From b7d5006de1afab266175288f41e5dc70e69cce33 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Wed, 27 Aug 2014 07:35:05 +0200
Subject: s390: remove unused MACHINE_FLAG_RRBM

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/early.c | 2 --
 1 file changed, 2 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/early.c b/arch/s390/kernel/early.c
index 0dff972a169c..f6c66b5d0cfa 100644
--- a/arch/s390/kernel/early.c
+++ b/arch/s390/kernel/early.c
@@ -390,8 +390,6 @@ static __init void detect_machine_facilities(void)
 		S390_lowcore.machine_flags |= MACHINE_FLAG_LPP;
 	if (test_facility(50) && test_facility(73))
 		S390_lowcore.machine_flags |= MACHINE_FLAG_TE;
-	if (test_facility(66))
-		S390_lowcore.machine_flags |= MACHINE_FLAG_RRBM;
 	if (test_facility(51))
 		S390_lowcore.machine_flags |= MACHINE_FLAG_TLB_LC;
 #endif
-- 
cgit v1.2.3


From 070b7be633dc33c0899e8c934b4d5fad046b06e8 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Fri, 29 Aug 2014 12:44:40 +0200
Subject: s390/vdso: replace stck with stcke

If gettimeofday / clock_gettime are called multiple times in a row
the STCK instruction will stall until a difference in the result is
visible. This unnecessarily slows down the vdso calls, use stcke
instead of stck to get rid of the stall.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/vdso32/clock_gettime.S | 8 ++++----
 arch/s390/kernel/vdso32/gettimeofday.S  | 4 ++--
 arch/s390/kernel/vdso64/clock_gettime.S | 8 ++++----
 arch/s390/kernel/vdso64/gettimeofday.S  | 4 ++--
 4 files changed, 12 insertions(+), 12 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/vdso32/clock_gettime.S b/arch/s390/kernel/vdso32/clock_gettime.S
index 7cf18f8d4cb4..4e20a9365bb7 100644
--- a/arch/s390/kernel/vdso32/clock_gettime.S
+++ b/arch/s390/kernel/vdso32/clock_gettime.S
@@ -30,8 +30,8 @@ __kernel_clock_gettime:
 1:	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
 	tml	%r4,0x0001			/* pending update ? loop */
 	jnz	1b
-	stck	24(%r15)			/* Store TOD clock */
-	lm	%r0,%r1,24(%r15)
+	stcke	24(%r15)			/* Store TOD clock */
+	lm	%r0,%r1,25(%r15)
 	s	%r0,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
 	sl	%r1,__VDSO_XTIME_STAMP+4(%r5)
 	brc	3,2f
@@ -72,8 +72,8 @@ __kernel_clock_gettime:
 11:	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
 	tml	%r4,0x0001			/* pending update ? loop */
 	jnz	11b
-	stck	24(%r15)			/* Store TOD clock */
-	lm	%r0,%r1,24(%r15)
+	stcke	24(%r15)			/* Store TOD clock */
+	lm	%r0,%r1,25(%r15)
 	s	%r0,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
 	sl	%r1,__VDSO_XTIME_STAMP+4(%r5)
 	brc	3,12f
diff --git a/arch/s390/kernel/vdso32/gettimeofday.S b/arch/s390/kernel/vdso32/gettimeofday.S
index fd621a950f7c..60def5f562db 100644
--- a/arch/s390/kernel/vdso32/gettimeofday.S
+++ b/arch/s390/kernel/vdso32/gettimeofday.S
@@ -29,8 +29,8 @@ __kernel_gettimeofday:
 	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
 	tml	%r4,0x0001			/* pending update ? loop */
 	jnz	1b
-	stck	24(%r15)			/* Store TOD clock */
-	lm	%r0,%r1,24(%r15)
+	stcke	24(%r15)			/* Store TOD clock */
+	lm	%r0,%r1,25(%r15)
 	s	%r0,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
 	sl	%r1,__VDSO_XTIME_STAMP+4(%r5)
 	brc	3,3f
diff --git a/arch/s390/kernel/vdso64/clock_gettime.S b/arch/s390/kernel/vdso64/clock_gettime.S
index 3f34e09db5f4..4add40b78ed5 100644
--- a/arch/s390/kernel/vdso64/clock_gettime.S
+++ b/arch/s390/kernel/vdso64/clock_gettime.S
@@ -33,10 +33,10 @@ __kernel_clock_gettime:
 0:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
 	tmll	%r4,0x0001			/* pending update ? loop */
 	jnz	0b
-	stck	48(%r15)			/* Store TOD clock */
+	stcke	48(%r15)			/* Store TOD clock */
 	lgf	%r2,__VDSO_TK_SHIFT(%r5)	/* Timekeeper shift */
 	lg	%r0,__VDSO_WTOM_SEC(%r5)
-	lg	%r1,48(%r15)
+	lg	%r1,49(%r15)
 	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
 	msgf	%r1,__VDSO_TK_MULT(%r5)		/*  * tk->mult */
 	alg	%r1,__VDSO_WTOM_NSEC(%r5)
@@ -58,9 +58,9 @@ __kernel_clock_gettime:
 5:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
 	tmll	%r4,0x0001			/* pending update ? loop */
 	jnz	5b
-	stck	48(%r15)			/* Store TOD clock */
+	stcke	48(%r15)			/* Store TOD clock */
 	lgf	%r2,__VDSO_TK_SHIFT(%r5)	/* Timekeeper shift */
-	lg	%r1,48(%r15)
+	lg	%r1,49(%r15)
 	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
 	msgf	%r1,__VDSO_TK_MULT(%r5)		/*  * tk->mult */
 	alg	%r1,__VDSO_XTIME_NSEC(%r5)	/*  + tk->xtime_nsec */
diff --git a/arch/s390/kernel/vdso64/gettimeofday.S b/arch/s390/kernel/vdso64/gettimeofday.S
index d0860d1d0ccc..7a344995a97f 100644
--- a/arch/s390/kernel/vdso64/gettimeofday.S
+++ b/arch/s390/kernel/vdso64/gettimeofday.S
@@ -28,8 +28,8 @@ __kernel_gettimeofday:
 	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
 	tmll	%r4,0x0001			/* pending update ? loop */
 	jnz	0b
-	stck	48(%r15)			/* Store TOD clock */
-	lg	%r1,48(%r15)
+	stcke	48(%r15)			/* Store TOD clock */
+	lg	%r1,49(%r15)
 	sg	%r1,__VDSO_XTIME_STAMP(%r5)	/* TOD - cycle_last */
 	msgf	%r1,__VDSO_TK_MULT(%r5)		/*  * tk->mult */
 	alg	%r1,__VDSO_XTIME_NSEC(%r5)	/*  + tk->xtime_nsec */
-- 
cgit v1.2.3


From b7eacb59cd7fb5e98852186e485c0c865f862645 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Fri, 29 Aug 2014 12:31:45 +0200
Subject: s390/vdso: add vdso support for coarse clocks

Add CLOCK_REALTIME_COARSE and CLOCK_MONOTONIC_COARSE optimization to
the 64-bit and 31-bit vdso.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/asm-offsets.c          |  7 +++++++
 arch/s390/kernel/time.c                 | 13 +++++++++++++
 arch/s390/kernel/vdso32/clock_getres.S  | 11 +++++++++--
 arch/s390/kernel/vdso32/clock_gettime.S | 24 ++++++++++++++++++++++++
 arch/s390/kernel/vdso64/clock_getres.S  |  8 +++++++-
 arch/s390/kernel/vdso64/clock_gettime.S | 24 ++++++++++++++++++++++++
 6 files changed, 84 insertions(+), 3 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/asm-offsets.c b/arch/s390/kernel/asm-offsets.c
index afe1715a4eb7..3e9e479d9b49 100644
--- a/arch/s390/kernel/asm-offsets.c
+++ b/arch/s390/kernel/asm-offsets.c
@@ -62,8 +62,12 @@ int main(void)
 	DEFINE(__VDSO_XTIME_STAMP, offsetof(struct vdso_data, xtime_tod_stamp));
 	DEFINE(__VDSO_XTIME_SEC, offsetof(struct vdso_data, xtime_clock_sec));
 	DEFINE(__VDSO_XTIME_NSEC, offsetof(struct vdso_data, xtime_clock_nsec));
+	DEFINE(__VDSO_XTIME_CRS_SEC, offsetof(struct vdso_data, xtime_coarse_sec));
+	DEFINE(__VDSO_XTIME_CRS_NSEC, offsetof(struct vdso_data, xtime_coarse_nsec));
 	DEFINE(__VDSO_WTOM_SEC, offsetof(struct vdso_data, wtom_clock_sec));
 	DEFINE(__VDSO_WTOM_NSEC, offsetof(struct vdso_data, wtom_clock_nsec));
+	DEFINE(__VDSO_WTOM_CRS_SEC, offsetof(struct vdso_data, wtom_coarse_sec));
+	DEFINE(__VDSO_WTOM_CRS_NSEC, offsetof(struct vdso_data, wtom_coarse_nsec));
 	DEFINE(__VDSO_TIMEZONE, offsetof(struct vdso_data, tz_minuteswest));
 	DEFINE(__VDSO_ECTG_OK, offsetof(struct vdso_data, ectg_available));
 	DEFINE(__VDSO_TK_MULT, offsetof(struct vdso_data, tk_mult));
@@ -73,8 +77,11 @@ int main(void)
 	/* constants used by the vdso */
 	DEFINE(__CLOCK_REALTIME, CLOCK_REALTIME);
 	DEFINE(__CLOCK_MONOTONIC, CLOCK_MONOTONIC);
+	DEFINE(__CLOCK_REALTIME_COARSE, CLOCK_REALTIME_COARSE);
+	DEFINE(__CLOCK_MONOTONIC_COARSE, CLOCK_MONOTONIC_COARSE);
 	DEFINE(__CLOCK_THREAD_CPUTIME_ID, CLOCK_THREAD_CPUTIME_ID);
 	DEFINE(__CLOCK_REALTIME_RES, MONOTONIC_RES_NSEC);
+	DEFINE(__CLOCK_COARSE_RES, LOW_RES_NSEC);
 	BLANK();
 	/* idle data offsets */
 	DEFINE(__CLOCK_IDLE_ENTER, offsetof(struct s390_idle_data, clock_idle_enter));
diff --git a/arch/s390/kernel/time.c b/arch/s390/kernel/time.c
index 4cef607f3711..69e980de0f62 100644
--- a/arch/s390/kernel/time.c
+++ b/arch/s390/kernel/time.c
@@ -232,6 +232,19 @@ void update_vsyscall(struct timekeeper *tk)
 		vdso_data->wtom_clock_nsec -= nsecps;
 		vdso_data->wtom_clock_sec++;
 	}
+
+	vdso_data->xtime_coarse_sec = tk->xtime_sec;
+	vdso_data->xtime_coarse_nsec =
+		(long)(tk->tkr.xtime_nsec >> tk->tkr.shift);
+	vdso_data->wtom_coarse_sec =
+		vdso_data->xtime_coarse_sec + tk->wall_to_monotonic.tv_sec;
+	vdso_data->wtom_coarse_nsec =
+		vdso_data->xtime_coarse_nsec + tk->wall_to_monotonic.tv_nsec;
+	while (vdso_data->wtom_coarse_nsec >= NSEC_PER_SEC) {
+		vdso_data->wtom_coarse_nsec -= NSEC_PER_SEC;
+		vdso_data->wtom_coarse_sec++;
+	}
+
 	vdso_data->tk_mult = tk->tkr.mult;
 	vdso_data->tk_shift = tk->tkr.shift;
 	smp_wmb();
diff --git a/arch/s390/kernel/vdso32/clock_getres.S b/arch/s390/kernel/vdso32/clock_getres.S
index 36aaa25d05da..eca3f001f081 100644
--- a/arch/s390/kernel/vdso32/clock_getres.S
+++ b/arch/s390/kernel/vdso32/clock_getres.S
@@ -19,14 +19,20 @@
 	.type  __kernel_clock_getres,@function
 __kernel_clock_getres:
 	.cfi_startproc
+	basr	%r1,0
+	la	%r1,4f-.(%r1)
 	chi	%r2,__CLOCK_REALTIME
 	je	0f
 	chi	%r2,__CLOCK_MONOTONIC
+	je	0f
+	la	%r1,5f-4f(%r1)
+	chi	%r2,__CLOCK_REALTIME_COARSE
+	je	0f
+	chi	%r2,__CLOCK_MONOTONIC_COARSE
 	jne	3f
 0:	ltr	%r3,%r3
 	jz	2f				/* res == NULL */
-	basr	%r1,0
-1:	l	%r0,4f-1b(%r1)
+1:	l	%r0,0(%r1)
 	xc	0(4,%r3),0(%r3)			/* set tp->tv_sec to zero */
 	st	%r0,4(%r3)			/* store tp->tv_usec */
 2:	lhi	%r2,0
@@ -35,5 +41,6 @@ __kernel_clock_getres:
 	svc	0
 	br	%r14
 4:	.long	__CLOCK_REALTIME_RES
+5:	.long	__CLOCK_COARSE_RES
 	.cfi_endproc
 	.size	__kernel_clock_getres,.-__kernel_clock_getres
diff --git a/arch/s390/kernel/vdso32/clock_gettime.S b/arch/s390/kernel/vdso32/clock_gettime.S
index 4e20a9365bb7..48c2206a3956 100644
--- a/arch/s390/kernel/vdso32/clock_gettime.S
+++ b/arch/s390/kernel/vdso32/clock_gettime.S
@@ -21,8 +21,12 @@ __kernel_clock_gettime:
 	.cfi_startproc
 	basr	%r5,0
 0:	al	%r5,21f-0b(%r5)			/* get &_vdso_data */
+	chi	%r2,__CLOCK_REALTIME_COARSE
+	je	10f
 	chi	%r2,__CLOCK_REALTIME
 	je	11f
+	chi	%r2,__CLOCK_MONOTONIC_COARSE
+	je	9f
 	chi	%r2,__CLOCK_MONOTONIC
 	jne	19f
 
