All of the infrastructure already exists to support read accesses
for platforms that support a read access independently of read/write
(such as in the case of the SuperH UBC). This just trivially hooks
up the read case by itself.
Signed-off-by: NPaul Mundt <lethal@linux-sh.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Li Zefan <lizf@cn.fujitsu.com>
Cc: Prasad <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Jan Kiszka <jan.kiszka@web.de>
Cc: Jiri Slaby <jirislaby@gmail.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Masami Hiramatsu <mhiramat@redhat.com>
Cc: Arjan van de Ven <arjan@linux.intel.com>
LKML-Reference: <20091109083733.GA25848@linux-sh.org>
Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>

This patch rebase the implementation of the breakpoints API on top of
perf events instances.

Each breakpoints are now perf events that handle the
register scheduling, thread/cpu attachment, etc..

The new layering is now made as follows:

       ptrace       kgdb      ftrace   perf syscall
          \          |          /         /
           \         |         /         /
                                        /
            Core breakpoint API        /
                                      /
                     |               /
                     |              /

              Breakpoints perf events

                     |
                     |

               Breakpoints PMU ---- Debug Register constraints handling
                                    (Part of core breakpoint API)
                     |
                     |

             Hardware debug registers

Reasons of this rewrite:

- Use the centralized/optimized pmu registers scheduling,
  implying an easier arch integration
- More powerful register handling: perf attributes (pinned/flexible
  events, exclusive/non-exclusive, tunable period, etc...)

Impact:

- New perf ABI: the hardware breakpoints counters
- Ptrace breakpoints setting remains tricky and still needs some per
  thread breakpoints references.

Todo (in the order):

- Support breakpoints perf counter events for perf tools (ie: implement
  perf_bpcounter_event())
- Support from perf tools

Changes in v2:

- Follow the perf "event " rename
- The ptrace regression have been fixed (ptrace breakpoint perf events
  weren't released when a task ended)
- Drop the struct hw_breakpoint and store generic fields in
  perf_event_attr.
- Separate core and arch specific headers, drop
  asm-generic/hw_breakpoint.h and create linux/hw_breakpoint.h
- Use new generic len/type for breakpoint
- Handle off case: when breakpoints api is not supported by an arch

Changes in v3:

- Fix broken CONFIG_KVM, we need to propagate the breakpoint api
  changes to kvm when we exit the guest and restore the bp registers
  to the host.

Changes in v4:

- Drop the hw_breakpoint_restore() stub as it is only used by KVM
- EXPORT_SYMBOL_GPL hw_breakpoint_restore() as KVM can be built as a
  module
- Restore the breakpoints unconditionally on kvm guest exit:
  TIF_DEBUG_THREAD doesn't anymore cover every cases of running
  breakpoints and vcpu->arch.switch_db_regs might not always be
  set when the guest used debug registers.
  (Waiting for a reliable optimization)

Changes in v5:

- Split-up the asm-generic/hw-breakpoint.h moving to
  linux/hw_breakpoint.h into a separate patch
- Optimize the breakpoints restoring while switching from kvm guest
  to host. We only want to restore the state if we have active
  breakpoints to the host, otherwise we don't care about messed-up
  address registers.
- Add asm/hw_breakpoint.h to Kbuild
- Fix bad breakpoint type in trace_selftest.c

Changes in v6:

- Fix wrong header inclusion in trace.h (triggered a build
  error with CONFIG_FTRACE_SELFTEST
Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: Prasad <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Arnaldo Carvalho de Melo <acme@redhat.com>
Cc: Steven Rostedt <rostedt@goodmis.org>
Cc: Ingo Molnar <mingo@elte.hu>
Cc: Jan Kiszka <jan.kiszka@web.de>
Cc: Jiri Slaby <jirislaby@gmail.com>
Cc: Li Zefan <lizf@cn.fujitsu.com>
Cc: Avi Kivity <avi@redhat.com>
Cc: Paul Mackerras <paulus@samba.org>
Cc: Mike Galbraith <efault@gmx.de>
Cc: Masami Hiramatsu <mhiramat@redhat.com>
Cc: Paul Mundt <lethal@linux-sh.org>

The compiler warns us about:

  kernel/trace/trace_ksym.c: In function ksym_hbp_handler:
  kernel/trace/trace_ksym.c:92: attention : passing argument 1 of trace_buffer_lock_reserve from incompatible pointer type
  kernel/trace/trace_ksym.c:106: attention : passing argument 1 of trace_buffer_unlock_commit from incompatible pointer type

Commit "e77405ad" (tracing: pass around ring buffer
instead of tracer) has changed the central tracing
APIs. And this change has updated every callsites of
these APIs except those that aren't in tracing/core,
such as the ksym tracer.

