[PATCH 0/7] tracing: seqfile fixes, take 2

[Li Zefan <lizf@cn.fujitsu.com>]

While testing syscall tracepoints proposed by Jason, I found some
entries were missing when reading available_events.

It turned out there's bug in seqfile handling. The bug is, it's
wrong to increment @pos in seq start().

The bug exists in some other places. I also fixed some other different
bugs.

All the patches have been reviewed and tested.

Thanks Liming Wang for the review.

[PATCH 1/7] tracing/events: don't increment @pos in s_start()
[PATCH 2/7] tracing_bprintk: don't increment @pos in t_start()
[PATCH 3/7] trace_stat: don't increment @pos in seq start()
[PATCH 4/7] tracing: reset iterator in t_start()
[PATCH 5/7] ftrace: don't increment @pos in g_start()
[PATCH 6/7] ftrace: don't manipulate @pos in t_start()
[PATCH 7/7] ftrace: fix t_hash_start()
---
 kernel/trace/ftrace.c       |   52 ++++++++++++++++++++++++-------------------
 kernel/trace/trace.c        |   18 +++-----------
 kernel/trace/trace_events.c |   28 ++++++++++++++++++-----
 kernel/trace/trace_printk.c |   26 +++++----------------
 kernel/trace/trace_stat.c   |    6 +----
 5 files changed, 62 insertions(+), 68 deletions(-)