@@ -68,6 +72,26 @@ __kernel_clock_gettime:
 	lhi	%r2,0
 	br	%r14
 
+	/* CLOCK_MONOTONIC_COARSE */
+9:	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
+	tml	%r4,0x0001			/* pending update ? loop */
+	jnz	9b
+	l	%r2,__VDSO_WTOM_CRS_SEC+4(%r5)
+	l	%r1,__VDSO_WTOM_CRS_NSEC+4(%r5)
+	cl	%r4,__VDSO_UPD_COUNT+4(%r5)	/* check update counter */
+	jne	9b
+	j	8b
+
+	/* CLOCK_REALTIME_COARSE */
+10:	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
+	tml	%r4,0x0001			/* pending update ? loop */
+	jnz	10b
+	l	%r2,__VDSO_XTIME_CRS_SEC+4(%r5)
+	l	%r1,__VDSO_XTIME_CRS_NSEC+4(%r5)
+	cl	%r4,__VDSO_UPD_COUNT+4(%r5)	/* check update counter */
+	jne	10b
+	j	17f
+
 	/* CLOCK_REALTIME */
 11:	l	%r4,__VDSO_UPD_COUNT+4(%r5)	/* load update counter */
 	tml	%r4,0x0001			/* pending update ? loop */
diff --git a/arch/s390/kernel/vdso64/clock_getres.S b/arch/s390/kernel/vdso64/clock_getres.S
index 34deba7c7ed1..c8513deb8c66 100644
--- a/arch/s390/kernel/vdso64/clock_getres.S
+++ b/arch/s390/kernel/vdso64/clock_getres.S
@@ -19,6 +19,12 @@
 	.type  __kernel_clock_getres,@function
 __kernel_clock_getres:
 	.cfi_startproc
+	larl	%r1,4f
+	cghi	%r2,__CLOCK_REALTIME_COARSE
+	je	0f
+	cghi	%r2,__CLOCK_MONOTONIC_COARSE
+	je	0f
+	larl	%r1,3f
 	cghi	%r2,__CLOCK_REALTIME
 	je	0f
 	cghi	%r2,__CLOCK_MONOTONIC
@@ -32,7 +38,6 @@ __kernel_clock_getres:
 	jz	2f
 0:	ltgr	%r3,%r3
 	jz	1f				/* res == NULL */
-	larl	%r1,3f
 	lg	%r0,0(%r1)
 	xc	0(8,%r3),0(%r3)			/* set tp->tv_sec to zero */
 	stg	%r0,8(%r3)			/* store tp->tv_usec */
@@ -42,5 +47,6 @@ __kernel_clock_getres:
 	svc	0
 	br	%r14
 3:	.quad	__CLOCK_REALTIME_RES
+4:	.quad	__CLOCK_COARSE_RES
 	.cfi_endproc
 	.size	__kernel_clock_getres,.-__kernel_clock_getres
diff --git a/arch/s390/kernel/vdso64/clock_gettime.S b/arch/s390/kernel/vdso64/clock_gettime.S
index 4add40b78ed5..9d9761f8e110 100644
--- a/arch/s390/kernel/vdso64/clock_gettime.S
+++ b/arch/s390/kernel/vdso64/clock_gettime.S
@@ -20,12 +20,16 @@
 __kernel_clock_gettime:
 	.cfi_startproc
 	larl	%r5,_vdso_data
+	cghi	%r2,__CLOCK_REALTIME_COARSE
+	je	4f
 	cghi	%r2,__CLOCK_REALTIME
 	je	5f
 	cghi	%r2,__CLOCK_THREAD_CPUTIME_ID
 	je	9f
 	cghi	%r2,-2		/* Per-thread CPUCLOCK with PID=0, VIRT=1 */
 	je	9f
+	cghi	%r2,__CLOCK_MONOTONIC_COARSE
+	je	3f
 	cghi	%r2,__CLOCK_MONOTONIC
 	jne	12f
 
@@ -54,6 +58,26 @@ __kernel_clock_gettime:
 	lghi	%r2,0
 	br	%r14
 
+	/* CLOCK_MONOTONIC_COARSE */
+3:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
+	tmll	%r4,0x0001			/* pending update ? loop */
+	jnz	3b
+	lg	%r0,__VDSO_WTOM_CRS_SEC(%r5)
+	lg	%r1,__VDSO_WTOM_CRS_NSEC(%r5)
+	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
+	jne	3b
+	j	2b
+
+	/* CLOCK_REALTIME_COARSE */
+4:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
+	tmll	%r4,0x0001			/* pending update ? loop */
+	jnz	4b
+	lg	%r0,__VDSO_XTIME_CRS_SEC(%r5)
+	lg	%r1,__VDSO_XTIME_CRS_NSEC(%r5)
+	clg	%r4,__VDSO_UPD_COUNT(%r5)	/* check update counter */
+	jne	4b
+	j	7f
+
 	/* CLOCK_REALTIME */
 5:	lg	%r4,__VDSO_UPD_COUNT(%r5)	/* load update counter */
 	tmll	%r4,0x0001			/* pending update ? loop */
-- 
cgit v1.2.3


From 2481a87b0250bbf429fc8cdc78331efbc44a0221 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Fri, 15 Aug 2014 12:33:46 +0200
Subject: s390/ftrace: optimize function graph caller code

When the function graph tracer is disabled we can skip three additional
instructions. So let's just do this.

So if function tracing is enabled but function graph tracing is
runtime disabled, we get away with a single unconditional branch.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/ftrace.c   | 24 ++++++++++++++++++++++++
 arch/s390/kernel/mcount64.S | 15 +++++++++------
 2 files changed, 33 insertions(+), 6 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c
index 54d6493c4a56..de55efa5b64e 100644
--- a/arch/s390/kernel/ftrace.c
+++ b/arch/s390/kernel/ftrace.c
@@ -170,6 +170,29 @@ out:
  * directly after the instructions. To enable the call we calculate
  * the original offset to prepare_ftrace_return and put it back.
  */
+
+#ifdef CONFIG_64BIT
+
+int ftrace_enable_ftrace_graph_caller(void)
+{
+	static unsigned short offset = 0x0002;
+
+	return probe_kernel_write((void *) ftrace_graph_caller + 2,
+				  &offset, sizeof(offset));
+}
+
+int ftrace_disable_ftrace_graph_caller(void)
+{
+	unsigned short offset;
+
+	offset = ((void *) &ftrace_graph_caller_end -
+		  (void *) ftrace_graph_caller) / 2;
+	return probe_kernel_write((void *) ftrace_graph_caller + 2,
+				  &offset, sizeof(offset));
+}
+
+#else /* CONFIG_64BIT */
+
 int ftrace_enable_ftrace_graph_caller(void)
 {
 	unsigned short offset;
@@ -188,5 +211,6 @@ int ftrace_disable_ftrace_graph_caller(void)
 				  &offset, sizeof(offset));
 }
 
+#endif /* CONFIG_64BIT */
 #endif /* CONFIG_DYNAMIC_FTRACE */
 #endif /* CONFIG_FUNCTION_GRAPH_TRACER */
diff --git a/arch/s390/kernel/mcount64.S b/arch/s390/kernel/mcount64.S
index c67a8bf0fd9a..5b33c83adde9 100644
--- a/arch/s390/kernel/mcount64.S
+++ b/arch/s390/kernel/mcount64.S
@@ -32,14 +32,17 @@ ENTRY(ftrace_caller)
 	lg	%r14,0(%r14)
 	basr	%r14,%r14
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
+# The j instruction gets runtime patched to a nop instruction.
+# See ftrace_enable_ftrace_graph_caller. The patched instruction is:
+#	j	.+4
+ENTRY(ftrace_graph_caller)
+	j	ftrace_graph_caller_end
 	lg	%r2,168(%r15)
 	lg	%r3,272(%r15)
-ENTRY(ftrace_graph_caller)
-# The bras instruction gets runtime patched to call prepare_ftrace_return.
-# See ftrace_enable_ftrace_graph_caller. The patched instruction is:
-#	bras	%r14,prepare_ftrace_return
-	bras	%r14,0f
-0:	stg	%r2,168(%r15)
+	brasl	%r14,prepare_ftrace_return
+	stg	%r2,168(%r15)
+ftrace_graph_caller_end:
+	.globl	ftrace_graph_caller_end
 #endif
 	aghi	%r15,160
 	lmg	%r2,%r5,32(%r15)
-- 
cgit v1.2.3


From 10dec7dbd50ab0be96dda085d625d54ce800e426 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Fri, 15 Aug 2014 13:01:46 +0200
Subject: s390/ftrace: add HAVE_DYNAMIC_FTRACE_WITH_REGS support

This code is based on a patch from Vojtech Pavlik.
http://marc.info/?l=linux-s390&m=140438885114413&w=2

The actual implementation now differs significantly:
Instead of adding a second function "ftrace_regs_caller" which would be nearly
identical to the existing ftrace_caller function, the current ftrace_caller
function is now an alias to ftrace_regs_caller and always passes the needed
pt_regs structure and function_trace_op parameters unconditionally.

