_stack - STACK_INIT_OFFSET;
+	mcck_stack = lc->mcck_stack - STACK_INIT_OFFSET;
 	pcpu_sigp_retry(pcpu, SIGP_SET_PREFIX, 0);
 	lowcore_ptr[pcpu - pcpu_devices] = NULL;
-	nmi_free_per_cpu(pcpu->lowcore);
+	pcpu->lowcore = NULL;
+	nmi_free_per_cpu(lc);
 	stack_free(async_stack);
 	stack_free(mcck_stack);
-	if (pcpu == &pcpu_devices[0])
-		return;
 	free_pages(nodat_stack, THREAD_SIZE_ORDER);
-	free_pages(lowcore, LC_ORDER);
+	free_pages((unsigned long) lc, LC_ORDER);
 }
 
 static void pcpu_prepare_secondary(struct pcpu *pcpu, int cpu)
-- 
cgit v1.2.3


From d2e834c62d7fa467a534758f7994e981ce163c80 Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@linux.ibm.com>
Date: Thu, 6 May 2021 16:26:54 +0200
Subject: s390/smp: remove redundant pcpu::lowcore member

Per-CPU pointer to lowcore is stored in global lowcore_ptr[]
array and duplicated in struct pcpu::lowcore member. This
update removes the redundancy.

Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/smp.c | 36 +++++++++++++++++++++---------------
 1 file changed, 21 insertions(+), 15 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index eaca08b1688f..e137c840a4d3 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -74,7 +74,6 @@ enum {
 static DEFINE_PER_CPU(struct cpu *, cpu_device);
 
 struct pcpu {
-	struct lowcore *lowcore;	/* lowcore page(s) for the cpu */
 	unsigned long ec_mask;		/* bit mask for ec_xxx functions */
 	unsigned long ec_clk;		/* sigp timestamp for ec_xxx */
 	signed char state;		/* physical cpu state */
@@ -214,7 +213,6 @@ static int pcpu_alloc_lowcore(struct pcpu *pcpu, int cpu)
 	if (nmi_alloc_per_cpu(lc))
 		goto out;
 	lowcore_ptr[cpu] = lc;
-	pcpu->lowcore = lc;
 	pcpu_sigp_retry(pcpu, SIGP_SET_PREFIX, (u32)(unsigned long) lc);
 	return 0;
 
@@ -230,14 +228,15 @@ static void pcpu_free_lowcore(struct pcpu *pcpu)
 {
 	unsigned long async_stack, nodat_stack, mcck_stack;
 	struct lowcore *lc;
+	int cpu;
 
-	lc = pcpu->lowcore;
+	cpu = pcpu - pcpu_devices;
+	lc = lowcore_ptr[cpu];
 	nodat_stack = lc->nodat_stack - STACK_INIT_OFFSET;
 	async_stack = lc->async_stack - STACK_INIT_OFFSET;
 	mcck_stack = lc->mcck_stack - STACK_INIT_OFFSET;
 	pcpu_sigp_retry(pcpu, SIGP_SET_PREFIX, 0);
-	lowcore_ptr[pcpu - pcpu_devices] = NULL;
-	pcpu->lowcore = NULL;
+	lowcore_ptr[cpu] = NULL;
 	nmi_free_per_cpu(lc);
 	stack_free(async_stack);
 	stack_free(mcck_stack);
@@ -247,7 +246,7 @@ static void pcpu_free_lowcore(struct pcpu *pcpu)
 
 static void pcpu_prepare_secondary(struct pcpu *pcpu, int cpu)
 {
-	struct lowcore *lc = pcpu->lowcore;
+	struct lowcore *lc = lowcore_ptr[cpu];
 
 	cpumask_set_cpu(cpu, &init_mm.context.cpu_attach_mask);
 	cpumask_set_cpu(cpu, mm_cpumask(&init_mm));
@@ -269,8 +268,11 @@ static void pcpu_prepare_secondary(struct pcpu *pcpu, int cpu)
 
 static void pcpu_attach_task(struct pcpu *pcpu, struct task_struct *tsk)
 {
-	struct lowcore *lc = pcpu->lowcore;
+	struct lowcore *lc;
+	int cpu;
 
+	cpu = pcpu - pcpu_devices;
+	lc = lowcore_ptr[cpu];
 	lc->kernel_stack = (unsigned long) task_stack_page(tsk)
 		+ THREAD_SIZE - STACK_FRAME_OVERHEAD - sizeof(struct pt_regs);
 	lc->current_task = (unsigned long) tsk;
@@ -286,8 +288,11 @@ static void pcpu_attach_task(struct pcpu *pcpu, struct task_struct *tsk)
 
 static void pcpu_start_fn(struct pcpu *pcpu, void (*func)(void *), void *data)
 {
-	struct lowcore *lc = pcpu->lowcore;
+	struct lowcore *lc;
+	int cpu;
 
+	cpu = pcpu - pcpu_devices;
+	lc = lowcore_ptr[cpu];
 	lc->restart_stack = lc->nodat_stack;
 	lc->restart_fn = (unsigned long) func;
 	lc->restart_data = (unsigned long) data;
@@ -372,7 +377,7 @@ void smp_call_online_cpu(void (*func)(void *), void *data)
  */
 void smp_call_ipl_cpu(void (*func)(void *), void *data)
 {
-	struct lowcore *lc = pcpu_devices->lowcore;
+	struct lowcore *lc = lowcore_ptr[0];
 
 	if (pcpu_devices[0].address == stap())
 		lc = &S390_lowcore;
@@ -585,18 +590,21 @@ EXPORT_SYMBOL(smp_ctl_clear_bit);
 
 int smp_store_status(int cpu)
 {
-	struct pcpu *pcpu = pcpu_devices + cpu;
+	struct lowcore *lc;
+	struct pcpu *pcpu;
 	unsigned long pa;
 
-	pa = __pa(&pcpu->lowcore->floating_pt_save_area);
+	pcpu = pcpu_devices + cpu;
+	lc = lowcore_ptr[cpu];
+	pa = __pa(&lc->floating_pt_save_area);
 	if (__pcpu_sigp_relax(pcpu->address, SIGP_STORE_STATUS_AT_ADDRESS,
 			      pa) != SIGP_CC_ORDER_CODE_ACCEPTED)
 		return -EIO;
 	if (!MACHINE_HAS_VX && !MACHINE_HAS_GS)
 		return 0;
-	pa = __pa(pcpu->lowcore->mcesad & MCESA_ORIGIN_MASK);
+	pa = __pa(lc->mcesad & MCESA_ORIGIN_MASK);
 	if (MACHINE_HAS_GS)
-		pa |= pcpu->lowcore->mcesad & MCESA_LC_MASK;
+		pa |= lc->mcesad & MCESA_LC_MASK;
 	if (__pcpu_sigp_relax(pcpu->address, SIGP_STORE_ADDITIONAL_STATUS,
 			      pa) != SIGP_CC_ORDER_CODE_ACCEPTED)
 		return -EIO;
@@ -997,7 +1005,6 @@ void __init smp_prepare_boot_cpu(void)
 