Cc: Steven Rostedt <srostedt@redhat.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

It's rather boring to clear symbol one by one in ksym_trace_filter
file, so, this patch will let ksym_trace_filter file support quickly
clear all break points. We can write "0" to this file and it will clear
all symbols

for example:
 # cat ksym_trace_filter
 ksym_filter_head:rw-
 global_trace:rw-
 # echo 0 > ksym_trace_filter
 # cat ksym_trace_filter
 #

Changelog v1->v2:
Add other ways to clear all breakpoints by writing NULL or "*:---"
to ksym_trace_filter file base on K.Prasad's suggestion
Signed-off-by: NXiao Guangrong <xiaoguangrong@cn.fujitsu.com>
LKML-Reference: <4A67E092.3080202@cn.fujitsu.com>
Signed-off-by: NSteven Rostedt <rostedt@goodmis.org>

This patch fix 2 bugs:
- fix the return value of ksym_trace_filter_write() when we want to
  clear symbol in ksym_trace_filter file
  for example:
  # echo global_trace:rw- > /debug/tracing/ksym_trace_filter
  # echo global_trace:--- > /debug/tracing/ksym_trace_filter
  -bash: echo: write error: Invalid argument
  # cat /debug/tracing/ksym_trace_filter
  #
  We want to clear 'global_trace' in ksym_trace_filter, it complain
  with "Invalid argument", but the operation is successful

- the "r--" access types is not allowed, but ksym_trace_filter file think
  it OK
  for example:
  # echo ksym_tracer_mutex:r-- > ksym_trace_filter
  -bash: echo: write error: Resource temporarily unavailable
  # dmesg
  ksym_tracer request failed. Try again later!!

  The error occur at register_kernel_hw_breakpoint(), but It's should
  at access types parser
Signed-off-by: NXiao Guangrong <xiaoguangrong@cn.fujitsu.com>
LKML-Reference: <4A66863D.5090802@cn.fujitsu.com>
Signed-off-by: NSteven Rostedt <rostedt@goodmis.org>

Fix the output format of ksym tracer, make it properly aligned

Befor patch:
# tracer: ksym_tracer
#
#       TASK-PID      CPU#      Symbol         Type    Function
#          |           |          |              |         |
bash            1378  1   ksym_tracer_mutex     W  mutex_lock+0x11/0x27
bash            1378  1   ksym_filter_head      W  process_new_ksym_entry+0xd2/0x10c
bash            1378  1   ksym_tracer_mutex     W  mutex_unlock+0x12/0x1b
cat             1429  0   ksym_tracer_mutex     W  mutex_lock+0x11/0x27

After patch:
# tracer: ksym_tracer
#
#       TASK-PID   CPU#      Symbol                    Type    Function
#          |        |          |                        |         |
        cat-1423  [000] ksym_tracer_mutex               RW mutex_lock+0x11/0x27
        cat-1423  [000] ksym_filter_head                RW ksym_trace_filter_read+0x6e/0x10d
        cat-1423  [000] ksym_tracer_mutex               RW mutex_unlock+0x12/0x1b
        cat-1423  [000] ksym_tracer_mutex               RW mutex_lock+0x11/0x27
        cat-1423  [000] ksym_filter_head                RW ksym_trace_filter_read+0x6e/0x10d
        cat-1423  [000] ksym_tracer_mutex               RW mutex_unlock+0x12/0x1b
Signed-off-by: NXiao Guangrong <xiaoguangrong@cn.fujitsu.com>
LKML-Reference: <4A6685BB.2090809@cn.fujitsu.com>
Signed-off-by: NSteven Rostedt <rostedt@goodmis.org>

- make ksym_tracer_stat_start() return head->first instead of
  &head->first
- make the output properly aligned

Before:

   Access type                Symbol                     Counter
     NA                   <NA>                              0
     RW                   pid_max                               0

After:

  Access Type   Symbol                                       Counter
  -----------   ------                                       -------
  RW            pid_max                                            0
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: "K.Prasad" <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A52E346.5050608@cn.fujitsu.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

- When remove a filter, we leak entry->ksym_hbp->info.name.