Besides that also use asm offsets to correctly allocate and access the new
struct pt_regs on the stack.

While at it we can make use of new instruction to get rid of some indirect
loads if compiled for new machines.

The passed struct pt_regs can be changed by the called function and it's new
contents will replace the current contents.

Note: to change the return address the embedded psw member of the pt_regs
structure must be changed. The psw member is right now incomplete, since
the mask part is missing. For all current use cases this should be sufficent.
Providing and restoring a sane mask would mean we need to add an epsw/lpswe
pair to the mcount code. Only these two instruction would cost us ~120 cycles
which currently seems not necessary.

Cc: Vojtech Pavlik <vojtech@suse.cz>
Cc: Jiri Kosina <jkosina@suse.cz>
Cc: Jiri Slaby <jslaby@suse.cz>
Cc: Steven Rostedt <rostedt@goodmis.org>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/ftrace.c   |  7 +++++++
 arch/s390/kernel/mcount64.S | 43 +++++++++++++++++++++++++++++++------------
 2 files changed, 38 insertions(+), 12 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c
index de55efa5b64e..14b61954d5a8 100644
--- a/arch/s390/kernel/ftrace.c
+++ b/arch/s390/kernel/ftrace.c
@@ -107,6 +107,13 @@ asm(
 
 #endif /* CONFIG_64BIT */
 
+#ifdef CONFIG_64BIT
+int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr,
+		       unsigned long addr)
+{
+	return 0;
+}
+#endif
 
 int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec,
 		    unsigned long addr)
diff --git a/arch/s390/kernel/mcount64.S b/arch/s390/kernel/mcount64.S
index 5b33c83adde9..4a65dabae851 100644
--- a/arch/s390/kernel/mcount64.S
+++ b/arch/s390/kernel/mcount64.S
@@ -8,28 +8,47 @@
 #include <linux/linkage.h>
 #include <asm/asm-offsets.h>
 #include <asm/ftrace.h>
+#include <asm/ptrace.h>
 
 	.section .kprobes.text, "ax"
 
 ENTRY(ftrace_stub)
 	br	%r14
 
+#define STACK_FRAME_SIZE  (STACK_FRAME_OVERHEAD + __PT_SIZE)
+#define STACK_PARENT_IP   (STACK_FRAME_SIZE + 8)
+#define STACK_PTREGS	  (STACK_FRAME_OVERHEAD)
+#define STACK_PTREGS_GPRS (STACK_PTREGS + __PT_GPRS)
+#define STACK_PTREGS_PSW  (STACK_PTREGS + __PT_PSW)
+
 ENTRY(_mcount)
 #ifdef CONFIG_DYNAMIC_FTRACE
 	br	%r14
 
 ENTRY(ftrace_caller)
+	.globl	ftrace_regs_caller
+	.set	ftrace_regs_caller,ftrace_caller
 #endif
-	stmg	%r2,%r5,32(%r15)
-	stg	%r14,112(%r15)
 	lgr	%r1,%r15
-	aghi	%r15,-160
+	aghi	%r15,-STACK_FRAME_SIZE
 	stg	%r1,__SF_BACKCHAIN(%r15)
+	stg	%r1,(STACK_PTREGS_GPRS+15*8)(%r15)
+	stmg	%r0,%r13,STACK_PTREGS_GPRS(%r15)
+	stg	%r14,(STACK_PTREGS_PSW+8)(%r15)
+#ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
+	aghik	%r2,%r14,-MCOUNT_INSN_SIZE
+	lgrl	%r4,function_trace_op
+	lgrl	%r14,ftrace_trace_function
+#else
 	lgr	%r2,%r14
-	lg	%r3,168(%r15)
 	aghi	%r2,-MCOUNT_INSN_SIZE
+	larl	%r4,function_trace_op
+	lg	%r4,0(%r4)
 	larl	%r14,ftrace_trace_function
 	lg	%r14,0(%r14)
+#endif
+	lg	%r3,STACK_PARENT_IP(%r15)
+	la	%r5,STACK_PTREGS(%r15)
 	basr	%r14,%r14
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
 # The j instruction gets runtime patched to a nop instruction.
@@ -37,16 +56,16 @@ ENTRY(ftrace_caller)
 #	j	.+4
 ENTRY(ftrace_graph_caller)
 	j	ftrace_graph_caller_end
-	lg	%r2,168(%r15)
-	lg	%r3,272(%r15)
+	lg	%r2,STACK_PARENT_IP(%r15)
+	lg	%r3,(STACK_PTREGS_PSW+8)(%r15)
 	brasl	%r14,prepare_ftrace_return
-	stg	%r2,168(%r15)
+	stg	%r2,STACK_PARENT_IP(%r15)
 ftrace_graph_caller_end:
 	.globl	ftrace_graph_caller_end
 #endif
-	aghi	%r15,160
-	lmg	%r2,%r5,32(%r15)
-	lg	%r14,112(%r15)
+	lmg	%r0,%r13,STACK_PTREGS_GPRS(%r15)
+	lg	%r14,(STACK_PTREGS_PSW+8)(%r15)
+	aghi	%r15,STACK_FRAME_SIZE
 	br	%r14
 
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
@@ -54,10 +73,10 @@ ftrace_graph_caller_end:
 ENTRY(return_to_handler)
 	stmg	%r2,%r5,32(%r15)
 	lgr	%r1,%r15
-	aghi	%r15,-160
+	aghi	%r15,-STACK_FRAME_OVERHEAD
 	stg	%r1,__SF_BACKCHAIN(%r15)
 	brasl	%r14,ftrace_return_to_handler
-	aghi	%r15,160
+	aghi	%r15,STACK_FRAME_OVERHEAD
 	lgr	%r14,%r2
 	lmg	%r2,%r5,32(%r15)
 	br	%r14
-- 
cgit v1.2.3


From 5d6a0163494c78ad7b6de733c8793e66b5da9212 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Fri, 15 Aug 2014 13:16:09 +0200
Subject: s390/ftrace: enforce DYNAMIC_FTRACE if FUNCTION_TRACER is selected

We have too many combinations for function tracing. Lets simply stick to
the most advanced option, so we don't have to care of other combinations.

This means we always select DYNAMIC_FTRACE if FUNCTION_TRACER is selected.

In the s390 Makefile also remove CONFIG_FTRACE_SYSCALLS since that
functionality got moved to architecture independent code in the meantime.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/Makefile   | 4 +---
 arch/s390/kernel/ftrace.c   | 6 ------
 arch/s390/kernel/mcount.S   | 2 --
 arch/s390/kernel/mcount64.S | 2 --
 4 files changed, 1 insertion(+), 13 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile
index a95c4ca99617..d44245d4df37 100644
--- a/arch/s390/kernel/Makefile
+++ b/arch/s390/kernel/Makefile
@@ -53,9 +53,7 @@ obj-$(CONFIG_COMPAT)		+= compat_wrapper.o $(compat-obj-y)
 obj-$(CONFIG_STACKTRACE)	+= stacktrace.o
 obj-$(CONFIG_KPROBES)		+= kprobes.o
 obj-$(CONFIG_FUNCTION_TRACER)	+= $(if $(CONFIG_64BIT),mcount64.o,mcount.o)
-obj-$(CONFIG_DYNAMIC_FTRACE)	+= ftrace.o
-obj-$(CONFIG_FUNCTION_GRAPH_TRACER) += ftrace.o
-obj-$(CONFIG_FTRACE_SYSCALLS)  += ftrace.o
+obj-$(CONFIG_FUNCTION_TRACER)	+= ftrace.o
 obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
 
 ifdef CONFIG_64BIT
diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c
index 14b61954d5a8..f908e42e11c4 100644
--- a/arch/s390/kernel/ftrace.c
+++ b/arch/s390/kernel/ftrace.c
@@ -17,8 +17,6 @@
 #include <asm/asm-offsets.h>
 #include "entry.h"
 
-#ifdef CONFIG_DYNAMIC_FTRACE
-
 void ftrace_disable_code(void);
 void ftrace_enable_insn(void);
 
@@ -142,8 +140,6 @@ int __init ftrace_dyn_arch_init(void)
 	return 0;
 }
 
-#endif /* CONFIG_DYNAMIC_FTRACE */
-
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
 /*
  * Hook the return address and push it in the stack of return addresses
@@ -169,7 +165,6 @@ out:
 	return parent;
 }
 
-#ifdef CONFIG_DYNAMIC_FTRACE
 /*
  * Patch the kernel code at ftrace_graph_caller location. The instruction
  * there is branch relative and save to prepare_ftrace_return. To disable
@@ -219,5 +214,4 @@ int ftrace_disable_ftrace_graph_caller(void)
 }
 
 #endif /* CONFIG_64BIT */
-#endif /* CONFIG_DYNAMIC_FTRACE */
 #endif /* CONFIG_FUNCTION_GRAPH_TRACER */
diff --git a/arch/s390/kernel/mcount.S b/arch/s390/kernel/mcount.S
index 433c6dbfa442..be6dbd9a81a7 100644
--- a/arch/s390/kernel/mcount.S
+++ b/arch/s390/kernel/mcount.S
@@ -15,11 +15,9 @@ ENTRY(ftrace_stub)
 	br	%r14
 
 ENTRY(_mcount)
-#ifdef CONFIG_DYNAMIC_FTRACE
 	br	%r14
 
 ENTRY(ftrace_caller)
-#endif
 	stm	%r2,%r5,16(%r15)
 	bras	%r1,1f
 0:	.long	ftrace_trace_function
diff --git a/arch/s390/kernel/mcount64.S b/arch/s390/kernel/mcount64.S
index 4a65dabae851..8cf976f83a10 100644
--- a/arch/s390/kernel/mcount64.S
+++ b/arch/s390/kernel/mcount64.S
@@ -22,13 +22,11 @@ ENTRY(ftrace_stub)
 #define STACK_PTREGS_PSW  (STACK_PTREGS + __PT_PSW)
 