 	WARN_ON(!cpu_present(0) || !cpu_online(0));
 	pcpu->state = CPU_STATE_CONFIGURED;
-	pcpu->lowcore = (struct lowcore *)(unsigned long) store_prefix();
 	S390_lowcore.percpu_offset = __per_cpu_offset[0];
 	smp_cpu_set_polarization(0, POLARIZATION_UNKNOWN);
 }
@@ -1264,7 +1271,6 @@ static int __init smp_reinit_ipl_cpu(void)
 	S390_lowcore.async_stack = async_stack + STACK_INIT_OFFSET;
 	S390_lowcore.mcck_stack = mcck_stack + STACK_INIT_OFFSET;
 	lowcore_ptr[0] = lc;
-	pcpu_devices[0].lowcore = lc;
 	local_mcck_enable();
 	local_irq_restore(flags);
 
-- 
cgit v1.2.3


From 0a500447b847f87b06ae814253376d983ef98e04 Mon Sep 17 00:00:00 2001
From: Sven Schnelle <svens@linux.ibm.com>
Date: Mon, 10 May 2021 21:10:18 +0200
Subject: s390: use struct tpi_info in lowcore.h

Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/include/asm/lowcore.h | 13 +++++++++----
 arch/s390/kernel/irq.c          |  4 ++--
 drivers/s390/cio/trace.h        |  6 ++----
 3 files changed, 13 insertions(+), 10 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/include/asm/lowcore.h b/arch/s390/include/asm/lowcore.h
index 3c85c60f56ac..adf7f473ad8f 100644
--- a/arch/s390/include/asm/lowcore.h
+++ b/arch/s390/include/asm/lowcore.h
@@ -48,10 +48,15 @@ struct lowcore {
 	__u8	pad_0x00a4[0x00a8-0x00a4];	/* 0x00a4 */
 	__u64	trans_exc_code;			/* 0x00a8 */
 	__u64	monitor_code;			/* 0x00b0 */
-	__u16	subchannel_id;			/* 0x00b8 */
-	__u16	subchannel_nr;			/* 0x00ba */
-	__u32	io_int_parm;			/* 0x00bc */
-	__u32	io_int_word;			/* 0x00c0 */
+	union {
+		struct {
+			__u16	subchannel_id;	/* 0x00b8 */
+			__u16	subchannel_nr;	/* 0x00ba */
+			__u32	io_int_parm;	/* 0x00bc */
+			__u32	io_int_word;	/* 0x00c0 */
+		};
+		struct tpi_info	tpi_info;	/* 0x00b8 */
+	};
 	__u8	pad_0x00c4[0x00c8-0x00c4];	/* 0x00c4 */
 	__u32	stfl_fac_list;			/* 0x00c8 */
 	__u8	pad_0x00cc[0x00e8-0x00cc];	/* 0x00cc */
diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c
index 169258012530..c0df4060d28d 100644
--- a/arch/s390/kernel/irq.c
+++ b/arch/s390/kernel/irq.c
@@ -146,8 +146,8 @@ void noinstr do_io_irq(struct pt_regs *regs)
 		account_idle_time_irq();
 
 	do {
-		memcpy(&regs->int_code, &S390_lowcore.subchannel_id, 12);
-		if (S390_lowcore.io_int_word & BIT(31))
+		regs->tpi_info = S390_lowcore.tpi_info;
+		if (S390_lowcore.tpi_info.adapter_IO)
 			do_irq_async(regs, THIN_INTERRUPT);
 		else
 			do_irq_async(regs, IO_INTERRUPT);
diff --git a/drivers/s390/cio/trace.h b/drivers/s390/cio/trace.h
index 4803139bce14..86993de25345 100644
--- a/drivers/s390/cio/trace.h
+++ b/drivers/s390/cio/trace.h
@@ -168,10 +168,8 @@ TRACE_EVENT(s390_cio_tpi,
 			memset(&__entry->tpi_info, 0, sizeof(struct tpi_info));
 		else if (addr)
 			__entry->tpi_info = *addr;
-		else {
-			memcpy(&__entry->tpi_info, &S390_lowcore.subchannel_id,
-			       sizeof(struct tpi_info));
-		}
+		else
+			__entry->tpi_info = S390_lowcore.tpi_info;
 		__entry->cssid = __entry->tpi_info.schid.cssid;
 		__entry->ssid = __entry->tpi_info.schid.ssid;
 		__entry->schno = __entry->tpi_info.schid.sch_no;
-- 
cgit v1.2.3


From 20232b18e5345385851ae7e2b7ef9a7cf983333a Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@linux.ibm.com>
Date: Mon, 17 May 2021 08:18:13 +0200
Subject: s390/mcck: cleanup use of cleanup_sie_mcck

cleanup_sie_mcck label is called from a single location only
and thus does not need to be a subroutine. Move the labelled
code to the caller - by doing that the SIE critical section
checks appear next to each other and the SIE cleanup becomes
bit more readable.

Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/entry.S | 22 ++++++++++------------
 1 file changed, 10 insertions(+), 12 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index a070a5d10409..7cfa08b67cae 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -214,7 +214,7 @@ ENTRY(sie64a)
 # are some corner cases (e.g. runtime instrumentation) where ILC is unpredictable.
 # Other instructions between sie64a and .Lsie_done should not cause program
 # interrupts. So lets use 3 nops as a landing pad for all possible rewinds.
-# See also .Lcleanup_sie_mcck/.Lcleanup_sie_int
+# See also .Lcleanup_sie
 .Lrewind_pad6:
 	nopr	7
 .Lrewind_pad4:
@@ -399,7 +399,7 @@ ENTRY(\name)
 	lghi	%r13,.Lsie_done - .Lsie_gmap
 	clgr	%r14,%r13
 	jhe	0f
-	brasl	%r14,.Lcleanup_sie_int
+	brasl	%r14,.Lcleanup_sie
 #endif
 0:	CHECK_STACK __LC_SAVE_AREA_ASYNC
 	aghi	%r15,-(STACK_FRAME_OVERHEAD + __PT_SIZE)
@@ -547,7 +547,13 @@ ENTRY(mcck_int_handler)
 	lghi	%r13,.Lsie_done - .Lsie_gmap
 	clgr	%r14,%r13
 	jhe	.Lmcck_stack
-	brasl	%r14,.Lcleanup_sie_mcck
+	larl	%r13,.Lsie_entry
+	slgr	%r9,%r13
+	lghi	%r13,.Lsie_skip - .Lsie_entry
+	clgr	%r9,%r13
+	jhe	5f
+	oi	__LC_CPU_FLAGS+7, _CIF_MCCK_GUEST
+5:	brasl	%r14,.Lcleanup_sie
 #endif
 	j	.Lmcck_stack
 .Lmcck_user:
@@ -649,21 +655,13 @@ ENDPROC(stack_overflow)
 #endif
 
 #if IS_ENABLED(CONFIG_KVM)
-.Lcleanup_sie_mcck:
-	larl	%r13,.Lsie_entry
-	slgr	%r9,%r13
-	lghi	%r13,.Lsie_skip - .Lsie_entry
-	clgr	%r9,%r13
-	jhe	.Lcleanup_sie_int
-	oi	__LC_CPU_FLAGS+7, _CIF_MCCK_GUEST
-.Lcleanup_sie_int:
+.Lcleanup_sie:
 	BPENTER	__SF_SIE_FLAGS(%r15),(_TIF_ISOLATE_BP|_TIF_ISOLATE_BP_GUEST)
 	lg	%r9,__SF_SIE_CONTROL(%r15)	# get control block pointer
 	ni	__SIE_PROG0C+3(%r9),0xfe	# no longer in SIE
 	lctlg	%c1,%c1,__LC_KERNEL_ASCE
 	larl	%r9,sie_exit			# skip forward to sie_exit
 	BR_EX	%r14,%r13
-
 #endif
 	.section .rodata, "a"
 #define SYSCALL(esame,emu)	.quad __s390x_ ## esame
-- 
cgit v1.2.3


From b5415c8f9755069640aad184293198bcf794f66d Mon Sep 17 00:00:00 2001
From: Alexander Gordeev <agordeev@linux.ibm.com>
Date: Mon, 7 Jun 2021 12:01:49 +0200
Subject: s390/entry.S: factor out OUTSIDE macro

Introduce OUTSIDE macro that checks whether an instruction
address is inside or outside of a block of instructions.