- With CONFIG_FTRAC_SELFTEST enabled, we leak ->info.name:
    # echo ksym_tracer > current_tracer
    # echo 'ksym_selftest_dummy:rw-' > ksym_trace_filter
    # echo nop > current_tracer
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: "K.Prasad" <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A52E328.8010200@cn.fujitsu.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

When access type is changed, the hw break point will be
unregistered and then be registered again with new access
type. But the registration may fail, in this case, -errno
should be returned.
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: "K.Prasad" <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A52E314.7070004@cn.fujitsu.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Make sure the user input string is NULL-terminated.
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: "K.Prasad" <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A52E300.7020601@cn.fujitsu.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Don't take newline into account, otherwise:

 # echo 'pid_max:-w-' > ksym_trace_filter
 # echo -n 'pid_max:rw-' > ksym_trace_filter
 bash: echo: write error: Invalid argument
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: "K.Prasad" <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A52E2EB.9070503@cn.fujitsu.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

# echo 'pid_max:rw-' > ksym_trace_filter
 # cat ksym_trace_filter
 pid_max:rw-
 # echo 'pid_max:ww-' > ksym_trace_filter
 (should return -EINVAL)
 # cat ksym_trace_filter
 (but it ended up removing filter entry)
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: "K.Prasad" <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A52E2CE.6080409@cn.fujitsu.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

Reading ksym_trace_filter gave me some arbitrary characters,
when it should show nothing. It's because buf is not initialized
when there's no filter.

Also reduce stack usage by about 512 bytes.
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: "K.Prasad" <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A52E2B4.6030706@cn.fujitsu.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

struct trace_ksym is used as an entry in hbp list, and is also
used as trace_entry stored in ring buffer.

This is not necessary and is a waste of memory in ring buffer.

There is also a bug that dereferencing field->ksym_hbp in
ksym_trace_output() can be invalid.
Signed-off-by: NLi Zefan <lizf@cn.fujitsu.com>
Acked-by: NFrederic Weisbecker <fweisbec@gmail.com>
Cc: "K.Prasad" <prasad@linux.vnet.ibm.com>
Cc: Alan Stern <stern@rowland.harvard.edu>
Cc: Steven Rostedt <rostedt@goodmis.org>
LKML-Reference: <4A52E2A4.4050007@cn.fujitsu.com>
Signed-off-by: NIngo Molnar <mingo@elte.hu>

ksym_tracer_mutex is declared inside an #ifdef CONFIG_PROFILE_KSYM_TRACER
section. This makes it unavailable for the hardware breakpoint tracer if it is
configured without the breakpoint profiler.

This patch fixes the following build error:

kernel/trace/trace_ksym.c: In function ‘ksym_trace_filter_read’:
kernel/trace/trace_ksym.c:226: erreur: ‘ksym_tracer_mutex’ undeclared (first use in this function)
kernel/trace/trace_ksym.c:226: erreur: (Each undeclared identifier is reported only once
kernel/trace/trace_ksym.c:226: erreur: for each function it appears in.)
kernel/trace/trace_ksym.c: In function ‘ksym_trace_filter_write’:
kernel/trace/trace_ksym.c:273: erreur: ‘ksym_tracer_mutex’ undeclared (first use in this function)
kernel/trace/trace_ksym.c: In function ‘ksym_trace_reset’:
kernel/trace/trace_ksym.c:335: erreur: ‘ksym_tracer_mutex’ undeclared (first use in this function)
make[1]: *** [kernel/trace/trace_ksym.o] Erreur 1

[ Impact: fix a build error ]
Reported-by: NIngo Molnar <mingo@elte.hu>
Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>

This patch adds an ftrace plugin to detect and profile memory access over kernel
variables. It uses HW Breakpoint interfaces to 'watch memory addresses.
Signed-off-by: NK.Prasad <prasad@linux.vnet.ibm.com>
Signed-off-by: NFrederic Weisbecker <fweisbec@gmail.com>