 ENTRY(_mcount)
-#ifdef CONFIG_DYNAMIC_FTRACE
 	br	%r14
 
 ENTRY(ftrace_caller)
 	.globl	ftrace_regs_caller
 	.set	ftrace_regs_caller,ftrace_caller
-#endif
 	lgr	%r1,%r15
 	aghi	%r15,-STACK_FRAME_SIZE
 	stg	%r1,__SF_BACKCHAIN(%r15)
-- 
cgit v1.2.3


From ea2f47699082b971769be8b8f38c08b49219f471 Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Wed, 3 Sep 2014 10:37:40 +0200
Subject: s390/kprobes: remove unused jprobe_return_end()

Even if it has a __used annotation it is actually unused.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/kprobes.c | 5 -----
 1 file changed, 5 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c
index bc71a7b95af5..c48a00c9f351 100644
--- a/arch/s390/kernel/kprobes.c
+++ b/arch/s390/kernel/kprobes.c
@@ -789,11 +789,6 @@ void __kprobes jprobe_return(void)
 	asm volatile(".word 0x0002");
 }
 
-static void __used __kprobes jprobe_return_end(void)
-{
-	asm volatile("bcr 0,0");
-}
-
 int __kprobes longjmp_break_handler(struct kprobe *p, struct pt_regs *regs)
 {
 	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
-- 
cgit v1.2.3


From 3d1e220d08c6a00ffa83d39030b8162f66665b2b Mon Sep 17 00:00:00 2001
From: Heiko Carstens <heiko.carstens@de.ibm.com>
Date: Wed, 3 Sep 2014 13:26:23 +0200
Subject: s390/ftrace: optimize mcount code

Reduce the number of executed instructions within the mcount block if
function tracing is enabled. We achieve that by using a non-standard
C function call ABI. Since the called function is also written in
assembler this is not a problem.
This also allows to replace the unconditional store at the beginning
of the mcount block with a larl instruction, which doesn't touch
memory.

In theory we could also patch the first instruction of the mcount block
to enable and disable function tracing. However this would break kprobes.
This could be fixed with implementing the "kprobes_on_ftrace" feature;
however keeping the odd jprobes working seems not to be possible without
a lot of code churn. Therefore keep the code easy and simply accept one
wasted 1-cycle "larl" instruction per function prologue.

Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/ftrace.c   | 57 +++++++++++++++++++++++++++++++--------------
 arch/s390/kernel/mcount64.S | 30 +++++++++++-------------
 2 files changed, 54 insertions(+), 33 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c
index f908e42e11c4..fcb009d3edde 100644
--- a/arch/s390/kernel/ftrace.c
+++ b/arch/s390/kernel/ftrace.c
@@ -1,7 +1,7 @@
 /*
  * Dynamic function tracer architecture backend.
  *
- * Copyright IBM Corp. 2009
+ * Copyright IBM Corp. 2009,2014
  *
  *   Author(s): Heiko Carstens <heiko.carstens@de.ibm.com>,
  *		Martin Schwidefsky <schwidefsky@de.ibm.com>
@@ -17,6 +17,7 @@
 #include <asm/asm-offsets.h>
 #include "entry.h"
 
+void mcount_replace_code(void);
 void ftrace_disable_code(void);
 void ftrace_enable_insn(void);
 
@@ -24,38 +25,50 @@ void ftrace_enable_insn(void);
 /*
  * The 64-bit mcount code looks like this:
  *	stg	%r14,8(%r15)		# offset 0
- * >	larl	%r1,<&counter>		# offset 6
- * >	brasl	%r14,_mcount		# offset 12
+ *	larl	%r1,<&counter>		# offset 6
+ *	brasl	%r14,_mcount		# offset 12
  *	lg	%r14,8(%r15)		# offset 18
- * Total length is 24 bytes. The middle two instructions of the mcount
- * block get overwritten by ftrace_make_nop / ftrace_make_call.
+ * Total length is 24 bytes. The complete mcount block initially gets replaced
+ * by ftrace_make_nop. Subsequent calls to ftrace_make_call / ftrace_make_nop
+ * only patch the jg/lg instruction within the block.
+ * Note: we do not patch the first instruction to an unconditional branch,
+ * since that would break kprobes/jprobes. It is easier to leave the larl
+ * instruction in and only modify the second instruction.
  * The 64-bit enabled ftrace code block looks like this:
- *	stg	%r14,8(%r15)		# offset 0
+ *	larl	%r0,.+24		# offset 0
  * >	lg	%r1,__LC_FTRACE_FUNC	# offset 6
- * >	lgr	%r0,%r0			# offset 12
- * >	basr	%r14,%r1		# offset 16
- *	lg	%r14,8(%15)		# offset 18
- * The return points of the mcount/ftrace function have the same offset 18.
- * The 64-bit disable ftrace code block looks like this:
- *	stg	%r14,8(%r15)		# offset 0
+ *	br	%r1			# offset 12
+ *	brcl	0,0			# offset 14
+ *	brc	0,0			# offset 20
+ * The ftrace function gets called with a non-standard C function call ABI
+ * where r0 contains the return address. It is also expected that the called
+ * function only clobbers r0 and r1, but restores r2-r15.
+ * The return point of the ftrace function has offset 24, so execution
+ * continues behind the mcount block.
+ *	larl	%r0,.+24		# offset 0
  * >	jg	.+18			# offset 6
- * >	lgr	%r0,%r0			# offset 12
- * >	basr	%r14,%r1		# offset 16
- *	lg	%r14,8(%15)		# offset 18
+ *	br	%r1			# offset 12
+ *	brcl	0,0			# offset 14
+ *	brc	0,0			# offset 20
  * The jg instruction branches to offset 24 to skip as many instructions
  * as possible.
  */
 asm(
 	"	.align	4\n"
+	"mcount_replace_code:\n"
+	"	larl	%r0,0f\n"
 	"ftrace_disable_code:\n"
 	"	jg	0f\n"
-	"	lgr	%r0,%r0\n"
-	"	basr	%r14,%r1\n"
+	"	br	%r1\n"
+	"	brcl	0,0\n"
+	"	brc	0,0\n"
 	"0:\n"
 	"	.align	4\n"
 	"ftrace_enable_insn:\n"
 	"	lg	%r1,"__stringify(__LC_FTRACE_FUNC)"\n");
 
+#define MCOUNT_BLOCK_SIZE	24
+#define MCOUNT_INSN_OFFSET	6
 #define FTRACE_INSN_SIZE	6
 
 #else /* CONFIG_64BIT */
@@ -116,6 +129,16 @@ int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr,
 int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec,
 		    unsigned long addr)
 {
+#ifdef CONFIG_64BIT
+	/* Initial replacement of the whole mcount block */
+	if (addr == MCOUNT_ADDR) {
+		if (probe_kernel_write((void *) rec->ip - MCOUNT_INSN_OFFSET,
+				       mcount_replace_code,
+				       MCOUNT_BLOCK_SIZE))
+			return -EPERM;
+		return 0;
+	}
+#endif
 	if (probe_kernel_write((void *) rec->ip, ftrace_disable_code,
 			       MCOUNT_INSN_SIZE))
 		return -EPERM;
diff --git a/arch/s390/kernel/mcount64.S b/arch/s390/kernel/mcount64.S
index 8cf976f83a10..07abe8d464d4 100644
--- a/arch/s390/kernel/mcount64.S
+++ b/arch/s390/kernel/mcount64.S
@@ -16,7 +16,6 @@ ENTRY(ftrace_stub)
 	br	%r14
 
 #define STACK_FRAME_SIZE  (STACK_FRAME_OVERHEAD + __PT_SIZE)
-#define STACK_PARENT_IP   (STACK_FRAME_SIZE + 8)
 #define STACK_PTREGS	  (STACK_FRAME_OVERHEAD)
 #define STACK_PTREGS_GPRS (STACK_PTREGS + __PT_GPRS)
 #define STACK_PTREGS_PSW  (STACK_PTREGS + __PT_PSW)
@@ -31,40 +30,39 @@ ENTRY(ftrace_caller)
 	aghi	%r15,-STACK_FRAME_SIZE
 	stg	%r1,__SF_BACKCHAIN(%r15)
 	stg	%r1,(STACK_PTREGS_GPRS+15*8)(%r15)
-	stmg	%r0,%r13,STACK_PTREGS_GPRS(%r15)
-	stg	%r14,(STACK_PTREGS_PSW+8)(%r15)
+	stg	%r0,(STACK_PTREGS_PSW+8)(%r15)
+	stmg	%r2,%r14,(STACK_PTREGS_GPRS+2*8)(%r15)
 #ifdef CONFIG_HAVE_MARCH_Z196_FEATURES
-	aghik	%r2,%r14,-MCOUNT_INSN_SIZE
+	aghik	%r2,%r0,-MCOUNT_INSN_SIZE
 	lgrl	%r4,function_trace_op
-	lgrl	%r14,ftrace_trace_function
+	lgrl	%r1,ftrace_trace_function
 #else
-	lgr	%r2,%r14
+	lgr	%r2,%r0
 	aghi	%r2,-MCOUNT_INSN_SIZE
 	larl	%r4,function_trace_op
 	lg	%r4,0(%r4)
-	larl	%r14,ftrace_trace_function
-	lg	%r14,0(%r14)
+	larl	%r1,ftrace_trace_function
+	lg	%r1,0(%r1)
 #endif
-	lg	%r3,STACK_PARENT_IP(%r15)
+	lgr	%r3,%r14
 	la	%r5,STACK_PTREGS(%r15)
-	basr	%r14,%r14
+	basr	%r14,%r1
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
 # The j instruction gets runtime patched to a nop instruction.
 # See ftrace_enable_ftrace_graph_caller. The patched instruction is:
 #	j	.+4
 ENTRY(ftrace_graph_caller)
 	j	ftrace_graph_caller_end
-	lg	%r2,STACK_PARENT_IP(%r15)
+	lg	%r2,(STACK_PTREGS_GPRS+14*8)(%r15)
 	lg	%r3,(STACK_PTREGS_PSW+8)(%r15)
 	brasl	%r14,prepare_ftrace_return
-	stg	%r2,STACK_PARENT_IP(%r15)
+	stg	%r2,(STACK_PTREGS_GPRS+14*8)(%r15)
 ftrace_graph_caller_end:
 	.globl	ftrace_graph_caller_end
 #endif
-	lmg	%r0,%r13,STACK_PTREGS_GPRS(%r15)
-	lg	%r14,(STACK_PTREGS_PSW+8)(%r15)
-	aghi	%r15,STACK_FRAME_SIZE
-	br	%r14
+	lg	%r1,(STACK_PTREGS_PSW+8)(%r15)
+	lmg	%r2,%r15,(STACK_PTREGS_GPRS+2*8)(%r15)
+	br	%r1
 
 #ifdef CONFIG_FUNCTION_GRAPH_TRACER
 
-- 
cgit v1.2.3


From 8f933b1043e1e51f4776fc1ffe86752c7785fd4e Mon Sep 17 00:00:00 2001
From: Ralf Hoppe <rhoppe@de.ibm.com>
Date: Mon, 8 Apr 2013 09:52:57 +0200
Subject: s390/hmcdrv: HMC drive CD/DVD access

This device driver allows accessing a HMC drive CD/DVD-ROM.
It can be used in a LPAR and z/VM environment.