Signed-off-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/entry.S | 48 +++++++++++++++++++++++++-----------------------
 1 file changed, 25 insertions(+), 23 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/entry.S b/arch/s390/kernel/entry.S
index 7cfa08b67cae..187f704cc9e9 100644
--- a/arch/s390/kernel/entry.S
+++ b/arch/s390/kernel/entry.S
@@ -129,6 +129,27 @@ _LPP_OFFSET	= __LC_LPP
 		    "jnz .+8; .long 0xb2e8d000", 82
 	.endm
 
+#if IS_ENABLED(CONFIG_KVM)
+	/*
+	 * The OUTSIDE macro jumps to the provided label in case the value
+	 * in the provided register is outside of the provided range. The
+	 * macro is useful for checking whether a PSW stored in a register
+	 * pair points inside or outside of a block of instructions.
+	 * @reg: register to check
+	 * @start: start of the range
+	 * @end: end of the range
+	 * @outside_label: jump here if @reg is outside of [@start..@end)
+	 */
+	.macro OUTSIDE reg,start,end,outside_label
+	lgr	%r14,\reg
+	larl	%r13,\start
+	slgr	%r14,%r13
+	lghi	%r13,\end - \start
+	clgr	%r14,%r13
+	jhe	\outside_label
+	.endm
+#endif
+
 	GEN_BR_THUNK %r14
 	GEN_BR_THUNK %r14,%r13
 
@@ -319,12 +340,7 @@ ENTRY(pgm_check_handler)
 .Lpgm_skip_asce:
 #if IS_ENABLED(CONFIG_KVM)
 	# cleanup critical section for program checks in sie64a
-	lgr	%r14,%r9
-	larl	%r13,.Lsie_gmap
-	slgr	%r14,%r13
-	lghi	%r13,.Lsie_done - .Lsie_gmap
-	clgr	%r14,%r13
-	jhe	1f
+	OUTSIDE	%r9,.Lsie_gmap,.Lsie_done,1f
 	lg	%r14,__SF_SIE_CONTROL(%r15)	# get control block pointer
 	ni	__SIE_PROG0C+3(%r14),0xfe	# no longer in SIE
 	lctlg	%c1,%c1,__LC_KERNEL_ASCE	# load primary asce
@@ -393,12 +409,7 @@ ENTRY(\name)
 	tmhh	%r8,0x0001			# interrupting from user ?
 	jnz	1f
 #if IS_ENABLED(CONFIG_KVM)
-	lgr	%r14,%r9
-	larl	%r13,.Lsie_gmap
-	slgr	%r14,%r13
-	lghi	%r13,.Lsie_done - .Lsie_gmap
-	clgr	%r14,%r13
-	jhe	0f
+	OUTSIDE	%r9,.Lsie_gmap,.Lsie_done,0f
 	brasl	%r14,.Lcleanup_sie
 #endif
 0:	CHECK_STACK __LC_SAVE_AREA_ASYNC
@@ -541,17 +552,8 @@ ENTRY(mcck_int_handler)
 	tmhh	%r8,0x0001			# interrupting from user ?
 	jnz	.Lmcck_user
 #if IS_ENABLED(CONFIG_KVM)
-	lgr	%r14,%r9
-	larl	%r13,.Lsie_gmap
-	slgr	%r14,%r13
-	lghi	%r13,.Lsie_done - .Lsie_gmap
-	clgr	%r14,%r13
-	jhe	.Lmcck_stack
-	larl	%r13,.Lsie_entry
-	slgr	%r9,%r13
-	lghi	%r13,.Lsie_skip - .Lsie_entry
-	clgr	%r9,%r13
-	jhe	5f
+	OUTSIDE	%r9,.Lsie_gmap,.Lsie_done,.Lmcck_stack
+	OUTSIDE	%r9,.Lsie_entry,.Lsie_skip,5f
 	oi	__LC_CPU_FLAGS+7, _CIF_MCCK_GUEST
 5:	brasl	%r14,.Lcleanup_sie
 #endif
-- 
cgit v1.2.3


From 0c4f2623b95779fe8cfb277fa255e4b91c0f96f0 Mon Sep 17 00:00:00 2001
From: Vasily Gorbik <gor@linux.ibm.com>
Date: Tue, 6 Oct 2020 22:12:39 +0200
Subject: s390: setup kernel memory layout early

Currently there are two separate places where kernel memory layout has
to be known and adjusted:
1. early kasan setup.
2. paging setup later.

Those 2 places had to be kept in sync and adjusted to reflect peculiar
technical details of one another. With additional factors which influence
kernel memory layout like ultravisor secure storage limit, complexity
of keeping two things in sync grew up even more.

Besides that if we look forward towards creating identity mapping and
enabling DAT before jumping into uncompressed kernel - that would also
require full knowledge of and control over kernel memory layout.

So, de-duplicate and move kernel memory layout setup logic into
the decompressor.

Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/boot/boot.h           |  1 +
 arch/s390/boot/ipl_parm.c       |  8 ++--
 arch/s390/boot/startup.c        | 88 +++++++++++++++++++++++++++++++++++++++++
 arch/s390/boot/uv.c             | 25 ++++++++++++
 arch/s390/include/asm/kasan.h   |  1 -
 arch/s390/include/asm/pgtable.h | 13 +++---
 arch/s390/include/asm/setup.h   |  1 -
 arch/s390/kernel/setup.c        | 50 ++---------------------
 arch/s390/kernel/uv.c           |  8 +---
 arch/s390/mm/kasan_init.c       | 35 +++-------------
 10 files changed, 135 insertions(+), 95 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/boot/boot.h b/arch/s390/boot/boot.h
index 8b50967f5804..ae04e1c93764 100644
--- a/arch/s390/boot/boot.h
+++ b/arch/s390/boot/boot.h
@@ -24,6 +24,7 @@ void __printf(1, 2) decompressor_printk(const char *fmt, ...);
 
 extern const char kernel_version[];
 extern unsigned long memory_limit;
+extern unsigned long vmalloc_size;
 extern int vmalloc_size_set;
 extern int kaslr_enabled;
 
diff --git a/arch/s390/boot/ipl_parm.c b/arch/s390/boot/ipl_parm.c
index 3485a1d4e97b..bc1f973e656a 100644
--- a/arch/s390/boot/ipl_parm.c
+++ b/arch/s390/boot/ipl_parm.c
@@ -12,13 +12,13 @@
 #include "boot.h"
 
 char __bootdata(early_command_line)[COMMAND_LINE_SIZE];
+int __bootdata(noexec_disabled);
+
+unsigned int __bootdata_preserved(zlib_dfltcc_support) = ZLIB_DFLTCC_FULL;
 struct ipl_parameter_block __bootdata_preserved(ipl_block);
 int __bootdata_preserved(ipl_block_valid);
-unsigned int __bootdata_preserved(zlib_dfltcc_support) = ZLIB_DFLTCC_FULL;
-
-unsigned long __bootdata(vmalloc_size) = VMALLOC_DEFAULT_SIZE;
-int __bootdata(noexec_disabled);
 
