Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux

Pull s390 updates from Martin Schwidefsky:
 "This patch set contains the main portion of the changes for 3.18 in
  regard to the s390 architecture.  It is a bit bigger than usual,
  mainly because of a new driver and the vector extension patches.

  The interesting bits are:
   - Quite a bit of work on the tracing front.  Uprobes is enabled and
     the ftrace code is reworked to get some of the lost performance
     back if CONFIG_FTRACE is enabled.
   - To improve boot time with CONFIG_DEBIG_PAGEALLOC, support for the
     IPTE range facility is added.
   - The rwlock code is re-factored to improve writer fairness and to be
     able to use the interlocked-access instructions.
   - The kernel part for the support of the vector extension is added.
   - The device driver to access the CD/DVD on the HMC is added, this
     will hopefully come in handy to improve the installation process.
   - Add support for control-unit initiated reconfiguration.
   - The crypto device driver is enhanced to enable the additional AP
     domains and to allow the new crypto hardware to be used.
   - Bug fixes"

* 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/s390/linux: (39 commits)
  s390/ftrace: simplify enabling/disabling of ftrace_graph_caller
  s390/ftrace: remove 31 bit ftrace support
  s390/kdump: add support for vector extension
  s390/disassembler: add vector instructions
  s390: add support for vector extension
  s390/zcrypt: Toleration of new crypto hardware
  s390/idle: consolidate idle functions and definitions
  s390/nohz: use a per-cpu flag for arch_needs_cpu
  s390/vtime: do not reset idle data on CPU hotplug
  s390/dasd: add support for control unit initiated reconfiguration
  s390/dasd: fix infinite loop during format
  s390/mm: make use of ipte range facility
  s390/setup: correct 4-level kernel page table detection
  s390/topology: call set_sched_topology early
  s390/uprobes: architecture backend for uprobes
  s390/uprobes: common library for kprobes and uprobes
  s390/rwlock: use the interlocked-access facility 1 instructions
  s390/rwlock: improve writer fairness
  s390/rwlock: remove interrupt-enabling rwlock variant.
  s390/mm: remove change bit override support
  ...

1 files changed, 54 insertions, 85 deletions

diff --git a/arch/s390/kernel/ftrace.c b/arch/s390/kernel/ftrace.c
index 54d6493c4a56..51d14fe5eb9a 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>
@@ -17,100 +17,76 @@
 #include <asm/asm-offsets.h>
 #include "entry.h"
 
-#ifdef CONFIG_DYNAMIC_FTRACE
-
+void mcount_replace_code(void);
 void ftrace_disable_code(void);
 void ftrace_enable_insn(void);
 
-#ifdef CONFIG_64BIT
 /*
- * The 64-bit mcount code looks like this:
+ * The 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.
- * The 64-bit enabled ftrace code block looks like this:
- *	stg	%r14,8(%r15)		# offset 0
+ * 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 enabled ftrace code block looks like this:
+ *	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 */
-/*
- * The 31-bit mcount code looks like this:
- *	st	%r14,4(%r15)		# offset 0
- * >	bras	%r1,0f			# offset 4
- * >	.long	_mcount			# offset 8
- * >	.long	<&counter>		# offset 12
- * > 0:	l	%r14,0(%r1)		# offset 16
- * >	l	%r1,4(%r1)		# offset 20
- *	basr	%r14,%r14		# offset 24
- *	l	%r14,4(%r15)		# offset 26
- * Total length is 30 bytes. The twenty bytes starting from offset 4
- * to offset 24 get overwritten by ftrace_make_nop / ftrace_make_call.
- * The 31-bit enabled ftrace code block looks like this:
- *	st	%r14,4(%r15)		# offset 0
- * >	l	%r14,__LC_FTRACE_FUNC	# offset 4
- * >	j	0f			# offset 8
- * >	.fill	12,1,0x07		# offset 12
- *   0:	basr	%r14,%r14		# offset 24
- *	l	%r14,4(%r14)		# offset 26
- * The return points of the mcount/ftrace function have the same offset 26.
- * The 31-bit disabled ftrace code block looks like this:
- *	st	%r14,4(%r15)		# offset 0
- * >	j	.+26			# offset 4
- * >	j	0f			# offset 8
- * >	.fill	12,1,0x07		# offset 12
- *   0:	basr	%r14,%r14		# offset 24
- *	l	%r14,4(%r14)		# offset 26
- * The j instruction branches to offset 30 to skip as many instructions
- * as possible.
- */
-asm(
-	"	.align	4\n"
-	"ftrace_disable_code:\n"
-	"	j	1f\n"
-	"	j	0f\n"
-	"	.fill	12,1,0x07\n"
-	"0:	basr	%r14,%r14\n"
-	"1:\n"
-	"	.align	4\n"
-	"ftrace_enable_insn:\n"
-	"	l	%r14,"__stringify(__LC_FTRACE_FUNC)"\n");
-
-#define FTRACE_INSN_SIZE	4
-
-#endif /* CONFIG_64BIT */
-
+int ftrace_modify_call(struct dyn_ftrace *rec, unsigned long old_addr,
+		       unsigned long addr)
+{
+	return 0;
+}
 
 int ftrace_make_nop(struct module *mod, struct dyn_ftrace *rec,
 		    unsigned long addr)
 {
+	/* 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;
+	}
 	if (probe_kernel_write((void *) rec->ip, ftrace_disable_code,
 			       MCOUNT_INSN_SIZE))
 		return -EPERM;
@@ -135,8 +111,6 @@ 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
@@ -162,31 +136,26 @@ 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
- * the call to prepare_ftrace_return we patch the bras offset to point
- * directly after the instructions. To enable the call we calculate
- * the original offset to prepare_ftrace_return and put it back.
+ * there is branch relative on condition. To enable the ftrace graph code
+ * block, we simply patch the mask field of the instruction to zero and
+ * turn the instruction into a nop.
+ * To disable the ftrace graph code the mask field will be patched to
+ * all ones, which turns the instruction into an unconditional branch.
  */
 int ftrace_enable_ftrace_graph_caller(void)
 {
-	unsigned short offset;
+	u8 op = 0x04; /* set mask field to zero */
 
-	offset = ((void *) prepare_ftrace_return -
-		  (void *) ftrace_graph_caller) / 2;
-	return probe_kernel_write((void *) ftrace_graph_caller + 2,
-				  &offset, sizeof(offset));
+	return probe_kernel_write(__va(ftrace_graph_caller)+1, &op, sizeof(op));
 }
 
 int ftrace_disable_ftrace_graph_caller(void)
 {
-	static unsigned short offset = 0x0002;
+	u8 op = 0xf4; /* set mask field to all ones */
 
-	return probe_kernel_write((void *) ftrace_graph_caller + 2,
-				  &offset, sizeof(offset));
+	return probe_kernel_write(__va(ftrace_graph_caller)+1, &op, sizeof(op));
 }
 
-#endif /* CONFIG_DYNAMIC_FTRACE */
 #endif /* CONFIG_FUNCTION_GRAPH_TRACER */