Reviewed-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
Reviewed-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Ralf Hoppe <rhoppe@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/irq.c | 1 +
 1 file changed, 1 insertion(+)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c
index 8eb82443cfbd..051574ee5366 100644
--- a/arch/s390/kernel/irq.c
+++ b/arch/s390/kernel/irq.c
@@ -70,6 +70,7 @@ static const struct irq_class irqclass_sub_desc[NR_ARCH_IRQS] = {
 	{.irq = IRQEXT_CMS, .name = "CMS", .desc = "[EXT] CPU-Measurement: Sampling"},
 	{.irq = IRQEXT_CMC, .name = "CMC", .desc = "[EXT] CPU-Measurement: Counter"},
 	{.irq = IRQEXT_CMR, .name = "CMR", .desc = "[EXT] CPU-Measurement: RI"},
+	{.irq = IRQEXT_FTP, .name = "FTP", .desc = "[EXT] HMC FTP Service"},
 	{.irq = IRQIO_CIO,  .name = "CIO", .desc = "[I/O] Common I/O Layer Interrupt"},
 	{.irq = IRQIO_QAI,  .name = "QAI", .desc = "[I/O] QDIO Adapter Interrupt"},
 	{.irq = IRQIO_DAS,  .name = "DAS", .desc = "[I/O] DASD"},
-- 
cgit v1.2.3


From d59b93da5e572703e1a7311c13dd3472a4e56e30 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Fri, 19 Sep 2014 14:29:31 +0200
Subject: s390/rwlock: use directed yield for write-locked rwlocks

Add an owner field to the arch_rwlock_t to be able to pass the timeslice
of a virtual CPU with diagnose 0x9c to the lock owner in case the rwlock
is write-locked. The undirected yield in case the rwlock is acquired
writable but the lock is read-locked is removed.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/smp.c | 6 ------
 1 file changed, 6 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index 243c7e512600..abec97b4ddbf 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -333,12 +333,6 @@ int smp_vcpu_scheduled(int cpu)
 	return pcpu_running(pcpu_devices + cpu);
 }
 
-void smp_yield(void)
-{
-	if (MACHINE_HAS_DIAG44)
-		asm volatile("diag 0,0,0x44");
-}
-
 void smp_yield_cpu(int cpu)
 {
 	if (MACHINE_HAS_DIAG9C)
-- 
cgit v1.2.3


From b881dcfbf7fd89b2be801843b060b9ad77cc77e7 Mon Sep 17 00:00:00 2001
From: Christian Borntraeger <borntraeger@de.ibm.com>
Date: Fri, 19 Sep 2014 15:46:02 +0200
Subject: s390/head.s: use zero as address for stfl

The architecture suggests to use address 0 as parameter for stfl,
to allow for future extensions. Using __LC_STFL_FAC_LIST (0x200)
shows which address is used, but might be not future proof.

Signed-off-by: Christian Borntraeger <borntraeger@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/head.S | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/head.S b/arch/s390/kernel/head.S
index e88d35d74950..d62eee11f0b5 100644
--- a/arch/s390/kernel/head.S
+++ b/arch/s390/kernel/head.S
@@ -398,7 +398,7 @@ ENTRY(startup_kdump)
 	xc	__LC_STFL_FAC_LIST(8),__LC_STFL_FAC_LIST
 #ifndef CONFIG_MARCH_G5
 	# check capabilities against MARCH_{G5,Z900,Z990,Z9_109,Z10}
-	.insn	s,0xb2b10000,__LC_STFL_FAC_LIST	# store facility list
+	.insn	s,0xb2b10000,0		# store facilities @ __LC_STFL_FAC_LIST
 	tm	__LC_STFL_FAC_LIST,0x01	# stfle available ?
 	jz	0f
 	la	%r0,1
-- 
cgit v1.2.3


From 975fab17399a2b29985166181ad80e5f50fa42e9 Mon Sep 17 00:00:00 2001
From: Jan Willeke <willeke@de.ibm.com>
Date: Mon, 22 Sep 2014 16:37:27 +0200
Subject: s390/uprobes: common library for kprobes and uprobes

This patch moves common functions from kprobes.c to probes.c.
Thus its possible for uprobes to use them without enabling kprobes.

Signed-off-by: Jan Willeke <willeke@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/kprobes.c | 154 +--------------------------------------------
 1 file changed, 3 insertions(+), 151 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/kprobes.c b/arch/s390/kernel/kprobes.c
index c48a00c9f351..27ae5433fe4d 100644
--- a/arch/s390/kernel/kprobes.c
+++ b/arch/s390/kernel/kprobes.c
@@ -58,161 +58,13 @@ struct kprobe_insn_cache kprobe_dmainsn_slots = {
 	.insn_size = MAX_INSN_SIZE,
 };
 
-static int __kprobes is_prohibited_opcode(kprobe_opcode_t *insn)
-{
-	if (!is_known_insn((unsigned char *)insn))
-		return -EINVAL;
-	switch (insn[0] >> 8) {
-	case 0x0c:	/* bassm */
-	case 0x0b:	/* bsm	 */
-	case 0x83:	/* diag  */
-	case 0x44:	/* ex	 */
-	case 0xac:	/* stnsm */
-	case 0xad:	/* stosm */
-		return -EINVAL;
-	case 0xc6:
-		switch (insn[0] & 0x0f) {
-		case 0x00: /* exrl   */
-			return -EINVAL;
-		}
-	}
-	switch (insn[0]) {
-	case 0x0101:	/* pr	 */
-	case 0xb25a:	/* bsa	 */
-	case 0xb240:	/* bakr  */
-	case 0xb258:	/* bsg	 */
-	case 0xb218:	/* pc	 */
-	case 0xb228:	/* pt	 */
-	case 0xb98d:	/* epsw	 */
-		return -EINVAL;
-	}
-	return 0;
-}
-
-static int __kprobes get_fixup_type(kprobe_opcode_t *insn)
-{
-	/* default fixup method */
-	int fixup = FIXUP_PSW_NORMAL;
-
-	switch (insn[0] >> 8) {
-	case 0x05:	/* balr	*/
-	case 0x0d:	/* basr */
-		fixup = FIXUP_RETURN_REGISTER;
-		/* if r2 = 0, no branch will be taken */
-		if ((insn[0] & 0x0f) == 0)
-			fixup |= FIXUP_BRANCH_NOT_TAKEN;
-		break;
-	case 0x06:	/* bctr	*/
-	case 0x07:	/* bcr	*/
-		fixup = FIXUP_BRANCH_NOT_TAKEN;
-		break;
-	case 0x45:	/* bal	*/
-	case 0x4d:	/* bas	*/
-		fixup = FIXUP_RETURN_REGISTER;
-		break;
-	case 0x47:	/* bc	*/
-	case 0x46:	/* bct	*/
-	case 0x86:	/* bxh	*/
-	case 0x87:	/* bxle	*/
-		fixup = FIXUP_BRANCH_NOT_TAKEN;
-		break;
-	case 0x82:	/* lpsw	*/
-		fixup = FIXUP_NOT_REQUIRED;
-		break;
-	case 0xb2:	/* lpswe */
-		if ((insn[0] & 0xff) == 0xb2)
-			fixup = FIXUP_NOT_REQUIRED;
-		break;
-	case 0xa7:	/* bras	*/
-		if ((insn[0] & 0x0f) == 0x05)
-			fixup |= FIXUP_RETURN_REGISTER;
-		break;
-	case 0xc0:
-		if ((insn[0] & 0x0f) == 0x05)	/* brasl */
-			fixup |= FIXUP_RETURN_REGISTER;
-		break;
-	case 0xeb:
-		switch (insn[2] & 0xff) {
-		case 0x44: /* bxhg  */
-		case 0x45: /* bxleg */
-			fixup = FIXUP_BRANCH_NOT_TAKEN;
-			break;
-		}
-		break;
-	case 0xe3:	/* bctg	*/
-		if ((insn[2] & 0xff) == 0x46)
-			fixup = FIXUP_BRANCH_NOT_TAKEN;
-		break;
-	case 0xec:
-		switch (insn[2] & 0xff) {
-		case 0xe5: /* clgrb */
-		case 0xe6: /* cgrb  */
-		case 0xf6: /* crb   */
-		case 0xf7: /* clrb  */
-		case 0xfc: /* cgib  */
-		case 0xfd: /* cglib */
-		case 0xfe: /* cib   */
-		case 0xff: /* clib  */
-			fixup = FIXUP_BRANCH_NOT_TAKEN;
-			break;
-		}
-		break;
-	}
-	return fixup;
-}
-
-static int __kprobes is_insn_relative_long(kprobe_opcode_t *insn)
-{
-	/* Check if we have a RIL-b or RIL-c format instruction which
-	 * we need to modify in order to avoid instruction emulation. */
-	switch (insn[0] >> 8) {
-	case 0xc0:
-		if ((insn[0] & 0x0f) == 0x00) /* larl */
-			return true;
-		break;
-	case 0xc4:
-		switch (insn[0] & 0x0f) {
-		case 0x02: /* llhrl  */
-		case 0x04: /* lghrl  */
-		case 0x05: /* lhrl   */
-		case 0x06: /* llghrl */
-		case 0x07: /* sthrl  */
-		case 0x08: /* lgrl   */
-		case 0x0b: /* stgrl  */
-		case 0x0c: /* lgfrl  */
-		case 0x0d: /* lrl    */
-		case 0x0e: /* llgfrl */
-		case 0x0f: /* strl   */
-			return true;
-		}
-		break;
-	case 0xc6:
-		switch (insn[0] & 0x0f) {
-		case 0x02: /* pfdrl  */
-		case 0x04: /* cghrl  */
-		case 0x05: /* chrl   */
-		case 0x06: /* clghrl */
-		case 0x07: /* clhrl  */
-		case 0x08: /* cgrl   */
-		case 0x0a: /* clgrl  */
-		case 0x0c: /* cgfrl  */
-		case 0x0d: /* crl    */
-		case 0x0e: /* clgfrl */
-		case 0x0f: /* clrl   */
-			return true;
-		}
-		break;
-	}
-	return false;
-}
-
 static void __kprobes copy_instruction(struct kprobe *p)
 {
 	s64 disp, new_disp;
 	u64 addr, new_addr;
 
 	memcpy(p->ainsn.insn, p->addr, insn_length(p->opcode >> 8));
-	if (!is_insn_relative_long(p->ainsn.insn))
+	if (!probe_is_insn_relative_long(p->ainsn.insn))
 		return;
 	/*
 	 * For pc-relative instructions in RIL-b or RIL-c format patch the
@@ -276,7 +128,7 @@ int __kprobes arch_prepare_kprobe(struct kprobe *p)
 	if ((unsigned long) p->addr & 0x01)
 		return -EINVAL;
 	/* Make sure the probe isn't going on a difficult instruction */
-	if (is_prohibited_opcode(p->addr))
+	if (probe_is_prohibited_opcode(p->addr))
 		return -EINVAL;
 	if (s390_get_insn_slot(p))
 		return -ENOMEM;
@@ -605,7 +457,7 @@ static void __kprobes resume_execution(struct kprobe *p, struct pt_regs *regs)
 {
 	struct kprobe_ctlblk *kcb = get_kprobe_ctlblk();
 	unsigned long ip = regs->psw.addr & PSW_ADDR_INSN;
-	int fixup = get_fixup_type(p->ainsn.insn);
+	int fixup = probe_get_fixup_type(p->ainsn.insn);
 
 	if (fixup & FIXUP_PSW_NORMAL)
 		ip += (unsigned long) p->addr - (unsigned long) p->ainsn.insn;
-- 
cgit v1.2.3