+unsigned long vmalloc_size = VMALLOC_DEFAULT_SIZE;
 unsigned long memory_limit;
 int vmalloc_size_set;
 int kaslr_enabled;
diff --git a/arch/s390/boot/startup.c b/arch/s390/boot/startup.c
index 61a8ac4067e5..bbe4df6c2f8b 100644
--- a/arch/s390/boot/startup.c
+++ b/arch/s390/boot/startup.c
@@ -5,6 +5,7 @@
 #include <asm/sections.h>
 #include <asm/cpu_mf.h>
 #include <asm/setup.h>
+#include <asm/kasan.h>
 #include <asm/kexec.h>
 #include <asm/sclp.h>
 #include <asm/diag.h>
@@ -15,6 +16,12 @@
 extern char __boot_data_start[], __boot_data_end[];
 extern char __boot_data_preserved_start[], __boot_data_preserved_end[];
 unsigned long __bootdata_preserved(__kaslr_offset);
+unsigned long __bootdata_preserved(VMALLOC_START);
+unsigned long __bootdata_preserved(VMALLOC_END);
+struct page *__bootdata_preserved(vmemmap);
+unsigned long __bootdata_preserved(vmemmap_size);
+unsigned long __bootdata_preserved(MODULES_VADDR);
+unsigned long __bootdata_preserved(MODULES_END);
 unsigned long __bootdata(ident_map_size);
 
 u64 __bootdata_preserved(stfle_fac_list[16]);
@@ -172,6 +179,86 @@ static void setup_ident_map_size(unsigned long max_physmem_end)
 #endif
 }
 
+static void setup_kernel_memory_layout(void)
+{
+	bool vmalloc_size_verified = false;
+	unsigned long vmemmap_off;
+	unsigned long vspace_left;
+	unsigned long rte_size;
+	unsigned long pages;
+	unsigned long vmax;
+
+	pages = ident_map_size / PAGE_SIZE;
+	/* vmemmap contains a multiple of PAGES_PER_SECTION struct pages */
+	vmemmap_size = SECTION_ALIGN_UP(pages) * sizeof(struct page);
+
+	/* choose kernel address space layout: 4 or 3 levels. */
+	vmemmap_off = round_up(ident_map_size, _REGION3_SIZE);
+	if (IS_ENABLED(CONFIG_KASAN) ||
+	    vmalloc_size > _REGION2_SIZE ||
+	    vmemmap_off + vmemmap_size + vmalloc_size + MODULES_LEN > _REGION2_SIZE)
+		vmax = _REGION1_SIZE;
+	else
+		vmax = _REGION2_SIZE;
+
+	/* keep vmemmap_off aligned to a top level region table entry */
+	rte_size = vmax == _REGION1_SIZE ? _REGION2_SIZE : _REGION3_SIZE;
+	MODULES_END = vmax;
+	if (is_prot_virt_host()) {
+		/*
+		 * forcing modules and vmalloc area under the ultravisor
+		 * secure storage limit, so that any vmalloc allocation
+		 * we do could be used to back secure guest storage.
+		 */
+		adjust_to_uv_max(&MODULES_END);
+	}
+
+#ifdef CONFIG_KASAN
+	if (MODULES_END < vmax) {
+		/* force vmalloc and modules below kasan shadow */
+		MODULES_END = min(MODULES_END, KASAN_SHADOW_START);
+	} else {
+		/*
+		 * leave vmalloc and modules above kasan shadow but make
+		 * sure they don't overlap with it
+		 */
+		vmalloc_size = min(vmalloc_size, vmax - KASAN_SHADOW_END - MODULES_LEN);
+		vmalloc_size_verified = true;
+		vspace_left = KASAN_SHADOW_START;
+	}
+#endif
+	MODULES_VADDR = MODULES_END - MODULES_LEN;
+	VMALLOC_END = MODULES_VADDR;
+
+	if (vmalloc_size_verified) {
+		VMALLOC_START = VMALLOC_END - vmalloc_size;
+	} else {
+		vmemmap_off = round_up(ident_map_size, rte_size);
+
+		if (vmemmap_off + vmemmap_size > VMALLOC_END ||
+		    vmalloc_size > VMALLOC_END - vmemmap_off - vmemmap_size) {
+			/*
+			 * allow vmalloc area to occupy up to 1/2 of
+			 * the rest virtual space left.
+			 */
+			vmalloc_size = min(vmalloc_size, VMALLOC_END / 2);
+		}
+		VMALLOC_START = VMALLOC_END - vmalloc_size;
+		vspace_left = VMALLOC_START;
+	}
+
+	pages = vspace_left / (PAGE_SIZE + sizeof(struct page));
+	pages = SECTION_ALIGN_UP(pages);
+	vmemmap_off = round_up(vspace_left - pages * sizeof(struct page), rte_size);
+	/* keep vmemmap left most starting from a fresh region table entry */
+	vmemmap_off = min(vmemmap_off, round_up(ident_map_size, rte_size));
+	/* take care that identity map is lower then vmemmap */
+	ident_map_size = min(ident_map_size, vmemmap_off);
+	vmemmap_size = SECTION_ALIGN_UP(ident_map_size / PAGE_SIZE) * sizeof(struct page);
+	VMALLOC_START = max(vmemmap_off + vmemmap_size, VMALLOC_START);
+	vmemmap = (struct page *)vmemmap_off;
+}
+
 /*
  * This function clears the BSS section of the decompressed Linux kernel and NOT the decompressor's.
  */
@@ -211,6 +298,7 @@ void startup_kernel(void)
 	parse_boot_command_line();
 	setup_ident_map_size(detect_memory());
 	setup_vmalloc_size();
+	setup_kernel_memory_layout();
 
 	random_lma = __kaslr_offset = 0;
 	if (IS_ENABLED(CONFIG_RANDOMIZE_BASE) && kaslr_enabled) {
diff --git a/arch/s390/boot/uv.c b/arch/s390/boot/uv.c
index 87641dd65ccf..82b99b916243 100644
--- a/arch/s390/boot/uv.c
+++ b/arch/s390/boot/uv.c
@@ -44,3 +44,28 @@ void uv_query_info(void)
 		prot_virt_guest = 1;
 #endif
 }
+
+#if IS_ENABLED(CONFIG_KVM)
+static bool has_uv_sec_stor_limit(void)
+{
+	/*
+	 * keep these conditions in line with setup_uv()
+	 */
+	if (!is_prot_virt_host())
+		return false;
+
+	if (is_prot_virt_guest())
+		return false;
+
+	if (!test_facility(158))
+		return false;
+
+	return !!uv_info.max_sec_stor_addr;
+}
+
+void adjust_to_uv_max(unsigned long *vmax)
+{
+	if (has_uv_sec_stor_limit())
+		*vmax = min_t(unsigned long, *vmax, uv_info.max_sec_stor_addr);
+}
+#endif
diff --git a/arch/s390/include/asm/kasan.h b/arch/s390/include/asm/kasan.h
index 76f351bd6645..2768d5db181f 100644
--- a/arch/s390/include/asm/kasan.h
+++ b/arch/s390/include/asm/kasan.h
@@ -16,7 +16,6 @@
 extern void kasan_early_init(void);
 extern void kasan_copy_shadow_mapping(void);
 extern void kasan_free_early_identity(void);
-extern unsigned long kasan_vmax;
 