From 2a0a5b2299b9bef76123fac91e68d39cb361c33e Mon Sep 17 00:00:00 2001
From: Jan Willeke <willeke@de.ibm.com>
Date: Mon, 22 Sep 2014 16:39:06 +0200
Subject: s390/uprobes: architecture backend for uprobes

Signed-off-by: Jan Willeke <willeke@de.ibm.com>
Signed-off-by: Heiko Carstens <heiko.carstens@de.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/Makefile  |   1 +
 arch/s390/kernel/entry.h   |   1 +
 arch/s390/kernel/entry64.S |  17 ++-
 arch/s390/kernel/ptrace.c  |   5 +-
 arch/s390/kernel/traps.c   |  33 +++--
 arch/s390/kernel/uprobes.c | 332 +++++++++++++++++++++++++++++++++++++++++++++
 6 files changed, 376 insertions(+), 13 deletions(-)
 create mode 100644 arch/s390/kernel/uprobes.c

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile
index d44245d4df37..3249e1f36d55 100644
--- a/arch/s390/kernel/Makefile
+++ b/arch/s390/kernel/Makefile
@@ -55,6 +55,7 @@ obj-$(CONFIG_KPROBES)		+= kprobes.o
 obj-$(CONFIG_FUNCTION_TRACER)	+= $(if $(CONFIG_64BIT),mcount64.o,mcount.o)
 obj-$(CONFIG_FUNCTION_TRACER)	+= ftrace.o
 obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
+obj-$(CONFIG_UPROBES)		+= uprobes.o
 
 ifdef CONFIG_64BIT
 obj-$(CONFIG_PERF_EVENTS)	+= perf_event.o perf_cpum_cf.o perf_cpum_sf.o \
diff --git a/arch/s390/kernel/entry.h b/arch/s390/kernel/entry.h
index 1aad48398d06..58541633b8d6 100644
--- a/arch/s390/kernel/entry.h
+++ b/arch/s390/kernel/entry.h
@@ -45,6 +45,7 @@ void transaction_exception(struct pt_regs *regs);
 void translation_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);
diff --git a/arch/s390/kernel/entry64.S b/arch/s390/kernel/entry64.S
index f2e674c702e1..7b2e03afd017 100644
--- a/arch/s390/kernel/entry64.S
+++ b/arch/s390/kernel/entry64.S
@@ -42,7 +42,8 @@ STACK_SHIFT = PAGE_SHIFT + THREAD_ORDER
 STACK_SIZE  = 1 << STACK_SHIFT
 STACK_INIT = STACK_SIZE - STACK_FRAME_OVERHEAD - __PT_SIZE
 
-_TIF_WORK	= (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_NEED_RESCHED)
+_TIF_WORK	= (_TIF_SIGPENDING | _TIF_NOTIFY_RESUME | _TIF_NEED_RESCHED | \
+		   _TIF_UPROBE)
 _TIF_TRACE	= (_TIF_SYSCALL_TRACE | _TIF_SYSCALL_AUDIT | _TIF_SECCOMP | \
 		   _TIF_SYSCALL_TRACEPOINT)
 _CIF_WORK	= (_CIF_MCCK_PENDING | _CIF_ASCE)
@@ -265,6 +266,10 @@ sysc_work:
 	jo	sysc_mcck_pending
 	tm	__TI_flags+7(%r12),_TIF_NEED_RESCHED
 	jo	sysc_reschedule
+#ifdef CONFIG_UPROBES
+	tm	__TI_flags+7(%r12),_TIF_UPROBE
+	jo	sysc_uprobe_notify
+#endif
 	tm	__PT_FLAGS+7(%r11),_PIF_PER_TRAP
 	jo	sysc_singlestep
 	tm	__TI_flags+7(%r12),_TIF_SIGPENDING
@@ -322,6 +327,16 @@ sysc_notify_resume:
 	larl	%r14,sysc_return
 	jg	do_notify_resume
 
+#
+# _TIF_UPROBE is set, call uprobe_notify_resume
+#
+#ifdef CONFIG_UPROBES
+sysc_uprobe_notify:
+	lgr	%r2,%r11		# pass pointer to pt_regs
+	larl	%r14,sysc_return
+	jg	uprobe_notify_resume
+#endif
+
 #
 # _PIF_PER_TRAP is set, call do_per_trap
 #
diff --git a/arch/s390/kernel/ptrace.c b/arch/s390/kernel/ptrace.c
index 5dc7ad9e2fbf..fe99d6b3f185 100644
--- a/arch/s390/kernel/ptrace.c
+++ b/arch/s390/kernel/ptrace.c
@@ -84,7 +84,8 @@ void update_cr_regs(struct task_struct *task)
 	new.end = thread->per_user.end;
 
 	/* merge TIF_SINGLE_STEP into user specified PER registers. */
-	if (test_tsk_thread_flag(task, TIF_SINGLE_STEP)) {
+	if (test_tsk_thread_flag(task, TIF_SINGLE_STEP) ||
+	    test_tsk_thread_flag(task, TIF_UPROBE_SINGLESTEP)) {
 		if (test_tsk_thread_flag(task, TIF_BLOCK_STEP))
 			new.control |= PER_EVENT_BRANCH;
 		else
@@ -93,6 +94,8 @@ void update_cr_regs(struct task_struct *task)
 		new.control |= PER_CONTROL_SUSPENSION;
 		new.control |= PER_EVENT_TRANSACTION_END;
 #endif
+		if (test_tsk_thread_flag(task, TIF_UPROBE_SINGLESTEP))
+			new.control |= PER_EVENT_IFETCH;
 		new.start = 0;
 		new.end = PSW_ADDR_INSN;
 	}
diff --git a/arch/s390/kernel/traps.c b/arch/s390/kernel/traps.c
index c5762324d9ee..e3e06a4fdfce 100644
--- a/arch/s390/kernel/traps.c
+++ b/arch/s390/kernel/traps.c
@@ -58,15 +58,10 @@ int is_valid_bugaddr(unsigned long addr)
 	return 1;
 }
 
-static void __kprobes do_trap(struct pt_regs *regs,
-			      int si_signo, int si_code, char *str)
+void do_report_trap(struct pt_regs *regs, int si_signo, int si_code, char *str)
 {
 	siginfo_t info;
 
-	if (notify_die(DIE_TRAP, str, regs, 0,
-		       regs->int_code, si_signo) == NOTIFY_STOP)
-		return;
-
 	if (user_mode(regs)) {
 		info.si_signo = si_signo;
 		info.si_errno = 0;
@@ -90,6 +85,15 @@ static void __kprobes do_trap(struct pt_regs *regs,
         }
 }
 
+static void __kprobes do_trap(struct pt_regs *regs, int si_signo, int si_code,
+			      char *str)
+{
+	if (notify_die(DIE_TRAP, str, regs, 0,
+		       regs->int_code, si_signo) == NOTIFY_STOP)
+		return;
+	do_report_trap(regs, si_signo, si_code, str);
+}
+
 void __kprobes do_per_trap(struct pt_regs *regs)
 {
 	siginfo_t info;
@@ -178,6 +182,7 @@ void __kprobes illegal_op(struct pt_regs *regs)
 	siginfo_t info;
         __u8 opcode[6];
 	__u16 __user *location;
+	int is_uprobe_insn = 0;
 	int signal = 0;
 
 	location = get_trap_ip(regs);
@@ -194,6 +199,10 @@ void __kprobes illegal_op(struct pt_regs *regs)
 				force_sig_info(SIGTRAP, &info, current);
 			} else
 				signal = SIGILL;
+#ifdef CONFIG_UPROBES
+		} else if (*((__u16 *) opcode) == UPROBE_SWBP_INSN) {
+			is_uprobe_insn = 1;
+#endif
 #ifdef CONFIG_MATHEMU
 		} else if (opcode[0] == 0xb3) {
 			if (get_user(*((__u16 *) (opcode+2)), location+1))
@@ -219,11 +228,13 @@ void __kprobes illegal_op(struct pt_regs *regs)
 #endif
 		} else
 			signal = SIGILL;
-	} else {
-		/*
-		 * If we get an illegal op in kernel mode, send it through the
-		 * kprobes notifier. If kprobes doesn't pick it up, SIGILL
-		 */
+	}
+	/*
+	 * We got either an illegal op in kernel mode, or user space trapped
+	 * on a uprobes illegal instruction. See if kprobes or uprobes picks
+	 * it up. If not, SIGILL.
+	 */
+	if (is_uprobe_insn || !user_mode(regs)) {
 		if (notify_die(DIE_BPT, "bpt", regs, 0,
 			       3, SIGTRAP) != NOTIFY_STOP)
 			signal = SIGILL;
diff --git a/arch/s390/kernel/uprobes.c b/arch/s390/kernel/uprobes.c
new file mode 100644
index 000000000000..956f4f7a591c
--- /dev/null
+++ b/arch/s390/kernel/uprobes.c
@@ -0,0 +1,332 @@
+/*
+ *  User-space Probes (UProbes) for s390
+ *
+ *    Copyright IBM Corp. 2014
+ *    Author(s): Jan Willeke,
+ */
+
+#include <linux/kprobes.h>
+#include <linux/uaccess.h>
+#include <linux/uprobes.h>
+#include <linux/compat.h>
+#include <linux/kdebug.h>
+#include <asm/switch_to.h>
+#include <asm/facility.h>
+#include <asm/dis.h>
+#include "entry.h"
+
+#define	UPROBE_TRAP_NR	UINT_MAX
+
+int arch_uprobe_analyze_insn(struct arch_uprobe *auprobe, struct mm_struct *mm,
+			     unsigned long addr)
+{
+	return probe_is_prohibited_opcode(auprobe->insn);
+}
+
+int arch_uprobe_pre_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+	if (psw_bits(regs->psw).eaba == PSW_AMODE_24BIT)
+		return -EINVAL;
+	if (!is_compat_task() && psw_bits(regs->psw).eaba == PSW_AMODE_31BIT)
+		return -EINVAL;
+	clear_pt_regs_flag(regs, PIF_PER_TRAP);
+	auprobe->saved_per = psw_bits(regs->psw).r;
+	auprobe->saved_int_code = regs->int_code;
+	regs->int_code = UPROBE_TRAP_NR;
+	regs->psw.addr = current->utask->xol_vaddr;
+	set_tsk_thread_flag(current, TIF_UPROBE_SINGLESTEP);
+	update_cr_regs(current);
+	return 0;
+}
+
+bool arch_uprobe_xol_was_trapped(struct task_struct *tsk)
+{
+	struct pt_regs *regs = task_pt_regs(tsk);
+
+	if (regs->int_code != UPROBE_TRAP_NR)
+		return true;
+	return false;
+}
+
+int arch_uprobe_post_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+	int fixup = probe_get_fixup_type(auprobe->insn);
+	struct uprobe_task *utask = current->utask;
+
+	clear_tsk_thread_flag(current, TIF_UPROBE_SINGLESTEP);
+	update_cr_regs(current);
+	psw_bits(regs->psw).r = auprobe->saved_per;
+	regs->int_code = auprobe->saved_int_code;
+
+	if (fixup & FIXUP_PSW_NORMAL)
+		regs->psw.addr += utask->vaddr - utask->xol_vaddr;
+	if (fixup & FIXUP_RETURN_REGISTER) {
+		int reg = (auprobe->insn[0] & 0xf0) >> 4;
+
+		regs->gprs[reg] += utask->vaddr - utask->xol_vaddr;
+	}
+	if (fixup & FIXUP_BRANCH_NOT_TAKEN) {
+		int ilen = insn_length(auprobe->insn[0] >> 8);
+
+		if (regs->psw.addr - utask->xol_vaddr == ilen)
+			regs->psw.addr = utask->vaddr + ilen;
+	}
+	/* If per tracing was active generate trap */
+	if (regs->psw.mask & PSW_MASK_PER)
+		do_per_trap(regs);
+	return 0;
+}
+
+int arch_uprobe_exception_notify(struct notifier_block *self, unsigned long val,
+				 void *data)
+{
+	struct die_args *args = data;
+	struct pt_regs *regs = args->regs;
+
+	if (!user_mode(regs))
+		return NOTIFY_DONE;
+	if (regs->int_code & 0x200) /* Trap during transaction */
+		return NOTIFY_DONE;
+	switch (val) {
+	case DIE_BPT:
+		if (uprobe_pre_sstep_notifier(regs))
+			return NOTIFY_STOP;
+		break;
+	case DIE_SSTEP:
+		if (uprobe_post_sstep_notifier(regs))
+			return NOTIFY_STOP;
+	default:
+		break;
+	}
+	return NOTIFY_DONE;
+}
+
+void arch_uprobe_abort_xol(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+	clear_thread_flag(TIF_UPROBE_SINGLESTEP);
+	regs->int_code = auprobe->saved_int_code;
+	regs->psw.addr = current->utask->vaddr;
+}
+
+unsigned long arch_uretprobe_hijack_return_addr(unsigned long trampoline,
+						struct pt_regs *regs)
+{
+	unsigned long orig;
+
+	orig = regs->gprs[14];
+	regs->gprs[14] = trampoline;
+	return orig;
+}
+
+/* Instruction Emulation */
+
+static void adjust_psw_addr(psw_t *psw, unsigned long len)
+{
+	psw->addr = __rewind_psw(*psw, -len);
+}
+
+#define EMU_ILLEGAL_OP		1
+#define EMU_SPECIFICATION	2
+#define EMU_ADDRESSING		3
+
+#define emu_load_ril(ptr, output)			\
+({							\
+	unsigned int mask = sizeof(*(ptr)) - 1;		\
+	__typeof__(*(ptr)) input;			\
+	int __rc = 0;					\
+							\
+	if (!test_facility(34))				\
+		__rc = EMU_ILLEGAL_OP;			\
+	else if ((u64 __force)ptr & mask)		\
+		__rc = EMU_SPECIFICATION;		\
+	else if (get_user(input, ptr))			\
+		__rc = EMU_ADDRESSING;			\
+	else						\
+		*(output) = input;			\
+	__rc;						\
+})
+
+#define emu_store_ril(ptr, input)			\
+({							\
+	unsigned int mask = sizeof(*(ptr)) - 1;		\
+	int __rc = 0;					\
+							\
+	if (!test_facility(34))				\
+		__rc = EMU_ILLEGAL_OP;			\
+	else if ((u64 __force)ptr & mask)		\
+		__rc = EMU_SPECIFICATION;		\
+	else if (put_user(*(input), ptr))		\
+		__rc = EMU_ADDRESSING;			\
+	__rc;						\
+})
+
+#define emu_cmp_ril(regs, ptr, cmp)			\
+({							\
+	unsigned int mask = sizeof(*(ptr)) - 1;		\
+	__typeof__(*(ptr)) input;			\
+	int __rc = 0;					\
+							\
+	if (!test_facility(34))				\
+		__rc = EMU_ILLEGAL_OP;			\
+	else if ((u64 __force)ptr & mask)		\
+		__rc = EMU_SPECIFICATION;		\
+	else if (get_user(input, ptr))			\
+		__rc = EMU_ADDRESSING;			\
+	else if (input > *(cmp))			\
+		psw_bits((regs)->psw).cc = 1;		\
+	else if (input < *(cmp))			\
+		psw_bits((regs)->psw).cc = 2;		\
+	else						\
+		psw_bits((regs)->psw).cc = 0;		\
+	__rc;						\
+})
+
+struct insn_ril {
+	u8 opc0;
+	u8 reg	: 4;
+	u8 opc1 : 4;
+	s32 disp;
+} __packed;
+
+union split_register {
+	u64 u64;
+	u32 u32[2];
+	u16 u16[4];
+	s64 s64;
+	s32 s32[2];
+	s16 s16[4];
+};
+
+/*
+ * pc relative instructions are emulated, since parameters may not be
+ * accessible from the xol area due to range limitations.
+ */
+static void handle_insn_ril(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+	union split_register *rx;
+	struct insn_ril *insn;
+	unsigned int ilen;
+	void *uptr;
+	int rc = 0;
+
+	insn = (struct insn_ril *) &auprobe->insn;
+	rx = (union split_register *) &regs->gprs[insn->reg];
+	uptr = (void *)(regs->psw.addr + (insn->disp * 2));
+	ilen = insn_length(insn->opc0);
+
+	switch (insn->opc0) {
+	case 0xc0:
+		switch (insn->opc1) {
+		case 0x00: /* larl */
+			rx->u64 = (unsigned long)uptr;
+			break;
+		}
+		break;
+	case 0xc4:
+		switch (insn->opc1) {
+		case 0x02: /* llhrl */
+			rc = emu_load_ril((u16 __user *)uptr, &rx->u32[1]);
+			break;
+		case 0x04: /* lghrl */
+			rc = emu_load_ril((s16 __user *)uptr, &rx->u64);
+			break;
+		case 0x05: /* lhrl */
+			rc = emu_load_ril((s16 __user *)uptr, &rx->u32[1]);
+			break;
+		case 0x06: /* llghrl */
+			rc = emu_load_ril((u16 __user *)uptr, &rx->u64);
+			break;
+		case 0x08: /* lgrl */
+			rc = emu_load_ril((u64 __user *)uptr, &rx->u64);
+			break;
+		case 0x0c: /* lgfrl */
+			rc = emu_load_ril((s32 __user *)uptr, &rx->u64);
+			break;
+		case 0x0d: /* lrl */
+			rc = emu_load_ril((u32 __user *)uptr, &rx->u32[1]);
+			break;
+		case 0x0e: /* llgfrl */
+			rc = emu_load_ril((u32 __user *)uptr, &rx->u64);
+			break;
+		case 0x07: /* sthrl */
+			rc = emu_store_ril((u16 __user *)uptr, &rx->u16[3]);
+			break;
+		case 0x0b: /* stgrl */
+			rc = emu_store_ril((u64 __user *)uptr, &rx->u64);
+			break;
+		case 0x0f: /* strl */
+			rc = emu_store_ril((u32 __user *)uptr, &rx->u32[1]);
+			break;
+		}
+		break;
+	case 0xc6:
+		switch (insn->opc1) {
+		case 0x02: /* pfdrl */
+			if (!test_facility(34))
+				rc = EMU_ILLEGAL_OP;
+			break;
+		case 0x04: /* cghrl */
+			rc = emu_cmp_ril(regs, (s16 __user *)uptr, &rx->s64);
+			break;
+		case 0x05: /* chrl */
+			rc = emu_cmp_ril(regs, (s16 __user *)uptr, &rx->s32[1]);
+			break;
+		case 0x06: /* clghrl */
+			rc = emu_cmp_ril(regs, (u16 __user *)uptr, &rx->u64);
+			break;
+		case 0x07: /* clhrl */
+			rc = emu_cmp_ril(regs, (u16 __user *)uptr, &rx->u32[1]);
+			break;
+		case 0x08: /* cgrl */
+			rc = emu_cmp_ril(regs, (s64 __user *)uptr, &rx->s64);
+			break;
+		case 0x0a: /* clgrl */
+			rc = emu_cmp_ril(regs, (u64 __user *)uptr, &rx->u64);
+			break;
+		case 0x0c: /* cgfrl */
+			rc = emu_cmp_ril(regs, (s32 __user *)uptr, &rx->s64);
+			break;
+		case 0x0d: /* crl */
+			rc = emu_cmp_ril(regs, (s32 __user *)uptr, &rx->s32[1]);
+			break;
+		case 0x0e: /* clgfrl */
+			rc = emu_cmp_ril(regs, (u32 __user *)uptr, &rx->u64);
+			break;
+		case 0x0f: /* clrl */
+			rc = emu_cmp_ril(regs, (u32 __user *)uptr, &rx->u32[1]);
+			break;
+		}
+		break;
+	}
+	adjust_psw_addr(&regs->psw, ilen);
+	switch (rc) {
+	case EMU_ILLEGAL_OP:
+		regs->int_code = ilen << 16 | 0x0001;
+		do_report_trap(regs, SIGILL, ILL_ILLOPC, NULL);
+		break;
+	case EMU_SPECIFICATION:
+		regs->int_code = ilen << 16 | 0x0006;
+		do_report_trap(regs, SIGILL, ILL_ILLOPC , NULL);
+		break;
+	case EMU_ADDRESSING:
+		regs->int_code = ilen << 16 | 0x0005;
+		do_report_trap(regs, SIGSEGV, SEGV_MAPERR, NULL);
+		break;
+	}
+}
+
+bool arch_uprobe_skip_sstep(struct arch_uprobe *auprobe, struct pt_regs *regs)
+{
+	if ((psw_bits(regs->psw).eaba == PSW_AMODE_24BIT) ||
+	    ((psw_bits(regs->psw).eaba == PSW_AMODE_31BIT) &&
+	     !is_compat_task())) {
+		regs->psw.addr = __rewind_psw(regs->psw, UPROBE_SWBP_INSN_SIZE);
+		do_report_trap(regs, SIGILL, ILL_ILLADR, NULL);
+		return true;
+	}
+	if (probe_is_insn_relative_long(auprobe->insn)) {
+		handle_insn_ril(auprobe, regs);
+		return true;
+	}
+	return false;
+}
-- 
cgit v1.2.3


From 48e9a6c1f54695609b709bf674aac133794ada00 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 24 Sep 2014 16:37:20 +0200
Subject: s390/topology: call set_sched_topology early

The call to topology_init is too late for the set_sched_topology call.
The initial scheduling domain structure has already been established
with default topology array. Use the smp_cpus_done() call to get the
s390 specific topology array registered early enough.