 /*
  * Estimate kasan memory requirements, which it will reserve
diff --git a/arch/s390/include/asm/pgtable.h b/arch/s390/include/asm/pgtable.h
index 9512f6820ead..5677be473261 100644
--- a/arch/s390/include/asm/pgtable.h
+++ b/arch/s390/include/asm/pgtable.h
@@ -17,6 +17,7 @@
 #include <linux/page-flags.h>
 #include <linux/radix-tree.h>
 #include <linux/atomic.h>
+#include <asm/sections.h>
 #include <asm/bug.h>
 #include <asm/page.h>
 #include <asm/uv.h>
@@ -86,16 +87,16 @@ extern unsigned long zero_page_mask;
  * happen without trampolines and in addition the placement within a
  * 2GB frame is branch prediction unit friendly.
  */
-extern unsigned long VMALLOC_START;
-extern unsigned long VMALLOC_END;
+extern unsigned long __bootdata_preserved(VMALLOC_START);
+extern unsigned long __bootdata_preserved(VMALLOC_END);
 #define VMALLOC_DEFAULT_SIZE	((512UL << 30) - MODULES_LEN)
-extern struct page *vmemmap;
-extern unsigned long vmemmap_size;
+extern struct page *__bootdata_preserved(vmemmap);
+extern unsigned long __bootdata_preserved(vmemmap_size);
 
 #define VMEM_MAX_PHYS ((unsigned long) vmemmap)
 
-extern unsigned long MODULES_VADDR;
-extern unsigned long MODULES_END;
+extern unsigned long __bootdata_preserved(MODULES_VADDR);
+extern unsigned long __bootdata_preserved(MODULES_END);
 #define MODULES_VADDR	MODULES_VADDR
 #define MODULES_END	MODULES_END
 #define MODULES_LEN	(1UL << 31)
diff --git a/arch/s390/include/asm/setup.h b/arch/s390/include/asm/setup.h
index 29baab03f091..a8b75da3c1b8 100644
--- a/arch/s390/include/asm/setup.h
+++ b/arch/s390/include/asm/setup.h
@@ -89,7 +89,6 @@ extern unsigned int zlib_dfltcc_support;
 
 extern int noexec_disabled;
 extern unsigned long ident_map_size;
-extern unsigned long vmalloc_size;
 
 /* The Write Back bit position in the physaddr is given by the SLPC PCI */
 extern unsigned long mio_wb_bit_mask;
diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index 146d01700a55..9145ef983f38 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -96,7 +96,6 @@ unsigned long int_hwcap = 0;
 
 int __bootdata(noexec_disabled);
 unsigned long __bootdata(ident_map_size);
-unsigned long __bootdata(vmalloc_size);
 struct mem_detect_info __bootdata(mem_detect);
 
 struct exception_table_entry *__bootdata_preserved(__start_dma_ex_table);
@@ -545,53 +544,10 @@ static void __init setup_resources(void)
 #endif
 }
 
-static void __init setup_ident_map_size(void)
+static void __init setup_memory_end(void)
 {
-	unsigned long vmax, tmp;
-
-	/* Choose kernel address space layout: 3 or 4 levels. */
-	tmp = ident_map_size / PAGE_SIZE;
-	tmp = tmp * (sizeof(struct page) + PAGE_SIZE);
-	if (tmp + vmalloc_size + MODULES_LEN <= _REGION2_SIZE)
-		vmax = _REGION2_SIZE; /* 3-level kernel page table */
-	else
-		vmax = _REGION1_SIZE; /* 4-level kernel page table */
-	/* module area is at the end of the kernel address space. */
-	MODULES_END = vmax;
-	if (is_prot_virt_host())
-		adjust_to_uv_max(&MODULES_END);
-#ifdef CONFIG_KASAN
-	vmax = _REGION1_SIZE;
-	MODULES_END = kasan_vmax;
-#endif
-	MODULES_VADDR = MODULES_END - MODULES_LEN;
-	VMALLOC_END = MODULES_VADDR;
-	VMALLOC_START = VMALLOC_END - vmalloc_size;
-
-	/* Split remaining virtual space between 1:1 mapping & vmemmap array */
-	tmp = VMALLOC_START / (PAGE_SIZE + sizeof(struct page));
-	/* vmemmap contains a multiple of PAGES_PER_SECTION struct pages */
-	tmp = SECTION_ALIGN_UP(tmp);
-	tmp = VMALLOC_START - tmp * sizeof(struct page);
-	tmp &= ~((vmax >> 11) - 1);	/* align to page table level */
-	tmp = min(tmp, 1UL << MAX_PHYSMEM_BITS);
-	vmemmap = (struct page *) tmp;
-
-	/* Take care that ident_map_size <= vmemmap */
-	ident_map_size = min(ident_map_size, (unsigned long)vmemmap);
-#ifdef CONFIG_KASAN
-	ident_map_size = min(ident_map_size, KASAN_SHADOW_START);
-#endif
-	vmemmap_size = SECTION_ALIGN_UP(ident_map_size / PAGE_SIZE) * sizeof(struct page);
-#ifdef CONFIG_KASAN
-	/* move vmemmap above kasan shadow only if stands in a way */
-	if (KASAN_SHADOW_END > (unsigned long)vmemmap &&
-	    (unsigned long)vmemmap + vmemmap_size > KASAN_SHADOW_START)
-		vmemmap = max(vmemmap, (struct page *)KASAN_SHADOW_END);
-#endif
-	max_pfn = max_low_pfn = PFN_DOWN(ident_map_size);
 	memblock_remove(ident_map_size, ULONG_MAX);
-
+	max_pfn = max_low_pfn = PFN_DOWN(ident_map_size);
 	pr_notice("The maximum memory size is %luMB\n", ident_map_size >> 20);
 }
 
@@ -1132,7 +1088,7 @@ void __init setup_arch(char **cmdline_p)
 	remove_oldmem();
 
 	setup_uv();
-	setup_ident_map_size();
+	setup_memory_end();
 	setup_memory();
 	dma_contiguous_reserve(ident_map_size);
 	vmcp_cma_reserve();
diff --git a/arch/s390/kernel/uv.c b/arch/s390/kernel/uv.c
index 370f664580af..6be2167943bb 100644
--- a/arch/s390/kernel/uv.c
+++ b/arch/s390/kernel/uv.c
@@ -52,7 +52,7 @@ void __init setup_uv(void)
 	unsigned long uv_stor_base;
 
 	/*
-	 * keep these conditions in line with kasan init code has_uv_sec_stor_limit()
+	 * keep these conditions in line with has_uv_sec_stor_limit()
 	 */
 	if (!is_prot_virt_host())
 		return;
@@ -91,12 +91,6 @@ fail:
 	prot_virt_host = 0;
 }
 
-void adjust_to_uv_max(unsigned long *vmax)
-{
-	if (uv_info.max_sec_stor_addr)
-		*vmax = min_t(unsigned long, *vmax, uv_info.max_sec_stor_addr);
-}
-
 /*
  * Requests the Ultravisor to pin the page in the shared state. This will
  * cause an intercept when the guest attempts to unshare the pinned page.
diff --git a/arch/s390/mm/kasan_init.c b/arch/s390/mm/kasan_init.c
index db4d303aaaa9..a0fdc6dc5f9d 100644
--- a/arch/s390/mm/kasan_init.c
+++ b/arch/s390/mm/kasan_init.c
@@ -13,7 +13,6 @@
 #include <asm/setup.h>
 #include <asm/uv.h>
 
-unsigned long kasan_vmax;
 static unsigned long segment_pos __initdata;
 static unsigned long segment_low __initdata;
 static unsigned long pgalloc_pos __initdata;
@@ -251,28 +250,9 @@ static void __init kasan_early_detect_facilities(void)
 	}
 }
 
-static bool __init has_uv_sec_stor_limit(void)
-{
-	/*
-	 * keep these conditions in line with setup_uv()
-	 */
-	if (!is_prot_virt_host())
-		return false;
-
-	if (is_prot_virt_guest())
-		return false;
-
-	if (!test_facility(158))
-		return false;
-
-	return !!uv_info.max_sec_stor_addr;
-}
-
 void __init kasan_early_init(void)
 {
-	unsigned long untracked_mem_end;
 	unsigned long shadow_alloc_size;
-	unsigned long vmax_unlimited;
 	unsigned long initrd_end;
 	unsigned long memsize;
 	unsigned long pgt_prot = pgprot_val(PAGE_KERNEL_RO);
@@ -306,9 +286,6 @@ void __init kasan_early_init(void)
 	BUILD_BUG_ON(!IS_ALIGNED(KASAN_SHADOW_START, P4D_SIZE));
 	BUILD_BUG_ON(!IS_ALIGNED(KASAN_SHADOW_END, P4D_SIZE));
 	crst_table_init((unsigned long *)early_pg_dir, _REGION2_ENTRY_EMPTY);
-	untracked_mem_end = kasan_vmax = vmax_unlimited = _REGION1_SIZE;
-	if (has_uv_sec_stor_limit())
-		kasan_vmax = min(vmax_unlimited, uv_info.max_sec_stor_addr);
 
 	/* init kasan zero shadow */
 	crst_table_init((unsigned long *)kasan_early_shadow_p4d,
@@ -375,18 +352,18 @@ void __init kasan_early_init(void)
 	 */
 	/* populate kasan shadow (for identity mapping and zero page mapping) */
 	kasan_early_pgtable_populate(__sha(0), __sha(memsize), POPULATE_MAP);
-	if (IS_ENABLED(CONFIG_MODULES))
-		untracked_mem_end = kasan_vmax - MODULES_LEN;
 	if (IS_ENABLED(CONFIG_KASAN_VMALLOC)) {
-		untracked_mem_end = kasan_vmax - vmalloc_size - MODULES_LEN;
 		/* shallowly populate kasan shadow for vmalloc and modules */
-		kasan_early_pgtable_populate(__sha(untracked_mem_end), __sha(kasan_vmax),
+		kasan_early_pgtable_populate(__sha(VMALLOC_START), __sha(MODULES_END),
 					     POPULATE_SHALLOW);
 	}
 	/* populate kasan shadow for untracked memory */
-	kasan_early_pgtable_populate(__sha(ident_map_size), __sha(untracked_mem_end),
+	kasan_early_pgtable_populate(__sha(ident_map_size),
+				     IS_ENABLED(CONFIG_KASAN_VMALLOC) ?
+						   __sha(VMALLOC_START) :
+						   __sha(MODULES_VADDR),
 				     POPULATE_ZERO_SHADOW);
-	kasan_early_pgtable_populate(__sha(kasan_vmax), __sha(vmax_unlimited),
+	kasan_early_pgtable_populate(__sha(MODULES_END), __sha(_REGION1_SIZE),
 				     POPULATE_ZERO_SHADOW);
 	/* memory allocated for identity mapping structs will be freed later */
 	pgalloc_freeable = pgalloc_pos;
-- 
cgit v1.2.3


From 6a9100ad132c61e4ff345277862e3fecfb7cdf0e Mon Sep 17 00:00:00 2001
From: Vasily Gorbik <gor@linux.ibm.com>
Date: Thu, 8 Oct 2020 15:07:27 +0200
Subject: s390/setup: cleanup reserve/remove_oldmem

Since OLDMEM_BASE/OLDMEM_SIZE is already taken into consideration and is
reflected in ident_map_size. reserve/remove_oldmem() is no longer needed
and could be removed.

Reviewed-by: Alexander Gordeev <agordeev@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/setup.c | 26 --------------------------
 1 file changed, 26 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index 9145ef983f38..b58ee83f30e3 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -585,30 +585,6 @@ static void __init reserve_above_ident_map(void)
 	memblock_reserve(ident_map_size, ULONG_MAX);
 }
 
-/*
- * Make sure that oldmem, where the dump is stored, is protected
- */
-static void __init reserve_oldmem(void)
-{
-#ifdef CONFIG_CRASH_DUMP
-	if (OLDMEM_BASE)
-		/* Forget all memory above the running kdump system */
-		memblock_reserve(OLDMEM_SIZE, (phys_addr_t)ULONG_MAX);
-#endif
-}
-
-/*
- * Make sure that oldmem, where the dump is stored, is protected
- */
-static void __init remove_oldmem(void)
-{
-#ifdef CONFIG_CRASH_DUMP
-	if (OLDMEM_BASE)
-		/* Forget all memory above the running kdump system */
-		memblock_remove(OLDMEM_SIZE, (phys_addr_t)ULONG_MAX);
-#endif
-}
-
 /*
  * Reserve memory for kdump kernel to be loaded with kexec
  */
@@ -1074,7 +1050,6 @@ void __init setup_arch(char **cmdline_p)
 
 	/* Do some memory reservations *before* memory is added to memblock */
 	reserve_above_ident_map();
-	reserve_oldmem();
 	reserve_kernel();
 	reserve_initrd();
 	reserve_certificate_list();
@@ -1085,7 +1060,6 @@ void __init setup_arch(char **cmdline_p)
 	memblock_add_mem_detect_info();
 
 	free_mem_detect_info();
-	remove_oldmem();
 
 	setup_uv();
 	setup_memory_end();
-- 
cgit v1.2.3


From d2beeb3bc74ec897ced7309c3a104fa4b3be0ac3 Mon Sep 17 00:00:00 2001
From: Peter Oberparleiter <oberpar@linux.ibm.com>
Date: Wed, 9 Jun 2021 14:05:56 +0200
Subject: s390/debug: Remove pointer obfuscation

When read via debugfs, s390dbf debug-views print the kernel address of
the call-site that created a trace entry. The kernel's %p pointer
hashing feature obfuscates this address, and commit 860ec7c6e21c
("s390/debug: use pK for kernel pointers") made this obfuscation
configurable via the kptr_restrict sysctl.

Obfuscation of kernel address data printed via s390dbf debug-views does
not add any additional protection since the associated debugfs files are
only accessible to the root user that typically has enough other means
to obtain kernel address data.

Also trace payload data may contain binary representations of kernel
addresses as part of logged data structues. Requiring such payload data
to be obfuscated as well would be impractical and greatly diminish the
use of s390dbf.

Therefore completely remove pointer obfuscation from s390dbf
debug-views.