Cc: stable@vger.kernel.org # v3.16+
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/topology.c | 18 ++++++++++--------
 1 file changed, 10 insertions(+), 8 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/topology.c b/arch/s390/kernel/topology.c
index 355a16c55702..b93bed76ea94 100644
--- a/arch/s390/kernel/topology.c
+++ b/arch/s390/kernel/topology.c
@@ -464,15 +464,17 @@ static struct sched_domain_topology_level s390_topology[] = {
 
 static int __init topology_init(void)
 {
-	if (!MACHINE_HAS_TOPOLOGY) {
+	if (MACHINE_HAS_TOPOLOGY)
+		set_topology_timer();
+	else
 		topology_update_polarization_simple();
-		goto out;
-	}
-	set_topology_timer();
-out:
-
-	set_sched_topology(s390_topology);
-
 	return device_create_file(cpu_subsys.dev_root, &dev_attr_dispatching);
 }
 device_initcall(topology_init);
+
+static int __init early_topology_init(void)
+{
+	set_sched_topology(s390_topology);
+	return 0;
+}
+early_initcall(early_topology_init);
-- 
cgit v1.2.3


From 242a112af62ea73ce507cbe76c2c944c23b6a1e3 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Fri, 26 Sep 2014 09:23:20 +0200
Subject: s390/setup: correct 4-level kernel page table detection

Fix calculation to decide if a 4-level kernel page table is required.
Git commit c972cc60c23f5a63 "s390/vmalloc: have separate modules area"
added the separate module area which reduces the size of the vmalloc
area but fails to take it into account for the 3 vs 4 level page table
decision.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/setup.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/setup.c b/arch/s390/kernel/setup.c
index 82bc113e8c1d..cdfc060dd319 100644
--- a/arch/s390/kernel/setup.c
+++ b/arch/s390/kernel/setup.c
@@ -452,8 +452,8 @@ static void __init setup_memory_end(void)
 #ifdef CONFIG_64BIT
 	vmalloc_size = VMALLOC_END ?: (128UL << 30) - MODULES_LEN;
 	tmp = (memory_end ?: max_physmem_end) / PAGE_SIZE;
-	tmp = tmp * (sizeof(struct page) + PAGE_SIZE) + vmalloc_size;
-	if (tmp <= (1UL << 42))
+	tmp = tmp * (sizeof(struct page) + PAGE_SIZE);
+	if (tmp + vmalloc_size + MODULES_LEN <= (1UL << 42))
 		vmax = 1UL << 42;	/* 3-level kernel page table */
 	else
 		vmax = 1UL << 53;	/* 4-level kernel page table */
-- 
cgit v1.2.3


From a9b1649917f0d2058022eda06082f9d299a06354 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 1 Oct 2014 10:44:40 +0200
Subject: s390/vtime: do not reset idle data on CPU hotplug

The sysfs attributes /sys/devices/system/cpu/cpu0/idle_count and
/sys/devices/system/cpu/cpu0/idle_time_us are reset to zero every
time a CPU is set online. The idle and iowait fields in /proc/stat
corresponding to idle_time_us are not reset. To make things
consistent do not reset the data for the sys attributes.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/processor.c | 2 --
 1 file changed, 2 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/processor.c b/arch/s390/kernel/processor.c
index 24612029f450..32587cc76306 100644
--- a/arch/s390/kernel/processor.c
+++ b/arch/s390/kernel/processor.c
@@ -23,7 +23,6 @@ static DEFINE_PER_CPU(struct cpuid, cpu_id);
  */
 void cpu_init(void)
 {
-	struct s390_idle_data *idle = &__get_cpu_var(s390_idle);
 	struct cpuid *id = &__get_cpu_var(cpu_id);
 
 	get_cpu_id(id);
@@ -31,7 +30,6 @@ void cpu_init(void)
 	current->active_mm = &init_mm;
 	BUG_ON(current->mm);
 	enter_lazy_tlb(&init_mm, current);
-	memset(idle, 0, sizeof(*idle));
 }
 
 /*
-- 
cgit v1.2.3


From fe0f49768d807a8fe6336b097feb8c4441951710 Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Tue, 30 Sep 2014 17:37:52 +0200
Subject: s390/nohz: use a per-cpu flag for arch_needs_cpu

Move the nohz_delay bit from the s390_idle data structure to the
per-cpu flags. Clear the nohz delay flag in __cpu_disable and
remove the cpu hotplug notifier that used to do this.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/irq.c   |  2 +-
 arch/s390/kernel/smp.c   |  1 +
 arch/s390/kernel/vtime.c | 19 +------------------
 3 files changed, 3 insertions(+), 19 deletions(-)

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/irq.c b/arch/s390/kernel/irq.c
index 051574ee5366..1b8a38ab7861 100644
--- a/arch/s390/kernel/irq.c
+++ b/arch/s390/kernel/irq.c
@@ -259,7 +259,7 @@ static irqreturn_t do_ext_interrupt(int irq, void *dummy)
 
 	ext_code = *(struct ext_code *) &regs->int_code;
 	if (ext_code.code != EXT_IRQ_CLK_COMP)
-		__get_cpu_var(s390_idle).nohz_delay = 1;
+		set_cpu_flag(CIF_NOHZ_DELAY);
 
 	index = ext_hash(ext_code.code);
 	rcu_read_lock();
diff --git a/arch/s390/kernel/smp.c b/arch/s390/kernel/smp.c
index abec97b4ddbf..46317d6951c4 100644
--- a/arch/s390/kernel/smp.c
+++ b/arch/s390/kernel/smp.c
@@ -720,6 +720,7 @@ int __cpu_disable(void)
 	cregs[6]  &= ~0xff000000UL;	/* disable all I/O interrupts */
 	cregs[14] &= ~0x1f000000UL;	/* disable most machine checks */
 	__ctl_load(cregs, 0, 15);
+	clear_cpu_flag(CIF_NOHZ_DELAY);
 	return 0;
 }
 
diff --git a/arch/s390/kernel/vtime.c b/arch/s390/kernel/vtime.c
index 8c34363d6f1e..40709821abde 100644
--- a/arch/s390/kernel/vtime.c
+++ b/arch/s390/kernel/vtime.c
@@ -163,7 +163,7 @@ void __kprobes vtime_stop_cpu(void)
 	/* Wait for external, I/O or machine check interrupt. */
 	psw_mask = PSW_KERNEL_BITS | PSW_MASK_WAIT | PSW_MASK_DAT |
 		PSW_MASK_IO | PSW_MASK_EXT | PSW_MASK_MCHECK;
-	idle->nohz_delay = 0;
+	clear_cpu_flag(CIF_NOHZ_DELAY);
 
 	/* Call the assembler magic in entry.S */
 	psw_idle(idle, psw_mask);
@@ -378,25 +378,8 @@ void init_cpu_vtimer(void)
 	set_vtimer(VTIMER_MAX_SLICE);
 }
 
-static int s390_nohz_notify(struct notifier_block *self, unsigned long action,
-			    void *hcpu)
-{
-	struct s390_idle_data *idle;
-	long cpu = (long) hcpu;
-
-	idle = &per_cpu(s390_idle, cpu);
-	switch (action & ~CPU_TASKS_FROZEN) {
-	case CPU_DYING:
-		idle->nohz_delay = 0;
-	default:
-		break;
-	}
-	return NOTIFY_OK;
-}
-
 void __init vtime_init(void)
 {
 	/* Enable cpu timer interrupts on the boot cpu. */
 	init_cpu_vtimer();
-	cpu_notifier(s390_nohz_notify, 0);
 }
-- 
cgit v1.2.3


From b5f87f15e20092c060f465b283b07a76af7f2e5f Mon Sep 17 00:00:00 2001
From: Martin Schwidefsky <schwidefsky@de.ibm.com>
Date: Wed, 1 Oct 2014 10:57:57 +0200
Subject: s390/idle: consolidate idle functions and definitions

Move the C functions and definitions related to the idle state handling
to arch/s390/include/asm/idle.h and arch/s390/kernel/idle.c. The function
s390_get_idle_time is renamed to arch_cpu_idle_time and vtime_stop_cpu to
enabled_wait.

Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>
---
 arch/s390/kernel/Makefile      |   2 +-
 arch/s390/kernel/asm-offsets.c |   2 +-
 arch/s390/kernel/entry.h       |   2 +-
 arch/s390/kernel/idle.c        | 124 +++++++++++++++++++++++++++++++++++++++++
 arch/s390/kernel/process.c     |  24 --------
 arch/s390/kernel/smp.c         |  39 +------------
 arch/s390/kernel/vtime.c       |  60 +-------------------
 7 files changed, 130 insertions(+), 123 deletions(-)
 create mode 100644 arch/s390/kernel/idle.c

(limited to 'arch/s390/kernel')

diff --git a/arch/s390/kernel/Makefile b/arch/s390/kernel/Makefile
index 3249e1f36d55..c249785669f3 100644
--- a/arch/s390/kernel/Makefile
+++ b/arch/s390/kernel/Makefile
@@ -28,7 +28,7 @@ CFLAGS_ptrace.o		+= -DUTS_MACHINE='"$(UTS_MACHINE)"'
 
 CFLAGS_sysinfo.o += -Iinclude/math-emu -Iarch/s390/math-emu -w
 
-obj-y	:= traps.o time.o process.o base.o early.o setup.o vtime.o
+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	+= debug.o irq.o ipl.o dis.o diag.o sclp.o vdso.o
 obj-y	+= sysinfo.o jump_label.o lgr.o os_info.o machine_kex