Reviewed-by: Steffen Maier <maier@linux.ibm.com>
Acked-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Peter Oberparleiter <oberpar@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/debug.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/debug.c b/arch/s390/kernel/debug.c
index bb958d32bd81..09b6c6402f9b 100644
--- a/arch/s390/kernel/debug.c
+++ b/arch/s390/kernel/debug.c
@@ -1418,7 +1418,7 @@ int debug_dflt_header_fn(debug_info_t *id, struct debug_view *view,
 	else
 		except_str = "-";
 	caller = (unsigned long) entry->caller;
-	rc += sprintf(out_buf, "%02i %011ld:%06lu %1u %1s %04u %pK  ",
+	rc += sprintf(out_buf, "%02i %011ld:%06lu %1u %1s %04u %px  ",
 		      area, sec, usec, level, except_str,
 		      entry->cpu, (void *)caller);
 	return rc;
-- 
cgit v1.2.3


From ddd38fd261e7cd7772caea0128d0bee65bf1b159 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <hca@linux.ibm.com>
Date: Wed, 9 Jun 2021 22:59:13 +0200
Subject: s390/smp: use register pair instead of register asm

Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/smp.c | 22 +++++++++++-----------
 1 file changed, 11 insertions(+), 11 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index e137c840a4d3..c8493e7205ce 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -1233,21 +1233,21 @@ subsys_initcall(s390_smp_init);
 
 static __always_inline void set_new_lowcore(struct lowcore *lc)
 {
-	struct lowcore *old_lc = &S390_lowcore;
-	struct lowcore *new_lc = lc;
+	union register_pair dst, src;
 	u32 pfx;
-	register struct lowcore *reg2 asm ("2") = new_lc;
-	register unsigned long	 reg3 asm ("3") = sizeof(*reg2);
-	register struct lowcore *reg4 asm ("4") = old_lc;
-	register unsigned long	 reg5 asm ("5") = sizeof(*reg4);
+
+	src.even = (unsigned long) &S390_lowcore;
+	src.odd  = sizeof(S390_lowcore);
+	dst.even = (unsigned long) lc;
+	dst.odd  = sizeof(*lc);
+	pfx = (unsigned long) lc;
 
 	asm volatile(
-		"	st	2,%[pfx]\n"
-		"	mvcl	2,4\n"
+		"	mvcl	%[dst],%[src]\n"
 		"	spx	%[pfx]\n"
-		: "+&d" (reg2), "+&d" (reg3),
-		  "+&d" (reg4), "+&d" (reg5), [pfx] "=Q" (pfx)
-		: : "memory", "cc");
+		: [dst] "+&d" (dst.pair), [src] "+&d" (src.pair)
+		: [pfx] "Q" (pfx)
+		: "memory", "cc");
 }
 
 static int __init smp_reinit_ipl_cpu(void)
-- 
cgit v1.2.3


From 3c45a07bee619bb997b75b323f8d014e3efb36f8 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <hca@linux.ibm.com>
Date: Wed, 9 Jun 2021 22:59:13 +0200
Subject: s390/diag: use register pair instead of register asm

Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/diag.c | 11 +++++------
 1 file changed, 5 insertions(+), 6 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/diag.c b/arch/s390/kernel/diag.c
index b8b0cd7b008f..a3f47464c3f1 100644
--- a/arch/s390/kernel/diag.c
+++ b/arch/s390/kernel/diag.c
@@ -141,16 +141,15 @@ EXPORT_SYMBOL(diag14);
 
 static inline int __diag204(unsigned long *subcode, unsigned long size, void *addr)
 {
-	register unsigned long _subcode asm("0") = *subcode;
-	register unsigned long _size asm("1") = size;
+	union register_pair rp = { .even = *subcode, .odd = size };
 
 	asm volatile(
-		"	diag	%2,%0,0x204\n"
+		"	diag	%[addr],%[rp],0x204\n"
 		"0:	nopr	%%r7\n"
 		EX_TABLE(0b,0b)
-		: "+d" (_subcode), "+d" (_size) : "d" (addr) : "memory");
-	*subcode = _subcode;
-	return _size;
+		: [rp] "+&d" (rp.pair) : [addr] "d" (addr) : "memory");
+	*subcode = rp.even;
+	return rp.odd;
 }
 
 int diag204(unsigned long subcode, unsigned long size, void *addr)
-- 
cgit v1.2.3


From 6a7b4e4ee1d9f7444dc7fe3ec1b3cf441a3ce9b4 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <hca@linux.ibm.com>
Date: Wed, 9 Jun 2021 22:59:13 +0200
Subject: s390/sthyi: use register pair instead of register asm

Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/sthyi.c | 13 ++++++-------
 1 file changed, 6 insertions(+), 7 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/sthyi.c b/arch/s390/kernel/sthyi.c
index 888cc2f166db..4d141e2c132e 100644
--- a/arch/s390/kernel/sthyi.c
+++ b/arch/s390/kernel/sthyi.c
@@ -395,19 +395,18 @@ out:
 
 static int sthyi(u64 vaddr, u64 *rc)
 {
-	register u64 code asm("0") = 0;
-	register u64 addr asm("2") = vaddr;
-	register u64 rcode asm("3");
+	union register_pair r1 = { .even = 0, }; /* subcode */
+	union register_pair r2 = { .even = vaddr, };
 	int cc;
 
 	asm volatile(
-		".insn   rre,0xB2560000,%[code],%[addr]\n"
+		".insn   rre,0xB2560000,%[r1],%[r2]\n"
 		"ipm     %[cc]\n"
 		"srl     %[cc],28\n"
-		: [cc] "=d" (cc), "=d" (rcode)
-		: [code] "d" (code), [addr] "a" (addr)
+		: [cc] "=&d" (cc), [r2] "+&d" (r2.pair)
+		: [r1] "d" (r1.pair)
 		: "memory", "cc");
-	*rc = rcode;
+	*rc = r2.odd;
 	return cc;
 }
 
-- 
cgit v1.2.3


From 0a9d947fbe3e89e44758e1b8bf0098e98b3ccd78 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <hca@linux.ibm.com>
Date: Wed, 9 Jun 2021 22:59:13 +0200
Subject: s390/cpcmd: use register pair instead of register asm

Remove register asm usage from diag8_noresponse() since it wasn't
needed at all. There is no requirement for even/odd register pairs for
diag 0x8.

For diag_response() use register pairs to fulfill the rx+1 and ry+1
requirements as required if a response buffer is specified. Also
change the inline asm to return the condition code of the diagnose
instruction and do the conditional handling of response length
calculation in C.

Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/cpcmd.c | 42 ++++++++++++++++++++++--------------------
 1 file changed, 22 insertions(+), 20 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/cpcmd.c b/arch/s390/kernel/cpcmd.c
index 2da027359798..54efc279f54e 100644
--- a/arch/s390/kernel/cpcmd.c
+++ b/arch/s390/kernel/cpcmd.c
@@ -26,33 +26,35 @@ static char cpcmd_buf[241];
 
 static int diag8_noresponse(int cmdlen)
 {
-	register unsigned long reg2 asm ("2") = (addr_t) cpcmd_buf;
-	register unsigned long reg3 asm ("3") = cmdlen;
-
 	asm volatile(
-		"	diag	%1,%0,0x8\n"
-		: "+d" (reg3) : "d" (reg2) : "cc");
-	return reg3;
+		"	diag	%[rx],%[ry],0x8\n"
+		: [ry] "+&d" (cmdlen)
+		: [rx] "d" ((addr_t) cpcmd_buf)
+		: "cc");
+	return cmdlen;
 }
 
 static int diag8_response(int cmdlen, char *response, int *rlen)
 {
-	unsigned long _cmdlen = cmdlen | 0x40000000L;
-	unsigned long _rlen = *rlen;
-	register unsigned long reg2 asm ("2") = (addr_t) cpcmd_buf;
-	register unsigned long reg3 asm ("3") = (addr_t) response;
-	register unsigned long reg4 asm ("4") = _cmdlen;
-	register unsigned long reg5 asm ("5") = _rlen;
+	union register_pair rx, ry;
+	int cc;
 
+	rx.even = (addr_t) cpcmd_buf;
+	rx.odd	= (addr_t) response;
+	ry.even = cmdlen | 0x40000000L;
+	ry.odd	= *rlen;
 	asm volatile(
-		"	diag	%2,%0,0x8\n"
-		"	brc	8,1f\n"
-		"	agr	%1,%4\n"
-		"1:\n"
-		: "+d" (reg4), "+d" (reg5)
-		: "d" (reg2), "d" (reg3), "d" (*rlen) : "cc");
-	*rlen = reg5;
-	return reg4;
+		"	diag	%[rx],%[ry],0x8\n"
+		"	ipm	%[cc]\n"
+		"	srl	%[cc],28\n"
+		: [cc] "=&d" (cc), [ry] "+&d" (ry.pair)
+		: [rx] "d" (rx.pair)
+		: "cc");
+	if (cc)
+		*rlen += ry.odd;
+	else
+		*rlen = ry.odd;
+	return ry.even;
 }
 
 /*
-- 
cgit v1.2.3


From 5fe29839deb49ae5e9af32c1d344c867398b33eb Mon Sep 17 00:00:00 2001
From: Heiko Carstens <hca@linux.ibm.com>
Date: Mon, 14 Jun 2021 19:54:42 +0200
Subject: s390/sysinfo: get rid of register asm

Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/sysinfo.c | 19 +++++++++++--------
 1 file changed, 11 insertions(+), 8 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/sysinfo.c b/arch/s390/kernel/sysinfo.c
index 2ac3c9b56a13..ef3f2659876c 100644
--- a/arch/s390/kernel/sysinfo.c
+++ b/arch/s390/kernel/sysinfo.c
@@ -25,19 +25,22 @@ int topology_max_mnest;
 
 static inline int __stsi(void *sysinfo, int fc, int sel1, int sel2, int *lvl)
 {
-	register int r0 asm("0") = (fc << 28) | sel1;
-	register int r1 asm("1") = sel2;
+	int r0 = (fc << 28) | sel1;
 	int rc = 0;
 
 	asm volatile(
-		"	stsi	0(%3)\n"
+		"	lr	0,%[r0]\n"
+		"	lr	1,%[r1]\n"
+		"	stsi	0(%[sysinfo])\n"
 		"0:	jz	2f\n"
-		"1:	lhi	%1,%4\n"
-		"2:\n"
+		"1:	lhi	%[rc],%[retval]\n"
+		"2:	lr	%[r0],0\n"
 		EX_TABLE(0b, 1b)
-		: "+d" (r0), "+d" (rc)
-		: "d" (r1), "a" (sysinfo), "K" (-EOPNOTSUPP)
-		: "cc", "memory");
+		: [r0] "+d" (r0), [rc] "+d" (rc)
+		: [r1] "d" (sel2),
+		  [sysinfo] "a" (sysinfo),
+		  [retval] "K" (-EOPNOTSUPP)
+		: "cc", "0", "1", "memory");
 	*lvl = ((unsigned int) r0) >> 28;
 	return rc;
 }
-- 
cgit v1.2.3


From 5a4e0f58e2d959e2de0f0f1ddaa169e60711d2f0 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <hca@linux.ibm.com>
Date: Mon, 14 Jun 2021 22:32:21 +0200
Subject: s390/ipl: use register pair instead of register asm

Signed-off-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/ipl.c | 14 ++++++++------
 1 file changed, 8 insertions(+), 6 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/ipl.c b/arch/s390/kernel/ipl.c
index dba04fbc37a2..9ae33977e0b7 100644
--- a/arch/s390/kernel/ipl.c
+++ b/arch/s390/kernel/ipl.c
@@ -162,16 +162,18 @@ static bool reipl_ccw_clear;
 
 static inline int __diag308(unsigned long subcode, void *addr)
 {
-	register unsigned long _addr asm("0") = (unsigned long) addr;
-	register unsigned long _rc asm("1") = 0;
+	union register_pair r1;
 
+	r1.even = (unsigned long) addr;
+	r1.odd	= 0;
 	asm volatile(
-		"	diag	%0,%2,0x308\n"
+		"	diag	%[r1],%[subcode],0x308\n"
 		"0:	nopr	%%r7\n"
 		EX_TABLE(0b,0b)
-		: "+d" (_addr), "+d" (_rc)
-		: "d" (subcode) : "cc", "memory");
-	return _rc;
+		: [r1] "+&d" (r1.pair)
+		: [subcode] "d" (subcode)
+		: "cc", "memory");
+	return r1.odd;
 }
 
 int diag308(unsigned long subcode, void *addr)
-- 
cgit v1.2.3


From c74d3c182ab4a3db6c3c2a6c4b3c8b9a7f1feb1a Mon Sep 17 00:00:00 2001
From: Kees Cook <keescook@chromium.org>
Date: Wed, 16 Jun 2021 13:18:23 -0700
Subject: s390/speculation: Use statically initialized const for instructions

In preparation for FORTIFY_SOURCE performing compile-time and run-time
field bounds checking for memcpy(), memmove(), and memset(), avoid
confusing the checks when using a static const source.

Move the static const array into a variable so the compiler can perform
appropriate bounds checking.

Signed-off-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20210616201823.1245603-1-keescook@chromium.org
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>
---
 arch/s390/kernel/nospec-branch.c | 3 ++-
 1 file changed, 2 insertions(+), 1 deletion(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/nospec-branch.c b/arch/s390/kernel/nospec-branch.c
index 2c5c3756644b..250e4dbf653c 100644
--- a/arch/s390/kernel/nospec-branch.c
+++ b/arch/s390/kernel/nospec-branch.c
@@ -99,6 +99,7 @@ early_param("spectre_v2", spectre_v2_setup_early);
 static void __init_or_module __nospec_revert(s32 *start, s32 *end)
 {
 	enum { BRCL_EXPOLINE, BRASL_EXPOLINE } type;
+	static const u8 branch[] = { 0x47, 0x00, 0x07, 0x00 };
 	u8 *instr, *thunk, *br;
 	u8 insnbuf[6];
 	s32 *epo;
@@ -128,7 +129,7 @@ static void __init_or_module __nospec_revert(s32 *start, s32 *end)
 		if ((br[0] & 0xbf) != 0x07 || (br[1] & 0xf0) != 0xf0)
 			continue;
 
-		memcpy(insnbuf + 2, (char[]) { 0x47, 0x00, 0x07, 0x00 }, 4);
+		memcpy(insnbuf + 2, branch, sizeof(branch));
 		switch (type) {
 		case BRCL_EXPOLINE:
 			insnbuf[0] = br[0];
-- 
cgit v1.2.3